Report of the National Lipid Association`s Statin Safety Task Force

Transcription

Report of the National Lipid Association`s Statin Safety Task Force
SUPPLEMENT TO VOLUME 97, NUMBER 8
APRIL 17, 2006
A Symposium:
Report of the National Lipid Association’s
Statin Safety Task Force
GUEST EDITOR:
James M. McKenney, PharmD
Professor Emeritus
Virginia Commonwealth University
President and CEO
National Clinical Research, Inc.
Richmond, Virginia
ELSEVIER INC.
April 17, 2006 VOL 97 (8A)
A Symposium:
Report of the National Lipid Association’s
Statin Safety Task Force
GUEST EDITOR:
James M. McKenney, PharmD
Professor Emeritus
Virginia Commonwealth University
President and CEO
National Clinical Research, Inc.
Richmond, Virginia
NLA STATIN SAFETY TASK FORCE:
James M. McKenney, PharmD, Chairman
Virginia Commonwealth University
Richmond, Virginia
John R. Guyton, MD
Duke University
Durham, North Carolina
Michael H. Davidson, MD
Rush University
Chicago, Illinois
Terry A. Jacobson, MD
Emory University
Atlanta, Georgia
This supplement is based in part on a symposium held July 17–19, 2005, in Washington, DC. The
symposium and publication of these proceedings were supported by unrestricted educational grants
from Abbott Laboratories, AstraZeneca LP, Kos Pharmaceuticals, Inc., Merck/Schering-Plough, and
Sanyko Pharma, Inc. Editorial support was provided by Conexus Health, Inc., Tampa, Florida.
SENIOR EDITOR
Craig Smith
SENIOR PRODUCTION EDITOR
Mickey Kramer
EDITOR IN CHIEF
William C. Roberts, MD
EXECUTIVE PUBLISHER
David Dionne
PROOF/PRODUCTION EDITOR
Mary Crowell
The opinions expressed in this supplement are those of the panelists and are not attributable
to the sponsor or the publisher, editor, or editorial board of The American Journal of
Cardiology. Clinical judgment must guide each physician in weighing the benefits of treatment
against the risk of toxicity. References made in the articles may indicate uses of drugs at
dosages, for periods of time, and in combinations not included in the current prescribing
information.
Editor’s suggestion: The symposium issues for a full year should be bound together separately from
the regular issues.
THE AMERICAN JOURNAL OF CARDIOLOGY姞
Publisher: The American Journal of Cardiology 姞 (ISSN 00029149, GST 128741063, IPM 0388173) is published 2 times
in January, February, March, April, May, June, July, August,
September, October, November, and December, by Elsevier
Inc., 360 Park Avenue South, New York, NY 10010. Telephone
(212) 462-1933, FAX (212) 462-1935.
Trademarks: THE AMERICAN JOURNAL OF CARDIOLOGY姞 is
a registered trademark of Elsevier, used under license.
Librarians: To date the following supplements have been
published for the year 2006: January 16, 2006, Vol 97 (2A);
April 3, 2006, Vol 97 (7A); April 17, 2006, Vol 97 (8A).
Copyright: Copyright © 2006 by Elsevier. All rights reserved.
This journal and the individual contributions contained in it
are protected by the copyright of Elsevier, and the following
terms and conditions apply to their use:
Photocopying: Single photocopies of single articles may be
made for personal use as allowed by national copyright laws.
Permission of the publisher and payment of a fee is required for
all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale,
and all forms of document delivery. Special rates are available
for educational institutions that wish to make photocopies for
non-profit educational classroom use.
Court Rd, London W1P OLP, UK; phone: (⫹44) 20 7631
5555; fax: (44) 20 7631 5500. Other countries may have a
local reprographic mgnts agency for payments.
All inquiries regarding copyright material from this publication, other than reproduction through the Copyright Clearance Center, should be directed to Elsevier Science Ltd.,
P.O. Box 800, Oxford OX5 1DX, U.K. Phone: 44-1865
843830; Fax: 44-1865 853333; E-mail: permissions@
elsevier.co.uk
Derivative Works: Subscribers may reproduce tables of contents or prepare lists of articles including abstracts for internal
circulation within their institutions. Permission of the publisher
is required for resale or distribution outside the institution.
Permission of the publisher is required for all other derivative
works, including compilations and translations.
Electronic Storage: Permission of the publisher is required
to store electronically any material contained in this journal,
including any article or part of an article. Contact the publisher at the address indicated.
Except as outlined above, no part of this publication may be
reproduced, stored in a retrieval system, or transmitted in
any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the publisher.
Permissions may be sought directly from Elsevier’s Rights
Department in Philadelphia, PA. USA: phone (⫹1) 215-2387869, fax (⫹1) 215-238-2239. e-mail: healthpermissions@
elsevier.com. Requests may also be completed online via
the Elsevier homepage (http://www.elsevier.com/locate/
permissions).
Address permissions requests to: Elsevier Science Rights &
Permissions Department, at the mail, fax, and e-mail addresses noted above.
In the USA, users may clear permissions and make payment
through the Copyright Clearance Center, Inc., 222 Rosewood
Drive, Danvers, MA 01923; telephone (978) 750-8400; fax
(978) 750-4744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Totteham
Notice: No responsibility is assumed by the Publisher for any
injury and/or damage to persons or property as a matter of
products liability, negligence or otherwise, or from any use or
operation of any methods, products, instructions or ideas
contained in the material herein. Because of rapid advances
in the medical sciences, independent verification of diagnoses and drug dosages should be made.
Reprint inquiries should be addressed to Anne Rosenthal,
The American Journal of Cardiology, 360 Park Avenue
South, New York, NY 10010. FAX (212) 633-3820. Email:
a.rosenthal@elsevier.com.
Subscriptions: Solicited only from cardiologists and other physicians. Publisher reserves the right to refuse unqualified subscribers. Please address all subscription mail to THE AMERICAN JOURNAL OF CARDIOLOGY姞, Elsevier Science,
Subscription Customer Service, 6277 Sea Harbor Drive, Orlando, Florida 32887-4800. For customer service in the USA,
phone (877) 839-7126; for customers outside the USA: (407)
345-4020; FAX: (407) 363-1354. E-mail usjcs@elsevier.com.
Single copy price: Regular issues $23; symposia and special
issues $23. Subscription rates for individuals in field of publication for 2006 are: U.S. 1 year $121. All other countries
(prices include air-speeded delivery) 1 year $292. U.S. institutions 1 year $382. All other countries (prices include airspeeded delivery) 1 year $576. The Journal is not responsible
for replacing missing issues unless the Circulation Department
is notified of nonreceipt within 3 months of issue date for
domestic addresses and 6 months for foreign addresses. Circulation records are maintained at Excerpta Medica, Inc., Madison Square Station, P.O. Box 882, New York, NY 10159-0882.
Send address changes to: The American Journal of Cardiology 姞, Elsevier Science, Subscription Customer Service,
6277 Sea Harbor Drive, Orlando, Florida 32887-4800.
Please include a copy of the old address label. Changes of
address must reach the Journal 1 month preceding month of
issue.
Postmaster: Send address changes to The American Journal of Cardiology 姞, Elsevier Science, Subscription Customer
Service, 6277 Sea Harbor Drive, Orlando, Florida 328874800. Periodicals postage paid at New York, NY and at
additional mailing offices.
APRIL 17, 2006 VOL 97 (8A)
A SYMPOSIUM: REPORT OF THE NATIONAL LIPID ASSOCIATION’S STATIN SAFETY
TASK FORCE
1C
Introduction
James M. McKenney
3C
Statins, Cardiovascular Disease, and Drug Safety
Antonio M. Gotto, Jr.
THE EVIDENCE:
6C
Statin Safety: An Overview and Assessment of the Data—2005
Harold Bays
27C
Statin Safety and Drug Interactions: Clinical Implications
Michael B. Bottorff
32C
Statin Safety: An Appraisal from the Adverse Event Reporting System
Michael H. Davidson, John A. Clark, Lucas M. Glass, and Anju Kanumalla
44C
Statin Safety: Lessons from New Drug Applications for Marketed Statins
Terry A. Jacobson
52C
Statin Safety: A Systematic Review
Malcolm Law and Alicja R. Rudnicka
61C
Statin Safety: An Assessment Using an Administrative Claims Database
Mark J. Cziraky, Vincent J. Willey, James M. McKenney, Siddhesh A. Kamat, Maxine D. Fisher, John R. Guyton,
Terry A. Jacobson, and Michael H. Davidson
THE ANALYSIS:
69C
An Assessment of Statin Safety by Muscle Experts
Paul D. Thompson, Priscilla M. Clarkson, and Robert S. Rosenson
77C
An Assessment of Statin Safety by Hepatologists
David E. Cohen, Frank A. Anania, and Naga Chalasani
82C
An Assessment of Statin Safety by Nephrologists
Bertram L. Kasiske, Christoph Wanner, and W. Charles O’Neill
86C
An Assessment of Statin Safety by Neurologists
Lawrence M. Brass,† Mark J. Alberts, and Larry Sparks
CONCLUSIONS:
89C
Final Conclusions and Recommendations of the National Lipid Association Statin Safety
Assessment Task Force
James M. McKenney, Michael H. Davidson, Terry A. Jacobson, and John R. Guyton
96C
Benefit versus Risk in Statin Treatment
John R. Guyton
†
Deceased.
Report of the National Lipid Association’s Statin Safety Task Force
Guest Editor
James M. McKenney, PharmD
Professor Emeritus
Virginia Commonwealth University
President and CEO
National Clinical Research, Inc.
Richmond, Virginia
Faculty
Mark J. Alberts, MD
Northwestern University
Chicago, Illinois
David E. Cohen, MD, PhD
Brigham and Women’s Hospital
Boston, Massachusetts, USA
Frank A. Anania, MD
Emory University School of Medicine
Atlanta, Georgia
Mark J. Cziraky, PharmD
HealthCore, Inc.
Wilmington, Delaware
Harold Bays, MD
Louisville Metabolic and Atherosclerosis
Research Center
Louisville, Kentucky
Michael H. Davidson, MD
Rush University Medical Center
Chicago, Illinois
Michael B. Bottorff, PharmD
University of Cincinnati
Cincinnati, Ohio
†
Lawrence M. Brass, MD
Yale University
New Haven, Connecticut
Naga Chalasani, MD
Indiana University School of Medicine
Indianapolis, Indiana
John A. Clark, MD, MSPH
Galt Associates
Blue Bell, Pennsylvania
Priscilla M. Clarkson, PhD
University of Massachusetts
Amherst, Massachusetts
†
Deceased.
Maxine D. Fisher, PhD
HealthCore, Inc.
Wilmington, Delaware
Lucas M. Glass, BA
Galt Associates
Blue Bell, Pennsylvania
Antonio M. Gotto, Jr., MD, DPhil
Weill Medical College of Cornell University
New York, New York
John R. Guyton, MD
Duke University Medical Center
Durham, North Carolina
Terry A. Jacobson, MD
Emory University
Atlanta, Georgia
Siddhesh A. Kamat, MS
HealthCore, Inc.
Wilmington, Delaware
Anju Kanumalla, MS
Galt Associates
Blue Bell, Pennsylvania
Bertram L. Kasiske, MD
University of Minnesota
Minneapolis, Minnesota
Malcolm Law, MD
Wolfson Institute of Preventive Medicine
Barts and The London School of Medicine
London, United Kingdom
W. Charles O’Neill, MD
Emory University
Atlanta, Georgia
Robert S. Rosenson, MD
Northwestern University
The Feinburg School of Medicine
Chicago, Illinois
Alicja R. Rudnicka, PhD
Wolfson Institute of Preventive Medicine
Barts and The London School of Medicine
London, United Kingdom
Lary Sparks, PhD
Sun Health Research Institute
Sun City, Arizona
Paul D. Thompson, MD
Hartford Hospital
Hartford, Connecticut
Christoph Wanner, MD
University of Würzburg
Würzburg, Germany
Vincent J. Willey, PharmD
HealthCore, Inc.
Wilmington, Delaware
Faculty Disclosures
Naga Chalasani, MD, has no financial arrangement or affiliation with a corporate organization or a manufacturer of a product disMark J. Alberts, MD, is a member of the cussed in this supplement.
Speakers’ Bureau for AstraZeneca and BristolMyers Squibb; and serves as a consultant to
AstraZeneca, Bristol-Myers Squibb, and Pfizer John A. Clark, MD, MSPH, holds stock in
Pfizer Inc., and Schering-Plough.
Inc.
The authors who contributed to this publication
have disclosed the following industry relationships:
Frank A. Anania, MD, has no financial ar- Priscilla M. Clarkson, PhD, serves as a conrangement or affiliation with a corporate orga- sultant to Merck & Co. and has received renization or a manufacturer of a product dis- search/grant support from Merck & Co.
cussed in this supplement.
Harold Bays, MD, is a member of the Speakers’ Bureau for AstraZeneca, Kos Pharmaceuticals, Merck & Co., Reliant Pharmaceuticals,
and Schering-Plough; serves as a consultant to
AstraZeneca, Kos Pharmaceuticals, Merck &
Co., Microbia, Pfizer Inc., Sankyo Pharma, and
Schering-Plough; and has received research/
grant support from Abbott Laboratories, AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, Kos Pharmaceuticals, Merck & Co.,
Novartis, Pfizer Inc., Reliant Pharmaceuticals,
Sankyo Pharma, and Schering-Plough.
David E. Cohen, MD, PhD, is a member of the
Speakers’ Bureau for Merck & Co., ScheringPlough, Merck/Schering-Plough, sanofi-aventis, and Schering-Plough; serves as a consultant
to Merck & Co. and Schering-Plough; and has
received research/grant support from Pfizer Inc.
Michael B. Bottorff, PharmD, is a member of
the Speakers’ Bureau for AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, Kos
Pharmaceuticals, Novartis, Pfizer Inc., and
sanofi-aventis.
Michael H. Davidson, MD, is a member of the
Speakers’ Bureau for Abbott Laboratories, AstraZeneca, Bristol-Myers Squibb, Kos Pharmaceuticals, Merck & Co., Pfizer Inc., Reliant
Pharmaceuticals, Sankyo Pharma, ScheringPlough, Sumitomo Pharmaceuticals, and
Takeda; serves as a consultant to Abbott Laboratories, AstraZeneca, Bristol-Myers Squibb,
Kos Pharmaceuticals, Merck & Co., Pfizer
Inc., Reliant Pharmaceuticals, Sankyo Pharma,
Schering-Plough, Sumitomo Pharmaceuticals,
and Takeda Pharmaceuticals America, and
has received research/grant support from Abbott Laboratories, AstraZeneca, Bristol-Myers
Squibb, Kos Pharmaceuticals, Merck & Co.,
Pfizer Inc., Reliant Pharmaceuticals, Sankyo
Pharma, Schering-Plough, Sumitomo Pharmaceuticals, and Takeda Pharmaceuticals America.
Lawrence M. Brass, MD,† was a member
of the Speakers’ Bureau for Bristol-Myers
Squibb, sanofi-aventis, Solvay Pharmaceuticals, and Wyeth, served as a consultant to AstraZeneca, Bayer, Bristol-Myers Squibb, Merck
& Co., Ono Pharmaceuticals, sanofi-aventis,
Solvay Pharmaceuticals, and Wyeth, and received research/grant support from BristolMyers Squibb and sanofi-aventis.
†
Deceased.
Mark J. Cziraky, PharmD, has no financial
arrangement or affiliation with a corporate organization or a manufacturer of a product discussed in this supplement.
Maxine D. Fisher, PhD, has no financial arrangement or affiliation with a corporate organization or a manufacturer of a product discussed in this supplement.
Anju Kanumalla, MS, has no financial arrangement or affiliation with a corporate organization or a manufacturer of a product discussed in this supplement.
Lucas M. Glass, BA, has no financial arrangement or affiliation with a corporate organization or a manufacturer of a product discussed in
this supplement.
Bertram L. Kasiske, MD, serves as a consultant to Wyeth; and has received research/grant
support from Bristol-Myers Squibb and Merck/
Schering-Plough.
Antonio M. Gotto, Jr., MD, DPhil, serves as
a consultant to Bristol-Myers Squibb, Johnson
& Johnson, Kos Pharmaceuticals, KOWA
Pharmaceuticals, Merck & Co., Merck/Schering-Plough, Novartis, Reliant Pharmaceuticals,
and Pfizer Inc.; and serves on the Board of
Directors for Medtronic.
Malcolm Law, MD, has no financial arrangement or affiliation with a corporate organization or a manufacturer of a product discussed in
this supplement.
James M. McKenney, PharmD, is a member
of the Speakers’ Bureau for AstraZeneca, Kos
Pharmaceuticals, Merck & Co., Pfizer Inc., Reliant Pharmaceuticals, and Schering-Plough;
serves as a consultant to AstraZeneca, Kos
Pharmaceuticals, Microbia, Pfizer Inc., and
Sankyo Pharma; and has received research/
grant support from AstraZeneca, GlaxoSmithKline, Kos Pharmaceuticals, Merck & Co., Reliant Pharmaceuticals, Hoffmann-La Roche,
Pfizer Inc., Schering-Plough, and Takeda Pharmaceuticals America.
John R. Guyton, MD, is a member of the
Speakers’ Bureau for Abbott Laboratories, AstraZeneca, Kos Pharmaceuticals, Merck & Co.,
Pfizer Inc., Schering-Plough, and Takeda Pharmaceuticals America; serves as a consultant to
Merck/Schering-Plough, Oryx Pharmaceuticals, Sankyo Pharma, and Takeda Pharmaceuticals America, and Sankyo Pharma; has
received research/grant support from AstraZeneca, Kos Pharmaceuticals, Merck & Co.,
Hoffmann-LaRoche, and Pfizer Inc., and holds W. Charles O’Neill, MD, has no financial
stock in Merck & Co. and Eli Lilly and Com- arrangement or affiliation with a corporate orpany.
ganization or a manufacturer of a product discussed in this supplement.
Terry A. Jacobson, MD, is a member of the
Speakers’ Bureau for Abbott Laboratories, AstraZeneca, Merck & Co., Merck/ScheringPlough, Reliant Pharmaceuticals, and Takeda
Pharmaceuticals America; and serves as a consultant to Abbott Laboratories, AstraZeneca,
Kos Pharmaceuticals, Pfizer Inc., and Reliant
Pharmaceuticals.
Robert S. Rosenson, MD, is a member of the
Speakers’ Bureau for Abbott Laboratories, AstraZeneca, Kos Pharmaceuticals, Merck & Co.,
and Sankyo Pharma; and has received research
grant support from Abbott Laboratories, AstraZeneca, Kos Pharmaceuticals, Merck & Co.,
and Sankyo Pharma.
Siddhesh A. Kamat, MS, has no financial arrangement or affiliation with a corporate organization or a manufacturer of a product discussed in this supplement.
Alicja R. Rudnicka, PhD, has no financial
arrangement or affiliation with a corporate organization or a manufacturer of a product discussed in this supplement.
Larry Sparks, PhD, serves as a consultant to Christoph Wanner, MD, serves as a consulPfizer Inc.
tant to Genzyme, and has received research
grant support from Genzyme; and has received
Paul D. Thompson, MD, is a member of the honoraria from Pfizer Inc.
Speakers’ Bureau for AstraZeneca, Merck &
Co., Pfizer Inc., and Schering-Plough; serves as Vincent J. Willey, PharmD, has no financial
a consultant to AstraZeneca and Pfizer Inc; arrangement or affiliation with a corporate orreceived research/grant support from Kos Phar- ganization or a manufacturer of a product dismaceuticals, Merck & Co., and Pfizer Inc; and cussed in this supplement.
holds stock in Merck & Co., Pfizer Inc., and
Schering-Plough.
Introduction
James M. McKenney, PharmD
Health professionals who seek to reduce the consequences of atherosclerotic vascular disease in their patients
should be enjoying the best of times. An explosion of
discovery and new knowledge has helped define the pathogenesis of atherosclerosis and provide new insights into its
insidious effects. New risk markers, new diagnostic approaches, new treatment guidelines, and new lipid-altering
drug therapies have advanced our ability to reduce the risk
of coronary artery disease (CAD) events, the nation’s number 1 killer. In particular, the availability of the 3-hydroxy3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors, or statins, with their superior ability to modify lipid
levels as well as mechanisms of disease, has given us the
power to reduce CAD events and strokes and to improve the
length and quality of the lives of our patients by substantial
margins.
But the question arises: Are the statins safe? Or, better, is
the ratio of safety to benefit associated with statin therapy
sufficient to justify their widespread use? These questions,
which are being asked by many health professionals and
patients, are the subject of this supplement to The American
Journal of Cardiology.
In recent years, many patients and health professionals
have questioned the safety of statins. Reports from the field
suggest that some patients are refusing to initiate statin
therapy, whereas others are choosing to withdraw from
long-term statin treatment out of concerns about safety.
These concerns appear to have arisen from information that
individuals obtain from the news media, direct-to-consumer
advertising, and the Internet. Health professionals are also
expressing concerns, perhaps out of a response to their
patients’ sentiments but also because of the withdrawal of
cerivastatin from the market due to serious adverse experiences, recent publications reporting statin-related adverse
effects, and the constant threat of litigation from malpractice
lawyers.
Although the National Lipid Association (NLA) is unable to accurately gauge the depth and breadth of these
concerns, it is assumed that they are common and may be
discouraging the use of a potentially effective, life-saving
treatment. Specialists in the field hold that when statins and
other lipid-altering agents are used properly and in conjunction with lifestyle modification, the quality and length of
patient’s lives can be significantly improved.
Virginia Commonwealth University, Richmond, Virginia, USA.
Address for reprints: James M. McKenney, PharmD, National Clinical
Research, Virginia Commonwealth University, 2809 Emerywood Parkway, Suite 140 Richmond, Virginia 23294.
E-mail address: jmckenney@ncrinc.net.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.004
To address the concerns about the safety of lipid-altering
therapies, the NLA appointed a Safety Assessment Task
Force to evaluate statin safety and, in a second report, the
safety of nonstatin lipid-altering drugs. This initiative is
undertaken as a service to its members—physicians, nurses,
pharmacists, dietitians, and the many other health professionals who strive daily to reduce the risk of CAD in their
patients through prescription medications and recommendations for lifestyle changes.
The charge given to the Safety Assessment Task Force
was to conduct a rigorous, scholarly, up-to-date, and unbiased assessment of the safety of statins and statin combination therapy and, at a later date, to conduct a similar assessment of nonstatin therapy. The Task Force presented their
findings regarding statin safety at a meeting convened from
July 17–19, 2005, at the Mandarin Hotel in Washington,
DC. The report from this meeting is published in the present
supplement. The members of this Task Force were Dr.
James M. McKenney (Chair; Virginia Commonwealth University, Richmond, VA), Dr. Michael H. Davidson (Rush
University, Chicago, IL), Dr. Terry A. Jacobson (Emory
University, Atlanta, GA), and Dr. John R. Guyton (Duke
University, Durham, NC).
To assure a rigorous, comprehensive assessment of statin
safety, the Task Force further commissioned reviews of the
specialist literature on adverse reaction (Dr. Harold Bays)
and drug interaction (Dr. Michael B. Bottorff). The Task
Force also commissioned new research to be undertaken.
Reports of this work include an up-to-the-minute systematic
review of published cohort and clinical trial data on statin
safety by Drs. Malcolm Law and Alicja R. Rudnicka; an
assessment of the most recently available data from the
James M. McKenney, PharmD
www.AJConline.org
2C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
FDA’s Adverse Event Reporting System (AERS) by Dr.
Davidson and coworkers; an inspection of the data contained in the NewDrug Applications (NDAs) and the FDA’s
Summary Basis of Approvals for marketed statins by Dr.
Jacobson; and an analysis of statin use and associated adverse health events in a 22-million person managed care
database by Dr. Mark J. Cziraky and associates. The Task
Force further appointed expert panels made up of highly
credentialed medical subspecialists to independently examine the evidence, answer specific questions about safety
posed by the Task Force, and provide recommendations to
health professionals. Panel members selected for this appointment were men and women of letters, established
scholars from the major universities of the world who are
recognized for their expertise and professional records.
There were 4 panels assembled, composed of 3 members
each, focusing on muscle, liver, renal, and neurologic effects of statins; the reports of these respective expert panels
are contained in this supplement in the articles by Drs. Paul
D. Thompson, David E. Cohen, Bertram L. Kasiske, and
Larry Brass and their colleagues. It is with regret that the
NLA acknowledges the death of Dr. Brass during the development of this supplement. We are indebted to him for
the leadership and expert contribution he made. We also
wish to acknowledge with gratitude the work of his colleague Dr. Mark J. Alberts, who helped draft and edit the
final manuscript during Dr. Brass’ illness.
The supplement begins with a commentary on drug
surveillance in the United States by Dr. Antonio M.
Gotto, Jr. and concludes with an evaluation of the riskbenefit considerations in statin therapy by Dr. Guyton.
The final conclusions and recommendations from the
NLA Statin Safety Task Force are an attempt to glean key
information from commissioned reviews and research
reports, and the expert panel assessments into a concluding summary statement.
It is our hope that this supplement will provide health
readers with a rigorous, scholarly, up-to-date, and unbiased
assessment of the safety of statins so that, if warranted,
health professionals and their patients can execute effective
cardiovascular risk–reducing therapies with new knowledge
and confidence.
Statins, Cardiovascular Disease, and Drug Safety
Antonio M. Gotto, Jr., MD, DPhil
Available for almost 2 decades, the 3-hydroxy-3-methylglutaryl coenzyme A reductase inhibitors, or statins, have emerged at the forefront of preventive drugs for
cardiovascular disease because of a substantial clinical trial database demonstrating
that statins reduce the risk for coronary artery disease morbidity and death across a
broad range of at-risk patient cohorts. Although generally well tolerated, statins may
be associated with infrequent adverse events that warrant serious and frank discussion, including myopathy and rhabdomyolysis. In 2005, the National Lipid Association (NLA), a multidisciplinary, nonprofit association of healthcare providers and
researchers in the lipid field, convened a Safety Task Force to undertake an intensive,
fair-minded evaluation of available data on the effects of statins on muscle, liver,
kidneys, and the brain. In the end, physicians and patients must weigh the potential
clinical benefits of statin treatment against the potential risks when deciding whether
to initiate treatment. © 2006 Elsevier Inc. All rights reserved. (Am J Cardiol 2006;
97[suppl]:3C–5C)
Available for almost 2 decades, the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors, or
statins, have emerged at the forefront of preventive drugs for
cardiovascular disease (CVD) because of a substantial clinical trial database that demonstrates without question that,
compared with placebo, statins reduce the risk for coronary
artery disease (CAD) morbidity and death across a broad
range of at-risk patient cohorts.1 The withdrawal of 1 agent
of this class in 2001 because of excess fatal toxicity compelled the media and consumer interest groups to increase
their scrutiny of statin safety, of potential conflicts of interest with the pharmaceutical industry, and of the independence of government regulatory agencies.2,3 This negative
attention, especially when considered with an expert panel’s
coincidental release of recommendations for broader use of
statins in at-risk individuals,4 has understandably generated
some confusion. Are the risks of statins excessive? If so,
then why do recommendations indicate that more patients
should be treated with them? The cognitive dissonance
means that there is a clear necessity to provide clinicians
with balanced, unambiguous, and comprehensive guidance
regarding the safety of these lipid-modifying drugs.
To help meet that challenge, the National Lipid Association (NLA), a multidisciplinary, nonprofit association of
healthcare providers and researchers in the lipid field, recently convened a Safety Task Force to undertake an intensive, fair-minded evaluation of available data about the
effects of statins on muscle, liver, kidneys, and the brain.
Weill Medical College of Cornell University, New York, New York,
USA.
Address for reprints: Antonio M. Gotto, Jr., MD, DPhil, Weill Medical
College of Cornell University, 445 East 69th Street, OH205, New York,
New York 10021.
E-mail address: amg_editorial@med.cornell.edu.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.005
The Task Force’s important discussions about statin safety
were designed to bring forth insights into both the unique
properties of the statins themselves and the systemic challenges that confront the monitoring of drug safety. While all
would agree that medicines that are made available to the
public should be both efficacious and safe, concerns about
the toxicity of statins should be considered in the overall
balance of the potential harm versus the potential benefit.
STATINS: A VIEW ON SAFETY AND CLINICAL
BENEFIT
Recognized early on as the main hazards of these drugs, the
low but real risks for myopathy or liver toxicity did little to
dampen the enthusiasm for the clinical potential of this class
as positive results from landmark clinical trials followed
one after another from 1994 to 2001.1 Myotoxicity has
garnered the most attention as an important adverse reaction
of this drug class, because of cerivastatin’s removal due to
its disproportionately greater risk for fatal rhabdomyolysis
compared with the other statins. Small studies and anecdotal
evidence have raised anxieties about potential effects on
kidney function (such as proteinuria) and on cognition (such
as unusual reports of memory loss) with statin use, but
investigators have yet to establish conclusively the clinical
relevance or pathologic mechanisms of these outcomes.5,6
The US Food and Drug Administration (FDA), in its review
of safety studies related to rosuvastatin, noted that available
data showed no consistent pattern of clinical presentation of
renal failure or renal injury that clearly indicate causation by
statins.7
The incidence of muscle toxicity increases with increasing dosage of statin monotherapy. Concurrent use of certain
drugs such as fibrates, erythromyocin, itraconazole, and
www.AJConline.org
4C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
immunosuppressive drugs such as cyclosporine can increase
blood levels of statins and likewise raise the risk for myopathy. Indeed, ⬎33% of deaths from rhabdomyolysis that led
to the withdrawal of cerivastatin occurred where gemfibrozil use was also present. Cerivastatin’s withdrawal was
based on 31 known deaths among several million patients
who had received the drug; by January 2002, approximately
100 deaths had been attributed to cerivastatin.2 An analysis
of pharmacy benefit data from 11 managed-care health
plans concluded that while the rhabdomyolysis rate with
cerivastatin was comparatively higher (5.34 per 10,000 person-years of treatment), it was low and similar between
atorvastatin, pravastatin, and simvastatin (average incidence
of 0.44 per 10,000 person-years of treatment).8
Consider this risk in the context of the incidence of CAD
in the United States. Within the next year, approximately
700,000 individuals in the United States will have a new
coronary attack and about 500,000 will have a recurrent
attack9; coronary disease caused 1 in 5 deaths in 2002. Also
consider that most of the clinical trials of statins report an
approximately 30% reduction in relative risk for coronary
events with statin treatment compared with placebo.1 Moreover, 3 of the landmark statin trials reported a reduction in
the relative risk for all-cause mortality: a 30% reduction in
the Scandinavian Simvastatin Survival Study (4S)10; a 22%
reduction in the Long-term Intervention with Pravastatin in
Ischaemic Disease (LIPID)11; and a 13% reduction in the
Heart Protection Study (HPS).12 Some trials also have reported a reduction in stroke risk. Therefore, the potential of
statins to protect patients from atherosclerosis appears to be
far greater than the risks for its most serious adverse event.
ISSUES IN PHARMACOVIGILANCE
Bringing a drug to market in the United States requires that
drug companies undertake several phases of clinical trials
with several thousands of subjects to demonstrate efficacy,
tolerability, and safety before the FDA will consider approving it for sale. However, as in the case of cerivastatin,
millions of patients may need to be exposed to a drug before
the rate of a rare side effect can be assessed, and this may
be feasible only after the drug has been released.13 The
current US postmarketing system for monitoring drug side
effects is the FDA’s Adverse Event Reporting System
(AERS). The AERS database has helped investigators uncover the greater hazards associated with a number of drugs,
including cerivastatin. This system may provide a long-term
perspective about a drug’s safety record, including capturing many potential signals of adverse reactions, and its use
may be readily integrated into the routine operations of a
medical office.14
Nevertheless, the AERS database is imperfect. Submission of cases is voluntary and may not occur in a well-timed
manner, the protocol for documenting an adverse event may
capture only a superficial amount of information, and reports are often incomplete and may require contacting the
event reporter, who may have little time or inclination for
further follow-up. Also, spontaneous reporting captures
only a small fraction of the adverse events that actually
occur.15 Although the exact rate of underreporting is unknown, a rough estimate suggests that only about 10% of
adverse events are reported.13
A recent analysis of the AERS database that suggested a
greater adverse event rate with rosuvastatin compared with
other statins helps exemplify some of these potential biases.
The detected increase may have reflected a selective increase in reporting of cases due to greater press coverage of
rosuvastatin’s safety compared with other statins in recent
years, rather than a true difference in rosuvastatin compared
with the other drugs.16,17 High reporting rates may indicate
a culture committed to identifying and reducing errors and
adverse events, rather than a truly high rate.14
Although AERS may help identify signals of enhanced
toxicity, AERS cannot predict which signals will warrant
more attention than others for any single drug. It is very
easy to see what should have been done with a safety signal
after information becomes available (hindsight bias), but it
is very difficult to decide which signals to follow up in order
to get that additional information.18 Edwards has criticized
safety process decisions by governmental regulatory agencies as superficial because they do not offer either detailed
comparisons of drugs or more than basic information on risk
and effectiveness.18 Furthermore, the FDA has limited powers to enforce its requests to pharmaceutical companies to
publicize safety concerns, and its mechanisms for alerting
physicians to new safety information (e.g., letters to physicians, “black-box” warnings) may not induce the desired
increase in awareness.13 Increased federal funding for drug
safety monitoring has been proposed for 2006, and the FDA
has moved to implement a new Drug Safety Oversight
Board to help identify, track, and oversee the management
of important drug safety issues in a timely and independent
manner.
Exploring alternative sources of data besides AERS has
been put forth as another solution; such an approach might
include, for example, reviewing patient records and monitoring other databases, such as Medicare or Medicaid
claims, but these strategies also may be problematic.14
Record review is labor intensive and entries may be biased
or incomplete. Administrative claims data also may be incomplete, are divorced from clinical context, and may be
biased by reimbursement policies and regulations that provide incentives to code for conditions and complications
that increase payments to hospitals.14 Ultimately, an approach that incorporates multiple surveillance modalities
may prove to be the best means of monitoring a drug’s
safety after its release.
Gotto, Jr./Statins, Cardiovascular Disease, and Drug Safety
CONCLUSION
Recent clinical trial data suggest that aggressive low-density
lipoprotein (LDL) cholesterol reduction with statins lowers
CVD risk to a greater extent than do moderate approaches in
patients with stable coronary disease and acute coronary
syndromes. These findings have led to a revised statement
from the National Cholesterol Education Program (NCEP)
that advocates even lower LDL-cholesterol goal options in
the highest-risk and moderate-risk patients.19 A call for
more aggressive LDL-cholesterol targets that will increase
the dosages of statin monotherapy or the use of combination
treatments may increase the risk for adverse events. Therefore, a clear understanding of all of the issues surrounding
statin safety is needed.
On the whole, statins have a very good safety profile.
Such an assurance, however, is little comfort to those few
patients who have experienced a negative outcome while on
a drug. The issue is one of both education and transparency
for physicians and patients. Once made aware of the possible dangers of the drugs, as well as the factors that may
increase the risk for an adverse reaction,20 the patient will be
better prepared to make informed decisions about his or her
own treatment. Although the determination of an acceptable
risk-to-benefit ratio is, without a doubt, the concern of
federal drug regulators as well as pharmaceutical companies, it is also an individual judgment call that must be made
by each patient in conjunction with the advice of his or her
physician. In my opinion, assessing how much risk a patient
is willing to tolerate for the benefit of a treatment has been
underappreciated as a clinical skill that physicians should
develop. In the end, the dangers of drug toxicity should be
balanced against the dangers of withholding drug therapy
with proven efficacy. In the case of statins, the prevention of
the physical, economic, and psychological burden of symptomatic CVD for many patients in the face of the distress of
those few patients who have been harmed by the drugs
makes a frank discussion of the potential side effects of
therapy essential.
1. Vaughan CJ, Gotto AM Jr. Update on statins: 2003. Circulation
2004;110:886 – 892.
2. Staffa JA, Chang J, Green L. Cerivastatin and reports of fatal rhabdomyolysis. N Engl J Med 2002;346:539 –540.
3. Wolfe SM. Remedies needed to address the pathology in reporting
adverse reactions and Food and Drug Administration use of reports.
J Gen Intern Med 2003;18:72–73.
4. Expert Panel on Detection, Evaluation, and Treatment of High Blood
Cholesterol in Adults. Executive Summary of the Third Report of the
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
5C
National Cholesterol Education Program (NCEP) Expert Panel on
Detection, Evaluation, and Treatment of High Blood Cholesterol in
Adults (Adult Treatment Panel III). JAMA 2001;285:2486 –2497.
Verhulst A, D’Haese PC, De Broe ME. Inhibitors of HMG-CoA
reductase reduce receptor-mediated endocytosis in human kidney
proximal tubular cells. J Am Soc Nephrol 2004;15:2249 –2257.
Wagstaff LR, Mitton MW, Arvik BM, Doraiswamy PM. Statin-associated memory loss: analysis of 60 case reports and review of the
literature. Pharmacotherapy 2003;23:871– 880.
US Food and Drug Administration. FDA Public Health Advisory on
Crestor (rosuvastatin) [FDA Web site]. Available at: http://www.
fda.gov/cder/drug/advisory/crestor_3_2005.htm. Accessed August 19,
2005.
Graham DJ, Staffa JA, Shatin D, Andrade SE, Schech SD, La Grenade
L, Gurwitz JH, Chan KA, Goodman MJ, Platt R. Incidence of hospitalized rhabdomyolysis in patients treated with lipid-lowering drugs.
JAMA 2004;292:2585–2590.
American Heart Association. Heart and Stroke Facts 2005 Statistical
Update [American Heart Association Web site]. Available at: http://
www.americanheart.org/presenter.jhtml?identifier⫽30000902004.
Accessed July 15, 2005.
Scandinavian Simvastatin Survival Study Group. Randomised trial of
cholesterol lowering in 4444 patients with coronary heart disease: the
Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:
1383–1389.
The Long-Term Intervention with Pravastatin in Ischaemic Disease
(LIPID) Study Group. Prevention of cardiovascular events and death
with pravastatin in patients with coronary heart disease and a broad
range of initial cholesterol levels. N Engl J Med 1998;339:1349 –1357.
Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk
individuals: a randomised placebo-controlled trial. Lancet 2002;360:
7–22.
Zielinski SL. FDA attempting to overcome major roadblocks in monitoring drug safety. J Natl Cancer Inst 2005;97:872– 873.
Thomas EJ, Petersen LA. Measuring errors and adverse events in
health care. J Gen Intern Med 2003;18:61– 67.
Ahmad SR. Adverse drug event monitoring at the Food and Drug
Administration. J Gen Intern Med 2003;18:57– 60.
Alsheikh-Ali AA, Ambrose MS, Kuvin JT, Karas RH. The safety of
rosuvastatin as used in common clinical practice: a postmarketing
analysis. Circulation 2005;111: 3051–3057.
Grundy SM. The issue of statin safety: where do we stand? Circulation
2005;111: 3016 –3019.
Edwards IR. What are the real lessons from Vioxx? Drug Saf 2005;
28:651– 658.
Grundy SM, Cleeman JI, Bairey Merz N, Brewer B Jr, Clark LT,
Hunninghake DB, Pasternak RC, Smith SC Jr, Stone NJ. Implications
of recent clinical trials for the National Cholesterol Education Program
Adult Treatment Panel III guidelines. Circulation 2004;110:227–239.
Pasternak RC, Smith SC Jr, Bairey-Merz CN, Grundy SM, Cleeman
JI, Lenfant C, for the American College of Cardiology; American
Heart Association, and the National Heart, Lung and Blood Institute.
ACC/AHA/NHLBI Clinical Advisory on the Use and Safety of
Statins. Circulation 2002;106:1024 –1028.
Statin Safety: An Overview and Assessment of the Data—2005
Harold Bays, MD
The 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors, or
statin drugs, have been studied in numerous controlled human research trials involving hundreds of thousands of study participants. Statins have been prescribed for
millions of patients. Based on this vast research and clinical experience, statins have
been shown to improve lipid blood levels and reduce atherosclerotic coronary artery
disease (CAD) risk, resulting in reduced CAD morbidity and mortality, and in several
studies, reduced overall (“all-cause”) mortality. From a safety perspective, both
research trial evidence and clinical practice experience have demonstrated that
statins are generally well tolerated. However, as with all pharmaceuticals, safety
considerations exist with both monotherapy and combination statin therapy, mainly
involving potential adverse effects on muscle, liver, kidney, and the nervous system.
The evidence supporting statin-related potential adverse experiences on these organ
systems is sometimes strong and based on clear clinical trial evidence (such as the
increased risk of muscle enzyme elevation with higher statin doses). The evidence is
at other times more speculative, being based on case reports and inconclusive clinical
trial data (such as possible favorable or unfavorable effects of statins on cognition).
Because the use of statins is so widespread, it is useful for the clinician to understand
statin safety issues and the level of available evidence supporting the contention that
various adverse effects are caused by statins. This review presents an assessment of
statin safety based on an overview of the current statin safety data and their clinical
implications. © 2006 Elsevier Inc. All rights reserved. (Am J Cardiol 2006;
97[suppl]:6C–26C)
The safety and tolerability of pharmaceutical agents is perhaps the most important consideration in their clinical use
(primum non nocere, or “First, do no harm”). In fact, in the
research and development of novel drugs, it is often said
that safety trumps efficacy. Introduced in 1987, lovastatin
was the first 3-hydroxy-3-methylglutaryl coenzyme A
(HMG-CoA) reductase inhibitor, or statin, approved for use
in the United States. Since then, decades of clinical trial
evidence involving hundreds of thousands of study participants, and the practical clinical experiences of millions of
patients treated with statins, have demonstrated that statins
are generally well tolerated.1 However, as with any pharmaceutical agent, adverse experiences have been associated
with statins, both in monotherapy and in combination therapy with other agents.2– 6
Safety concerns regarding the use of statin treatment
were heightened by the withdrawal of cerivastatin from the
world market in 2001, owing to a rate of fatal rhabdomyolysis that, in postmarketing voluntary reports to the US
Food and Drug Administration (FDA), was found to be
much more frequent than with other statins.7,8 This isolated
Louisville Metabolic and Atherosclerosis Research Center, Louisville,
Kentucky, USA.
Address for reprints: Harold Bays MD, Louisville Metabolic and Atherosclerosis Research Centre, 3288 Illinois Avenue, Louisville, Kentycky
40213.
E-mail address: hbaysmd@aol.com.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.006
withdrawal of a previously approved statin drug suggests
that the degree of risk of potential adverse experiences (in
this case rhabdomyolysis) varies between statins. These
differences in safety risk are based on the marketed statin
doses and the statin pharmacology profile such as bioavailability, metabolism, excretion rate and mode, as well as the
patient population treated and the concurrent use of agents
having a potential for drug interactions.3,5,9 –13 The circumstance surrounding the withdrawal of cerivastatin also illustrates that even when early clinical trials suggest reasonable
safety, independent postmarketing surveillance reports are
critical to detecting potential severe adverse experiences
that may be revealed only after millions of patients have
been exposed to the drug.7,8,14,15
Multiple data sources must be evaluated to best understand and assess potential safety issues of pharmaceuticals.
Such data sources may include prior animal studies, earlyphase clinical trials (such as phase 1 studies and phase 2
dose-ranging studies), drug interaction studies, results of
later clinical trials (such as phase 3 and phase 4 studies),
meta-analyses of a number of clinical trials, clinician reporting of adverse human experiences to regulatory agencies (such as voluntary adverse event reporting to the FDA),
analysis of managed care database claims,16 isolated case
reports, as well as the unpublished knowledge of research
investigators, impressions and experiences of clinicians, and
the drug prescribing information. Statin safety has been
assessed by all of these data sources.
www.AJConline.org
Bays/Statin Safety: Overview of the Data
7C
Table 1
Evidence grading system describing the level of evidence that potential adverse experiences are
associated with statin use
Level of
evidence
Supporting Evidence That Potential Human Adverse Experience
Is Related to Use of Statins
A
Clear evidence from well-conducted, generalizable, randomized, controlled
trials that are adequately powered including safety evidence derived from:
● Well-conducted prospective, randomized multicenter clinical trials
● Meta-analysis of such trials that incorporated safety assessments in the
analysis
Supporting evidence from:
● Well-conducted, prospective cohort studies or registry that included safety
assessments
● Well-conducted meta-analysis of cohort studies that included safety
assessments
● Well-conducted retrospective case-control studies that included safety
assessments
● Managed care claims database analyses of safety issues with highly
statistically significant results
● Reports to regulatory agencies of “hard” safety end points (ie, death) that
clearly exceed that of population averages and/or comparator treatments*
Supporting safety evidence from poorly controlled or uncontrolled studies:
● Randomized clinical trials with major or minor flaws that could invalidate
the results
● Observational studies with high potential for bias, such as case series with
comparison to historical controls
● Case series or case reports
● Nondefinitive safety trends from well-conducted clinical trials, managed
care database analyses, or safety reports to regulatory agencies
● Expert consensus or experience of clinicians
● No evidence, or evidence to the contrary
● Unknown
B
C
E
F
U
* Data derived from voluntary reports to regulatory agencies are by their nature, inherently suspect.
Nonetheless, a “B”-rating is given in this review for such reports of a specific, and otherwise rare,
safety adverse experience (rhabdomyolysis), that results in a severe “hard” safety end point that
would be difficult to misdiagnose (death), that is clearly excessive, and especially if it prompts
intervention from a regulatory agency (such as a warning letter from the US Food and Drug
Administration), and eventual withdrawal of the drug from the market. In this case, clinicians might
reasonably and practically conclude that this safety concern is supported by a high “level of
evidence,” when engaged in the decision-making process directed toward the day-to-day care of their
patients. (Adapted from the American Diabetes Association Clinical Practice Recommendation for
standards of Medical Care in Diabetes.17)
Grading of the Scientific Evidence
The strength of the current scientific evidence supporting
the association of statins with specific potential adverse
experiences varies with the adverse experience being considered. In some cases, the clinical trial evidence is clear; in
other cases, potential adverse experiences are unproved,
having only been described in isolated case reports. It therefore is useful to assess the level of scientific evidence of
potential safety issues associated with statins. Although no
universally accepted grading system is available to assess
the evidence of statin safety, the grading in this overview is
derived and adapted from the annual American Diabetes
Association (ADA) Position Statement regarding the Evidence Grading System for Clinical Practice Recommendations.17 Grades of A, B, C, E, F, or U are given depending
on the quality of evidence (Table 1).
In this overview, if a specific potential adverse experi-
ence is given an “A” rating, it suggests that the evidence to
support the contention that the statin caused or contributed
to this potential adverse experience is convincing and firmly
based on well-conducted clinical trial evidence, including
early phase dose-ranging studies that largely determine the
dose and safety risk of any drug. A “B” rating also suggests
strong evidence of an association between the statin and a
potential adverse experience, and is based on data that may
be sometimes difficult to obtain through controlled clinical
trials. A “C” rating denotes less convincing evidence, such
as through poorly controlled or uncontrolled studies and
isolated case reports. An “E” rating indicates a safety or
tolerability concern that is derived more from the “expert”
opinion or consensus of many investigators, or the “experience” of clinicians, as opposed to definitive confirmation
by clinical trials that have not always been designed to
measure specific adverse experiences. An “F”-level rating
means that the safety statement has no evidence, or has
8C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 2
Clinical trial definitions of potential muscle adverse experiences due to statins
Potential Muscle
Adverse Experience
Myalgias
Myopathy
Rhabdomyolysis
Definitions Used in this Overview
Muscle ache, pain, or weakness with or without CK elevation
Otherwise unexplained elevations in CK ⱖ10⫻ the ULN,
associated with muscle symptoms (myalgias)
Marked CK elevation, typically substantially ⬎10⫻ the ULN
and with creatinine elevation (usually with brown urine and
urinary myoglobin). Elevations in other muscle enzymes may
also occur, as well as the following:
● Hyperkalemia
● Hypocalcemia
● Hyperphosphatemia
● Hyperuricemia
● Metabolic acidosis
● Renal failure
● Death
● Symptoms of muscle weakness may be present, but perhaps
only 50% of the time.20
CK ⫽ creatine kinase; ULN ⫽ upper limit of normal.
evidence contrary to that which would support the claim.
Finally, a “U” rating suggests that the information is unknown or unclear.
Potential Muscle Adverse Experiences
Definitions: Before a drug can be said to be associated
with a potential toxicity, it is helpful to have a clear understanding of what defines the adverse experience. Unfortunately, the definitions ascribed to muscle adverse experiences have not always been consistent. In 2002, the
American College of Cardiology/American Heart Association/National Heart, Lung, and Blood Institute (ACC/AHA/
NHLBI) issued a clinical advisory statement on the use and
safety of statins.18 In this advisory statement, myalgia was
defined as muscle ache or weakness without creatine kinase
(CK) elevation. Myopathy was defined as any disease of the
muscles, and myositis was defined as muscle symptoms
with increased CK levels. Rhabdomyolysis was defined as
muscle symptoms associated with marked CK elevations,
typically substantially ⬎10 times the upper limit of normal
(ULN) and with creatinine elevation and the usual presence
of brown urine with urinary myoglobin.
However, in the reporting of clinical trials (and often in
the day-to-day conduct of clinical practice), alternative
muscle adverse experience definitions are often used for
each of these terms. From a research investigator and clinician standpoint, myalgia is frequently defined as muscle
pain, irrespective of whether muscle enzymes are elevated
(myo is a Greek derivative meaning “muscle” and algia is a
Greek derivative from algos, meaning “pain”). In the reporting of statin clinical trials, myopathy is routinely defined
as elevations in CK levels ⱖ10 times the ULN19 associated
with muscle symptoms that are not attributable to other
causes. (Admittedly, it is unclear why muscle symptoms
have been a necessary component for the clinical trial definition of myopathy, because there is no evidence to support
the contention that a finding of a CK value ⬎10 times the
ULN without symptoms is less pathologic to muscle than
the same CK value associated with symptoms.)
In both clinical trials and clinical practice, rhabdomyolysis is often defined similarly to the ACC/AHA/NHLBI
definition, except that whereas muscle symptoms are part of
the ACC/AHA/NHLBI rhabdomyolysis definition, symptoms are not necessarily required in clinical trial or clinical
practice reporting. This is because the initial symptoms of
muscle pain may occur in only 50% of cases of rhabdomyolysis.20 So, from a practical standpoint, if a clinician (or an
author drafting a report to a regulatory agency21) encounters
a patient with marked CK elevations ⬎10 times the ULN
accompanied by the subsequent need for hospitalization,
hydration, and subsequent renal failure with characteristic
multiple electrolyte and metabolic abnormalities, then it is
likely that this will be diagnosed and reported as rhabdomyolysis, irrespective of whether muscle symptoms were
present. The definitions used in the present overview are
these later clinical trial definitions, and are listed in Table 2.
Aspects of the relation between statins and potential
muscle adverse experiences are summarized in Table
3.2– 4,6 –9,12,13,18,22–56 Some outstanding questions regarding
the potential effect of statins on muscle include (1) the
frequency of myalgia and the source of muscle pain or
weakness, (2) whether statins differ in their potential for
muscle adverse experiences, and (3) the mechanism by
which statins may cause potential muscle adverse experiences.
Muscle symptoms and muscle laboratory abnormalities: In clinical practice, a number of muscle complaints,12,24 including pain to muscle areas and muscle
Bays/Statin Safety: Overview of the Data
9C
Table 3
Level of evidence that potential muscle adverse experiences are associated with statin use
Level of
Evidence
Potential Statin Adverse Experiences
A*
Elevations in muscle enzymes are a potential adverse
experience of statins
Rhabdomyolysis (fatal and nonfatal) is a potential
adverse experience of statins†
Myalgias/muscle weakness is a potential adverse
experience of statins
Muscle adverse experiences are more common at
higher statin doses
Some statins are safer than others with regard to
potential adverse muscle experiences‡
The combined use of statins with bile acid
sequestrants increases the risk of muscle adverse
experiences§
The combination of statins with fish oils increases
the risk of muscle adverse experiences§
The combined use of statins with niacin increases the
risk of muscle adverse experiences储
The combined use of statins with gemfibrozil
increases the risk of muscle adverse experiences
The combined use of statins with fenofibrate
increases the risk of muscle adverse experiences#
The combined use of statins with ezetimibe increases
the risk of muscle adverse experiences¶
B
E
A*
B
F
F
C
B
C
C
Select
References
3,4,12,13,18
7,18,22,23
3,4,12,18,24
3,6,9,13,25,26,27
7,9,28,29
30
No reports
31–38
7–9,28,39,40
2,41–46
47–52, 56
* Adverse experiences are often found in early, dose-ranging, multicenter trials.
The finding of rhabdomyolysis in a patient treated with statins does not necessarily always mean
that the rhabdomyolysis was caused by the statin. Rhabdomyolysis has numerous other potential
causes that may occur in patients treated with statins.53,54
‡
Some statins may be safer than other statins (eg, cerivastatin) based on marketed dose, pharmacology profile, and the patient population treated. Of the currently marketed statins, the risk of
rhabdomyolysis risk is similar among all statins, and low for statin monotherapy, with increased risk
observed when combined with fibrates or when used in patients with comorbidities such as diabetes
mellitus.55
§
An “F”-level of evidence means that there is no evidence or evidence to the contrary.
储
Although older, rare case reports suggested that niacin may increase the risk of muscle adverse
experiences with statins, controlled clinical trials of an extended-release formulation have not
supported an increased muscle adverse experience risk when combined with statin alone.
#
Fibrates other than gemfibrozil, such as fenofibrate, may have less risk of drug interaction with
statins, and thus may have a reduced risk of drug-related muscle adverse experiences.
¶
Muscle adverse experiences have been described in patients treated with ezetimibe,56 however
this does not mean that ezetimibe caused the adverse muscle adverse experience. Controlled trials of
the addition of ezetimibe to statins have not shown an increase in muscle adverse experiences
compared with statin alone.
†
weakness, with or without elevations in muscle enzyme CK
levels, are often described by patients treated with statins. In
an analysis of 468 patients in a managed health organization
who were identified as having “myopathy,” only 61 had
received statin therapy before the diagnosis and only 41 had
confirmed myopathy based on documentation of significant
elevations in CK levels. Of these 41 patients with confirmed
myopathy, only 17 had no other plausible clinical explanation such as muscle injury. The conclusion in this analysis
of a “real-world” clinical setting was that although the risk
of myopathy with statins is slightly increased when used in
combination with fibrates (0.12% vs 0.22%, respectively),
the risk for myopathy with either monotherapy or combination therapy is very low at ⬍1%.57
In clinical trials, the incidence of nonspecific muscle
aches or joint pains, or muscle weakness unassociated with
elevations in CK levels, has been reported in only about 5%
of study participants,18 with a range as wide as 0.3%–
33%.58 In clinical trials, the reports of muscle complaints
are typically found to be similar to the rates found in
comparator placebo groups.18 However, it should be understood that most statin clinical trials were not designed to
specifically assess muscle-related complaints. In fact, given
that myalgias and muscle weakness are among the most
common adverse experiences described by statin-treated
patients, it is curious that no validated survey or standardized diagnostic instrument or criterion has yet been established to assess myalgias or muscle weakness in statin trials.
Statin clinical trial investigators and protocols have often
excluded patients with prior intolerance to statins. Also,
10C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
patients with prior intolerance to statins have frequently
chosen not to participate in statin clinical trials.3 Thus,
complaints of statin-associated myalgias or muscle weakness may be found more frequently in the real world of
clinical practice than in clinical trials, because actual patients encountered in clinical practice may not always reflect
the same population as research study participants.3 Perhaps
most importantly, although it is true that clinical trial data
provide no statistical consensus regarding the incidence of
myalgias beyond that of placebo, and although it is true that
many cases of myalgias in statin-treated patients are entirely
unrelated to statin use,12 it is also true that it is the consensus
of many clinicians (and many patients) that statins clearly
do have the potential to cause myalgias, even when muscle
enzymes are not elevated.3,4,12,24
Finally, the muscle complaints often described with
statins are not always of definitive origin. In the event that
profound elevations in CK levels occur after statin use (with
no other obvious causes), muscle biopsies have shown muscle anatomic abnormalities, including pathologic findings of
necrotizing myopathy and inflammation.59 In these cases, it
is reasonable to conclude that muscle pathology is the etiology of the muscle complaints. However, the etiology of
myalgias associated with normal CK levels is not as clear.
The profound myalgias and profound muscle weakness,
sometimes described by patients with normal muscle enzyme levels who are taking statins, are rarely accompanied
by physical examination findings of muscle damage (such
as fasciculations or muscle wasting). It therefore is not
known whether all cases of statin-induced myalgias accompanied by normal CK levels are muscle related, or rather,
whether there may be some other possible statin-mediated
mechanism (eg, a neurologic cause or another rheumaticologic60 etiology).
To address this concern, in a description of 4 patients who
were reportedly able to distinguish statin therapy from placebo
due to muscle complaints, and who had normal CK levels,
muscle biopsies demonstrated “mitochondrial dysfunction, including abnormally increased lipid stores, fibers that did not
stain for cytochrome oxidase activity, and ragged red fibers.”61
These findings reversed in the 3 patients who had repeat
biopsies when they were not receiving statins (1 patient elected
not to have a repeat biopsy).12,61 It should be kept in mind that
this was a very small number of study participants, and they
were not chosen in a randomized fashion. Instead, they were
selectively identified as being among the first 20 study subjects
who had been otherwise assigned to another ongoing clinical
trial. There was no control group, and the pathologists reading
the biopsy results were not blinded.61 Thus, it cannot be said
with certainty that these biopsy findings are representative, or
even common, in all statin-treated patients who experience
myalgias without elevations in CK levels. However, this report
does provide a hypothesis that patients with normal CK levels
who complain of myalgias or muscle weakness may indeed be
experiencing statin-induced muscle pathology. It is only
through larger and better designed trials that such a relation can
be more definitively determined.
Cerivastatin and muscle adverse experiences: As to
whether all statins are the same with regard to potential
muscle effects, the situation surrounding cerivastatin’s
withdrawal confirms that some statins at marketed doses
have shown a greater risk for muscle adverse experiences
when compared with other statins at their marketed doses.
To a large degree, these safety differences are owing to
differences in the safety windows, which are related to the
blood levels achieved with the individual statins. Thus, the
safety of statins is largely based on the market dose, pharmacology profile, potential for drug interactions, and the
patient population treated.
It is illustrative to note that the initial clinical development of cerivastatin involved dose-ranging studies of 0.025
mg, 0.05 mg, 0.1 mg, and 0.2 mg per day.62 After yet further
studies, cerivastatin was initially approved in 1997 at the
0.2- and 0.3-mg daily doses,63 and then later approved at the
0.4- and 0.8-mg daily doses.64 Of the reported cases of fatal
rhabdomyolysis resulting in cerivastatin’s withdrawal from
market in 2001, 12 deaths were associated with the 0.8-mg/
day dose, 6 deaths were associated with the 0.4-mg/day
dose, 1 death occurred in a patient taking an unknown
cerivastatin dose, and 12 deaths were associated with the
concomitant use of cerivastatin with gemfibrozil. Gemfibrozil has been shown to increase cerivastatin blood levels,
significantly39 and to a greater degree than with other statins.28 It therefore may be reasonable to conclude that if the
marketed dose of cerivastatin had been limited to no more
than the 0.2-mg/day dose, had not been used concomitantly
with drugs known to increase its blood levels (such as
gemfibrozil), and if cerivastatin had not been used in patients at high risk for potential toxicities,9 then many of the
cases of nonfatal and fatal rhabdomyolysis would have been
avoided. Admittedly, cerivastatin 0.2 mg/day reduced lowdensity lipoprotein (LDL) cholesterol levels by only 30.5%,
and lower cerivastatin doses affected LDL to an even lesser
degree.62 Therefore, cerivastatin, at these lower doses, may
not have been commercially viable. However, this example
does illustrate that the potential adverse experiences of
statins should best be viewed in relation to their lipidaltering efficacy.
Moreover, if the marketed dose of cerivastatin had been
limited to 0.2 mg, if it had never been used concomitantly
with agents that have potential drug interactions and never
been used in patients at high risk for cerivastatin toxicity,
and if, then, the subsequent reported cases of nonfatal and
fatal rhabdomyolysis were found to be less than reported
with other statins, this still would not mean that cerivastatin
was an inherently safer molecule. It would simply mean that
cerivastatin was safer at lower marketed doses and when
used under recommended conditions. This is not surprising
given that the potential muscle adverse experiences with
statins are a function of dose, rather than a function of the
Bays/Statin Safety: Overview of the Data
degree of LDL cholesterol reduction.13,65 It is precisely such
safety issues as potential muscle adverse experiences that
have limited higher doses of, and significantly determined
the current doses of, marketed statins.25,26 In other words, if
a higher statin dose is only marginally more efficacious, and
if this higher dose is associated with even just a mild further
increase in muscle adverse experiences, then from a risk–
benefit standpoint, a lower top dose might be a more appropriate dose to market.66 It is also the recognition that the
highest statin doses are most closely associated with potential adverse experiences3,13 that has largely prompted the
development of combination lipid-altering agents.
Combination Lipid-Altering Drug Therapy
Combination lipid-altering drug treatment is a well-established therapeutic approach for (1) patients with severe
hypercholesterolemia, (2) those with more complicated
lipid abnormalities, or (3) patients who need aggressive
lipid management.19,67– 69 Clinical trials have shown that the
combination of bile acid sequestrants (such as colesevelam)
with statins lowers cholesterol levels more than does the
same dose of statin monotherapy, without an increased risk
of muscle adverse experiences.30 Similarly, an increase in
the frequency of muscle adverse experiences with fish oils
high in omega-3 fatty acids (used to lower triglyceride
levels) is not known to be a potential adverse experience,
either alone or when fish oils are used in combination with
statins.2,70
Cerivastatin was withdrawn from the market owing to an
increased risk of rhabdomyolysis; in many cases the drug
was found to have been given in concurrence with the
fibrate gemfibrozil. Gemfibrozil appears to be somewhat
unique in that it significantly interferes with statin metabolism through glucuronidation, and this interference appears
to be greater with cerivastatin than with other statins.5,28,39
Even so, retrospective studies have suggested that the combination of fibrates (including gemfibrozil) with statins
(other than cerivastatin) is relatively safe if patients are
adequately counseled to report muscle aches, malaise, or
other potential drug interaction symptoms to their clinician.71
Thus, it could be argued that 1 of the most important
measures that could be taken to reduce the risk of potential
muscle adverse experiences when statins are used in combination with fibrates has already been taken—ie, the withdrawal of cerivastatin from the market. An additional measure to reduce risk is appropriate patient education
regarding the signs and symptoms of potential muscle adverse experiences.
Yet another measure involves the choice of fibrate. Although the use of gemfibrozil with statins (other than cerivastatin) may be relatively safe under specified conditions,
the potential for impaired metabolism of statins with gemfibrozil40 may be greater than with other fibrates, such as
11C
fenofibrate.72 In fact, pharmacokinetic studies have shown
that the combination of fenofibrate with statins is associated
with minimal differences in the concentrations of either
fenofibrate or statin.41 In contrast, the concurrent use of
certain statins with gemfibrozil has shown a 2- to 3-fold
increase in statin levels.73 Possibly as a result of these
metabolic differences, long-term case studies have failed to
show an increase in CK levels with a fenofibrate-statin
combination.42 Analyses of the US FDA Adverse Event
Reporting System (AERS) have suggested that the use of
fenofibrate with statins results in fewer reports of rhabdomyolysis per million prescriptions than does the use of
gemfibrozil with statins.40,43 Data such as these have led
some researchers to recommend that when statins and fibrates are used concomitantly, fibrates such as fenofibrate
should be administered in preference to gemfibrozil.2,44 Also, because the highest risk of potential drug
interactions tends to occur at the highest dose of statins and,
theoretically, the higher doses of fibrates, then it may be
best to limit the use of the higher doses of statins. If
fenofibrate is to be used, it may likewise be theoretically
better to use fenofibrate formulations that use the lowest
fenofibrate doses.45,46
The combined use of niacin with statins results in more
global improvement in lipid parameters than does the use of
statins (or niacin) alone,74,75 and thus may allow better
achievement of lipid treatment goals at lower statin doses.
This is because the combination of niacin and statins represents the concomitant use of different lipid-altering drugs
with different and complementary beneficial lipid-altering
effects. However, just as the efficacy reflects the beneficial
actions of both agents, the potential adverse experiences
also reflect the safety and tolerability profile of both agents.
For example, study participants administered a fixed-dose,
extended-release niacin-lovastatin preparation reported
more flushing and other niacin-related adverse experiences
than did participants administered statin alone.31 Strictly
from a muscle standpoint, however, clinical trials of a fixed
dose, extended-release niacin-statin combination (as well as
other trials of niacin and statins76) have not supported an
increased risk of adverse muscle experiences when compared with the use of statins alone.32,33
Similarly, statins and ezetimibe have complementary
actions toward lowering LDL cholesterol levels,47 and
this combination may have other potential CAD benefits (such as a reduction in remnant lipoproteins and
a reduced intestinal absorption of phytostanols and
phytosterols, which are thought to be potentially atherogenic).77 Isolated case reports have suggested that
ezetimibe may increase the risk of muscle adverse experiences used alone or in combination with statins,48 –50
and muscle adverse experiences have been described with
the ezetimibe-simvastatin agent, which is not unexpected
given that simvastatin (a statin) is a component of this
treatment. In contrast to case reports, clinical trials have
not shown an increase in muscle adverse experiences
12C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
with ezetimibe plus statins. In a study of 1,528 women
and men comparing placebo, ezetimibe, simvastatin, and
the ezetimibe-simvastatin combination formulation, none
of the ezetimibe-simvastatin combination subjects experienced myopathy or rhabdomyolysis.47 Admittedly,
studies such as this do not definitively disprove possible
rare ezetimibe-related adverse muscle experiences. However, clinical trials do support the contention that the
addition of ezetimibe to ongoing statin treatment and the
use of the combination ezetimibe-simvastatin agent as
first-line therapy are both therapeutic treatment options
that might better achieve LDL cholesterol treatment targets at lower statin doses. This may have therapeutic
implications for patients who have statin-related adverse
experiences (including muscle adverse experiences) that
are thought to be dose-related or who may be at risk for
such potential adverse experiences.
Currently Marketed Statin Monotherapy and Clinical
Trial Assessment of Muscle Adverse Experiences
With regard to currently marketed statins used as monotherapy at recommended doses, and administered under
recommended conditions, no conclusive comparative evidence exists that these statins differ with regard to potential
muscle adverse experiences. The exception to this, theoretically, might be in cases where potential drug interactions
may occur.6 For example, pravastatin78 and rosuvastatin79
are not significantly metabolized through the cytochrome
P450 enzyme (CYP) system. Therefore, pravastatin and
rosuvastatin blood levels may not increase to the same
degree as do other statins when administered concomitantly
with drugs known to be inhibitors of CYP enzymes responsible for metabolism of other statins. Otherwise, whereas
muscle adverse experiences occur with all statins24 (including rare cases of rhabdomyolysis80), clinical trials have
shown that statins are very safe with regard to potential
muscle adverse experiences.
In the Extended Clinical Evaluation of Lovastatin (EXCEL) trial involving 8,245 hypercholesterolemic patients,81,82 myopathy (as defined in Table 2) was reported in
only 1 patient (0.1%) receiving lovastatin 40 mg/day and 4
patients receiving lovastatin 80 mg/day (0.2%). In an analysis of 44 completed clinical trials involving 9,416 study
participants administered atorvastatin 10 – 80 mg, the incidence of myalgia was 1.9% (compared with 0.8% for placebo); only 1 case of asymptomatic CK elevation persistently ⬎10 times the ULN was found, and no case of
rhabdomyolysis was reported in the clinical trials analyzed.83 In the landmark CAD outcome study, commonly
known as Treating to New Targets (TNT),84 3 cases of
rhabdomyolysis were observed in the atorvastatin 10 mg
group (n ⫽ 5,006), and 2 cases of rhabdomyolysis were
observed in the atorvastatin 80 mg group (n ⫽ 4,995). None
of the cases of rhabdomyolysis were thought by the investigators to be causally related to atorvastatin.
During the same reporting period in which cerivastatin
was withdrawn because of 31 cases of reported fatal rhabdomyolysis, fluvastatin had no cases of reported fatal rhabdomyolysis.7 Furthermore, in an analysis of all fluvastatin
trials conducted by the manufacturer (Lescol and Lescol
XL; Novartis Pharmaceuticals, East Hanover, NJ) from
1987–2001 and involving 8,951 study participants, 5 cases
of CK ⬎10 times the ULN were found in placebo groups;
only 4 (0.2%), 13 (0.3%), and 0 (0.0%) cases were observed
with daily fluvastatin at 20-mg, 40-mg, and 80-mg doses
respectively. This suggested that although often less efficacious than other statins in lowering LDL cholesterols, fluvastatin has perhaps the least propensity to cause myotoxicity.85
Three of the major pravastatin trials were the West of
Scotland Coronary Prevention Study (WOSCOPS),86 the
Cholesterol and Recurrent Events (CARE)87 study, and the
Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID)88 study. These studies involved ⱕ5 years of
drug exposure, 19,592 study participants who were administered active treatment or placebo, and ⬎243,000 blood
samples. An analysis of these studies for potential adverse
muscle experiences revealed that the incidence of myalgias
was comparable between groups administered active drug
and placebo, with the withdrawal of blinded study drug due
to CK elevations found in only 3 subjects who received
pravastatin and 7 subjects who were given placebo. There
were no cases of myopathy (as defined in Table 2) and no
cases of confirmed rhabdomyolysis.89
In an evaluation of 4 main pivotal studies of 1,936 study
participants receiving the higher marketed doses of simvastatin, myopathy (as defined in Table 2) occurred in only
0.2% of the 40-mg and 0.6% of the 80-mg simvastatin
groups.90 There were no reported cases of rhabdomyolysis.
In the A to Z Trial, patients with acute coronary syndrome
were administered either simvastatin 40 mg/day for 1 month
and then titrated to 80 mg/day, or administered placebo for
4 months followed by simvastatin 20 mg/day. Follow-up
was from 6 –24 months. In this large trial of a very high
CAD risk population, 34 of 2,230 (1.5%) study participants
in the placebo followed by simvastatin group, and 41 of
2,263 (1.8%) in the simvastatin alone group discontinued
study drug owing to a muscle-related adverse event, which
did not represent a statistical difference. A total of 10
patients developed myopathy (as defined in Table 2). One
study participant was in the simvastatin 20 mg/day group
and 9 were in the simvastatin 80 mg/day group (p ⫽ 0.02).
In all, 3 of the 9 patients with myopathy met the definition
for rhabdomyolysis. Of these 3 study participants, 1 had
contrast-induced renal failure, 1 was receiving concomitant
verapamil, and 1 had a history of alcohol abuse.91 No cases
of rhabdomyolysis were observed with simvastatin 20 mg or
40 mg. These findings are consistent with the notion that
rhabdomyolysis can occur for reasons other than statin
Bays/Statin Safety: Overview of the Data
use,53,54 and that when muscle adverse experiences such as
rhabdomyolysis are thought to be statin-related (as is true
with most other statin adverse experiences for that matter92),
they most often occur at the higher statin doses, and/or
occur as the result of a drug interaction or in the presence of
comorbidities. Finally, the Heart Protection Study (HPS)93
was a multicenter, placebo-controlled, double-blind study of
20,536 study participants treated with simvastatin 40 mg
over 5 years. HPS involved men and women ⱕ80 years.
Entry criteria required that study participants had preexisting CAD, diabetes mellitus, stroke (or other cerebrovascular
disease), peripheral vascular disease, or hypertension in
men ⱖ65 years. Despite the massive nature of this longterm study involving high CAD risk patients (and thus
including a sicker population than many other studies), only
5 cases of rhabdomyolysis were reported in the simvastatin
40 mg group (n ⫽ 10,269) compared with 3 cases in the
placebo group (n ⫽ 10,267). It is relevant to note, however,
that as part of the exclusion criteria, study participants were
excluded from entering the prerandomization portion of the
study if they had muscle disease or “evidence of muscle
problems.” Furthermore, 36% of the 32,145 study participants who entered the prerandomization phase were excluded during the prerandomization “run-in” period (4
weeks of placebo, followed by 4 – 6 weeks of simvastatin 40
mg/day). Specifically, 11,609 potential study subjects who
originally entered the trial were not eligible for randomization or withdrew for a variety of other reasons. A full 3% of
the total excluded potential study subjects were withdrawn
specifically due to elevated liver enzymes, creatinine, or CK
in their pretreatment screening blood sample. Additionally,
2 potential study subjects were withdrawn before randomization due to myopathy. This is highly illustrative of how
statin clinical trials are often designed to specifically ensure
exclusion of study participants with intolerance to, or potential safety issues with, statins. Thus, although such trial
designs are well suited to evaluate statin efficacy in an
unbiased manner, they are often not well suited to objectively assess statin safety and tolerability.
Yet another analysis was done of 5 landmark statin CAD
outcomes trials87,88,94 –96 reported between 1966 –1998 (with
the selection of trials based on criteria suitable for a metaanalysis97). This analysis involved 30,817 study participants
administered the statins pravastatin, simvastatin, or lovastatin. In a meta-analysis of these studies, 50 subjects administered statin (0.16%) were reported to have had asymptomatic elevations of CK ⬎10 times the ULN compared
with 40 such subjects (0.13%) in placebo control groups.
Finally, largely because of statin safety concerns raised
after the withdrawal of cerivastatin, the most recently introduced statin, rosuvastatin, has undergone intense scrutiny
and regulatory review with regard to safety issues.98 To
some degree, it may be reasonable to conclude that the
increased postmarketing reports of muscle adverse experiences with rosuvastatin in its first year of marketing (compared with other statins in their first year of marketing)
13C
might be largely due to the heightened awareness or increased publicity of potential statin-related safety issues.99
However, although postmarketing data are helpful in alerting researchers, clinicians, regulatory agencies, and patients
to potential adverse experiences that are revealed only after
large numbers of patients have had the drug prescribed, this
type of data relies solely on reports and not actual event
rates and is limited by lack of confirmation of causality or
control of potential confounders.99 This is in contrast to
findings of controlled clinical trials. In an analysis of 27
phase 2/3 controlled clinical trials of rosuvastatin involving
the 5– 40 mg doses (conducted from 1996 –2000), administered to 12,400 study participants, the finding of CK ⬎10
times the ULN occurred in 0.4% of those administered 5
mg, 0.2% of those administered 10 mg, 0.2% of those
administered 20 mg, and 0.4% of those administered 40 mg.
In this dose range, myopathy was considered to possibly be
related to use of rosuvastatin in only 1 study subject administered rosuvastatin 20 mg. No cases of rhabdomyolysis
were reported. It has been suggested that this rate was
actually lower than found in clinical trials of other statins
during their development.13 In postmarketing studies, only 2
of 17,800 study participants (0.01%) administered 5– 40 mg
rosuvastatin experienced myopathy.13 As a result, the US
FDA issued a 2005 Information Sheet alerting healthcare
professionals that all statins have a low incidence of rhabdomyolysis, and that data from controlled trials and postmarketing safety information indicated that the risk of serious muscle damage with rosuvastatin was similar to that
with other statins.22
Thus, the data support that within the confines of the
clinical trial setting, and therefore presumably in the context
of most appropriate use, the occurrence of significant muscle adverse experiences with currently marketed statins at
available doses is rare and statins are very safe with regard
to potential muscle adverse experiences.4 It is in the less
controlled, nonresearch setting of clinical practice, involving the medical management of millions of patients and a
wide range of patient populations, that findings of muscle
adverse experiences are most frequently encountered with
statin treatment.
With regard to causality, speculation has been offered as
to the mechanism by which statins may cause muscle adverse experiences.4,5,10,24,29 The exact cause remains largely
unknown.18
Potential Liver Adverse Experiences
As with muscle, statins have a low risk for potential liver
adverse experiences. The levels of evidence for determining
whether such an adverse event is associated with statin use
are presented in Table 4.27,34,77,100 –115 Severe potential liver
adverse experiences have been described in isolated case
reports (such as rare reports of possible statin-related cholestatic hepatotoxicity, autoimmune hepatitis, fulminant
14C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 4
Level of evidence that potential liver adverse experiences are associated with statin use
Level of
Evidence
A
F
C
F
A
F
Potential Statin Adverse Experience
Mild, asymptomatic elevations in liver enzymes are a potential
adverse experience of statins
Liver dysfunction is an adverse consequence of the mild,
asymptomatic elevations in liver enzymes found in patients
treated with statins*
Significant or severe liver toxicity or damage is a potential
adverse experience of statins
The elevations in liver enzymes related to statin use are due
to, and proportional to, the reduction in LDL-C levels*
Asymptomatic elevations in liver enzymes are more common
at higher statin doses
Some statins are safer than others with regard to potential
liver adverse experiences†
Select
References
100
100
100–114
27,77,115
34,100,114
27,100
LDL-C ⫽ low-density lipoprotein cholesterol.
* An “F”-level of evidence means that there is no evidence or evidence to the contrary.
†
The mild-to-moderate elevations in liver enzymes sometimes observed with treatment have not
been shown to affect liver function adversely. Head-to-head, comparative controlled clinical trials of
existing statins at marketed doses have not consistently and statistically shown that any statin is more
or less likely to affect liver function adversely as compared with other statins.
hepatitis, and cirrhosis100 –113). Perhaps what might be more
clinically relevant than case reports is an evaluation of
23,000 patients who were treated with statins in a large
health maintenance organization (HMO), and who had alanine aminotransferase (ALT) level tests performed. A total
of 62 (0.3%) individuals were found to have ALT levels
⬎10 times the ULN; in 17 of these 62 patients, elevated
ALT levels were thought to be caused by statin treatment.
Of these 17 cases of marked elevations in ALT that were
thought to be statin related, 13 were associated with potential drug interactions. Of the 4 cases that did not appear to
be potentially associated with drug interactions, 3 were
cases of heart disease, diabetes, or both. The 1 remaining
case was of a 71-year-old woman treated with atorvastatin
80 mg.92 In 16 of the 17 patients with statin-associated
marked elevations in ALT, the transaminase levels resolved
upon statin discontinuation (with the remaining case being
an 80-year-old woman with heart failure who died shortly
after starting the statin). This evaluation of a large number
of patients in a “real-life” practice setting suggests that
marked elevations in liver enzymes are rare and are most
likely to occur when (1) potential drug interactions exist, (2)
comorbidities are present (including preexisting liver disease),11 or (3) the highest dose of statin is used.92 In the vast
majority of cases, if the marked elevations in ALT levels
occur and are thought to be statin related, these elevations
resolve upon discontinuation of statin therapy; moreover,
and as shown in clinical trials, the problem often resolves
even with continuation of statin therapy. The fact that higher
statin doses are most associated with liver adverse experiences is also supported by trials such as the Myocardial
Ischemia Reduction with Aggressive Cholesterol Lowering
(MIRACL) trial, which was a 16-week trial in which patients were administered atorvastatin 80 mg or placebo
within 24 –96 hours of admission for an acute coronary
artery syndrome.114 Three of the 38 study participants who
received high-dose atorvastatin (80 mg) were hospitalized
with elevated liver enzymes and a diagnosis of hepatitis.
In contrast to the very rare cases of marked elevations in
liver enzymes, it is the mild-to-moderate, asymptomatic
elevations in liver enzymes (which occur with all statins116)
that are most commonly seen in clinical practice. In fact, it
is the mild-to-moderate elevations in liver enzymes in statin-treated patients that represent 1 of the more common
reasons why clinicians refer patients to lipid experts, lipid
specialty clinics, gastroenterologists, and hepatologists.116
Management of the statin-treated patient with mild-tomoderate elevated liver enzymes presents a challenge, because this is often the population at highest risk for CAD.
Thus, in many respects, it is the very population that might
best benefit from statin therapy that has the highest risk of
liver enzyme elevations unrelated to statin use. Many patients with CAD, or at risk for CAD, may be overweight,
may have diabetes, or may be otherwise predisposed to the
development of hepatosteatosis (“fatty liver”), which is
commonly seen in clinical practice.116 –118
Additionally, patients treated with statin are often sicker,
older, and may be on multiple other concurrent medications
that commonly affect the liver119; they may be susceptible
to multiple causes of liver enzyme elevation, irrespective of
statin use.120 Thus, the elevations in liver enzymes found in
patients treated with statin may not always be caused by the
statin. It is the complexity of the etiology of elevations in
liver enzymes that may help to explain why postmarketing
studies have shown that elevations in liver enzymes found
in patients treated with statin spontaneously resolve in approximately 70% of cases, even when statin therapy is
continued.11
Bays/Statin Safety: Overview of the Data
It has been suggested that because the elevations in liver
enzymes occur more often at the higher statin doses, the
elevations in liver enzymes are a physiologic consequence
of lowering LDL cholesterol levels.121 However, if this
were true, then one might reasonably expect that the greatest
degree of increase in liver enzymes compared with placebo
would be found at the lower statin doses, where the greatest
degree of LDL cholesterol lowering (per milligram of statin) occurs. It would also be reasonable to conclude that
trials of statins with more LDL cholesterol lowering would
show greater liver enzyme elevations than statins with less
LDL cholesterol lowering. However, neither appears to be
true. In a meta-analysis of 13 trials involving 49,275 study
participants, low-to-moderate doses of pravastatin, lovastatin, and simvastatin were not associated with significant
increases in liver enzymes compared with placebo.115 In
head-to-head statin trials, elevations in liver enzymes have
not been shown to be more frequent with statins with greater
LDL cholesterol lowering efficacy versus statins with lesser
LDL cholesterol lowering efficacy. In 1 example, a headto-head, 6-week clinical trial of 2,431 patients with hypercholesterolemia who were administered rosuvastatin, atorvastatin, simvastatin, or pravastatin across marketed dose
ranges, found no significant differences in liver enzyme
elevations. Specifically, ALT elevation ⬎3 times the ULN
at 2 consecutive visits was found in only 5 study participants on atorvastatin 80 mg (n ⫽ 1), atorvastatin 20 mg (n
⫽ 2), simvastatin 40 mg (n ⫽ 1), and simvastatin 80 mg (n
⫽ 1), but not in any patient given pravastatin or rosuvastatin. There was no relation between the degree of ALT
elevations and the level of LDL cholesterol reduction. Specifically, in this comparator study, rosuvastatin 40 mg lowered LDL cholesterol more than any of the other comparator
statins at any dose, yet had no significant elevations in ALT,
and certainly no more than comparator statins despite improved LDL cholesterol lowering efficacy.27
The clinical trial evidence suggests little relation between the degree of LDL cholesterol lowering and the
degree of elevations in liver enzymes at lower statin doses.
In contrast, it is when the statin is doubled from the second
highest to the very highest marketed dose that elevated liver
enzymes are most frequent, despite the general finding that
this doubling results in only an additional 5%– 6% further
LDL cholesterol lowering.30 Thus, the evidence suggests
that the increase in transaminase levels with statins has
more to do with the individual statin at its highest marketed
doses, and more to do with increases in statin blood levels
(as also might occur through drug interactions) than with the
degree of cholesterol lowering77,34—a point that may also
apply to other statin dose-related adverse experiences, such
as potential muscle adverse experiences.
Overall, statin clinical trials have demonstrated that the
significant elevations in liver transaminase levels (defined
as ⱖ3 times the ULN on consecutive measurements) somewhere between 0.5%–5% of study participants,18,100 and
these liver enzyme elevations are statin dose related.18,121
15C
The clinical significance of these elevations is unclear,
given that these elevations often resolved with continued
statin treatment,18,121–124 and may have sometimes been due
to nonstatin etiologies. There is little to no evidence that
mild elevations in liver enzymes observed in statin-treated
patients adversely affect liver function. Nonetheless, it is
currently recommended that testing of liver transaminases
occur before and during statin therapy. The baseline measurement of liver enzyme testing may be most beneficial for
future diagnostic and comparative purposes,121 because the
pretreatment liver enzyme levels have not been shown to be
predictive of potential liver injury of acute hepatocellular
reactions.125 In other words, even when baseline elevations
in liver enzymes are present, this does not necessarily mean
that patients are at higher risk of hepatotoxicity as compared
with those who have normal baseline liver enzymes.126,127
Finally, not only has the frequency of significant liver
enzymes been found to be similar to placebo in many
clinical trials, but the clinical importance of statin-induced
elevated liver enzymes, when it occurs, has been questioned. Because of the unclear clinical relevance of the
elevations of mild-to-moderate liver enzymes in patients
treated with statins, and because liver blood testing is not
predictive of future statin-induced severe liver toxicity, it
has been suggested that lower-dose statins may be safe
enough for over-the-counter marketing.121 Whereas continued liver enzyme monitoring seems to be clinically warranted for patients on concomitant medications, with comorbid conditions, or otherwise at risk,92 the routine liver
enzyme testing in many other statin-treated patients may
need to be reconsidered, as the clinical trial evidence suggests that such testing and monitoring may not be clinically
necessary.116,128 –130
Potential Renal Adverse Experiences
Levels of evidence for determining whether potential renal
adverse experiences are associated with use of statins are
shown in Table 5.7,18, 21–23,131–134
Proteinuria and hematuria has been described as rare,
potential kidney effects associated with all statins (Table
6).21,135 Although these renal findings had previously been
recognized with statin therapy, this issue resurfaced during
the vast rosuvastatin development program,131 most likely
because (1) rosuvastatin was developed after the withdrawal
of cerivastatin, and the circumstance regarding the withdrawal of cerivastatin heightened the attention paid to all
statin safety issues, particularly with regard to muscle and
kidney; (2) the premarketing experience of rosuvastatin had
dwarfed that of any other statin approved to date, having
been studied in ⬎12,000 study participants13; (3) most of
the rosuvastatin development program had no upper age
limit and allowed a substantial number of study participants
with mild-to-moderate renal impairment; (4) the 80-mg
rosuvastatin dose (originally planned to be a marketed dose)
16C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 5
Level of evidence that potential kidney adverse experiences are associated with statin use
Level of
Evidence
B
B
B
F
B
B
U
Potential Statin Adverse Experience
Rhabdomyolysis (fatal and nonfatal) is a potential adverse
experience of statins and is more common at higher
statin doses*
Proteinuria may rarely occur with statin use
Hematuria may rarely occur with statin use
In the absence of rhabdomyolysis, statin administration of
existing statins at marketed doses results in progressive
kidney disease or progressive decrease in kidney
function
Severe kidney damage (renal failure) is a potential
adverse experience of statins†
Nonrhabdomyolysis kidney adverse experiences are more
common at higher statin doses
Some currently marketed statins are safer than others with
regard to potential adverse kidney experiences‡
Select
References
7,18,22,23
21,131
21,131
132–134
7,18,21–23
21
21
* Rhabdomyolysis can have numerous potential causes.53,54 Therefore, a report of rhabdomyolysis
in a patient on statin therapy does not necessarily mean that the statin caused the rhabdomyolysis.
†
Rhabdomyolysis, by definition, involves some degree of renal failure.
‡
All statins have the potential to result in rhabdomyolysis, with only minor differences of currently
marketed statins expressed per million prescriptions, as based on voluntary reports to regulatory
agencies. The determination as to whether the statin-induced mild proteinuria and hematuria has the
potential to lead to renal damage or renal dysfunction is unknown. Thus, it is unknown whether the
higher rate of proteinuria and hematuria found with higher marketed doses of some statins is less safe
than the lower rate of these urinary findings found with other statins.
was withdrawn from development because of the reports of
myotoxicity, including 7 cases of rhabdomyolysis, which
was a rate higher than observed over the entire range of the
approved rosuvastatin 5– 40 mg doses (1 rhabdomyolysis
case)21; and (5) isolated cases of renal failure were described, not always associated with myopathy.21,27,136
It should be understood that the basis behind the withdrawal of the development of the rosuvastatin 80-mg dose
was not unprecedented, given that it is below the point
where the risks of higher statin doses are thought to outweigh the potential benefits of further LDL cholesterol
lowering that has largely determined the eventual approved
doses of all statins. Thus the rationale of the withdrawal of
the rosuvastatin 80-mg dose was not substantially different
than the rationale behind the withdrawal of development of
higher doses of other marketed statins.25,26,66 Nonetheless,
given the backdrop of the cerivastatin withdrawal, as previously described, much attention was paid to the rosuvastatin development program, and thus the potential renal
effects of statins.
Preclinical animal studies have demonstrated renal tubular degeneration and other renal tubular toxicities in high
doses of all statins.131 In contrast, it has been suggested that
the mild proteinuria found in humans treated with statins
may not be a toxic effect, but rather a physiologic response.131 Given that albumin uptake in the proximal tubular cells is dependent on receptor-mediated endocytosis
(RME), and given that RME is dependent on megalin and
cubulin receptors, which in turn require the presence of
guanosine triphosphate (GTP)– binding proteins; given that
GTP-binding proteins require isoprenoid pyrophosphates,
and given that the generation of isoprenoid pyrophosphates
requires mevalonate, then a reduction in mevalonate by
statins has been suggested to impair RME and thus reduce
albumin uptake in the proximal renal tubules.131,137,138 This
potential mechanism suggests that the proteinuria found
with statins is not necessarily a toxic effect, but rather a
physiologic response to the HMG-CoA reductase inhibition
that is central to the cholesterol-lowering effect of statins.
This theory is supported by human kidney cell studies in
which the impairment of RME induced by statins (simvastatin, rosuvastatin, and pravastatin) was reversed by the
administration (or perhaps replenishment) of mevalonate.139
Other studies have also supported the contention that it is
mostly a physiologic effect (ie, it does not adversely affect
glomerular filtration) that results in the occasional finding of
mild proteinuria in statin-treated patients. Gel electrophoresis of urine of rosuvastatin-treated patients demonstrates a
“tubular pattern” of urinary protein excretion, which suggests a lack of tubular uptake of low molecular weight
proteins, possibly through the mechanism previously described.131,136
Hematuria that may be related to statin use is often
difficult to assess during the conduct of clinical trials.
From a practical standpoint, urinalyses are often routinely done for safety purposes only and not for the
purpose of evaluating any particular sign or symptom
from study participants. With such indiscriminant testing,
it is common to find hematuria on routine study visits
that, on repeat testing, returns to normal. Frequent, non–
Bays/Statin Safety: Overview of the Data
17C
Table 6
Frequency and associated statin doses that demonstrated the highest percent of urine abnormalities in rosuvastatin comparator clinical trials, as reported
near rosuvastatin approval date in 2003*
Urine Abnormality (statin dose/s)†
Statin
Dietary lead-in
Placebo
Pravastatin (20–40 mg)‡
Simvastatin (20–80 mg)
Atorvastatin (10–80 mg)
Rosuvastatin (5–40 mg)§
Patients,
N (dose)
5,811
372
191 (20 mg)
67 (40 mg)
517 (20 mg)
356 (40 mg)
337 (80 mg)
710 (10 mg)
667 (20 mg)
245 (40 mg)
377 (80 mg)
653 (5 mg)
1,202 (10 mg)
1,460 (20 mg)
2,384 (40 mg)
Urine Dipstick
Proteinuria ⱖ ⫹⫹
Urine Dipstick
Hematuria ⱖ ⫹
Proteinuria ⱖ ⫹⫹
and Hematuria ⱖ ⫹
1%
3%
1% (20 mg)
3%
5%
7% (20 mg)
0.1%
0.0%
0.5% (20 mg)
4% (20 mg)
8% (80 mg)
0.8% (40 mg)
2% (10 and 20 mg)
4% (10 mg)
0.6% (10 mg)
10% (40 mg)
1.3% (40 mg)
4% (40 mg)
* The finding of both proteinuria and hematuria is a common, incidental finding during the conduct of any clinical research trial, statin or otherwise. In
this table, the finding of proteinuria and hematuria in statin-treated study participants was not markedly higher, and in some cases, was lower than the same
urinary findings in study participants undergoing dietary lead-in, or who were administered placebo. The 2005 Prescribing Information for rosuvastatin states:
“In the rosuvastatin clinical trial program, dipstick-positive proteinuria and microscopic hematuria were observed among rosuvastatin-treated patients,
predominantly in patients dosed above the recommended dose range (ie, 80 mg). However, this finding was more frequent in patients taking rosuvastatin
40 mg, when compared to lower doses of rosuvastatin or comparator statins, though it was generally transient and was not associated with worsening renal
function. Although the clinical significance of this finding is unknown, a dose reduction should be considered for patients on rosuvastatin 40 mg therapy with
unexplained persistent proteinuria during routine urinalysis testing.”135 Although the percentages listed in this table do not suggest consistent differences in
the rates of proteinuria or hematuria based on the relative (higher or lower) doses of comparator statins, this table is consistent with the prescribing information
that states that the highest rates of proteinuria and hematuria with rosuvastatin were found at the highest (40 mg) dose.
†
Dose in parentheses indicates statin dose with the most frequent finding of urine abnormality.
‡
Pravastatin 80 mg was not studied, as this was not an approved pravastatin dose during the time of the rosuvastatin development program.
§
This table does not include data from the nonapproved rosuvastatin 80-mg dose, nor does it include data from the rosuvastatin open-label extension
analysis.
Adapted from FDA Briefing Document NDA 21-366 for the use of Crestor.21
study-related (and thus nonstatin-related) causes of hematuria include urologic or prostate abnormalities, menstrual or postmenopausal bleeding, urinary tract
infections, idiopathic (unexplained) hematuria, and exercise-induced hematuria. In the rare case where an association of hematuria with statins may be present, the
cause has yet to be explained.131
For clinicians, an important question is whether the rare
findings of urine abnormalities in clinical trials with statins
mean that statins are detrimental to the short- or long-term
renal function of patients. To address this, a prospective,
open-label study was conducted comparing the 1-year effects of atorvastatin therapy versus no treatment in 56 patients with renal disease who were already using angiotensin-converting enzyme (ACE) inhibitors or angiotensin-1
receptor blockers (ARBs). It was concluded that atorvastatin may reduce proteinuria and the progression of kidney
disease in patients with chronic kidney disease, proteinuria,
and hypercholesterolemia, and that atorvastatin may have a
beneficial additive effect to treatment with ACE inhibitors
and ARBs.132 It should be kept in mind that atorvastatin is
somewhat unique in its metabolism, in that it has the least
degree of renal excretion (⬍2%), followed by fluvastatin
(5%), rosuvastatin (10%), lovastatin (10%), simvastatin
(13%), and pravastatin (20%).140 This low rate of renal
excretion may help to explain why atorvastatin is generally
well tolerated in patients with impaired renal function. The
Deutsche Diabetes Dialyse Studie (4D Study) was a multicenter, randomized, double-blind, prospective study of
1,255 subjects with type 2 diabetes receiving maintenance
hemodialysis who were randomly assigned to receive atorvastatin 20 mg/day (n ⫽ 619) or matching placebo (n ⫽
636) for a mean duration of about 4 years.141,142 No cases of
rhabdomyolysis or severe liver disease were detected in
either group. The study medication was discontinued by the
investigators in 1 patient because of a report of myalgia in
combination with elevated CK levels, but this patient had
been given placebo. Otherwise, myalgia or myopathy was
found in 5 individuals (1%) given placebo and 7 patients
(1%) given atorvastatin. CK levels 3–5 times the ULN were
found in 3 (0.5%) individuals given placebo and 11 (2%)
patients who received atorvastatin; CK levels ⬎5–10 times
the ULN were found in 1 individual (0.2%) given placebo
and 1 patient (0.2%) given atorvastatin. An ALT level ⬎4
18C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 7
Level of evidence that potential nervous system adverse experiences are associated with statin use
Level of
Evidence
A
B
B
F
B
U
C
Potential Statin Adverse Experience
Statins reduce the risk of stroke
A decrease in cognition or memory is a potential adverse experience
of statins*
A decrease in cognition or memory is not a potential adverse
experience of statins*
Statins may worsen dementia or Alzheimer disease†
Statins may improve dementia or Alzheimer disease
Some statins are safer than others with regard to potential adverse
neurologic adverse experiences‡
Peripheral neuropathy is a potential adverse experience of statins
Select
References
144, 145
146–150
151–155
156
157–160
152,161–163
164–175
* Clinical trials have shown conflicting results, with some demonstrating possible mild decreases
in cognition, while other trials have demonstrated no such effect.
†
An “F”-level of evidence means that there is no evidence or evidence to the contrary.
‡
Although studies have suggested differing degrees of central nervous system statin exposure
based on the lipophilicity of different statins and other aspects of transport (see text), no large,
prospectively designed clinical trial has definitively demonstrated any increased or decreased risk of
neurologic adverse experiences. Older case reports have suggested that pravastatin may be less likely
to interfere with sleep.152,161 An increased risk of depression has also been described in older reports
with statin use, with both lipophilic161 and nonlipophilic statins (including pravastatin),162 which was
not confirmed.163 Yet other studies have not demonstrated any abnormalities in sleep with use of
lipophilic or hydrophilic statins.187,188
times the ULN was found in 1 individual (0.2%) given
placebo and 5 patients (1%) given atorvastatin.141
Thus the use of statins (with the notable, but very rare
exception of rhabdomyolysis), may be considered generally
safe with regard to the kidneys and kidney function. In fact,
in a meta-analysis of 13 clinical trials examining the effects
of lipid-altering drugs in general on renal function, the
conclusion was that lipid-altering therapies may actually
preserve glomerular filtration rate and decrease proteinuria
in patients with renal disease.133
Additionally, a retrospective assessment was done of the
rosuvastatin development program134 examining study participants administered rosuvastatin (N ⫽ 8,135), as well as
comparator atorvastatin (N ⫽ 3,793), simvastatin (N ⫽
2,417), pravastatin (N ⫽ 1,278), and placebo (N ⫽ 382)
treatments. After 8 weeks, there were no significant changes
in urine dipstick protein results comparing baseline with
final-visit values with any of the statin or placebo treatments. Combined proteinuria and hematuria occurred in
0%– 0.3% of study participants, with no significant differences between statins or between doses of statins. Creatinine levels were essentially unchanged as well. In a follow-up of 1,929 study participants with diabetes, as well as
8,722 study participants without diabetes, who entered into
a rosuvastatin extension study for up to 3.8 years, no progressive decline in renal function (as assessed by creatinine
blood levels) was observed in study participants receiving
long-term rosuvastatin treatment, irrespective of rosuvastatin dose and even when preexisting renal insufficiency was
present.
Perhaps the best conclusion the clinician can draw from
these data is summarized by the FDA in their 2005 “Public
Health Advisory on Crestor (rosuvastatin),” which states:
“Various forms of kidney failure have been reported in
patients taking Crestor (rosuvastatin), as well as with other
statins. Renal failure due to other factors is known to occur
at a higher rate in patients who are candidates for statin
therapy (eg, patients with diabetes, hypertension, atherosclerosis, heart failure). No consistent pattern of clinical
presentation or of renal injury (ie, pathology) is evident
among the cases of renal failure reported to date that clearly
indicates causation by Crestor (rosuvastatin) or other statins. Mild, transient proteinuria (or protein in the urine,
usually from the tubules), with and without microscopic
hematuria (minute amounts of blood in the urine), occurred
with Crestor (rosuvastatin), as it has with other statins, in
Crestor’s (rosuvastatin’s) pre-approval trials. The frequency
of occurrence of proteinuria appeared dose-related. In clinical trials with doses from 5 to 40 mg daily, this effect was
not associated with renal impairment or renal failure (ie,
damage to the kidneys). It is recommended, nevertheless,
that a dose reduction and an investigation into other potential causes be considered if a patient on Crestor (rosuvastatin) develops unexplained, persistent proteinuria.”143
Potential Neurologic Adverse Experiences
Levels of evidence supporting that potential neurologic adverse experiences may be related to statin use are listed in
Table 7.144 –175 Although much is known about the effects of
statins on muscle, liver, and even kidney, less is known
about potential statin-related neurologic adverse experi-
Bays/Statin Safety: Overview of the Data
ences. Neurologic conditions that have been described in
association with statin use include neuropsychiatric disorders (eg, decrease in cognition, and uncontrolled case reports of severe irritability176,177), and peripheral nervous
system disorders (eg, peripheral neuropathy).
Before addressing potential statin neurologic adverse experiences, it is important to note the beneficial effects of
statins on cerebral stroke, which is among the most common
and devastating illnesses encountered in the clinical practice
of medicine.178 For a number of reasons, the direct correlation of elevated cholesterol levels to the incidence of
stroke has not been consistent.2 However, what has been
generally consistent is the finding that statins reduce the rate
of stroke. A meta-analysis of 58 trials, corroborated by the
results from 9 cohort studies, showed that statins reduced
the relative risk of stroke by as much as 17%, depending on
the level of LDL cholesterol lowering.179 Similarly, the
combined data from 9 statin trials including 70,020 study
participants indicate that statins reduce the relative and
absolute risk reductions for stroke at 21% and 0.9%, respectively.144,145 A reduction in nonhemorrhagic stroke is therefore a clear beneficial effect of statins with regard to the
central nervous system (CNS), and thus is potentially beneficial in maintenance of mental status.180
CNS: One of the issues raised regarding the potential for
CNS effects has been the lipophilicity of statins.181 The
more lipophilic, the more potential there is for permeability
across the blood-brain barrier. Lovastatin and simvastatin
have been described as being the most lipophilic, followed
by atorvastatin, fluvastatin, rosuvastatin, and pravastatin.182,183 Animal studies have demonstrated that lipophilic
statins such as lovastatin and simvastatin cross the bloodbrain barrier, whereas less lipophilic statins such as pravastatin have minimal permeability.184 This may help to explain why lovastatin has been shown to be detectible in
cerebrospinal fluid, whereas pravastatin has not.185 However, determining the degree of brain tissue exposure to
statins involves more complicated considerations than assessment of lipophilicity alone. In addition to differences in
permeability and influx, statins may also undergo different
degrees of CNS efflux, based on the activity of transporters.186 So the actual exposure of brain tissue is a balance
between the blood-brain barrier diffusion into the CNS, and
the movement out of the CNS by transporters. From a
practical standpoint, it remains to be definitively proven that
the lipophilicity of statins has any effect on their clinical
efficacy or safety.
Case reports146,147 and clinical trials have suggested that
statins may impair cognitive function, which may be of
safety concern, particularly in older individuals.148 In a
double-blind study of 209 generally healthy hypercholesterolemic adults randomly assigned to 6-month treatment with
lovastatin 20 mg or placebo, assessments were made of
neuropsychologic performance, depression, hostility, and
quality of life. Lovastatin did not cause psychologic distress
19C
or substantially alter cognitive function, but it did result in
small performance decrements on neuropsychologic tests of
attention and psychomotor speed, which were concluded to
be of uncertain clinical importance.149 In a similar follow-up
study of 308 adults with hypercholesterolemia, a randomized, double-blind, placebo-controlled trial of simvastatin
10 mg or 40 mg for 6 months provided partial support for
minor decrements in cognitive functioning with statins.150
In contrast, however, an evaluation of simvastatin and
pravastatin’s effect on cognitive function in a double-blind,
placebo-controlled, crossover study of 36 patients, 4 weeks
of statin therapy resulted in no significant differences in any
cognitive measure compared with placebo.151 Likewise, in a
double-blind, placebo-controlled, randomized crossover
study of 25 healthy volunteers, 4 weeks of simvastatin or
pravastatin resulted in no increase in electroencephalogramevoked potentials, power spectral analysis, Hospital Anxiety Depression Scale (HADS), or Digit-Symbol Substitution Test (DSST). On the sleep measure of the Leeds Sleep
Questionnaire, study participants reported significantly
greater difficulty in getting to sleep while on simvastatin
compared with pravastatin, but neither score differed from
placebo.152 However, even this finding has been called into
question, because other studies have shown no abnormalities in sleep with use of lipophilic or hydrophilic
statins.187,188
Similarly, in the largest statin trial conducted specifically
in older study participants, the Pravastatin in Elderly Individuals at Risk of Vascular Disease (PROSPER) trial, 5,804
men and women aged 70 – 82 years with a history of, or risk
factors for, vascular disease were evaluated for mental
changes. After an average of 3.2 years, pravastatin 40 mg/
day was found to have no significant effect on cognitive
function or disability compared with placebo,156 as assessed
by diagnostic instruments such as the Mini-Mental State
Examination (MMSE), Stroop word learning, letter-digit
coding, activities of daily living.189 Similarly, the massive
HPS investigation, involving 20,536 study participants, did
not find that simvastatin decreased cognitive function.93,148
However, no initial cognitive assessment was conducted in
this study, and cognitive impairment was assessed by a
final-visit telephone interview for cognitive status
questionnaire.189
With specific regard to dementia (which may include
Alzheimer disease), an epidemiologic study of the potential
effect of statins and other lipid-lowering agents was done in
a nested case-control design study with information derived
from patients (284 with dementia and 1,080 controls) aged
ⱖ50 years from 368 medical practices. This revealed that
individuals who were prescribed statins actually had a substantially lowered risk of developing dementia, independent
of the presence or absence of untreated hyperlipidemia.157 A
4-year observational study of 1,037 older women without
dementia who were treated with statins, and assessed by the
Modified MMSE, revealed that when compared with nonusers, statin users had higher mean Modified MMSE scores
20C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
and a trend for a lower likelihood of cognitive impairment
that seemed to be independent of effects on lipid levels.153
A case-control, retrospective cohort study of a communitybased ambulatory primary care geriatric practice of 655
patients demonstrated that, compared with controls, statins
users had an improvement in their Mini-MMSE score and
scored higher on the Clock Drawing Test (CDT). The conclusion was that the use of statins is associated with a lower
prevalence of dementia and has a positive impact on the
progression of cognitive impairment.158
However, not all studies have confirmed that statins have
a favorable or beneficial effect on dementia. A cross-sectional study of the prevalence and incidence of Alzheimer
disease and a prospective study of the incidence of dementia
and Alzheimer disease among 5,092 residents of Cache
County, Utah, aged ⱖ65 years, revealed no association
between statin use and the subsequent onset of dementia or
Alzheimer disease.153 Similarly, a prospective, cohort study
of statin use and incident dementia and probable Alzheimer
disease involving 2,356 cognitively intact persons aged
ⱖ65 years, as selected from an HMO, found no significant
association between statin use and incident dementia or
probable Alzheimer disease.155
Statins have been suggested to specifically improve Alzheimer disease, which is a CNS condition. For example, a
1-year pilot, proof-of-concept, double-blind, placebo-controlled study of 67 randomized patients with mild-to-moderate Alzheimer disease demonstrated that atorvastatin
treatment may be of some clinical benefit as assessed
through validated diagnostic instruments (Alzheimer’s Disease Assessment Scale-cognitive subscale [ADAS-cog],
Clinical Global Impression of Change Scale, MMSE, Geriatric Depression Scale [GDS], Neuropsychiatric Inventory
[NPI], and the Alzheimer’s Disease Cooperative StudyActivities of Daily Living Inventory [ADCS-ADL]).159
Similarly, in a nested case-control study at the Veterans
Affairs Medical Center in Birmingham, Alabama, 309 patients newly diagnosed with Alzheimer disease were compared with 3,088 age-matched controls without Alzheimer
disease. This study revealed that statin users had a 39%
lower risk of Alzheimer disease relative to nonstatin users.160 Various mechanisms have been suggested as to why
statins may have a beneficial effect on Alzheimer disease,
such as a decrease in the output of ␤-amyloid (a peptide
toxic to neurons and thought to be the prime cause of the
neurodegeneration seen in Alzheimer disease),190 –192 as
well as other possible effects.193–201
In summary, there are theoretical concerns about statin
use and the potential for neurologic adverse experiences on
the CNS.202 This has been supported by case reports203 and
somewhat supported by some of the inconclusive clinical
trials described above. Conversely, other studies have not
substantiated these adverse CNS experiences, and there are
clinical data that suggest that statins may have beneficial
effects on CNS disorders such as Alzheimer disease and
dementia.204 Yet other studies have demonstrated no CNS
effects of statins, either favorable or unfavorable. Perhaps
the best that can be said at this point is that until definitive
clinical trials have been completed, and until the results are
known,159,201,205–208 the totality of the existing data does not
support the contention that statins worsen either Alzheimer
disease or dementia.
Peripheral nervous system: Case reports,164 –170 and a
small number of case-control171,172 and cohort studies,173
have suggested that statins may be associated with peripheral nervous system adverse experiences.174 However, from
a review of the literature, it is reasonable to conclude that
any potential risk of peripheral neuropathy with statin use is
very small.175
This is important because peripheral neuropathy is
among the more common neurologic disorders encountered
in clinical practice, and it has a number of potential etiologies209—regardless of treatment with statins. Some perspective may be helpful for the clinician in the evaluation of the
finding of peripheral neuropathy in a patient who is treated
with a statin drug. The electronic pharmacy claims of
915,066 patients in a healthcare system in Utah and its
neighboring states identified 272 patients who met the criteria for idiopathic polyneuropathy not due to known secondary causes of peripheral neuropathy, such as diabetes,
renal insufficiency, alcohol abuse, cancer, hypothyroidism,
acquired immunodeficiency syndrome, Lyme disease, or
heavy metal intoxication. Each of these patients was
matched with a control. The mean age of this population
was 47 ⫾ 13 years, and 57% were women. Statin drugs
were found to have been prescribed for about 9% of the 272
patients with idiopathic peripheral neuropathy and for about
7% for controls, for a mean duration of 10 months (thus
confirming that idiopathic peripheral neuropathy can be and
is diagnosed in patients both with and without concurrent
statin use). The average statin dose given to patients with
idiopathic peripheral neuropathy and controls was similar.
In this study, there was no significant association between
idiopathic peripheral neuropathy and statin use. The authors
concluded that their study was not consistent with a 4- to
14-fold increased risk of peripheral neuropathy in patients
treated with statins, as had been suggested in another cohort
study.172 Rather, they stated: “Our findings are supported by
the 20,536-patient [HPS], which did not report an excess of
idiopathic peripheral neuropathy during an average of 5
years of therapy with simvastatin 40 mg/day.”93,210
Admittedly, studies such as these cannot exclude the
potential rare case of a statin-associated peripheral neuropathy adverse experience. It may therefore be reasonable to
consider statins as a potential cause of peripheral neuropathy when other etiologies have been excluded.175 However,
when a rare case arises in which peripheral neuropathy is
suspected to be causally related to a statin, how might such
a patient best be evaluated and treated? A very small trial in
3 patients has suggested that serial sympathetic skin responses may be of value in the electrophysiologic assess-
Bays/Statin Safety: Overview of the Data
ment and follow-up of possible statin small-fiber neuropathy, which may reveal abnormalities within a month after
start and may result in resolution of symptoms and objective
neurologic testing shortly after discontinuing the statin. One
patient developed both small- and large-fiber neuropathy
upon rechallenge.211 Thus, a stepwise approach to the patient with a potential statin-related peripheral neuropathy
adverse experience may be to (1) ensure that other secondary causes have been evaluated; (2) perform a neurologic
physical examination and attempt to objectively quantify
abnormal neurologic physical findings; (3) obtain appropriate diagnostic neurologic studies; and (4) stop administering
the statin. If objective abnormalities are found on physical
examination and diagnostic neurologic testing, and if the
neuropathic symptoms resolve upon discontinuing the statin, it may then be useful to repeat the objective evaluations
to see whether the resolution of symptoms correlates with
the resolution of objective diagnostic neurologic testing. If
resolution of symptoms or objective neurologic testing does
not occur after withdrawal of statin therapy, then the diagnosis of idiopathic peripheral neuropathy unrelated to statin
use should be considered. Conversely, if symptoms and
objective neurologic testing resolve, then the clinician can
best decide whether the benefits of a rechallenge of statin
drug exceeds the potential risks.212
Finally, as noted above under “Potential Muscle Adverse
Experiences,” it is possible that many cases of myalgias
described by patients might represent neuropathic complaints. Therefore, when symptoms of isolated or generalized pain are reported by patients treated with statins, it is
important for the clinician to take a detailed history to
determine whether the complaints might be neurologic in
nature.
1. Black DM. A general assessment of the safety of HMG CoA reductase inhibitors (statins). Curr Atheroscler Rep 2002;4:34 – 41.
2. Bays H, Stein EA. Pharmacotherapy for dyslipidaemia— current
therapies and future agents. Expert Opin Pharmacother 2003;
4:1901–1938.
3. Ballantyne CM, Corsini A, Davidson MH, Holdaas H, Jacobson TA,
Leitersdorf E, Marz W, Reckless JP, Stein EA. Risk for myopathy
with statin therapy in high-risk patients. Arch Intern Med 2003;163:
553–564.
4. Thompson PD, Clarkson P, Karas RH. Statin-associated myopathy.
JAMA 2003;289:1681–1690.
5. Rosenson RS. Current overview of statin-induced myopathy. Am J
Med 2004; 116:408 – 416.
6. Ucar M, Mjorndal T, Dahlqvist R. HMG-CoA reductase inhibitors
and myotoxicity. Drug Saf 2000;22:441– 457.
7. Staffa JA, Chang J, Green L. Cerivastatin and reports of fatal rhabdomyolysis. N Engl J Med 2002;346:539 –540.
8. Psaty BM, Furberg CD, Ray WA, Weiss NS. Potential for conflict of
interest in the evaluation of suspected adverse drug reactions: use of
cerivastatin and risk of rhabdomyolysis. JAMA 2004;292:2622–2631.
9. Davidson MH. Controversy surrounding the safety of cerivastatin.
Expert Opin Drug Saf 2002;1:207–212.
10. Evans M, Rees A. The myotoxicity of statins. Curr Opin Lipidol
2002;13:415– 420.
11. Bottorff MB. Safety and statins: pharmacologic and clinical perspectives. Am J Manag Care 2004;4(suppl 2):S27–S37.
21C
12. Grundy SM. Can statins cause chronic low-grade myopathy [editorial]? Ann Intern Med 2002;137:617– 618.
13. Grundy SM. The issue of statin safety: where do we stand [editorial]?
Circulation 2005;111:3016 –3019.
14. Strom BL. Potential for conflict of interest in the evaluation of
suspected adverse drug reactions: a counterpoint. JAMA 2004;292:
2643–2646.
15. Piorkowski JD Jr. Bayer’s response to “potential for conflict of
interest in the evaluation of suspected adverse drug reactions: use of
cerivastatin and risk of rhabdomyolysis.” JAMA 2004;292:2655–
2657.
16. Andrade SE, Graham DJ, Staffa JA, Schech SD, Shatin D, La GL,
Goodman MJ, Platt R, Gurwitz JH, Chan KA. Health plan administrative databases can efficiently identify serious myopathy and rhabdomyolysis. J Clin Epidemiol 2005;58:171–174.
17. American Diabetes Association. Position Statement: Standards of
Medical Care in Diabetes [American Diabetes Association Clinical
Practice Recommendation]. Diabetes Care 2005;28(suppl 1):S4 –
S36.
18. Pasternak RC, Smith SC Jr, Bairey-Merz CN, Grundy SM, Cleeman
JI, Lenfant C. ACC/AHA/NHLBI clinical advisory on the use and
safety of statins. J Am Coll Cardiol 2002;40:567–572.
19. Bays HE. Combination lipid-altering drug therapy with statins: an
update [slide talk]. [Lipids Online Website.] Available at: http://
www.lipidsonline.org/slides/talk_cme_activity.cfm?tk⫽33&cme_
activityid⫽94&cmepage⫽cme_info. Accessed February 6, 2006.
20. Gabow PA, Kaehny WD, Kelleher SP. The spectrum of rhabdomyolysis. Medicine (Baltimore) 1982;61:141–152.
21. FDA Advisory Committee Meeting. CRESTOR (ZD4522, rosuvastatin calcium) TABLETS. Briefing Document NDA 21-366 for the
Use of CRESTOR. June 11, 2003. [US Food and Drug Administration
Website.] Available at: http://www.fda.gov/ohrms/dockets/ac/03/
briefing/3968B1_02_A-FDA-Clinical%20Review.pdf. Accessed
February 6, 2006.
22. US Food and Drug Administration. Alert for Healthcare Professional:
Crestor (Rosuvastatin Calcium). [US Food and Drug Administration
Website.] United States Department of Health and Human Services,
Center for Drug Evaluation and Research March 2005. Available at:
http://www.fda.gov/cder/drug/InfoSheets/HCP/RosuvastatinHCP.htm.
Accessed February 6, 2006.
23. Jamal SM, Eisenberg MJ, Christopoulos S. Rhabdomyolysis associated with hydroxymethylglutaryl-coenzyme A reductase inhibitors.
Am Heart J 2004;147:956 –965.
24. Tomlinson SS, Mangione KK. Potential adverse effects of statins on
muscle. Phys Ther 2005;85:459 – 465.
25. Nissen SE. High-dose statins in acute coronary syndromes: not just
lipid levels. JAMA 2004;292:1365–1367.
26. Mitchel YB. High-dose statins in acute coronary syndromes. JAMA
2005;293:37–39.
27. Jones PH, Davidson MH, Stein EA, Bays HE, McKenney JM, Miller
E, Cain VA, Blasetto JW. Comparison of the efficacy and safety of
rosuvastatin versus atorvastatin, simvastatin, and pravastatin across
doses (STELLAR Trial). Am J Cardiol 2003;92:152–160.
28. Prueksaritanont T, Zhao JJ, Ma B, Roadcap BA, Tang C, Qiu Y, Liu
L, Lin JH, Pearson PG, Baillie TA. Mechanistic studies on metabolic
interactions between gemfibrozil and statins. J Pharmacol Exp Ther
2002;301:1042–1051.
29. Evans M, Rees A. Effects of HMG-CoA reductase inhibitors on
skeletal muscle: are all statins the same? Drug Saf 2002;25:649 – 663.
30. Bays H, Dujovne C. Colesevelam HCl: a non-systemic lipid-altering
drug. Expert Opin Pharmacother 2003;4:779 –790.
31. Bays HE, Dujovne CA, McGovern ME, White TE, Kashyap ML,
Hutcheson AG, Crouse JR. Comparison of once-daily, niacin extended-release/lovastatin with standard doses of atorvastatin and simvastatin (the ADvicor Versus Other Cholesterol-Modulating Agents
Trial Evaluation [ADVOCATE]). Am J Cardiol 2003;91:667– 672.
22C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
32. Bays HE. Extended-release niacin / lovastatin: the first combination
product for dyslipidemia. Expert Rev Cardiovasc Ther 2004;2:485–
501.
33. Bays H. Trial finds that simvastatin plus niacin is safe in people with
coronary artery disease and low HDL cholesterol [commentary].
Evidence-Based Cardiovascular Medicine 2004;8:173–176.
34. Davidson MH. Combination therapy for dyslipidemia: safety and
regulatory considerations. Am J Cardiol 2002;90(suppl 10B):50K–
60K.
35. Reaven P, Witztum JL. Lovastatin, nicotinic acid, and rhabdomyolysis. Ann Intern Med 1988;109:597–598.
36. Guyton JR, Capuzzi DM. Treatment of hyperlipidemia with combined niacin-statin regimens. Am J Cardiol 1998;82(suppl 12A):
82U– 84U.
37. Rubenfire M. Safety and compliance with once-daily niacin extended-release/lovastatin as initial therapy in the Impact of Medical Subspecialty on Patient Compliance to Treatment (IMPACT) study. Am J
Cardiol 2004;94:306 –311.
38. Kashyap ML, McGovern ME, Berra K, Guyton JR, Kwiterovich PO,
Harper WL, Toth PD, Favrot LK, Kerzner B, Nash SD, Bays HE,
Simmons PD. Long-term safety and efficacy of a once-daily niacin/
lovastatin formulation for patients with dyslipidemia. Am J Cardiol
2002;89:672– 678.
39. Shitara Y, Hirano M, Sato H, Sugiyama Y. Gemfibrozil and its
glucuronide inhibit the organic anion transporting polypeptide
2 (OATP2/OATP1B1:SLC21A6)-mediated hepatic uptake and
CYP2C8-mediated metabolism of cerivastatin: analysis of the mechanism of the clinically relevant drug-drug interaction between cerivastatin and gemfibrozil. J Pharmacol Exp Ther 2004;311:228 –236.
40. Alsheikh-Ali AA, Kuvin JT, Karas RH. Risk of adverse events with
fibrates. Am J Cardiol 2004;94:935–938.
41. Martin PD, Dane AL, Schneck DW, Warwick MJ. An open-label,
randomized, three-way crossover trial of the effects of coadministration of rosuvastatin and fenofibrate on the pharmacokinetic properties
of rosuvastatin and fenofibric acid in healthy male volunteers. Clin
Ther 2003;25:459 – 471.
42. Ellen RL, McPherson R. Long-term efficacy and safety of fenofibrate
and a statin in the treatment of combined hyperlipidemia. Am J
Cardiol 1998;81(suppl 4A):60B– 65B.
43. Jones PH, Davidson MH. Reporting rate of rhabdomyolysis with
fenofibrate ⫹ statin versus gemfibrozil ⫹ any statin. Am J Cardiol
2005;95:120 –122.
44. Rosenson RS. Statins and the risk of rhabdomyolysis [editorial]. J
Endocrinol Invest 2002;25:577.
45. Davidson MH, Bays H, Rhyne J, Stein E, Rotenberg K, Doyle R.
Efficacy and safety profile of fenofibrate-coated microgranules 130
mg, with and without food, in patients with hypertriglyceridemia and
the metabolic syndrome: an 8-week, randomized, double-blind, placebo-controlled study. Clin Ther 2005;27:1–13.
46. Tricor [fenofibrate] prescribing information. North Chicago, IL:
Abbott Laboratories; 2004. Available at: http://rxabbott.com/pdf/
tricorpi.pdf. Accessed February 6, 2006.
47. Bays HE, Ose L, Fraser N, Tribble DL, Quinto K, Reyes R, JohnsonLevonas AO, Sapre A, Donahue SR. A multicenter, randomized,
double-blind, placebo-controlled, factorial design study to evaluate
the lipid-altering efficacy and safety profile of the ezetimibe/simvastatin tablet compared with ezetimibe and simvastatin monotherapy
in patients with primary hypercholesterolemia. Clin Ther 2004;26:
1758 –1773.
48. Phillips PS. Ezetimibe and statin-associated myopathy [letter]. Ann
Intern Med 2004; 19;141:649.
49. Fux R, Morike K, Gundel UF, Hartmann R, Gleiter CH. Ezetimibe
and statin-associated myopathy. Ann Intern Med 2004;20;140:671–
672.
50. Health Canada PaFB. Association of Ezetrol (ezetimibe) with myalgia, rhabdomyolysis, hepatitis, pancreatitis, and thrombocytopenia.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
Available at: http://hc-sc.gc.ca/dhp-mps/medeff/advisories-avis/prof/
2005/ezetrol_hpc-cps_e.html. Accessed February 6, 2006.
Vytorin [ezetimibe-simvastatin] prescribing information. North
Wales, PA: Merck/Schering-Plough; 2005. [Vytorin Web site.]
Available at: http://www.vytorin.com/vytorin/shared/documents/
vytorin_pi.pdf. Accessed February 6, 2006.
Gagne C, Bays HE, Weiss SR, Mata P, Quinto K, Melino M, Cho M,
Musliner TA, Gumbiner B. Efficacy and safety of ezetimibe added to
ongoing statin therapy for treatment of patients with primary hypercholesterolemia. Am J Cardiol 2002;90:1084 –1091.
Wilson R., Shah BR. Rhabdomyolysis. emedicine [serial online]
2004; May 26:1–16. Available at: http://www.emedicine.com/ped/
topic2003.htm.
Craig S. Rhabdomyolysis. emedicine [serial online] 2004;December
29:1–15. Available at: http://www.emedicine.com/emerg/topic508.htm.
Graham DJ, Staffa JA, Shatin D, Andrade SE, Schech SD, La GL,
Gurwitz JH, Chan KA, Goodman MJ, Platt R. Incidence of hospitalized rhabdomyolysis in patients treated with lipid-lowering drugs.
JAMA 2004;292:2585–2590.
Zetia [ezetimibe] prescribing information. North Wales, PA: Merck/
Schering-Plough; 2005. Available at: http://www.zetia.com/zetia/
shared/documents/zetia_pi.pdf. Accessed February 6, 2006.
Shanahan RL, Kerzee JA, Sandhoff BG, Carroll NM, Merenich JA.
Low myopathy rates associated with statins as monotherapy or combination therapy with interacting drugs in a group model health
maintenance organization. Pharmacotherapy 2005;25:345–351.
Daugird AJ, Crowell K, Saseen J. Clinical inquiries. Do statins cause
myopathy? J Fam Pract 2003;52:973–977.
McKelvie PA, Dennett X. Myopathy associated with HMG-CoA
reductase inhibitors (statins): a series of 10 patients and review of the
literature. Journal of Clinical Neuromuscular Disease 2002;3:143–
148.
Chazerain P, Hayem G, Hamza S, Best C, Ziza JM. Four cases of
tendinopathy in patients on statin therapy. Joint Bone Spine 2001;68:
430 – 433.
Phillips PS, Haas RH, Bannykh S, Hathaway S, Gray NL, Kimura BJ,
Vladutiu GD, England JD. Statin-associated myopathy with normal
creatine kinase levels. Ann Intern Med 2002;137:581–585.
Hunninghake DB. Clinical efficacy of cerivastatin: phase IIa doseranging and dose-scheduling studies. Am J Cardiol 1998;82(suppl
4B):26J–31J.
Cheng-Lai A. Cerivastatin. Heart Dis 2000;2:93–99.
Isaacsohn J, Insull WJr, Stein E, Kwiterovich P, Patrick MA, Brazg
R, Dujovne CA, Shan M, Shugrue-Crowley E, Ripa S, Tota R.
Long-term efficacy and safety of cerivastatin 0.8 mg in patients with
primary hypercholesterolemia. Clin Cardiol 2001;24(suppl):IV1–
IV9.
Evans E. Statin safety in perspective—maximising the risk:benefit.
British Journal of Cardiology 2004;11:449 – 454.
Rosenson RS, Bays HE. Results of two clinical trials on the safety
and efficacy of pravastatin 80 and 160 mg per day. Am J Cardiol
2003 1;91:878 – 881.
Xydakis AM, Ballantyne CM. Combination therapy for combined
dyslipidemia. Am J Cardiol 2002;90(suppl 10B):21K–29K.
Vega GL. Management of atherogenic dyslipidemia of the metabolic
syndrome: evolving rationale for combined drug therapy. Endocrinol
Metab Clin North Am 2004;33:525–544, vi.
Jones PH. Statins as the cornerstone of drug therapy for dyslipidemia:
monotherapy and combination therapy options. Am Heart J 2004;
148(suppl):S9 –S13.
Covington MB. Omega-3 fatty acids. Am Fam Physician 2004;70:
133–140.
Taher TH, Dzavik V, Reteff EM, Pearson GJ, Woloschuk BL, Francis GA. Tolerability of statin-fibrate and statin-niacin combination
therapy in dyslipidemic patients at high risk for cardiovascular
events. Am J Cardiol 2002;89:390 –394.
Bays/Statin Safety: Overview of the Data
72. VHA Pharmacy Benefits Management–Strategic Healthcare Group
and the Medical Advisory Panel. Statin-Fibrate Report: Focus on
Safety. Available at: http://www.pbm.va.gov/safety%20reports/
87ry38statin-fibrate-final.pdf. Accessed February 6, 2006.
73. Rosenson RS. Cholesterol lowering in diabetes: new evidence supports aggressive LDL-C targets. Postgrad Med 2005;117:17–20, 23–
27.
74. Bays HE, McGovern ME. Once-daily niacin extended release/lovastatin combination tablet has more favorable effects on lipoprotein
particle size and subclass distribution than atorvastatin and simvastatin. Prev Cardiol 2003;6:179 –188.
75. Bays HE, Dujovne CA, McGovern ME, White TE, Kashyap ML,
Hutcheson AG, Crouse JR. Comparison of once-daily, niacin extended-release/lovastatin with standard doses of atorvastatin and simvastatin (the ADvicor Versus Other Cholesterol-Modulating Agents
Trial Evaluation [ADVOCATE]). Am J Cardiol 2003;91:667– 672.
76. Zhao XQ, Morse JS, Dowdy AA, Heise N, DeAngelis D, Frohlich J,
Chait A, Albers JJ, Brown BG. Safety and tolerability of simvastatin
plus niacin in patients with coronary artery disease and low highdensity lipoprotein cholesterol (The HDL Atherosclerosis Treatment
Study). Am J Cardiol 2004;93:307–312.
77. Bays H. Ezetimibe. Expert Opin Investig Drugs 2002;11:1587–1604.
78. Jacobson TA. Comparative pharmacokinetic interaction profiles of
pravastatin, simvastatin, and atorvastatin when coadministered with
cytochrome P450 inhibitors. Am J Cardiol 2004;94:1140 –1146.
79. McKenney JM. Efficacy and safety of rosuvastatin in treatment of
dyslipidemia. Am J Health-Syst Pharm 2005;62:1033–1047.
80. Omar MA, Wilson JP. FDA adverse event reports on statin-associated rhabdomyolysis. Ann Pharmacother 2002;36:288 –295.
81. Dujovne CA, Chremos AN, Pool JL, Schnaper H, Bradford RH,
Shear CL, Higgins J, Downton M, Franklin FA, Nash DT. Expanded
clinical evaluation of lovastatin (EXCEL) study results. IV. Additional perspectives on the tolerability of lovastatin. Am J Med 1991;
91(suppl 1B):25S–30S.
82. Bradford RH, Shear CL, Chremos AN, Dujovne CA, Franklin FA,
Grillo RB, Higgins J, Langendorfer A, Nash DT, Pool JL. Expanded
Clinical Evaluation of Lovastatin (EXCEL) study results: two-year
efficacy and safety follow-up. Am J Cardiol 1994;74:667– 673.
83. Newman CB, Palmer G, Silbershatz H, Szarek M. Safety of atorvastatin derived from analysis of 44 completed trials in 9,416 patients.
Am J Cardiol 2003; 92:670 – 676.
84. LaRosa JC, Grundy SM, Waters DD, Shear C, Barter P, Fruchart JC,
Gotto AM, Greten H, Kastelein JJ, Shepherd J, Wenger NK. Intensive lipid lowering with atorvastatin in patients with stable coronary
disease. N Engl J Med 2005;352:1425–1435.
85. De AG. The influence of statin characteristics on their safety and
tolerability. Int J Clin Pract 2004;58:945–955.
86. Shepherd J, Cobbe SM, Ford I, Isles CG, Lorimer AR, MacFarlane
PW, McKillop JH, Packard CJ, for the West of Scotland Coronary
Prevention Study Group. Prevention of coronary heart disease with
pravastatin in men with hypercholesterolemia. N Engl J Med 1995;
333:1301–1307.
87. Sacks FM, Pfeffer MA, Moye LA, Rouleau JL, Rutherford JD, Cole
TG, Brown L, Warnica JW, Arnold JM, Wun CC, Davis BR, Braunwald E, for the Cholesterol and Recurrent Events Trial investigators.
The effect of pravastatin on coronary events after myocardial infarction in patients with average cholesterol levels. N Engl J Med 1996;
335:1001–1009.
88. The Long-Term Intervention with Pravastatin in Ischaemic Disease
(LIPID) Study Group. Prevention of cardiovascular events and death
with pravastatin in patients with coronary heart disease and a broad
range of initial cholesterol levels. N Engl J Med 1998;339:1349 –
1357.
89. Pfeffer MA, Keech A, Sacks FM, Cobbe SM, Tonkin A, Byington
RP, Davis BR, Friedman CP, Braunwald E. Safety and tolerability of
pravastatin in long-term clinical trials: prospective Pravastatin Pooling (PPP) Project. Circulation 2002; 105:2341–2346.
23C
90. Davidson MH, Stein EA, Hunninghake DB, Ose L, Dujovne CA,
Insull W Jr, Bertolami M, Weiss SR, Kastelein JJ, Scott RS, et al.
Lipid-altering efficacy and safety of simvastatin 80 mg/day: worldwide long-term experience in patients with hypercholesterolemia.
Nutr Metab Cardiovasc Dis 2000;10:253–262.
91. de Lemos JA, Blazing MA, Wiviott SD, Lewis EF, Fox KA, White
HD, Rouleau JL, Pedersen TR, Gardner LH, Mukherjee R, et al.
Early intensive vs a delayed conservative simvastatin strategy in
patients with acute coronary syndromes: phase Z of the A to Z trial.
JAMA 2004;292:1307–1316.
92. Charles EC, Olson KL, Sandhoff BG, McClure DL, Merenich JA.
Evaluation of cases of severe statin-related transaminitis within a
large health maintenance organization. Am J Med 2005;118:618 –
624.
93. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet 2002;360:7–22.
94. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study
(4S). Lancet 1994; 344:1383–1389.
95. Downs JR, Clearfield M, Weis S, Whitney E, Shapiro DR, Beere PA,
Langendorfer A, Stein EA, Kruyer W, Gotto AM Jr. Primary prevention of acute coronary events with lovastatin in men and women
with average cholesterol levels: results of AFCAPS/TexCAPS [Air
Force/Texas Coronary Atherosclerosis Prevention Study]. JAMA
1998;279:1615–1622.
96. The Long-Term Intervention with Pravastatin in Ischaemic Disease
(LIPID) Study Group. Prevention of cardiovascular events and death
with pravastatin in patients with coronary heart disease and a broad
range of initial cholesterol levels. N Engl J Med 1998;339:1349 –
1357.
97. LaRosa JC, He J, Vupputuri S. Effect of statins on risk of coronary
disease: a meta-analysis of randomized controlled trials. JAMA 1999;
282:2340 –2346.
98. US Food and Drug Administration. Response of the United States
Department of Health & Human Services to Public Citizen’s March
4, 2004 petition from Sidney Wolfe. March 11, 2005. [US Food and
Drug Administration Web site.] Available at: http://www.fda.gov/
cder/drug/infopage/rosuvastatin/crestor_CP.pdf. Accessed February
6, 2006.
99. Alsheikh-Ali AA, Ambrose MS, Kuvin JT, Karas RH. The safety of
rosuvastatin as used in common clinical practice: a postmarketing
analysis. Circulation 2005; 111:3051–3057.
100. Parra JL, Reddy KR. Hepatotoxicity of hypolipidemic drugs. Clin
Liver Dis 2003;7:415– 433.
101. Heuer T, Gerards H, Pauw M, Gabbert HE, Reis HE. Toxic liver
damage caused by HMG-CoA reductase inhibitor [in German]. Med
Klin (Munich) 2000;95:642– 644.
102. Pelli N, Setti M. Atorvastatin as a trigger of autoimmune hepatitis
[letter]. J Hepatol 2004;40:716.
103. Graziadei IW, Obermoser GE, Sepp NT, Erhart KH, Vogel W.
Drug-induced lupus-like syndrome associated with severe autoimmune hepatitis. Lupus 2003;12:409 – 412.
104. Wolters LM, Van Buuren HR. Rosuvastatin-associated hepatitis with
autoimmune features. Eur J Gastroenterol Hepatol 2005;17:589 –
590.
105. Pedersen TR, Tobert JA. Benefits and risks of HMG-CoA reductase
inhibitors in the prevention of coronary heart disease: a reappraisal.
Drug Saf 1996;14:11–24.
106. Hartleb M, Rymarczyk G, Januszewski K. Acute cholestatic hepatitis
associated with pravastatin. Am J Gastroenterol 1999;94:1388 –1390.
107. Punthakee Z, Scully LJ, Guindi MM, Ooi TC. Liver fibrosis attributed to lipid lowering medications: two cases. J Intern Med 2001;
250:249 –254.
108. Jimenez-Alonso J, Osorio JM, Gutierrez-Cabello F, Lopez de la OA,
Leon L, Mediavilla Garcia JD, for the Grupo Lupus Virgen de las
Nieves. Atorvastatin-induced cholestatic hepatitis in a young woman
24C
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
125.
126.
127.
128.
129.
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
with systemic lupus erythematosus. Arch Intern Med
1999;159:1811–1812.
Nakad A, Bataille L, Hamoir V, Sempoux C, Horsmans Y. Atorvastatin-induced acute hepatitis with absence of cross-toxicity with simvastatin. Lancet 1999;353:1763–1764.
Caldwell SH, Hespenheide EE, von Borstel RW. Myositis, microvesicular hepatitis, and progression to cirrhosis from troglitazone added
to simvastatin. Dig Dis Sci 2001;46:376 –378.
Bruguera M, Joya P, Rodes J. Hepatitis associated with treatment
with lovastatin. Presentation of 2 cases [in Spanish]. Gastroenterol
Hepatol 1998;21:127–128.
Huchzermeyer H, Munzenmaier R. Lovastatin-induced acute cholestatic hepatitis [in German]. Dtsch Med Wochenschr 1995;120:252–
256.
Grimbert S, Pessayre D, Degott C, Benhamou JP. Acute hepatitis
induced by HMG-CoA reductase inhibitor, lovastatin. Dig Dis Sci
1994;39:2032–2033.
Schwartz GG, Olsson AG, Ezekowitz MD, Ganz P, Oliver MF,
Waters D, Zeiher A, Chaitman BR, Leslie S, Stern T. Effects of
atorvastatin on early recurrent ischemic events in acute coronary
syndromes: the MIRACL study: a randomized controlled trial. JAMA
2001;285:1711–1718.
de DS, Spinler SA, Miller K, Peterson AM. Statins and liver toxicity:
a meta-analysis. Pharmacotherapy 2004;24:584 –591.
Chalasani N. Statins and hepatotoxicity: focus on patients with fatty
liver. Hepatology 2005;41:690 – 695.
Bays H, Abate N, Chandalia M. Adiposopathy: Sick fat causes high
blood sugar, high blood pressure, and dyslipidemia. Future Cardiology 2005;1:39 –59.
Bays H. Adiposopathy, metabolic syndrome, quantum physics, general relativity, chaos and the Theory of Everything. Expert Rev
Cardiovasc Ther 2005;3:393– 404.
Chitturi S, George J. Hepatotoxicity of commonly used drugs: nonsteroidal anti-inflammatory drugs, antihypertensives, antidiabetic
agents, anticonvulsants, lipid-lowering agents, psychotropic drugs.
Semin Liver Dis 2002;22:169 –183.
Mallory MA, Lee SW, Kowdley KV. Abnormal liver test results on
routine screening. How to evaluate, when to refer for a biopsy.
Postgrad Med 2004;115:53–56, 59 – 62, 66.
Gotto AM Jr. The case for over-the-counter statins. Am J Cardiol
2004;94:753–756.
Downs JR, Clearfield M, Tyroler HA, Whitney EJ, Kruyer W, Langendorfer A, Zagrebelsky V, Weis S, Shapiro DR, Beere PA, Gotto
AM. Air Force/Texas Coronary Atherosclerosis Prevention Study
(AFCAPS/TEXCAPS): additional perspectives on tolerability of
long-term treatment with lovastatin. Am J Cardiol 2001;87:1074 –
1079.
Cressman MD, Hoogwerf BJ, Moodie DS, Olin JW, Weinstein CE.
HMG-CoA reductase inhibitors: a new approach to the management
of hypercholesterolemia. Cleve Clin J Med 1988;55:93–100.
Hunninghake DB. Drug treatment of dyslipoproteinemia. Endocrinol
Metab Clin North Am 1990;19:345–360.
Tolman KG. The liver and lovastatin. Am J Cardiol 2002;89:1374 –
1380.
Vuppalanchi R, Teal E, Chalasani N. Patients with elevated baseline
liver enzymes do not have higher frequency of hepatotoxicity from
lovastatin than those with normal baseline liver enzymes. Am J Med
Sci 2005;329:62– 65.
Dujovne CA. Side effects of statins: hepatitis versus “transaminitis”myositis versus “CPKitis.” Am J Cardiol 2002;89:1411–1413.
Sniderman AD. Is there value in liver function test and creatine
phosphokinase monitoring with statin use? Am J Cardiol 2004;
94(suppl 9A):30F–34F.
Smith CC, Bernstein LI, Davis RB, Rind DM, Shmerling RH.
Screening for statin-related toxicity: the yield of transaminase and
creatine kinase measurements in a primary care setting. Arch Intern
Med 2003;163:688 – 692.
130. Gotto AM Jr. Safety and statin therapy: reconsidering the risks and
benefits. Arch Intern Med 2003;163:657– 659.
131. Agarwal R. Statin induced proteinuria: renal injury or renoprotection? J Am Soc Nephrol 2004;15:2502–2503.
132. Bianchi S, Bigazzi R, Caiazza A, Campese VM. A controlled, prospective study of the effects of atorvastatin on proteinuria and progression of kidney disease. Am J Kidney Dis 2003;41:565–570.
133. Fried LF, Orchard TJ, Kasiske BL. Effect of lipid reduction on the
progression of renal disease: a meta-analysis. Kidney Int 2001;59:
260 –269.
134. Vidt DG, Cressman MD, Harris S, Pears JS, Hutchinson HG. Rosuvastatin-induced arrest in progression of renal disease. Cardiology
2004;102:52– 60.
135. CRESTOR [rosuvastatin calcium] prescribing information. Wilmington, DE: AstraZeneca Pharmaceuticals LP; 2005. Available at: http://
www.astrazeneca-us.com/pi/crestor.pdf. Accessed August 2005.
136. US Department of Health and Human Services. US Food and Drug
Administration Center for Drug Evaluation and Research, Transcript
of meeting of the Endocrinologic and Metabolic Drugs Advisory
Committee. July 9, 2003. [FDA Web site.] Available at: http://
www.fda.gov/ohrms/dockets/ac/03/transcripts/3968T1.doc. Accessed February 6, 2006.
137. Christensen EI. Pathophysiology of protein and vitamin handling in
the proximal tubule. Nephrol Dial Transplan 2002;17(suppl 9):57–
58.
138. Sidaway JE, Davidson RG, McTaggart F, Orton TC, Scott RC, Smith
GJ, Brunskill NJ. Inhibitors of 3-hydroxy-3-methylglutaryl-CoA reductase reduce receptor-mediated endocytosis in opossum kidney
cells. J Am Soc Nephrol 2004;15:2258 –2265.
139. Verhulst A, D’Haese PC, De Broe ME. Inhibitors of HMG-CoA
reductase reduce receptor-mediated endocytosis in human kidney
proximal tubular cells. J Am Soc Nephrol 2004;15:2249 –2257.
140. McKenney JM. Defining the pharmacological profile of rosuvastatin:
a look at statin therapy for dyslipidemia [poster online]. France
Foundation Web Site Poster: Advances in Managing Hyperlipidemia
online continuing medical education activity. July 24, 2005. Available at: http://www.francefoundation.com/cme/posttest/pages/amh_
posters.pdf. Accessed February 6, 2006.
141. Wanner C, Krane V, Marz W, Olschewski M, Mann JF, Ruf G, Ritz
E. Atorvastatin in patients with type 2 diabetes mellitus undergoing
hemodialysis. N Engl J Med 2005;353:238 –248.
142. Bays HE. Collaborative Atorvastatin Diabetes Study (CARDS). Lipids Online Commentaries on Research 2005. Available at: http://
www.lipidsonline.org. Accessed February 6, 2006.
143. US Food and Drug Administration. FDA Public Health Advisory
on Crestor (rosuvastatin). Center for Drug Evaluation and Research.
[FDA Web site.] Available at: http://www.fda.gov/cder/drug/
advisory/crestor_3_2005.htm. Accessed February 6, 2006.
144. Amarenco P, Tonkin AM. Statins for stroke prevention: disappointment and hope. Circulation 2004;109(suppl 1):III44 –III49.
145. Amarenco P, Lavallee P, Touboul PJ. Stroke prevention, blood cholesterol, and statins. Lancet Neurol 2004;3:271–278.
146. King DS, Wilburn AJ, Wofford MR, Harrell TK, Lindley BJ, Jones
DW. Cognitive impairment associated with atorvastatin and simvastatin. Pharmacotherapy 2003;23:1663–1667.
147. Orsi A, Sherman O, Woldeselassie Z. Simvastatin-associated memory loss. Pharmacotherapy 2001;21:767–769.
148. Golomb BA. Implications of statin adverse effects in the elderly.
Expert Opin. Drug Saf 2005;4:389 –397.
149. Muldoon MF, Barger SD, Ryan CM, Flory JD, Lehoczky JP, Matthews KA, Manuck SB. Effects of lovastatin on cognitive function
and psychological well-being. Am J Med 2000;108:538 –546.
150. Muldoon MF, Ryan CM, Sereika SM, Flory JD, Manuck SB. Randomized trial of the effects of simvastatin on cognitive functioning in
hypercholesterolemic adults. Am J Med 2004;117:823– 829.
151. Cutler N, Sramek J, Veroff A, Block G, Stauffer L, Lines C. Effects
of treatment with simvastatin and pravastatin on cognitive function in
Bays/Statin Safety: Overview of the Data
152.
153.
154.
155.
156.
157.
158.
159.
160.
161.
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.
173.
patients with hypercholesterolaemia. Br J Clin Pharmacol
1995;39:333–336.
Harrison RW, Ashton CH. Do cholesterol-lowering agents affect
brain activity? A comparison of simvastatin, pravastatin, and placebo
in healthy volunteers. Br J Clin Pharmacol 1994;37:231–236.
Yaffe K, Barrett-Connor E, Lin F, Grady D. Serum lipoprotein levels,
statin use, and cognitive function in older women. Arch Neurol
2002;59:378 –384.
Zandi PP, Sparks DL, Khachaturian AS, Tschanz J, Norton M,
Steinberg M, Welsh-Bohmer KA, Breitner JC. Do statins reduce risk
of incident dementia and Alzheimer disease? The Cache County
Study. Arch Gen Psychiatry 2005; 62:217–224.
Li G, Higdon R, Kukull WA, Peskind E, Van Valen MK, Tsuang D,
van Belle G, McCormick W, Bowen JD, Teri L, Schellenberg GD,
Larson EB. Statin therapy and risk of dementia in the elderly: a
community-based prospective cohort study. Neurology 2004;
63:1624 –1628.
Shepherd J, Blauw GJ, Murphy MB, Bollen EL, Buckley BM, Cobbe
SM, Ford I, Gaw A, Hyland M, Jukema JW, et al. Pravastatin in
Elderly Individuals at Risk of Vascular Disease (PROSPER): a randomised controlled trial. Lancet 2002;360:1623–1630.
Jick H, Zornberg GL, Jick SS, Seshadri S, Drachman DA. Statins and
the risk of dementia. Lancet 2000;356:1627–1631.
Hajjar I, Schumpert J, Hirth V, Wieland D, Eleazer GP. The impact
of the use of statins on the prevalence of dementia and the progression of cognitive impairment. J Gerontol A Biol Sci Med Sci 2002;
57:M414 –M418.
Sparks DL, Sabbagh MN, Connor DJ, Lopez J, Launer LJ, Browne P,
Wasser D, Johnson-Traver S, Lochhead J, Ziolwolski C. Atorvastatin
for the treatment of mild to moderate Alzheimer disease: preliminary
results. Arch Neurol 2005;62:753–757.
Zamrini E, McGwin G, Roseman JM. Association between statin use
and Alzheimer’s disease. Neuroepidemiology 2004;23:94 –98.
Rosenson RS, Goranson NL. Lovastatin-associated sleep and mood
disturbances. Am J Med 1993;95:548 –549.
Lechleitner M, Hoppichler F, Konwalinka G, Patsch JR, Braunsteiner
H. Depressive symptoms in hypercholesterolaemic patients treated
with pravastatin [letter]. Lancet 1992;340:910.
Kassler-Taub K, Woodward T, Markowitz JS. Depressive symptoms
and pravastatin. Lancet 1993;341:371–372.
Jeppesen U, Gaist D, Smith T, Sindrup SH. Statins and peripheral
neuropathy. Eur J Clin Pharmacol 1999;54:835– 838.
Jacobs MB. Atorvastatin-induced polyneuropathy [letter]. Ann Intern
Med 2004;141:77.
Jacobs MB. HMG-CoA reductase inhibitor therapy and peripheral
neuropathy [letter]. Ann Intern Med 1994;120:970.
Ahmad S. Lovastatin and peripheral neuropathy [letter]. Am Heart J
1995;130:1321.
Phan T, McLeod JG, Pollard JD, Peiris O, Rohan A, Halpern JP.
Peripheral neuropathy associated with simvastatin. J Neurol Neurosurg Psychiatry 1995;58:625– 628.
Rajabally YA, Varakantam V, Abbott RJ. Disorder resembling Guillain-Barre syndrome on initiation of statin therapy. Muscle Nerve
2004;30:663– 666.
Scola RH, Trentin AP, Germiniani FM, Piovesan EJ, Werneck LC.
Simvastatin-induced mononeuropathy multiplex: case report. Arq
Neuropsiquiatr 2004;62:540 –542.
Corrao G, Zambon A, Bertu L, Botteri E, Leoni O, Contiero P. Lipid
lowering drugs prescription and the risk of peripheral neuropathy: an
exploratory case-control study using automated databases. J Epidemiol Community Health 2004;58:1047–1051.
Gaist D, Jeppesen U, Andersen M, Garcia Rodriguez LA, Hallas J,
Sindrup SH. Statins and risk of polyneuropathy: a case-control study.
Neurology 2002;58:1333–1337.
Gaist D, Garcia Rodriguez LA, Huerta C, Hallas J, Sindrup SH. Are
users of lipid-lowering drugs at increased risk of peripheral neuropathy? Eur J Clin Pharmacol 2001;56:931–933.
25C
174. Backes JM, Howard PA. Association of HMG-CoA reductase inhibitors with neuropathy. Ann Pharmacother 2003;37:274 –278.
175. Chong PH, Boskovich A, Stevkovic N, Bartt RE. Statin-associated
peripheral neuropathy: review of the literature. Pharmacotherapy
2004;24:1194 –1203.
176. Golomb BA, Kane T, Dimsdale JE. Severe irritability associated with
statin cholesterol-lowering drugs. QJM 2004;97:229 –235.
177. Waters D. Statins and safety: applying the results of randomized trials
to clinical practice. Am J Cardiol 2003;92:692– 695.
178. American Stroke Association. Impact of stroke. [American Heart
Association Web site.] Available at: http://www.strokeassociation.
org/presenter.jhtml?identifier⫽1033. Accessed February 6, 2006.
179. Law MR, Wald NJ, Rudnicka AR. Quantifying effect of statins on
low density lipoprotein cholesterol, ischaemic heart disease, and
stroke: systematic review and meta-analysis. BMJ 2003;326:1423–
1430.
180. Collins R, Armitage J, Parish S, Sleight P, Peto R. Effects of cholesterol-lowering with simvastatin on stroke and other major vascular
events in 20536 people with cerebrovascular disease or other highrisk conditions. Lancet 2004;363:757–767.
181. Sparks DL, Connor DJ, Browne PJ, Lopez JE, Sabbagh MN. HMGCoA reductase inhibitors (statins) in the treatment of Alzheimer’s
disease and why it would be ill-advise to use one that crosses the
blood-brain barrier. J Nutr Health Aging 2002;6:324 –331.
182. Chong PH, Seeger JD, Franklin C. Clinically relevant differences
between the statins: implications for therapeutic selection. Am J Med
2001;111:390 – 400.
183. Schachter M. Chemical, pharmacokinetic and pharmacodynamic
properties of statins: an update. Fundam Clin Pharmacol 2005;19:
117–125.
184. Saheki A, Terasaki T, Tamai I, Tsuji A. In vivo and in vitro bloodbrain barrier transport of 3-hydroxy-3-methylglutaryl coenzyme A
(HMG-CoA) reductase inhibitors. Pharm Res 1994;11:305–311.
185. Botti RE, Triscari J, Pan HY, Zayat J. Concentrations of pravastatin
and lovastatin in cerebrospinal fluid in healthy subjects. Clin Neuropharmacol 1991;14:256 –261.
186. Kikuchi R, Kusuhara H, Abe T, Endou H, Sugiyama Y. Involvement
of multiple transporters in the efflux of 3-hydroxy-3-methylglutarylCoA reductase inhibitors across the blood-brain barrier. J Pharmacol
Exp Ther 2004;311:1147–1153.
187. Eckernas SA, Roos BE, Kvidal P, Eriksson LO, Block GA, Neafus
RP, Haigh JR. The effects of simvastatin and pravastatin on objective
and subjective measures of nocturnal sleep: a comparison of two
structurally different HMG CoA reductase inhibitors in patients with
primary moderate hypercholesterolaemia. Br J Clin Pharmacol 1993;
35:284 –289.
188. Keech AC, Armitage JM, Wallendszus KR, Lawson A, Hauer AJ,
Parish SE, Collins R, for the Oxford Cholesterol Study Group. Absence of effects of prolonged simvastatin therapy on nocturnal sleep
in a large randomized placebo-controlled study. Br J Clin Pharmacol
1996;42:483– 490.
189. Sano M. Current concepts in the prevention of Alzheimer’s disease.
CNS Spectr 2003;8:846 – 853.
190. Sidera C, Parsons R, Austen B. The regulation of beta-secretase by
cholesterol and statins in Alzheimer’s disease. J Neurol Sci 2005;
229230:269 –273.
191. Buxbaum JD, Cullen EI, Friedhoff LT. Pharmacological concentrations of the HMG-CoA reductase inhibitor lovastatin decrease the
formation of the Alzheimer beta-amyloid peptide in vitro and in
patients. Front Biosci 2002;7:a50 –a59.
192. Simons M, Schwarzler F, Lutjohann D, von Bergmann K, Beyreuther
K, Dichgans J, Wormstall H, Hartmann T, Schulz JB. Treatment with
simvastatin in normocholesterolemic patients with Alzheimer’s disease: a 26-week randomized, placebo-controlled, double-blind trial.
Ann Neurol 2002;52:346 –350.
193. Cole SL, Grudzien A, Manhart IO, Kelly BL, Oakley H, Vassar R.
Statins cause intracellular accumulation of amyloid precursor protein,
26C
194.
195.
196.
197.
198.
199.
200.
201.
202.
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
beta-secretase-cleaved fragments, and amyloid beta-peptide via
an isoprenoid-dependent mechanism. J Biol Chem 2005;280:
18755–18770.
Wolozin B. Cholesterol, statins and dementia. Curr Opin Lipidol
2004;15:667– 672.
Crisby M, Carlson LA, Winblad B. Statins in the prevention and
treatment of Alzheimer disease. Alzheimer Dis Assoc Disord 2002;
16:131–136.
Kirsch C, Eckert GP, Koudinov AR, Muller WE. Brain cholesterol,
statins and Alzheimer’s Disease. Pharmacopsychiatry 2003;36(suppl
2):S113–S119.
Blauw GJ, Shepherd J, Murphy MB, for the PROSPER study group.
Dementia and statins: PROSPER study group [letter]. Lancet 2001;
881.
Kast RE. Dementia and statins [letter]. Lancet 2001;357:881.
Lowe G, Rumley A, Packard C, Shepherd J. Dementia and statins
[letter]. Lancet 2001;357:881.
Ledesma MD, Dotti CG. The conflicting role of brain cholesterol in
Alzheimer’s disease: lessons from the brain plasminogen system. In:
Lipids, Rafts and Traffic. Biochemistry Society Symposia, Vol. 72.
London: Portland Press; 2005:129 –138.
Hartman T. Cholesterol and Alzheimer’s disease: statins, cholesterol
depletion in APP processing and Abeta generation. Subcell Biochem
2005;38:365–380.
Algotsson A, Winblad B. Patients with Alzheimer’s disease may be
particularly susceptible to adverse effects of statins. Dement Geriatr
Cogn Disord 2004;17:109 –116.
203. Wagstaff LR, Mitton MW, Arvik BM, Doraiswamy PM. Statinassociated memory loss: analysis of 60 case reports and review of the
literature. Pharmacotherapy 2003;23:871– 880.
204. Etminan M, Gill S, Samii A. The role of lipid-lowering drugs in
cognitive function: a meta-analysis of observational studies. Pharmacotherapy 2003;23:726 –730.
205. Caballero J, Nahata M. Do statins slow down Alzheimer’s disease? A
review. J Clin Pharm Ther 2004;29:209 –313.
206. Rockwood K, Darvesh S. The risk of dementia in relation to statins
and other lipid lowering agents. Neurol Res 2003;25:601– 604.
207. Ebrahim S, Shlomo YB, Smith GD, Whincup P, Emberson J. Dementia and statins [letter]. Lancet 2001;357:882.
208. Golomb BA, Criqui MH, White HL, Dimsdale JE. The UCSD Statin
Study: a randomized controlled trial assessing the impact of statins on
selected noncardiac outcomes. Control Clin Trials 2004;25:178 –202.
209. Pascuzzi RM. Peripheral neuropathies in clinical practice. Med Clin
North Am 2003;87:697–724.
210. Anderson JL, Muhlestein JB, Bair TL, Morris S, Weaver AN, Lappe
DL, Renlund DG, Pearson RR, Jensen KR, Horne BD. Do statins
increase the risk of idiopathic polyneuropathy? Am J Cardiol 2005;
95:1097–1099.
211. Lo YL, Leoh TH, Loh LM, Tan CE. Statin therapy and small fibre
neuropathy: a serial electrophysiological study. J Neurol Sci 2003;
208:105–108.
212. Donaghy M. Assessing the risk of drug-induced neurologic disorders:
statins and neuropathy. Neurology 2002;58:1321–1322.
Statin Safety and Drug Interactions: Clinical Implications
Michael B. Bottorff, PharmD
The risks of muscle adverse events related to use of the 3-hydroxy-3-methylglutaryl
coenzyme A (HMG-CoA) reductase inhibitors, or statins, increase significantly with
the addition of interacting drugs to a patient’s therapy. The mechanism for most
statin drug interactions involves the cytochrome P-450 system, which provides an
indication of which drugs may interact. However, it is difficult to predict the probability of a drug interaction in a given patient because there are individual differences
in sensitivity to increased statin drug levels. Drug metabolism studies show simvastatin and lovastatin to be especially sensitive to the inhibiting effects of other drugs
on the cytochrome P-450 3A4 (CYP3A4) isoenzyme. Atorvastatin metabolism is less
affected by inhibitors of this isoenzyme. Case reports, postmarketing surveillance,
and clinical trial data demonstrate the clinical effect of CYP3A4 inhibitors on statins.
Also, through possible inhibition of statin biliary excretion and glucuronidation,
gemfibrozil given concomitantly with rosuvastatin, lovastatin, and simvastatin significantly increases the risk of myopathy and rhabdomyolysis, a potentially lifethreatening consequence of statin drug interactions. © 2006 Elsevier Inc. All rights
reserved. (Am J Cardiol 2006;97[suppl]:27C–31C)
Risks of adverse drug reactions (ADR), especially those
resulting from drug interactions (DI) are greater with higher
doses of 3-hydroxy-3-methylglutaryl coenzyme A (HMGCoA) reductase inhibitors, or statins, and with each additional medication a patient is prescribed. Statins have been
administered to thousands of study participants and prescribed to millions of patients, with outcomes demonstrating that they are well tolerated with good safety profiles.1– 4
However, withdrawal of cerivastatin from world markets as
a result of a high rate of fatal rhabdomyolysis and changing
clinical practice standards for cholesterol management have
refocused concerns on the safety of statins.5 The revised
National Cholesterol Education Panel Adult Treatment
Panel (NCEP ATP III) guidelines expanded the population
of patients targeted for cholesterol-lowering therapy to patients who are often prescribed multiple medications including elderly individuals (ⱖ65 years of age) and those with
metabolic syndrome.6 In addition, several recently published studies indicate target cholesterol levels should be
lower than those recommended by the NCEP ATP III,
which not only increases the number of individuals potentially prescribed a statin, but increases the likelihood that
statins will be used at higher doses.7–10
The effective doses for statins are rarely associated with
significant adverse events.11 However, only small incremental gains in efficacy occur with increased doses, compared
with the increased risk of adverse effects.12,13 With higher
Division of Clinical Pharmacy, College of Pharmacy, University of
Cincinnati, Cincinnati, Ohio, USA.
Address for reprints: Michael B. Bottorff, PharmD, Division of Clinical
Pharmacy, College of Pharmacy, University of Cincinnati, 3232 Eden
Avenue, Mail Location #4, Cincinnati, Ohio 45267-0004.
E-mail address: Michael.Bottorff@uc.edu.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.007
than approved doses, or with a DI, the risk of adverse events
can be significant. The mechanism for most statin DIs involves the cytochrome P-450 system, making it relatively
easy to predict which drugs may interact but difficult to
predict the probability of a DI in a given patient owing to
individual differences in sensitivity to increased statin drug
levels.11 Also, the magnitude of change in statin drug levels
shows as much as 10-fold variability between patients with
a specific DI.14 Because the risks of myopathy and rhabdomyolysis appear to be concentration-dependent, the potential for life-threatening consequences exists with statin DIs.
This review focuses on the clinical impact of statin DIs.
Drug Interaction Evidence
Controlled clinical trials designed with DI-related adverse
effects as a primary outcome are unethical. Therefore, identifying specific drugs with the potential to interact with
statins and defining the risks of the interactions relies on less
definitive types of evidence. These types of evidence include individual case reports, in vitro and in vivo drug
metabolism studies, postmarketing surveillance reports, and
reports from clinical efficacy trials. Each of these evidence
sources has significant limitations. Case reports do not help
in early identification of a DI, there are no comparative or
control data, and underreporting is a significant problem.
Drug metabolism studies are helpful in understanding DI
mechanisms. Unfortunately, the primary outcome of these
studies is the change in drug concentrations, which do not
correlate well with incidence and severity of DI-related
clinical events. Postmarketing surveillance reports can provide a large database of patients using statins concurrently
with other drugs, but underreporting from lack of documenwww.AJConline.org
28C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Figure 1. Sites of statin metabolism and elimination, which represent potential sites of drug interactions through inhibition. AT ⫽ atorvastatin; FV ⫽
fluvastatin; OATP ⫽ organic anion transporting polypeptide; P450 ⫽ CYP450 isoenzyme; PGP ⫽ P-glycoprotein; PV ⫽ pravastatin; RS ⫽ rosuvastatin; SV
⫽ simvastatin.
tation is a limitation with this type of data. Also, the number
of drugs that can be reported is sometimes limited by the
system, resulting in drugs that potentially interact with
statins being omitted from the report. Secondary outcomes
or post-hoc analyses from controlled clinical trials provide a
better incidence of a specific DI, but they are biased by
exclusion criteria that eliminate patients taking specific
drugs or classes of drugs.
Mechanisms of Drug Interactions
In general, DIs result from a change in the concentration of
either or both drugs in the body (pharmacokinetic interaction) or from a change in the relation between drug concentration and the response of the body to the drug (pharmacodynamic interaction).11 Pharmacokinetic interactions can
involve alterations of normal absorption, distribution, metabolism, or excretion of the substrate drug. Clinically significant DIs with statins are thought to result from altered
pharmacokinetics, primarily metabolism, as these drugs are
highly selective inhibitors of HMG-CoA reductase with no
known effects on other receptors, making pharmacodynamic interactions less likely.11 However, fibrates and niacin individually can cause myopathy rarely; when added to
statin therapy, an increased additive risk may be anticipated.15 Furthermore, the risk with gemfibrozil monotherapy appears to be higher than that seen with fenofibrate.4 In addition, owing to a recognized drug interaction,
the risk for myopathy with gemfibrozil and most statins
would be the highest, because there would be both a pharmacokinetic and pharmacodynamic interaction with this
particular combination. Recent reports do not as strongly
implicate niacin as a significant cause of statin-associated
myopathy or rhabdomyolysis.16
Approximately 80% of drugs require biotransformation
to hydrophilic metabolites for renal elimination, with about
50% of these drugs undergoing metabolism by the cytochrome P-450 3A4 (CYP3A4) isoenzyme, which is the
major liver microsomal (60%) and intestinal wall (70%)
isoenzyme of the cytochrome P-450 system.17
Concomitant administration of compounds metabolized
by this system can result in inhibition of enzyme activity by
the inhibiting drug or compound with an increased plasma
level of the substrate drug and an increased potential for
ADRs. In addition to inhibition of the liver P-450 system,
other sites of potential statin DIs include inhibiting metabolism by intestinal P-450 isoenzymes, preventing P-glycoprotein (PGP) transfer across the intestinal wall, blocking
organic anion transporting polypeptide (OATP)–mediated
hepatic uptake, and decreasing renal elimination of hydrophilic metabolites (Figure 1).18 –21 Each of the marketed
statins differ in their pharmacokinetic profile, which affects
the potential mechanisms and sites for DIs.11,22 In addition,
genetic variability results in individual differences in expression of specific cytochrome P-450 isoenzymes, which
can significantly alter drug disposition, affecting efficacy
and risks of ADRs and DIs.23,24
Drug-specific interactions with each statin are dependent
on the metabolic pathway of the statin. Lovastatin, simvastatin, and atorvastatin undergo metabolism by CYP3A4
isoenzyme creating the potential for DIs with a significant
number of widely used medications.11 Lovastatin and simvastatin are highly reliant on CYP3A4 for elimination,
Bottorff/Statin Safety and Drug Interactions: Clinical Implications
29C
Table 1
Rhabdomyolysis and statins: published report, 1985–2000
Statin
Monotherapy (n)
Combination
Therapy, n (%)
Total (n)
0
0
0
5
5
5
2 (100)
3 (100)
0
30 (86)
5 (50)
19 (79)
2
3
0
35
10
24
Atorvastatin
Cerivastatin
Fluvastatin
Lovastatin
Pravastatin
Simvastatin
Adapted from Ann Pharmacother.34
whereas only about 20% of atorvastatin is metabolized by
this isoenzyme.25–29 Pravastatin and rosuvastatin do not
undergo significant metabolism, however gemfibrozil and
cyclosporine can increase concentrations of these statins by
possibly blocking their biliary excretion.11,30 Gemfibrozil
has no effect on CYP3A4, but inhibits CYP2C9, which
contributes to ⬍10% of rosuvastatin metabolism. In vitro
drug metabolism studies of the less potent CYP3A4 inhibitors erythromycin and verapamil show that simvastatin
levels increase 4- or 5-fold; with the more potent inhibitors
(eg, example itraconazole), concentrations of simvastatin
increase 10- to 20-fold.25–27 Similar results have been
shown with lovastatin and CYP3A4 inhibitor.27 Gemfibrozil
also inhibits glucuronidation, which affects primarily the
acid form of statins.31 Concentrations of lovastatin and
simvastatin increase 3-fold and rosuvastatin increase 2-fold.
Grapefruit juice contains 6=,7=-dihydroxybergamottin a
furanocoumarin compound that may inhibit CYP3A4; statin-associated rhabdomyolysis as a result of grapefruit juice
consumption has been reported.32,33 The oral bioavailability
of drugs including statins that are metabolized by intestinal
wall CYP3A4 is thought to increase with grapefruit juice
ingestion due to a loss of this isoenzyme function within the
intestinal epithelium.
Fluvastatin is metabolized by the liver cytidylyltransferase 2C9 (CYP2C9) isoenzyme and has been shown to
significantly increase the concentration of diclofenac, indicating fluvastatin is a potent inhibitor of CYP2C9.11 Other
drugs metabolized by CYP2C9 with documented increased
effect when concomitantly administered with fluvastatin
include warfarin and phenytoin. Rosuvastatin undergoes
minimal metabolism by CYP2C9, since 90% of an oral dose
of this drug is eliminated as the parent compound in the
feces.
Clinical Impact of Statin Drug Interactions
Rarely, DIs can be beneficial through the increased efficacy
of a drug given concomitantly with a second drug. However, because the dose-response curve for statins is relatively flat across recommended doses, the clinical impact of
statin DIs usually is an increased risk of adverse events
rather than an increased therapeutic effect. In an analysis of
published reports of statin-related rhabdomyolysis cases
from 1985–2000, Omar and colleagues34 demonstrated the
significant negative impact of statin DIs (Table 1). It was
found that ⱖ50% of cases for any specific statin were
associated with a potential DI with either a fibrate or a
recognized CYP3A4 inhibitor. The rate of myopathy was
reported as 0.12% (n ⫽ 17) in a large cohort of patients in
a health maintenance organization whose formulary allowed
either simvastatin or lovastatin.35 The rate was 2-fold
greater (0.22%) in statin combination therapy with potential
inhibitors. In all, 14 of the 17 patients with statin-associated
myopathy were taking either gemfibrozil or a CYP3A4
inhibitor.
Selection of gemfibrozil over fenofibrate for combination
statin lipid-lowering therapy can result in a significant increase in risk, as gemfibrozil is a potent inhibitor of several
components of statin metabolism (conjugation and biliary
excretion), while fenofibrate does not appear to interact with
statins through these mechanisms.36 The increased risk from
a statin– gemfibrozil DI is shown in an evaluation by Jones
and Davidson.36 With fenofibrate and cerivastatin the number of rhabdomyolysis cases reported to the US Food and
Drug Administration (FDA) was 14, for an estimated 140
cases per 1 million prescriptions dispensed. Cerivastatin
combined with gemfibrozil accounted for 533 cases of rhabdomyolysis, for an estimated rate of 4,600 cases per
1 million prescriptions. A significant increase was also seen
with other statins combined with gemfibrozil (57 reports;
estimated 8.6 cases per 1 million prescriptions) compared
with statin–fenofibrate combinations (2 reports; estimated
0.58 cases per 1 million prescriptions). For adverse events
reported within the first year of marketing of each statin,
overall 60% were in the presence of interacting drugs.37 The
incidence of myopathy with simvastatin 80 mg and amiodarone was reported as 6% in 1 clinical trial.38 In an analysis
of 25,248 clinical trial participants receiving 20 mg– 80 mg
of simvastatin, the incidence of myopathy in those also
receiving verapamil was 0.63% (4 of 635) compared with
0.061% (13 of 21,224) of those who did not receive verapamil.39
Graham and colleagues4 estimated the rate of statinrelated hospitalized rhabdomyolysis cases from managed
30C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
care claims data of 252,460 patients treated with a lipidlowering agent. Some 24 patients required hospitalization
for rhabdomyolysis during treatment with an average
incidence per 10,000 person-years of 0.44 (95% confidence interval [CI], 0.20 – 0.84) for atorvastatin, pravastatin, or simvastatin monotherapy compared with an
incidence of 5.98 (95% CI, 0.72–216.0) for each of these
statins when used in combination with a fibrate. The
incidence for cerivastatin monotherapy was 5.34 (95%
CI, 1.46 to 13– 68) per 10,000 person-years, and the
incidence was 1,035 (95% CI, 389 –2,117) for cerivastatin–fibrate combinations, which is a risk of approximately
1 in 10 treated patients per year.
Conclusion
The risks of statin-related muscle adverse events increase
significantly with the addition of interacting drugs to a
patient’s therapy. Drug metabolism studies show simvastatin and lovastatin to be especially sensitive to the inhibiting
effects of other drugs on the CYP3A4 isoenzyme. Atorvastatin metabolism is less affected by inhibitors of this isoenzyme. Case reports, postmarketing surveillance, and clinical
trial data demonstrate that CYP3A4 inhibitors can have an
important negative effect on statin therapy. Also, through
inhibition of statin biliary excretion, gemfibrozil given concomitantly with all statins significantly increases the risk of
myopathy and rhabdomyolysis. Therefore, the consequences of statin DIs can be severe and life-threatening.
1. Pasternak RC, Smith SC, Bairey-Merz CN, Grundy SM, Cleeman JI,
Cleeman JI, Lenfant C, for the American College of Cardiology, the
American Heart Association, and the National Heart, Lung and Blood
Institute. ACC/AH/NHLBI Clinical Advisory on the use and safety of
statins. Stroke 2002;33:2337–2341.
2. Black DM. A general assessment of the safety of HMG CoA reductase
inhibitors (statins). Curr Atheroscler Rep 2002;4:34 – 41.
3. Lopez LM. Rosuvastatin: A high potency 3-hydroxy-3-methylglutaryl
coenzyme A reductase inhibitor. J Am Pharm Assoc 2005;45:503–13.
4. Graham DJ, Staffa JA, Shatin D, Andrade SE, Schech SD, La Grenade
L, Gurwitz JH, Chan KA, Goodman MJ, Platt R. Incidence of hospitalized rhabdomyolysis in patients treated with lipid-lowering drugs.
JAMA 2004;292:2585–2590.
5. Staffa JA, Chang J, Green L. Cerivastatin and reports of fatal rhabdomyolysis [letter]. N Engl J Med 2002;346:539 –540.
6. National Cholesterol Education Program (NCEP) Expert Panel on
Detection, Evaluation, and Treatment of High Blood Cholesterol in
Adults (Adult Treatment Panel III). Third report of the National
Cholesterol Education Program (NCEP) Expert Panel on Detection,
Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult
Treatment Panel III): final report. Circulation 2002;106:3143–3421.
7. Cannon CP, Braunwald E, McCabe CH, Rader DJ, Rouleau JL, Belder
R, Joyal SV, Hill KA, Pfeffer MA, Skene AM, for the Pravastatin or
Atorvastatin Evaluation and Infection Therapy-Thrombolysis in Myocardial Infarction 22 Investigators. Intensive versus moderate lipid
lowering with statins after acute coronary syndromes. N Engl J Med
2004;350:1495–1504.
8. Nissen SE, Tuzcu EM, Schoenhagen P, Brown BG, Ganz P, Vogel
RA, Crowe T, Howard G, Cooper CJ, Brodie B, Grines CL, DeMaria
AN, for the REVERSAL [Reversal of Atherosclerosis with Aggressive
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
Lipid Lowering] Investigators. Effect of intensive compared with
moderate lipid-lowering therapy on progression of coronary
atherosclerosis: a randomized controlled trial. JAMA 2004;291:1071–
1080.
Koren MJ, Hunninghake DB, and the ALLIANCE Investigators. Clinical outcomes in managed-care patients with coronary heart disease
treated aggressively in lipid-lowering disease management clinics: the
ALLIANCE [Aggressive Lipid-Lowering Initiation Abates New Cardiac Events] study. J Am Coll Cardiol 2004;44:1772–1779.
Downs JR, Clearfield M, Weis S, Whitney E, Shapiro DR, Beere PA,
Langendorfer A, Stein EA, Kruyer W, Gotto AM Jr. Primary prevention of acute coronary events with lovastatin in men and women with
average cholesterol levels: results of AFCAPS/TexCAPS [Air Force/
Texas Coronary Atherosclerosis Prevention Study]. JAMA 1998;279:
1615–1622.
Bellosta S, Paoletti R, Corsini A. Safety of statins: focus on clinical
pharmacokinetics and drug interactions. Circulation 2004:109(suppl
III):III50 –III57.
Jones P, Kafonek S, Laurora I, Hunninghake D, for the CURVES
Investigators. Comparative dose efficacy study of atorvastatin versus
simvastatin, pravastatin, lovastatin, and fluvastatin in patients with
hypercholesterolemia: the CURVES Study. Am J Cardiol 1998;81:
582–587.
Brewer HB Jr. Benefit-risk assessment of rosuvastatin 10 to 40 milligrams. Am J Cardiol 2003;92:23K–29K.
Olbricht C, Wanner C, Eisenhauer T, Kliem V, Doll R, Boddaert M,
O’Grady P, Krekler M, Mangold B, Christians U. Accumulation of
lovastatin, but not pravastatin, in the blood of cyclosporine-treated
kidney graft patients after multiple doses. Clin Pharmacol Ther 1997;
62:311–321.
Tobert JA. Efficacy and long-term adverse effect pattern of lovastatin.
Am J Cardiol 1988;62:28J–34J.
Omar MA, Wilson JP. FDA adverse event reports on statin-associated
rhabdomyolysis. Ann Pharmacother 2002;36:288 –295.
Wilkinson GR. Pharmacokinetics: the dynamics of drug absorption,
distribution, and elimination. In: Hardman JG, Limbird LE, Gilman
AG (eds). Goodman and Gilman’s The Pharmacological Basis of
Therapeutics. 10th Ed. New York: McGraw Hill; 2001:971–1002.
Wang E, Casciano CN, Clement RP, Johnson WW. HMG-CoA reductase inhibitors (statins) characterized as direct inhibitors of P-glycoprotein. Pharm Res 2001;18:800 – 806.
Obach RS, Zhang QY, Dunbar D, Kaminsky LS. Metabolic characterization of the major human small intestinal cytochrome P450s. Drug
Metab Dispos 2001;29:347–352.
Hsiang B, Zhu Y, Wang Z, Wu Y, Sasseville V, Yang WP, Kirchgessner TG. A novel human hepatic organic anion transporting
polypeptide (OATP2). Identification of a liver-specific human organic
anion transporting polypeptide and identification of rat and human
hydroxymethylglutaryl-CoA reductase inhibitor transporters. J Biol
Chem 1999;274:37161–37168.
Kyrklund C, Backman JT, Neuvonen M, Neuvonen PJ. Gemfibrozil
increases plasma pravastatin concentrations and reduces pravastatin
renal clearance. Clin Pharmacol Ther 2003;73:538 –544.
Lennernas H, Fager G. Pharmacodynamics and pharmacokinetics of
the HMG-CoA reductase inhibitors: similarities and differences. Clin
Pharmacokinet 1997;32:403– 425.
Mulder AB, van Lijf HJ, Bon MA, van den Bergh FA, Touw DJ, Neef
C, Vermes I. Association of polymorphism in the cytochrome
CYP2D6 and the efficacy and tolerability of simvastatin. Clin Pharmacol Ther 2001;70:546 –551.
Vermes A, Vermes I. Genetic polymorphisms in cytochrome P450
enzymes: effect on efficacy and tolerability of HMG-CoA reductase
inhibitors. Am J Cardiovasc Drugs 2004;4:247–255.
Neuvonen PJ, Kantola T, Kivisto KT. Simvastatin but not pravastatin
is very susceptible to interaction with the CYP3A4 inhibitor itraconazole. Clin Pharmacol Ther 1998;63:332–341.
Bottorff/Statin Safety and Drug Interactions: Clinical Implications
26. Kantola T, Kivisto KT, Neuvonen PJ. Erythromycin and verapamil
considerably increase serum simvastatin and simvastatin acid concentrations. Clin Pharmacol Ther 1998;64:177–182.
27. Neuvonen PJ, Jalava KM. Itraconazole drastically increases plasma
concentrations of lovastatin and lovastatin acid. Clin Pharmacol Ther
1996;60:54 – 61.
28. Azie NE, Brater DC, Becker PA, Jones DR, Hall SD. The interaction
of diltiazem with lovastatin and pravastatin. Clin Pharmacol Ther
1998;64:369 –377.
29. Kantola T, Kivisto KT, Neuvonen PJ. Effect of itraconazole on the
pharmacokinetics of atorvastatin. Clin Pharmacol Ther 1998;64:58 –
65.
30. Everett DW, Chando TJ, Didonato GC, Singhvi SM, Pan HY, Weinstein SH. Biotransformation of pravastatin sodium in humans. Drug
Metab Dispos 1991;19:740 –748.
31. Prueksaritanont T, Zhao JJ, Ma B, Roadcap BA, Tang C, Qiu Y, Liu
L, Lin JH, Pearson PG, Baillie TA. Mechanistic studies on metabolic
interactions between gemfibrozil and statins. J Pharmacol Exp Ther
2002;301:1042–1051.
32. Schmiedlin-Ren P, Edwards DJ, Fitzsimmons ME, He K, Lown KS,
Woster PM, Rahman A, Thummel KE, Fisher JM, Hollenberg PF,
33.
34.
35.
36.
37.
38.
39.
31C
Watkins PB. Mechanisms of enhanced oral availability of CYP3A4
substrates by grapefruit constituents: decreased enterocyte CYP3A4
concentration and mechanism-based inactivation by furanocoumarins.
Drug Metab Dispos 1997;25:1228 –1233.
Dreier JP, Endres M. Statin-associated rhabdomyolysis triggered by
grapefruit consumption [letter]. Neurology 2004;62:670.
Omar MA, Wilson JP, Cox TS. Rhabdomyolysis and HMG-CoA
reductase inhibitors. Ann Pharmacother 2001;35:1096 –1107.
Shanahan RL, Kerzee JA, Sandhoff BG, Carroll NM, Merenich JA.
Low myopathy rates associated with statins as monotherapy or combination therapy with interacting drugs in a group model health maintenance organization. Pharmacotherapy 2005;25:345–351.
Jones PH, Davidson MH. Reporting rate of rhabdomyolysis with
fenofibrate ⫹ statin versus gemfibrozil ⫹ any statin. Am J Cardiol
2005;95:120 –122.
Alsheikh-Ali AA, Ambrose MS, Kuvin JT, Karas RH. The safety of
rosuvastatin as used in common clinical practice. A post-marketing
analysis. Circulation 2005;111:3051–3057.
Alsheikh-Ali AA, Karas RH. Adverse events with concomitant amiodarone and statin therapy. Prev Cardiol 2005;8:95–97.
Zocor [package insert]. Whitehouse Station, NJ: Merck & Co.; 2004.
Statin Safety: An Appraisal from the Adverse Event
Reporting System
Michael H. Davidson, MD,a,* John A. Clark, MD, MSPH,b Lucas M. Glass, BA,b and
Anju Kanumalla, MSb
The adverse event (AE) profiles of 3-hydroxy-3-methylglutaryl coenzyme A (HMGCoA) reductase inhibitor (statin) agents are of great interest, in particular the most
recently approved statin, rosuvastatin. The forwarding of reports of AEs has been
shown to be influenced by several reporting biases, including secular trend, the new
drug reporting effect, product withdrawals, and publicity. Comparative assessments
that use AE reporting rates are difficult to interpret under these circumstances,
because such effects can themselves lead to marked increases in AE reporting.
Consequently, many comparative reporting rate analyses are best carried out in
conjunction with other metrics that put reporting burden into context, such as report
proportion. All-AE reporting rates showed a temporal profile that resembled those of
other statins when marketing cycle and secular trend were taken into account. A
before-and-after cerivastatin withdrawal comparison showed a substantial increase in
the reporting of AEs of interest for the statin class overall. Report proportion analyses
indicated that the burden of rosuvastatin-associated AEs was similar to that for other
statin agents. Analyses of monthly reporting rates showed that the reporting of
rosuvastatin-associated rhabdomyolysis and renal failure have increased following
AE-specific mass media publicity. Postrosuvastatin AE reporting patterns were comparable to those seen with other statins and did not resemble cerivastatin. © 2006
Elsevier Inc. All rights reserved. (Am J Cardiol 2006;97[suppl]:32C– 43C)
Pharmaceutical manufacturers design and conduct largescale clinical development programs to examine both the
effectiveness and the safety of new products. However,
quite often, the efficacy and safety profiles and, therefore
the benefit-risk profile of a drug, continue to be defined in
the months and years following approval. Numerous clinical trials have demonstrated that the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitor,
or statin, drug class can improve the lipid profile of
patients with dyslipidemia, bring large numbers of these
patients within treatment guidelines, and lower their risk
of cardiovascular-related morbidity and mortality.1
Moreover, statins are generally well tolerated and occurrences of serious adverse events (AEs) are generally
rare.2 However, in recent years, statin-associated myotoxicity, including skeletal muscle necrosis that may result in life-threatening rhabdomyolysis, has become a
subject of increased interest to regulators, healthcare provider (HCPs), and patients. The recent heightened awareness of rhabdomyolysis is related, in part, to the clinical
experience with cerivastatin (Baycol; Bayer Corp., West
Haven, CT), a statin that was associated with an in-
a
Rush University Medical Center, Chicago, Illinois, USA; and bGalt
Associates, Blue Bell, Pennsylvania, USA.
*Address for reprints: Michael H. Davidson, MD, 515 North State
Street, Suite 2700, Chicago, Illinois 60035.
E-mail address: michaeldavidson@radiantresearch.com.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.008
creased reporting rate of fatal rhabdomyolysis nearly 80
times higher than rates reported for the other statins
available at the time, specifically atorvastatin (Lipitor;
Pfizer Inc, New York, NY), fluvastatin (Lescol and Lescol XL; Novartis, East Hanover, NJ), lovastatin (Mevacor; Merck & Co., West Point, PA), pravastatin (Pravachol; Bristol-Myers Squibb, Princeton, NJ), and
simvastatin (Zocor; Merck & Co.).3,4 This risk of myopathy was markedly increased when cerivastatin was combined with gemfibrozil. Gemfibrozil inhibits the glucuronidation-mediated clearance of statins and is also an
inhibitor of the cytochrome P-450 2C8 (CYP2C8) isoenzyme, both of which are major pathways for cerivastatin
catabolism. Fenofibrate does interfere with statin glucuronidation of CYP2C8 clearance and has a much lower
rate of reported myopathy in combination with statins
than does gemfibrozil. The markedly increased reporting
rate of rhabdomyolysis and, in particular, fatal rhabdomyolysis associated with cerivastatin resulted in its manufacturer’s voluntary withdrawal of the drug from the US
market in August 2001 and an ensuing worldwide withdrawal.
Although the reported occurrence of muscle-related AEs
in clinical trials of statins is low,2 few clinical trials are of
sufficient size, duration, or design to detect rhabdomyolysis;
moreover, differing definitions of this AE are used. In the
5-year Heart Protection Study (HPS),5 which is the largest
clinical trial of statin therapy to date, 5 cases (0.05%) of
www.AJConline.org
Davidson et al/Statin Safety: Appraisal from the AERS
nonfatal rhabdomyolysis (defined as muscle symptoms plus
creatine kinase ⬎40 times the upper limit of normal [ULN])
were reported in patients receiving simvastatin 40 mg, compared with 3 cases (0.03%) in patients receiving placebo.
Moreover, despite extensive clinical experience with cerivastatin and the other statins, relatively little quantitative
information regarding statin-associated rhabdomyolysis has
been published.3,6 –9 The paucity of reported incidence data
makes the benefit–risk ratio of individual statins difficult to
assess by regulators and HCPs.
The US Food and Drug Administration (FDA) maintains the largest publicly available databases of AE reports in the world, and these databases are used for
postmarketing surveillance of drugs for product-associated AEs.10 Our objectives in this article are to quantify
and evaluate the results of an adverse event trend analysis
of statin-associated total and fatal rhabdomyolysis, and
other serious statin associated side effects such as hepatitis, peripheral neuropathy, and renal failure, using reports forwarded to the FDA’s AE databases, and to
present postmarketing surveillance data for rosuvastatin
(Crestor; AstraZeneca, Wilmington, DE), as the first statin introduced since the withdrawal of cerivastatin. Additionally, we discuss possible historical, temporal, and
secular factors that appeared to affect the reporting rates
of statin-associated rhabdomyolysis.
METHODS
Data Sources: The AE data used in these analyses were
obtained from the FDA’s MedWatch reporting program,
which is provided to the general public under the Freedom of
Information Act (FOI), which was amended in 1996. AE
reports may be submitted voluntarily by either an HCP or a
non-HCP reporter, including patients, family members, and
legal professionals. At the time of this analysis the FDA had
released AE report information up to and including December
31, 2004.
The FOI database consists of 2 parts: the legacy database, called the Spontaneous Reporting System (SRS),
and the current database, called the Adverse Event Reporting System (AERS). SRS contains data from 1965 to
October 1997, whereas AERS was implemented in November 1997 and contains data from that point forward.
In the SRS database, AEs were originally coded using the
Coding Symbols for Thesaurus of Adverse Reaction
Terms (COSTART) dictionary; in the AERS database,
AEs have always been coded using versions of the Medical Dictionary for Regulatory Activities (MedDRA)
dictionary (International Federation of Pharmaceutical
Manufacturers and Associations [IFPMA], Geneva,
Switzerland).
US prescriptions (new plus refill) were used as a
measure of exposure. Prescription data for all drugs were
obtained from IMS Health (Plymouth Meeting, PA).
33C
Definitions for Report Collections: AE reports were
included in this analysis according to the criteria that (1) the
report listed a drug from the following list as a suspect
medication: atorvastatin (Lipitor), ezetimibe (Zetia; Merck/
Schering-Plough Pharmaceuticals, North Wales, PA), fluvastatin (Lescol, Lescol XL), lovastatin (Mevacor, Altoprev
[Andrx Pharmaceuticals, Inc., Fort Lauderdale, FL]), pravastatin (Pravachol), rosuvastatin (Crestor), simvastatin (Zocor), or an ezetimibe plus simvastatin combination product
(Vytorin; Merck/Schering-Plough); and (2) the report did
not originate from a study, the medical literature, or a
reporting territory outside the United States. Definition of
report collections by drug exposure used searches for exact
spelling text strings for trade and generic names as well as
for a variety of close misspellings.
Definitions for Medical Events: AEs in the FOI database are coded using terminology (called preferred terms)
derived from the MedDRA dictionary. The following 7 AEs
of interest were examined in this study: rhabdomyolysis,
myopathy, myositis, renal failure, hepatitis/liver failure, peripheral neuropathy/polyneuropathy, and peripheral demyelinating neuropathy. Each of these AEs was defined using
either a single preferred term or a group of clinically related
preferred terms (Table 1). In addition to preferred term
inclusion criteria, the definition for renal failure also specified that any reports containing the preferred term “rhabdomyolysis” be excluded.
Calculation of US Spontaneous Reporting Rates and
Report Proportions: Reporting rates were calculated by
dividing the number of spontaneous case reports for all AEs
or for an AE of interest that fell into a particular period by
the number of prescriptions that were dispensed in the
United States for the same period. Cases were assigned to
time periods based on the date of data entry into the FOI
database for all analyses except the manufacturer’s HCP
expedited publicity analysis, which used the manufacturer’s
receipt date. Reporting rates were expressed as cases per
million prescriptions. Report proportions were calculated by
dividing the number of spontaneous case reports for all AEs
or for an event of interest that fell into a particular period by
the total number of spontaneous case reports for the same
period of time. US spontaneous expedited HCP and directto-FDA reporting rates were calculated similarly except that
the numerator counts were further restricted by reporting
route and reporter type.
Analysis of All-AE Reporting Over Time: Annual
all-AE reporting rates were calculated for each study product
beginning with the first month that prescriptions were first
written in the United States. This approach produced market
cycle–adjusted annual rates that were plotted beginning with
the first year of each product’s marketing cycle (rather than by
calendar year).
34C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 1
Definitions for conditions of interest
Condition of Interest
Rhabdomyolysis
Renal failure*
Myopathy
Myositis
Hepatitis/liver failure
Peripheral neuropathy/polyneuropathy
MedDRA Preferred Terms Used in Definition
Rhabdomyolysis
Renal failure
Renal failure acute
Renal failure acute or chronic
Renal failure aggravated
Renal failure chronic
Muscle necrosis
Myopathy
Myositis
Hepatitis
Hepatitis acute
Hepatitis aggravated
Hepatitis granulomatous
Hepatitis granulomatous NOS
Hepatitis neonatal
Hepatitis NOS
Hepatitis toxic
Hepatitis cholestatic
Cytolytic hepatitis
Hepatitis acute toxic
Neuritis
Neuritis NOS
Neuropathy
Neuropathy NOS
Neuritis motor
Neuropathy peripheral
Peripheral motor neuropathy
Peripheral nerve palsy
Peripheral neuropathy aggravated
Peripheral demyelinating neuropathy
Peripheral neuropathy NEC
Peripheral neuropathy NOS
Chronic inflammatory demyelinating
polyradiculoneuropathy
Demyelinating polyneuropathy
Demyelinating polyneuropathy NEC
Renal failure chronic aggravated
Renal failure neonatal
Renal failure NOS
Renal tubular necrosis
Renal insufficiency
Myopathy aggravated
Myopathy toxic
Myositis-like syndrome
Hepatitis fulminant
Chronic hepatitis
Hepatitis chronic active
Hepatitis chronic active aggravated
Hepatitis chronic NOS
Hepatitis chronic persistent
Autoimmune hepatitis
Hepatic encephalopathy
Hepatic failure
Hepatic necrosis
Hepatitis fulminant
Peripheral neuropathy NOS
Peripheral sensorimotor neuropathy
Peripheral sensory neuropathy
Polyneuropathy
Polyneuropathy NOS
Polyneuropathy idiopathic
progressive
Polyneuropathy toxic
Polyneuropathy toxic NEC
Autonomic neuropathy
Autonomic neuropathy NOS
Demyelinating polyneuropathy
NOS
Guillain-Barré syndrome
Guillain-Barré syndrome
MedDRA ⫽ Medical Dictionary for Regulatory Activities (International Federation of Pharmaceutical Manufacturers and Associations [IFPMA], Geneva,
Switzerland); NEC ⫽ not elsewhere classified; NOS ⫽ not otherwise specified.
*Application of the definition for renal failure excluded any report with the term rhabdomyolysis.
Analysis of the Effect of Cerivastatin Withdrawal
from the US Market: The effect on statin reporting of the
withdrawal of cerivastatin from the US market was examined by comparing 3-year US spontaneous reporting
rates and report proportions before and after the withdrawal announcement (August 8, 2001). Reporting rates
and report proportions were calculated for individual
statin agents and for all statins as a group for 2 intervals
of 3 years each, the 3 years before the year of cerivastatin
withdrawal (1998 –2000), and the 3 years after the year of
cerivastatin withdrawal (2002–2004).
Analysis of the Effect of Intermittent Publicity: The
effect of intermittent publicity on the AE reporting of
rosuvastatin versus atorvastatin was evaluated by selecting 5 publicity events as points of reference in the 4-year
interval 2001–2004 and comparing them with direct-toFDA and expedited HCP reporting rate trend lines for
possible effect. Direct-to-FDA and expedited reports
were used because both have relatively short processing
lag times that make these trend lines suitable for monthly
analyses. Because reports from HCP reporters are considered to be of higher quality than those sent by nonHCP reporters, only the HCP fraction of expedited reports was used.
Before initiation of this analysis, a total of 5 publicity
events were chosen based on their expected effect on
spontaneous AE reporting rates in the United States.
They were as follows: (1) withdrawal of cerivastatin from
the US market (August 8, 2001); (2) the launch of rosuvastatin in the United States (August 13, 2003); (3) the
First Public Citizen petition for removal of rosuvastatin
from the US market11 (March 4, 2004); (4) publication of
an editorial by Sidney M. Wolfe in The Lancet12 (June
26, 2004); and (5) testimony before the US Congress by
Davidson et al/Statin Safety: Appraisal from the AERS
35C
Table 2
Reporting characteristics
N (%)
All non-US reports
US nonspontaneous reports
US spontaneous reports
Direct-to-FDA reports
HCP
Non-HCP
Manufacturer’s expedited
reports
HCP
Non-HCP
Manufacturer’s periodic
reports
HCP
Non-HCP
Atorvastatin
(N ⫽ 17,396)
Cerivastatin
(N ⫽ 3,073)
Fluvastatin
(N ⫽ 2,018)
Lovastatin
(N ⫽ 13,770)
Pravastatin
(N ⫽ 6,849)
Rosuvastatin
(N ⫽ 2,944)
Simvastatin
(N ⫽ 12,336)
5,701 (32.8)
136 (0.8)
11,559 (66.4)
1,246 (7.2)
101 (0.6)
1,145 (6.6)
4,736 (27.2)
591 (19.2)
45 (1.5)
2,437 (79.3)
711 (23.1)
55 (1.8)
656 (21.3)
1,063 (34.6)
498 (24.7)
24 (1.2)
1,496 (74.1)
214 (10.6)
107 (5.3)
107 (5.3)
486 (24.1)
347 (2.5)
108 (0.8)
13,315 (96.7)
798 (5.8)
495 (3.6)
303 (2.2)
185 (1.3)
1,317 (19.2)
65 (0.9)
5,467 (79.8)
431 (6.3)
191 (2.8)
240 (3.5)
1,171 (17.1)
346 (11.8)
7 (0.2)
2,591 (88.0)
191 (6.5)
1 (0.0)
190 (6.5)
263 (8.9)
2,105 (17.1)
119 (1.0)
10,112 (82.0)
1,481 (12.0)
228 (1.8)
1,253 (10.2)
2,671 (21.7)
2,557 (14.7)
2,179 (12.5)
5,489 (31.6)
920 (29.9)
143 (4.7)
663 (21.6)
199 (9.9)
287 (14.2)
763 (37.8)
102 (0.7)
83 (0.6)
11,118 (80.7)
516 (7.5)
655 (9.6)
3,578 (52.2)
218 (7.4)
45 (1.5)
2,137 (72.6)
1,058 (8.6)
1,613 (13.1)
5,644 (45.8)
2,902 (16.7)
2,587 (14.9)
462 (15.0)
201 (6.5)
646 (32.0)
117 (5.8)
8,716 (63.3)
2,402 (17.4)
1,708 (24.9)
1,870 (27.3)
807 (27.4)
1,330 (45.2)
2,933 (23.8)
2,711 (22.0)
FDA ⫽ US Food and Drug Administration; HCP ⫽ healthcare provider.
Figure 1. New drug reporting effect analysis. Asterisks indicate that points for which there were ⬍4 months of data were excluded. Rx ⫽ prescriptions.
a high-ranking safety specialist at the FDA (November
18, 2004).13
RESULTS
Distribution of AE Reports for Non-Combination
Statin Products: Table 2 shows the cumulative number of
AE reports for 7 statin products by reporting route, report
environment, and reporter type. All statin products sold in
the United States had ⱖ1,000 US spontaneous reports over
this period (range, 1,496 for fluvastatin to 11,559 for atorvastatin). Up to December 2004, there were well over 2,000
rosuvastatin spontaneous reports from the FDA. For all
36C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 3
Change in report proportions* and reporting rates* before and after withdrawal of cerivastatin
% Change
†
All AEs
Report proportion‡
Reporting rate
Fatal AEs
Report proportion
Reporting rate
Serious AEs
Report proportion
Reporting rate
Rhabdomyolysis
Report proportion
Reporting rate
Renal failure
Report proportion
Reporting rate
Myopathy
Report proportion
Reporting rate
Myositis
Report proportion
Reporting rate
Liver failure/hepatitis
Report proportion
Reporting rate
Peripheral neuropathy/polyneuropathy
Report proportion
Reporting rate
Peripheral demyelinating neuropathy
Report proportion
Reporting rate
All Statins
Atorvastatin
0.0
⫺15.3
0.0
⫺51.4
63.6
41.0
Fluvastatin
Lovastatin
Pravastatin
Simvastatin
0.0
74.4
0.0
⫺54.1
0.0
361.5
0.0
⫺44.1
80.8
⫺13.3
266.7
535.7
51.1
⫺29.6
⫺36.1
197.4
206.7
76.8
97.7
69.7
76.4
⫺14.4
133.3
309.3
80.0
⫺17.6
⫺52.7
118.9
255.6
100.3
292.9
232.7
468.8
183.1
606.7
1,128.6
170.5
25.5
⫺39.3
181.0
566.7
275.4
137.5
106.7
133.3
8.6
40.0
142.9
633.3
260.0
⫺25.0
215.4
316.7
150.0
130.0
94.7
357.1
126.9
⫺66.7
⫺41.1
176.9
25.0
⫺70.6
38.9
350.0
147.2
63.6
32.6
66.7
⫺15.6
566.7
1,072.7
⫺13.5
⫺59.7
⫺62.5
76.0
200.0
68.1
66.7
37.7
83.3
⫺9.0
69.4
194.0
⫺15.4
⫺62.5
⫺69.4
43.4
310.0
120.9
63.6
32.6
314.3
92.6
91.7
236.4
⫺3.8
⫺56.3
11.8
422.2
46.2
⫺20.9
100.0
66.7
0.0
⫺75.0
§
§
§
§
50.0
333.3
200.0
120.0
AE ⫽ adverse event.
* Per million prescriptions.
†
Except cerivastatin.
‡
Report proportions for all AEs are calculated by dividing the total report count by itself, and therefore equal to 1.00. As such, no change is expected in
the report proportions for all AEs.
§
Reporting rates for both fluvastatin and lovastatin changed from 0.000 to 0.004. The report proportion for fluvastatin changed from 0.00 to 0.12. The
report proportion for lovastatin changed from 0.00 to 0.06.
products, there were significant proportions of expedited
reports that came from HCP reporters. A high percentage of
direct-to-FDA reports came from non-HCPs for atorvastatin, cerivastatin, rosuvastatin, and simvastatin.
Market Cycle Adjusted All-AE Reporting Rates: Figure 1 shows the all-AE reporting rate for lipid-lowering
agents adjusted for marketing cycle. When sufficient data
were available, all products except fluvastatin, cerivastatin,
and ezetimibe-simvastatin showed a characteristic new drug
reporting effect in which the highest reporting rate was
recorded during year 1 of marketing. Rosuvastatin also
appeared to be following a reporting pattern typical of the
new drug reporting effect, although only 2 points of observation were available for analysis. Only a single observation
was available for the combination product ezetimibe-simvastatin. In contrast to the other products, cerivastatin reporting peaked in year 2 and did not undergo the classic
sharp decline that is characteristic of AE reporting.
Ezetimibe, a drug introduced within 1 year of rosuvastatin,
exhibits a similarly high level of AE reporting, consistent
with secular trend bias.13
Effect of Cerivastatin Withdrawal on Reporting
Rates and Report Proportions: Tables 3 and 4 present the
reporting rates and report proportions for 7 statin agents and
all statins combined (including rosuvastatin and excluding
cerivastatin) for the periods before and after the withdrawal
of cerivastatin from the US market. Data were available for
5 agents during both periods. Data for rosuvastatin (not
marketed before cerivastatin withdrawal) and cerivastatin
are provided for reference.
There was a decrease in the US spontaneous all-AE
reporting rate for all statins combined that was attributable
to marked decreases for atorvastatin and simvastatin. However, for all fatal events, all serious events, and the 7 AEs of
interest, both report proportions and reporting rates increased for all statins combined. These effects were most
—
—
—
—
—
—
—
—
—
—
—
—
—
—
0.022
4.84
0.027
5.98
0.045
9.96
0.032
7.12
0.008
1.71
0.002
0.43
—
—
0.845
187.9
0
0.328
72.88
—
—
—
—
0.045
9.96
1.00
222.4
9.0
2002–
2004
0.001
0.03
0.011
0.43
0.018
0.69
0.011
0.43
0.010
0.38
0.008
0.30
0.028
1.07
0.44
16.65
0.022
0.83
1.00
38.15
1998–
2000
0.002
0.05
0.018
0.57
0.030
0.95
0.018
0.57
0.023
0.74
0.019
0.62
0.110
3.56
0.87
28.25
0.036
1.17
1.00
32.32
2002–
2004
All Statins†
0.001
0.04
0.007
0.27
0.018
0.67
0.009
0.32
0.007
0.26
0.009
0.35
0.016
0.59
0.55
20.69
0.026
0.98
1.00
37.79
1998–
2000
0.001
0.01
0.029
0.52
0.033
0.61
0.015
0.27
0.032
0.59
0.021
0.38
0.091
1.67
0.97
17.72
0.047
0.85
1.00
18.36
2002–
2004
Atorvastatin
AE ⫽ adverse event.
* Per million prescriptions.
†
Except cerivastatin.
‡
All values are equal to the report count divided by itself, and therefore equal to 1.00.
Reporting rate
Rhabdomyolysis
Report proportion
Reporting rate
Renal failure
Report proportion
Reporting rate
Myopathy
Report proportion
Reporting rate
Myositis
Report proportion
Reporting rate
Liver failure/hepatitis
Report proportion
Reporting rate
Peripheral
neuropathy/
polyneuropathy
Report proportion
Reporting rate
Peripheral
demyelinating
neuropathy
Report proportion
Reporting rate
Reporting rate
Fatal AEs
Report proportion
Reporting rate
Serious AEs
Report proportion
All AEs
Report proportion‡
1998–
2000
Cerivastatin
0.000
0.00
0.012
0.22
0.036
0.67
0.006
0.11
0.039
0.73
0.015
0.28
0.015
0.28
0.42
7.76
0.015
0.28
1.00
18.60
1998–
2000
0.004
0.12
0.023
0.74
0.061
1.97
0.040
1.29
0.013
0.43
0.021
0.68
0.106
3.44
0.98
31.76
0.055
1.78
1.00
32.43
2002–
2004
Fluvastatin
Table 4
Comparison of report proportions* and reporting rates*† before and after withdrawal of cerivastatin
0.000
0.00
0.026
0.96
0.026
0.96
0.037
1.34
0.013
0.48
0.003
0.10
0.061
2.20
0.50
18.25
0.045
1.62
1.00
36.30
1998–
2000
0.004
0.06
0.025
0.42
0.022
0.36
0.032
0.54
0.036
0.60
0.022
0.36
0.165
2.76
0.90
15.04
0.068
1.14
1.00
16.66
2002–
2004
Lovastatin
0.002
0.03
0.017
0.18
0.072
0.76
0.024
0.25
0.034
0.36
0.012
0.13
0.056
0.58
0.91
9.59
0.036
0.38
1.00
10.50
1998–
2000
0.003
0.13
0.019
0.94
0.022
1.09
0.009
0.44
0.010
0.50
0.009
0.41
0.034
1.63
0.43
20.99
0.023
1.13
1.00
48.46
2002–
2004
Pravastatin
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
1998–
2000
0.000
0.00
0.003
0.92
0.011
3.68
0.007
2.37
0.007
2.23
0.020
6.83
0.040
13.54
0.99
336.7
4.0
0.012
4.21
1.00
340.5
5.0
2002–
2004
Rosuvastatin
0.001
0.05
0.013
0.86
0.010
0.67
0.011
0.72
0.008
0.53
0.006
0.36
0.036
2.32
0.27
17.48
0.015
0.95
1.00
65.08
1998–
2000
0.003
0.11
0.019
0.68
0.041
1.48
0.033
1.21
0.036
1.31
0.025
0.90
0.240
8.71
0.96
35.01
0.046
1.68
1.00
36.35
2002–
2004
Simvastatin
Davidson et al/Statin Safety: Appraisal from the AERS
37C
38C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Figure 2. Expedited healthcare provider reports for all adverse events. Rx ⫽ prescriptions.
Figure 3. Direct reports for all adverse events. Rx ⫽ prescriptions.
Davidson et al/Statin Safety: Appraisal from the AERS
39C
Figure 4. Expedited healthcare provider reports for rhabdomyolysis. Rx ⫽ prescriptions.
pronounced for rhabdomyolysis and renal failure. Table 4
also demonstrates a dramatic increase in the all-AE reporting rate for rosuvastatin versus other statin agents, whereas
the AE-specific report proportions for rosuvastatin resembled those seen for the statin class as a whole. This effect
can be seen in the form of higher reporting rates for all key
AEs of interest for this product.
Effects of Intermittent Publicity: Total AE reporting
for atorvastatin shows a response only to the withdrawal of
cerivastatin. No other response to publicity is seen for any
of the conditions of interest for either direct reporting or
expedited HCP reporting (Figures 2 and 3).
For expedited HCP reporting of rosuvastatin-associated
events, each of the last 4 publicity points was associated
with a spike in the reporting of all AEs (Figure 2), rhabdomyolysis (Figure 4), and renal failure excluding rhabdomyolysis (Figure 5). These responses appear to also mimic the
new drug reporting effect. A response appears to have
occurred in the expedited HCP reporting of myopathy to the
Wolfe editorial12 and the FDA staffer’s testimony before
Congress (data not shown).13 Expedited HCP reporting of
myositis and of liver failure and hepatitis appears to have
shown a response to the petition to remove rosuvastatin
from the market and to the Wolfe editorial (data not shown).
Neither category of neuropathies shows any response to
publicity, possibly owing to the sparseness of data (data not
shown).
For direct reporting of rosuvastatin-associated events,
each of the last 4 publicity points was associated with a
spike in the reporting of all AEs (Figure 3), rhabdomyolysis
(Figure 6), and renal failure excluding rhabdomyolysis (Figure 7). A response appears to have occurred in the expedited
HCP reporting of myopathy and myositis to the FDA
staffer’s testimony before Congress (data not shown). Expedited HCP reporting of liver failure and hepatitis appears
to have shown a response to the petition to remove rosuvastatin from the market and to the FDA staffer’s testimony
before Congress (data not shown). As with expedited HCP
reporting, neither category of neuropathies showed any response to publicity (data not shown).
DISCUSSION
Postmarketing surveillance systems are used by worldwide
regulatory authorities to monitor AEs after the introduction
of new products. Since 1963, when such monitoring systems were first proposed, the assessments derived from this
kind of observational data have repeatedly been shown to be
valuable in identifying the presence and extent of druginduced AEs.14 –16 In the last 20 years, a trend toward
increased AE reporting to the FDA has been noted.10 The
present analysis used surveillance methodology to analyze
recent trends in statin-associated AEs such as rhabdomyol-
40C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Figure 5. Expedited healthcare provider reports for renal failure excluding rhabdomyolysis. Rx ⫽ prescriptions.
ysis.17,18 Reporting analyses are particularly useful in the
assessment of serious AEs that occur at a sufficiently low
rate so that detection and characterization of the AEs in
clinical trials often are not possible.
The main report-based metric that has been used to
gauge the magnitude of AE reporting in AE surveillance
systems is the reporting rate.19 Reporting rates are formed
by dividing an incompletely ascertained numerator (number
of forwarded AE reports) by a completely ascertained measure of patient use (eg, number of prescriptions issued over
the same period). In stable reporting situations, comparison
of the reporting rates for a particular AE across a therapeutic
class can be a useful way to locate outlier products that may
be particularly prone to cause an important AE.14 However,
in the presence of dynamic effects such as publicity, the
incompletely ascertained numerator of a reporting rate can
be markedly increased, while the completely ascertained
denominator remains the same. This shortcoming makes
reporting rate comparisons very sensitive to differential
effects that affect the AE reporting behavior for a product in
comparison with its peer group.
These data show that 3 such differential factors have
affected the AE reporting for rosuvastatin: (1) the new drug
reporting effect (Weber effect20); (2) the withdrawal of
cerivastatin from the US market in August 2001; and (3)
intense recent publicity regarding the side-effect profile for
rosuvastatin. As a result, although the reporting rates for
several postrosuvastatin AEs of interest are substantially
higher than for the statin class as a whole, this does not
necessarily indicate an increase in the incidence of these
events.
The all-AE reporting rate analysis for statin products and
ezetimibe presented in this article showed that 7 (including
rosuvastatin) of the 9 examined products (the exceptions are
fluvastatin and cerivastatin) demonstrated a new drug reporting effect. The new drug reporting curves for the 2 most
recently approved agents, the nonstatin ezetimibe and rosuvastatin, were of greatest magnitude—an observation in
keeping with increasing levels of AE reporting over the last
decade (ie, US secular AE reporting trend). In contrast the
cerivastatin all-AE reporting curve did not show a new drug
reporting effect, but was instead elevated and sustained.
This latter pattern of all-AE reporting may be an important
differentiating feature of drug products that possess particularly disadvantageous AE profiles, and, thus far, has not
been seen with rosuvastatin.
For statin agents, there was widespread upward movement in the percentage changes for report proportions and
reporting rates between the pre- and postcerivastatin withdrawal periods. For all categories and key AEs that were
examined, both the report proportion and the reporting rate
increased for the statin class as a whole, with changes in
report proportion somewhat exceeding those for reporting
rate. The increase in reporting rate was particularly notable
for rhabdomyolysis. In contrast, the all-AE reporting rate
over the same period decreased. This indicates that ceriv-
Davidson et al/Statin Safety: Appraisal from the AERS
41C
Figure 6. Direct reports for rhabdomyolysis. Rx ⫽ prescriptions.
astatin’s withdrawal from the US market likely provoked
heightened concern about the statin class as a whole that
manifested in increased reporting activity for particular AEs
of interest, with special emphasis on rhabdomyolysis.
An examination of the postcerivastatin period showed
that rosuvastatin’s reporting rates for both all-AEs and AEs
of interest (other than neurologic AEs) were about 3–12
times greater than the comparable rates for the statin class as
a whole. However, rosuvastatin’s proportionate reporting
for the same events was about the same or lower than that
seen for all statins. This implies that rosuvastatin, which
commenced marketing in the postcerivastatin period, has
been subject to a generalized effect in which all aspects of
its AE reporting have been affected. As a result, even
though rosuvastatin’s proportionate reporting of key AEs
has been lower than for the statin class as a whole, its
reporting rates have been far higher. Such a disparity can
only occur in situations where overall AE reporting is markedly increased when measured against a comparator, while
the proportion of reporting attributable to a particular AE is
relatively unaffected.
The chronologic displays presented in this article show
that, in addition to the new drug reporting effect and more
active generalized reporting dynamics after the withdrawal
of cerivastatin, mass media publicity regarding rosuvastatin
has also been a major contributor to elevated rosuvastatin
AE reporting. When 2 indices that respond quickly to publicity were followed over time (manufacturer’s expedited
and direct-to-FDA report counts), it was seen that known
major publicity points were closely followed by prominent
increases in reporting activity. Although documented for all
AEs, rhabdomyolysis, and renal failure in this article, the
phenomenon was also seen with virtually any analysis of an
important AE in the rosuvastatin report set (data not shown).
In contrast to cerivastatin withdrawal, the effect of mass
media publicity on rosuvastatin reporting has been highly
focused only on this agent, because the comparable longterm reporting trends for atorvastatin remained relatively
unaffected. The several episodes of negative, product-specific publicity that appear to have increased the AE reporting for rosuvastatin include the petition by Public Citizen to
remove rosuvastatin from the US market (March 4, 2004),11
an editorial in a medical journal by Sidney M. Wolfe, Public
Citizen spokesperson, that called for the removal of rosuvastatin from the US market (June 26, 2004),12 and FDA
staffer David Graham’s congressional testimony, in which
rosuvastatin was identified as 1 of 5 problem drugs that
should be removed from the market (November 18, 2004).13
Although reporting rate trend analyses are an important postmarketing surveillance tool for monitoring product-associated AEs, their interpretation should take documented reporting biases into account. In addition to the
effects of publicity, reporting rates can be affected by the
previously mentioned new drug reporting effect, secular
trends in AE reporting, and interproduct differences that
arise from variability in the data collection infrastructure
42C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Figure 7. Direct reports for renal failure excluding rhabdomyolysis
of different companies.10,20,21 It also should be emphasized that AE reporting rates differ from the incidence
rates that are reported in clinical trials as well as from
epidemiologic studies in which all new AEs and patient
exposure data are available. Instead, spontaneous postmarketing reporting relies primarily on the recognition
by the HCP or patient that an AE is associated with a
particular drug, and their willingness to report the case to
a pharmaceutical company or regulatory authority. As a
result of this susceptibility to highly variable behaviors,
trend analysis based on reporting rates should be considered hypothesis-generating, and is best interpreted within
the overall context of an existing body of scientific and
public health knowledge.
In summary, our analyses indicate that rosuvastatin’s
reporting patterns do not differ from those of other wellestablished statin products when typical differential reporting effects are taken into account. Examination of the
report proportion metric is particularly useful in evaluating elevated AE reporting rates that have been subject
to upward reporting biases. In the case of rosuvastatin,
the report proportion metric is reassuring because, unlike
cerivastatin, the proportionate reporting of key AEs has
been comparable to the remainder of the statin class. In
conclusion, the initial US reporting rates for total and
fatal rhabdomyolysis for rosuvastatin are consistent with
the preapproval safety database for rosuvastatin and the
results seen for the other currently available statins in the
period following cervastatin’s withdrawal. Continued
surveillance monitoring for all statins will be useful in
determining future reporting trends for serious AEs.
1. National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults
(Adult Treatment Panel III). Third Report of the National Cholesterol
Education Program (NCEP) Expert Panel on Detection, Evaluation, and
Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel
III) final report. Circulation 2002;106:3143–3421.
2. Davidson MH. Safety profiles for the HMG-CoA reductase inhibitors:
treatment and trust. Drugs 2001;61:197–206.
3. Staffa JA, Chang J, Green L. Cerivastatin and reports of fatal rhabdomyolysis. N Engl J Med 2002;346:539 –540.
4. Davidson MH. Controversy surrounding the safety of cerivastatin.
Expert Opin Drug Saf 2002;1:207–212.
5. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection
Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet 2002;360:7–22.
6. Omar MA, Wilson JP, Cox TS. Rhabdomyolysis and HMG-CoA
reductase inhibitors. Ann Pharmacother 2001;35:1096 –1107.
7. Black C, Jick H. Etiology and frequency of rhabdomyolysis. Pharmacotherapy 2002;22:1524 –1526.
8. Omar MA, Wilson JP. FDA adverse event reports on statin-associated
rhabdomyolysis. Ann Pharmacother 2002;36:288 –295.
9. Thompson PD, Clarkson P, Karas RH. Statin-associated myopathy.
JAMA 2003;289:1681–1690.
10. Rodriguez EM, Staffa JA, Graham DJ. The role of databases in drug
postmarketing surveillance. Pharmacoepidemiol Drug Saf 2001;10:
407– 410.
11. First Petition for Removal of Rosuvastatin from US Market (letter from
Sidney M. Wolfe, Director of Health Research Group to the FDA). [US
Davidson et al/Statin Safety: Appraisal from the AERS
12.
13.
14.
15.
16.
Food and Drug Administration Web site]. March 10, 2005. Available at
http://www.fda.gov/ohrms/dockets/dockets/04p0113/04p-0113-let0002.
pdf. Accessed August 18, 2005.
Wolfe SM. Dangers of rosuvastatin identified before and after FDA
approval. Lancet 2004;363:2189 –2190.
Testimony of David Graham, MD, MPH, Associate Director for Science and Medicine, US Food and Drug Administration, before the US
Senate Committee on Finance. November 18, 2004. Available at:
http://senate.gov/⬃finance/hearings/testimony/2004test/111804dgtest.
pdf. Accessed March 9, 2006.
Rossi AC, Hsu JP, Faich GA. Ulcerogenicity of piroxicam: an analysis
of spontaneously reported data. BMJ 1987;294:147–150.
Clark JA, Klincewicz SL, Stang PE. Spontaneous adverse event signaling methods: classification and use with health care treatment products. Epidemiol Rev 2001;23:191–210.
Woods SW, Martin A, Spector SG, McGlashan TH. Effects of development on olanzapine-associated adverse events. J Am Acad Child
Adolesc Psychiatry 2002;41:1439 –1446.
43C
17. Physicians’ Desk Reference, 58th Edition. Montvale, NJ: Thompson
PDR, 2004.
18. Pasternak RCD, Smith SC Jr, Bairey-Merz CN, Grundy SM, Cleeman
JI, Lenfant C, for the American College of Cardiology, the American
Heart Association, and the National Heart, Lung, and Blood Institute.
ACC/AHA/NHLBI advisory on the use and safety of statins. J Am Coll
Cardiol 2002;40:567–572.
19. US Food and Drug Administration. Guidance for industry: good pharmacovigilance practices and pharmacoepidemiologic assessment [US
FDA Web site]. March 2005. Available at: http://www.fda.gov/cder/
guidance/6359OCC.pdf. Accessed March 9, 2006.
20. Weber JCP. Epidemiology of adverse reactions to nonsteroidal antiinflammatory drugs. In: Rainsford KD, Velo GP (eds). Side Effects of
Antiinflammatory/Analgesic Drugs: Advances in Inflammation Research. Vol. 6. New York: Raven Press; 1984:1–7.
21. Tsong Y. Comparing the reporting rates of adverse events between
drugs with adjustment for year of marketing and secular trends in total
reporting. J Biopharm Stat 1995;5:95–114.
Statin Safety: Lessons from New Drug Applications for
Marketed Statins
Terry A. Jacobson, MD
Safety has become a central issue in the management of dyslipidemia with statins. A
review of New Drug Applications (NDAs) and the US Food and Drug Administration
(FDA) Web site was conducted for all 3-hydroxy-3-methylglutaryl coenzyme A
(HMG-CoA) reductase inhibitors, or statins, with a major focus on cerivastatin and
rosuvastatin. The findings provide insight into the incidence of adverse events for this
class of drugs and support the significant benefits of statins relative to associated
risks. These data delineate the nature of statin associated liver, muscle, and renal
adverse events. Although transaminase levels increase in a dose-related fashion with
statins, a definitive correlation between statin therapy and hepatotoxicity is not
supported by statin NDA data. Statin-induced myopathy is a relatively rare event (1
in 1,000) and rhabdomyolysis is even rarer (1 in 10,000). The cerivastatin NDA, along
with its supplementary NDA, was the first to demonstrate a clear statin dose-response
relation with myopathy and a threshold effect above which myotoxicity increases
significantly. Proteinuria was identified as a consequence of statin therapy with data
from the rosuvastatin NDA, and subsequent analysis suggests a class effect that is
dose related but transient. Studies in cell culture suggest the mechanism is a pharmacologic effect on the proximal renal tubule. The available evidence suggests no
clear renal toxicity with currently approved statins, because no declines in renal
function or glomerular filtration rate have been documented over time. Overall,
currently marketed statins have a very favorable benefit-to-risk relation with respect
to liver, muscle, and renal issues. © 2006 Elsevier Inc. All rights reserved. (Am J
Cardiol 2006;97[suppl]:44C–51C)
After the introduction of lovastatin in 1987, the management of dyslipidemia was focused on the lipid-lowering
efficacy of the 3-hydroxy-3-methylglutaryl coenzyme A
(HMG-CoA) reductase inhibitors, or statins, and the significant impact these agents have on decreasing cardiovascular
morbidity and mortality.1–3 However, with the withdrawal
of cerivastatin from the market in 2001, safety became a
central issue in the use of statins, as the US Food and Drug
Administration (FDA) drug review process appeared to
have missed the significant risk of muscle toxicity with
cerivastatin at higher doses.4 –7 Consequently, approval of
the next statin, rosuvastatin, resulted in generation of a
database containing 4 times the number of patients of that
for any previously approved statin (Table 1). For a New
Drug Application (NDA), the FDA generally recommends
that approximately 3,000 patients be studied to expose an
adverse event with an incidence rate of 1:1,000 with 95%
confidence intervals.8 Therefore, rare events (eg, statinrelated myopathy) may not be well characterized, and very
rare events (eg, statin-related rhabdomyolysis) may not occur in the patient population studied for approval.
The vast amount of information collected and reviewed
by the FDA for each drug undergoing the approval process
reflects the tremendous effort required by the sponsor and
the FDA to bring safe and effective medications to the
market. The FDA approval process requires a drug’s sponsor (usually the manufacturer) to submit an NDA9 that
contains comprehensive reports of all studies to facilitate
FDA evaluation of the data. The entire database is used to
identify the risks and rate of adverse effects. From analyses
of these data, FDA reviewers assess the relation of benefits
to risks, and an FDA advisory committee determines
whether a drug should be recommended for approval. Analysis of the FDA database can provide significant insight into
the safety profile of a drug. Thus, to better understand the
safety profile of statins an analysis of all statin NDA submissions was undertaken.
Methodology: US Food and Drug Administration
Sources of Information
Office of Health Promotion and Disease Prevention, Emory University,
Atlanta, Georgia, USA.
Address for reprints: Terry A. Jacobson, MD, Office of Health Promotion and Disease Prevention, Emory University, Faculty Office Building,
49 Jesse Hill Jr Drive SE, Atlanta, Georgia 30303.
E-mail address: tjaco02@emory.edu.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.009
The information for this report was obtained through 2
primary sources (1) the FDA Center for Drug Evaluation
and Research (CDER) Web site,9,10 which provides an extensive database of information for individual drugs, and (2)
www.AJConline.org
Jacobson/Statin Safety: Lessons from New Drug Applications
45C
Table 1
Timeline of statin New Drug Application (NDA) approvals
Statin
Patients in NDA Database (N)
Approval Date
873
1,925
2,423
2,342
1,965
2,815
12,569
August 1987
October 1991
December 1991
December 1993
December 1996
June 1997
August 2003
Lovastatin
Pravastatin
Simvastatin
Fluvastatin
Atorvastatin
Cerivastatin
Rosuvastatin
Freedom of Information (FOI) requests, which provided the
NDAs and Summary Basis for Approval documents5,11–16
for each of the statins. Package inserts were also reviewed
and compared with information in the specific drug’s
NDA.17–22 Package inserts are readily available and widely
used by practitioners, but are not as comprehensive as the
NDA.
An NDA contains a large amount of information that is
generally not readily available in the public domain, and
often is not published comprehensively in the medical literature. Postmarketing adverse side effect reporting is also
required by the FDA and can be found on the FDA Web
site.9,10 The FDA Web site archives communications between the sponsor and the FDA reviewers. The site includes
such items as chemistry reviews, pharmacology reviews,
statistical reviews, and labeling guidelines for each drug.
One of the most clinically relevant reports is the medical
review, which gives a very detailed analysis of the data by
FDA reviewers, including categorizing the data as acceptable or unacceptable, providing interpretations, and requesting further data or studies
Statin New Drug Application Submissions: Historical
Perspective
Lovastatin was the first statin submitted for approval and
was approved with an NDA containing data on 873 patients;
the primary safety concerns at the time were lens opacities
and liver toxicity.15 Slit-lamp examinations were recommended on an annual basis to detect development of cataracts (now known not to be a risk with statin therapy), and
serum transaminase levels were required every 12 weeks.
Initially, risk for myopathy or myalgia was not considered a
potentially significant consequence of lovastatin therapy.
Following the approval of lovastatin, the number of patients
studied for each statin NDA increased to approximately
2,000 –3,000, with the exception of the rosuvastatin NDA
(Table 1).11 Although a sufficient number of patients was
included in the cerivastatin NDA (N ⫽ 2,815) to identify the
occurrence of myopathy, the significant increase in risk for
myopathy relative to other statins and the increased risk of
the more serious and rare rhabdomyolysis was not fully
appreciated until after approval of the initial NDA. The
supplemental NDAs for cerivastatin at the 0.4-mg and
0.8-mg dose levels began to suggest additional myotoxicity.6
Cerivastatin approval history: Review of the cerivastatin approval process and postmarketing data is an excellent case study illustrating how rare, potentially serious adverse events can be missed in the approval process.
The sequence of events with cerivastatin begins in June
1997, when the 0.2-mg and 0.3-mg doses were approved
and the risk of rhabdomyolysis was added as a warning to
the approved label in July.23 In August 1998 a supplemental NDA was submitted requesting approval of a
0.4-mg dose (Table 2), and soon after the first case of a
cerivastatin and gemfibrozil interaction associated with
rhabdomyolysis was published.24 A change was made to
the 0.4-mg dose NDA in May 1999, adding a warning
regarding concomitant use with gemfibrozil. The NDA
for the 0.8-mg dose was submitted in September 1999,
followed by a letter to practitioners in December warning
of the contraindication for using gemfibrozil with cerivastatin. As with cerivastatin, sponsors often request approval of the lowest doses initially, and supplemental
NDAs are submitted afterward to request approval of
higher doses. Often the higher doses are studied in fewer
patients. Of note, an increased risk of myopathy in thin,
elderly women given the 0.8-mg dose was recognized and
reported by an FDA medical reviewer but, in the final
analysis, this was not considered significant enough to
prevent approval.6 Cerivastatin was voluntarily withdrawn from the market in August 2001 by Bayer, Inc.
(West Haven, CT) because of a significantly higher rate
of rhabdomyolysis than was observed with other statins.6,23
Rosuvastatin approval history: The initial rosuvastatin NDA was submitted in June 2001 after the withdrawal
of cerivastatin.11,25 The cerivastatin experience significantly increased the initial awareness of safety issues for
all of the statins and rosuvastatin’s NDA contained data
on 3,903 patients. The FDA ultimately denied approval
of the 80-mg dose because the lipid-lowering benefits
were outweighed by the increased risks for renal toxicity
and myotoxicity. With the denial, the FDA also requested
46C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 2
Cerivastatin history and timeline
Date
June 1997
July 1998
August 1998
April 1999
May 1999
September 1999
December 1999
July 6, 2000
July 11, 2000
April 2001
August 2001
Event
NDA approved for cerivastatin 0.2 mg and 0.3 mg
Label update: rhabdomyolysis added to warnings
Supplemental NDA for cerivastatin 0.4 mg submitted
First published case report of rhabdomyolysis with cerivastatin–gemfibrozil
coprescription
Supplemental NDA for cerivastatin 0.4 mg approved; label update:
additional gemfibrozil warnings
Supplemental NDA for cerivastatin 0.8 mg submitted
Label update: gemfibrozil coprescription contraindication
FDA medical review of cerivastatin 0.8 mg supplement identified thin,
elderly women at increased risk of CK elevation to ⬎10 ⫻ ULN
Supplemental NDA for cerivastatin 0.8 mg approved
Label update: cerivastatin starting dose should be 0.4 mg
Cerivastatin withdrawn form US market
CK ⫽ creatine kinase; FDA ⫽ US Food and Drug Administration;
NDA ⫽ New Drug Application; ULN ⫽ upper limit of normal.
Adapted from JAMA.23
more safety data from AstraZeneca (Wilmington, DE) on
rosuvastatin 20 mg and 40 mg, because the initial NDA
was heavily weighted toward the 10-mg and 80-mg
doses. As a result, additional studies were completed, and
12,569 patients were included in the revised NDA for
rosuvastatin submitted in February 2003. Approximately
4,000 patients were treated with the 40-mg dose alone, a
greater number of patients than for all doses of any other
statin NDA. The rosuvastatin NDA provided a database
of approximately 4 times the number of patients of any
previous statin NDA, allowing for significantly better
characterization of adverse events. However, ⬍1 year
after approval of rosuvastatin in March 2004, a Public
Citizen petition was submitted to the FDA requesting
removal of rosuvastatin from the market. The petition
was subsequently rejected by the FDA in March 2005,8
primarily because of the extensive database provided in
the NDA. The FDA’s response is available on their Web
site8 and is an in-depth review of FDA data available on
statin safety. However, the FDA did recommend additional collection of postmarketing pharmacoepidemiologic data on rosuvastatin.8 In addition to the large number of patients in the rosuvastatin database, the additional
patients more accurately reflected the population treated
with statins. The mean age was 58 years, with ⬎33% of
these patients ⬎65 years of age—much older than patients evaluated in other statin NDAs. Approximately
50% of the patients had renal impairment defined by their
glomerular filtration rate using the Cockcroff-Gault equation. The patients had significant comorbidities, including hypertension (51%), cardiovascular disease (36%),
and diabetes mellitus (16%). Drug exposure data were
also greater in the rosuvastatin NDA, because 4,000
patients received the 40-mg dose for ⱖ1 year and 1,100
patients received it for 2 years.
Statin Safety: Perspectives from the New Drug
Applications
Current concerns for adverse events with statins involve
liver, muscle, and kidney. The NDAs for cerivastatin and
rosuvastatin, in conjunction with other FDA documents,
provide significant insight into the characterization of these
adverse events for all statins.
Hepatotoxicity: Liver toxicity has been a concern with
statin therapy beginning with lovastatin, and monitoring of
serum transaminases has been recommended for all
statins.15,17–22 Initial monitoring guidelines suggested measuring transaminase levels before initiating therapy, again at 12
weeks, and then periodically thereafter depending on dose
escalation. Currently, recommendations suggest assessing levels before initiating therapy, 12 weeks after initiating therapy,
and after a dose increase.17–22,26 A pooled analysis of all statin
NDA data showed no correlation between persistent increased
alanine aminotransferase (ALT) concentrations and statin-induced lowering of low-density lipoprotein (LDL) cholesterol
(Figure 1).5,11–22 However, statins do increase transaminases in
a dose-response fashion, with transaminitis defined as 3 times
the upper limit of normal (ULN). From Figure 1, it is clear that
the highest doses of statins generally have higher rates of
transaminitis. This level of transaminitis does not necessarily
reflect liver damage and may not represent just liver synthesized enzymes, as aspartate aminotransferase (AST) and to a
lesser extent ALT are released by damaged muscle cells.26 Of
critical importance is that a cardinal sign of drug-induced
hepatic damage, transaminitis accompanied by bilirubin elevation, was absent or rarely seen in most of the statin NDAs. A
review of the NDAs does not support a causal relation between
statin therapy and liver damage, suggesting monitoring of
transaminases may not provide clinical benefit.5,11–16 Also notable in the NDA submissions is the frequency with which
Jacobson/Statin Safety: Lessons from New Drug Applications
47C
Figure 1. Percent alanine aminotransferase (ALT) elevations ⬎3 times (3⫻) the upper limit of normal
(ULN) relative to (A) percent low-density lipoprotein cholesterol (LDL-C) reduction and percent
ALT elevations and (B) statin dose. There is no correlation between the percent decrease in LDL-C
and persistent ALT elevations ⬎3⫻ the ULN (A). However, there is a correlation between persistent
ALT concentrations ⬎3⫻ the ULN and increasing statin dose (B). (Adapted from New Drug
Applications5,11–16 and package inserts17–22 for the various statins.)
transaminase elevations are observed: approximately 20% of
patients taking statins having an increase in ALT and AST ⬎2
times the ULN. However, long-term followup of increased
liver enzymes shows progression to liver failure to be exceedingly rare, and concerns regarding statin-induced hepatotoxicity from existing NDAs appear to be overstated.
Muscle toxicity: A primary problem in determining the
absolute risk of muscle-related adverse events with statins is
a lack of standardized definitions and a failure to carefully
characterize adverse events reported in the literature. The
FDA defines myopathy as creatine kinase (CK) concentrations ⱖ10 times the ULN and defines rhabdomyolysis by
associated clinical features, such as renal failure requiring
hospitalization, or intravenous hydration, or myoglobinuria.
In a pooled analysis of all statin NDAs, the increase in CK
concentrations was not found to be related to the effectiveness of the reduction of LDL cholesterol (Figure 2).5,6,11–22
However, a direct correlation between increasing dose and
CK levels is observed. The cerivastatin NDA provided the
first example of how statin-induced muscle toxicity may
exhibit a threshold effect.5,6 For the low doses (0.2 mg and
0.3 mg cerivastatin), myopathy was not significant. However, at 0.4 mg the incidence increases dramatically to
1.5%, higher than for any marketed statin, suggesting the
threshold for this adverse event had been reached. With the
0.8-mg dose the incidence of myopathy was ⬎2%. As a
result of these numbers, an FDA subanalysis was undertaken, which showed a very high incidence of myopathy in
48C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Figure 2. Percent creatine kinase (CK) ⬎10 times (10⫻) the upper limit of normal (ULN) relative to
(A) the percent reduction in low-density lipoprotein cholesterol (LDL-C) and (B) the percent CK
⬎10⫻ the ULN relative to the statin dose. There is no correlation between the percent decrease in
LDL-C and CK elevations (A). However, there is a correlation between increased CK concentrations
and statin dose (B). Cerivastatin is plotted at 100 times the dose given. (Adapted from New Drug
Applications5,11–16 and package inserts17–22 for the various statins.)
certain subpopulations including thin, elderly women. In
elderly women, the 0.4-mg dose (n ⫽ 27) had a myopathy
incidence of 7.4%, and the 0.8-mg dose (n ⫽ 90) had an
incidence of 5.6%.6 Regardless of these data, the doses were
subsequently approved for marketing.
Other characteristics of myopathy identified in the cerivastatin NDA were that 45% (9 of 20) of patients with CK
elevations ⬎10 times ULN were asymptomatic, with only 1
patient complaining of muscle weakness. In addition CK elevations could occur at any time during therapy, in contrast to
transaminase elevations, which generally occur early in therapy or after an increase in dose. Of note, increases in CK
concentrations were often accompanied by increases in
transaminase concentrations, supporting the theory of muscle
cell contribution to transaminase elevations in statin therapy.
The now obvious conclusion from the cerivastatin expe-
rience is that as the statin dose (or more likely statin serum
concentration) increases, the risk of CK elevation increases
to the point where a threshold level is reached. Above this
level, myotoxicity begins to accelerate to levels beyond
acceptable risk– benefit ratios.5,6,27 From the NDA data and
additional postmarketing FDA data, the cerivastatin threshold dose appears to be at the 0.4 mg dose. For statins
presently on the market, the threshold concentrations appear
to be above currently approved doses, although in certain
populations or with significant drug– drug interactions, statin concentrations can be elevated to myotoxic levels.
The submission of the rosuvastatin NDA after withdrawal
of cerivastatin resulted in a more critical review of the rosuvastatin data in comparison with the approval process for other
statins.25,28 For example, the review questioned whether the
incidence of myotoxicity per percent of LDL lowering was
Jacobson/Statin Safety: Lessons from New Drug Applications
49C
Figure 3. Plasma rosuvastatin concentrations by dose in 6 patients with rhabdomyolysis or renal
toxicity. (Adapted from rosuvastatin New Drug Application.25)
similar with other statins and whether the risk was adequately
evaluated. The sponsor was asked to show data from enough
patients and to include patients in potentially high-risk populations to ensure that the risks were well characterized. The
rosuvastatin NDA shows the incidence of myopathy to be as
low as, or lower than, for other statins; for the various doses,
rates were 0.1% (10 mg), 0.1% (20 mg), and 0.4% (40 mg).
The incidence of myopathy was 0.9% with the 80-mg dose that
did not receive FDA approval and may represent the threshold
dose for rosuvastatin. The complete rosuvastatin NDA with 6
months of additional postmarketing data (N ⫽ 13,395) demonstrated 1 case of rhabdomyolysis, which occurred in a patient taking the 20-mg dose (for an overall incidence of
0.01%).25 The data for the NDA for all statins suggest that
myopathy occurs in approximately 1 in 1,000 patients, and that
the rate of rhabdomyolysis is very rare, occurring in about 1 in
10,000 patients.
The rosuvastatin NDA provides a limited amount of
data regarding serum statin concentrations and muscle
toxicity.8,11,25 The dose-concentration curve was generally linear across the 20-mg to 80-mg dose range (Figure
3).25 In all, 6 patients with either rhabdomyolysis or renal
insufficiency, who received 80 mg of rosuvastatin daily,
had significantly elevated serum concentrations ⬎50 ng/
mL. Among asymptomatic patients treated with 80 mg of
rosuvastatin, 33% had serum concentrations ⬎50 ng/mL.
These data suggests a potential threshold dose (80 mg) at
which the risks of muscle and renal toxicity are increased. The rosuvastatin NDA gives a rare glimpse of
statin myotoxicity above currently approved doses.
Renal toxicity: The primary concern with rosuvastatin 80
mg was an incidence of proteinuria of 12%–15%, which is
significantly higher than seen at lower doses (Table 3).8,11,25,28
Proteinuria had not been associated with statin therapy in
previous NDAs, leading the FDA to question whether the
sponsor had adequately addressed the clinical safety findings
of rosuvastatin-associated proteinuria. In addition, the FDA
asked its advisory committee to determine whether the risk of
renal function impairment was adequately investigated,
whether proteinuria was a statin class effect, and, if so, whether
the potential for proteinuria with rosuvastatin was similar to
that for other statins. Finally, the FDA asked if there was any
need for proteinuria monitoring in clinical practice for rosuvastatin or any other statin. A review of all statin NDAs suggests
a possible class effect, because proteinuria is documented in
each application, but proteinuria had not been directly related
to statin administration before the rosuvastatin NDA. An exception appears in a published letter indicating that simvastatin
may cause proteinuria.29 A major limitation of the statin-associated proteinuria data was the lack of a direct placebo comparator, because the data were generated, in part, during the
open-label, follow-up studies after patients were switched from
placebo to active drug. Also, the methods used to measure
proteinuria were urine dipstick or spot urine analyses, but did
not include 24-hour urine collection or uniform times of
collection.
Evaluation of hematuria with rosuvastatin 80 mg showed
an incidence of 12%, and the incidence of hematuria and
proteinuria occurring in the same patients was 6.1%. As
with the proteinuria, these represented incidences significantly higher than with any other marketed statin. Comprehensive proteinuria analyses were conducted in 50 patients.
An analysis of the renal source of the proteinuria indicated
the majority was tubular in origin, as it included low-
50C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 3
Frequency of proteinuria, hematuria, and proteinuria/hematuria
Treatment
Dose
(mg)
N
Urine Protein
ⱖ2⫹ (%)
Urine Blood
ⱖ1⫹ (%)
Proteinuria
ⱖ2⫹ and
Hematuria
ⱖ1⫹ (%)
—
5
10
20
40
80
10
20
40
80
20
40
80
20
40
372
653
1,202
1,460
2,384
804
710
667
245
377
517
356
337
191
67
3.0
1.0
2.0
2.0
4.0
12.0
2.0
2.0
0.4
0.5
4.0
2.0
0.6
1.0
0
5.0
6.0
7.0
4.0
10.0
12.0
4.0
3.0
2.0
2.0
5.0
5.0
8.0
7.0
4.0
0
0
0.3
0.3
1.3
6.1
0.6
0.3
0.4
0
0.6
0.8
0.3
0.5
0
Placebo
Rosuvastatin
Atorvastatin
Simvastatin
Pravastatin
Adapted from rosuvastatin New Drug Application.11
Table 4
Patients with proteinuria* at ⱖ96 weeks of rosuvastatin treatment†
Proteinuria, n (%)
Dose
(mg)
N
Any Visit
Last Visit
5
10
20
40
80
ⱖ40§
261
838
112
100
590
807
3 (1.1)
17 (2.0)
5 (4.5)
4 (4.0)
99 (16.8)
136 (16.9)
0
4 (0.5)
1 (0.9)
2 (2.0)
37 (6.3)
10 (1.2)
Creatinine ⬎30%‡
0
0
0
0
7
0
* Proteinuria defined as “none or trace” to “2⫹ or greater.”
†
Combined all controlled/uncontrolled and real-time lab data pool.
‡
At last visit.
§
Includes patients who backtitrated from the 80-mg dose.
Adapted from rosuvastatin New Drug Application11,25
molecular-weight proteins.25,28 Proteinuria resulting from
glomerular damage was evident in some of the patients;
however, in this small sample size, a significant number of
the patients had diabetes and hypertension. To further elucidate the mechanism of the proteinuria, opossum OK cells
were studied, because they are a well-accepted animal
model of proximal tubular absorption. In the presence of
each of the statins, the OK cells exhibited an inhibition in
the uptake of albumin in a dose-dependent fashion, suggesting that the mechanism of low-molecular-weight protein
loss in the urine is a class effect and is dependent on drug
concentration. Further supporting this mechanism, the addition of mevalonate completely reversed the inhibition of
protein uptake in the OK cells. The production of mevalonate is inhibited by the pharmacologic action of statins.
The rosuvastatin NDA data were analyzed to determine
whether an association existed between proteinuria, hematuria, and advancing renal disease by examining the incidence of serum creatinine increase ⱖ30% of baseline.25 The
clinical justification for using this metric of advancing renal
disease is not established, but it was requested by the FDA.
The incidence of creatinine increasing in patients with proteinuria only was greatest with the 40-mg dose, although it
occurred in only 0.4% of patients on this dose. For patients
with proteinuria and hematuria the highest incidence of a
30% creatinine increase was 0.3%, which occurred in patients given the 40-mg dose. Follow-up data including ⬎96
weeks of patients with proteinuria at any visit given the
40-mg dose showed no increase in serum creatinine ⱖ30%
(Table 4).11,25 For patients with or without impaired renal
function at baseline, serum creatinine levels decreased, suggesting a potential renal-protective effect.
Conclusion
The NDAs contain significant and important safety information that is often unpublished, as highlighted by the cases
Jacobson/Statin Safety: Lessons from New Drug Applications
of the rosuvastatin and cerivastatin NDAs. These NDAs
demonstrate statin myopathy and myotoxicity increase with
dose and suggest a threshold effect for statin-related muscle
toxicity. Although transaminases increase with increasing
dose, data from the NDAs do not support a causal relation
between statins and liver toxicity or liver failure. The rosuvastatin NDA indicated tubular proteinuria may be a side
effect of all statin therapy that had not previously been
recognized; however, long-term data suggest that tubular
proteinuria with statin therapy is transient and does not
progress to chronic renal disease. The mechanism of proteinuria appears to be a pharmacologic effect on proximal
renal tubule function that is dose dependent. Consistent with
a dose-related pharmacologic effect, the effect is reversible
with a decrease in dose. In summary, the NDAs for all
approved statins demonstrate a very favorable benefit-torisk relation with respect to liver, muscle, and renal issues.
With all statins, there appears to be a threshold level of dose
where risks exceed benefits.
10.
11.
12.
13.
14.
15.
16.
17.
1. Downs JR, Clearfield DO, Weis S, Whitney E, Shapiro DR, Beere PA,
Langendorfer A, Stein EA, Kruyer W, Gotto AM Jr, for the AFCAPS/
TexCAPS [Air Force/Texas Coronary Atherosclerosis Prevention
Study] Research Group. Primary prevention of acute coronary events
with lovastatin in men and women with average cholesterol levels.
JAMA 1998;279:1615–1622.
2. Scandinavian Simvastatin Survival Study Group. Randomized trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383–1389.
3. The Long-Term Intervention with Pravastatin in Ischaemic Disease
(LIPID) Study Group. Prevention of cardiovascular events and deaths
with pravastatin in patients with coronary heart disease and a broad
range of initial cholesterol levels. N Engl J Med 1998;339:1349 –1357.
4. Davidson MH. Controversy surrounding the safety of cerivastatin.
Expert Opin Drug Saf 2002;1:207–212.
5. US Food and Drug Administration. Summary Basis for Approval of
the Cerivastatin New Drug Application #020740. Washington, DC: US
Food and Drug Administration; May 24, 1999.
6. US Food and Drug Administration. Baycol Supplementary NDA (0.8
mg dose). FDA Medical Review. Cited July 19, 2000. [US Food and
Drug Administration Web site.] Available at: http://www.fda.gov/
cder/foi/nda/2000/20-740S008_Baycol_medr.pdf. Accessed August
30, 2005.
7. Staffa JA, Chang J, Green I. Cerivastatin and reports of fatal rhabdomyolysis. N Engl J Med 2002;346:539 –540.
8. US Food and Drug Administration. FDA Rejection of Citizen Petition.
Cited March 11, 2005. Docket No. 2004P-01131CP1. [US Food and Drug
Administration Web site.] Available at: http://www.fda.gov/cder/drug/
infopage/rosuvastatin/crestor_CP.pdf. Accessed August 30, 2005.
9. US Food and Drug Administration.1999 FDA/CDER resource pages.
[US Food and Drug Administration Web site.] Available at: http://
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
51C
www.fda.gov/cder/regulatory/applications/NDA.htm. Accessed August 30, 2005.
US Food and Drug Administration. Center for Drug Evaluation and Research
(CDER). [NDA Approval History Web site.] Available at: http://www.
accessdata.fda.gov/scripts/cder/drugsatfda/. Accessed August 30, 2005.
US Food and Drug Administration, Summary Basis for Approval of
the Rosuvastatin New Drug Application #021366. Washington, DC:
US Food and Drug Administration; July 2003.
US Food and Drug Administration, Summary Basis for Approval of
the Atorvastatin New Drug Application #020702. Washington, DC:
US Food and Drug Administration; December 17, 1996.
US Food and Drug Administration, Summary Basis for Approval of
the Fluvastatin New Drug Application #020261. Washington, DC: US
Food and Drug Administration; December 31, 1993.
US Food and Drug Administration. NDA for Lescol XL 80mg. [US
Food and Drug Administration Web site.] Available at: http://www.fda.
gov/cder/foi/nda/2000/21-192_Lescol.htm. Accessed August 30, 2005.
US Food and Drug Administration. Summary Basis for Approval of
the lovastatin New Drug Application #019643. Washington, DC: US
Food and Drug Administration; August 1987.
US Food and Drug Administration, Summary Basis for Approval of
the simvastatin New Drug Application #019766/S-028. Washington,
DC: US Food and Drug Administration; December 1991.
Crestor (rosuvastatin) [package insert]. Princeton, NJ: Bristol-Myers
Squibb; 2005.
Lescol/ Lescol XL (fluvastatin) [package insert]. East Hanover, NJ:
Novartis Pharmaceuticals/Reliant Pharmaceuticals; 2005.
Lipitor (atorvastatin) [package insert]. New York: Pfizer Inc; 2005.
Mevacor (lovastatin) [package insert]. West Point, PA: Merck & Co.;
2005.
Pravachol (pravastatin) [package insert]. Princeton, NJ: Bristol-Myers
Squibb; 2005.
Zocor (simvastatin) [package insert]. Whitehouse Station, NJ: Merck
& Co.; 2005.
Psaty BM, Furberg CD, Ray WA, Weiss NS. Potential for conflictof
interest in the evaluation of suspected adverse drug reactions: use of
cerivastatin and risk of rhabdomyolysis. JAMA 2004;292:2622–2631.
Pogson GW, Kindred LH, Carper BG. Rhabdomyolysis and renal
failure associated with cerivastatin-gemfibrozil combination therapy
[letter]. Am J Cardiol 1999;83:1146.
US Food and Drug Administration. Crestor [rosuvastatin] NDA. Medical Review. [US Food and Drug Administration Website.] Available
at: http://www.fda.gov/cder/foi/nda/2003/21-366_Crestor_Medr_P1.
pdf cited June 11, 03. Accessed August 30, 2005.
Chalasani N. Statins and hepatotoxicity: focus on patients with fatty
liver. Hepatology 2005;41:690 – 695.
Jacobson TA. Combination lipid lowering therapy with statins: safety
issues in the post-cerivastatin era. Expert Opin Drug Saf 2003;2:269 –286.
US Food and Drug Administration. Crestor Advisory Committee
Briefing Document. Cited June 11, 2003. [US Food and Drug Administration Web site.] Available at: http://www.fda.gov/ohrms/dockets/
ac/03/briefing/3968B1_02_A-FDA-Clinical%20Review.pdf. Accessed August 30, 2005.
Deslypere JP, Delanghe J, Vermeulen A. Proteinuria as complication
of simvastatin treatment [letter]. Lancet 1990;336:1453.
Statin Safety: A Systematic Review
Malcolm Law, MD, Alicja R. Rudnicka, PhD*
A systematic review of cohort studies, randomized trials, voluntary notifications to
national regulatory authorities, and published case reports was undertaken to assess
the incidence and characteristics of adverse effects in patients treated with 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors, or statins. For
statins other than cerivastatin, the incidence of rhabdomyolysis in 2 cohort studies
was 3.4 (1.6 to 6.5) per 100,000 person-years, an estimate supported by data from 20
randomized controlled trials. Case fatality was 10%. Incidence was about 10 times
greater when gemfibrozil was used in combination with statins. Incidence was higher
(4.2 per 100,000 person-years) with lovastatin, simvastatin, or atorvastatin (which are
oxidized by cytochrome P450 3A4 [CYP3A4], which is inhibited by many drugs) than
pravastatin or fluvastatin (which are not oxidized by CYP3A4). In persons taking
simvastatin, lovastatin, or atorvastatin, 60% of cases involved drugs known to inhibit
CYP3A4 (especially erythromycin and azole antifungals), and 19% involved fibrates,
principally gemfibrozil. The incidence of myopathy in patients treated with statins,
estimated from cohort studies supported by randomized trials, was 11 per 100,000
person-years. For liver disease, randomized trials reported fewer hepatobiliary disorders in patients allocated statins than in those allocated placebo. The notification
rate of liver failure to regulatory authorities was about 1 per million person-years of
statin use. Randomized trials show no excess of renal disease or proteinuria in
statin-allocated participants, and the decline in glomerular filtration rate was smaller
with statins than with placebo. Evidence from 4 cohort studies and case reports
suggests that statins cause peripheral neuropathy, but the attributable risk is small
(12 per 100,000 person-years). No change in cognitive function was found in randomized trials of statins in elderly patients. © 2006 Elsevier Inc. All rights reserved.
(Am J Cardiol 2006;97[suppl]:52C– 60C)
Concern over the safety of the 3-hydroxy-3-methylglutaryl
coenzyme A (HMG-CoA) reductase inhibitors, or statins,
followed the worldwide withdrawal in 2001 of cerivastatin,
a drug that had been thought to be relatively free of serious
adverse effects over the 4 years it was marketed.1 Further
concern followed documentation of the hazards of rosuvastatin after regulatory approval by the US Food and Drug
Administration (FDA) and marketing.2 These episodes
prompted the present article, which is a systematic review of
published safety data on all statins.
Methods
Data on safety were gathered from the following 4 sources:
(1) Cohort studies, in which persons taking and not taking
statins are identified and followed prospectively for disease
occurrence. Such studies have the advantage of large size
Wolfson Institute of Preventive Medicine, Barts and The London
School of Medicine, London, United Kingdom.
Reprints are not available.
*Address for correspondence: Malcolm Law, MD, Wolfson Institute of
Preventive Medicine, Barts and The London School of Medicine, Charterhouse Square, London EC1M 6BQ, United Kingdom.
E-mail address: m.r.law@qmul.ac.uk.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.010
but the disadvantage (because they are derived from electronic healthcare databases) of uncertainty that all disease
events have been recorded. (2) Randomized placebo-controlled trials of statins, typically 2–5 years in duration and
conducted primarily to determine the reduction in cardiovascular disease events.3 These trials have the advantage
that they avoid bias and are unlikely to miss disease events
(because participants are followed closely), but they have
the disadvantage of relatively small size. (3) Voluntary
notifications to national regulatory authorities of adverse
events occurring in patients taking statins. These sources
have the advantage of recording information on very large
numbers of people with serious adverse effects, but they
have the disadvantage of underestimating incidence because
notification is incomplete. (4) Published individual case
reports (n ⫽ 152) were identified. Examined collectively,
these provide evidence on the prevalence of various cofactors when patients taking statins develop adverse effects;
however, they are not necessarily representative of all cases.
These 4 sources together generated sufficient data to
allow reasonably definitive conclusions on the safety of
lovastatin, simvastatin, atorvastatin, fluvastatin, and pravastatin. Qualified conclusions could be made regarding rosuvastatin (for which there currently are relatively few published data from these sources).4
www.AJConline.org
Law and Rudnicka/Statin Safety: Evidence from Published Literature
53C
Table 1
Causal factors in 25 patients with rhabdomyolysis: results from 25 million person-years
of follow-up (1990 –1999) in the UK General Practice Research Database (GPRD).
Drug overdose*
Alcohol excess
Infection
Trauma, exercise
Epilepsy (convulsions)
Genetic predisposition
Hypothermia
Lipid-lowering drugs†
No recognized cause
7
2
6
4
2
1
1
1
1
* Opiates, amphetamines, or antipsychotics.
†
The 1 case was in a man who had taken a statin and a fibrate concurrently for 3 years.
Adapted from Pharmacotherapy.7
Studies dating from 1980 to 2005 were identified on the
Medline database, using the following Medical Subject
Heading (MeSH) terms: muscular diseases, liver diseases,
kidney diseases, peripheral nervous system diseases, or
polyneuropathies (all /chemically induced or /epidemiology); hydroxymethylglutaryl-CoA reductase inhibitors, antilipemic agents, or the names of individual statins (all
/adverse effects or /toxicity). The citations of each article
identified and of review articles were also examined.
Muscle Diseases
Quantifying the incidence of rhabdomyolysis: The defining features of rhabdomyolysis in epidemiologic studies
were physician diagnosis, hospital admission,5 muscle
symptoms, and a serum concentration of creatine kinase
(CK) ⬎10,000 U/L.6 Rhabdomyolysis may occur at any
time an individual is taking a statin (ie, cases do not concentrate in a short period after the initiation of therapy5–7).
Incidence (proportion developing illness per year) rather
than prevalence (proportion developing illness without
specifying time) is therefore the appropriate measure.
COHORT STUDIES. A total of 3 cohorts were available.5,7,8
Of these cohorts, 2 were useful in deriving estimates of
incidence; both are research databases based on electronic
health records of millions of people for whom drug prescription and disease occurrance are recorded. Until recently
the common disease coding systems lacked specific codes
for rhabdomyolysis and other muscle diseases. This problem was circumvented by using large numbers of nonspecific search terms and examining many individual case
records. The first cohort is the UK General Practice Research Database (GPRD), which contains computerized
medical information entered by family practitioners in the
United Kingdom since 1988. It includes 2.5 million persons
aged 20 –75 years with information in the database for the
decade 1990 –1999 (25 million person-years of observation).7 In all, 25 persons with a first-time diagnosis of
rhabdomyolysis of any cause were identified over the decade, an incidence of 1 per 1 million person-years. Table 1
lists the recognized causal factors identified in these 25
cases.7 Only 1 of the 25 cases occurred among the 52,000
persons in the cohort who took lipid-lowering drugs, and
only 1 case had no recognized cause.
The second cohort, from the United States, pools data
from 11 separate health maintenance organizations (HMOs)
and similar organizations; a cohort of 252,000 individuals
taking statins or fibrates (which also cause rhabdomyolysis)
was identified.5 Table 2 summarizes data from this cohort
(based on hospital admissions), and the more limited data
from the UK GPRD,7 on the incidence and mortality of
rhabdomyolysis in persons taking statins, with data stratified separately for use of cerivastatin, for all statins other
than cerivastatin, and for the statin taken alone or together
with gemfibrozil. The rarity of rhabdomyolysis in persons
not taking lipid-lowering drugs (Table 1) means that it is
reasonable to attribute all cases of rhabdomyolysis observed
in persons taking statins or fibrates to those drugs.
For persons taking statins other than cerivastatin, the
cohort study data indicate a low incidence of rhabdomyolysis of 3.4 per 100,000 person-years, with a narrow 95%
confidence interval (CI) of 1.6 – 6.5 per 100,000 personyears. The estimated mortality is 0.3 per 100,000 personyears (10% case fatality; see Table 2). These rates were
about 10 times higher for cerivastatin and for statins other
than cerivastatin when taken with gemfibrozil. For cerivastatin taken with gemfibrozil, the incidence of rhabdomyolysis was about 2,000 times higher, an absolute annual incidence of about 10%. The reduction in low-density
lipoprotein (LDL) cholesterol is scarcely greater with the
cerivastatin-gemfibrozil combination than with atorvastatin,
so the 3,000-fold risk difference indicates that cholesterol
reduction is not the cause of the rhabdomyolysis.
The third cohort, from Japan, of 51,000 persons taking
simvastatin with 175,000 person-years of follow-up,8 recorded no cases of rhabdomyolysis (6 cases would have
been expected based on the above-mentioned rate of 3.4 per
100,000). However, the dose of simvastatin (5 mg/day) was
lower than generally used in Western countries. This cohort
suggests low risk at low doses.
54C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 2
Cases of rhabdomyolysis occurring in cohort studies of patients taking statins or fibrates, or both, with estimates of incidence and mortality
Monotherapy
Statins Other than
Cerivastatin
Cerivastatin
Cases
5
US HMOs
UK GPRD7
Both
Incidence per 100,000
person-yrs†
4
0
4
Combined Therapy
Person-Years
Cases
Person-Years
7,486
1,140
8,626
9
0
9
207,523
54,250
261,773
Cerivastatin plus
Gemfibrozil*
Gemfibrozil*
Cases
3
0
3
Person-Years
Cases
Person-Years
8,102
2,700
10,802
6
58
6
58
Statins Other than
Cerivastatin Plus
Gemfibrozil*
Cases
1
0
1
Person-Years
2,474
400
2,874
Rate
(95% CI)
Rate
(95% CI)
Rate
(95% CI)
Rate
(95% CI)
Rate
(95% CI)
46
(13–120)
3.4
(1.6–6.5)
28
(6–81)
10,300
(3,800–22,500)
35
(1–194)
CI ⫽ confidence interval; UK GPRD ⫽ United Kingdom General Practice Research Database; US HMOs ⫽ United States health maintenance
organizations.
* For fibrates other than gemfibrozil there were no cases in 28,589 person-years; incidence 0 (95% CI, 0 to 13) per 100,000 person-years.
†
Mortality is 10% of incidence, based on an average case fatality of 10% (96 of 935 cases from 3 sources, 2 published case series,5,6 and examination
of the 152 published case reports).
Adapted from JAMA5 and Pharmacotherapy.7
Table 39 –13 shows the numbers of
cases of rhabdomyolysis reported in the randomized trials of
statins and cardiovascular disease (CVD) events, recording
in total about 180,000 person-years of follow-up in both
statin and placebo groups.3 None of the trials tested cerivastatin or a statin-fibrate combination. The randomized trials have the advantage that no cases of rhabdomyolysis are
likely to have been missed on follow-up, but their disadvantage is poorly specified case definition. Rhabdomyolysis
was defined using a low threshold of CK (2,000 U/L; 10,000
U/L has been used elsewhere6) and no clinical criteria were
specified. In total there were 8 cases reported in the statin
groups and 5 in the placebo groups, but from the estimated
incidence of 1 per 1 million person-years in the general
population (Table 1) no cases would have been expected in
the placebo group; therefore, rhabdomyolysis was probably
overdiagnosed in the trials. The difference in incidence
between the statin-treated and placebo groups, however, is
likely to provide a valid estimate of the incidence of rhabdomyolysis attributable to statins. This difference was 1.6
(95% CI, ⫺2.4 to 5.5) per 100,000 person-years.
This estimate from randomized trials of statins other than
cerivastatin is similar to the corresponding estimate from
the cohort studies of 3.4 (95% CI, 1.6 – 6.5) per 100,000
person-years (Table 2). Although the trial result is not statistically significant, it provides reassurance against a background of concern that the cohort studies may have failed to
detect all cases. The randomized trials confirm that the
estimate from the cohort studies is unlikely to be too low.
RANDOMIZED TRIALS.
VOLUNTARY
NOTIFICATIONS
TO
REGULATORY
AUTHORITIES. Table 4 shows data from the US FDA Ad-
verse Effects Reporting System (AERS).6,14,15 The estimates of incidence are substantially lower than those from
the cohort studies (the estimate for cerivastatin is approximately 50% that in Table 2; that for the other statins is
approximately 20%), confirming the expected undernotification in voluntary systems. The notification rates of rhabdomyolysis to the UK Medicines Control Agency16 showed
a comparable degree of undernotification. The data in Table
4, however, confirm the cohort study result that risk is far
greater when a statin is taken with gemfibrozil than when it
is taken alone. The FDA AERS data show higher risk at
higher doses of statins,15 as do the cohort study data.5
Incidence of rhabdomyolysis for different statins: The
incidence of rhabdomyolysis may be higher among persons
taking lovastatin, simvastatin, and atorvastatin (because they
are metabolized by cytochrome P450 3A4 [CYP3A4]-mediated oxidation, which is inhibited by several commonly used
drugs17–19), than fluvastatin (oxidized by CYP2A9) or pravastatin (not oxidized by the CYP450 system because it is water
soluble). The notification rate of rhabdomyolysis to the FDA
AERS was about 4 times higher for monotherapy with lovastatin, simvastatin, and atorvastatin (mean rate, 0.73; 95% CI,
0.64 – 0.82 per 1 million prescriptions [264 cases]) than for
monotherapy with pravastatin and fluvastatin (mean rate, 0.15;
95% CI, 0.09 – 0.24 per 1 million prescriptions [18 cases]), p
⬍0.001.6 A similar difference, also statistically significant, was
seen in notifications to the UK Medicines Control Agency.10 In
the randomized trials no cases of rhabdomyolysis occurred in
the treated or placebo groups in trials of pravastatin or fluvastatin (Table 3). In the cohort studies (Table 2) none of the
cases of rhabdomyolysis occurred in persons taking fluvastatin
or pravastatin. The mean incidence of rhabdomyolysis among
persons taking lovastatin, simvastatin, or atorvastatin in the 2
cohort studies was 4.2 (95% CI, 1.9 – 8.0) per 100,000 personyears (20% higher than the rates for all statins other than
cerivastatin shown in Table 2). This difference between statins
in the cohort study data was not statistically significant, however, because relatively few person-years of follow-up were
10,269
6,582
5,168
4,512
3,304
3,302
2,891
2,221
2,081
1,538
844
800
530
460
450
409
224
203
193
187
157
HPS9,10
EXCEL11
ASCOT
LIPID
AFCAPS/TexCAPS
WOSCOPS
PROSPER12
4S13
CARE
MIRACL
LIPS
GREACE
PMSG
ACAPS
REGRESS
FLARE
KAPS
LRT
MAAS
Riegger et al
LCAS
All trials
10,267
7
1,663
5,137
4,502
3,301
3,293
2,913
2,223
2,078
1,548
833
800
532
459
434
425
223
201
188
178
164
Pl
L
A
P
L
P
P
S
P
A
F
A
P
L
P
F
P
L
S
F
F
S
Statin
0.9
3.3
6.1
5.2
4.9
3.2
5.4
5.0
0.3
3.9
3.0
0.5
3.0
2.0
0.8
3.0
0.5
4.0
2.5
0.9
5.3
Duration (yr)
1.6
(⫺2.4 to 5.5)
4.4
0
1
0
1
0
0
1
0
0
0
0
0
0
0
0
—
—
0
0
0
5
Rx
†
2.8
0
0
0
2
0
0
0
0
0
0
0
0
0
0
0
—
—
0
0
0
3
Pl
Rhabdomyolysis
5
(⫺17 to 27)
92
10
9
19
32
8
4
—
—
—
3
—
0
3
0
—
—
—
—
8
11
20
36
11
0
—
—
—
2
—
1
7
0
—
—
—
—
97
0
50
Pl
‡
5
49
Rx
Myopathy
Pl
4,960
—
—
83
—
64
—
—
—
—
—
—
—
—
—
—
—
—
—
125
3,359
190
(⫺38 to 410)
5,150
—
—
77
—
71
—
—
—
—
—
—
—
—
—
—
—
—
—
512
3,330
Rx
Minor Muscle
Pain§
60
—
21
1
0
1
7
—
3
—
0
—
0
0
—
0
—
1
2
7
—
Pl
23
(⫺4 to 50)
83
—
21
3
0
6
12
—
0
—
0
—
0
0
—
3
—
0
1
17
—
Rx
Single
Measure
0
—
0
—
—
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
Rx
0
0
—
0
—
—
0
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
Pl
2 Consecutive
Measures
200
86
—
12
1
32
73
9
—
—
1
6
1
3
3
1
0
0
—
15
32
Pl
100
(64 to 140)
300
95
—
16
1
46
66
38
—
—
6
6
0
7
4
3
0
0
—
95
43
Rx
Single
Measure
70
(50 to 90)
110
—
18
—
—
14
—
—
10
—
0
—
—
0
—
—
—
—
2
45
9
Rx
40
—
11
—
—
12
—
—
3
—
0
—
—
0
—
—
—
—
0
2
4
Pl
2 Consecutive
Measures
Elevated ALT¶
A ⫽ atorvastatin; ACAPS ⫽ Asymptomatic Carotid Artery Progression Study; AFCAPS/TexCAPS ⫽ Air Force/Texas Coronary Atherosclerosis Prevention Study; ASCOT ⫽ Anglo-Scandinavian Cardiac Outcomes Trial;
CARE ⫽ Cholesterol and Recurrent Events Trial; CI ⫽ confidence interval; EXCEL ⫽ Expanded Clinical Evaluation of Lovastatin; F ⫽ fluvastatin; FLARE ⫽ Fluvastatin Angiographic Restenosis Trial; 4S ⫽ Scandinavian
Simvastatin Survival Study; GREACE ⫽ Greek Atorvastatin and Coronary Heart Disease Evaluation Study; HPS ⫽ Heart Protection Study; KAPS ⫽ Kuopio Atherosclerosis Prevention Study; L ⫽ lovastatin; LCAS ⫽
Lipoprotein Coronary Atherosclerosis Study; LIPID ⫽ Long-Term Intervention with Pravastatin in Ischaemic Disease study; LIPS ⫽ Lescol Intervention Prevention Study; LRT ⫽ Lovastatin Restenosis Study; MAAS ⫽
Multicentre Antiatheroma Study; MIRACL ⫽ Myocardial Ischemia Reduction with Aggressive Cholesterol Lowering Trial; P ⫽ pravastatin; Pl ⫽ placebo; PMSG ⫽ Pravastatin Multinational Study Group; PROSPER ⫽
Prospective Study of Pravastatin in the Elderly at Risk; REGRESS ⫽ Regression Growth Evaluation Statin Study; Rx ⫽ treated group; S ⫽ simvastatin; ULN ⫽ upper limit of normal; WOSCOPS ⫽ West of Scotland Coronary
Prevention Study.
* Citations to 4 trials are published here; the others are cited in Law et al. BMJ 2003;326:1423–1430.3
†
Definition poorly specified, except CK ⱖ10⫻ ULN (or ⱖ2,000 U/L).
‡
Muscle pain, tenderness, or weakness sufficient to consult a physician or to stop taking prescribed tablets.
§
Muscle pain, tenderness, or weakness elicited on questionnaire but insufficient to consult a physician or to stop taking prescribed tablets.
储
Elevated CK defined as ⱖ10⫻ ULN (or ⱖ2,000 U/L).
¶
Elevated ALT defined as ⱖ3⫻ ULN (or ⱖ120 U/L).
Incidence per
100,000
person-yrs
Treated minus
placebo, per
100,000
person-yrs
(95% CI)
Rx
Study*
Participants (N)
Elevated CK储
Trial Participants with Disorder (n)
Table 3
Participants in randomized trials who developed muscle disorders, elevated creatine kinase (CK), or elevated alanine aminotransferase (ALT)
Law and Rudnicka/Statin Safety: Evidence from Published Literature
55C
56C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 4
Data from the US Food and Drug Administration (FDA) Adverse Events Reporting System (AERS) on numbers of notifications of rhabdomyolysis,*
with estimates of incidence per 100,000 years derived from numbers of prescriptions
Monotherapy
Combined with Gemfibrozil
Statins Other than
Cerivastatin
Cerivastatin
Statins Other than
Cerivastatin
Cerivastatin
Cases
Prescriptions
(millions)
Cases
Prescriptions
(millions)
Cases
Prescriptions
(millions)
Cases
Prescriptions
(millions)
200
11
282
484
279
0.022
105
6.0
Rate
(95% CI)
Rate per 1 million
prescriptions
18
(15–21)
Estimated
incidence per
100,000 personyrs†
21
(19–25)
Rate
(95% CI)
Rate
(95% CI)
Rate
(95% CI)
0.58
(0.52–0.66)
13,000
(1,000–14,000)
17
(14–21)
0.70
(0.62–0.79)
15,000
(13,000–17,000)
21
(17–25)
CI ⫽ confidence interval.
* As defined in Pharmacoepidemiol Drug Saf.6
†
Based on the estimate that a prescription is for 1 month on average.14
recorded for fluvastatin and pravastatin.
These observations might be attributable to a higher rate of
undernotification for pravastatin and fluvastatin in both voluntary reporting systems and to chance in the trials and cohort
studies, but the likely interpretation is that the incidence of
rhabdomyolysis in persons taking lovastatin, simvastatin, or
atorvastatin is indeed about 4 per 100,000 person-years; for
fluvastatin and pravastatin it is about 4 times lower (around 1
per 100,000 person-years). This safety advantage does not
mean, however, that fluvastatin or pravastatin should be preferred in general clinical use, because they are less effective at
lowering LDL cholesterol.3
For rosuvastatin there are no data on the incidence of
rhabdomyolysis from cohort studies or randomized trials.
Uncontrolled trials and postmarketing surveillance indicated that the highest dosage (80 mg/day), now withdrawn,
caused rhabdomyolysis significantly more frequently than
lovastatin, simvastatin, and atorvastatin, but that with lower
(ⱕ40 mg/day) dosages the incidence is similar.4
Drugs that inhibit the CYP3A4 enzyme system: Notifications to the FDA AERS18 and analysis of the 152
published case reports both provide data on the proportion
of cases of rhabdomyolysis in which drugs known to inhibit
the CYP3A4 enzyme system were taken with statins. Such
drugs include diltiazem and other nondihydropyridine calcium-channel blockers, ritonavir and other protease inhibitors, cyclosporine, erythromycin and other macrolides,
azole antifungals, and other drugs.17 As expected, these
drugs were taken more frequently (60%) when the statin
was simvastatin, lovastatin, or atorvastatin (oxidized by
CYP3A4) than when it was pravastatin or fluvastatin (7%).
Table 5 summarizes these data. Grapefruit juice also inhibits
the CYP3A4 system, but these sources provide no data on
its consumption in people who develop rhabdomyolysis;
health databases generally record data on drugs but rarely
information on diet.
The interaction between statins and fibrates: Both
fibrates and statins cause rhabdomyolysis when given as
monotherapy (Table 2), probably through separate mechanisms. However, the risk of rhabdomyolysis with statins
and fibrates in combination is substantially higher than
the sum of the risks associated with each class of drug
taken as monotherapy (Table 2). Among fibrates, the
incidence of rhabdomyolysis is higher with gemfibrozil
than with other fibrates1 (see footnote to Table 2). For
statins other than cerivastatin combined with fibrates, the
FDA AERS data showed a 15-fold higher incidence (p
⬍0.001) when the fibrate was gemfibrozil than when it
was fenofibrate.19 This is consistent with experimental
studies showing that in persons taking statins the plasma
concentration of the statin increases when it is taken with
gemfibrozil but not when it is taken with other fibrates.20
For gemfibrozil combined with statins the incidence of
rhabdomyolysis was 500 times greater when the statin
was cerivastatin than when it was another statin.19 Gemfibrozil (and other fibrates) are substrates but not inhibitors of the CYP3A4 system, but gemfibrozil inhibits the
glucuronidation of statins.21 Experimentally, gemfibrozil
increases the plasma concentration of cerivastatin about
5-fold,22 which may be caused by this inhibition of glucuronidation, but it is difficult to see that this effect alone
could account for the extraordinarily high risk of rhabdomyolysis with cerivastatin and gemfibrozil combined
(about 10% per year in the cohort studies). It is possible
that ⬎1 mechanism may be involved. The reasons for the
greater toxicity of gemfibrozil than other fibrates and
cerivastatin than other statins are poorly understood. An
estimated 19% of cases of statin-related rhabdomyolysis
Law and Rudnicka/Statin Safety: Evidence from Published Literature
57C
Table 5
Cases of rhabdomyolysis in patients taking simvastatin, lovastatin, or atorvastatin in which drugs
known to inhibit cytochrome P450 3A4 (CYP3A4) or fibrates were also taken, from the US Food
and Drug Administration (FDA) Adverse Effects Reporting System (AERS)18 and analysis of
individual published case reports
Drugs known to inhibit CYP3A4
Gemfibrozil
Other fibrates
FDA AERS Data
(n ⫽ 328), n (%)
Published Case Reports
(n ⫽ 102), n (%)
Both Combined
(n ⫽ 430), n (%)
195 (59)
65 (64)
30 (29)
2 (2)
260 (60)
48 (15)
occurred in patients also taking fibrates, principally gemfibrozil (Table 5).
Monitoring serum CK concentration: Table 3 shows
data from the 13 randomized trials in which serum CK was
measured, generally at regular intervals throughout the trial.
CK was found to be ⱖ10 times the upper limit of normal
(ULN), or ⱖ2,000 U/L, on a single measure in 83 statinallocated participants per 100,000 person-years and 60 placebo-allocated participants per 100,000 person-years; the
difference was not statistically significant. In 2 trials together recording 30,000 person-years’ observation of statintreated patients, none had CK elevated on 2 consecutive
measures. Hence, elevations of CK observed in persons
taking statins would likely have occurred had the person not
been taking a statin; and these elevations disappear on
repeat measurement.
Myopathy: Myopathy is defined as diffuse muscle
symptoms (pain, tenderness, weakness) with elevated CK,16
sufficient to consult a physician but insufficient to warrant
hospital admission. Statins and fibrates cause myopathy as
well as rhabdomyolysis; it is probable that the 2 terms
describe less severe and more severe cases of the same
muscle disorder, rhabdomyolysis being characterized by
myoglobinuria and other features. Data on myopathy are
available from 1 cohort study and from randomized trials.
COHORT STUDY. The incidence of myopathy in persons
taking statins or fibrates was estimated from the UK General
Practice Research Database.23 There were 2 cases in 17,056
person-years of follow-up in patients taking statins other
than cerivastatin, 5 in 9,136 person-years in people taking
fibrates, and 4 in 357,186 person-years among people taking
no lipid-lowering drugs.23 Adjusting for incidence in the
untreated group, these rates are equivalent to a mean incidence of 11 (95% CI, 4 –27) per 100,000 person-years in
people taking statins other than cerivastatin, and a mean
incidence of 54 (95% CI, 6 –102) in persons taking fibrates.
There was no estimate for cerivastatin because of scanty
data (it was little used in the United Kingdom). Estimates
from a large US HMO have also been published, but the
definition of myopathy was based on elevated CK alone
with few details on clinical symptoms available; moreover
there were no data on duration of use (prevalence, not
80 (19)
incidence, was estimated).24
RANDOMIZED TRIALS. Table 3 shows data from the randomized trials on the numbers of participants with muscle
symptoms generally of sufficient severity to either consult a
physician or to stop taking the allocated tablets (data on CK
were not always available). As with rhabdomyolysis there
was poorly specified case definition and overdiagnosis in
the trials, but cases of myopathy are not likely to have been
missed, and the difference in incidence between the statintreated and placebo groups is likely to provide a valid
estimate of the incidence of myopathy attributable to statins.
This mean difference was 5 (95% CI, ⫺17 to 27) per
100,000 person-years. As with rhabdomyolysis, while the
trial result is not statistically significant it supports the
estimate from the cohort study of a mean 11 (95% CI, 4 –27)
per 100,000 person-years and provides reassurance that it is
unlikely that the incidence of myopathy was substantially
underestimated in the cohort study.
Four randomized trials also recorded the numbers of
participants with minor degrees of muscle pain, elicited on
a questionnaire and insufficient to consult a physician or to
stop taking allocated tablets. The incidence of such pain was
similar in treated and placebo groups on average (about 5%
per year). The difference between the 2 groups of 190 (95%
CI, ⫺38 to 410) per 100,000 person-years (or about 0.2%
per year) was not statistically significant. Hence minor muscle pain attributable to statins, if it occurs at all, is
uncommon.
Although the importance of fibrates and of drugs that
inhibit the CYP3A4 enzyme system has not been investigated with respect to myopathy, it is probable that myopathy, like rhabdomyolysis, is more common with high serum concentration of statins (which these drugs cause).
Thus, it seems likely that measures to avoid coprescribing
statins with these other drugs would prevent myopathy as
well as rhabdomyolysis.
Liver Diseases
Drugs in general are an important cause of liver disease, and
drug-induced hepatotoxicity may mimic almost any type of
hepatobiliary disease from fulminant liver failure to chronic
liver disease with cirrhosis.25 Despite case reports of liver
58C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
disease in persons taking statins, however,25 the evidence
indicates that liver disease attributable to statins is rare. In
pooled data from 3 randomized trials of pravastatin, recording 45,000 person-years follow-up of both statin-treated and
placebo groups, both gall bladder disorders (186 vs 208
[1.9% vs 2.1%]) and other hepatobiliary disorders (69 vs 89
[0.7% vs 0.9%]) were less common in patients treated with
statin than in participants who received placebo.26 In the
FDA AERS, 38 cases of liver failure in persons taking
statins were reported up to the end of 1999 (8 cases where
other causes of liver failure were also present and 30 cases
with no other recognized causes).27 This is equivalent to a
notification rate of about 0.1 per 100,000 person-years of
use. Another reporting system has independently generated
the same estimate.28 With undernotification similar to that
of rhabdomyolysis the true rate would be about 0.5 per
100,000 person-years of statin use, an extremely low incidence that is little or no greater than the risk of liver failure
in the general population among persons not taking
statins.28
Monitoring serum concentration of liver enzymes in
persons taking statins: Table 3 shows data from randomized trials on alanine aminotransferase (ALT). Results were
similar for aspartate aminotransferase. Elevations of ALT
(defined as ⱖ 3 times the ULN, or ⱖ120 U/L) were found
in 300 statin-allocated and 200 placebo-allocated participants per 100,000 person-years. There was statistically significant heterogeneity across trials (p ⫽ 0.002), reflecting
the fact that trials measuring liver enzymes more frequently
found elevated values. Elevated ALT on 2 consecutive
measures was found in 110 statin-allocated and 40 placeboallocated participants per 100,000 person-years; ALT elevations were observed more frequently at higher doses of
drug.11,29 The incidence of liver disease (see above) is 0.5
per 100,000 person-years. Hence in 100,000 person-years of
statin use, denying 300 persons with elevated ALT the
benefit of a statin (or 110 persons if repeat measures were
used) would prevent liver disease in ⬍1 person.
Kidney Diseases
Despite evidence of renal disease and proteinuria in persons
taking the highest (80 mg/day) dosage of rosuvastatin (now
withdrawn),2,4 there is no evidence that lower doses of
rosuvastatin or other statins cause renal disease. Combined
data from 3 trials of pravastatin showed that renal failure or
other renal disease designating a serious adverse event occurred in 48 (0.5%) participants allocated pravastatin and 78
(0.8%) allocated placebo.26 None of the randomized trials
reported renal disease or proteinuria occurring significantly
more frequently in patients allocated statins than in participants given placebo. In a meta-analysis of 13 trials of
lipid-lowering drugs in patients with renal disease (statins in
10 trials, other drugs in 3 trials) there was a lower rate of
decline in glomerular filtration rate in treated participants
than in controls (p ⫽ 0.008)30; the difference in favor of
treatment was equivalent to about 3% of baseline glomerular filtration rate per year. In the Assessment of Lescol
[fluvastatin; Novartis, East Hanover, NJ] in Renal Transplantation (ALERT) trial of fluvastatin in renal transplant
recipients, the incidence of either graft loss or doubling of
serum creatinine did not significantly differ between participants allocated fluvastatin or placebo (relative risk [RR],
1.10; CI, 0.89 –1.36).31 There is no indication that any statin
at any currently marketed dose causes renal disease.
Neurologic Diseases
Hemorrhagic stroke: COHORT STUDIES. In an analysis
of 9 cohort studies of serum cholesterol and stroke that
distinguished thromboembolic and hemorrhagic strokes (using computed tomography or postmortem findings), a lower
LDL cholesterol level was associated with a higher risk of
hemorrhagic stroke.3 The RR was 1.19 (95% CI, 1.10 –1.29;
p ⬍0.001) for a 1.0 mmol/L decrease in serum LDL cholesterol concentration. There was, however, a lower risk of
thromboembolic stroke (RR, 0.85; 95% CI, 0.79 – 0.94).3
The 2 opposing effects resulted in no material association
between serum cholesterol and all stroke in cohort studies in
which hemorrhagic and thromboembolic strokes were not
distinguished.32
RANDOMIZED TRIALS. The randomized trials of serum
cholesterol reduction did not show an increase in hemorrhagic stroke in treated patients.3 Importantly, however, the
number of hemorrhagic strokes recorded in the trials was
small: 149 hemorrhagic strokes, compared with 1,204
thromboembolic strokes and 1,966 strokes of undetermined
subtype. Accordingly the 95% CI on the risk estimate for
hemorrhagic stroke was wide, from a 35% reduction to a
47% increase for a 1.0 mmol/L decrease in LDL cholesterol.3 The trial data are therefore relatively uninformative
and do not exclude the 19% increase in hemorrhagic stroke
shown in the cohort studies.
The interpretation of the increase in hemorrhagic stroke
shown in the cohort studies is uncertain. The evidence is
insufficient to attribute it to cause and effect, and no mechanism is apparent, nor can it readily be ascribed to confounding or bias. This uncertainty should not affect the use
of statins to prevent CVD because the possible excess of
hemorrhagic stroke is greatly outweighed by the protective
effect against coronary artery disease (CAD) and thromboembolic stroke.3 However, patients who have had a hemorrhagic stroke should not be given cholesterol-lowering
drugs.
Peripheral neuropathy: A total of 16 published case reports of peripheral neuropathy in patients taking statins have
been identified.33 The symptoms generally developed 1–2
months after the start of therapy, and usually resolved after
discontinuation of the statin. There have been 4 cohort studies
Law and Rudnicka/Statin Safety: Evidence from Published Literature
59C
Figure 1. Statins and peripheral neuropathy shown as odds ratios (95% confidence interval) estimated from 4 cohort studies,34 –37 with summary estimate,
and from a randomized trial.9
of peripheral neuropathy in persons taking or not taking
statins34 –37; all 4 studies were based on large electronic healthcare databases (in 3 studies the analysis was of nested casecontrol design); Figure 1 summarizes their results.34 –37 There
was heterogeneity between them (␹23 ⫽ 10, p ⫽ 0.02), based
on 1 study with a more extreme result than the others. The
summary odds ratio (random effects analysis) is 1.8 (95% CI,
1.1–3.0; p ⬍0.001). Peripheral neuropathy was recorded in 1
large trial with 11 cases in patients treated with statin and 8
cases in participants given placebo.9 This result is also shown
in Figure 1; the 95% CI is wide and consistent with both the
association shown in the cohort studies and with no association, so the trial is not discriminatory.
It is probable that the association in the cohort studies is
cause and effect. It is difficult to explain the association
through bias or confounding (for example, patients with
diabetes mellitus may be more likely to develop peripheral
neuropathy and to take statins, but the association was little
different on adjustment for diabetes35). In 1 case report 4
different statins were introduced and discontinued in succession; peripheral neuropathy appeared with each drug in
turn and resolved when the drug was discontinued.38 Also,
there are plausible mechanisms for cause and effect.33 Even
if statins do cause peripheral neuropathy however, the attributable risk is small. The excess risk of 80% derived from
the meta-analysis of the 4 cohort studies in Figure 1, applied
to the risk in placebo participants in the large trial,9 indicate
an incidence of 12 per 100,000 person-years, or a prevalence of 60 per 100,000 persons, attributable to statins. This
minor hazard is no reason to limit the use of statins, except
that they should be discontinued if peripheral neuropathy
develops.
Cognitive function: Some case reports have suggested
that statins accelerate decline in cognitive function. Cognitive status was measured in the Heart Protection Study
(HPS) (the largest randomized trial) after 5 years of taking
simvastatin (n ⫽ 10,269) or placebo (n ⫽ 10,267).9,10 Cognitive impairment was detected in similar proportions of
participants allocated simvastatin (23.7%) and placebo
(24.2%), as was dementia (0.3% and 0.3%). In the Prospec-
tive Study of the Elderly at Risk (PROSPER) trial in participants aged 70 – 82 years, cognitive function declined at
the same rate in 2,891 patients allocated pravastatin and
2,913 persons allocated placebo.12 These measurements in
large numbers of participants in randomized trials establish
beyond doubt that statins cause no perceptible decline in
cognitive function.
Conclusion
Despite the high risk with cerivastatin, the incidence of
rhabdomyolysis is low in patients taking simvastatin, lovastatin, atorvastatin, pravastatin, or fluvastatin— estimated as
3 per 100,000 person-years and unlikely to exceed 7 per
100,000 person-years. Myopathy attributable to these statins
is also rare (11 per 100,000 person-years). Most muscle
symptoms in patients taking statins are not attributable to
the statins. Rare as muscle disease caused by statins is,
many instances could be prevented. Drugs that inhibit
CYP3A4 are taken by about 60% of persons using simvastatin, lovastatin, or atorvastatin who develop rhabdomyolysis; thus using statins in low dose in these circumstances
(or suspending them while a patient takes a course of erythromycin or other macrolide) would be expected to prevent
many cases. Alternatively, pravastatin could be used in
these situations. Coprescription of a statin with gemfibrozil
is also a preventable cause of many cases of rhabdomyolysis.
Liver disease attributable to statins, if it occurs at all, is
rare. There is no indication that statins cause renal disease or
cognitive decline. Statins are probably a cause of peripheral
neuropathy, but the attributable risk is small (12 per 100,000
person-years). The evidence for hemorrhagic stroke is uncertain (cohort studies show an association and randomized
trials are uninformative because the CI on the summary
estimate is too wide); the possible risk, however, is greatly
outweighed by the protective effect against thromboembolic
stroke and CAD. By any standard, statins are remarkably
safe drugs.
60C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
1. Omar MA, Wilson J, Cox TS. Rhabdomyolysis and HMG-CoA reductase inhibitors. Ann Pharmacother 2001;35:1096 –1107.
2. Wolfe SM. Dangers of rosuvastatin identified before and after FDA
approval. Lancet 2004;363:2189 –2190.
3. Law MR, Wald NJ, Rudnicka AR. Quantifying effect of statins on low
density lipoprotein cholesterol, ischaemic heart disease, and stroke:
systematic review and meta-analysis. BMJ 2003;326:1423–1430.
4. Davidson MH. Rosuvastatin safety: lessons from the FDA review and
post-approval surveillance. Expert Opin Drug Saf 2004;3:547–557.
5. Graham DJ, Staffa JA, Shatin D, Andrade SE, Schech SD, La Grenade
L, Gurwitz JH, Chan KA, Goodman MJ, Platt R. Incidence of hospitalized rhabdomyolysis in patients treated with lipid-lowering drugs.
JAMA 2004;292:2585–2590.
6. Chang JT, Staffa JA, Parks M, Green L. Rhabdomyolysis with HMGCoA reductase inhibitors and gemfibrozil combination therapy. Pharmacoepidemiol Drug Saf 2004;13:417– 426.
7. Black C, Jick H. Etiology and frequency of rhabdomyolysis. Pharmacotherapy 2002;22:1524 –1526.
8. Matsuzawa Y, Kita T, Mabuchi H, Matsuzaki M, Nakaya N, Oikawa
S, Saito Y, Sasaki J, Shimamoto K, Itakura H. Sustained reduction of
serum cholesterol in low-dose 6-year simvastatin treatment with minimum side effects in 51,321 Japanese hypercholesterolemic patients.
Circ J 2003;67:287–294.
9. Heart Protection Study Collaborative Group. Effects of cholesterollowering with simvastatin on stroke and other major vascular events in
20,536 people with cerebrovascular disease or other high-risk conditions. Lancet 2004;363:757–767.
10. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk
individuals: a randomised placebo-controlled trial. Lancet 2002;360:
7–22.
11. Bradford RH, Shear CL, Chremos AN, Dujovne CA, Franklin FA,
Grillo RB, Higgins J, Langendorfer A, Nash DT, Pool JL, Schnaper H.
Expanded clinical evaluation of lovastatin (EXCEL) study results:
2-year efficacy and safety follow-up. Am J Cardiol 1994;74:667– 673.
12. PROspective Study of Pravastatin in the Elderly at Risk (PROSPER)
Study Group. Pravastatin in elderly individuals at risk of vascular
disease (PROSPER): a randomised controlled trial. Lancet 2002;360:
1623–1630.
13. Pedersen TR, Berg K, Cook TJ, Færgeman O, Haghfelt T, Kjekshus J,
Miettinen T, Musliner TA, Olsson AG, Pyorala K, et al. Safety and
tolerability of cholesterol lowering with simvastatin during 5 years in
the Scandinavian Simvastatin Survival Study. Arch Intern Med 1996;
156:2085–2092.
14. Wysowski DK, Kennedy DL, Gross TP. Prescribed use of cholesterollowering drugs in the United States, 1978 through 1988. JAMA 1990;
263:2185–2188.
15. Staffa JA, Chang J, Green L. Cerivastatin and reports of fatal rhabdomyolysis [letter]. N Engl J Med 2002;346:539 –540.
16. Evans M, Rees A. Effects of HMG-CoA reductase inhibitors on
skeletal muscle. Drug Saf 2002;25:649 – 663.
17. Zhou S, Chan SY, Cher Goh B, Chan E, Duan W, Huang M, McLeod
HL. Mechanism-based inhibition of cytochrome P450 3A4 by therapeutic drugs. Clin Pharmacokinet 2005;44:279 –304.
18. Omar MA, Wilson JP. FDA adverse event reports on statin-associated
rhabdomyolysis. Ann Pharmacother 2002;36:288 –295.
19. Jones PH, Davidson MH. Reporting rate of rhabdomyolysis with
fenofibrate ⫹ statin versus gemfibrozil ⫹ any statin. Am J Cardiol
2005;95:120 –122.
20. Kyrklund C, Backman JT, Kivistö KT, Neuvonen M, Laitila J, Neuvonen PJ. Plasma concentrations of active lovastatin acid are markedly
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
increased by gemfibrozil but not by bezafibrate. Clin Pharmacol Ther
2001;69:340 –345.
Prueksaritanont T, Zhao JJ, Ma B, Roadcap BA, Tang C, Qiu Y, Liu
L, Lin JH, Pearson PG, Baillie TA. Mechanistic studies on metabolic
interactions between gemfibrozil and statins. J Pharmacol Exp Ther
2002;301:1042–1051.
Backman JT, Kyrklund C, Neuvonen M, Neuvonen PJ. Gemfibrozil
greatly increases plasma concentrations of cerivastatin. Clin Pharmacol Ther 2002;72:685– 691.
Gaist D, Rodríguez G, Huerta C, Hallas J, Sindrup SH. Lipid-lowering
drugs and risk of myopathy: a population-based follow-up study.
Epidemiology 2001;12:565–569.
Shanahan RL, Kerzee JA, Sandhoff BG, Carroll NM, Merenich JA, for
the Clinical Pharmacy Cardiac Risk Service (CPCRS) Study Group.
Low myopathy rates associated with statins as monotherapy or combination therapy with interacting drugs in a group model health maintenance organization. Pharmacotherapy 2005;25:345–351.
Kinnman N, Hultcrantz R. Lipid lowering medication and hepatotoxicity. J Intern Med 2001;250:183–185.
Pfeffer MA, Keech A, Sacks FM, Cobbe SM, Tonkin A, Byington RP,
Davis BR, Friedman CP, Braunwald E. Safety and tolerability of
pravastatin in long-term clinical trials. Circulation 2002;105:2341–
2346.
US Food and Drug Administration, Center for Drug Evaluation and
Research. Statins and hepatotoxicity. [US Food and Drug Administration Web site] Available at: http://www.fda.gov/ohrms/dockets/ac/00/
backgrd/3622b2b_safety_review.pdf. Accessed August 15, 2005.
Tolman KG. The liver and lovastatin. Am J Cardiol 2002;89:1374 –
1380.
Newman CB, Palmer G, Silbershatz H, Szarek M. Safety of atorvastatin derived from analysis of 44 completed trials in 9,416 patients.
Am J Cardiol 2003;92:670 – 676.
Fried LF, Orchard TJ, Kasiske BL, for the Lipids and Renal Disease
Progression Meta-Analysis Study Group. Effect of lipid reduction on
the progression of renal disease: a meta-analysis. Kidney Int 2001;59:
260 –269.
Fellstrom B, Holdaas H, Jardine AG, Holme I, Nyberg G, Fauchald P,
Gronhagen-Riska C, Madsen S, Neumayer HH, Cole E, et al, for the
Assessment of Lescol in Renal Transplantation Study Investigators.
Effect of fluvastatin on renal end points in the Assessment of Lescol in
Renal Transplant (ALERT) trial. Kidney Int 2004;66:1549 –1555.
Prospective Studies Collaboration. Cholesterol, diastolic blood pressure and stroke: 13,000 strokes in 450,000 people in 45 prospective
cohorts. Lancet 1995;346:1647–1653.
Chong PH, Boskovich A, Stevkovic N, Bartt RE. Statin-associated
peripheral neuropathy: review of literature. Pharmacotherapy 2004;
24:1194 –1203.
Anderson JLK, Muhlestein JB, Bair TL, Morris S, Weaver AN, Lappé
DL, Renlund DG, Pearson RR, Jensen KR, Horne BD. Do statins
increase the risk of idiopathic polyneuropathy? Am J Cardiol 2005;
95:1097–1099.
Corrao G, Zambon A, Bertù L, Botteri E, Leoni O, Contiero P.
Lipid-lowering drugs prescription and the risk of peripheral neuropathy: an exploratory case-control study using automated databases. J
Epidemiol Community Health 2004;58:1047–1051.
Gaist D, Jeppesen U, Andersen M, Rodriguez G, Hallas J, Sindrup SH.
Statins and risk of polyneuropathy. Neurology 2002;58:1333–1337.
Gaist D, Rodrigues LAG, Huerta C, Hallas J, Sindrup SH. Are users of
lipid-lowering drugs at increased risk of peripheral neuropathy? Eur
J Clin Pharmacol 2001;56:931–933.
Ziajka PE, Wehmeier T. Peripheral neuropathy and lipid-lowering
therapy. Southern Med J 1998;91:667– 668.
Statin Safety: An Assessment Using an Administrative Claims
Database
Mark J. Cziraky, PharmD,a,* Vincent J. Willey, PharmD,a James M. McKenney, PharmD,b
Siddhesh A. Kamat, MS,a Maxine D. Fisher, PhD,a John R. Guyton, MD,c
Terry A. Jacobson, MD,d and Michael H. Davidson, MDe
The large administrative databases of health plans contain information on drugrelated medical adverse events (AE) and constitute an increasingly powerful tool for
the assessment of drug safety. We conducted a retrospective observational study
using an administrative managed care claims database covering 9 million members
from diverse regions of the United States. Patients aged >18 years who received >2
prescriptions for lipid-lowering drugs between July 1, 2000 and December 1, 2004
were included in the study. Hospitalizations with diagnosis codes (International
Classification of Diseases, 9th Revision, Clinical Modification [ICD-9]) related to
muscle, kidney, and liver were determined for patients exposed to 3-hydroxy-3methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors (statins), fibrates, extended-release niacin, cholesterol absorption inhibitors, or statin combination therapy. A total of 473,343 patients contributed 490,988 person-years of monotherapy and
11,624 person-years of combination dyslipidemia therapy. Rates of hospitalization
due to AEs in patients on monotherapy with currently available statins were similar,
whereas the incidence of hospitalization for muscle disorders increased 6.7-fold with
cerivastatin therapy. Patients who received a lipid-lowering medication with a concomitant cytochrome P450 3A4 (CYP3A4) inhibitor had a 6-fold increased rate of
muscle disorders, including rhabdomyolysis. Hypertension was associated with a
5-fold increase in both muscle and renal events, whereas patients with diabetes
mellitus had a 2.5-fold increased risk of renal events. No hospitalized cases of the
index AEs were observed in study subjects during the 6-month period before
initiation of the lipid-lowering drug. Statin monotherapy as currently prescribed
is generally well tolerated and safe. © 2006 Elsevier Inc. All rights reserved. (Am
J Cardiol 2006;97[suppl]:61C– 68C)
The 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA)
reductase inhibitors, or statins, have been demonstrated in
numerous clinical trials to have beneficial effects on cardiovascular morbidity and mortality and hence have continued
to be recommended as first-line dyslipidemia therapy in the
latest National Cholesterol Education Program (NCEP)
guidelines.1 However, certain medical adverse events (AEs)
related to the skeletal muscle, renal, and hepatic systems
have been associated with statins and other lipid-lowering
drugs, even though the cases of serious AEs associated with
statin use have been extremely limited.2–5 The most life
threatening of these serious AEs is rhabdomyolysis, a disorder that is often linked to myoglobinuria, myoglobinemia,
and acute renal failure. Following recent withdrawals of
drugs previously approved by the US Food and Drug ada
HealthCore, Inc., Wilmington, Delaware, USA; bNational Clinical
Research, Inc., Richmond, Virginia, USA; cDuke University Medical Center, Durham, North Carolina, USA; dEmory University, Atlanta, Georgia,
USA; eRush University Medical Center, Chicago, Illinois, USA.
*Address for reprints: Mark J. Cziraky, PharmD, HealthCore, Inc., 800
Delaware Avenue, 5th Floor, Wilmington, Delaware 19801.
E-mail address: mcziraky@healthcore.com.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.011
ministration (FDA), in particular the withdrawal of cerivastatin from the US market in August 2001,6 concerns
have arisen among clinicians and patients regarding the
safety of lipid-lowering agents,7 especially when used in
combination.
Currently, most data regarding drug-associated AEs result from randomized clinical trials, the FDA Adverse Event
Reporting System (AERS) case reports, and observational
cohort studies. Large-scale population-based studies using
administrative and prescription claims data allow the creation of a much broader demographic portrait of reported
AEs than can be achieved by either the FDA AERS or
clinical trials. Indeed, 2 recent studies used prescription
claims data and confirmatory review of medical charts and
records from geographically dispersed healthcare management organizations. These analyses were designed to investigate the incidence rates of myopathy and rhabdomyolysis
associated with use of statins and/or fibrates and showed a
similar risk of both AEs with use of atorvastatin, pravastatin, or simvastatin. Additionally, the evaluations also demonstrated an increased risk with combination statin-fibrate
use.8,9 The large administrative databases of health plans
www.AJConline.org
62C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
can provide a very useful reservoir of data regarding drugrelated AEs, and with the growth of such organizations and
related large-scale integrated databases, the use of such data
can constitute an increasingly powerful tool for the assessment of drug safety.8 The present study evaluates the incidence of hospitalizations for myopathy, renal medical
events, and hepatic medical events in patients on statins and
other lipid-lowering therapies in a “real-world” clinical
practice setting via health plan administrative claims data.
Methods
Data source: This retrospective observational study was
conducted using administrative claims data from diverse
regions in the United States, including western, midwestern,
mid-Atlantic, and southeastern states, covering ⬎9 million
commercially insured lives. The data consisted of automated health plan enrollment, medical, and pharmacy administrative claims files. All study materials were handled
in compliance with Health Insurance Portability and Accountability Act of 1996 (HIPAA) regulations, and the
analysis was conducted on a limited dataset.
Patient cohort: Adult patients (aged ⱖ18 years) who
received ⱖ2 prescriptions for a lipid-lowering drug between
July 1, 2000 and December 1, 2004 were included in the
study. Using pharmacy claims data, inception cohorts were
identified for statin monotherapy (atorvastatin, cerivastatin,
fluvastatin, lovastatin, pravastatin, rosuvastatin, and simvastatin), nonstatin monotherapy (ezetimibe, fenofibrate, gemfibrozil, and extended-release [niacin-ER]), and combinations of nonstatins with any other statin. Patients were
included in the study cohort if their first dispensed lipidlowering drug prescription (index drug fill) was preceded by
a 6-month period without any lipid-lowering drug dispensed
prescriptions (fills). Continuous drug therapy was defined as
consecutive lipid-lowering drug fills occurring within the
timeframe consisting of the days’ supply of the previous fill
plus an additional 30 days. The 30-day buffer was used as
per Graham and colleagues9 to account for imperfect adherence to therapy and gaps between drug fills. Lipid-lowering
drug exposure time was estimated for each patient based on
the number of days for which the patient received continuous lipid-lowering drug therapy and was reported in personyears. As a result of medication additions and switches,
patients could contribute to multiple cohorts. Patient medical claims were used to identify characteristics such as age
and sex as well as the presence of comorbidities including
hypertension, diabetes mellitus, and coronary artery disease
(CAD).
Identification of adverse medical events: This study
evaluated the incidence rates of hospitalizations for myopathies (including rhabdomyolysis), renal medical events,
and hepatic medical events using International Classification of Diseases, 9th Revision, Clinical Modification
(ICD-9) codes. Codes on inpatient claims were used to
provide greater specificity to the event identification criteria. Myopathy events requiring hospitalization were identified by ICD-9 codes used in previous studies8 and included
myoglobinuria (791.3x), disorders of the muscle, ligament,
and fascia (728.89), and rhabdomyolysis (728.88). Renal
events requiring hospitalization were identified using codes
for acute renal failure/acute tubular necrosis (584.xx), and
acute glomerulonephritis (580.xx). Hepatic events requiring
hospitalization were captured using codes for acute/subacute necrosis of liver (570.xx), hepatitis (573.3x), other
specified disorders of the liver (573.8x), and unspecified
disorders of liver (573.9x). These ICD-9 codes were selected based on previous reports,4 AEs mentioned in the
prescribing information of the lipid-lowering drug, and the
advice of expert panels of clinicians convened by the National Lipid Association (NLA).
Outcome metrics: Incidence rates for medical AEs were
determined per 10,000 person-years with 95% confidence
intervals (CI). Within the lipid-lowering drug inception cohort, a nested case-control analysis for each AE was conducted to estimate the risk of AEs associated with different
lipid-lowering drug therapies.10,11 The case group consisted
of patients who experienced the AE of interest. The control
group was identified from a pool of patients receiving lipidlowering drugs who did not experience AEs. Cases and
controls were randomly matched for age, sex, geographic
region, length of follow-up, and time of index drug fill. For
myopathy and hepatic events requiring hospitalization, 12
controls were selected per case because, provided sufficient
controls are available for matching, the use of more than the
conventional 4 controls per case has been suggested to
increase power of the statistical tests.12 However, the renal
event hospitalization analysis included up to 4 controls per
case owing to a greater renal event hospitalization rate and
limited availability of potential controls.
Based on criteria for adequate model fit, Poisson or
negative binomial regression was used to estimate relative
risk (RR) for inpatient encounters of AEs between different
lipid-lowering drug therapies. Robust standard errors were
used to account for matched pairs. For the purpose of all
multivariable comparisons, atorvastatin monotherapy was
used as the reference group because it is the most widely
prescribed statin and it was also used as the reference group
in a similar statin safety study conducted by Staffa and
coworkers.13 All analyses were conducted using Stata Version 8.2 (StataCorp, College Station, TX) and SAS 9.1
(SAS Institute Inc., Cary, NC) software.
Results
A total of 473,343 patients contributed 490,988 personyears of monotherapy and 11,624 person-years of combination dyslipidemia therapy (Table 1). Consistent with general
use of dyslipidemia therapy in the United States,14,15 86% of
Cziraky et al/Statin Safety: Assessment Using a Claims Database
63C
Table 1
Description of patient cohorts using dyslipidemia therapy
Person-years (n)
Monotherapy
Combination therapy*
Demographics
Age, yr (mean ⫾ SD)
Male (%)
Hypertension (%)
Diabetes mellitus (%)
CAD (%)
Renal disease (%)
CYP3A4 inhibitor† (%)
Rhabdomyolysis AE cases (n)
Monotherapy
Combination therapy*
Renal AE cases (n)
Monotherapy
Combination therapy*
Hepatic AE cases (n)
Monotherapy
Combination therapy*
Atorvastatin
(n ⫽ 264,399)
Cerivastatin
(n ⫽ 11,879)
Fluvastatin
(n ⫽ 17,761)
Lovastatin
(n ⫽ 39,624)
Pravastatin
(n ⫽ 70,811)
Rosuvastatin
(n ⫽ 18,584)
Simvastatin
(n ⫽ 66,757)
261,567
6,544
4,719
25
12,635
226
26,122
547
64,254
2,241
8,213
434
54,394
1,607
54.6 ⫾ 11.9
54.9
69.9
26.3
31.6
2.1
0.88
57.1 ⫾ 12.9
50.5
69.6
26.6
32.2
1.9
0.69
57.3 ⫾ 13.2
49
71.9
27.5
29.4
2
0.7
55.6 ⫾ 12.9
47.3
68.8
29.2
23.3
1.5
0.55
55.4 ⫾ 12.1
51.4
72.2
28.4
33.3
2.3
1.11
53.6 ⫾ 11.1
54.8
69.4
23.9
31
1.5
1.06
57.5 ⫾ 12.4
54.4
76.6
28.2
44.2
2.7
0.68
64
2
5
0
2
0
6
0
22
0
2
0
19
1
810
47
15
0
37
0
78
1
202
10
22
1
297
12
257
9
3
0
8
1
16
0
69
0
7
0
70
0
AE ⫽ medical adverse events requiring hospitalization; CAD ⫽ coronary artery disease; CYP3A4 ⫽ cytochrome P-450 3A4.
* Combination therapy is only described in terms of the statin component.
†
Concomitant use of any of the following CYP3A4 inhibitors: cyclosporine, ketoconazole, clarithromycin, human immunodeficiency virus protease
inhibitors, itraconazole, erythromycin, telithromycin, or nefazodone.
therapy was statin monotherapy. Use of atorvastatin monotherapy was 4 times greater than that of the next mostprescribed therapy and represented ⬎50% of the overall
person-years in the study. The overall mean age of the
cohort was 54.7 years, and 55.2% were men (Table 2).
Patients in the niacin-ER and fibrate cohorts were younger
and contained a greater percentage of women in comparison
with the overall cohort. Hypertension (71.2%), diabetes
(27.9%), and CAD (32.8%) were observed to be prevalent
in the overall cohort population. Diabetes was more common among fibrate users (35.2%), which is consistent with
previous research and the use of these drugs in this patient
population.9,16
The total cases of hospitalization for myopathy (n ⫽
144), renal events (n ⫽ 1,786), and hepatic events (n ⫽ 518)
were extracted from the administrative claims database.
Incidence of hospitalization for myopathy (including rhabdomyolysis) in patients treated with monotherapy ranged
from 1.58 with fluvastatin to 10.59 with cerivastatin per
10,000 person-years (Table 3). The incidence of myopathy
cases requiring hospitalization after use of cerivastatin was
significantly greater than with the most commonly prescribed statin monotherapy (p ⬍0.01). There were no significant differences among other statins with respect to the
incidence of myopathy events requiring hospitalization.
The incidence of renal events requiring hospitalization in
patients treated with monotherapy ranged from 26.79 with
rosuvastatin to 54.6 with simvastatin per 10,000 personyears. Simvastatin monotherapy and gemfibrozil monotherapy were associated with the highest incidence of renal
events requiring hospitalization and were significantly
higher than the most commonly prescribed statin (p ⬍0.01
for both). There were no significant differences among the
other monotherapies with respect to the incidence of renal
events requiring hospitalization.
The incidence of hepatic events requiring hospitalization
ranged from 6.13 with lovastatin to 16.32 with ezetimibe
per 10,000 person-years. There were no significant differences between monotherapies with respect to the incidence
of these hepatic events.
Combination therapy, even after aggregating all statins
together by nonstatin therapy, contributed to a low number of
patient-years of observation, relatively few hospitalizations for
myopathy, renal events, and hepatic events, and a wide 95% CI
around the incident rate point estimates (Table 3).
To establish baseline rates of AEs, the 6-month period
before initiation of lipid-lowering drugs was reviewed for
all patients. During the 240,193 person-years of unexposed
time, no cases of hospitalization for any of the aforementioned AEs were observed. The RR for each AE was estimated using atorvastatin monotherapy as the reference
group. Other covariates of clinical significance that were not
included in the match criteria were also explored.
Risk of hospitalization due to myopathy was increased
5.13 times (95% CI, 2.42–10.85) in the presence of hypertension and 6.01 times (95% CI, 2.08 –17.38) in the presence of the coadministration of cytochrome P450 3A4
(CYP3A4) inhibitors (Figure 1). In terms of therapy, only
cerivastatin (RR, 6.69; 95% CI, 2.14 –20.99) was clearly
associated with increased risk of hospitalization due to my-
64C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 2
Description of patient cohorts* using dyslipidemia therapy
Person-years (n)
Monotherapy
Combination therapy‡
Demographics
Age, yr (mean ⫾ SD)
Male (%)
Hypertension (%)
Diabetes mellitus (%)
CAD (%)
Renal disease (%)
CYP3A4 inhibitor§ (%)
Rhabdomyolysis AE cases (n)
Monotherapy
Combination therapy‡
Renal AE cases (n)
Monotherapy
Combination therapy‡
Hepatic AE cases (n)
Monotherapy
Combination therapy‡
Ezetimibe
(n ⫽ 16,604)
Fenofibrate
(n ⫽ 30,059)
Gemfibrozil
(n ⫽ 28,919)
Niacin-ER
(n ⫽ 18,392)
Total†
(n ⫽ 473,343)
9,192
NA
21,063
NA
19,471
NA
9,358
NA
490,988
11,624
56.2 ⫾ 11.5
47.2
73.9
25.9
38.9
2.3
1.03
51.0 ⫾ 11.8
65.2
71.6
35.2
28.4
2.2
1.02
51.6 ⫾ 12.4
62.7
68.7
35.1
23.6
1.9
0.98
53.1 ⫾ 12.2
71.1
68.5
23.6
35.6
2.3
0.73
54.7 ⫾ 12.1
55.2
71.2
27.9
32.8
2.1
0.87
3
NA
6
7
NA
25
NA
NA
82
106
NA
15
NA
NA
25
141
3
41
1,715
71
12
508
10
NA
26
NA
5
NA
NA
NA
AE ⫽ medical adverse events requiring hospitalization; CAD ⫽ coronary artery disease; ER ⫽ extended release; NA ⫽ not available.
* Patients could contribute to multiple cohorts.
†
Total refers to all statin and nonstatin therapies.
‡
Combination therapy is only described in terms of the statin component.
§
Concomitant use of any of the following CYP3A4 inhibitors: cyclosporine, ketoconazole, clarithromycin, human immunodeficiency virus protease
inhibitors, itraconazole, erythromycin, telithromycin, nefazodone.
Table 3
Incident rates of medical adverse events in patients receiving lipid-lowering drugs (per 10,000 person-years)
Therapy
Monotherapy
Atorvastatin
Cerivastatin
Fluvastatin
Lovastatin
Pravastatin
Rosuvastatin
Simvastatin
Ezetimibe
Fenofibrate
Gemfibrozil
Niacin-ER
Combination therapy
Statin ⫹ ezetimibe
Statin ⫹ fenofibrate
Statin ⫹ gemfibrozil
Statin ⫹ niacin-ER
Person-Years
261,567
4,719
12,635
26,122
64,254
8,213
54,394
9,192
21,063
19,471
9,358
2,903
3,854
1,894
2,973
Myopathies
(95% CI)
2.45 (1.9–3.1)
10.59 (3.4–24.7)
1.58 (0.2–5.7)
2.3 (0.8–4.5)
3.42 (2.1–5.2)
2.44 (0.3–8.8)
3.49 (2.1–5.5)
3.26 (0.7–9.5)
2.85 (1.0–6.2)
3.6 (1.4–7.4)
5.34 (1.7–12.5)
0.0 (0–12.7)
5.19 (0.6–18.7)
0.0 (0–19.5)
3.36 (0.08–187.3)
Renal Events
(95% CI)
30.97 (28.8–33.2)
31.78 (17.8–52.4)
29.28 (20.6–40.3)
29.86 (23.6–37.3)
31.44 (27.3–36.1)
26.79 (16.8–40.5)
54.6 (48.6–61.2)
27.2 (17.6–40.1)
38.93 (30.9–48.3)
54.44 (44.6–65.8)
43.81 (31.5–59.4)
37.89 (18.9–67.8)
70.05 (46.2–101.8)
137.31 (89.7–201.1)
23.55 (9.5–48.5)
Hepatic Events
(95% CI)
9.83 (8.7–11.1)
6.36 (1.3–18.6)
6.33 (2.7–12.5)
6.13 (3.5–9.9)
10.74 (8.3–13.6)
8.52 (3.4–17.6)
12.87 (10–16.3)
16.32 (9.2–26.9)
11.87 (7.7–17.5)
13.35 (8.7–19.6)
12.82 (6.6–22.4)
3.44 (0.09–191.9)
7.78 (1.6–22.7)
21.12 (5.8–54.1)
6.73 (0.8–24.3)
CI ⫽ confidence interval; ER ⫽ extended release.
opathy. Fenofibrate (RR, 4.32; 95% CI 1.06 –17.68) was
marginally associated with increased risk of hospitalization
for myopathy compared with atorvastatin.
Risk of renal events requiring hospitalization was increased over 7-fold in patients with hypertension (RR, 7.02;
95% CI, 3.69 –13.35) and 2.8-fold in patients with diabetes
(RR, 2.83; 95% CI, 2.44 –3.30) (Figure 2). With regard to
lipid-lowering drugs, only marginal differences were evident for renal events requiring hospitalization. The relative
risk estimates were lower in patients on lovastatin monotherapy (RR, 0.58; 95% CI, 0.44 – 0.77) and ezetimibe (RR,
0.57; 95% CI, 0.33– 0.99) and higher in patients on simva-
Cziraky et al/Statin Safety: Assessment Using a Claims Database
65C
Figure 1. Relative risk (RR) estimate of myopathies with 95% confidence intervals in parentheses. A total of 144 cases were matched to 1,728 controls by
age, sex, geographic region, length of follow-up, and time of index drug fill. RR was estimated using Poisson regression with robust standard errors. CYP3A4
⫽ cytochrome P450 3A4.
statin monotherapy (RR, 1.31; 95% CI, 1.08 –1.59), combination statin-gemfibrozil therapy (RR, 1.54; 95% CI, 1.06 –
2.23), and combination statin-fenofibrate therapy (RR, 1.66;
95% CI, 1.05–2.62).
Hypertension (RR, 2.55; 95% CI, 1.73–3.70) and diabetes (RR, 1.84; 95% CI, 1.45–2.32) also increased the risk of
hospitalization for hepatic events (Figure 3). In terms of the
effects of therapy, the relative risk estimate was marginally
lower for lovastatin (RR, 0.53; 95% CI, 0.30 – 0.95) and
pravastatin (RR, 0.69; 95% CI, 0.50 – 0.95) and marginally
higher for the combination of statin and gemfibrozil (RR,
3.83; 95% CI, 1.23–11.95) compared with the atorvastatin
reference group.
Discussion
The focus of this analysis was to evaluate the RR for AEs in
patients treated with statins and other lipid-lowering therapies
in a “real-world” clinical practice setting. The selection of
target AEs was based on clinical importance, and included
events in muscle, kidney, and liver. In our analysis, a total of
473,343 patients contributed 490,988 person-years of monotherapy and 11,624 person-years of combination therapy with
a statin and another lipid-lowering drug. To our knowledge,
this is the first administrative claims data analysis of serious
AEs to include the more recent market entries of rosuvastatin
and ezetimibe. In addition, our analysis examined mono-
therapy and combined statin-nonstatin therapy with fenofibrate, gemfibrozil, ezetimibe, and niacin-ER.
Our results demonstrated that the rates of muscle abnormalities requiring hospitalization in patients who received
pravastatin, rosuvastatin, lovastatin, simvastatin, fluvastatin,
or atorvastatin were similar. Consistent with previous research and reports,9,13 cerivastatin was associated with a
6.7-fold increase in myopathy when compared to monotherapy with atorvastatin. Moreover, in patients who received a lipid-lowering medication with a concomitant
CYP3A4 inhibitor, a 6-fold increased risk of myopathy was
observed. The mechanism for this drug-drug interaction
with this class of medications is thought to be through the
inhibition of the CYP450 enzyme system and resultant
elevation in serum statin levels. Specifically, the CYP3A4
system is responsible for the metabolism of the majority of
medications including statins.17 In the statin class, simvastatin, lovastatin, cerivastatin, and atorvastatin are the most
extensively metabolized through this pathway.17 Concomitant medications that inhibit this metabolic pathway such as
those used in our analysis can lead to an increase in the
concentrations of the concomitantly administered statin and
increase the potential for the development of myopathy and
other concentration related AEs.
The use of simvastatin as monotherapy and the use of
any statin combined with gemfibrozil or fenofibrate were
associated with a greater number of renal AEs and a marginally increased renal risk. These results should be re-
66C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Figure 2. Relative risk (RR) estimate of renal adverse medical events with 95% confidence intervals in parentheses. A total of 1,785 cases were matched to
7,140 controls by age, sex, geographic region, length of follow-up, and time of index drug fill. One case was dropped owing to unavailability of controls
available for matching. RR was estimated using negative binomial regression with robust standard errors. CYP3A4 ⫽ cytochrome P450 3A4.
garded as hypothesis-generating, not definitive. Simvastatin
and lovastatin are divergent in this analysis, but the drugs
are very similar both structurally and pharmacologically, as
detailed in the prescribing information for these products.
The association of simvastatin monotherapy with renal
events is not consistent with previous clinical experience.18,19 This may be related in part to a difference in the
populations that make up the administrative claims database
and those that are generally included in a clinical trial. The
claims environment may be reflecting prescribing patterns
that support more aggressive management of patients with
elevated cholesterol levels as well as patients with greater
comorbidity than seen in healthier clinical trial volunteers.
In addition, while matching and multivariate analysis techniques were used, these techniques cannot account for patient characteristics that were not studied and reported in the
claims data.
The results showed no increase in risk of hepatic AEs
with the use of any statin monotherapy, which was consistent with the clinical trial literature.1 The combination of a
statin with gemfibrozil was associated with an increased risk
of hepatic AEs, which was also consistent with clinical trial
data.1
Commonly associated comorbidities were also shown to
increase the risk of myopathy in our study. Hypertension
was associated with a 5-fold increase in both myopathy and
renal AEs, whereas patients with diabetes had a 2.5-fold
increased risk of renal AEs.
Historically, various types of study designs have been
used to capture data related to the relative safety of specific
treatment regimens. Some of the more common approaches
include randomized clinical trials, postmarketing surveillance using patient registries, spontaneous reporting systems such as the FDA AERS system, and administrative
claims analyses.8,9,18 –22 Each methodology has various
strengths and weaknesses when considering the overall assessment of safety. Understanding the strengths and weaknesses of these various sources of information, as well as
their potential for application to a specific population, is
critical when interpreting results.
Our study used an administrative claims database that
provided a large, general clinical practice setting–treated
population for analysis. This type of study design can prove
advantageous, specifically in comparison with clinical trials,
because of its ability to detect rare drug-related AEs such as
rhabdomyolysis.8,9 Inpatient claims were used in this analysis to provide greater specificity to the potential AE. To
define myopathy within the administrative claims database
for our study, previous research from Andrade and associates8 that explored the positive predictive value of multiple
claims-based definitions was used.
Although there are advantages to administrative claims
Cziraky et al/Statin Safety: Assessment Using a Claims Database
67C
Figure 3. Relative risk (RR) estimate of hepatic adverse medical events with 95% confidence intervals in parentheses. A total of 518 cases were matched
to 6,216 controls by age, sex, geographic region, length of follow-up, and time of index drug fill. RR was estimated using Poisson regression with robust
standard errors. CYP3A4 ⫽ cytochrome P450 3A4.
data analyses, there also are limitations to the methodology
applied and population used in this analysis that should be
considered as these data are interpreted. The database used
for this analysis is not representative of the entire US population and is deficient in certain sociodemographic variables such as race/ethnicity, family history of cardiovascular disease, income status, and education, which may have
an effect on the outcomes analyzed. Our analysis was based
on the claims submitted with appropriate statistical analysis
using a nested case control design for each associated AE
applied.11 Chart verification was not performed9 and the
retrospective study design does not allow the causality of
AEs to be determined. However, RR or association can be
defined. Because the databases analyzed reflect a commercially insured population, caution should be used when
applying these results beyond this population. Despite these
limitations, the consistency between the findings in our
report with prior studies supports the reasonable validity of
the methodologic approaches applied.
Statins continue to be the most commonly prescribed
class of medications to manage patients with dyslipidemia
and are the most effective class at reducing low-density
lipoprotein (LDL) cholesterol levels, as well as morbidity
and mortality in patients with dyslipidemia.1 In both the
current voluntary reporting system used by the FDA
(AERS)22 and our most recent analysis using commercial
health plan databases, the incidence of liver and renal ab-
normalities and of myopathy requiring hospitalization is
extremely low. Prevention and awareness of the potential
for drug-related AEs are the most important approaches to
be used to limit the risks associated with these medications.
Owing to the increasing prevalence of metabolic syndrome23 and diabetes in the US population, the clinical need
has emerged to treat the full range of lipid abnormalities in
patients, including high levels of LDL cholesterol and triglycerides and low levels of high-density lipoprotein cholesterol, using combination lipid-lowering therapy. As use
of combination lipid-lowering medications in these individuals increases, the potential for associated AEs increases
and appropriate monitoring becomes necessary.
Conclusion
Our study corroborates previous findings that statin monotherapy as currently prescribed is generally well tolerated
and safe. Comorbidities such as hypertension and diabetes,
certain combination therapies, and the coadministration of
CYP3A4 inhibitors increase the risk of AEs. As newer
guidelines continue to recommend more aggressive lipidlowering targets, higher doses and combinations of lipidlowering medications will be required and used to meet
these recommended goals. Appropriate monitoring will be
important to minimize the consequences of AEs in these
68C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
patients at higher risk. Further study should focus on ascertaining the risk from therapy as patients are treated more
aggressively.
1. Third Report of the National Cholesterol Education Program (NCEP)
Expert Panel on Detection, Evaluation, and Treatment of High Blood
Cholesterol in Adults (Adult Treatment Panel III) final report Circulation 2002;106:3143–3421.
2. Davidson MH. Safety profiles for the HMG-CoA reductase inhibitors:
treatment and trust. Drugs 2001;61:197–206.
3. Evans M, Rees A. The myotoxicity of statins. Curr Opin Lipidol
2002;13:415– 420.
4. Tomlinson B, Chan P, Lan W. How well tolerated are lipid-lowering
drugs? Drugs Aging 2001;18:665– 683.
5. Wortmann RL. Lipid-lowering agents and myopathy. Curr Opin Rheumatol 2002;14:643– 647.
6. Furberg CD, Pitt B. Withdrawal of cerivastatin from the world market.
Curr Control Trials Cardiovasc Med 2001;2:205–207.
7. Fontanarosa PB, Rennie D, DeAngelis CD. Postmarketing surveillance—lack of vigilance, lack of trust. JAMA 2004;292:2647–2650.
8. Andrade SE, Graham DJ, Staffa JA, Scheck SD, Shatin D, La Grenade
L Goodman MJ, Platt R, Gurwitz JH, Chan KA. Health plan administrative databases can efficiently identify serious myopathy and rhabdomyolysis. J Clin Epidemiol 2005;58:171–174.
9. Graham DJ, Staffa JA, Shatin D, Andrade SE, Schech SD, La Grenade
L, Gurwitz JH, Chan KA, Goodman MJ, Platt R. Incidence of hospitalized rhabdomyolysis in patients treated with lipid-lowering drugs.
JAMA 2004;292:2585–2590.
10. Ernster VL. Nested case-control studies. Prev Med 1994;23:587–590.
11. Graham DJ, Campen D, Hui R, Spence M, Cheetham C, Levy G,
Shoor S, Ray WA. Risk of acute myocardial infarction and sudden
cardiac death in patients treated with cyclo-oxygenase 2 selective and
non-selective non-steroidal anti-inflammatory drugs: nested case-control study. Lancet 2005;365:475– 481.
12. Pang D. A relative power table for nested matched case-control studies. Occup Environ Med 1999;56:67– 69.
13. Staffa JA, Chang J, Green L. Cerivastatin and reports of fatal rhabdomyolysis. N Engl J Med 2002;346:539 –540.
14. Mosca L, Merz NB, Blumenthal RS, Cziraky MJ, Fabunmi RP, Sarawate C, Watson KE, Willey VJ, Stanek EJ. Opportunity for intervention to achieve American Heart Association guidelines for optimal
lipid levels in high-risk women in a managed care setting. Circulation
2005;111:488 – 493.
15. Zielinski SL. FDA attempting to overcome major roadblocks in monitoring drug safety. J Natl Cancer Inst 2005;97:872– 873.
16. Cottrell DA, Marshall BJ, Falko JM. Therapeutic approaches to dyslipidemia in diabetes mellitus and metabolic syndrome. Curr Opin
Cardiol 2003;18:301–308.
17. Thummel KE, Wilkinson GR. In vitro and in vivo drug interactions
involving human CYP3A. Annu Rev Pharmacol Toxicol 1998;38:389 –
430.
18. Randomised trial of cholesterol lowering in 4444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S)
Lancet 1994;344:1383–1389.
19. Collins R, Armitage J, Parish S, Sleigh P, Peto R, for the Heart
Protection Study Collaborative Group. MRC/BHF Heart Protection
Study of cholesterol-lowering with simvastatin in 5963 people with
diabetes: a randomised placebo-controlled trial. Lancet 2003;361:
2005–2016.
20. Alsheikh-Ali AA, Ambrose MS, Kuvin JT, Karas RH. The safety of
rosuvastatin as used in common clinical practice: a postmarketing
analysis. Circulation 2005;111:3051–3057.
21. Cheung BM, Lauder IJ, Lau CP, Kumana CR. Meta-analysis of large
randomized controlled trials to evaluate the impact of statins on cardiovascular outcomes. Br J Clin Pharmacol 2004;57:640 – 651.
22. Omar MA, Wilson JP. FDA adverse event reports on statin-associated
rhabdomyolysis. Ann Pharmacother 2002;36:288 –295.
23. Ford ES, Giles WH, Dietz WH. Prevalence of the metabolic syndrome
among US adults: findings from the third National Health and Nutrition Examination Survey. JAMA 2002;287:356 –359.
An Assessment of Statin Safety by Muscle Experts
Paul D. Thompson, MD,a,* Priscilla M. Clarkson, PhD,b and Robert S. Rosenson, MDc
The National Lipid Association’s (NLA) Muscle Safety Expert Panel was charged
with the duty of examining the definitions, causative factors, and management of
statin myopathy. The Panel was asked to use its evidence-based findings to form
recommendations in response to a series of specific questions posed by the Task Force.
The panel was composed of a clinical cardiologist, an exercise physiologist and
skeletal muscle expert, and an expert in preventive cardiology who also examined
skeletal muscle complications of statin use. © 2006 Elsevier Inc. All rights reserved.
(Am J Cardiol 2006;97[suppl]:69C–76C)
Questions Posed by the National Lipid Association to
the Muscle Expert Panel
Are the definitions for muscle complaints specific and
adequate?
●
●
Response: No
Confidence/level of evidence: 1A (Table 1)
The Muscle Expert Panel found that 1 of its
most difficult tasks was to evaluate data developed using
different definitions. Myalgia is a nonspecific, common
complaint and there are no objective, validated measurement tools for 3-hydroxy-3-methylglutaryl coenzyme A
(HMG-CoA) reductase inhibitor (statin) myalgia. Myalgia
has rarely been examined in statin clinical trials.1 Myopathy
has been examined in clinical trials, but the definition of
myopathy varies and has been used to refer to all muscle
complaints2 or to creatine kinase (CK) levels ⬎10 times the
upper limits of normal (ULN) (approximately ⬎2,000 IU/
L),1 with or without associated muscle symptoms. Myositis
has been defined as muscle symptoms with increased CK
levels.2 This term implies muscle inflammation, but inflammatory infiltrates are not present early during statin-induced
muscle injury and appear to be a secondary event associated
with the healing process.3
Rhabdomyolysis by strict definition exists whenever
there is evidence of muscle damage, such as a mildly elevated CK level. Rhabdomyolysis is used clinically, however, to refer to severe muscle damage, but with varying
definitions for severity, and usually associated with renal
dysfunction. The National Cholesterol Education Program
(NCEP) Advisory Panel defines rhabdomyolysis as a CK
level ⬎10 times the ULN with renal compromise,2 whereas
the US Food and Drug Administration (FDA) requires a CK
level ⬎50 times normal (or ⬎10,000 IU/L) with organ
RATIONALE.
a
Division of Cardiology, Hartford Hospital, Hartford, Connecticut,
USA; bUniversity of Massachusetts at Amherst, Amherst, Massachusetts,
USA; cNorthwestern University, Chicago, Illinois, USA.
*Address for reprints: Paul D. Thompson, MD, Division of Cardiology,
Hartford Hospital, 80 Seymour Street, Hartford, Connecticut 06102.
E-mail address: pthomps@harthosp.org.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.013
damage, usually renal compromise,4 and a recent trial5 required only a CK level ⬎50 times normal (10,000 IU/L)
with or without azotemia. Consequently, the Muscle Expert
Panel believes a reexamination of the definitions used to
define muscle complaints with lipid-lowering therapy is
advisable and suggests the following schema:
●
Myopathy should be used as the general term for all
the potential muscle problems listed below.
● Symptomatic myopathy should be used to refer to complaints referable to skeletal muscle including myalgia
(muscle pain), weakness (by complaint or objective
testing), and cramps.
● Asymptomatic myopathy should be used to refer to CK
elevations without symptoms or objective evidence of
weakness.
● Clinically important rhabdomyolysis should be used to
refer to any evidence of muscle cell destruction or
enzyme leakage, regardless of the CK level when
measured, considered to be causally related to a
change in renal function. Some degree of muscle
breakdown or rhabdomyolysis exists whenever there is
any evidence of muscle cell destruction or enzyme
leakage evidenced most commonly by an increase in
serum CK levels above normal. Consequently, the
Muscle Expert Panel suggests that the term rhabdomyolysis, which has been variously defined, be replaced by classes of absolute CK elevation including:
— Mild CK increase should be used to refer to CK levels
greater than normal, but ⬍10 times the ULN.
— Moderate CK increase should be used to refer to CK
levels ⱖ10 times the ULN but ⬍50 times the ULN.
— Marked CK increase should be used to refer to CK
levels ⱖ50 times the ULN. (The Muscle Expert Panel
would like to emphasize that even CK elevations ⱖ50
times the ULN do not necessarily portend a serious
outcome, and that such elevations are observed, for
example, after muscle-damaging exercise,6,7 often
with no ill effects.8)
Use of such a CK classification would eliminate different
definitions of rhabdomyolysis and increase the capture of
www.AJConline.org
70C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 1
Scales for assigning confidence and type of evidence* codes to the
answers given to task force questions
Scale
Description
Confidence
1
2
3
4
Type of
evidence
A
B
C
D
U
Very confident
Confident
Marginally confident
Not confident
● Well-designed RCTs, including RCTs conducted in
patients who reported adverse experiences
● Single RCT with a highly statistically significant result
● Well-conducted retrospective case-control studies with
adverse experiences as primary end points
● Managed care claims database analysis with a highly
statistically significant result
● Reports to regulatory agencies judged to exceed
population averages and reporting bias
● Multiple case studies with nonblinded dechallenge and
rechallenge
● Strong trends, not reaching statistical significance, for
safety issues in large RCTs
● Well-conducted prospective cohort study giving a
result that is statistically well above population
average
● Metabolic or clinically surrogate studies
● Undocumented opinion of experienced research
investigators and clinicians
● Poorly controlled or uncontrolled studies
● Nondefinitive evidence from regulatory agency
reporting systems or managed care claims databases
● Unknown, no appropriate evidence, or evidence
considered subject to bias
RCT ⫽ randomized controlled clinical trial.
*Support for evidence for or against contention that a potential human
adverse experience is related to use of statins.
mild CK increases in clinical trials. This group is presently
ignored or not reported in most study designs.
The Muscle Expert Panel recognizes that average baseline CK levels are higher in African Americans9 and in men9
and recommends that race/ethnicity– and sex-specific values be used for ULN.
Are the muscle complaints related to statin therapy
manifestations of the same process?
●
●
Response: Yes
Confidence/level of evidence: 3D
RATIONALE. The Muscle Expert Panel finds that there is
insufficient evidence to address this question with conviction. Nevertheless, it would be unusual for different pathologic processes to produce the various muscle manifestations, and more likely that the different manifestations are
caused by individual differences in susceptibility, pain tolerance, and other factors.
Is the myopathy and rhabdomyolysis associated with
statin therapy a class effect?
●
●
Response: Yes
Confidence/level of evidence: 1C
RATIONALE. The Muscle Expert Panel believes that these
adverse events are a class effect as demonstrated by the
observation that muscle toxicity has been reported with all
of the currently available statins as well as with cerivastatin,1 which was withdrawn from the US market in August
2003.10 Reports of myopathy with other lipid-lowering
agents, including rare reports with niacin,11 fibric acid derivatives,12–14 and even ezetimibe,15 used as monotherapy
also raise the possibility that muscle toxicity is an effect of
lipid reduction per se and is not limited to statin therapy, but
the frequency of such problems with these drugs in clinical
practice is much less than with statins. The experience with
cerivastatin suggests that statins do vary in their myotoxicity, but there are no direct comparisons among statins as to
their myotoxic potential, nor are there comparisons between
statins and other classes of lipid-lowering agents. Consequently, conclusions as to the relative toxicity of lipidlowering therapies are not possible, and the absence of
controlled clinical trials comparing statins, dictates a 1C
level of evidence.
Is the myopathy and rhabdomyolysis associated with
statin therapy dependent on the following?
● Statin dose
— Response: Yes
— Confidence/level of evidence: 1A
● Blood levels
— Response: Yes
— Confidence/level of evidence: 1C
● Hydrophilicity
— Response: Uncertain
— Confidence/level of evidence: 4C
● Cytochrome metabolism
— Response: Yes
— Confidence/level of evidence: 2B
● Glucuronidation, half-life
— Response: Yes
— Confidence/level of evidence: 3B
● Degree of low-density lipoprotein
(LDL) cholesterol
reduction
— Response: No
— Confidence/level of evidence: 1A
RATIONALE. The Muscle Expert Panel affirms that statinassociated muscle complaints have been documented to
increase with increasing serum concentration in humans16
and in animal models. Interestingly, the Panel could find no
direct evidence relating intramuscular statin concentrations
to myopathy, even though most experts consider intramuscular statin levels critical to the myopathic process. Nevertheless, factors increasing statin concentrations in blood,
and possibly in muscle, are likely to increase statin-related
muscle complaints. These include the statin dose and concomitant medications interfering with statin metabolism via
Thompson et al/Report of the Muscle Expert Panel
71C
Table 2
Evaluation of Statin Prescribing information
Patients aged ⬎70 yr
Impaired renal function ⬍30 mL
Impaired hepatic function
Concurrent Therapy with CYP3A4
Inhibitor
Diabetes with proteinuria
Baseline CK 2–5⫻ ULN
I/S
Vigorous exercise
Asian ethnic groups
Lovastatin
Pravastatin
Simvastatin
Fluvastatin
Atorvastatin
Rosuvastatin
✓
✓
✓
✓
NA
NA
✓
NA†
0
✓
✓
✓
NA
NA
✓
NA†
✓
NA
✓
✓
NA
✓
✓
NA†
0
0
✓S
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
✓I
0
0
0
0
✓I
0
✓
CK ⫽ creatine kinase; CYP3A4 ⫽ cytochrome P450 3A4; I ⫽ illness; NA ⫽ not available; S ⫽ surgery; ULN ⫽ upper limit of normal; ✓ ⫽ addressed
in package insert.
*Table was created August 2005 from package inserts provided for the respective branded statins.
†
Statin levels are increased with cyclosporin therapy, probably via a non-CYP3A4 mechanism.
either the cytochrome P450 (CYP) or glucuronidation processes. Simvastatin,17 pravastatin,18 and rosuvastatin19 at
doses double those currently marketed produced unacceptable rates of muscle damage and provide unequivocal evidence of the adverse effect of increasing drug dose on
skeletal muscle. Marked increases in CK levels were also
more frequent with simvastatin when used in patients post–
myocardial infarction at 80 mg/day vs 40 mg/day.5
Statins are transported into the liver by the organic anion
transporters (OAT).20 Transport into muscle, however, requires passive diffusion through the lipid-rich sarcolemma
because muscle cells lacks OAT.20 Consequently, hydrophilicity should theoretically decrease statin entry into the
skeletal muscle and reduce muscle damage, although there
are no direct comparisons of muscle complaints with hydrophilic and lipophilic statins to confirm this hypothesis.
Also, cases of moderate rhabdomyolysis have been reported
with both pravastatin1,18 and rosuvastatin,19 the 2 most hydrophilic statins, attesting to the fact that hydrophilicity
does not guarantee protection against muscle damage.
The magnitude of LDL cholesterol reduction has not
predicted the frequency of statin-associated muscle complaints in clinical trials or clinical experience.
Does the current labeling appropriately describe the
risk of myopathy and rhabdomyolysis under the following circumstances?
● Elderly patients (ie, ⬎70 years of age)
— Response: Yes
— Confidence/level of evidence: 2C
● Impaired renal function (ie, creatinine clearance
⬍30
mg/dL)
— Response: Yes
— Confidence/level of evidence: 1C
● Impaired liver function
— Response: Yes
— Confidence/level of evidence: 1C
● Concurrent therapy with CYP3A4 –inhibiting drugs or
substrates
Response: Yes
Confidence/level of evidence: 1C
● Diabetic proteinuria
— Response: Appropriately not addressed by labeling
— Confidence/level of evidence: 2C
● Baseline CK levels 2–5 times the ULN
— Response: Appropriately not addressed by labeling
— Confidence/level of evidence: 3C
● During acute illness or major surgery
— Response: Appropriately not addressed by labeling
— Confidence/level of evidence: 3C
● During vigorous exercise (ie, marathon)
— Response: Appropriately not addressed by labeling
— Confidence/level of evidence: 4C
● In certain ethnic/racial groups, such as Asians
— Response: No
— Confidence/level of evidence: 2C
—
—
The Muscle Expert Panel finds that the
available statins appear to be appropriately labeled based on
the available evidence for most of the above potential risk
factors. Labeling varies among the statins because the
statins differ in their metabolism and their risks for muscle
damage are not similarly affected by the above factors
(Table 2). None of the statin labels addressed the issue of
diabetic proteinuria, primarily because this is not considered
a risk factor for myotoxicity. No label addressed mild baseline CK elevations or the theoretical risk of vigorous, sustained endurance exercise.21–23 Several labels address the
potential risk of concurrent illness and some address the
potential risk of surgery, although the latter is based primarily on anecdotal reports.24
RATIONALE.
Is an elevation of CK without other evidence of muscle injury, such as weakness, indicative of statin-induced
muscle damage?
●
●
Response: Yes
Confidence/level of evidence: 3D
72C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Is muscle weakness or pain despite normal CK levels
indicative of stain-induced muscle damage?
●
●
Response: Yes
Confidence/level of evidence: 3D
Is a muscle biopsy recommended to determine
whether muscle damage has occurred in patients with
either muscle symptoms or increased CK levels or
both?
●
●
Response: No
Confidence/level of evidence: 3D
RATIONALE. The Muscle Expert Panel believes that a CK
elevation, even in the absence of other symptoms such as
myalgia or weakness, does represent muscle damage. It is
not known, however, whether such findings portend more
serious muscle problems in the future, although clinical
experience has not demonstrated that CK elevations alone
are associated with adverse long-term sequelae.
Muscle biopsies in a very small group of selected patients (n ⫽ 3) with muscle weakness and myalgia demonstrated intramuscular lipid accumulation as well as histologic and mitochondrial changes suggestive of a
mitochondrial myopathy.25 These results require confirmation in a larger sample.
There is insufficient information to recommend muscle
biopsies to determine muscle damage in patients with possible statin myopathy, and the Muscle Expert Panel recommends a clinical approach to treating and evaluating myopathic patients. This clinical approach includes cessation of
statin therapy, observation for symptom and CK resolution,
and possible repeat challenge to determine whether symptoms reappear. Muscle biopsies in patients with persistent
myopathy after statin withdrawal may be useful, depending
on the clinical circumstances and in consultation with experts in muscle diseases.
Is the risk of causing myopathy or rhabdomyolysis in
a patient using statin increased with the addition of any
of the following drugs to the regimen?
● Gemfibrozil
— Response: Yes
— Confidence/level of
● Fenofibrate
— Response: Yes
— Confidence/level of
● Bile acid resins
— Response: No
— Confidence/level of
● Crystalline niacin,
evidence: 1B
evidence: 4C
evidence: 1C
slow-release niacin, extended-
release niacin
— Response: Yes
— Confidence/level of evidence: 4D
● Cholesterol absorption inhibitor (ezetimibe)
— Response: No
— Confidence/level of evidence: 4D
●
Omega-3 fatty acids
Response: No
— Confidence/level of evidence: 1C
● Plant sterol and stanols
— Response: No
— Confidence/level of evidence: 2C
—
RATIONALE. The Muscle Expert Panel believes that the
best way to determine whether other agents interact with
statins to increase myopathy is via randomized, controlled
clinical trials with subjects assigned to statin treatment with
or without the potentially exacerbating agent. Such trials are
virtually nonexistent. In addition, among available clinical
trials, serious statin-induced muscle damage, such as that
usually reported in such trials, is a rare event even with
medications that increase the risk, and participants in such
trials are selected to minimize the use of concomitant medications affecting statin metabolism. Consequently, most of
the above conclusions on the increased statin myopathic risk
with concomitant medication use are based on data from the
FDA’s Adverse Event Reporting System (AERS)1 and on
examination of managed care databases.
There is strong epidemiologic evidence from clinical
reports,26 AERS,1 and managed care databases13 buttressed
by metabolic study results,27,28 that the risks of muscle
injury during statin therapy increase with concomitant gemfibrozil therapy. There also is evidence that fenofibrate increases the risk of statin myopathy and rhabdomyolysis.
These data are less conclusive, in part, because the epidemiologic results are not supported by metabolic pathway
data.27
Graham and colleagues13 examined claims data from 11
managed care health plans that included 252,460 patients
treated with lipid-lowering drugs, and identified only 24
cases of lipid-lowering, drug-associated rhabdomyolysis. Of
these, 16 cases were associated with statin monotherapy and
8 cases were associated with combination statin-fibrate therapy. The risk of rhabdomyolysis with fibrate monotherapy
was 5.5 times higher than the risk with statin monotherapy.
This risk was entirely due to gemfibrozil use, because there
were no cases of rhabdomyolysis with fenofibrate. There
were cases of combination statin-fenofibrate therapy, but
the number of such cases was not provided. Similarly, Gaist
and colleagues,14 using data from general practices in the
United Kingdom, concluded that the incidence rate of myopathy with lipid-lowering therapy was 5.6 times higher
with fibrate monotherapy than with statin monotherapy and
that fenofibrate monotherapy was associated with the greatest risk.
These conclusions, that the risk of myopathy is greatest
with fibrate monotherapy and that fenofibrate is as dangerous as gemfibrozil in combination therapy, differ from most
experts’ clinical experience and from physiologic studies
suggesting that fenofibrate, in contrast to gemfibrozil, does
not impede statin glucuronidation.27 The explanation for the
observation of increased risk with fibrate monotherapy in
Thompson et al/Report of the Muscle Expert Panel
general and fenofibrate in particular is unclear, but it may
represent unreported concomitant statin use. For example,
in the Graham report, 7 additional cases of muscle damage
were excluded from the analysis because statin use was not
recorded in the pharmacy database, although all 7 patients
were receiving statin therapy at the time of their muscle
injury according to the medical record.13 Such unreported
use can result from physician samples and the use by individuals of other patient’s prescriptions. Nevertheless, such
data cannot be discounted without additional studies, and
they account for the Expert Panel’s decision to acknowledge
additional risk from concomitant fibrate therapy.
Evidence supporting an increased risk with niacin preparations29,30 and ezetimibe15 are weak, largely anecdotal,
and not confirmed by larger studies or the weight of evidence. To the Panel’s knowledge, there is no evidence of
increased risk for bile acid sequestrants, omega-3 fatty acids, or plant sterol and stanols.
Recommendations of the Muscle Expert Panel
Recommendations to regulatory authorities: The
Muscle Expert Panel’s primary recommendation to regulatory authorities, and specifically the FDA, is that the agency
should require industry to evaluate more carefully minor
statin-related muscle problems including myalgia, weakness, and cramps, even when they occur in the absence of
CK elevations. The Panel believes that this evaluation
would facilitate comparisons among the statins for these
important, albeit not life-threatening, side effects. The minor effects of myalgia and cramps, for example, affect
quality of life as well as adherence to these beneficial drugs.
The Muscle Expert Panel is concerned that collection of
only marked CK elevations, a notoriously rare event, may
obscure minor differences among the drugs. In addition, the
Panel suggests that CK elevations in clinical studies be
recorded and reported, again to facilitate comparison among
the statins and to better evaluate the frequency of this
problem.
Recommendations to healthcare professionals: DEFIThe Muscle Expert Panel recommends that the
definitions of statin-associated muscle complaints be standardized using the recommendations presented in this report. It is extremely difficult to define muscle problems and
to compare results among studies when the definitions vary
and when data for mild symptoms and small CK elevations
are not collected and reported.
PATIENT MONITORING. The Muscle Expert Panel did not
consider a baseline CK level absolutely necessary.31 Baseline CK values may be useful to determine whether increased CK levels on statin therapy are due to the drug or to
other causes,2 but similar information can be obtained by
statin withdrawal. Baseline CK values should be strongly
considered for patients at increased risk for myopathy such
as those with renal or hepatic dysfunction or those on
medications that might affect statin metabolism.
NITIONS.
73C
The Muscle Expert Panel recommends that myalgia
symptoms be monitored in patients during statin therapy.
The Panel suggests some caution with using this approach
as a routine procedure, however, because frequent inquires
may prompt symptoms in suggestible patients. The Panel
also recommends that healthcare providers be aware that
myalgia is a side effect of statin therapy and that statins
should be considered in the differential diagnosis of the
cause of myalgia in symptomatic patients. The Panel considered myalgia to refer to muscle discomfort that was
provoked by statin therapy and that resolved within 2
months of discontinuation of the medication. Based on
clinical experience, statin-related myalgia is usually symmetrical, involves large proximal muscle groups, and resolves with discontinuation of the medication. The discomfort can appear anytime during statin therapy, even years
after initiation of treatment. There is insufficient evidence to
conclude whether myalgia that persists after the cessation of
statin therapy is caused by the medications.
The Muscle Expert Panel does not advocate routinely
measuring or monitoring CK levels in asymptomatic patients because marked, clinically important CK elevations
from statins alone are rare; most CK elevations during statin
therapy are benign and related to such factors as recent
physical exertion, and there is no evidence that the added
cost of such monitoring improves medical care.31
The Muscle Expert Panel advocates CK measurement in
symptomatic subjects to gauge the severity of muscle damage and to facilitate a decision about whether to continue
therapy. The Panel also recommends that all symptomatic
patients on statin therapy have an evaluation of thyroid
function, because hypothyroidism can decrease statin catabolism, as well as a search for exacerbating factors such as
concomitant medications that reduce statin metabolism,
over-the-counter herbal remedies such as red rice fungus,
which contains lovastatin and can produce myopathy,32 and
grapefruit juice consumption, which impedes statin
catabolism.33
In the absence of exacerbating factors, and if the patient
has intolerable muscle symptoms, the Muscle Expert Panel
recommends that the statin be discontinued regardless of
CK level until the patient is asymptomatic. Once the patient
is asymptomatic, the same statin can then be restarted at the
same dose to test the reproducibility of symptoms, at a
lower dose with or without other lipid-lowering mediations,
or another statin can be used instead of the offending agent.
Reoccurrence of symptoms with multiple statins and statin
doses requires the use of other lipid-lowering agents. There
is no direct comparison of tolerability among the statins and
therefore no definitive evidence to recommend a given statin on the basis of its chemical properties.
If the patient has tolerable muscle complaints and no
(CK less than the ULN) or mild CK elevation (CK ⬍10
times the ULN) as defined above, the Muscle Expert Panel
recommends that statin therapy can be continued at the
same or reduced doses with symptoms used as the clinical
74C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
guide to stop or continue therapy. If the patient has tolerable
muscle complaints but moderate or severe CK elevations or
clinically important rhabdomyolysis as previously defined,
statin therapy should be stopped and the risks and benefits
of statin therapy carefully reconsidered. In rare instances,
such as those in which the patient has another cause of
muscle injury, the physician and patient may decide to
continue statin therapy despite the CK increase because the
benefits are deemed to outweigh the risks.
STRATEGIES TO PREVENT MUSCLE INJURY. The Muscle
Expert Panel does not recommend prophylactic therapy
with coenzyme Q10 to reduce muscle injury for the following reasons. Statins and HMG-CoA reductase inhibition
block production of farnesyl pyrophosphate (FPP).34 FPP is
an intermediary for the production of ubiquinone, or coenzyme Q10, a steroid isoprenoid that participates in electron
transport during oxidative phosphorylation in mammalian
mitochondria. Serum ubiquinone levels decrease with statin
treatment because ubiquinone is transported in the LDL
particle.35 Serum and intramuscular ubiquinone levels do
not correlate with each other, suggesting different regulatory mechanisms for blood and muscle ubiquinone levels.36
Intramuscular levels of ubiquinone are not usually reduced
by low-dose statin treatment (simvastatin 20 mg37 or 20 – 40
mg38 daily) in nonmyopathic patients,37–39 although a recent
report comparing 80-mg doses of simvastatin and atorvastatin found reductions in mitochondrial volume and intramuscular ubiquinone levels with the simvastatin therapy.40
Also, 17 of 36 of patients (47%) referred for myopathic
complaints attributed to statin therapy had intramuscular
ubiquinone levels ⬎2 standard deviations below the reference mean (G. Vladitu, personal communication, August
2005).
It is not clear from these reports, however, whether the
reduction in intramuscular ubiquinone concentrations is
caused by loss of mitochondria volume or is the actual cause
of the mitochondrial dysfunction.40 In animal models of
statin myopathy, muscle degeneration precedes the mitochondrial dysfunction,41 suggesting that changes in ubiquinone levels are a consequence of the basic myopathic process. Also, intramuscular ubiquinone levels are not
necessarily different between myopathic and nonmyopathic
animals.41 To our knowledge, only 1 study has examined
the effect of ubiquinone replacement therapy on symptoms
in patients treated with statins. Kelly and colleagues42 randomly assigned 41 patients with statin myalgia to daily
doses of vitamin E 400 mg or coenzyme Q10 100 mg. After
30 days of therapy, there was no change in pain in the
vitamin E group, but there was a significant reduction in
pain in the subjects receiving coenzyme Q10.42 Pain was
reduced in 3 of the 20 vitamin E group patients and 18 of the
21 coenzyme Q10 group subjects. These results have only
been presented in preliminary form and require confirmation. The Muscle Expert Panel’s clinical experience suggests that the response to coenzyme Q10 therapy is variable,
and that this therapy cannot be recommended with confidence at the present time.
As discussed, there are theoretical arguments supporting
the concept that hydrophilic agents have less muscle toxicity, but no direct comparisons of statins to justify recommending hydrophilic agents over lipophilic agents.
Recommendations to patients: Statins are used to reduce blood cholesterol and have been shown in multiple
studies to reduce the risk of heart attacks, heart attack
deaths, and strokes. The benefits of treating blood cholesterol with statins have been demonstrated in a wide variety
of patient groups, including healthy patients, patients with
previous heart disease, diabetes mellitus, and hypertension,
patients with a recent heart attack, patients after angioplasty,
and elderly patients. In all patient groups treated to date, the
health benefits of these medications outweigh their risks.
The major medical problem with statins is that they can
produce muscle aching, weakness, and cramps in some
patients. These symptoms are usually mild and can be tolerated, but should be discussed with a physician if they
occur. These complaints rarely lead to any important medical problem, but once in a great while these drugs produce
such severe muscle damage that they can cause kidney
failure and even death. The best estimate suggests that for
every 15 million prescriptions there is only 1 occurrence of
severe muscle damage. Nevertheless, patients should discuss any new muscle discomforts, muscle weakness, or
muscle cramps with their physician. Also, a patient started
on new medications, should inform his or her physician and
pharmacist about the use of the statin, because some medications can increase the risk of muscle injury with statins.
Some over-the-counter medications, specifically Chinese
red rice fungus, contain statins and should not be taken with
the prescription medication. In addition, patients should
avoid drinking or eating a lot of grapefruit products, because
grapefruit can increase statin blood levels.
Recommendations to researchers, funding agencies,
and pharmaceutical companies: The Panel makes the following recommendations to researchers, funding agencies,
and the pharmaceutical industry:
●
●
●
More research attention in clinical trials should be
given to the more mild symptoms of statin myopathy
including myalgia, weakness, cramps, and CK increases ⬍10 times the ULN, because these problems
are considerably more frequent than more severe muscle injury, and because statin tolerability greatly affects patient adherence to treatment and patient quality
of life.
A validated measurement instrument for mild statin
myalgia and other muscle complaints must be developed to facilitate examination of this problem in future
clinical trials.
Precise measurements of muscle strength such as
handgrip, elbow flexor, and knee extensor strength
Thompson et al/Report of the Muscle Expert Panel
●
●
●
●
●
●
●
●
should be incorporated into research programs evaluating statin therapy. Without such measurements in
multiple patient groups, it will be difficult to determine
the frequency with which statins affect muscular performance.
Comparisons among statins should be performed in
patients groups enriched for the potential of muscle
toxicity to determine if different statins have different
frequencies of mild muscle complaints.
Measurement of muscle injury, in addition to CK values, must be evaluated and used in clinical trials to
detect muscle damage not necessarily reflected in increased CK levels.
Additional studies are needed to determine whether
this problem is restricted to, or primarily affects, patients treated with statins, because database examinations and anecdotal reports suggest that myopathy can
occur with other lipid therapies.
Mechanistic studies must be developed to determine
the mechanism by which statins, and possibly other
lipid-lowering agents, affect skeletal muscle.
Clinical models of statin-induced muscle injury, such
as the combination of exercise and statins,23 should be
validated as indicative of spontaneous muscle injury
and used to evaluate individual susceptibility, differences among treatment strategies, and prevention.
Because muscle weakness is a major problem in an
increasingly aged population, long-term studies of statin
treatment and muscle performance are required to determine whether prolonged statin use is ultimately associated with muscle weakness, physical frailty, or disability.
Possible preventive techniques for statin myopathy
such as ubiquinone supplementation should be rigorously evaluated in controlled clinical trials.
Additional statins that do not enter, and therefore affect, skeletal muscle must be developed to reduce
further the possibility of muscle intolerance and to
permit effective treatment of patients who presently
cannot tolerate these medications.
1. Thompson PD, Clarkson P, Karas RH. Statin-associated myopathy.
JAMA 2003;289:1681–1690.
2. Pasternak RC, Smith SC, Bairey-Merz CN, Grundy SM, Cleeman JI,
Lenfant C. ACC/AHA/NHLBI clinical advisory on the use and safety
of statins. J Am Coll Cardiol 2002;40:567–572.
3. Pierce LR, Wysowski DK, Gross TP. Myopathy and rhabdomyolysis
associated with lovastatin-gemfibrozil combination therapy. JAMA
1990;264:71–75.
4. Ballantyne CM, Corsini A, Davidson MH, Holdaas H, Jacobson TA,
Leitersdorf E, Marz W, Reckless JP, Stein EA. Risk for myopathy with
statin therapy in high-risk patients. Arch Intern Med 2003;163:553–
564.
5. de Lemos JA, Blazing MA, Wiviott SD, Lewis EF, Fox KA, White
HD, Rouleau JL, Pedersen TR, Gardner LH, Mukherjee R, et al. Early
intensive vs a delayed conservative simvastatin strategy in patients
with acute coronary syndromes: phase Z of the A to Z trial. JAMA
2004;292:1307–1316.
6. Clarkson PM, Hubal MJ. Exercise-induced muscle damage in humans.
Am J Phys Med Rehabil 2002;81:S52–S69.
75C
7. Sayers SP, Clarkson PM, Rouzier PA, Kamen G. Adverse events
associated with eccentric exercise protocols: six case studies. Med Sci
Sports Exerc 1999;31:1697–1702.
8. Clarkson PM, Kearns AK, Rouzier P, Rubin R, Thompson PD. Serum
creatine kinase levels and renal function measures in exertional muscle
damage. Med Sci Sports Exerc. In press.
9. Black HR, Quallich H, Gareleck CB. Racial differences in serum
creatine kinase levels. Am J Med 1986;81:479 – 487.
10. Fuhrmans V. Bayer discloses higher death toll from Baycol. Wall
Street Journal. January 21, 2002: A10.
11. Gharavi AG, Diamond JA, Smith DA, Phillips RA. Niacin-induced
myopathy. Am J Cardiol 1994;74:841– 842.
12. Langer T, Levy R. Acute muscular syndrome associated with administration of clofibrate. N Engl J Med 1968;279:856 – 858.
13. Graham DJ, Staffa JA, Shatin D, Andrade SE, Schech SD, La Grenade
L, Gurwitz JH, Chan KA, Goodman MJ, Platt R. Incidence of hospitalized rhabdomyolysis in patients treated with lipid-lowering drugs.
JAMA 2004;292:2585–2590.
14. Gaist D, Rodriguez LA, Huerta C, Hallas J, Sindrup SH. Lipidlowering drugs and risk of myopathy: a population-based follow-up
study. Epidemiology 2001;12:565–569.
15. Fux R, Morike K, Gundel UF, Hartmann R, Gleiter CH. Ezetimibe and
statin-associated myopathy. Ann Intern Med 2004;140:671– 672.
16. East C, Alivizatos PA, Grundy SM, Jones PH, Farmer JA. Rhabdomyolysis in patients receiving lovastatin after cardiac transplantation.
N Engl J Med 1988;318:47– 48.
17. Davidson MH, Stein EA, Dujovne CA, Hunninghake DB, Weiss SR,
Knopp RH, Illingworth DR, Mitchel YB, Melino MR, Zupkis RV, et
al. The efficacy and six-week tolerability of simvastatin 80 and 160
mg/day. Am J Cardiol 1997;79:38 – 42.
18. Rosenson RS, Bays HE. Results of two clinical trials on the safety and
efficacy of pravastatin 80 and 160 mg per day. Am J Cardiol 2003;
91:878 – 881.
19. The statin wars: why AstraZeneca must retreat [editorial]. Lancet
2003;362:1341.
20. Hsiang B, Zhu Y, Wang Z, Wu Y, Sasseville V, Yang WP, Kirchgessner TG. A novel human hepatic organic anion transporting
polypeptide (OATP2): identification of a liver-specific human organic
anion transporting polypeptide and identification of rat and human
hydroxymethylglutaryl-CoA reductase inhibitor transporters. J Biol
Chem 1999;274:37161–37168.
21. Thompson PD, Nugent AM, Herbert PN. Increases in creatine kinase
after exercise in patients treated with HMG Co-A reductase inhibitors
[letter]. JAMA 1990;264:2992.
22. Thompson PD, Gadaleta PA, Yurgalevitch S, Cullinane E, Herbert PN.
Effects of exercise and lovastatin on serum creatine kinase activity.
Metabolism 1991;40:1333–1336.
23. Thompson PD, Zmuda JM, Domalik LJ, Zimet RJ, Staggers J, Guyton
JR. Lovastatin increases exercise-induced skeletal muscle injury. Metabolism 1997;46:1206 –1210.
24. Rosenberg AD, Neuwirth MG, Kagen LJ, Singh K, Fischer HD,
Bernstein RL. Intraoperative rhabdomyolysis in a patient receiving
pravastatin, a 3-hydroxy-3-methylglutaryl coenzyme A (HMG CoA)
reductase inhibitor. Anesth Analg 1995;81:1089 –1091.
25. Phillips PS, Haas RH, Bannykh S, Hathaway S, Gray NL, Kimura BJ,
Vladutiu GD, England JD. Statin-associated myopathy with normal
creatine kinase levels. Ann Intern Med 2002;137:581–585.
26. Shek A, Ferrill MJ. Statin-fibrate combination therapy. Ann Pharmacother 2001;35:908 –917.
27. Prueksaritanont T, Zhao JJ, Ma B, Roadcap BA, Tang C, Qiu Y, Liu
L, Lin JH, Pearson PG, Baillie TA. Mechanistic studies on metabolic
interactions between gemfibrozil and statins. J Pharmacol Exp Ther
2002;301:1042–1051.
28. Prueksaritanont T, Tang C, Qiu Y, Mu L, Subramanian R, Lin JH.
Effects of fibrates on metabolism of statins in human hepatocytes.
Drug Metab Dispos 2002;30:1280 –1287.
76C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
29. Hill MD, Bilbao JM. Case of the month: February 1999 —54 year old
man with severe muscle weakness. Brain Pathol 1999;9:607– 608.
30. Reaven P, Witztum JL. Lovastatin, nicotinic acid, and rhabdomyolysis. Ann Intern Med 1988;109:597–598.
31. Smith CC, Bernstein LI, Davis RB, Rind DM, Shmerling RH. Screening for statin-related toxicity: the yield of transaminase and creatine
kinase measurements in a primary care setting. Arch Intern Med
2003;163:688 – 692.
32. Smith DJ, Olive KE. Chinese red rice-induced myopathy. South Med
J 2003;96:1265–1267.
33. Dahan A, Altman H. Food-drug interaction: grapefruit juice augments
drug bioavailability—mechanism, extent and relevance. Eur J Clin
Nutr 2004;58:1–9.
34. Flint OP, Masters BA, Gregg RE, Durham SK. Inhibition of cholesterol synthesis by squalene synthase inhibitors does not induce myotoxicity in vitro. Toxicol Appl Pharmacol 1997;145:91–98.
35. Ghirlanda G, Oradei A, Manto A, Lippa S, Uccioli L, Caputo S, Greco
AV, Littarru GP. Evidence of plasma CoQ10-lowering effect by
HMG-CoA reductase inhibitors: a double-blind, placebo-controlled
study. J Clin Pharmacol 1993;33:226 –229.
36. Laaksonen R, Riihimaki A, Laitila J, Martensson K, Tikkanen MJ,
Himberg JJ. Serum and muscle tissue ubiquinone levels in healthy
subjects. J Lab Clin Med 1995;125:517–521.
37. Laaksonen R, Jokelainen K, Laakso J, Sahi T, Harkonen M, Tikkanen
MJ, Himberg JJ. The effect of simvastatin treatment on natural antioxidants in low-density lipoproteins and high-energy phosphates and
ubiquinone in skeletal muscle. Am J Cardiol 1996;77:851– 854.
38. Laaksonen R, Ojala JP, Tikkanen MJ, Himberg JJ. Serum ubiquinone
concentrations after short- and long-term treatment with HMG-CoA
reductase inhibitors. Eur J Clin Pharmacol 1994;46:313–317.
39. Laaksonen R, Jokelainen K, Sahi T, Tikkanen MJ, Himberg JJ. Decreases in serum ubiquinone concentrations do not result in reduced
levels in muscle tissue during short-term simvastatin treatment in
humans. Clin Pharmacol Ther 1995;57:62– 66.
40. Paiva H, Thelen KM, Van Coster R, Smet J, De Paepe B, Mattila KM,
Laakso J, Lehtimaki T, von Bergmann K, Lutjohann D, Laaksonen R.
High-dose statins and skeletal muscle metabolism in humans: a randomized, controlled trial. Clin Pharmacol Ther 2005;78:60 – 68.
41. Schaefer WH, Lawrence JW, Loughlin AF, Stoffregen DA, Mixson
LA, Dean DC, Raab CE, Yu NX, Lankas GR, Frederick CB. Evaluation of ubiquinone concentration and mitochondrial function relative
to cerivastatin-induced skeletal myopathy in rats. Toxicol Appl Pharmacol 2004;194:10 –23.
42. Kelly P, Vasu S, Getato M, McNurlan M, Lawson WE. Coenzyme
Q10 improves myopathic pain in statin treated patients [abstract]. J Am
Coll Cardiol 2005;45:3A.
An Assessment of Statin Safety by Hepatologists
David E. Cohen, MD, PhD,a,* Frank A. Anania, MD,b Naga Chalasani, MDc
The purpose of the Liver Expert Panel was to provide advice to the National Lipid
Association’s (NLA) Safety Task Force in response to specific questions concerning
liver-associated risks of statin therapy. The panel was composed of academic hepatologists with clinical and research interests in nonalcoholic fatty liver disease, lipid
metabolic disorders, and drug hepatotoxicity. © 2006 Elsevier Inc. All rights reserved. (Am J Cardiol 2006;97[suppl]:77C– 81C)
Questions Posed by the National Lipid Association to
the Liver Expert Panel
Are elevations in serum aminotransferase levels associated with 3-hydroxy-3-methylglutaryl coenzyme A
(HMG-CoA) reductase inhibitor, or statin, therapy?
●
●
Response: Yes
Confidence/level of evidence: 1A (Table 1)
RATIONALE. The Liver Expert Panel of the National
Lipid Association (NLA) affirms that there is a relation
between statin therapy and elevations in serum aminotransferase levels (alanine aminotransferase [ALT] and aspartate
aminotransferase [AST]). This has been consistently demonstrated in clinical trials performed during statin phase 2
and 3 development programs and in long-term, end point
trials.1 The prescribing information for each statin cites
these associations. Aminotransferase elevations ⬎3 times
the upper limit of normal generally occur in ⬍1% of patients across the dose range for marketed statins; the exceptions are aminotransferase elevations of this magnitude that
occur in 2%–3% of patients receiving atorvastatin 80 mg/
day or the combination of ezetimibe and a statin.2– 4
Although the relation between statin therapy and aminotransferase elevations appears to be clear-cut, it is difficult
to conclude with certainty that statins are causally related to
these elevations or to be precise about the exact incidence.
Considerable spontaneous fluctuations in aminotransferase
levels occur over time in a population. In multiple studies,
the incidence of aminotransferase elevations was similar in
patients treated with statin or placebo patients. Moreover,
nearly 50% of hyperlipidemic patients have coexisting non-
a
Division of Gastroenterology, Brigham and Women’s Hospital, Harvard Medical School and Harvard-Massachusetts Institute of Technology,
Division of Health Sciences and Technology, Boston, Massachusetts,
USA; bEmory University School of Medicine, Atlanta, Georgia, USA; and
c
Indiana University School of Medicine, Indianapolis, Indiana, USA.
*Address for reprints: David E. Cohen, MD, PhD, Division of Gastroenterology, Brigham and Women’s Hospital, Thorn 1405, 75 Francis
Street, Boston, Massachusetts 02115.
E-mail address: dcohen@partners.org.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.014
alcoholic fatty liver disease (NAFLD), and it is well known
that aminotransferase levels fluctuate in NAFLD.5
Are statin-associated elevations in aminotransferase
levels indicative of liver damage or dysfunction?
●
●
Response: No
Confidence/level of evidence: 2C
RATIONALE. Isolated elevations of aminotransferases in the
absence of increased bilirubin levels have not been linked
clinically or histologically with evidence of acute or chronic
liver injury.6 – 8 Other mechanisms have been proposed that
could explain commonly observed aminotransferase elevations
in individuals treated with statins, including a transient pharmacologic effect secondary to cholesterol reduction in hepatocytes, comorbid conditions such as diabetes mellitus and obesity, and the consumption of alcohol or nonstatin medications.6
Are statin-associated elevations in aminotransferases
a class effect?
●
●
Response: Yes
Confidence/level of evidence: 1A
RATIONALE. The Liver Expert Panel affirms that elevations in aminotransferase levels have been reported with all
doses of all marketed statins and that no particular statin
appears to cause these elevations more frequently than others. This observation is supported by the official product
labeling for each marketed statin2,3,9 –13 and by long-term
randomized end point trials.1 A recent meta-analysis of 13
of these clinical trials, involving 49,275 patients, supports
this assertion.14 Whereas fluvastatin demonstrated statistically significant higher aminotransferase elevations at certain doses in this meta-analysis, the Panel was not persuaded
that this difference is clinically significant.
Does statin therapy increase the incidence of liver
failure, liver transplants or death associated with liver
failure in the general population?
●
●
Response: Yes
Confidence/level of evidence: 2D
www.AJConline.org
78C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 1
Scales for assigning confidence and type of evidence* codes to the
answers given to task force questions
Scale
Confidence
1
2
3
4
Type of
evidence
A
B
C
D
U
Description
Very confident
Confident
Marginally confident
Not confident
● Well-designed RCTs, including RCTs conducted in
patients who reported adverse experiences
● Single RCT with a highly statistically significant result
● Well-conducted retrospective case-control studies with
adverse experiences as primary end points
● Managed care claims database analysis with a highly
statistically significant result
● Reports to regulatory agencies judged to exceed
population averages and reporting bias
● Multiple case studies with nonblinded dechallenge and
rechallenge
● Strong trends, not reaching statistical significance, for
safety issues in large RCTs
● Well-conducted prospective cohort study giving a
result that is statistically well above population
average
● Metabolic or clinically surrogate studies
● Undocumented opinion of experienced research
investigators and clinicians
● Poorly controlled or uncontrolled studies
● Nondefinitive evidence from regulatory agency
reporting systems or managed care claims databases
● Unknown, no appropriate evidence, or evidence
considered subject to bias
RCT ⫽ randomized controlled clinical trial.
*Support for evidence for or against contention that a potential human
adverse experience is related to use of statins.
RATIONALE. Very rare case reports of liver failure have
occurred in patients receiving statin therapy.1,15 Because the
association between statin therapy and liver failure is rare, it
is impossible to directly attribute liver failure to statin usage. Nevertheless, it is possible that these cases do represent
an idiosyncratic reaction to the statin.
Significant liver damage appears to be extremely uncommon with statins, especially when one considers the magnitude of their use worldwide. Based on 232 cases of acute
liver injury potentially associated with lovastatin reported to
Merck’s Worldwide Adverse Event Database (WAES), it
was estimated that risk of liver failure attributable to lovastatin was 2 in 1 million patients.16 In an article in this
supplement, Law and Rudnicka17 estimate that the incidence of statin-associated liver failure is about 1 per million
person-years of use. Of the 51,741 patients who underwent
liver transplantation in the United States between 1990 and
2002, there were 3 patients in whom the procedure was
performed for acute liver failure presumably caused by
statins.15 Of these 3 patients, 2 had acute liver failure while
receiving cerivastatin and 1 had liver failure that was ap-
parently associated with simvastatin. After an extensive
review of the literature, the Liver Expert Panel could find no
direct evidence of death due to liver failure caused by statin
therapy.
The mechanism that underlies the rare association of
acute liver failure and statin therapy is not clear. Statins
have been reported to unmask autoimmune type liver pathology in genetically predisposed individuals,18,19 but this
appears to be very rare. Furthermore, the rare occurrence of
liver failure due to an idiosyncratic reaction is not specific to
statins and has been reported with a number of other commonly used medications (eg, isoniazid, nitrofurantoin).20
In this Panel’s opinion, the evidence presented indicates
that liver failure may occur very rarely with statin therapy.
Reports have described liver failure requiring transplantation, but not deaths due to liver failure. Reports from spontaneous reporting systems and other sources also suggest
that the risk of liver failure is present with any statin, but the
risk is quite remote.
Should liver enzymes and liver function tests be monitored in patients receiving long-term statin therapy?
●
●
Response: No
Confidence/level of evidence: 2B
RATIONALE. The Liver Expert Panel does not believe
that the available scientific evidence supports the routine
monitoring of liver biochemistries in asymptomatic patients
receiving statins.1,6 – 8,15,16,21 The Panel makes this recommendation because (1) irreversible liver damage resulting
from statins is exceptionally rare and is likely idiosyncratic
in nature, and (2) no data exist to show that routine monitoring of liver biochemistries is effective in identifying the
very rare individual who may develop significant liver injury from ongoing statin therapy. In the view of the Panel,
routine monitoring will instead identify patients with isolated increased aminotransferase levels, which could motivate physicians to alter or discontinue statin therapy,
thereby placing patients at increased risk for cardiovascular
events (see “Recommendations to Healthcare Professionals” below).
Whereas it is not fruitful to measure aminotransferase
levels in order to detect an adverse reaction to statin therapy,
it may be prudent to obtain these tests during routine medical evaluations of patients. In this setting, if a patient
receiving a statin is found to have an elevated aminotransferase, it is essential for the physician to exclude other
etiologies such as viral hepatitis, alcohol consumption, or
other medication-related causes (eg, use of nonsteroidal
anti-inflammatory drugs).
Are any of the following conditions a contraindication
for statin therapy?
● Chronic liver disease
— Response: No
— Confidence/level of evidence:
2B
Cohen et al/Report of the Liver Expert Panel
● Compensated cirrhosis
— Response: No
— Confidence/level of evidence: 3D
● Decompensated cirrhosis or acute liver
— Response: Yes
— Confidence/level of evidence: 2D
liver enzymes.7,8 Furthermore, small studies have shown
that statins may actually improve liver histology in patients
with NASH.24 –26
failure
RATIONALE. The Liver Expert Panel believes that neither
chronic liver disease nor compensated cirrhosis should be
considered a contraindication for statin therapy. This position is supported by studies demonstrating that the frequency and degree of aminotransferase elevations were the
same in patients with 1 of these conditions, regardless of
whether they received statin therapy.1,6
Compensated cirrhosis is considered to be present when
individuals have histologic or clinical evidence of cirrhosis
but their liver function is preserved. The prevalence of
compensated cirrhosis in adults in the United States is
estimated to be ⬍1%.22 The Liver Expert Panel did not
identify any scientific evidence to support consideration of
compensated cirrhosis as a contraindication for statin usage.
Several studies have shown that the pharmacokinetics of
various statins are not significantly altered in patients with
Child’s class A cirrhosis.2,3,10 –13 Because patients with compensated cirrhosis may have normal aminotransferase levels, it is the opinion of the Liver Expert Panel that a substantial number of individuals with unsuspected but
compensated cirrhosis have already taken statins over the
years without excessive toxicity.
The Liver Expert Panel believes that decompensated
cirrhosis (ie, cirrhosis associated with impaired liver function) or acute liver failure should remain a contraindication
for statin therapy. However, it is not likely that statin therapy would be indicated in either of these conditions because
lipid-lowering therapy would not likely be considered a relevant option in patients with such a life-threatening illness.21
Can statins be used in patients with NAFLD or nonalcoholic steatohepatitis (NASH)?
●
●
79C
Response: Yes
Confidence/level of evidence: 1B
RATIONALE. The Liver Expert Panel believes that statins
can be used safely in patients with either NAFLD or NASH.
Moreover, individuals with NAFLD or NASH should be
considered important targets for statin therapy because of
their significantly increased cardiovascular risk. There is a
high prevalence of suspected or unsuspected NAFLD in
patients with hyperlipidemia, and it is not uncommon that
aminotransferase levels are normal in patients with
NAFLD.23 Therefore, it is this Panel’s opinion that a large
number of hyperlipidemic patients with unsuspected
NAFLD have already been treated with statins over the
years without significant toxicity. Recent case-control studies have shown that individuals with elevated baseline liver
enzymes and presumed NAFLD are not at higher risk for
statin hepatotoxicity than are those with normal baseline
Recommendations of the Liver Expert Panel
Recommendations to regulatory authorities: Because
there is no evidence that a relation exists between elevated
serum aminotransferase levels and significant liver injury,
or that routine monitoring of liver biochemistries will identify individuals likely to develop rare cases of idiosyncratic
liver failure, the requirement for routine liver biochemistry
monitoring in patients receiving any of the currently marketed statin therapies should be reexamined.1,6 – 8,15,16,21,27
The Liver Expert Panel is concerned that isolated elevations
in aminotransferases may prompt health professionals to
discontinue statin therapy inappropriately in patients otherwise at increased risk for an adverse cardiovascular event.
The Panel is also concerned that patients may be unduly
alarmed by the perceived implications of monitoring and
may choose to discontinue or refuse statin therapy. Finally,
preliminary estimates suggest that the costs associated with
monitoring are very high.6
Recommendations to healthcare professionals: PABefore instituting any type of medical
therapy, it is advisable for the clinician to perform a complete and systematic history, physical examination, and pertinent laboratory testing. If, in the course of this workup,
elevated aminotransferase levels are identified, they should
be investigated in an appropriate fashion. Patients with
chronically abnormal liver biochemical tests should undergo a thorough medical evaluation, and, if indicated, be
referred to a gastroenterologist or hepatologist.
Outside of measuring liver biochemistries for the purpose of periodically updating a patient’s medical history, we
can find no scientific or medical basis for monitoring aminotransferase levels during long-term statin therapy as a
measure to enhance patient safety. We acknowledge that the
Panel’s recommendations are at odds with current prescribing information for marketed statins; however, we are optimistic that the regulatory agencies and pharmaceutical
industry will update their recommendations to be consistent
with evidence-based data cited in this article.
EVALUATION OF A POTENTIAL ADVERSE EVENT. When a
healthcare professional is concerned about the possible occurrence of a hepatotoxic reaction due to statin therapy (eg,
because the patient reports jaundice, malaise, fatigue, lethargy, or related symptoms during treatment), the Liver Expert Panel believes that an assessment of fractionated bilirubin level is advisable. In the absence of biliary
obstruction, bilirubin is a more reliable prognosticator of
liver injury in the setting of drug toxicity.27,28 If the direct
fraction of bilirubin is found to be increased in association
with elevated aminotransferases, it is reasonable to assume
that there is ongoing liver injury and further appropriate
testing should be undertaken to ascertain the etiology.
TIENT MONITORING.
80C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
WHEN TO REDUCE A STATIN DOSE OR DISCONTINUE
THERAPY. There is no evidence that statin therapy should be
altered or discontinued solely on the basis of elevated aminotransferase levels in an asymptomatic patient. Should
more objective evidence of hepatic dysfunction be identified, such as hepatomegaly, clinical evidence of jaundice,
elevated direct bilirubin level, or increased prothrombin
time,20,29 statin therapy should be discontinued. The patient
should be evaluated appropriately and referred to a gastroenterologist or hepatologist, if necessary.
NEW-ONSET LIVER DISEASE IN A PATIENT RECEIVING ONGOING STATIN THERAPY. Once a systematic and complete
medical evaluation reveals significant liver disease in a
patient receiving statin therapy, the etiology should be established. If a causal relation between significant liver injury
and statin therapy cannot be excluded, then reinitiation of
statin therapy is not recommended and alternative lipidlowering strategies should be considered.
ANTICOAGULANT THERAPY AND ELEVATED AMINOTRANSFERASES DUE TO STATINS. There is no evidence that ele-
vated aminotransferase levels due to statins affect response
to anticoagulant therapy. Therefore, no modification in statin therapy is recommended.
STATIN THERAPY AND ALCOHOL CONSUMPTION. Mild-tomoderate alcohol consumption (ie, up to 1–2 drinks per day)
is not a contraindication for statin therapy.23
Recommendations to patients: The class of cholesterol-lowering medications called statins has the ability to
lower the risk of a heart attack, stroke, and the need for
hospital-based heart procedures by 25%–50%. Fortunately,
the side effects associated with these drugs occur very
infrequently. Side effects that affect the liver are rare. While
taking statin medications, some blood tests traditionally
obtained by physicians to monitor the liver may be elevated,
but these test results do not indicate that the statin medication is causing serious liver problems. Serious liver damage
due to statins is exceptionally rare. It is important to appreciate that a number of other commonly prescribed medications can cause similar reactions (eg, antibiotics, seizure
medications).
Recommendations to researchers, funding agencies,
and pharmaceutical companies: To promote and optimize
appropriate use of statins in the dyslipidemic population, the
Liver Expert Panel encourages research in the following
areas:
●
●
●
Pharmacogenomics of statin-associated aminotransferase elevations to clarify why some patients experience
elevations and others do not
Potential benefits of statin therapy in fatty liver disease, as demonstrated in preliminary studies24 –26
Impact of statin therapy on the natural progression of
cirrhosis and fibrosis
As indicated above, it is important that pharmaceutical
manufacturers of statin products work with regulatory au-
thorities to modify recommendations for patient monitoring.
In addition, pharmaceutical companies should carefully assess both the positive and negative effects that direct-toconsumer advertising has on the patient’s understanding of
statins.
Pharmaceutical companies are encouraged to release the
results of clinical research performed with lipid-altering
therapies as a part of the New Drug Application (NDA) for
market approval. Future drug development should include a
process for clinician-based peer-review of suspected hepatic
events identified in clinical trials. This process would include but not be limited to the following:
●
●
Communication with general practitioners regarding
suspected hepatic adverse events and encouraging
them to report adverse events (AEs) to the US Food
and Drug Administration (FDA) via MedWatch
(http://www.fda.gov/medwatch/)
Education concerning AEs that would incorporate (1)
procedures for systematic evaluation of an AE, (2)
possible etiology of an AE, and (3) correct use of
validated instruments in the assessment of causality
1. Chalasani N. Statins and hepatotoxicity: focus on patients with fatty
liver. Hepatology 2005;41:690 – 695.
2. Lipitor (atorvastatin) [package insert]. New York, NY: Pfizer Inc.;
2004.
3. Vytorin (ezetimibe-simvastatin) [package insert]. North Wales, PA:
Merck/Schering-Plough Pharmaceuticals; 2005.
4. Zetia (ezetimibe) [package insert]. North Wales, PA: Merck/ScheringPlough Pharmaceuticals; 2005.
5. Mofrad P, Contos MJ, Haque M, Sargeant C, Fisher RA, Luketic VA,
Sterling RK, Shiffman ML, Stravitz RT, Sanyal AJ. Clinical and
histologic spectrum of nonalcoholic fatty liver disease associated with
normal ALT values. Hepatology 2003;37:1286 –1292.
6. Sniderman AD. Is there value in liver function test and creatine
phosphokinase monitoring for statin use? Am J Cardiol 2004;
94(suppl):30F–34F.
7. Vuppalanchi R, Teal E, Chalasani N. Patients with elevated baseline
liver enzymes do not have higher frequency of hepatoxicity from
lovastatin than those with normal baseline liver enzymes. Am J Med
Sci 2005:329:62– 65.
8. Chalasani N, Aljadhey H, Kesterson J, Murray MD, Hall SD. Patients
with elevated liver enzymes are not at higher risk for statin hepatotoxicity. Gastroenterology 2004;126:1287–1292.
9. Mevacor (lovastatin) [package insert]. Whitehouse Station, NJ: Merck
& Co. Inc.; 2005.
10. Zocor (simvastatin) [package insert]. Whitehouse Station, NJ: Merck
& Co. Inc.; 2004.
11. Lescol (fluvastatin) [package insert]. East Hanover, NJ: Novartis;
2003.
12. Pravachol (pravastatin) [package insert]. Princeton, NJ: Bristol-Myers
Squibb; 2004.
13. Crestor (rosuvastatin) [package insert]. Wilmington, DE: AstraZeneca;
2005.
14. de Denus S, Spinler SA, Miller K, Peterson AM. Statins and liver
toxicity: a meta analysis. Pharmacotherapy 2004;24:584 –591.
15. Russo MW, Galanko JA, Shrestha R, Fried MW, Watkins P. Liver
transplantation for acute liver failure from drug induced liver injury in
the United States. Liver Transpl 2004;10:1018 –1023.
16. Gotto AM. Safety and statin therapy. Reconsidering the risks and
benefits. Arch Intern Med 2003;163:657– 659.
Cohen et al/Report of the Liver Expert Panel
17. Law MR, Rudnicka AR. Statin safety: evidence from the published
literature. Am J Cardiol 2006;97(suppl 8A):52C– 60C.
18. Pelli N, Setti M, Ceppa P, Toncini C, Indiveri F. Autoimmune hepatitis
revealed by atorvastatin. Eur J Gastroenterol Hepatol 2003;15:921–924.
19. Siddiqui J, Raina D, Abraham A, Alla V, Chaslasani N, Wu GY,
Bonkovsky HL. Autoimmune hepatitis induced by statins [abstract].
Gastroenterology 2005;128:A171.
20. Tolman KG. The liver and lovastatin. Am J Cardiol 2002;89:1374–1380.
21. Smith CC, Bernstein LI, Davis RB, Rind DM, Shmerling RH. Screening for statin-related toxicity: The yield of transaminase and creatine
kinase measurements in a primary care setting. Arch Intern Med
2003;163:688 – 692.
22. Sandler RS, Everhart JE, Donowitz M, Adams E, Cronin K, Goodman C,
Gemmen E, Shah S, Avdic A, Rubin R. The burden of selected digestive
diseases in the United States. Gastroenterology 2002;122:1500 –1511.
23. Neuschwanter-Tetri BA, Caldwell SH. Nonalcoholic steatohepatitis: summary of an AASLD single topic conference. Hepatology 2003;37:1202–1219.
81C
24. Rallidis LS, Drakoulis CK, Parasi AS. Pravastatin in patients with
nonalcoholic steatohepatitis: results of a pilot study. Atherosclerosis
2004;174:193–196.
25. Horlander JC, Kwo PY, Cummings OW, Koukoulis G. Atorvastatin
for the treatment of NASH [abstract]. Gastroenterology 2001;120:
A544.
26. Kiyici M, Gulten M, Gurel S, et al. Ursodeoxycholic acid and atorvastatin in the treatment of nonalcoholic steatohepatitis. Can J Gastroenterol 2003;17:713–718.
27. Zimmerman HJ. Hepatotoxicity. The Adverse Effects of Drugs and
Other Chemicals on the Liver. New York: Appleton-Century-Crofts;
1978:181–185, 363–364.
28. Bjornsson E, Olsson R. Outcome and prognostic markers in severe
drug-induced liver disease. Hepatology 2005;42:481– 489.
29. Senior JR. Regulatory perspective. In: Kaplowitz N, DeLeve LD eds.
Drug Induced Liver Disease. New York: Marcel Dekker, Inc; 2003:
739 –754.
An Assessment of Statin Safety by Nephrologists
Bertram L. Kasiske, MD,a,* Christoph Wanner, MD,b and W. Charles O’Neill, MDc
Recently, concerns regarding potential adverse effects of the statins on the kidney
have been raised. The Kidney Expert Panel of the National Lipid Association’s (NLA)
Safety Task Force, made up of 3 nephrologists, was convened to review all of the
currently available evidence pertinent to determining whether statins cause kidney
injury, independent of the known, rare mechanisms of rhabdomyolysis and allergic,
drug-induced, interstitial nephritis. The Panel reviewed published and unpublished
evidence and found none that suggested that statins, when used in doses currently
approved by the US Food and Drug Administration (FDA), cause kidney
injury. © 2006 Elsevier Inc. All rights reserved. (Am J Cardiol 2006;97[suppl]:
82C– 85C)
Questions Posed the National Lipid Association to the
Renal Expert Panel:
Do the 3-hydroxy-3-methylglutaryl coenzyme A
(HMG-CoA) reductase inhibitors, or statins, cause acute
renal failure (ARF) or renal insufficiency not associated
with rhabdomyolysis or severe myopathy?
●
●
Response: No
Confidence/level of evidence: 1B (Table 1)
RATIONALE. The Renal Expert Panel finds no evidence
that statins cause ARF or renal insufficiency not associated with rhabdomyolysis. One case study1 reported
acute interstitial nephritis in a patient taking a statin.
Although case reports rarely prove cause and effect, it is
possible that a statin could cause acute, allergic, interstitial nephritis. However, the major cardiovascular disease
end point trials2 have not reported ARF as an adverse
event associated with statins, and we are not aware of
trials showing an increased incidence of ARF compared
with placebo. The US Food and Drug Administration
(FDA) New Drug Application (NDA) data3 and data on
adverse events4 provide little convincing evidence that
statins cause ARF or renal insufficiency that is not associated with rhabdomyolysis.
Do statins cause proteinuria?
●
●
Response: No
Confidence/level of evidence: 3C
RATIONALE. The Renal Expert Panel finds no convincing evidence in humans linking proteinuria with the use
of statins that are currently approved by the FDA. How-
a
University of Minnesota, Minneapolis, Minnesota, USA; bUniversity
of Würzburg, Würzburg, Germany; and cEmory University, Atlanta, Georgia, USA.
*Address for reprints: Bertram L. Kasiske, MD, Department of Medicine, Hennepin County Medical Center, 701 Park Avenue, Minneapolis,
Minnesota 55415.
E-mail address: kasis001@umn.edu.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.015
ever, few if any studies have systematically collected
data on urinary protein excretion; screening for proteinuria with dipsticks may miss mild degrees of proteinuria.
One case study5 linked the use of simvastatin with proteinuria, although baseline proteinuria was not reported,
and whether simvastatin caused the proteinuria could not
be determined in this study. Cardiovascular disease end
point trials have generally not mentioned the occurrence
of proteinuria,2 but it is not well documented how often
proteinuria was measured in these studies. Data from
NDAs do not convincingly show an association between
statins and proteinuria among FDA-approved statins at
their approved doses.
In NDA data, rosuvastatin 80 mg/day was associated
with an increased incidence of ⱖ2⫹ dipstick-positive proteinuria, when compared with placebo, lower doses of rosuvastatin, and other statins.3 In some cases, increased urine
protein excretion was confirmed by measuring protein/creatinine ratios (29% had a protein/creatinine ratio ⬎500
mg/g creatinine). On electrophoresis, performed in some
cases, the proteinuria resembled tubular proteinuria. Studies
in cultured tubular cells indicate that statins may block
tubular absorption of protein,6,7 although the relevance of
these studies to clinical proteinuria has yet to be demonstrated. While the rosuvastatin 80-mg/day dose is twice the
highest FDA-approved 40-mg/day dose, the data with the
80-mg/day dose suggest the need for additional studies on
the incidence of proteinuria among patients treated with
40 mg/day.
Do statins cause renal tubular damage?
●
●
Response: No
Confidence/level of evidence: 2C
RATIONALE. The Renal Expert Panel finds no association
between renal tubular damage and statin use. There have
been no case reports linking statins to renal tubular acidosis
or other measures of tubular damage.
www.AJConline.org
Kasiske et al/Report of the Renal Expert Panel
Table 1
Scales for assigning confidence and type of evidence* codes to the
answers given to task force questions
Scale
Description
Confidence
1
2
3
4
Type of
evidence
A
B
C
D
U
Very confident
Confident
Marginally confident
Not confident
● Well-designed RCTs, including RCTs conducted in
patients who reported adverse experiences
● Single RCT with a highly statistically significant result
● Well-conducted retrospective case-control studies with
adverse experiences as primary end points
● Managed care claims database analysis with a highly
statistically significant result
● Reports to regulatory agencies judged to exceed
population averages and reporting bias
● Multiple case studies with nonblinded dechallenge and
rechallenge
● Strong trends, not reaching statistical significance, for
safety issues in large RCTs
● Well-conducted prospective cohort study giving a
result that is statistically well above population
average
● Metabolic or clinically surrogate studies
● Undocumented opinion of experienced research
investigators and clinicians
● Poorly controlled or uncontrolled studies
● Nondefinitive evidence from regulatory agency
reporting systems or managed care claims databases
● Unknown, no appropriate evidence, or evidence
considered subject to bias
RCT ⫽ randomized controlled clinical trial.
*Support for evidence for or against contention that a potential human
adverse experience is related to use of statins.
Do statins cause a reversible interference with protein
transport across tubular cells that may result in proteinuria?
●
●
Response: No
Confidence/level of evidence: 3U
RATIONALE. The Renal Expert Panel finds no convincing
data in humans suggesting that statin use causes a reversible
interference with protein transport across renal tubular cells
that may result in proteinuria. Although studies in cultured
tubular cells indicate that statins may interfere with protein
transport,6,7 there have not yet been in vivo studies to
determine whether this occurs in humans.
Do statins cause renal glomerular damage or dysfunction?
●
●
Response: No
Confidence/level of evidence: 1U
The Renal Expert Panel finds no evidence
supporting an association of statin use with renal glomerular
RATIONALE.
83C
damage or dysfunction. There have been no reports of
histologic changes in the renal glomerulus associated with
statin use. Similarly, there have been no reports of nephrotic-range proteinuria. It is reasonable to expect that if
statin use caused nephrotic-range proteinuria, this would
have been reported among the several thousand patients
treated with statins in clinical trials.
Do statins cause hematuria?
●
●
Response: No
Confidence/level of evidence: 2C
RATIONALE. The Renal Expert Panel finds no convincing
evidence linking hematuria with the use of statins that are
currently approved by the FDA. Cardiovascular disease end
point trials generally have not mentioned the occurrence of
hematuria,2 but it is not well documented how often tests for
hematuria were performed in these studies. Data from
NDAs do not convincingly demonstrate an association between statins and hematuria among FDA-approved statins at
their approved doses.
In NDA data, rosuvastatin 80 mg/day was associated
with an increased incidence of dipstick-positive hematuria,
when compared with placebo, lower doses of rosuvastatin,
and other statins.3 In some of these cases, hematuria was
confirmed by microscopic examination of the urine sediment. Although the rosuvastatin 80-mg/day dose is twice
the highest FDA-approved 40-mg/day dose, the data with
the 80-mg/day dose suggest the need for additional studies
on the incidence of hematuria among patients treated with
40 mg/day.
Is there evidence that statins cause chronic kidney
disease (CKD)?
●
●
Response: No
Confidence/level of evidence: 1B
RATIONALE. The Renal Expert Panel finds no evidence
that statins cause CKD. In fact, several small, randomized,
controlled trials have found that statins may slow the rate of
decline in kidney function.8 –10 In addition, post hoc analyses of large cardiovascular disease end point trials have also
suggested that statins may slow the rate of decline in function as estimated by serum creatinine.11–13
Should patients be routinely monitored for proteinuria and/or renal function while they are receiving a
statin?
●
●
Response: No
Confidence/level of evidence: 1C
RATIONALE. The Renal Expert Panel concludes that patients need not be routinely monitored for proteinuria and/or
renal function while they are receiving a statin.
Can statins be used safely in patients with CKD,
whether or not they are treated by hemodialysis? Are
certain statins preferred in CKD?
84C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 2
Lipid-lowering medication dose adjustments for reduced kidney function
Adjust for Reduced GFR (mL/min per 1.73 m2)
Agent
60–90
Atorvastatin
Fluvastatin
Lovastatin
Pravastatin
Simvastatin
Nicotinic acid
Cholestipol
Cholestyramine
Colesevelam
Bezafibrate
Clofibrate
Ciprofibrate
Fenofibrate
Gemfibrozil
No
No
No
No
?
No
No
No
No
2 to 50%
2 to 50%
?
2 to 50%
No
15–59
No
?
2 to
No
?
No
No
No
No
2 to
2 to
?
2 to
No
50%
25%
25%
25%
Notes
⬍15
No
?
2 to 50%
No
?
2 to 50%
No
No
No
Avoid
Avoid
?
Avoid
No
—
—
—
—
—
34% Kidney excretion
Not absorbed
Not absorbed
Not absorbed
May 1 serum creatinine
May 1 serum creatinine
May 1 serum creatinine
May 1 serum creatinine
May 1 serum creatinine
GFR ⫽ glomerular filtration rate; 2 ⫽ decrease; 1 ⫽ increase.
Adapted with permission from Am J Kidney Dis.14
●
●
Response: Yes
Confidence/level of evidence: 1B
RATIONALE. The Renal Expert Panel finds that statins
can be used safely in patients with CKD. Guidelines have
been established by the National Kidney Foundation (Table
2).14 Pharmacokinetic studies suggest that the blood levels
of some statins may be increased in patients with CKD and
recommend that dose adjustments should be made. In addition, caregivers should be vigilant for possible drug interactions that may increase blood levels of statins, with this
occurring in some statins more than others.14 Recent randomized controlled trials have confirmed the safety of fluvastatin in recipients of kidney transplant15 and of atorvastatin in patients with diabetes on hemodialysis.16
Recommendations of the Renal Expert Panel
Recommendations to regulatory authorities: The Renal Expert Panel recommends that, in the future, the assessment of statin safety should include more accurate and comprehensive testing for urine protein excretion, with spot
protein/creatinine ratios and/or albumin/creatinine ratios (or
timed urine collections to quantitate protein and/or albumin
excretion). In addition, patients with dipstick-positive hematuria should have a microscopic examination of the urinary
sediment for semiquantitative assessment of red blood cells.
Although existing data on rosuvastatin indicate that
40 mg/day is safe, the adverse effects of rosuvastatin
80 mg/day narrow the therapeutic window and increase the
possibility that adverse effects of 40 mg/day exist but have
not been detected. Therefore, the Renal Expert Panel recommends that additional data should be gathered to assess
the effects of rosuvastatin 40 mg/day, if any, on proteinuria
and hematuria.
Recommendations to healthcare professionals: PAMeasurement of albumin/creatinine ratio and serum creatinine at baseline to identify patients who
are at increased risk for cardiovascular disease and CKD
(not to assess the risk of an adverse event with statin
therapy) is recommended. Microalbuminuria (albumin/creatinine ratio ⬎30 mg/g creatinine) and/or elevated serum
creatinine are independent risk factors for cardiovascular
disease. Modifiable risk factors, such as dyslipidemia and
hypertension, should be treated more intensively in patients
who are at increased risk for cardiovascular disease. Routine
serum creatinine or other measurements of renal function,
including dipstick urine tests, are not required to monitor
chronic statin therapy.
EVALUATION OF A POTENTIAL ADVERSE EVENT. A clinician should respond to an increase in serum creatinine (or
glomerular filtration rate estimated with serum creatinine)
discovered in the routine care of a patient receiving statin
therapy with potential rhabdomyolysis in mind. The serum
creatinine measurement should be repeated, and, if it remains elevated and unexplained, referral to a nephrologist
should be considered for the diagnosis and management of
ARF. The statin dose need not be altered if there is no
evidence of rhabdomyolysis.
Should dipstick-positive proteinuria be detected in the
routine care of a patient receiving statin therapy, a spot
urinary protein quantification with protein/creatinine ratio
and/or albumin/creatinine ratio (or timed urine collection to
measure protein and/or albumin excretion) should be obtained. If quantification confirms clinical proteinuria (eg,
total protein ⬎500 mg/g creatinine or albumin ⬎300 mg/g
creatinine), referral to a nephrologist should be considered.
The statin dose need not be altered, nor is it necessary to
discontinue the statin therapy.
TIENT MONITORING.
Kasiske et al/Report of the Renal Expert Panel
WHEN TO REDUCE A STATIN DOSE OR DISCONTINUE
THERAPY. If there is no evidence of muscle damage due to
rhabdomyolysis, statin therapy need not be discontinued. It
is also not necessary to avoid statin therapy in CKD; however, the dose of some statins may need to be reduced
(Table 2). A change in kidney function during statin therapy
does not necessitate a change in the type of statin used.
Recommendations to patients: Statins used according
to directions do not appear to have any direct adverse effects
on the kidney. Statins can be used safely in patients with
CKD; however the dose of some statins may need to be
modified. The benefits of statin therapy far outweigh the
negative effects, if any, on the kidney.
Recommendations to researchers, funding agencies,
and pharmaceutical companies: The NDA data with regard to rosuvastatin 80 mg/day suggest that additional studies on hematuria and proteinuria for rosuvastatin 40 mg/day
would be prudent. These studies should include quantitative
testing of proteinuria and microscopic screening for hematuria. Similarly, any future statin NDAs should include
detailed and comprehensive data on proteinuria and
hematuria.
Studies in humans on the possible effects of statins on
tubular reabsorption of proteins and other tubular functions
are warranted. Studies on the safety and efficacy of statins
(both for reduction of risk of cardiovascular events and also
for slowing of kidney disease progression) in patients with
CKD are needed, and some are currently in progress.17,18
Future studies should include quantitative measurements of
proteinuria and screening for microscopic hematuria (not
just dipstick testing).
Acknowledgment
The authors thank James Bass for his expert assistance in
writing the manuscript.
1. van Zyk-Smit R, Firth JC, Duffield M, Marais AD. Renal tubular
toxicity of HMG-CoA reductase inhibitors. Nephrol Dial Transplant
2004;19:3176 –3179.
2. Cheung BM, Lauder IJ, Lau CP, Kumana CR. Meta-analysis of large
randomized controlled trials to evaluate the impact of statins on cardiovascular outcomes. Br J Clin Pharmacol 2004;15:640 – 651.
3. Jacobson TA. Statin safety: lessons from New Drug Applications for
marketed statins. Am J Cardiol 2006;97(suppl 8A):44C–51C.
85C
4. Law M, Rudnicka AR. Statin safety: evidence from the published
literature. Am J Cardiol 2006;97(suppl 8A):52C– 60C.
5. Deslypere JP, Delanghe J, Vermeulen A. Proteinuria as complication
of simvastatin treatment [letter]. Lancet 1990;336:1453.
6. Verhulst A, D’Haese PC, De Broe ME. Inhibitors of HMG-CoA
reductase reduce receptor-mediated endocytosis in human kidney
proximal tubular cells. J Am Soc Nephrol 2004;15:2249 –2257.
7. Sidaway JE, Davidson RG, McTaggart F, Orton TC, Scott RC, Smith
GJ, Brunskill NJ. Inhibitors of 3-hydroxy-3-methylglutaryl-CoA reductase reduce receptor-mediated endocytosis in opossum kidney
cells. J Am Soc Nephrol 2004;15:2258 –2265.
8. Fried LF, Orchard TJ, Kasiske BL. The effect of lipid reduction on
renal disease progression: a meta-analysis. Kidney Int 2001;59:260 –
269.
9. Bianchi S, Bigazzi R, Caiazza A, Campese VM. A controlled, prospective study of the effects of atorvastatin on proteinuria and progression of kidney disease. Am J Kidney Dis 2003;41:565–570.
10. Kano K, Nishikura K, Yamada Y, Arisaka O. Effect of fluvastatin and
dipyridamole on proteinuria and renal function in childhood IgA nephropathy with mild histological findings and moderate proteinuria.
Clin Nephrol 2003;60:85– 89.
11. Tonelli M, Moyé L, Sacks F, Cole T, Curhan GC, for the Cholesterol
and Recurrent Events (CARE) Trial Investigators. Effect of pravastatin
on loss of renal function in people with moderate chronic renal insufficiency and cardiovascular disease. J Am Soc Nephrol 2003;14:1605–
1613.
12. Athyros VG, Mikhailidis DP, Papageorgiou AA, Symeonidis AN,
Pehlivanidis AN, Bouloukos VI, Elisaf M. The effect of statins versus
untreated dyslipidaemia on renal function in patients with coronary
heart disease: a subgroup analysis of the Greek Atorvastatin and
Coronary Heart Disease Evaluation (GREACE) study. J Clin Pathol
2004;57:728 –734.
13. Vidt DG, Cressman MD, Harris S, Pears JS, Hutchinson HG. Rosuvastatin-induced arrest in progression of renal disease. Cardiology
2004;102:52– 60.
14. Kasiske B, Cosio FG, Beto J, Bolton K, Chavers BM, Grimm R Jr,
Levin A, Masri B, Parekh R, Wanner C, Wheeler DC, Wilson PW.
K/DOQI clinical practice guidelines for managing dyslipidemias in
chronic kidney disease. Am J Kidney Dis 2003;41(suppl 3):S1–S91.
15. Holdaas H, Fellstrom B, Jardine AG, Holme I, Nyberg G, Fauchald P,
Gronhagen-Riska C, Madsen S, Neumayer HH, Cole E, et al, for the
Assessment of LEscol in Renal Transplantation (ALERT) Study Investigators. Effect of fluvastatin on cardiac outcomes in renal transplant recipients: a multicentre, randomised, placebo-controlled trial.
Lancet 2003;361:2024 –2031.
16. Wanner C, Krane V, März W, Olschewski M, Mann JFE, Ruf G, Ritz
E, for the German Diabetes and Dialysis Study Investigators. Atorvastatin in patients with type 2 diabetes mellitus undergoing hemodialysis. N Engl J Med 2005;353:238 –248.
17. Fellstrom BC, Holdaas H, Jardine AG. Why do we need a statin trial
in hemodialysis patients? Kidney Int Suppl 2003;63(suppl 84S):S204 –
S206.
18. Baigent C, Landray M. Study of Heart and Renal Protection (SHARP).
Kidney Int Suppl 2003;63(suppl 84S):S207–S210.
An Assessment of Statin Safety by Neurologists
Lawrence M. Brass, MD,a,† Mark J. Alberts, MD,b,* and Larry Sparks, PhDc
The National Lipid Association’s (NLA) Statin Safety Task Force charged the Neurology Expert Panel with the task of reviewing the scientific evidence related to
adverse effects with statins and providing assessments and advice regarding the
safety of statins. The evidence included key adverse reaction statin literature identified via a Medline search by the Task Force and Panel members and the commissioned reviews and research presented in this supplement. Panel members were asked
to use this evidence to independently form explicit answers to a series of questions
posed by the Task Force. Panelists were asked to grade the type of literature and the
confidence they had in it in forming their answers using prescribed scales. Panelists
were encouraged to seek the highest level of evidence available to answer their
questions and to concentrate on literature involving humans. In addition, the Neurology Expert Panel was asked to propose recommendations to regulatory authorities,
health professionals, patients, researchers, and the pharmaceutical industry to address statin safety issues. © 2006 Elsevier Inc. All rights reserved. (Am J Cardiol
2006;97[suppl]:86C– 88C)
Questions Posed by the National Lipid Association to
the Neurology Expert Panel
Do the 3-hydroxy-3-methylglutaryl coenzyme A
(HMG-CoA) reductase inhibitors, or statins, cause peripheral neuropathy in some patients?
●
●
Response: No
Confidence/level of evidence: 2B (Table 1)
There is no evidence that statins are a common or significant cause of peripheral neuropathy. This
position is supported by the lack of an association found in
large, randomized, controlled, clinical trials including the
Heart Protection Study (HPS) and the Prospective Study of
Pravastatin in the Elderly at Risk (PROSPER).1,2 The HPS
was the largest statin trial to date (N ⫽ 20,536), and was a
placebo-controlled 5-year trial of simvastatin 40 mg/day, in
which there was no evidence of peripheral neuropathy.
PROSPER compared pravastatin 40 mg/day with placebo in
5,804 elderly patients aged 70 – 82 years and found no
evidence of statin-related peripheral neuropathy. A small
number of epidemiologic and case studies have suggested
an association between statins and peripheral neuropathy,
but, as noted above, this has not been shown in large clinical
trials.3– 6 It is possible that a rare case of peripheral neuropathy, which is otherwise unexplained, could occur in paRATIONALE.
Yale University, New Haven, Connecticut, USA; bNorthwestern University, Chicago, Illinois, USA; and cSun Health Research Institute, Sun
City, Arizona, USA.
†
Deceased.
*Address for reprints: Mark J. Alberts, MD, Northwestern University
Feinberg School of Medicine, 710 North Lake Shore Drive, Chicago,
Illinois 60611.
E-mail address: m-alberts@northwestern.edu.
tients administered statin therapy; this would most likely
represent an idiosyncratic reaction.
Do statins impair memory or cognition in some patients?
●
●
Response: No
Confidence/level of evidence: 1B
RATIONALE. There is no evidence of a causal relation
between impaired memory and/or cognition dysfunction
and statin therapy. This position is supported by the lack of
demonstration of an association in large, randomized, controlled, clinical trials, including the HPS and PROSPER.1,2
Two additional studies have specifically evaluated the effect
of statin therapy on patients with Alzheimer disease, a
population at risk for cognitive decline. In the first trial,
simvastatin 80 mg/day was administered to patients (n ⫽
44) with Alzheimer disease over 26 weeks with no evidence
of a change in cognitive function compared with placebo.7
In the second trial, atorvastatin 80 mg/day administered to
patients with Alzheimer disease (n ⫽ 71) for 3 months
showed a statistically significant reduction in the rate of
cognitive decline compared with placebo, suggesting a benefit for atorvastatin in Alzheimer disease.8 Case reports
suggest that a rare individual may experience memory or
cognitive impairment while receiving a statin, but this most
likely represents an idiosyncratic effect.9
a
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.017
Recommendations of the Neurology Expert Panel
Recommendations to regulatory authorities: Steps
should be taken to obtain more detail on specific adverse
event (AE) cases reported to regulatory agencies. This
would allow adjudication of causality between the AE and
www.AJConline.org
Brass et al/Report of the Neurology Expert Panel
Table 1
Scales for assigning confidence and type of evidence* codes to the
answers given to task force questions
Scale
Confidence
1
2
3
4
Type of
evidence
A
B
C
D
U
Description
Very confident
Confident
Marginally confident
Not confident
● Well-designed RCTs, including RCTs conducted in
patients who reported adverse experiences
● Single RCT with a highly statistically significant result
● Well-conducted retrospective case-control studies with
adverse experiences as primary end points
● Managed care claims database analysis with a highly
statistically significant result
● Reports to regulatory agencies judged to exceed
population averages and reporting bias
● Multiple case studies with nonblinded dechallenge and
rechallenge
● Strong trends, not reaching statistical significance, for
safety issues in large RCTs
● Well-conducted prospective cohort study giving a
result that is statistically well above population
average
● Metabolic or clinically surrogate studies
● Undocumented opinion of experienced research
investigators and clinicians
● Poorly controlled or uncontrolled studies
● Nondefinitive evidence from regulatory agency
reporting systems or managed care claims databases
● Unknown, no appropriate evidence, or evidence
considered subject to bias
RCT ⫽ randomized controlled clinical trial.
*Support for evidence for or against contention that a potential human
adverse experience is related to use of statins.
a specific drug. The Neurology Expert Panel believes that to
improve standardization and reporting of potential neurologic complications, specific and precise definitions for cognitive impairment and peripheral neuropathy should be developed. We believe that combining all cognitive
impairment complaints or neuropathic symptoms is likely to
mask a true effect or any signal, should one exist.
Recommendations to healthcare professionals: PATIENT MONITORING. Specific or routine neurologic monitoring of patients administered statin therapy for neurologic
changes indicative of peripheral neuropathy or impaired
cognition is not recommended. However, should symptoms
occur in patients receiving statins, a thorough neurologic
evaluation is indicated, with due consideration of preexisting or comorbid conditions.
EVALUATION OF A POTENTIAL AE. Patients experiencing
peripheral neuropathy or impaired cognition while receiving
a statin should undergo a thorough neurologic evaluation,
including consideration of referral to a neurologist to spe-
87C
cifically assess the symptoms being experienced. This evaluation should seek to determine the cause of the neurologic
symptom or establish it as an idiopathic reaction.
If another etiology of the neurologic symptoms is not
identified, it is appropriate to withdraw statin therapy for a
time to establish whether an apparent association with statin
therapy exists. Because reversible peripheral neuropathies
can take weeks or months to resolve, the patient should
remain off statin therapy for 3– 6 months. For patients with
impaired cognition, we recommend discontinuing statin
therapy for 1–3 months.
If the patient’s symptoms improve while off statin therapy, a presumptive diagnosis of peripheral neuropathy or
impaired cognition associated with statin therapy might be
made. Because of the proven benefit of statin therapy, reinitiation of such therapy, preferably with another statin,
should be considered.
However, if the patient’s neurologic symptoms do not
improve after statin therapy has been withdrawn for the
specified period, the symptoms may be categorized as idiopathic or unrelated to statin therapy. Therefore, statin therapy should be restarted based on a risk-benefit analysis, as
for any clinical circumstance. The potential benefits of statin
therapy should be strongly considered in such cases.
It should be kept in mind that there are multiple potential
causes of peripheral neuropathy and impaired cognition in
patients likely to be treated with statins, including advanced
vascular disease, advancing age, diabetes mellitus, or insulin resistance. Further study is needed to more fully understand the potential relation among peripheral neuropathy,
cognitive impairment, and statin therapy. Some of these
trials are in the planning phase. Clinicians should consider
the well-established potential benefit of statin therapy while
making decisions to withdraw patients from statin therapy
for an interim period. This is especially important when
considering restarting statin therapy at some point in the
future. For patients with neurologic symptoms in whom a
statin agent is reinitiated, the Neurology Expert Panel recommends using pravastatin based solely on expert opinion
(3D). There are no data that support this, and other statin
agents may be reasonable options.
Recommendations to patients: Large scientific studies
do not show that statin therapy causes nerve problems,
impairs memory, or decreases mental function.
Recommendations to researchers, funding agencies,
and pharmaceutical companies: Investigations of statins
should include secondary outcome measures that address
potential AEs. Trial designs should incorporate the use of
standardized instruments to detect and measure the occurrence of AEs. Clinical trials should use appropriate and
well-validated neurologic assessment scales to measure
AEs. Standardized definitions applied with objectivity
should be used in clinical trials and in reporting AEs. Sufficient clinical information should be collected to allow
adequate assessment of the relation between the AE and
88C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
statin therapy. With sufficient detail and objectivity, reports
of AEs could and should be adjudicated with rigor.
5.
Hemorrhagic Stroke and Depression
6.
Although the Neurology Expert Panel was not specifically
charged with evaluating the role of statins in hemorrhagic
stroke or depression, the members believed that these conditions should be addressed. The large landmark statin trials
do not support that lowering lipids with statins is associated
with an increase in risk for cerebral hemorrhage.10 –12
Depression has been suggested as a consequence of lipidlowering therapy, and very low lipid levels have been reported
to increase the risk for suicide.13 For some patients the clinical
conditions necessitating the use of statin therapy are wellknown risks for depression.14 These include stroke and myocardial infarction. Approximately 33%–50% of patients with
stroke will experience cognitive changes or depression after a
stroke. These changes should be completely evaluated in the
context of the patient’s clinical condition but are not by themselves an indication to discontinue statin therapy.
1. Heart Protection Study Collaborative Group. Effects of cholesterol
lowering with simvastatin on stroke and other major vascular events in
20,536 people with cerebrovascular disease or other high-risk conditions. Lancet 2004;363:757–767.
2. Shepherd, J, Blauw GJ, Murphy MB, Bollen ELEM, Buckley BM,
Cobbe SM, Ford I, Gaw A, Hyland M, Jukema JW, et al, for the
PROSPER study group. Pravastatin in Elderly Individuals at Risk of
Vascular Disease (PROSPER): a randomised controlled trial. Lancet
2002;360:1623–1630.
3. Chong PH, Boskovich A, Stevkovic N, Bartt RE. Statin-associated
peripheral neuropathy: review of the literature. Pharmacotherapy
2004;24:1194 –1203.
4. Corrao G, Zambon A, Bertu, Botteri E, Leoni O. Lipid-lowering drugs
prescription and the risk of peripheral neuropathy: an exploratory
7.
8.
9.
10.
11.
12.
13.
14.
case-control study using automated databases. J Epidemiol Community
Health 2004;58:1047–1051.
Gaist D, Garcia Rodriguez LA, Huerta C, Hallas J, Sindrup SH. Are
users of lipid-lowering drugs at increased risk of peripheral neuropathy? Eur J Clin Pharmacol 2001;56:931–933.
Gaist D, Jeppesen U, Andersen M, Garcia Rodriguez LA, Hallas J,
Sindrup SH. Statins and risk of polyneuropathy: a case-control study.
Neurology 2002;58:1333–1337.
Simons M, Schwarzler F, Lutjohann D, von Bergmann K, Beyreuther
K, Dichgans J, Wormstall H, Hartmann T, Schulz JB. Treatment with
simvastatin in normocholesterolemic patients with Alzheimer’s disease: a 26-week randomized, placebo-controlled, double-blind trial.
Ann Neurol 2002;52:346 –350.
Sparks DL, Sabbagh MN, Connor DJ, Lopez J, Launer LJ, Browne P,
Wasser D, Johnson-Traver S, Lochhead J, Ziolwolski C. Atorvastatin
for the treatment of mild to moderate Alzheimer disease. Arch Neurol
2005;62:753–757.
Wagstaff LR, Mitton MW, Arvik BM, Doraiswamy PM. Statin-associated memory loss: analysis of 60 case reports and review of the
literature. Pharmacotherapy 2003;23:871– 880.
Downs JR, Clearfield DO, Weis S, Whitney E, Shapiro DR, Beere PA,
Langendorfer A, Stein EA, Kruyer W, Gotto AM Jr, for the AFCAPS/
TexCAPS [Air Force/Texas Coronary Atherosclerosis Prevention
Study] Research Group. Primary prevention of acute coronary events
with lovastatin in men and women with average cholesterol levels.
JAMA 1998;279:1615–1622.
Scandinavian Simvastatin Survival Study Group. Randomized trial of
cholesterol lowering in 4444 patients with coronary heart disease: the
Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:1383–
1389.
The Long-Term Intervention with Pravastatin in Ischaemic Disease
(LIPID) Study Group. Prevention of cardiovascular events and deaths
with pravastatin in patients with coronary heart disease and a broad
range of initial cholesterol levels. N Engl J Med 1998;339:1349 –1357.
Lindberg G, Rastam L, Gullberg B, Eklund GA. Low serum cholesterol concentration and short term mortality from injuries in men and
women. BMJ 1992;305:277–279.
Morris PL, Raphael B, Robinson RG. Clinical depression is associated
with impaired recovery from stroke. Med J Aust 1992;157:239 –242.
Final Conclusions and Recommendations of the National Lipid
Association Statin Safety Assessment Task Force
James M. McKenney, PharmD,a,* Michael H. Davidson, MD,b Terry A. Jacobson, MD,c and
John R. Guyton, MDd
This article summarizes the final conclusions of the National Lipid Association (NLA)
Statin Safety Task Force, based on a review and independent research of New Drug
Application (NDA) information, US Food and Drug Administration (FDA) Adverse
Event Reporting System (AERS) data, cohort and clinical trial results, and analysis
of administrative claims database information and the assessment of its 4 Expert
Panels, which focused on issues of statin safety with regard to liver, muscle, renal, and
neurologic systems. Practical guidance in the form of recommendations to health
professionals who manage the coronary artery disease risk of patients with statin
therapy is provided. © 2006 Elsevier Inc. All rights reserved. (Am J Cardiol 2006;
97[suppl]:89C–94C)
In the sections that follow, the National Lipid Association
(NLA) Statin Safety Task Force draws from the extensive
evidence so superbly presented and analyzed by the scientists and experts who authored the preceding articles in this
supplement. The Task Force herein offers what it believes to
be a summary of final conclusions that can be made based
on this evidence and provides practical guidance in the form
of recommendations to health professionals who manage
the coronary artery disease risk of patients with statin therapy.
The Liver and Statin Safety
Final conclusions: Asymptomatic elevations in alanine aminotransferase (ALT) or aspartate aminotransferase (AST) liver enzymes ⬎3 times the upper limit of
normal (ULN) are seen with all 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors, or
statins.1 According to data from New Drug Applications
(NDAs) and the prescribing information for each marketed statin, elevations of this magnitude are seen in
⬍1% of patients receiving initial and intermediate doses
and in 2%–3% of patients receiving 80 mg/day.1 It is
evident that these elevations are related to the dose of the
statin but not to the low-density lipoprotein (LDL) cholesterol reduction.2,3 It is also evident that an elevation of
a
National Clinical Research, Virginia Commonwealth University,
Richmond, Virginia, USA; bRush University Medical Center, Chicago,
Illinois, USA; cEmory University, Atlanta, Georgia, USA; and dDuke
University Medical Center, Durham, North Carolina, USA.
*Address for reprints: James M. McKenney, PharmD, National Clinical
Research, 2809 Emerywood Parkway, Suite 140, Richmond, Virginia
23294.
E-mail address: jmckenney@ncrinc.net.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2006.02.030
ALT and/or AST ⬎3 times the ULN is most often transient and will resolve spontaneously in 70% of cases even
if the statin and dose are continued unchanged.2,4 To
more accurately identify patients with a persistent liver
test abnormality, some investigators have adopted a more
rigorous definition, eg, ALT or AST ⬎3 times the ULN
on 2 consecutive occasions. When this definition is applied, the number of patients with a significant elevation
drops from 300 per 100,000 person-years to 110 per
100,000 person-years.4 Reduction in the dose or withdrawal of the statin regularly results in a return of the
elevated enzyme levels to normal without adverse
sequelae.
The cause of an elevation in liver transaminase levels
during statin therapy has not been determined. Generally
in clinical trials, the proportion of patients experiencing
elevations is greater when individuals are given a statin
than when they receive placebo, thus supporting the argument for a statin effect. However, confounding this is
the fact that the population most likely to receive statin
therapy is also the population most likely to experience
liver function changes, including patients with diabetes
mellitus or obesity, older individuals, and patients taking
multiple medications.2
The most relevant question with regard to the liver and
statin safety is not whether statins cause a significant
increase in liver function test results, but whether they
cause serious liver dysfunction or failure. The answer to
this question is not clear, owing in part to the rarity of
these events among statin users. A handful of case reports
have been published that describe liver failure in patients
receiving statin therapy, but a causal relation cannot be
established from these data alone.1 Data from the US
Food and Drug Administration (FDA) Adverse Event
www.AJConline.org
90C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 1
Recommendations to healthcare professionals regarding the liver and statin safety1–5
1. During the routine general evaluation of patients being considered for statin and other lipid-lowering therapy, it is advisable to obtain liver
transaminase levels. If these tests are found to be abnormal, further investigation should be performed to determine the etiology of the abnormal test
results.
2. Until there is a change in the FDA-approved prescribing information for statins, it is appropriate to continue to measure transaminase levels before
starting therapy, 12 weeks after initiating therapy, after a dose increase, and periodically thereafter. However, routine monitoring of liver function
tests is not supported by the available evidence and the current recommendation for monitoring needs to be reconsidered by the FDA.
3. The clinician should be alert to patient reports of jaundice, malaise, fatigue, lethargy, and related symptoms in patients taking statin therapy as
a signal of potential hepatotoxicity. Evidence for hepatotoxicity includes jaundice, hepatomegaly, increased indirect bilirubin level and
elevated prothrombin time (rather than simple elevations in liver transaminase levels).
4. The preferred biochemical test to ascertain significant liver injury is fractionated bilirubin, which, in the absence of biliary obstruction, is a more
accurate prognosticator of liver injury than isolated aminotransferase levels.
5. Should the clinician identify objective evidence of significant liver injury in a patient receiving a statin, the statin should be discontinued. The
etiology should be sought and, if indicated, the patient referred to a gastroenterologist or hepatologist.
6. If an isolated asymptomatic transaminase level is found to be elevated 1–3 times the ULN, there is no need to discontinue the statin.
7. If an isolated asymptomatic transaminase level is found to be ⬎3 times the ULN during a routine evaluation of a patient administering a statin, the
test should be repeated and, if still elevated, other etiologies should be ruled out. Consideration should be given to continuing the statin, reducing its
dose, or discontinuing it based on clinical judgment.
8. According to the Expert Liver Panel, patients with chronic liver disease, nonalcoholic fatty liver disease, or nonalcoholic steatohepatitis may safely
receive statin therapy.1
FDA ⫽ US Food and Drug Administration; ULN ⫽ upper limit of normal.
Reporting System (AERS) database through 1999 included 30 cases of liver failure in individuals taking
statins, for a reporting rate of 1 case per 1 million statin
prescriptions.4 The Merck Worldwide Adverse Event Database (WAES) included 22 cases of liver failure in
patients taking lovastatin, for a rate of 1 case per 1.14
million patients.1 Only 1 of the 51,741 patients who
underwent liver transplantation between 1990 and 2002
was taking a marketed statin.1 These data do not establish
causality. In fact, because the rate of liver failure in a
population not receiving statin therapy is about the same,
it may support a conclusion that there is no relation
between liver failure and statin therapy. Alternatively,
cases of liver failure may represent idiosyncratic reactions that occur very rarely in patients taking statin therapy. In either case, the routine monitoring of liver enzyme levels may not identify these patients.
Based on the evidence, one would have to monitor
transaminase levels in 100,000 patients each year for an
average of 3 years to detect 110 patients who have consecutive elevations in ALT in order to identify the statistical 0.1
person who may experience liver failure, assuming that
statins can cause liver failure in the first place.4 Unfortunately, routine monitoring may lead to a temporary or permanent withdrawal of statin treatment, thus depriving a
considerable number of patients of the life-protecting benefit of statin therapy.
Based on this, the NLA Statin Safety Assessment Task
Force can find no evidence to support the continued
monitoring of liver function tests in patients receiving
statin therapy. However, because of medical–legal issues,
we also believe that the cessation of liver function monitoring is not advisable until changes in the prescribing
information for marketed statins occur. Thus, we recommend a thorough and timely review of these data by
regulatory authorities and the statin manufacturers and, if
found warranted, removal of a recommendation for liver
function monitoring from the prescribing information.
Further, we believe that statin manufacturers who choose
to market their statin directly to the consumer should not
be required to include a caution regarding potential liver
adverse effects as a part of a fair-balance statement. This
only serves to confuse and unnecessarily alarm the public
and potentially to discourage them from pursuing this
life-sustaining therapy, an outcome that would not be in
the public’s best interest.
Recommendations: Table 1 presents our consensus recommendations to health professionals based on the evidence, interpretations, and assessments of liver issues and
statin safety presented in this supplement.1–5
The Muscle and Statin Safety
Final conclusions: Muscle symptoms (ie, pain, soreness, weakness, and/or cramps) or signs (creatine kinase
[CK] elevations) are arguably the most prevalent and
important adverse effect associated with statin therapy.
The occurrence of serious muscle toxicity with currently
marketed statins fortunately is rare.6 According to findings from 21 clinical trials providing 180,000 personyears of follow-up in patients treated with statin or placebo, myopathy (defined as muscle symptoms plus CK
⬎10 times the ULN) occurs in 5 patients per 100,000
person-years and rhabdomyolysis in 1.6 patients per
100,000 person-years (placebo corrected).4 This compares with the reporting rate of 0.3–2.2 cases of myopathy and 0.3–13.5 cases of rhabdomyolysis per million
statin prescriptions from the FDA’s AERS database7 and
with 1.6 –3.5 cases of hospitalized myopathy (including
McKenney et al/Conclusions and Recommendations
rhabdomyolysis) per 10,000 person-years from an analysis of an administrative managed care claims database.5
(Note that these latter data have not been verified with
chart review.) A CK level ⬎10 times the ULN, or ⬎2,000
U/L, was found in 23 patients per 100,000 person-years
in clinical trials; this rate fell to zero when repeat measures were recorded.4
The most common muscle side effects remain myalgia
(ie, muscle pain or soreness), weakness, and/or cramps
without CK elevations.2,4,6 These symptoms are most often
tolerable, but occasionally can be intolerable and debilitating, requiring the statin to be withdrawn. Muscle symptoms
have been reported in clinical trials to occur in 1.5%–3.0%
of patients receiving statin therapy, most often without an
elevation in the CK level, and at an equivalent rate in
patients given placebo.2.4 The incidence of muscle complaints among patients being treated in a practice setting
ranges from 0.3%–33%.2 The higher rate may occur partly
because statin-intolerant patients and those with risk factors
for muscle toxicity are more likely to be excluded from
clinical trials.
Among marketed statins, it appears that the risk of drugrelated muscle injury is roughly the same. All marketed
statins cause the spectrum of muscle injury, but they are
rarely severe, and very rarely progress to a life-threatening
situation.2,4,6,7 Fluvastatin and pravastatin, perhaps because
they are the weakest inhibitors of HMG-CoA reductase,
appear to cause the lowest frequency of rhabdomyolysis;
simvastatin 80 mg (but not lower doses) appears to be
associated with the highest frequency.2,4 The use of more
hydrophilic statins (ie, pravastatin and rosuvastatin) does
not offer protection from muscle toxicity as symptoms of
muscle damage and rhabdomyolysis have been reported
with these statins.2
Cerivastatin was unique among the marketed statins in
that it had unfavorable pharmacokinetic features, the potential for multiple drug interactions, and was marketed at a
dose that exceeded its safety threshold. It caused a 5- to
7-fold greater incidence of muscle damage sequelae, including rhabdomyolysis and death.2,4 Currently marketed statins
do not have the unfavorable features of cerivastatin.
The exact mechanism for muscle injury from statin
therapy is not known. However, it appears to be related to
the blood concentration of the statin, which is influenced
by the drug’s pharmacokinetics and its potential for drug
interactions, the statin dose, and the patient’s myopathic
risk factors (eg, age, renal disease, diabetes), but not by
the LDL cholesterol level achieved. The latter is influenced mostly by the potency of HMG-CoA reductase
inhibition in the hepatocyte.3 Although muscle adverse
effects can occur in patients taking the starting dose of a
statin, symptoms are much more likely to occur with
higher doses. Other situations that may raise the statin’s
blood levels include advanced age and frailty, small body
frame, deteriorating renal function, infection, untreated
hypothyroidism, interacting drugs—particularly with
91C
statins metabolized by the cytochrome P450 system and
gemfibrozil, perioperative periods, and alcohol abuse.2
The theory that these toxicities are related to a reduction
in muscle levels of ubiquinone has not been proved, and
attempts to reduce muscle symptoms with coenzyme Q10
prophylaxis have given equivocal results and cannot be
recommended.6
Recommendations: The NLA Task Force recommends that the generally accepted and widely used definition of myopathy be retained, namely, the presence of
muscle pain, soreness, weakness, and/or cramps plus a
CK level 10 times the ULN (see Table 2).6 Presentation
of muscle symptoms that cannot otherwise be explained
in a statin-taking patient should prompt the measurement
of a CK level. It is not necessary to monitor CK levels in
patients receiving statin therapy. If the CK level is ⬎10
times the ULN, a repeat measure is generally recommended to establish persistency. The Task Force is also
aware that an occasional patient will describe intolerable
muscle symptoms but not be found to have a CK level
⬎10 times the ULN. In this case, the patient may be
presumed to be experiencing myopathy for the purpose of
further evaluation and workup.
The Task Force offers a new definition for rhabdomyolysis. This definition is an attempt to integrate differing
definitions used by the FDA and clinical trialists. The definition is meant to identify the clinical situation where the
risk of acute renal failure and urgent medical intervention is
high. We chose a CK level of ⬎10,000 U/L, in accord with
the definition currently used by the FDA, regardless of
whether the patient has experienced a change in renal function, because such a CK level places the patient at high risk
of acute renal failure. A second component in our definition
is a CK ⬎10 times the ULN with worsening renal function
and/or a requirement for medical intervention with intravenous hydration therapy. We acknowledge that CK levels
may not always be ⬎10 times the ULN in cases of diminishing renal function, especially if the laboratory sample is
drawn some time after the event; thus, this should not be
taken as an absolute criterion.
In Table 3, we present our consensus recommendations
to health professionals based on the evidence, interpretations, and assessments of muscle issues and statin safety
presented in this supplement.2–5
The Kidney and Statin Safety Panel
Final conclusions: In the absence of rhabdomyolysis,
acute renal failure or insufficiency does not appear to be
caused by statin therapy.7 Although case reports of renal
failure have been reported in patients receiving statin
therapy, they are encountered as frequently in patients
receiving statins as in patients not receiving statins.2 In
the 3 pravastatin clinical trials (Cholesterol and Recurrent Events [CARE] trial, Long-Term Intervention with
92C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 2
New definitions to describe muscle findings in patients taking statins6
● Myopathy*
— Complaints of myalgia (muscle pain or soreness), weakness, and/or cramps, plus
— Elevation in serum CK ⬎10⫻ the ULN
● Rhabdomyolysis
— CK ⬎10,000 IU/L, or
— CK ⬎10⫻ the ULN plus an elevation in serum creatinine or medical intervention with IV hydration therapy†
CK ⫽ creatine kinase; IV ⫽ intravenous; ULN ⫽ upper limit of normal.
*A patient may describe intolerable muscle symptoms but not be found to have a CK level ⬎10 times the ULN. This patient may be considered to be
experiencing myopathy for the purposes of further evaluation.
†
The CK level may be ⬍10 times the ULN depending on the temporal relation between the event and the drawing of the laboratory sample.
Table 3
Recommendations to health professionals regarding the muscle and statin safety2–5
1. Whenever muscle symptoms or an increased CK level is encountered in a patient receiving statin therapy, health professionals should attempt to rule
out other etiologies, because these are most likely to explain the findings. Other common etiologies include increased physical activity, trauma, falls,
accidents, seizure, shaking chills, hypothyroidism, infections, carbon monoxide poisoning, polymyositis, dermatomyositis, alcohol abuse, and drug
abuse (cocaine, amphetamines, heroin, or PCP).
2. Obtaining a pretreatment, baseline CK level may be considered in patients who are at high risk of experiencing a muscle toxicity (eg, older
individuals or when combining a statin with an agent known to increase myotoxicity), but this is not routinely necessary in other patients.
3. It is not necessary to measure CK levels in asymptomatic patients during the course of statin therapy, because marked, clinically important CK
elevations are rare and are usually related to physical exertion or other causes.
4. Patients receiving statin therapy should be counseled about the increased risk of muscle complaints, particularly if the initiation of vigorous, sustained
endurance exercise or a surgical operation is being contemplated; they should be advised to report such muscle symptoms to a health professional.
5. CK measurements should be obtained in symptomatic patients to help gauge the severity of muscle damage and facilitate a decision of whether to
continue therapy or alter doses.
6. In patients who develop intolerable muscle symptoms with or without a CK elevation and in whom other etiologies have been ruled out, the statin
should be discontinued. Once asymptomatic, the same or different statin at the same or lower dose can be restarted to test the reproducibility of
symptoms. Recurrence of symptoms with multiple statins and doses requires initiation of other lipid-altering therapy.
7. In patients who develop tolerable muscle complaints or are asymptomatic with a CK ⬍10⫻ the ULN, statin therapy may be continued at the same or
reduced doses and symptoms may be used as the clinical guide to stop or continue therapy.
8. In patients who develop rhabdomyolysis (a CK ⬎10,000 IU/L or a CK ⬎10 times the ULN with an elevation in serum creatinine or requiring IV
hydration therapy), statin therapy should be stopped. IV hydration therapy in a hospital setting should be instituted if indicated for patients
experiencing rhabdomyolysis. Once recovered, the risk vs benefit of statin therapy should be carefully reconsidered.
CK ⫽ creatine kinase; IV ⫽ intravenous; PCP ⫽ phencyclidine; ULN ⫽ upper limit of normal.
Pravastatin in Ischaemic Disease [LIPID] study, and West
of Scotland Coronary Prevention Study [WOSCOPS]),
for example, renal failure and other renal diseases were
reported more frequently in patients who were given
placebo.4 None of the other end point clinical trials with
statins even report cases of renal disease.4 In the FDA
AERS database, the proportional reporting rate for renal
failure is low, generally 0.3– 0.9 cases per 1 million statin
prescriptions.8 In 2005 the FDA undertook the most comprehensive analysis of this topic to date, conducting a
case-by-case review of 38 reports it had received of renal
failure/insufficiency in patients receiving rosuvastatin.2
The FDA reported that it could find no convincing evidence that statin therapy was associated with serious
renal injury, concluding that “no consistent pattern of
clinical presentation or of renal injury (ie, pathology) is
evident among the cases of renal failure reported to date
that clearly indicate causation by Crestor (rosuvastatin;
AstraZeneca, Wilmington, DE) or other statins.”2
Although the evidence that statins cause renal failure
is sparse, other evidence including small randomized
controlled trials and post hoc analyses of large end point
trials suggest that statin therapy may slow the rate of
decline in renal function. For example, post hoc analysis
of the CARE and Greek Atorvastatin and Coronary Heart
Disease Evaluation (GREACE) trials of pravastatin and
atorvastatin, respectively, report improved glomerular
filtration rate (GFR) in patients treated with statin compared with controls.7 Additionally, observation of
⬎10,000 patients with and without diabetes receiving
open-label rosuvastatin for up to 3.8 years revealed no
progressive decline in renal function; instead there was
an improvement in serum creatinine and GFR values.7
These data suggest a potential renal-protective effect
with statins. A number of large end point trials involving
therapy with several different stains are under way in
patients with compromised renal function, including dialysis patients; these should help clarify the role of
statins in preserving renal function.
Proteinuria has been described only rarely with
statins,3 but recently it was found significantly more
frequently in patients receiving rosuvastatin 80 mg than
in patients given placebo.2,3,7 In the same study, the
frequency of proteinuria found with other doses of rosu-
McKenney et al/Conclusions and Recommendations
93C
Table 4
Recommendations to health professionals regarding the kidney and statin safety2– 4,7,8
1. During the management of patients with statin therapy, it is not necessary to carry out serum creatinine and proteinuria monitoring routinely for the
purpose of identifying an adverse effect, although an assessment of renal function is advisable before initiating statin therapy.
2. If serum creatinine becomes elevated in a patient without rhabdomyolysis while receiving statin therapy, there is generally no need to withdraw the
statin but in some cases, according to prescribing information, an adjustment in the statin dose may be required.
3. If unexpected proteinuria develops in a patient receiving a statin, there is no need to withdraw statin therapy or to alter the dose of the statin. An
investigation into the cause of the proteinuria is warranted, as is consideration of a change in the statin dose as guided by the prescribing information
for each statin.
4. Chronic kidney disease does not preclude the use of a statin. However, the dose of some statins should be adjusted in cases of moderate or severe
renal insufficiency7
vastatin (5– 40 mg) currently on the market, as well as
with marketed doses of atorvastatin, pravastatin, and simvastatin, was no different than that found with placebo
allocation.2,7 This latter observation supported our Renal
Expert Panel’s answer of “no” when asked whether
statins cause proteinuria in humans.7 Part of the explanation for proteinuria is that individuals who are candidates for statin therapy often are prone to proteinuria
owing to diabetes, hypertension, or advancing age. Further confounding the interpretation of these data is that
the proteinuria found in clinical trials is often detected
during random spot urine testing with a dipstick in patients participating in long-term, open-label studies that
often lack a placebo comparison group.2
Other evaluations support the suggestion that the proteinuria observed with statin therapy is the result of physiologic interference with protein uptake in renal tubules.2,3 In
vitro studies using an opossum proximal tubular epithelial
kidney cell line in culture demonstrated that all statins can
interfere with protein renal tubular uptake through a concentration-dependent inhibition of HMG-CoA reductase.3
Furthermore, when mevalonate is added to the culture, the
inhibition of protein uptake was reversed, further validating
that the mechanism of the statin’s effect on protein uptake is
dependent on HMG-CoA reductase inhibition.3 Consistent
with this proposed mechanism is the finding that low-molecular-weight protein of renal tubular origin is found in the
urine of these patients.2,3 These studies also illustrate that
proteinuria is at least possible with all statins at some
concentration, but is more likely to be seen with statins that
are potent inhibitors of HMG-CoA reductase.2 In their analysis of these data, the FDA concluded that proteinuria in
patients receiving statins is not associated with renal impairment or renal failure.2
Recommendations: Table 4 includes our consensus recommendations to health professionals based on the evidence, interpretations, and assessments presented in this
supplement regarding the kidney and statin safety.2– 4,7,8
Neurologic Disorders and Statin Safety
Final conclusions: The occurrence of peripheral neuropathy in patients taking a statin is very rare.9 A causal
relation is not supported by the Heart Protection Study
(HPS), a randomized, placebo-controlled clinical trial in
⬎20,000 individuals in which peripheral neuropathy was
recorded in 11 patients who received simvastatin and in
8 patients who received placebo.2 Another large randomized, placebo-controlled trial in elderly patients, the
Pravastatin in Elderly Individuals at Risk of Vascular
Disease (PROSPER) study, reported no evidence of peripheral neuropathy with pravastatin therapy.2 Case-control and cohort studies present conflicting findings, but,
according to a meta-analysis of 4 cohort studies, an odds
ratio of 1.8 (95% confidence interval, 1.1–3.0; p ⫽ 0.001)
was found, favoring the conclusion of a relation between
statin use and peripheral neuropathy.4 The association is
also supported by 16 case reports of peripheral neuropathy in patients taking statins; symptoms of peripheral
neuropathy generally appeared within 2 months of the
initiation of statin therapy and dissipated after withdrawal of the statin.4 In 1 case report, 4 different statins
were started and stopped in succession with the concurrent appearance and disappearance of symptoms.4
The conclusion from these data is that the potential risk of
peripheral neuropathy with statin therapy is very small, if it
exists at all. Our neurology experts do not believe that such a
relation exists and speculate that cases of peripheral neuropathy in patients taking statins are likely to be idiopathic in
nature.9 Given this background, it is reasonable to systematically evaluate patients who develop peripheral neuropathy
symptoms while taking a statin. The first step would be to rule
out secondary causes (ie, diabetes, renal insufficiency, alcohol
abuse, vitamin B12 deficiency, cancer, hypothyroidism, acquired immunodeficiency syndrome, Lyme disease, or heavy
metal intoxication).2 A second step would be to perform a
neurologic physical examination and obtain diagnostic neurologic studies to quantify neurologic abnormalities.2 If findings
are supportive of peripheral neuropathy with no other identified cause, it would be appropriate to withdraw the statin
(dechallenge), and if symptoms resolve, with the patient’s
permission, to restart therapy with another statin (rechallenge).4 The goal would be to find a way to continue to provide
the patient with the benefits of statin therapy, but without
adverse consequences, if possible.
As for dementia and cognitive impairment, there is practically no evidence to support a link with statin therapy. In
fact, statins may actually improve cognition.9
94C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
Table 5
Recommendations for health professionals regarding neurologic disorders and statin therapy2,4,9
1. Routine neurologic monitoring of patients administering statin therapy for changes indicative of peripheral neuropathy or impaired cognition is not
recommended.
2. Patients experiencing symptoms consistent with peripheral neuropathy while receiving a statin should be evaluated to rule out secondary causes (eg,
diabetes mellitus, renal insufficiency, alcohol abuse, vitamin B12 deficiency, cancer, hypothyroidism, acquired immunodeficiency syndrome, Lyme
disease, or heavy metal intoxication).
3. If another etiology of the neurologic symptoms is not identified, it is appropriate to withdraw statin therapy for a period of 3– 6 months to establish
whether an apparent association with statin therapy exists.
4. If the patient’s neurologic symptoms improve while off statin therapy, a presumptive diagnosis of statin-induced peripheral neuropathy might be
made. However, because of the proven benefit of statin therapy, reinitiation of statin therapy should be considered with a different statin and dose.
5. If the patient’s neurologic symptoms do not improve after statin therapy has been withdrawn for the specified period, statin therapy should be
restarted based on a risk– benefit analysis.
6. If the patient experiences impaired cognition while receiving statin therapy it is appropriate to follow a similar course of evaluation as
suggested above for peripheral neuropathy, ie, first rule out other etiologies, and if none are found, then withdraw the statin for 1–3 months. If
improvement is not seen, statin therapy should be restarted based on a risk– benefit analysis.
The most noteworthy evidence addressing dementia is
the large HPS, which studied 20,536 patients over a
5-year period and found no difference in the rate of
cognitive impairment (based on a phone interview at the
conclusion of the study) in patients receiving simvastatin
versus placebo.4,9 Similarly, the PROSPER study of patients aged 70 – 82 years reported no difference between
placebo and pravastatin therapy.2,4,9 One small proof-ofconcept randomized, placebo-controlled clinical trial in
patients with mild-to-moderate Alzheimer disease found
that patients treated with atorvastatin actually showed
improvement in state-of-the-art measures of cognition
compared with those who were given placebo.2,9 Additionally, several case-control and cohort studies suggest
statin benefit in lowering the risk of Alzheimer disease
and dementia.2 Only a handful of case reports suggests
worsening of cognition with statin therapy. While these
might be idiosyncratic reactions, the existence of such
reactions is not supported by any evidence from randomized clinical trials and cohort studies.
Recommendations: Table 5 shows our consensus recommendations to health professionals based on the evidence, interpretations, and assessments presented in this
supplement regarding the neurologic system and statin
safety.2,4,9
Acknowledgments
The authors are grateful for the generous expert advice we
received from Neil J. Stone, MD, Professor of Medicine,
Northwestern University School of Medicine and Harold
Bays, MD, President, Louisville Metabolic and Atherosclerosis Research Center.
1. Cohen DE, Anania FA, Chalasani N. An assessment of statin safety by
hepatologists. Am J Cardiol 2006;97(suppl 8A):77C– 81C.
2. Bays H. Statin safety: an overview and assessment of the data—2005.
Am J Cardiol 2006;97(suppl 8A):6C–26C.
3. Jacobson TA. Statin safety: lessons from New Drug Applications for
marketed statins. Am J Cardiol 2006;97(suppl 8A):44C–51C.
4. Law M, Rudnicka AR. Statin safety: evidence from the published
literature. Am J Cardiol 2006;97(suppl 8A):52C– 60C.
5. Cziraky MJ, Willey VJ, McKenney JM, Kamat SA, Fisher MD, Guyton
JR, Jacobson TA, Davidson MH. Statin safety: An assessment using an
administrative claims database. Am J Cardiol 2006;97(suppl 8A):61C–
68C.
6. Thompson PD, Clarkson PM, Rosenson RS. An assessment of statin
safety by muscle experts. Am J Cardiol 2006;97(suppl 8A):69C–76C.
7. Kasiske BL, Wanner C, O’Neill WC. An assessment of statin safety by
nephrologists. Am J Cardiol 2006;97(suppl 8A):82C– 85C.
8. Davidson MH, Clark JA, Glass LM, Kanumalla A. Statin safety: an
appraisal from the Adverse Event Reporting System (AERS). Am J
Cardiol 2006;97(suppl 8A):32C– 43C.
9. Brass LM, Alberts MJ, Sparks L. An assessment of statin safety by
neurologists. Am J Cardiol 2006;97(suppl 8A):86C– 88C.
Benefit versus Risk in Statin Treatment
John R. Guyton, MD
The Statin Safety Assessment Conference of the National Lipid Association (NLA),
reported in this supplement to The American Journal of Cardiology, provides a
comprehensive evaluation of old and new experience on adverse events associated
with the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase inhibitors,
or statins. To place these in context, one can express both the risk of side effects and
the benefits for cardiovascular disease in terms of events per person-year of statin
treatment. The mortality risk from fatal rhabdomyolysis is approximately 0.3 per
100,000 person-years, and the risks of nonfatal rhabdomyolysis and of putative
statin-attributable peripheral neuropathy are approximately 3 and 12 events, respectively, per 100,000 person-years. Reports of acute liver failure and acute or chronic
kidney disease give lower rate estimates that, even when corrected for underreporting, are approximately equal to the background rates of these conditions in the
general population, lending scant support for statin-attributable etiology. In contrast,
the benefit of statin use is to avert several hundred deaths and several hundred cases
each of heart and brain infarction per 100,000 person-years in appropriately treated
high-risk patients. Although population estimates such as these are useful, they must
be translated repeatedly to individual patient-provider encounters, where clinical skill
and art must combine with scientific evidence. The continued publication of individual case reports and small randomized trials among groups of patients with potential
side effects should be encouraged. Statins should not be used in situations where
minimal benefit is expected, as safety data and risk– benefit analysis must be meshed
with guidelines that help the clinician decide whom to treat and how aggressively to
treat. © 2006 Elsevier Inc. All rights reserved. (Am J Cardiol 2006;97[suppl]:
95C–97C)
The 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA)
reductase inhibitors, or statins, are an essential part of clinical lipid management aimed at prevention of atherosclerotic cardiovascular events. Although their efficacy is unquestioned, the occurrence and perception of side effects
from statins may limit their usefulness, as discussed by
Gotto.1 A Statin Safety Assessment Conference, sponsored
by the National Lipid Association (NLA) with funding from
4 pharmaceutical companies, was held July 17–19, 2005 in
Washington, DC. The NLA commissioned detailed analyses
of statin safety from the viewpoints of the clinical literature,
pharmacokinetics and drug interaction, premarketing pharmaceutical data, meta-analyses of cohort data and randomized clinical trials, spontaneous adverse event reports, and a
large healthcare claims database. At the conference, 4 panels in the fields of hepatology, nephrology, muscle disorders, and neurology, each composed of 3 expert reviewers,
evaluated the extensive evidence regarding statin-associated
adverse events. The panels were assigned specific questions
to answer, prepared by an NLA task force covering clinically important issues on statin safety and side effects. It is
Duke University Medical Center, Durham, North Carolina, USA.
Address for reprints: John R. Guyton, MD, Duke University Medical
Center, Box 3510, Durham, North Carolina 27710.
E-mail address: john.guyton@duke.edu.
0002-9149/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.
doi:10.1016/j.amjcard.2005.12.016
hoped that the extensive data and the expert panel evaluations presented in this supplement to The American Journal
of Cardiology will enhance the appropriate use of statins in
treating patients at risk for atherosclerotic cardiovascular
events.
Faced with this massive presentation of statin side effects, one might lose sight of the fact that these drugs are
lifesaving medications. To see the comparison more clearly,
we can express both risk and benefit in terms of mortality
change per person-year of statin treatment. The only substantial, well-defined mortality risk with statin therapy is
that of fatal rhabdomyolysis. From clinical trial and cohort
data, Law and Rudnicka2 estimate the rate of all cases of
rhabdomyolysis at 3 per 100,000 person-years during statin
treatment. The case fatality rate is about 9%, giving a
mortality risk from rhabdomyolysis of 0.3 per 100,000 person-years. Turning to survival benefit, a meta-analysis by
Wilt and colleagues3 of 17 placebo-controlled, secondaryprevention statin trials yielded an absolute reduction of
all-cause mortality of 1.8% over an average trial duration of
approximately 5 years. This is a rate of 360 per 100,000
person-years, due entirely to reduction of cardiovascular
mortality. Results from the Heart Protection Study (HPS)
suggest that statin treatment would be effective in the vast
majority of the 13 million people with prevalent coronary
artery disease (CAD) in the United States.4,5 In this group
www.AJConline.org
96C
The American Journal of Cardiology (www.AJConline.org) Vol 97 (8A) April 17, 2006
alone, widespread statin treatment could save 40,000 lives
each year. The actual potential for mortality reduction is
considerably greater, if one adds high-risk primary prevention, statin treatment in diabetes mellitus, and other factors.6
The risk of permanent organ damage can be analyzed
similarly to mortality risk. In secondary-prevention statin
trials, the composite end point of myocardial infarction
(MI), stroke, and coronary death shows absolute reductions
with statin treatment generally 2–3 times higher than mortality reductions alone.4,7,8 In subjects without prior clinical
atherosclerotic disease, randomized trials have shown that
statins reduce the composite end point of nonfatal MI and
coronary death with an absolute decrement of 1.6% over
approximately 4.3 years.9 –12 Among subjects with diabetes
in 1 study, the absolute reduction was 1.6% over 3.9 years,
and stroke was reduced by 1.0%.12 These benefits can be
expressed as 380 cases of MI or coronary death, 410 cases
of the same in patients with diabetes, and 260 cases of
stroke in patients with diabetes averted per 100,000 personyears of statin treatment.
In comparison, the risks of permanent organ damage
resulting from statin treatment are very small. The significant risks pertain to rhabdomyolysis and myopathy (although recovery is the usual course) and perhaps to peripheral nerve damage (most patients also recover).
Rhabdomyolysis risk, already cited, is about 3 cases per
100,000 person-years. Phillips and associates13 have described a statin-associated syndrome of weakness and
pathologic changes in muscle without serum creatine kinase
(CK) elevation, which deserves further investigation. However, because only 4 cases have been defined thus far, there
is no basis for assigning an occurrence rate other than
“rare.” This syndrome has not been apparent in randomized
clinical trials. Law and Rudnicka2 give an estimate for
statin-attributable peripheral neuropathy incidence of 12 per
100,000 person-years or prevalence of 60 per 100,000 persons. Acute liver failure has occurred in statin users with a
spontaneous reporting rate of 0.1 case per 100,000 personyears of treatment. Correction for underreporting might
increase this rate to 0.5–1 case per 100,000 person-years,
but this is approximately equal to the background rate of
liver failure in the general population.14 There is no evidence that statins cause acute or chronic kidney damage.
Therefore, when the benefits of averting infarction of the
myocardium and brain are considered, the risk– benefit ratio
for permanent organ damage with appropriately administered statin treatment is very low.
The Statin Safety Assessment Task Force covered many
additional topics, exemplifying the multilayered approach
necessary for safety and tolerability assessment of any medication class. In his compendious review, Bays15 provides
clinical insight particularly with regard to combination therapy. Bottorff16 highlights various drug-removal pathways,
including organic anion transport polypeptides, the importance of which have been recently recognized. From the
thousands of pages of New Drug Applications (NDAs)
reviewed by Jacobson,17 details on proteinuria emerge, as
well as a temporal correlation of aminotransferase and CK
elevations. Law and Rudnicka2 provide a systematic review
emphasizing the quantitation of adverse events in cohort
studies and randomized trials. Davidson18 looks in detail at
adverse event reports for rosuvastatin and cerivastatin compared with other statins, finding that reports for rosuvastatin
are similar to those for other statins, whereas use of cerivastatin resulted in far greater numbers of rhabdomyolysis
reports. The healthcare database analysis by Cziraky19 represents new data, corroborating the safety data on statins
while avoiding the uncertainties of self-selection of patients
for clinical trials and the vagaries of spontaneous adverse
event reporting.
One of the key questions addressed at this conference
was whether all currently marketed statins have a similar
very low risk of serious adverse effects. Based on the data
thus far available, the answer is yes. In particular, sufficient
data are available to say that rosuvastatin gives rates of
adverse events similar to those of other statins currently on
the market, affirming the recent well-documented judgment
of the US Food and Drug Administration (FDA).20
Evidence obtained from large randomized trials and surveillance studies on statin efficacy and safety is optimal for
public policy decisions, but how do we translate it to the
individual clinician-patient encounter? Specifically, when a
patient telephones the clinician to report an adverse experience or when laboratory abnormalities appear, how does
one respond? Risk– benefit assessment in an individual demands clinical skill and art in addition to the best scientific
evidence. The practical question is one of risk or discomfort
associated with the adverse experience versus the risk of
falsely diagnosing statin causation and thereby losing the
benefit of the drug.
Recommended strategies for the individual patient with
specific adverse experiences are presented by the Expert
Panels. It is recognized that this area needs further investigation. The conference highlighted the value of individual
published case reports and regulatory adverse-event reports,
particularly for adverse experiences that may be rare or less
well defined in randomized trials. The unique myopathic
potential of cerivastatin was discovered in this way and was
later confirmed by analysis of health claims databases.21,22
Reporting must continue and should be encouraged.
Another research strategy focuses on small groups of
patients presenting with specific, rare, or infrequent adverse
experiences while taking medication, as distinguished from
large segments of the population needing treatment with the
medication. Small randomized clinical trials performed in
patients who have reported possible side effects can be
highly definitive, giving “A”-grade or top-level evidence if
multiple studies are statistically significant and in agreement.23 The drawback of such small trials is that the frequency of the side effect in the whole treatment population
generally remains unknown, but a coordinated strategy of
Guyton/Benefit versus Risk in Statin Treatment
surveillance followed by randomized, blinded evaluation of
putative cases could assess both frequency and causation.
For the vast majority of patients needing statin therapy, it
suffices to know that these drugs are both very effective and
very safe. The official prescribing information for laboratory monitoring, special precautions, and drug interactions
should be followed. Statins, like any other class of drugs,
should not be used where minimal benefit is expected,
because safety data and risk– benefit analysis go hand-inhand with National Cholesterol Education Program (NCEP)
guidelines that help the clinician decide whom to treat and
how aggressively to treat.24,25
1. Gotto AM Jr. Statins, cardiovascular disease, and drug safety. Am J
Cardiol 2006;97(suppl 8A):3C–5C.
2. Law MR, Rudnicka AR. Statin safety: a systematic review. Am J
Cardiol 2006;97(suppl 8A):52C– 60C.
3. Wilt TJ, Bloomfield HE, MacDonald R, Nelson D, Rutks I, Ho M,
Larsen G, McCall A, Pineros S, Sales A. Effectiveness of statin
therapy in adults with coronary heart disease. Arch Intern Med 2004;
164:1427–1436.
4. Heart Protection Study Collaborative Group. MRC/BHF Heart Protection Study of antioxidant vitamin supplementation in 20,536 high-risk
individuals: a randomised placebo-controlled trial. Lancet 2002;360:
23–33.
5. American Heart Association. Heart and Stroke Facts 2005 Statistical
Update. [American Heart Association Web site.] Available at: http://
heart.org/presenter.jhtml?identifier⫽3000090. Accessed August 29,
2005.
6. Law MR, Wald NJ, Rudnicka AR. Quantifying effect of statins on low
density lipoprotein cholesterol, ischaemic heart disease, and stroke:
systematic review and meta-analysis [review]. BMJ 2003;326:1423.
7. Scandinavian Simvastatin Survival Study Group. Randomised trial of
cholesterol lowering in 4444 patients with coronary heart disease: the
Scandinavian Simvastatin Survival Study (4S). Lancet 1994;344:
1383–1389.
8. The Long-Term Intervention with Pravastatin in Ischaemic Disease
(LIPID) Study Group. Prevention of cardiovascular events and death
with pravastatin in patients with coronary heart disease and a broad
range of initial cholesterol levels. N Engl J Med 1998;339:1349 –1357.
9. Shepherd J, Cobbe SM, Ford I, Isles CG, Lorimer AR, MacFarlane
PW, Packard CJ, for the West of Scotland Coronary Prevention Study
Group. Prevention of coronary heart disease with pravastatin in men
with hypercholesterolemia. N Engl J Med 1995;333:1301–1307.
10. Downs JR, Clearfield M, Weis S, Whitney E, Shapiro DR, Beere PA,
Langendorfer A, Stein EA, Kruyer W, Gotto AM Jr. Primary prevention of acute coronary events with lovastatin in men and women with
average cholesterol levels: results of AFCAPS/TexCAPS. JAMA 1998;
279:1615–1622.
97C
11. Sever PS, Dahlof B, Poulter NR, Wedel H, Beevers G, Caulfield M,
Collins R, Kjeldsen SE, Kristinsson A, McInnes GT, et al, for the
ASCOT investigators. Prevention of coronary and stroke events with
atorvastatin in hypertensive patients who have average or lower-thanaverage cholesterol concentrations, in the Anglo-Scandinavian Cardiac
Outcomes Trial-Lipid Lowering Arm (ASCOT-LLA): a multicentre
randomised controlled trial. Lancet 2003;361:1149 –1158.
12. Colhoun HM, Betteridge DJ, Durrington PN, Hitman GA, Neil HA,
Livingstone SJ, Thomason MJ, Mackness MI, Charlton-Menys V,
Fuller JH. Primary prevention of cardiovascular disease with atorvastatin in type 2 diabetes in the Collaborative Atorvastatin Diabetes
Study (CARDS): multicentre randomised placebo-controlled trial.
Lancet 2004;364:685– 696.
13. Phillips PS, Haas RH, Bannykh S, Hathaway S, Gray NL, Kimura BJ,
Vladutiu GD, England JD. Statin-associated myopathy with normal
creatine kinase levels. Ann Intern Med 2002;137:581–585.
14. Tolman KG. The liver and lovastatin. Am J Cardiol 2002;89:1374 –
1380.
15. Bays H. Statin safety: An overview and assessment of the data—2005.
Am J Cardiol 2006;97(suppl 8A):6C–26C.
16. Bottorff MB. Statin safety and drug interactions: clinical implications.
Am J Cardiol 2006;97(suppl 8A):27C–31C.
17. Jacobson TA. Statin safety: lessons from New Drug Application submissions for marketed statins. Am J Cardiol 2006;97(suppl 8A):44C–
51C.
18. Davidson MH, Clark JA, Glass LM, Kanumalla A. Statin safety: an
appraisal from the Adverse Event Reporting System. Am J Cardiol
2006;97(suppl 8A):32C– 43C.
19. Cziraky M. An assessment of the safety of lipid-altering drugs using an
administrative claims database. Am J Cardiol 2006;97(suppl 8A):61C–
68C.
20. US Food and Drug Administration. FDA response to a Public Citizen
petition regarding Crestor. March 11, 2005. Available at: http://
www.fda.gov/cder/drug/infopage/rosuvastatin/crestor_CP.pdf. Accessed
August 29, 2005.
21. Staffa JA, Chang J, Green L. Cerivastatin and reports of fatal rhabdomyolysis. N Engl J Med 2002;346:539 –540.
22. Graham DJ, Staffa JA, Shatin D, Andrade SE, Schech SD, La GL,
Gurwitz JH, Chan KA, Goodman MJ, Platt R. Incidence of hospitalized rhabdomyolysis in patients treated with lipid-lowering drugs.
JAMA 2004;292:2585–2590.
23. Montori VM. Evidence-based endocrinology: how far have we come?
Treat Endocrinol 2004;3:1–10.
24. Expert Panel on Detection, Evaluation, and Treatment of High Blood
Cholesterol in Adults. Executive Summary of the Third Report of the
National Cholesterol Education Program (NCEP) Expert Panel on
Detection, Evaluation, and Treatment of High Blood Cholesterol in
Adults (Adult Treatment Panel III). JAMA 2001;285:2486 –2496.
25. Grundy SM, Cleeman JI, Merz CN, Brewer HB Jr, Clark LT, Hunninghake DB, Pasternak RC, Smith SC Jr, Stone NJ. Implications of
recent clinical trials for the National Cholesterol Education Program
Adult Treatment Panel III guidelines. Circulation 2004;110:227–239.
NOTES