0b0owyo

hreccermwweris

t

Journal of the American College of Cardiology Vol. 60, No. 8, 2012© 2012 by the American College of Cardiology Foundation ISSN 0735-1097/$36.00Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jacc.2012.05.012

EDITORIAL COMMENT

Danger Ahead: Watch Outfor Indirect Comparisons!*

Christopher P. Cannon, MD,† Payal Kohli, MD‡

Boston, Massachusetts

We have entered an exciting new era of anticoagulation—with multiple new oral anticoagulants becoming availablefor use as potential replacements for warfarin, the vitamin Kantagonist that has been the only oral agent available for 50years! The most prominent of the indications for anticoag-ulation is for stroke prevention in patients with atrialfibrillation, where long-term, often life-long therapy isneeded in millions of patients. Two agents have alreadybeen approved by the Food and Drug Administration(FDA), and a third is anticipated later this year. There are2 main classes of drugs that are undergoing testing or havebeen tested in large randomized trials: direct thrombininhibitors (dabigatran [1]) and factor Xa inhibitors (rivar-oxaban [2], apixaban [3], edoxaban [4]); in these trials, thesedrugs are being compared with warfarin.

See page 738

Three large randomized trials (totaling over 50,000 pa-tients with atrial fibrillation) comparing one of the neweroral anticoagulants to warfarin have already been published.Each new drug has shown equal or superior efficacy towarfarin, suggesting that these agents offer excellent alter-natives to warfarin for stroke prevention. But faced withthese growing new options, how do physicians choosebetween them? Are there differences between the drugs intheir efficacy and/or safety that would help inform ourclinical decision to prescribe one over the other? Unfortu-nately, no trials have been conducted (yet) that directlycompare one new agent versus another.

In this issue of the Journal, Lip et al. (5) have tried toanswer these questions by applying statistical methods to

*Editorials published in the Journal of the American College of Cardiology reflect theviews of the authors and do not necessarily represent the views of JACC or theAmerican College of Cardiology.

From the †Cardiovascular Division, Brigham and Women’s Hospital, Boston,Massachusetts; and the ‡Division of Cardiology, University of California SanFrancisco, San Francisco, California. Dr. Cannon is a member of the advisory boardsfor Alnylam, Bristol-Myers Squibb, and Pfizer; has received financial support fromthem, which has been donated to charity; and he has also received grant support fromAccumetrics, AstraZeneca, Essentialis, GlaxoSmithKline, Merck, Regeneron, Sanofi,

cTakeda; and has equity in Automedics Medical systems. Dr. Kohli is a member of theAdvisory Board to Daiichi Sankyo.

make indirect comparisons between the new agents. Be-cause each trial had a common comparator arm, one couldassume similarity in outcomes of that arm and then look atthe relative benefit (or harm) that each drug displays versusthe standard arm. They have used the Bucher method tocompare hazard ratios of the new agent to the comparator totry to determine the relative efficacy between the 2 neweragents.

For such comparisons, however, we need to proceed withextreme caution: attempts at indirect comparisons weremade among fibrinolytic agents—tissue plasminogen acti-vator, streptokinase, and anistreplase—where relative reduc-tions in mortality in placebo-controlled trials were 25%,25%, and 50%, respectively (6–8). On the basis of theseresults, one might be tempted to conclude that anistreplaseis superior, but in a direct comparison of these drugs inISIS-3 (Third International Study of Infarct Survival), nodifference in endpoints was seen (9). As such, indirectcomparisons can be misleading, and extreme caution shouldbe exercised when using such methods to draw definitiveconclusions.

After evaluating the efficacy of each new drug individu-ally, Lip et al. (5) then went on to calculate a weightedaverage of all the new agents combined to estimate therelative benefit of “any new oral anticoagulant” versuswarfarin. This latter approach is akin to pooling or meta-analysis of all the studies, and provides a reasonable over-view of the “class effect.” For all the new oral anticoagulants,they found that this class as a whole significantly reducedthe risk of: 1) stroke or systemic embolism by 21% (p �.001); 2) stroke by 23% (p � 0.001); 3) hemorrhagic strokey 53% (p � 0.001); and 4) all-cause mortality by 12% (p �.001). In addition, major bleeding was lower for any newral anticoagulant by 13% (p � 0.001) compared witharfarin. This straightforward and statistically sound anal-sis of the data strongly supports the use of the newer agentsver warfarin, as a class.For the indirect comparisons between individual agents,

owever, the picture remains a bit confusing. Lip et al.eport a “significantly lower risk of stroke or systemicmbolism (by 26%) for dabigatran 150 mg twice dailyompared with rivaroxaban.” On the other hand, theyonclude “we found no profound significant differences infficacy between apixaban, and dabigatran (both doses) orivaroxaban.” They also report no significant difference inyocardial infarction event rates among the 3 agents. Yet,ithin the individual trials, the rate of myocardial infarctionas definitely higher with dabigatran (1), but not with

ither of the factor Xa inhibitors (2,3). These conflictingesults lend support to the conclusion that such methods forndirect comparisons may not be the most accurate due toeveral sources of confounding.

The authors acknowledge that there are multiple limita-ions in comparing different trials, despite the common

omparator arm—notably the patient population differs.

DemTstweotiit

riFtssws(emsbsbdreWbr

ct

atnlwnptap

748 Cannon and Kohli JACC Vol. 60, No. 8, 2012Danger of Indirect Comparisons August 21, 2012:747–8

Variations in the baseline risk of patients in 1 study couldinfluence the absolute rates of the endpoints in that study,confounding the drug–drug comparison being studied. Forexample, the ROCKET trial (Rivaroxaban-once daily, oral,direct factor Xa inhibition compared with vitamin K antag-onism for prevention of stroke and Embolism Trial in AtrialFibrillation) trial enrolled older patients with higherCHADS2 (Congestive heart failure, Hypertension, Age,

iabetes, Prior Stroke) risk scores, and as such, would bexpected to have higher rates of bleeding and stroke—whichight influence the relative drug effect seen in this trial.herefore, 2 variables would differ when comparing a

imilar endpoint between these 2 trials—the drug beingested and the patient population—and it would be unclearhether any apparent difference observed in the relative

fficacy could be attributed to the drug alone. Similarly, theptimal use of warfarin (as measured by the “time inherapeutic range”) also differed between these trials, result-ng in another possible source of confounding that couldnfluence the relative efficacy of the drugs when comparinghe trials.

Traditional metrics for determining efficacy derive di-ectly from the results of the randomized trial. Indeed, thiss exactly what the FDA does in the prescribing insert: theDA approves and lists only those findings and indications

hat were directly reported in the randomized trials. Asuch, we would find that dabigatran 150 mg twice dailyignificantly reduced stroke or systemic embolism (andithin that endpoint, ischemic stroke alone), but with a

imilar rate of major bleeding to warfarin. The 110-mg doseunfortunately not approved by the FDA) had similarfficacy for stroke or systemic embolism but a lower risk ofajor bleeding compared with warfarin (1). For apixaban,

ignificant reductions in stroke or systemic embolism, majorleeding, and mortality were seen (3). Rivaroxaban showedimilar rates of stroke or systemic embolism and majorleeding versus warfarin (2). But whether apixaban andabigatran 150 mg are superior to rivaroxaban cannoteadily be concluded by comparing the relative efficacy ofach agent versus comparator within its own respective trial.

e can say though that these 2 agents have been shown toe superior for stroke prevention, and apixaban for mortalityeduction, whereas neither can be said for rivaroxaban.

So what are we to do? Should we use the indirectomparisons put forth by Lip et al. (5) since that provides

he only comparative data we have? In general, the authors

ppear to be saying that there are more similarities betweenhese agents than differences, as has also been previouslyoted by Mega (10). However, because of the statistical

imitations of such comparisons, although of some interest,e feel the differences they report on some endpoints areot robust enough to be relied upon for the clinical care ofatients. Instead, we would turn to direct evidence fromrials and the indications put forth by the FDA to select theppropriate agent, at the dose tested, for use in the patientopulation studied within the trial.

Reprint requests and correspondence: Dr. Christopher P. Can-non, TIMI Study Group, Brigham and Women’s Hospital, 350Longwood Avenue, 1st Floor, Boston, Massachusetts 02115.E-mail: cpcannon@partners.org.

REFERENCES

1. Connolly SJ, Ezekowitz MD, Yusuf S, et al. Dabigatran versuswarfarin in patients with atrial fibrillation. N Engl J Med 2009;361:1139–51.

2. Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin innonvalvular atrial fibrillation. N Engl J Med 2011;365:883–91.

3. Granger CB, Alexander JH, McMurray JJ, et al. Apixaban versuswarfarin in patients with atrial fibrillation. N Engl J Med 2011;365:981–92.

4. Ruff CT, Giugliano RP, Antman EM, et al. Evaluation of the novelfactor Xa inhibitor edoxaban compared with warfarin in patients withatrial fibrillation: design and rationale for the Effective aNticoaGula-tion with factor xA next GEneration in Atrial Fibrillation-Thrombolysis In Myocardial Infarction study 48 (ENGAGE AF-TIMI 48). Am Heart J 2010;160:635–41.

5. Lip GYH, Larsen TB, Skjøth F, Rasmussen KH. Indirect compari-sons of new oral anticoagulant drugs for efficacy and safety when usedfor stroke prevention in atrial fibrillation. J Am Coll Cardiol 2012;60:738–46.

6. Wilcox RG, von der Lippe G, Olsson CG, Jensen G, Skene AM,Hampton JR. Trial of tissue plasminogen activator for mortalityreduction in acute myocardial infarction. Anglo-Scandinavian Study ofEarly Thrombolysis (ASSET). Lancet 1988;2:525–30.

7. ISIS-2 (Second International Study of Infarct Survival) CollaborativeGroup. Randomised trial of intravenous streptokinase, oral aspirin,both, or neither among 17,187 cases of suspected acute myocardialinfarction: ISIS-2. Lancet 1988;2:349–60.

8. AIMS Trial Study Group. Long-term effects of intravenous anis-treplase in acute myocardial infarction: final report of the AIMS study.Lancet 1990;335:427–31.

9. ISIS-3 (Third International Study of Infarct Survival) CollaborativeGroup. ISIS-3: a randomised comparison of streptokinase vs tissueplasminogen activator vs anistreplase and of aspirin plus heparin vsaspirin alone among 41,299 cases of suspected acute myocardialinfarction. Lancet 1992;339:753–70.

10. Mega JL. A new era for anticoagulation in atrial fibrillation. N EnglJ Med 2011;365:1052–4.

Key Words: anticoagulation y atrial fibrillation y bleeding y stroke.