Meta-analysis Takes an Indirect Look at Differences Among Novel Anticoagulants

Download this article's Factoid in PDF (& PPT for Gold Subscribers)

With little prospect of head-to-head comparisons between the newer oral anticoagulants in clinical trials, a new meta-analysis based on indirect comparisons among the various agents suggests that not all of them provide the same safety and efficacy for preventing stroke in patients with atrial fibrillation (A-fib). The findings, published online August 21, 2012, in Circulation: Cardiovascular Quality and Outcomes, suggest more similarities than differences, but those differences may help clinicians individualize treatment.

For the study, William L. Baker, PharmD, of the University of Connecticut (Storrs, CT), and Olivia J. Phung, PharmD, of Western University of Health Sciences (Pomona, CA), identified 4 randomized clinical trials including 44,733 patients that compared the novel direct thrombin inhibitor dabigatran or the factor Xa antagonists apixaban or rivaroxaban with warfarin for stroke prophylaxis in patients with nonvalvular A-fib:

  • RE-LY: dabigatran (only data from the 150-mg dose arm included in the analysis, twice daily; n = 6,076)
  • PETRO: dabigatran (150 mg twice daily; n = 166)
  • ROCKET-AF: rivaroxaban (20 mg daily; n = 7,131)
  • ARISTOTLE: apixaban (5 mg twice daily; n = 9,120)

No Contest for New Anticoagulants vs. Warfarin

In a pooled analysis that compared the oral anticoagulants in general with warfarin, the newer drugs reduced the composite of stroke or systemic emboli as well as any stroke, hemorrhagic stroke, and all-cause mortality, with no differences for other outcomes (table 1).

Table 1. Efficacy of New Oral Anticoagulants vs. Warfarin

 

 

RR

95% CI

P Value

Stroke or Systemic Emboli

0.80

0.70-0.91

< 0.001

Any Stroke

0.77

0.64-0.92

0.0001

Hemorrhagic Stroke

0.46

0.27-0.77

0.0001

All-Cause Mortality

0.87

0.80-0.97

0.001

 However, significant heterogeneity among the studies was seen for the endpoints of any stroke, hemorrhagic stroke, and major and gastrointestinal bleeding.

Paired Comparisons Reveal Differences

Because warfarin was a common comparator in the randomized trials analyzed, the investigators were able to make adjusted, indirect comparisons among the 3 newer anticoagulants.

Compared with dabigatran, apixaban was associated with a lower risk of major bleeding (RR 0.75; 95% CI 0.62-0.92) and gastrointestinal bleeding (RR 0.60; 95% CI 0.43-0.84) and a trend toward increased hemorrhagic stroke (RR 1.93; 95% CI 0.93-4.02).

A comparison between rivaroxaban and dabigatran, meanwhile, found that dabigatran reduced the risk of the composite of stroke or systemic emboli (RR 0.76; 95% CI 0.58-0.99), ischemic stroke (RR 0.68; 95% CI 0.49-0.93), and hemorrhagic stroke (RR 0.45; 95% CI 0.21-0.98), while there was no difference between the anticoagulants in rates of any stroke, mortality, or major or gastrointestinal bleeding.

Finally, compared with rivaroxaban, apixaban increased the risk of systemic emboli (RR 3.85; 95% CI 1.20-12.36) but decreased the risk of major bleeding (RR 0.69; 95% CI 0.49-1.58) and gastrointestinal bleeding (RR 0.58; 95% CI 0.41-0.82), while other efficacy and safety endpoints were similar between the agents.

To put the results in context, the investigators calculated the clinical impact of using 1 new agent vs. another (table 2).

Table 2. Differences in Events per 1,000 Patients Treated

 

 

Apixaban vs. Dabigatran

Dabigatran vs. Rivaroxaban

Apixaban vs. Rivaroxaban

Stroke or Systemic Emboli

-5

-6

-1

Ischemic Stroke

4

-9

-5

Any Stroke

5

-5

-1

Mortality

1

-3

-2

Major Bleed

-11

-6

-16

Gastrointestinal Bleed

-12

0

-11

 

Though acknowledging that differences in stroke risk and warfarin management in the underlying trials could have confounded results, the investigators point out that meta-regression analysis showed no association between CHADS2 scores or warfarin time in therapeutic range and any of the efficacy or safety outcomes.

According to Drs. Baker and Phung, “[a]lthough direct head-to-head clinical trials are required to confirm the findings of this adjusted indirect comparison analysis, they are unlikely to be conducted. Data from real-world patient registries may shed light on differences between these agents.”

In the meantime, they advise that several other considerations should be factored into clinical decisions about which oral anticoagulant to select for a given patient, including convenience of administration, cost, ease of reversal, and possible adverse events.

Do Indirect Comparisons Supplement Trial Data?

In an e-mail correspondence with TCTMD, Gregory Y. H. Lip, MD, of the University of Birmingham (Birmingham, United Kingdom), confirmed that the current findings basically mirror those of a recent JACC study, for which he was the principal investigator (Lip GYH. J Am Coll Cardiol. 2012;60;738-746)

Such data may be useful because “one size does not fit all” when it comes to prescription of anticoagulants, he noted, adding that “with the availability of new agents, we can tailor the drug to the patient profile.”

However, Michael D. Ezekowitz, MD, PhD, of Thomas Jefferson Medical College (Philadelphia, PA), was more reserved regarding the value of such studies. “Indirect comparisons are a thought-provoking academic exercise,” he wrote in an e-mail correspondence with TCTMD. “[But] from a clinical standpoint, there are more similarities between the novel anticoagulants than differences.

“We are only beginning to understand these drugs,” he continued. “Of importance, none of the trials were stopped early because of more serious adverse events compared to warfarin. Thus, the objective of the immediate post-clinical-trial period is to stick to the inclusion and exclusion [criteria] of the respective trials and reduce the use of warfarin.”

While agreeing that indirect comparisons fall short, Dr. Lip suggested they may yet be clinically useful. “The trial data are the best, but in the absence of head-to-head trials [of the new anticoagulants], these indirect comparisons give some insight into the comparability of one drug against the other,” he concluded.

Source:

Baker WL, Phung OJ. Systematic review and adjusted indirect comparison meta-analysis of oral anticoagulants in atrial fibrillation. Circ Cardiovasc Qual Outcomes. 2012;Epub ahead of print.

 Related Stories:

Meta-analysis Takes an Indirect Look at Differences Among Novel Anticoagulants

With little prospect of head to head comparisons between the newer oral anticoagulants in clinical trials, a new meta analysis based on indirect comparisons among the various agents suggests that not all of them provide the same safety and efficacy
Disclosures
  • Drs. Baker and Phung report no relevant conflicts of interest.
  • Dr. Lip reports serving as a consultant or on the speaker’s bureau for multiple pharmaceutical companies.
  • Dr. Ezekowitz reports serving as co-principal investigator for the RE-LY trial, on the executive committee for the ENGAGE-AF TIMI 48 trial, and as a consultant for Bayer, Boehringer-Ingelheim, Bristol-Myers Squibb, and Pfizer.

Comments