RE-LY: Post Hoc Analysis Confirms Benefit of Dabigatran Relative to Warfarin at All INR Levels

Susan Jeffrey

November 17, 2009

November 17, 2009 — A new post hoc analysis of the RE-LY trial shows that dabigatran etexilate (Boehringer Ingelheim) is noninferior in the lower dose and superior in the higher dose to warfarin for stroke prevention in patients with atrial fibrillation (AF), regardless of the average international normalized ratio (INR) control with warfarin achieved across participating centers.

Even in centers where INR control was excellent, dabigatran was noninferior at the 110-mg dose and superior at the 150-mg dose to warfarin for prevention of stroke or systemic embolism, the primary outcome of the RE-LY, or Randomized Evaluation of Long-Term Anticoagulant Therapy, Warfarin, Compared With Dabigatran, trial. Rates of intracranial hemorrhage were superior at both doses of dabigatran, and major bleeding was at least noninferior to warfarin, regardless of INR control across centers.

Dr. Lars Wallentin

"Concerning the main outcome events that we aim to reduce, that is, stroke, intracranial hemorrhage, and major bleeding, the results were consistent, and there are advantages of dabigatran regardless of INR control," Lars Wallentin, MD, from Karolinska Institute in Stockholm, Sweden, said in an interview. "However, looking at secondary events such as overall mortality, if you count all events, then you have larger benefits at poorer INR control, which I think is expected."

The new analysis was presented here at the American Heart Association (AHA) 2009 Scientific Sessions. Main results of the RE-LY study were presented August 30 at the European Society of Cardiology Congress 2009 and published simultaneously in the New England Journal of Medicine (Connolly SJ, et al. N Engl J Med. 2009;361:1139–1151). The trial was funded by Boehringer Ingelheim.

Time in Treatment Range

The benefit of warfarin relates directly to the time in treatment range (TTR) or the time spent at the optimal INR between 2.0 and 3.0, Dr. Wallentin said. In this post hoc analysis, the researchers looked to see whether the quality of INR control at the centers involved in the RE-LY trial would affect the relative benefit seen in the overall trial with dabigatran.

Dabigatran is one of a number of oral anticoagulants being developed for the prevention of thromboembolism in the setting of AF and other conditions, such as venous thromboembolism (VTE), for which warfarin has long been the only choice. The last great hope was ximelagatran, which made it all the way to a Food and Drug Administration (FDA) review panel before being sidelined by liver toxic effects. Dabigatran is already available for VTE prevention during hip and knee replacement surgery in the European Union as Pradaxa and in Canada as Pradax.

The RE-LY trial was a direct comparison of dabigatran and warfarin conducted at 951 centers in 44 countries. A total of 18,113 patients with AF and at least 1 other risk factor for stroke were randomized to receive 110 or 150 mg of dabigatran twice daily or open-label warfarin therapy with a target INR of 2.0 to 3.0.

For this new analysis, investigators used the average TTR in the warfarin arm at each center as a proxy for INR control in all patients from that center and compared outcomes across quartiles of TTR vs dabigatran. No INR was available for dabigatran because one of its potential benefit is the lack of need for such monitoring. "We took the centers' INR control as an approximation of the INR control the dabigatran patients in that center would have had if they had been treated with warfarin," Dr. Wallentin noted.

Despite efforts made in the RE-LY study to improve the generally poor quality of INR control seen in many trials, there was still a wide variation in TTR across participating countries, from a high of 77% in Sweden to 41% in other countries, mostly in Eastern Europe, South America, and Asia.

To support their use of center-based TTRs as a proxy for INR control, the investigators first compared quartiles of TTR for individual patients versus center-based TTR. They found a similar gradient in outcomes of death, major bleeding, and stroke from poor to good INR control, although the gradient was somewhat blunted compared with results based on individual TTRs.

Results for the primary endpoint of prevention of stroke and systemic embolism mirrored those of the overall trial, showing noninferiority for the 110-mg dose and a significant 34% reduction with the 150-mg dose.

"When looking into the quartiles of center-based INR control, there was no indication of an interaction, no significant interaction, and you can see the same consistent results regardless of INR control concerning the primary endpoint," Dr. Wallentin said.

Risk (95% Confidence Interval) of Stroke or Systemic Embolism With Dabigatran in 2 Doses Versus Warfarin by Center-Based TTR

TTR 110 mg of Dabigatran P Value 150 mg of Dabigatran P Value
All patients 0.91 (0.74 – 1.11) .34 0.66 (0.53 – 0.82) < .001
<56.9% 1.1 (0.73 – 1.6)   0.61 (0.39 – 0.96)  
56.9% – 65.4% 0.74 (0.51 – 1.1)   0.48 (0.32 – 0.74)  
65.4% –72.4% 1.0 (0.65 – 1.5)   0.76 (0.48 – 1.21)  
>72% 0.88 (0.57 – 1.4)   0.88 (0.57 – 1.37)  
Interquartile P value   .27   .41

TTR = time in treatment range.

Overall results showed significant 70% and 60% reductions in intracranial hemorrhage with the 110-mg and 150-mg doses of dabigatran, respectively, vs warfarin, and similar results were seen across quartiles of center-based INR control with no sign of interaction, he noted.

Major bleeding was reduced in the overall trial by 20% with the 110-mg dose, and there was no difference with the 150-mg dose vs warfarin, and again, no statistically significant interaction with relation to center-based INR control was seen, although there was a trend to reduced major bleeding at poor INR control, he noted.

There was, however, a significant interaction for mortality with INR control, with a mortality reduction seen for those taking dabigatran in centers with poor INR control.

Despite the limitations of a post hoc analysis and the use of center-based INR control as a proxy for individual patients, Dr. Wallentin concluded that "for the primary efficacy and safety results, the main RE-LY study results are consistent, showing reductions in stroke and major bleeding with dabigatran compared to warfarin irrespective of center-based INR control.

"For secondary outcomes, such as all vascular events and mortality, the advantages of dabigatran may be greater at centers with poorer INR control," he added.

Center-Based INRs but Patient-Level Question

The invited discussant for this presentation was Daniel Singer, MD, from Massachusetts General Hospital, Boston. Dr. Singer said he was most impressed by the lower rate of intracranial hemorrhage in the dabigatran arms, "despite equal or better protection against ischemic stroke."

However, he said, the relevant clinical question for him is whether patients with very good INR control would benefit from switching to dabigatran, "absent issues of convenience and so on. This is an individual patient-level question."

Although the center-based approach preserves randomization and reduces confounding, he pointed out, it does not directly address the clinically relevant question. The analysis showed that relationships among TTR, stroke, and intracranial hemorrhage were somewhat blunted compared with the overall findings, whereas all-cause death and composite cardiovascular outcomes were stronger, "puzzling" findings that suggest death rates must include causes of death not sensitive to anticoagulants, he noted.

"Overall, RE-LY was a triumph for dabigatran," Dr. Singer concluded. "At the centers with the highest center-based TTR, dabigatran still appears more attractive than warfarin because of the reduced rate of intracranial hemorrhage, but the superiority in stroke prevention of the higher dose is no longer demonstrated.

I would like to see more individual-level comparisons with patients, not centers, grouped by TTR.

"I would like to see more individual-level comparisons with patients, not centers, grouped by TTR," he added. "Such comparisons will need sophisticated analyses to reduce confounding, but while such individual level analyses increase the risk of confounding, they also increase the relevance of the results to decision-making for individual patients."

Dr. Wallentin replied, "From our perspective, it's almost impossible to do the individual-level analysis because, who will the comparator be? Because you have no INRs in the dabigatran arm, and therefore we cannot identify the TTR and cannot expect a time in treatment range for the control patients. It might be possible if you go to patients who were on warfarin before randomization, but really, we don't have those data, and therefore, we thought this might be the most appropriate analysis that can be done."

From our perspective, it's almost impossible to do the individual-level analysis because, who will the comparator be?

Any comparison that was limited to the best-controlled patients — found generally in the same centers worldwide — would be "unfair to dabigatran," he said.

Dr. Wallentin told Medscape Neurology that he expects Boehringer Ingelheim will soon move for approval in the United States for both the prevention of VTE indication and stroke prevention in AF.

"We now have a new oral anticoagulant that does not need monitoring, that can be used more broadly in many patients, has few side effects, and provides both more effective and safer treatment," he said. Although those already well-controlled with warfarin might be hesitant to switch, Dr. Wallentin added, new patients, given the choice, would very likely choose dabigatran.

Changing Treatment

Asked for comment on the RE-LY findings, Ralph Sacco, MD, from the Miller School of Medicine at the University of Miami in Florida and president-elect of the AHA, said the results of the substudy presented here help substantiate the overall study results that show that this new agent is better than warfarin.

"Although warfarin has class I recommendations with grade A evidence, it has been underutilized, largely because of the safety issue and the narrow therapeutic index that this study continues to show," he told Medscape Neurology.

The RE-LY trial shows that dabigatran appears to be superior to warfarin and is reasonably safe, he noted. "It does not have FDA approval yet, and none of the AHA recommendations have yet reviewed [these] data, but I think it's a step in the right direction towards looking for new drugs that will be safer and perhaps more effective than warfarin."

"Helpful to hear," he added, is that there appears to be no sign of the liver toxic effects that derailed ximelagatran.

Philip B. Gorelick, MD, MPH, from the University of Illinois College of Medicine at Chicago, said the overall results of the RE-LY trial "are likely to substantially change the treatment paradigm for stroke prevention in atrial fibrillation."

Results showed that the 110-mg dose was associated with less bleeding than warfarin with similar efficacy and the 150-mg dose with significantly fewer strokes and no increase in major bleeding, he noted. "If approved for use by major regulatory agencies, higher-dose dabigatran may be used for AF patients at high risk of stroke who are in need of oral anticoagulation, and lower-dose dabigatran may be used in those AF patients not as high risk for stroke or those [whom] physicians believe may have an increased risk of major bleeding but who qualify for oral anticoagulation therapy."

However, gastrointestinal bleeding was significantly higher and myocardial infarction was somewhat increased with dabigatran, Dr. Gorelick pointed out. The study itself was unblinded and excluded patients with creatinine clearance of less than 30 mL/minute.

Despite these potential study limitations, many patients with AF and those who are at risk for stroke do not receive oral anticoagulation, he concluded. "Dabigatran therapy is not encumbered by limitations related to food-drug interaction, frequent blood test monitoring, and shows very little in the way of drug-drug interactions and will likely turn out to be a welcome addition to our treatment strategies for AF patients in need of oral anticoagulation."

The study was funded by Boehringer-Ingelheim. Dr. Wallentin has received institutional grants from Boehringer Ingelheim, Bristol-Myers Squibb, AstraZeneca, Schering-Plough, GlaxoSmithKline, and Eli Lilly. Dr. Singer has received significant research support from Daiichi Sankyo, is a consultant/advisory board member for Boehringer Ingelheim, Daiichi Sankyo, Merck and Co, and Sanofi Aventis; has received honoraria from Bristol-Myers Squibb and Pfizer; and is a member of the executive committee for the ROCKET-AF trial of rivaroxaban in AF. Dr. Sacco has consulted previously for Boehringer Ingelheim, but he relinquished the relationship in April 2009. Dr. Gorelick has served on advisory boards for Boehringer Ingelheim.

American Heart Association (AHA) 2009 Scientific Sessions: Abstract 101. Presented November 15, 2009.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.