AHA 2010 – Rocket data impresses but jury out on relative benefits


As details of the Rocket-AF trial, assessing Xarelto’s ability to prevent stroke in atrial fibrillation (AF) patients are digested by clinicians, analysts and investors alike, EP Vantagegarnered some initial reaction from the sidelines of the American Heart Association (AHA) meeting.

Consensus is emerging that Xarelto will become a much-needed alternative to warfarin. The jury is still out, and likely to be for some time to come, on the all-important question of how Xarelto compares to Boehringer Ingelheim’s Pradaxa. Dr Ralph Sacco, president of AHA, warns against cross-trial comparisons and suggests the focus instead should be on the pros and cons of the Rocket trial; the most contentious issue being the statistical analysis used to assess Xarelto’s performance.

Pros and cons

The key positives for Dr Sacco include the fact the Rocket trial was a double-blinded study and that Xarelto is a once-daily drug. In comparison, the Re-ly trial supporting Pradaxa’s approval was open-label, which can be subject to bias, and the Boehringer drug is dosed twice a day.

Dr Sacco’s concerns over the Rocket trial range from the relatively weak performance of warfarin in the study, the high-risk patient population and the lack of superiority for Xarelto on the intent-to-treat (ITT) analysis.

“It looks like warfarin was not used as effectively as we would like”, says Dr Sacco, pointing to the fact the median time spent within therapeutic range (TTR) for warfarin patients was low at just 57.8%. This was also a key concern raised by the trial’s independent reviewer, Dr Elaine Hylek of Boston University Medical School.

For reference the TTR in the Re-ly trial was 64%, more in line with clinical experience with warfarin.

High risk confusion

Meanwhile the fact the trial recruited such high-risk patients – average age of 73 years, 55% had prior stroke, 91% suffered hypertension, 62% had congestive heart failure and 39% were diabetic – appears to have led to different conclusions and some confusion.

On the one hand is Dr Robert Califf, co-principal investigator and vice chancellor for clinical research at Duke University School of Medicine. He believes tackling such a high risk population is to be commended, particularly as Xarelto was shown to be a lot safer on intracranial bleeding, one of the dreaded side effects of warfarin. Indeed Bernstein analyst Jack Scannell notes that some investors appear to have taken comfort from the fact Xarelto was tested in such a high-risk population, partly explaining Bayer’s 4% share price gain on Monday.

However, as Dr Sacco points out, using such a high risk population where event rates are higher, should actually have favoured Xarelto over warfarin. Further, Mr Scannell believes “there was no obvious relationship between stroke risk and efficacy in the Re-Ly trial (of Pradaxa)”, suggesting the risk profile of the patient population should not have an impact on the efficacy of a test drug. Yet Xarelto failed to show superiority on the so-called ‘gold standard’ ITT analysis.


The most contentious issue is the relative merits and validity of the patient population Bayer chose to highlight. The ITT population, which tracks all patients who enter the study for the entire duration, is considered a rigorous and conservative measure, but the company focused on a per protocol population it called "on treatment", which excluded certain patients from the analysis.

On both of these measures Xarelto was non-inferior to warfarin in preventing stroke and and most agree that the "on treatment" analysis is suitable for determining non-inferiority.

However, where opinions divide is whether this analysis can be extended to determine superiority. Under the "on treatment" analysis, Xarelto was shown to be superior to warfarin, resulting in a statistically significant 21% relative risk reduction, which became the headline act in bullish press releases issued by Xarelto’s developers, Bayer and Johnson & Johnson.

Yet on the ITT analysis, no superiority for Xarelto was established, with a non-statistically significant 12% relative risk reduction. In contrast the highest 150mg dose of Pradaxa demonstrated superiority with a 35% relative risk reduction using the ITT analysis.

Both Dr Sacco and Dr Hylek maintain that ITT is the gold standard in evaluating superiority claims, a view supported by Dr Harlan Krumholz, professor of medicine at Yale University School of Medicine: “I think you have to go with the intention to treat analysis. I think there was a little too much emphasis, at least in the press release (by Bayer and J&J), on the superiority claim”.

Of Xarelto’s safety, the drug reported a largely clean bill of health. Aside from the major boost from reduced intracranial bleeding, critical organ bleeds and bleeding-related deaths were also significantly lower with the Bayer drug. Overall, major and non-major bleeding events were comparable to warfarin. The only worrying signal was an increase in the number of haemorrhages requiring transfusion and a drop in haemoglobin levels associated with Xarelto.

Concerns about liver toxicity proved unfounded, with no increase in ALT levels, while the gastrointestinal profile of Xarelto was clean; in contrast Pradaxa was associated with higher levels of dyspepsia, or indigestion.

More data required

When presenting the Rocket data, Dr Kenneth Mahaffey of Duke Clinical Research Institute, stressed that the data was only unlocked on October 24 and that further analysis of the thousands of data points is required before the full extent of Xarelto’s efficacy and safety can be properly assessed.

For example, the nuances of how certain sub-groups, such as warfarin-naïve patients, responded to Xarelto still needs to be scrutinised.

Indeed Dr Krumholz believes speculation about the effectiveness of Xarelto and any comparison with Pradaxa is somewhat premature given the data remains at a top line level. “I think it’s really important to see what the paper looks like after it goes through peer review”, he says.

In the meantime and despite warnings about making indirect comparisons, the debate will rage as to the relative merits of Xarelto to Pradaxa, and even warfarin itself, which remains a very effective treatment for AF patients.

As Dr Sacco points out: “Warfarin works but it is underutilised, because of blood-testing, risk of bleeding and difficulty of food interactions.”

The new generation of oral anticoagulants may not blow warfarin out the water just yet, in terms of demonstrating they are much more effective at preventing strokes. But the fact they are safer and much more convenient, allowing more AF patients to be treated, supports their massive commercial potential.

This much the cardiology experts at AHA certainly agree on.

Trial ID NCT00403767 (Rocket-AF)

Share This Article