Professional Egos, Cultural Norms Clash At Xarelto AdComm

Let’s be honest – we look forward to some FDA advisory committee meetings more than others. Some meetings, such as those convened only because a product is a new molecular entity but with no controversial safety or efficacy issues attached, can be a bit of a yawn from a newsgathering perspective.

Give us a drug that has hit some regulatory bumps in a hot and highly lucrative therapeutic area, and it raises our journalistic antenna.

So, when we saw the FDA staff’s strongly negative review of the atrial fibrillation claim for Bayer/J&J’s Xarelto (rivaroxaban) – a review that went so far as to recommend a “complete response” action – we knew we were in for an interesting meeting of the agency’s cardio-renal committee on Sept. 8.

Couple that with the revelation that there were going to be some big names participating in the proceedings (Rob Califf! Steve Nissen! Sanjay Kaul! Tom Fleming! – none of whom could be accused of being afraid to share their views on a subject), we were practically salivating.

Our giddiness was further heightened when we saw FDA’s draft questions, which spanned six pages. FDA wanted to know whether rivaroxaban’s efficacy was appropriately studied against warfarin in the ROCKET AF trial, or whether the drug’s efficacy is best viewed against that of Pradaxa (dabigatran), Boehringer Ingelheim’s oral direct thrombin inhibitor that was approved last October.

Most of the committee members – and FDA’s cardiology guru himself, ODE I Director Bob Temple – were pretty clear in their views that the Xarelto sponsors shouldn’t be required to show rivaroxaban’s efficacy against that of another drug on the market for less than a year. (This week’s issue of “The Pink Sheet” offers in-depth analysis of how FDA might continue to use its “as effective” standard.)

Thanks to these views, drug developers appear to have escaped the potential calamity of having to go back to the drawing board to compare their late-stage compounds to the newest kid on the block for FDA approval. (Reimbursement is another story, but we digress.)

Nonetheless, the discussion over the appropriate comparator for rivaroxaban raised the lingering problem of how well clinical trials mimic real world use when it comes to establishing efficacy against another drug, and whether it’s wise to attempt to mimic actual medical practice when that offers a comparison that is far from ideal.

While warfarin may be the appropriate comparator in the eyes of most committee members and top FDA brass, a debate over how well warfarin is used in clinical practice and in the ROCKET AF trial led to a clash of egos and a lesson in cultural norms at the meeting.

One key issue before the committee was how well anticoagulation control with warfarin was managed in ROCKET AF, as gauged by International Normalized Ratio time in therapeutic range (TTR).

In their briefing documents, FDA reviewers took the ROCKET AF investigators to task for what they considered to be the less than “skillful” use of warfarin when compared to other anticoagulant studies.

The mean time in TTR was 55% across the ROCKET AF program, compared with 64% for dabigatran’s pivotal RE-LY trial. The lower TTR in ROCKET so distressed the agency reviewers that they used the word “skillfully” 16 times in their briefing document as they sought to drive home the point that it is difficult to ascertain rivaroxaban’s true efficacy when compared against warfarin that is well managed.

But some in the meeting room didn’t take too kindly to FDA’s characterization of unskillful use of a drug that has been a standard of care for stroke prevention in atrial fibrillation patients for years.

Duke University Vice Chancellor of Clinical Research Robert Califf, appearing on behalf of the sponsors and the study’s executive committee, maintained that the average TTR of 64% for North America was better than U.S. practice in general and comparable to the North American results in other anticoagulant studies.

“This is hardly a result that should be attributed to unskillful clinicians,” Califf said. “I have to say on behalf of the executive committee and over 1,000 investigators, we don’t think we’re unskilled. We think we’re pretty good.”

Califf maintained that the lower overall mean TTR in ROCKET AF was due in large part to the geographical diversity of the study population and the inclusion of patients at higher risk of stroke than in other anticoagulant trials. More than 38% of the patients were enrolled in Eastern Europe, where TTR was 49.7%. Another 15% came from the Asia-Pacific region, where TTR was only slightly better, 52.4%, according to FDA’s calculations.

Regional variations in how quickly patients were brought back for INR assessment resulted in an “artificial” lowering of TTRs in areas where there was a longer delay, such as in Asia-Pacific and Eastern Europe, Califf said.

Warfarin dosing was managed at each site according to local standards, and there was no centralized protocol for managing patients who fell outside of therapeutic range. Site investigators were educated on the importance of keeping INR within the 2.0-3.0 target range.

The geographical distribution of patients, coupled with the absence of a single warfarin dosing protocol, touched off one of the more interesting debates of the day between Califf and Cleveland Clinic’s Steven Nissen, a temporary voting member on the panel.

“Given the fact that I think you guys knew that the TTR was going to be a critical factor in the approvability of the drug, I’m just puzzled why you didn’t make an effort to give sites, particularly Third World sites, some guidance about what to do,” Nissen said to the sponsors. “Basically, you sort of left them on their own, and my concern is that’s why you got such a low TTR” across the ROCKET AF study.

At this point, Califf embarked on a lesson in global diplomacy.

“Dr. Nissen I’m sure you didn’t mean to use the term Third World, because I don’t think that’s an appropriate term anymore, and in fact if you look at the news today you would find what you’re calling Third World actually owns most of our national debt,” said Califf, whose remarks were met with laughter in the room.

“I think the investigators around the world will be highly offended by the sort of American exceptional tone of some of the discussion, not you personally, but – ”

At this point, Michael Lincoff, Nissen’s colleague at Cleveland Clinic and acting chairman of the committee, attempted to redirect the discussion away from geopolitical issues and back to anticoagulants.

“Let’s focus on the subject,” Lincoff said, before he was cut off by Nissen.

“Rob, you showed us that those sites in those countries didn’t do a very good job of managing warfarin. That’s where all this comes from,” Nissen said to Califf.

“But these two concepts are related,” Califf responded. “It’s not that doctors do a lousy job. This is a really hard drug to use and it’s especially really hard to use in various social, economic personality characteristic circumstances. You know that as well as I do.”

“What we felt obligated to do was to do better than standard of practice given the heterogeneity of the practices in those centers,” Califf said. “We spent a lot of time at every site saying how do you do it? Let’s make sure what you’re doing is OK, but let’s not force you to use one specified algorithm. Let’s make sure you make the appropriate dosing changes. I showed you that it wasn’t that they weren’t making the appropriate dosing changes. They just brought the patients back when it was feasible to do it.”

With the sparring over lesser developed countries behind them, the committee continued with its questions and discussion, though later in the meeting other panel members felt the need to weigh in on the skillfulness issue.

TTR is “certainly a marker of skillful care” but is not an effect modifier, Cedars-Sinai cardiologist Sanjay Kaul maintained.

Yet, fellow panel member and cardiologist Darren McGuire, University of Texas Southwestern Medical Center, was quick to take offense. McGuire conceded the TTR rate at his institution is only about 54%, but he noted that numerous patient level characteristics, and not the treatment provided by a given physician, play a role in how well TTR is maintained.

“I will agree with Dr. Califf’s comment earlier. I take some offense at the term unskillful use, because it’s a little bit accusatory and it’s a complicated therapeutic strategy.”

Some committee members said they appreciated the lack of a standardized dosing protocol in ROCKET AF because it more closely mirrored how warfarin is used in clinical practice, which is imperfectly.

The issue of how well INR was maintained came up in the advisory committee review of Pradaxa as well. Boehringer was lauded for the level of control achieved for warfarin in RE-LY, which had an open-label design for that reason (though the trial design also drew criticism). But the INR variability also kept FDA from granting an outright superiority claim; the agency included cautionary language in labeling noting that the benefit was driven by centers where INR control was poor.

The back and forth at Xarelto review begs the question of whether better data is gained by mimicking real world use, though it may mean imperfect use of a standard of care, or if tightly controlled trial practice is a better comparator for convincingly demonstrating efficacy in the minds of regulators and payers.

While answers may not be quickly or easily forthcoming, it’s an issue that other anticoagulant sponsors should be prepared to address.

– Sue Sutter

Photo, "Triplets of Cap Hill," by Flikr user daniel spils, reproduced under Creative Commons license