Merck’s Vorapaxar CV Outcome Follow-Up Success Not Ready To Become Standard
Executive Summary
Cardiovascular and Renal Drugs Advisory Committee says FDA should encourage lost-to-follow-up rates closer to what was seen in Merck’s TRA2P study of vorapaxar, but shouldn’t expect it.
FDA advisory committee members were impressed with the low lost-to-follow-up numbers in Merck & Co. Inc.’sTRA2P cardiovascular outcomes trial but were not ready to set it as the new expectation for cardiovascular trials going forward.
FDA may be inclined to advise sponsors to implement some of Merck’s methods for the trial of the antiplatelet agent Zontivity (vorapaxar), however, as part of an effort to improve patient follow-up in massive morbidity and mortality trials.
Missing vital status data has become an issue with many of the major cardiovascular adverse event trials conducted for the novel antithrombotics, at times affecting the confidence in the study results. For example, FDA reviewer Thomas Marciniak drew publicity when he questioned the data in the PLATO study used to approve AstraZeneca PLC’s Brilinta (ticagrelor) (Also see "Brilinta Review Sparks War Of Words Between FDA Cardio-Renal Division’s Marciniak, Stockbridge" - Pink Sheet, 1 Jan, 2012.).
Marciniak also found problems with the ATLAS trial used to support Bayer HealthCare AG/Janssen R&D LLC’s application for Xarelto (rivaroxaban) in acute coronary syndrome.
Vital status originally was missing in 8% to 9% of the more than 15,000 patients in ATLAS, and Marciniak argued that such a rate invalidates a CV outcomes trial (Also see "ATLAS Data Not Strong Enough To Support Xarelto’s Use In ACS, FDA Panel Says" - Pink Sheet, 23 May, 2012.). The extent of the missing data led to a “complete response” and another advisory committee review, held the day after the vorapaxar meeting. Merck’s contrasting success was brought up at that meeting (Also see "Janssen Faces Prospect Of New Trial For Xarelto ACS Claim" - Pink Sheet, 20 Jan, 2014.).
Janssen’s attempts to go back and track down vital status information on subjects who withdrew from ATLAS were not enough to persuade the committee that the drug should be approved in ACS (see sidebar).
Merck had admirable follow-up figures in both TRA2P and its TRACER trial. The two studies were the basis for the vorapaxar NDA seeking an indication for reduction of atherothrombotic events in patients with a history of myocardial infarction, with a demonstrated reduction in the rate of cardiovascular death, MI, stroke and urgent coronary revascularization. The committee on Jan. 15 voted overwhelmingly to recommend approval (Also see "Merck’s Vorapaxar Overcomes Trial Questions, Committee Recommends Approval" - Pink Sheet, 16 Jan, 2014.).
The company managed to lose only a handful of patients during the follow-up portion of TRA2P, a secondary prevention trial in patients with a history of atherosclerotic disease.
Of the more than 26,000 patients in the TRA2P placebo and vorapaxar arms, only 32 (0.1%) were considered lost to follow-up, according to FDA advisory committee briefing documents.
Cardiovascular and Renal Drugs Advisory Committee members said the rate was significantly better than other similar trials.
“It’s one of the lowest numbers of lost-to-follow-up that we’ve seen in recent trials,” said Sanjay Kaul, University of California at Los Angeles.
A New Standard?
The comments prompted FDA to ask the committee whether it should still “tolerate the usual” or whether TRA2P’s results should be the new expectation.
Committee members remained hesitant about demanding lost-to-follow-up rates that low but said the agency can expect better.
“I would say we should tolerate much more on this side than some of those large numbers that approach 10% we’ve seen on the other side,” said acting advisory committee Chairman Philip Sager of Stanford University.
TRA2P tested 2.5 mg vorapaxar once daily in addition to standard therapy, including other antiplatelet agents, compared to placebo. TRACER tested a 2.5 mg daily dose of vorapaxar after a 40 mg loading dose in addition to standard therapy, including other antiplatelet agents, versus placebo.
TRACER, in patients with acute coronary syndrome, was not as successful in its follow-up after the study was completed but was still laudable. Of the nearly 13,000 patients in that study, 180 (1.4%) were listed as lost to follow-up (see chart below).
No One-Size-Fits-All Approach
While the committee agreed that the follow-up was particularly good in TRA2P, members also warned the agency that it may not be easy to glean common practices from the trial.
Scott Emerson, a biostatistics professor at the University of Washington, said there was no “one-size-fits-all” solution for what causes problems with lost to follow-up.
Stakeholders have agreed that large clinical trials need simplifying. Typical trials now encompass numerous endpoints and lengthy protocols and case report forms. An Institute of Medicine report concluded that some policy changes were possible, but one of the issues that needs to be addressed is the gap between what data sponsors believe FDA requires and what agency officials actually want (Also see "Large Simple Trials Need FDA Boost, IoM Report Says" - Pink Sheet, 14 Oct, 2013.).
Committee industry representative Rob Scott, head of Amgen Inc.’s cardiovascular therapeutic area, said TRA2P should not be a bellwether of coming improvements.
“I think we either have to stop creating new drugs or we have to react to what’s happening,” he said. “There are very structural reasons why lost-to-follow-up rates are going up. I think this [TRA2P] is particularly good. We’ve seen some others that are particularly bad. But historically I think we’re going to see that the numbers continue to go in the wrong direction.”
FDA’s Advice: Pick The Right Countries
FDA already is advising sponsors on some clinical trial methods, like site selection, to improve follow-up and potentially overcome some of the structural issues that Scott hinted are causing problems.
Robert Temple, Center for Drug Evaluation and Research deputy director of clinical science, said the agency has recommended sponsors avoid some countries as clinical trial sites because they limit patient follow-up abilities.
“We’ve only been giving that advice for the last couple of years,” Temple said during the meeting. “As the people have been accepting it, sometimes the countries change their mind.”
FDA would not divulge the countries the agency recommends against using as clinical trial sites because “that information would vary.”
Temple said in a Jan. 24 emailed statement that sponsors should look to gain patient consent for follow-up.
“Our point is … that patients should be accounted for to the extent possible,” he said. “Vital status is one aspect, but we are also urging companies to try to gain [the] willingness of patients to be followed even when they have stopped the drug therapy.”
Merck officials said they structured their consent forms for the study to include allowing patient follow-up even if they withdrew from the dosing part of the study.
David Morrow, senior investigator for the TIMI Study Group, said it also was made clear to the investigators at the beginning of the trial that follow-up would be important.
“We exhausted efforts as a study team to help support them in doing that and also helped them, teach them, the sites and the coordinators, exactly what it meant to follow a patient adequately,” Morrow said. “I think there’s still old perspectives that don’t fully understand what drug consent means and what loss-of-contact means.”
CDER Office of Drug Evaluation I Director Ellis Unger said TRA2P seemed to include physicians who already had a relationship with the patients, which may have been a reason for the success in follow-up. But he acknowledged some advisory committee members did not agree.
Committee members also struggled with the vorapaxar labeling language, including how to make it useful to clinicians without unduly limiting the product’s use.
When it came to a potential problem among patients weighing less than 60 kg, the committee could not reach a consensus about how and where to mention the issue in the label, in part because weight alone does not drive all clinical decisions (Also see "“Use With Caution” Isn’t Specific Enough For Labeling, FDA Tells Merck At Vorapaxar Panel" - Pink Sheet, 20 Jan, 2014.).
Vorapaxar Trial Statistics | ||
Description | Placebo | Vorapaxar |
TRACER | ||
Total Patients | 6,471 | 6,473 |
Completed study therapy | 4,559 (70.8%) | 4,475 (69.4%) |
Prematurely discontinued follow-up | 409 (6.3%) | 352 (5.5%) |
Withdrew consent for follow-up | 315 (4.9%) | 266 (4.1%) |
Lost to follow-up | 94 (1.5%) | 86 (1.3%) |
TRA2P | ||
Total Patients | 13,224 | 13,225 |
Completed the study | 12,932 (97.8%) | 12,953 (97.9%) |
Prematurely discontinued follow-up | 292 (2.2%) | 272 (2.1%) |
Withdrew consent for follow-up | 277 (2.1%) | 255 (1.9%) |
Lost to follow-up | 15 (0.1%) | 17 (0.1%) |
Source: FDA advisory committee briefing materials | ||