Here’s my live-blog of the FDA’s meeting of the Circulatory System Devices Panel to consider the premarket approval (PMA) application for the Edwards Sapien Transcatheter Heart Valve for use in patients with symptomatic severe aortic stenosis who have high operative risk. As reported here earlier this week, FDA reviewers have raised a number of questions about the safety and efficacy of the Sapien heart valve system. Here is a link to the FDA and Edwards briefing documents.
6:00– The committee votes 10-2 that Sapien is safe, 12-0 that it is effective, and 11-0 (1 abstention) that the benefits outweigh the risks.
5:55– Here’s questions 2 and 3:
VOTING QUESTION 2: Is there reasonable assurance that the Edwards SAPIENTM Transcatheter Heart Valve is effective for use in patients who meet the criteria specified in the proposed indication?
VOTING QUESTION 3: Do the benefits of the Edwards SAPIENTM Transcatheter Heart Valve for use in patients who meet the criteria specified in the proposed indication outweigh the risks for use in patients who meet the criteria specified in the proposed indication?
5:54– Wow– computer voting system is broken! They’re going to vote on paper.
5:48– After they finish the legal mumbo jumbo the voting will begin.
5:43– Getting close to the voting questions. Here’s #1:
VOTING QUESTION 1: Is there reasonable assurance that the Edwards SAPIENTM Transcatheter Heart Valve is safe for use in patients who meet the criteria specified in the proposed indication?
5:39– Edwards via Craig Smith giving their summary. He says he’ll be brief.
5:38– Now talking about study data that should be included in the labeling.
5:30– Now talking about post-approval studies to get further insights into stroke, aortic regurgitation, long term effectiveness, valve-in-valve, and gender issues.
5:28– Now discussing the totality of the evidence. Chairman states that the committee agrees that Sapien is safe and effective, with a number of caveats, such as stroke, TA/TF, gender, AR. (Why bother voting now?)
5:20– Committee now discussing informed consent and whether a more detailed informed consent form should be required. An additional piece of paper is not really the ideal solution.
5:15– Chairman summarizes that the FDA is concerned that valve-in-valve use may occur, but there’s no data to support this. Committee strongly encourages getting data on this and it should not be endorsed on the label. Committee wants to motivate the company to perform a separate clinical trial for this indication.
5:10– Now discussing long-term durability. Everyone agrees there’s not enough data on this. But the 3 year data presented earlier was encouraging. Kandzari notes that in a population this old it will be difficult to gather enough 5 year followup data.
5:08– Chairman’s summary about the difference in gender: intrigued but don’t understand these results. At the moment it’s a post hoc observation that we’re having difficulty understanding clinically. Naftel objects that it’s not a post hoc analysis, it’s a necessary analysis, but that the trial is not powered to find a significant result in these subgroups.
5:07– Everyone agrees that it should be indicated for people with symptomatic severe aortic stenosis. Adding more details would interfere with the “art of medicine.”
5:00– To be clear, this panel is going to vote overwhelmingly to approve. The devil is in the details. They are all talking about with the assumption that this gets approved.
4:59– 94% of people in PARTNER were class 3 or 4. Panel thinks that TAVR might not be suitable for class 2 but Rick Lange and others argue that any symptomatic aortic stenosis needs to be fixed. Everyone agrees that people need to be symptomatic.
4:51– Now discussing indications for use. Panel agrees with CMS that in this population a cardiac surgeon also will need to be involved in the decision.
4:48– Everyone agrees that the transapical population is at higher risk and you would expect to see a higher event rate in this group. No one seems concerned by the finding that numerically there were more deaths in the TA arm with TAVR.
4:43– The most interesting thing I heard all day is that Adele’s “21” regained the #1 position on Billboard.
4:34– Everyone agrees that there were problems with the trial but these were probably inevitable and in the end they won’t make a difference in the decision.
4:30– Now moving to trial conduct questions.
4:22– Everyone agrees aortic regurgitation is a problem but, again, not enough data to form a strong opinion about it. Again, this will impact the labeling but doesn’t seem to provoke a big negative response.
4:08– Panel agrees stroke ascertainment in the trial was pretty good but not perfect. Now discussing antithrombotic regimen, but there’s so little data on this that no one has a strong opinion. Lack of data is a real drawback, says one panelist.
4:01– Chair: now we get to the meat of the matter. Our task is to answer the FDA questions. First set of questions concern safety. See the complete text of the questions here.
3:44– Break until 3:55. Then the panel will address the FDA questions.
3:40– Lots of uncertainty about the details but I don’t see a lot of negativity in the panel.
3:36– Panelist says women do better with TAVR because women are more likely to be at high surgical risk. But Rick Lange points out that the 1-year mortality is the same.
3:34– Nothing against these guys but they lack the fireworks of Steve Nissen and Sanjay Kaul.
3:32– Here’s the table on the difference in the primary endpoint between the TF and TA stratum: yfrog.com/odxdoicrj
3:24– Kandzari points out that differences between the randomized patients and CAP patients is partly due to the fact that some TAVR patients would not have been eligible for PARTNER because they would have refused surgery.
3:20– Panelist Naftel: even though you know it’s not powered for subgroups it’s impossible not to look at the subsets like men and women and TF and TA. He also notes that potential patients will need to be told that there was no “winner” in PARTNER, that it showed noninferiority. Really wants to see the registry in place to help prevent these sort of problems.
3:13– One panelist says he’s a little less concerned about stroke now than at the start of the meeting.
3:10– Questions to sponsor and FDA ends. Panel now begins deliberating among themselves.
3:05– Lots of talk about how to evaluate patient risk and concern over criteria creep. Little progress being made but hard to see how this will impact the committee vote, though it may have a big impact on labeling and post-marketing studies.
2:45– Question about early mortality benefit with TAVR, and why the benefit is lost over time. Leon says they were surprised when they first saw this, and speculates that these old, frail patients couldn’t hold up to the immediate rigors of surgery, but then the TAVR patients catch up over time.
2:23– Graphic of 3 year KM curves: yfrog.com/odwtupp
2:19– Leon presents actual treatment 3 year KM mortality curves: 46.5% TAVR versus 46.2%
2:16– Leon now addressing the comparability of randomized and CAP patients? Although somewhat different, there are no major differences and CAP patients also needed to be approved by a heart valve team at each institution.
2:11– Leon says that they have recently found that 3D annular sizing is critical to limit paravalvular leak, thereby suggesting that this problem can be solved, and produce results comparable to AVR.
2:02– Edwards, Leon, et al trying to answer some of the panel questions about stroke, but the data is just too limited to shed any new light on the topic.
1:57: Open public hearing over. Panel now begins deliberations.
1:53– I’ll go out on a limb and predict that the committee will vote in favor of TAVR. I think they will take the FDA review criticisms very seriously, but the overwhelming support from professional organizations, the TVT registry, and the other CMS safeguards will give them sufficient assurance that TAVR won’t run amok in the community. FWIW.
1:49– SCAI, ACC, STS all speaking in favor of TAVR, with qualifications.
1:31– Emory heart surgeon emphasizes that TAVR is a shared decision between patients, cardiologists and surgeons.
1:26– Jeffrey Moses now giving public comment, not as a Columbia interventional cardiologist but because his father received a Sapien under the continued access protocol. (“I wisely stayed out of it.”)
1:20– TAVR patient got her hair done, received a Sapien, and a week letter made her next hair appointment.
1:17– Finally! FDA posts the panel roster:
1:12– Lunch over, public comment time. 92-year-old survivor of the Bataan Death March and Japanese prison camp fought to get TAVR.
LUNCH BREAK UNTIL 1 PM
12:12– FDA says that the earlier question about preventing over enthusiasm for TAVR will be addressed by the ACC and STS in the public presentation after lunch.
12:07– Here’s a link to a good image of the FDA summary slide of the risk/benefit comparison of AVR versus TAVR: yfrog.com/kjzjsokp
12:05– Panel really wrestling with the problem of how to incorporate the CAP data with the randomized PARTNER data.
11:53– Panelist raises concern of “slippery slope” in which TAVR gets first approved for inoperable patients but now gets approved for operable patients as well, and patients will prefer TAVR over AVR in the same way that patients choose stents over CABG, despite the superiority of CABG. How to avoid the problem of a stampede to TAVR?
11:43– First two questions concern the differences between TA and TF approaches. A harbinger?
11:41– FDA summary: restating earlier concerns. No clear message or guidance here. Panel questions to FDA now begin.
11:29– FDA now talking about post-approval studies to evaluate long-term performance, TA/TF differences, gender differences, and importance of aortic regurgitation.
11:28– Now the ACC has posted on its website its comments: http://www.cardiosource.org/Advocacy/Issues/~/media/Files/Advocacy/FDA/TAVRwrittenstatementJune2012.ashx
11:25– Reviewer summarizing: trial met the statistical endpoint but the assessment of risk/benefit is difficult and requires consideration of neurological and other events. In addition, because we don’t have long term data, the durability of the device remains in question.
11:17– Reviewer says that we have no way of knowing whether the increased aortic regurgitation in the Sapien group is “a gift that may keep on giving” even after 2 years. Another red flag.
11:06– FDA reviewer talking about delay between randomization and procedure– much longer delays in the surgical group. Also control patients not undergoing AVR, and a very large variation in the sites in the number of patients randomized to AVR who did not undergo AVR.
11:04– Now getting into heterogeneity of treatment– differences between the two randomized groups that emerged during the trial, including crossovers. This will be a key issue.
11:00– “there are no perfect trials”. She points out that 80% of the TAVR patients received an incision, but many AVR patients get minimally invasive surgery. Her tone suggests that she is taking a critical view of PARTNER. Much of what she’s presenting was not published in NEJM or elsewhere.
10:58– FDA clinical reviewer makes the point that although the device improves over time (this is often stated) operative technique also improves over time.
10:56– SCAI just posted Augusto Pichard’s statement that I’m guessing he will present during the public session: http://t.co/eWiqegiO
10:50– Statistical reviewer reports very large upper confidence interval of 11.7% in the mortality difference in the TA approach.
10:35– FDA presentation now beginning.
PANEL ON BREAK UNTIL 10:30…
10:17–FDA says that it has trouble relying on OUS (outside US) data, and asks the panel to concentrate on the data in the briefing book, based on PARTNER and the US continued access protocol data.
10:13– David Kandzari asks about the differences in outcome between men and women, and wants to know if the distribution of TF/TA access was the same in the sexes. Leon said they will try to provide the answer in the afternoon.
10:06– Leon responds to question about larger proportion of TA (transapical) cases in the continued access protocol (CAP). In CAP there is a larger proportion of TA cases than in PARTNER, largely because the TA stratum didn’t start until 11 months after the start of the TF stratum. This may be a big issue in the afternoon, since the TA stratum did not do as well as the TF stratum.
9:53– Upon questioning from panelist Rick Lange, Leon states that the vast majority of early strokes were clinically serious or debilitating. Leon also confirms Lange’s suspicion that TAVR patients with no regurgitation had much better outcomes than TAVR patients who had regurgitation.
9:50– Discussion about the problem of comparing TAVR to AVR, because surgeons often perform concomitant bypass while performing AVR.
9:47– Interesting but confusing interchange about FDA problems with Edwards retrospective analysis of stroke data. Hard to know exactly what this means at this point.
9:43– Panel now has chance to ask clarifying questions to Edwards.
9:41– Addressing the important issue of aortic regurgitation after TAVR, Smith says this is being addressed by device improvements and greater experience and that the “consequences are not fully understood.”
9:36– Craig Smith now giving summary and conclusion of Edwards presentation.
9:32– Edwards VP Larry Wood now talking about the Edwards training program of cardiologists and surgeons.
9:30– I’m a bit surprised that Smith and Leon didn’t really address the issues raised by the FDA reviewers in the review documents (see above). I think the action will really get started when the FDA presentations begin.
9:27– Edwards VP Jodi Akin now reviewing ongoing and post-approval studies.
9:14– Leon now presenting the data on strokes– more strokes in the TAVR group in the first 30 days, similar number of strokes after 30 days. This will likely to be an important part of the discussion this afternoon.
9:06– In reviewing PARTNER results, Leon explores the difference in outcomes between men and women. Women generally did better than men with TAVR. Leon said the increased mortality in males with TAVR is probably caused by their co-morbidities. By contrast, it is well established that women experience more late complications after surgery.
8:52– PARTNER was “a bonding experience” between interventionists and surgeons.
8:47– Martin Leon now presenting PARTNER results.
8:46– 30-50% of patients with severe aortic stenosis are untreated.
8:41– Smith emphasizing that many patients who are eligible for surgery may be good candidates for TAVR. Surgery is technically feasible but difficult in many patients, making TAVR an attractive alternative even from the surgeon’s perspective.
8:35– Craig Smith will talk about the need for TAVR and Martin Leon will talk about PARTNER. Smith is the chair of the department of surgery at Columbia and is a co-PI of PARTNER.
8:28– Now moving to Edwards presentation. Edwards is reviewing the pivotal PARTNER A trial along with 1800+ patients in the continued access protocol (CAP).
8:22– TVT registry was a requirement of the November 2011 initial approval of Sapien for non-operable aortic stenosis patients. Represents a “paradigm shift” for post marketing surveillance. 42 sites currently enrolled in registry.
8:12– After some technical glitches, FDA director of epidemiology Danica Marinac-Dbic giving a postmarket update, focusing on the TVT registry. This is the STS/ACC registry (more information here) designed “to track patient safety and real-world outcomes related to new TVT procedures.”
8:03– Panel now underway. Introductions and formalities. Anyone know why the FDA didn’t post the roster of the panel participants?