July 31st, 2012
TAVI: Belgian Researchers Slam Evidence Base and Overuse in Europe
Larry Husten, PHD
The growing and enthusiastic adoption of transcatheter aortic valve implantation (TAVI) in Europe has no justification, according to three researchers who performed a health technology assessment for the Belgian government. In a paper published in BMJ, the authors from the Belgian Health Care Knowledge Centre conclude that TAVI should only be used in patients “who are deemed inoperable for technical reasons,” which is about 10% of patients “currently considered for treatment.”
In their paper, the writers do not identify any new concerns about TAVI, but they weave together the various threads of criticism that have been directed at the dissemination of TAVI. Hans Van Brabandt, Mattias Neyt, and Frank Hulstaert also charge that the evidence base for TAVI is deeply flawed as a result of an unpublished negative trial, serious baseline imbalances between controls and TAVI patients in the PARTNER trial, and unreported conflicts of interest by the principal investigator of PARTNER.
By the end of 2011, approximately 40,000 TAVI procedures had been performed, they write. Nearly all of the procedures were performed in Europe, where devices do not undergo the same regulatory scrutiny as drugs, needing only a CE mark to go on the market, “putting them on the same footing as domestic appliances such as toasters.” They note that TAVI was used in Europe for 4 years before the FDA approved its use in the U.S., but only for patients who were not surgical candidates and only when the transfemoral approach was used.
Despite the greater rigor of the U.S. regulatory process, the authors write that after their careful review of all the available evidence, “we remain far from convinced that it is adequate” and that “the arguments supporting the widespread use of TAVI do not stand up to scrutiny.”
Although in the PARTNER A trial TAVI met the predefined criteria for noninferiority when compared with surgery, the authors write that strokes, TIAs, and major vascular complications occurred more often with TAVI. TAVI performed better in the PARTNER B trial, which studied patients who were not eligible for surgery, but TAVI was still associated with a higher rate of stroke and vascular events. Results of the two trials “suggest that TAVI can be justified for inoperable patients on clinical grounds,” but, write the authors, this conclusion is weakened because the FDA, Edwards (manufacturer of the Sapien TAVI device), and trial investigators have failed to provide data to the authors, despite repeated requests, about a follow-up study comparing TAVI with standard therapy in 90 patients. The failure to share the data, they write, “is both ethically and scientifically unacceptable and should be legally regulated in future.”
The authors also maintain that the principal investigator of the PARTNER trial, Martin Leon, did not fully disclose his financial interest in TAVI. Although it was disclosed in the NEJM papers that Leon had received $6.9 million from Edwards when it purchased Percutaneous Valve Technologies, which Leon had cofounded, the articles did not mention that Leon “was to receive three further payments on the achievement of three milestones: successful treatment of 50 patients, regulatory approval in Europe, and limited approval in the US.”
The biggest concern about TAVI, write the authors, is widespread use of the transapical appraoch in patients in Europe, despite what they term a “dearth of evidence.” Overall, there is an absence of evidence to support usage patterns in Europe, they write. They note that the British Cardiovascular Intervention Society and the Society of Cardiothoracic Surgeons have called for randomized trials of TAVI, but “only when centres in the UK have got ‘beyond their learning curve.'” They go on to write: “Patients may be surprised to hear that trials are being delayed to allow cardiologists and surgeons time to learn the technique.”
The CMS had access to this HTA report as part of their due diligence on NCD for TAVR. While the authors raise several interesting points which are hard to disagree with, nonetheless they overlook the fact that the device passed the FDA’s robust regulatory muster for “inoperable” indication, a standard they wish the European device approval process would aspire to. Although randomization attempts to minimize imbalance amongst the treatment groups, it does not always ensure a “random” sample! Overall, in my opinion, none of the issues raised by the authors invalidate the results of PARTNER trial. While the benefit-risk balance in inoperable patients clearly favors TAVI over routine medical care, it is arguably questionable for high-risk patients who are otherwise suitable candidates for surgery. This then becomes more of an issue of patient preference and value judgment regarding what constitutes an acceptable benefit-risk tradeoff. I agree that “indication creep” is the elephant in the room. It is the collective responsibility of all the stakeholders to ensure appropriate, judicious and evidence-based use of this technology. The lack of access to Continuous Access Registry data is vexing and raises the challenge of reconciling proprietary information with timely and transparent public disclosure. The FDA does not have the legal authority to force the sponsor to make these data publicly available. However, perhaps the payers can help “concentrate” the sponsors mind!
Very disturbing is the financial conflict of interest!!!
TAVR represents a remarkable technological advance but the onus is on physicians who refer patients for this procedure to set realistic expectations. In PARTNER B, 30% of patients with TAVR diet after one year, and over 40% died at two years. This was better than medical therapy but certainly not a panacea. I agree with Dr. Kaul that what constitutes an acceptable benefit-risk tradeoff is critical in discussions with patients, given the attendant risks of stroke and vascular complications with TAVR, and the possibility that in certain very high-risk subgroups (yet to be defined) it may not confer meaningful clinical benefit. Registry data in the U.S. will help to describe how this technology disseminates into standard practice, and may capture the “indication creep” noted.