CMS’ proposal to limit coverage for unapproved transcatheter aortic valve replacement indications to clinical trials intended to prove superiority rather than non-inferiority to alternatives was widely panned by stakeholders in recent comments to the agency.

Some in the device industry worry the provision signals CMS’ intent to play a more active role in dictating clinical trial designs.

The Medicare agency proposed a national coverage policy for TAVR in February, just three months following PMA approval of Edwards Lifesciences’ Sapien valve for inoperable aortic valve disease patients. Specifically, CMS proposed a coverage-with-evidence-development (CED) policy, in which FDA-approved indications would automatically be covered by Medicare at facilities that meet certain credentialing requirements and if the beneficiary is entered into a registry that collects clinical data for long term follow up. (See Also see "Transcatheter Valve Coverage Proposal Reflects Stricter Medicare Environment" - Medtech Insight, 13 February, 2012..)

But until an indication is approved, it would only be covered in the context of a controlled clinical trial. Both clinicians and companies were caught off guard by the specific requirements the proposal includes on what will and will not be an acceptable trial design, and, in particular, the mandate for only superiority trials.

“For CMS to exclude unilaterally a particular clinical trial study design from consideration for CED coverage appears to be a radical change in policy that could deter enrollment of Medicare beneficiaries in future clinical trials,” writes St. Jude Medical exec Christopher Volker.

The proposal runs counter to previously stated CMS positions on the agency’s proper role in clinical trial design, groups commenting on the proposal point out. The requirement could also hinder important clinical research efforts for the promising new procedure and may even encourage subpar trial designs, they argue.

David Holmes, president of the American College of Cardiology, for instance, argues that positive results from Edwards’ PARTNER study make the design of a superiority study extremely difficult.

“We are concerned that such a requirement will effectively put an end to research into off-label uses in the U.S. because the thresholds for superiority trials are so high that studies would be too difficult to perform or would take many years,” he writes March 3.

FDA has already shifted its focus to non-inferiority trials. The PARTNER/cohort A study supporting Edwards’ pending PMA for Sapien in patients at high risk for surgery was a non-inferiority design comparing Sapien to open surgery. And both Edwards’ (PARTNER II) and Medtronic’s (CoreValve) ongoing trials for transcatheter valves involve non-inferiority designs. (See [A#111108g1].)

Non-inferiority and equivalence study designs, including prospective, registry-based observational studies, should be included in the national coverage allowances, ACC contends.

“The critical issue for CMS to evaluate in determining whether a study is qualified for CED will be whether the study design is appropriate for the research question proposed and feasible for the patient population selected,” Holmes writes.

A non-inferiority design can be, and often is, used to demonstrate advantages of one treatment over another, just not on the primary endpoint, argues Thomas Armitage, VP of cardiovascular clinical research at Medtronic. For instance, a new version of a device might be essentially equally effective to an older device, but may be more tolerable to a patient or safer by some measures.

“The presence of a non-inferiority hypothesis for the primary endpoint for any given clinical trial, therefore, does not simply signify the desire to only show equivalence of a candidate product to an extant standard of care on all levels that may provide value to the patient,” Armitage writes March 2.

Limiting reimbursement to only superiority trial designs “may lead to nonsensical endpoints just to meet the superiority definition,” contends Parashar Patel, global VP for health economics and reimbursement for Boston Scientific, which has TAVR technology in its pipeline. The restriction would add “a non-clinically driven influence to study design,” he said.

Further, CMS’ proposal breaks with its established policy, companies and industry groups argue. “For CMS to exclude unilaterally a particular clinical trial study design from consideration for CED coverage appears to be a radical change in policy that could deter enrollment of Medicare beneficiaries in future clinical trials,” writes Christopher Volker, St. Jude Medical’s VP of cardiovascular economics, in March 2 comments.

Volker and others cite CMS’ 2006 guidance document on coverage with evidence development, which states, “CMS will not routinely be involved in the design, review, or execution of research studies.”

In previous coverage-with-evidence-development coverage proposals, “CMS has not provided specific instructions on the design of the studies that would be eligible for coverage; and neither has CMS ruled out particular methodological approaches,” writes Ann-Marie Lynch, executive VP of payment and health care delivery policy for AdvaMed.

“If the agency intends to reverse direction on this policy, we believe there should be adequate engagement with a full range of stakeholders in setting parameters for the CED data collection,” Lynch said in Feb. 29 comments.

Broader CED Debate Continues

Opposition to the superiority trial requirement, which was among the most commonly cited issues in comments to CMS among companies and clinicians alike, dovetails with a general discomfort by industry with how the agency is pursuing its approach to CED requirements.

In the TAVR proposal, CMS requires beneficiaries treated with the procedure as part of an FDA-approved indication to be enrolled in a registry that tracks various endpoints, including stroke, vascular events and quality-of-life measures for at least five years. But device makers believe substantial details and constraints are lacking.

Companies ask the agency to collaborate with stakeholders to narrow the clinical questions to be addressed by a specific CED data collection; require the collection of only the minimum necessary data; and ensure that an appropriate endpoint is set in advance of CED data collection.

“As currently drafted, the registry requirements related to the proposed NCD do not appear to include any limitations on time, number of patients or follow-up period and appear to be open-ended,” AdvaMed’s Lynch writes.

The arguments match points made by companies earlier in the year in response to a general comment request from CMS on the CED program. (See Also see "Device Makers Want Say In Medicare 'Coverage-With-Evidence-Development' Policy" - Medtech Insight, 30 January, 2012..) CMS collected those comments as part of ongoing reassessment of the program that might lead to expanded use of CED. (See Also see "CMS Coverage-With-Evidence-Development Policy Could Face Revamp" - Medtech Insight, 14 November, 2011..)

And last week, as part of that effort, CMS announced that its Medicare Evidence Development and Coverage Advisory Committee will meet May 16 to review the “desirable characteristics” of evidence appropriate for the coverage-with-evidence-development program.

That meeting, however, will come after a final decision is expected on the TAVR coverage policy May 2.

Superiority Standard? Groups Bristle At CMS’ Transcatheter Valve Trial Proposal

More from Policy & Regulation

More from Medtech Insight