CMS should review the FDA’s Alzheimer drug decision

The Medicare Evidence Development & Coverage Advisory Committee should collect and analyze patient results

My grandfather died senile (that’s what they called it in 1959) at age 69. My father died at 76 in 1995 after a ten-year decline from Alzheimer’s disease. I turn 71 this month, so I have a stake in knowing if Biogen’s Aduhelm (generic name aducanumab) slows cognitive decline from AD.

I won’t review all the details about how the Food and Drug Administration’s scientific advisory committee rejected approving the drug after two inconclusive clinical trials; how the Alzheimer’s Association, which gets $24 million a year from drug firms including a quarter-to-half million dollars from Biogen, lobbied the FDA for its approval; or how the FDA at the last minute used the accelerated approval process to justify approving the drug. Under accelerated approval, the FDA relies on data showing an improvement in a surrogate marker – in this case, clearing plaque from the brain – and not clear-cut evidence of clinical improvement. The news and Twitter feeds are filled with that information today.

Rather, let me focus on where the government should go from here. Under accelerated approval, the FDA required Biogen to conduct another clinical trial to confirm that aducanumab actually slows cognitive decline. Many companies never complete these so-called Phase IV clinical trials, and when they do, they take years to complete.

But the FDA doesn’t have the final say on generating evidence of efficacy. The vast majority of patients who will be taking aducanumab will be on Medicare. Since it is an infusion drug (patients will visit their physicians’ offices once-a-month where they will be hooked up to an intravenous feed for their injections), it will be covered by Medicare Part B, which pays those clinics the average sales price of the drug plus 6% to cover the cost of infusion.

The Centers for Medicare and Medicaid Services (CMS) routinely pays for any FDA-approved drug. But it does have the option of referring payment approval to its Medicare Evidence Development & Coverage Advisory Committee (MEDCAC), which can recommend an independent review of existing evidence and seek additional data from companies and prescribing physicians as a condition of reimbursement.

“CMS is reviewing the FDA’s decision regarding aducanumab and will have more information soon,” an agency spokesperson said in response to my query. The statement also suggested the agency is leaning toward quick payment approval. “Today’s milestone provides hope to families, caregivers and the older Americans living with Alzheimer’s who rely on Medicare for health coverage.”

Register patients, collect data

Assuming CMS decides to pay for the drug, it can still use its authority to generate additional evidence ahead of the FDA-required clinical trial, whose results are at least a half decade away. The agency has the option of requiring every patient receiving the drug have their de-identified data entered into a registry.

Physicians will be seeing these patients every month. They can collect baseline data on their condition and closely track their outcomes over time. Researchers can then use this database to compare these patients to a comparable sample drawn from Medicare beneficiaries diagnosed as having early symptoms of AD who choose not to get the drug. Many won’t because of the side effects, which include brain bleeding and swelling in about 30% of patients.

These are called matched cohort studies. Regulators consider matched cohort studies inferior to randomized clinical trials (RCTs) for assessing the benefits and risks of drugs and other medical interventions. But they can generate valuable evidence from their use in real world settings, which in some ways is an improvement over RCTs, whose company sponsors often exclude complex patients, use endpoints like plaque-clearing that do not provide clear cut evidence of clinical benefit, and rarely follow patients after the end of the trial.

The registry will also be an valuable asset for evaluating the results of future studies of follow-on drugs in this new class of monoclonal antibodies. It will add significant new information for scientists seeking to further their understanding of the disease’s progression or identify other factors that might affect onset or progression.

The biggest complaint about using registries is that it imposes burdensome data collection requirements on the physicians and clinics administering the drugs. But we’re talking about a drug whose projected cost is minimally $10,000 a year, according to media reports. If only half of the two million people with early symptoms of AD take the drug, the potential revenue stream for Biogen will be $10 billion a year – or about $600 million a year to the prescribing clinicians.

Surely there’s adequate funding in that revenue stream to support a registry, especially given that data collection can be automated through the electronic health records that already exist thanks to generous federal financing over the past decade.