New FDA Documents Shed Light on Biogen Approval
Just-posted FDA review documents show how CDER management reconciled using the accelerated approval pathway for Biogen’s Alzheimer’s drug Aduhelm (aducanumab) and dismissed reviewers’ recommendations for traditional approval or outright denial of the BLA submission. A memo from CDER Office of New Drugs director Peter Stein noted that the primary clinical review and an Office of Clinical Pharmacology (OCP) review both initially supported traditional approval, whereas the statistical review did not recommend approval because of the two Phase 3 studies (Study 301 and Study 302) submitted — a conflict between one positive study and one negative study.
Stein said that despite the “reasonable arguments” to support standard approval, he agreed with the Office of Neuroscience that the evidence is not “sufficiently compelling or persuasive to meet the substantial evidence standard for standard approval.” Stein subsequently concurred with the accelerated approval decision, and agreed that the “application contains substantial evidence of effectiveness on the surrogate endpoint of reduction in brain amyloid plaque and that the surrogate endpoint is reasonably likely to predict clinical benefit.”
Stein said his assessment to support the accelerated approval pathway was based on the following reasons: “First, both Phase 3 studies were stopped for futility and were revisited after analyses showed disparate effects in the two trials. Conclusions from studies stopped for futility must be viewed with some caution, given the limits on strong control of alpha from what are, in essence, post-hoc analyses... Second, even though the results of Study 302 appear to be generally believable and supportive of effectiveness of aducanumab with respect to clinical endpoints, ... we are still left with a negative Phase 3 study (Study 301), a similarly designed study that failed to substantiate the findings of Study 302. Independent substantiation is the gold-standard basis for generating confidence in any initial scientific study finding.”
Because aducanumab presented an unusual situation — “where there is residual uncertainty regarding the clinical benefit of aducanumab but there is substantial evidence of effectiveness on an endpoint that is reasonably likely to predict such benefit”— consideration for accelerated approval was warranted, mainly due to patients having a fatal disease and they are desperate for treatments, according to Stein’s memo. Issuing a “complete response” on the BLA and requiring another trial would delay patients’ access for several years, he said. FDA staff commonly hear from patients that they would opt to accept the uncertainty and risks for a drug that is likely, but not confirmed, to provide clinical benefit, potentially slowing the progression of their disease, he said.
It is interesting to note that part of Stein’s argument contradicted aducanumab’s approved indication. Justifying the accelerated approval urgency, Stein argued that there may be a “window” for benefit, and patients “progressing to later stages of the disease over the next several years may no longer be eligible for treatment, if it is confirmed to provide clinical benefit, since the drug will be indicated for patients relatively early in the clinical course.” However, according to the approved label, there is no such restrictive indication and the drug is approved for use in all Alzheimer’s patients.
Stein acknowledged that allowing accelerated approval (AA) for aducanumab veered from its more usual scenario. “When drugs are approved using AA, even though there is substantial evidence of effectiveness with respect to the surrogate endpoint at the time of approval, there is a recognition that there remains residual uncertainty of clinical benefit, typically because the studies assessing clinical benefit are ongoing, but here because the results of the studies assessing clinical benefit strongly suggested but did not establish benefit,” he wrote in the memo.
“In summary, with respect to the use of AA for aducanumab,” Stein continued, “the criteria set out in statute, regulations, and discussed in the above referenced guidance are met: the setting is appropriate (serious disease with unmet need), the drug has the potential to provide a meaningful advantage over available therapy, there is substantial evidence of effectiveness on the surrogate, and the surrogate is reasonably likely to predict clinical benefit. A key element in AA is that there remains residual uncertainty at the time of approval as to whether the drug will provide clinical benefit, specifically whether the reasonably likely surrogate accurately predicts this benefit. As a key expectation of AA, the clinical benefit must be confirmed post-approval by the applicant based on additional study.”