See a paper from Stanford with several FDA authors on how FDA makes decisions when the evidence is close. The paper screens drug applications from 2013 to 2018, finding 912, of which 22 faced multiple review cycles including a flat-out initial denial.
The authors conclude that "the FDA has no mechanism to find, or tradition to cite, similar cases when weighing evidence." They argue that FDA decision pathways and evidence standards were "highly variable."
The authors argue that a better tradition of "case studies" and systematic approaches would enhance the transparency and reliability of these decisions.
Dr. Janiaud is at the Meta-Research Innovation Center at Stanford; Dr. Goodman is professor of epidemiology there. Goodman and John Ioannidis co-led the MRIC. Dr. Russek-Cohen is retired FDA; Dr. Irony is former FDA, now with Janssen.
Although it's a 2021 decision, outside their scope of 2013-2018 decisions, the authors make several references to the 2021 approval of aducanumab (Aduhelm). There is also an op ed from Washington University's Joel Perlmutter, who resigned in June from the FDA neurology committee.