PI3K Briefing Document Reiterates FDA Concerns

Share

A just-posted advisory committee briefing document reiterates FDA concerns with phosphatidylinositol 3-kinase (PI3K) inhibitors and their approval for treating hematological malignancies, setting the stage for a 4/21 discussion by the agency’s Oncologic Drugs Advisory Committee. Most of FDA’s concerns were outlined in a recent online Lancet Oncology article by FDA Oncology Center of Excellence director Richard Pazdur and other FDA officials, who dismissed (see earlier story) the “current paradigm of using overall response rates in single-arm trials to support accelerated approvals.”

 

The briefing document criticizes current PI3K inhibitor development programs that did not perform adequate dose exploration which in some cases was further complicated by selecting a single-arm study design for initial registration. “The safety analysis of data from single-arm trials,” it says, “is confounded by the absence of a control, which can pose challenges in accurately attributing AEs [adverse events] to either drug or the underlying disease. Because sponsors may select a dose to achieve the highest ORR [objective response rate] without carefully considering a benefit-risk analysis, toxicities are compounded and may contribute to the worrisome OS [overall survival] results seen repeatedly in multiple RCTs [randomized control trials] of these agents.”

 

Additionally, the briefing document endorses randomized trials as the “most effective method to control for confounding factors” and because they allow for assessing progression-free survival, overall survival (OS), and patient reported outcomes. It says that “OS data should be fully collected and analyzed, as OS is an important metric in the benefit-risk determination, especially for products with significant toxicity. Randomized trials should be prospectively planned to allow for an adequate assessment of survival and other measurements of clinical benefit. Future studies should be designed with additional opportunities to evaluate patient safety, andthere should be a high likelihood that when a study is completed it can indicate an acceptablebenefit-risk profile, which includes evaluation of the impact on survival as both a safety and efficacy metric.”

Read more