FDA sends a warning about perioperative cancer studies

AstraZeneca’s US filing for Imfinzi in neoadjuvant non-small cell lung cancer, due a mid-2024 decision, was delayed by the need to hold an advisory committee meeting. The concerns that prompted this 25 July adcom relate not only to Astra’s supporting Aegean trial, it’s now become clear, but also to other studies that similarly combine neoadjuvant and adjuvant stages.

Briefing documents revealed on Tuesday suggest that the FDA is losing patience with trials that include both pre and post-surgery settings – so-called periadjuvant – and don’t also split out the separate benefits of each. Beyond Astra this has immediate relevance for Bristol Myers Squibb’s Opdivo filing in a similar use, though Merck & Co managed to slip through the net last October.

Merck’s Keytruda was the first anti-PD-(L)1 drug approved for combined neoadjuvant followed by adjuvant use, specifically in triple-negative breast cancer in mid-2021. The adcom’s immediate focus, however, is NSCLC, where the Keynote-671 study backed last October’s approval of Keytruda plus chemo as neoadjuvant, followed by Keytruda monotherapy as adjuvant treatment.

While formally Astra has filed Imfinzi for neoadjuvant NSCLC, its supporting Aegean trial had a similar design to Keynote-671: Imfinzi plus chemo before surgery, and Imfinzi monotherapy after. Both trials had a simple two-arm design, with a straight comparison against chemo alone.

Enough already

Now the FDA seems to be saying enough is enough. The adcom briefing documents spell out the problem: Aegean made it impossible to assess how Imfinzi’s separate effects on the neoadjuvant and adjuvant stages contributed to the overall result. 

Doubts are raised about the need for checkpoint blockade in both phases, given concerns about overtreatment and resulting toxicities. Way back in November 2018, the documents reveal, the agency had suggested that Astra design a study including within-trial evaluation of contribution of phase, but “against this advice” the group started Aegean, a simple two-arm study. 

It’s a separate question, however, why the FDA apparently raised no such concerns when approving the Keytruda regimens as recently as last October. 

One clue, however, might be that, unlike Keytruda, Imfinzi isn’t approved in any other neo/adjuvant settings. Keytruda monotherapy got US approval for adjuvant NSCLC in January 2023, so its use in the perioperative Keynote-671 trial can be seen as just the addition of a chemo combo neoadjuvant stage to an already approved adjuvant setting.

And, if the FDA had doubts about Imfinzi’s contribution to the two stages, these were proved to be well founded when the Astra drug flunked no fewer than three trials testing it in adjuvant NSCLC: Mermaid-1 and 2, and most recently the academic-sponsored Adjuvant BR.31 study.

The adcom will discuss two separate issues: firstly whether Aegean supports the perioperative indication for Imfinzi, and as such whether the company needs to conduct an additional trial; and secondly whether future trials of any perioperative regimen should have a design that assesses contribution of treatment phase.

An “optimal” idea mooted in the briefing documents comprises a four-arm study, with two active and two control cohorts before surgery, followed by continued active treatment (representing the perioperative setting), continued control (standard of care), a switch from active to control (neoadjuvant) and a switch from control to active (adjuvant). Depending on powering assumptions this would necessitate between 960 and 2,420 patients.

And Opdivo?

If this is now a problem for Astra, consider Bristol, whose Opdivo regimen faces an 8 October PDUFA date for periadjuvant NSCLC, based on Checkmate-77T, a simple two-cohort study.

The twist here is that Opdivo plus chemo already carries a US label for neoadjuvant NSCLC, so it can be argued that the FDA can already see – albeit on a cross-trial basis – whether continuing with Opdivo monotherapy in the adjuvant stage adds a meaningful benefit.

Given Imfinzi’s earlier failures the same can’t be said for the Astra drug. The FDA could of course use Aegean to approve Imzinfi plus chemo solely for neoadjuvant NSCLC, but ignoring the fact that Aegean also gave Imfinzi after resection could be a leap too far.

Selected phase 3 trials in perioperative NSCLC

Source: OncologyPipeline.