Replimune's Third BLA: The Writing Playbook When the Agency Stands By Its Rejections

A Different Shape of Repeat-Submission Story

Last week we wrote about Outlook Therapeutics' fourth FDA cycle — three CRLs across manufacturing, inspection, and substantial-evidence findings, then a granted appeal in May 2026 that concluded substantial evidence of effectiveness had been established for LYTENAVA in wet AMD. The Outlook case is a repeat-submission program where the agency's position has demonstrably shifted.

The Replimune case is the other shape of the same problem.

On May 29, 2026, Replimune announced its third BLA resubmission for RP1 in combination with nivolumab for advanced melanoma, following a "productive discussion" with the FDA. The agency has agreed to prioritise the review as an urgent matter. But — and this is the central fact of the story — the FDA has not retracted its prior substantial-evidence finding. Coverage in OncLive describes the agency's position from the April 2026 CRL in plain language: the IGNYTE trial is "not considered to be an adequate and well-controlled clinical investigation that provides substantial evidence of effectiveness." The two CRLs — July 2025 and April 2026 — landed on the same axis.

For a writing organisation preparing the third cycle, the difference between Replimune's situation and Outlook's matters more than the surface similarity of "third re-submission." Both are repeat submissions. One has a granted appeal that recontextualises the prior CRLs. The other has an agency that has stood by its position twice.

What the FDA Has Said About IGNYTE

The agency's objection has two specific surfaces, and both are public.

First, the single-arm IGNYTE Phase 1/2 findings in the post-PD-1 melanoma population were deemed insufficient as the primary efficacy basis for approval. Single-arm evidence in a setting with active comparators available raises a recurring set of regulatory questions: comparator context, historical-control selection, response durability, and whether the population studied maps cleanly to the labelled population.

Second, the randomised IGNYTE-3 data comparing RP1 plus nivolumab against physician's choice after PD-1 and CTLA-4 therapy were also deemed insufficient to support approval. The agency's concern here is methodological — whether the trial as designed, with the comparator chosen and the population enrolled, constituted an "adequate and well-controlled clinical investigation" within the meaning of the substantial-evidence standard.

The third submission is being framed as supported by additional IGNYTE data and by FDA alignment on the path forward. The agency's willingness to prioritise the review is operational good news. But the underlying substantial-evidence question is the same one the agency has answered twice in the negative.

What a Third-Cycle Writing Team Has to Do When the Agency Has Not Moved

A third-cycle BLA submitted into an agency posture that has not changed is a different writing problem than one submitted into a reversed posture.

The substantial-evidence narrative cannot rest on volume or repetition. The temptation in a third cycle is to submit the same evidence with more emphasis — more tables, more analyses, more sub-population breakdowns. That is the wrong move when the agency has been clear that the issue is whether the trial as designed supports the claim, not whether the data has been adequately analysed. The Module 2.5 clinical overview has to make a new argument, not a louder version of the previous one. New evidence framing, new contextualisation against the post-PD-1 and post-CTLA-4 standard of care, and an explicit response to each prior FDA objection — anchored in what is in the dossier, not what the writing team would prefer the reviewer to read.

The IGNYTE-3 comparator argument needs explicit reconstruction. Physician's choice as a comparator in heavily pre-treated melanoma is defensible, but the defence has to be argued, not assumed. The third-cycle dossier should walk the reviewer through why this comparator was chosen, what alternatives were considered, what historical-control data exist for context, and how the observed treatment effect should be read against the comparator's expected performance. If the second CRL said the data did not support approval, the third submission's job is to make the data support the claim more visibly — and that is a writing problem, not a statistical one.

The CRL-to-resolution map is more important here than in Outlook's case. When the agency's position has shifted, the briefing document maps cycles 1 → 2 → 3 against a moving target. When the agency's position has held, the briefing document maps the same agency objection against three submission packages and has to show, line by line, why this package answers what the prior two did not. The reviewer is being asked to read a third response to the same question. The cost of forcing them to reconstruct the regulatory history themselves is now higher, not lower.

Provenance discipline is non-negotiable. Every substantial-evidence claim in the third-cycle Module 2.5 should be linked to its source — IGNYTE statistical output, IGNYTE-3 randomised analysis, prior FDA correspondence, the Type B meeting summary that produced the current submission alignment. Unsupported summary language in a third cycle reads, to a reviewer who has already issued two CRLs on the same axis, like the sponsor has not understood the objection.

Why This Shape of Case Matters

The Outlook and Replimune cases bookend the public set of repeat-submission scenarios a sponsor writing organisation can find itself in.

Outlook went through three CRLs and then a granted appeal that reframed the substantial-evidence question. The fourth-cycle writing problem there is about making the appeal outcome, the prior CRLs, and the new evidence read as one coherent argument inside a 60-day Class 1 review.

Replimune is going into a third cycle with the agency standing by its prior substantial-evidence findings, supported by FDA alignment on the path forward but not by a reversal of the underlying methodological objection. The third-cycle writing problem here is about making the substantial-evidence argument differently than it has been made twice before — and doing it inside a review the agency has agreed to expedite, which compresses the window even as the rhetorical burden increases.

Both are writing-organisation problems, and the answer in both cases is the same operating model: source-linked claims, CRL-to-resolution mapping, briefing documents that pre-answer the reviewer's reconstruction work, and a Module 2.5 that gets rebuilt rather than carried forward. We covered the architectural shape of this in an earlier piece on Atara's third shot with Pierre Fabre and in the Outlook post above.

What Replimune adds to the public case library is the version of the problem where the writing team has to make the case under exactly the conditions the second CRL described as insufficient. That is the harder version of the third-cycle writing problem — and the version most repeat-submission programs will recognise more easily than the appeal-reversal shape.

The agency's objection was specific. The third cycle's writing has to be specific back.