FDA Rejects Replimune’s RP1 Despite 34% Response Rate in Melanoma

01 — The Setup: A Drug That Made Clinical Sense

RP1, also known as vusolimogene oderparepvec, is Replimune’s lead oncolytic immunotherapy. It is based on an engineered HSV-1 platform designed to kill tumor cells locally and stimulate systemic anti-tumor immune activity.

The treatment was studied in combination with nivolumab in patients with advanced melanoma who had already progressed on anti-PD-1 therapy.

That is not a soft population. Patients with unresectable or metastatic melanoma after checkpoint inhibitor failure remain a high-unmet-need group. Many have exhausted the most important immunotherapy backbone available to them. For this kind of setting, accelerated approval logic can appear attractive: if a drug produces a durable response in a population with limited options, the pathway can theoretically move faster while confirmatory evidence matures.

RP1 looked, on paper, like the kind of biologic that could fit that logic.

It had a plausible mechanism. It had Breakthrough Therapy designation. It had a biologically coherent combination partner. It had a clinical signal in a post-PD-1 melanoma population.

But the problem was not the story. The problem was the structure of proof.

02 — The Data That Made the Rejection So Controversial

In the IGNYTE trial, RP1 plus nivolumab produced an objective response rate of approximately 34% in patients with confirmed progression on prior anti-PD-1 therapy. By RECIST 1.1, the response rate was reported at approximately 32.9%, with a complete response rate around 15%.

The median duration of response was reported at 24.8 months in the April 2026 CRL-related company disclosure.

For a post-PD-1 melanoma population, those numbers are not cosmetic. A one-in-three response rate with multi-year durability is clinically hard to ignore. Replimune also emphasized a favorable safety profile, and the regulatory debate did not center on a newly discovered toxicity problem.

The controversy is therefore simple: The drug produced a response profile that looked medically meaningful, but the FDA concluded that the trial could not reliably prove how much of that effect came from RP1 itself.

That is the heart of the case.

03 — Why FDA Said No

The FDA’s objection was not simply “the response rate is too low.” The objection was deeper and more structural.

The agency concluded that the IGNYTE dataset was not an adequate and well-controlled clinical investigation capable of providing substantial evidence of effectiveness. The main issues were:

The most important phrase in the RP1 case is contribution of components.

Because RP1 was given with nivolumab, the FDA wanted evidence showing that RP1 itself contributed to the observed clinical effect. In a single-arm trial, that question becomes difficult. Without a randomized comparator, the agency could not clearly determine whether the observed responses were due to RP1, nivolumab rechallenge activity, patient selection, response-assessment variability, or some combination of these factors.

For a sponsor, that is a brutal regulatory trap. A regimen can work as a whole. Patients can respond. The safety profile can look acceptable. The unmet need can be real. But if the agency cannot isolate the contribution of the new component, the application can still fail.

04 — The Point Was Not “No Signal.” The Point Was “Wrong Proof.”

This is where RP1 becomes more important than a normal FDA rejection.

The FDA’s position was not that RP1 had no biological activity. The agency’s position was that the submitted evidence did not meet the required legal and regulatory threshold for approval.

That is a very different message.

For developers, it means a clinically impressive single-arm signal can still collapse if the design does not answer the regulator’s specific causality question. In oncology, that is increasingly important because many new drugs are developed in combinations. The more combination-heavy the field becomes, the more dangerous the contribution-of-components problem becomes.

The sponsor sees a patient responding. The physician sees disease control. The investor sees an approvable unmet-need asset. The regulator asks: How do we know the new drug caused the effect?

If the trial cannot answer that cleanly, the entire approval case becomes vulnerable.

05 — The Regulatory Whiplash

RP1 is not a clean “FDA changed its mind overnight” story. The FDA’s April 2026 CRL makes clear that the agency had communicated concerns about single-arm design, patient heterogeneity, response interpretation, and contribution of components during earlier interactions.

That point matters because the article should not reduce the case to a simple accusation that the FDA invented new objections at the last minute.

The sharper point is more precise: FDA non-objection to a filing strategy is not the same as FDA agreement that the same strategy will satisfy approval standards.

That is the real business lesson.

Replimune was able to file. The BLA was reviewed. The agency engaged. But when it came time to decide whether the evidence met the standard for approval, the FDA concluded that the same evidence package was insufficient.

For biotech executives, this is the dangerous middle zone of regulatory development. A sponsor can interpret FDA engagement as practical alignment, while the agency later treats that engagement as procedural flexibility rather than substantive endorsement.

That distinction can decide the fate of a company.

06 — Then the Door Reopened

The RP1 story did not end with the April 2026 rejection.

After further FDA discussions, Replimune announced that it had aligned with the agency on a path forward for BLA resubmission and reconsideration. The company said it would resubmit the RP1 BLA, and that the FDA had indicated the resubmission would be treated as an urgent matter and prioritized for review.

That changed the market narrative immediately.

The same asset that had looked damaged by a second CRL suddenly had a third regulatory path.

This is why RP1 is such a valuable regulatory case study. It shows two risks at once.

First, single-arm evidence can fail even with a meaningful response rate.

Second, regulatory direction can change quickly when unmet need, public pressure, leadership turnover, and internal agency philosophy collide.

For investors, that makes the binary event harder to price. The data did not dramatically change between rejection and renewed resubmission pathway. The regulatory posture did (see also our analysis of China Biotech licensing risks).

That is not a small detail. That is the story.

07 — Why RP1 Fits the Larger Pattern

RP1 belongs in a broader pattern of FDA decisions where the central issue is not weak science alone, but the regulator’s interpretation of risk.

RP1 is the evidentiary-risk version of this pattern.

The agency did not need to say the drug was useless. It only needed to say the evidence did not prove enough.

That is a much more dangerous type of rejection because it can appear late, even after years of development confidence.

08 — The Investor Read

For investors, RP1 carries three direct lessons.

First, single-arm accelerated approval strategies now deserve a higher discount rate, especially when the program is a biologic, the population is heterogeneous, and the drug is used in combination with an active agent.

Second, Breakthrough Therapy designation should not be overread. It can signal urgency and regulatory interest, but it does not immunize a program from evidence-standard objections.

Third, FDA alignment is a soft asset, not a hard asset. Meeting minutes, filing acceptance, and agency engagement matter, but they do not guarantee approval. The final review can still turn on whether the agency believes the dataset satisfies substantial evidence requirements.

RP1 also shows that regulatory volatility can create asymmetric market moves. A second CRL can crush the asset, while a resubmission pathway can rapidly reprice it. The underlying clinical signal may remain broadly similar, but the perceived regulatory probability changes dramatically.

That is not normal clinical-risk pricing. That is regulatory-state pricing.

09 — The Developer Read

For developers, RP1 should change trial-design behavior.

If a new oncology drug is being developed in combination, the sponsor should assume the FDA will ask whether the new component’s contribution can be isolated.

That means future programs need to think earlier about:

• randomized evidence;
• control-arm feasibility;
• contribution-of-components strategy;
• patient-population homogeneity;
• response-assessment robustness;
• pre-specified analyses;
• external-control limitations;
• whether accelerated approval can survive without a clean comparator.

The most dangerous mistake is to build a beautiful single-arm dataset and only later discover that the agency does not believe the design can answer the decisive question.

RP1 is a warning that the trial architecture must be regulatory-grade from the beginning (as we discussed in the Pharma Tariff War analysis on policy and regulatory shifts).