Seven-valent mKRAS-specific lymph node targeted amphiphile immunotherapy versus observation in resected pancreatic ductal adenocarcinoma (PDAC): The AMPLIFY-7P randomized phase 2 study

AMPLIFY-7P

Patient / Population Intervention / Exposure Comparison Outcome

In patients with resected KRAS-mutated pancreatic ductal adenocarcinoma, the mKRAS-specific vaccine camupepimut did not improve investigator-assessed disease-free survival compared with observation.

N

144patients

Design

Randomised, open-label, phase 2 trial allocating 2:1, stratified by lymph node status, patients with stage I-III KRAS-mutated pancreatic ductal adenocarcinoma who had received curative-intent therapy and had no recurrence at ≤ 6 months to camupepimut (ELI-002 7P) or observation (NCT05726864).

Endpoint

Investigator-assessed disease-free survival, with success requiring a hazard ratio ≤ 0.70 at a 2-sided alpha of about 0.10.

ResultInvestigator-assessed DFS with camupepimut vs observation HR 0.85 (P = 0.539); the primary endpoint was not met. Pre-specified 6-month DFS 75.7% vs 61.7%, an advantage of 14% (HR 0.55; P = 0.0557). mKRAS-specific T cell responses correlated with efficacy (median DFS not reached vs 6.08 months; HR 0.22; P < 0.0001). mKRAS-specific T cell responses occurred in 99% (89/90). Grade ≥ 3 AEs 10.4% vs 14.6%; serious AEs 6.3% vs 8.3%.

O'Reilly EM, George B, Singh H, et al. Presented at ESMO Congress 2026, abstract LBA83, 23 October. Not yet published.

Discussion & critique

The trial missed its primary endpoint: the DFS hazard ratio of 0.85 fell well short of the 0.70 the design required. The positive signals rest on weaker ground. The 6-month DFS difference is borderline and secondary, the T cell responder comparison is not randomised and may reflect patients who were fitter or had better prognosis, and the ctDNA-negative, no-radiation and post-hoc multivariate analyses are hypothesis-generating. Baseline ctDNA positivity was imbalanced against the vaccine arm (20.8% vs 12.5%), and the design was open label with investigator-assessed DFS. The abstract gives no confidence interval for the primary hazard ratio, no follow-up duration and no overall survival data.