ESMO Congress 2026

In pancreatic ductal adenocarcinoma (PDAC), KRAS is the most solid target rationale: an integrated multi-platform analysis of 150 PDAC specimens showed KRAS as the most recurrent somatic mutation, and a subset of tumors even harbored multiple KRAS mutations ^[5], providing direct target logic for a multivalent mKRAS vaccine. On the immune front, PDAC stromal biology offers both rationale and constraint: in mice, depleting myofibroblasts paradoxically worsened immunosuppression and accelerated tumors, which anti-CTLA4 therapy could reverse ^[2]; single-cell work defined antigen-presenting CAFs expressing MHC II that activate CD4+ T cells in an antigen-specific fashion ^[3], indicating a workable interface for T cell-directed strategies in stroma-rich PDAC. With tumor- and stroma-specific molecular subtypes each carrying independent prognostic value ^[4], any signal from this class of therapy is more likely to be delivered in stratified populations.

At this ESMO meeting (#ESMO26), the PDAC readout was the AMPLIFY-7P randomized phase 2: camupepimut is a 7-valent mKRAS, lymph-node-targeted amphiphile vaccine, enrolled in patients with curative intent and no recurrence <6 months; T cell responses served as the primary endpoint and were negative; the signal presented was a DFS benefit of 14% at 6 months (p=0.056), with a phase 3 planned; the live post also flagged, with a question mark, that benefit may be concentrated in the ctDNA-negative and no-radiation subgroup, which is to be confirmed. A second rapid oral in the same session addressed adding SBRT to chemotherapy (mFOLFIRINOX ~90%, GemNab ~10%) in a mixed population of resectable, borderline resectable and locally advanced disease; the live post judged that better local control and any OS signal remain unclear.

In terms of forward movement, AMPLIFY-7P advances the mKRAS vaccine from target and immunobiological rationale ^[5][3] to a borderline clinical signal: despite a negative primary endpoint, the 14% DFS difference at 6 months and the planned phase 3 keep the program alive; if the ctDNA-negative subgroup benefit is confirmed, it would align with the established view that PDAC therapy should be stratified by tumor/stromal subtype ^[4]. The SBRT report echoes PDAC's systemic challenge — exosomal MIF from PDAC can pre-build a liver pre-metastatic niche ^[1] — so against this highly metastatic background, the value of local intensification alone is indeed hard to settle, consistent with the 'still unclear' view from the floor.

The limitations are equally clear: the primary endpoint was not met, p=0.056 does not cross conventional significance thresholds, and the subgroup information is exploratory and question-marked in the live coverage; sample size, follow-up duration and safety data for both reports were not disclosed, and conclusions should await the phase 3 and formal publication.