Pancreatic ductal adenocarcinoma

Combination cytotoxic chemotherapy

Thesis

[KNOWN] Modified FOLFIRINOX is an established standard systemic treatment option for fit patients with advanced or resectable PDAC. [INFERRED] Its value as the control backbone should be preserved while testing biomarker-selected additions rather than replacing it without evidence. Falsifier: [KNOWN] A randomized comparison fails to improve progression-free or overall survival over the appropriate contemporary control.

Mechanism

[KNOWN] Oxaliplatin causes DNA crosslinks, irinotecan inhibits topoisomerase I, fluorouracil inhibits thymidylate synthase and incorporates into RNA, and leucovorin enhances fluorouracil activity.

Kill experiment

In a randomized, adequately powered PDAC trial using the appropriate contemporary control, demonstrate no clinically meaningful overall-survival or progression-free-survival advantage for the specified modified FOLFIRINOX regimen over the comparator.

Combination cytotoxic chemotherapy

Thesis

[KNOWN] Gemcitabine plus nab-paclitaxel is an established treatment option for advanced PDAC. [INFERRED] It remains a pragmatic comparator or combination backbone for patients who are not candidates for FOLFIRINOX. Falsifier: [KNOWN] A prospectively defined combination produces no clinically meaningful benefit over gemcitabine plus nab-paclitaxel.

Mechanism

[KNOWN] Gemcitabine is a nucleoside analogue that disrupts DNA synthesis, while nab-paclitaxel stabilizes microtubules and prevents their normal disassembly.

Kill experiment

In the intended patient population, a randomized contemporary comparison showing materially inferior survival, response, or tolerability without a compensating clinical benefit would disprove its proposed pragmatic backbone role.

Second-line cytotoxic combination

Thesis

[KNOWN] Liposomal irinotecan with fluorouracil and leucovorin is an established post-gemcitabine treatment option for metastatic PDAC. [INFERRED] Improving patient selection or combining this backbone with a rational immune or DNA-damage strategy is more defensible than treating it as an unselected universal solution. Falsifier: [KNOWN] A biomarker-selected study shows no enrichment of response or survival.

Mechanism

[KNOWN] Irinotecan and its active metabolite inhibit topoisomerase I, producing replication-associated DNA damage; fluorouracil inhibits thymidylate synthase and disrupts RNA and DNA metabolism.

Kill experiment

In a prospectively biomarker-defined post-gemcitabine PDAC cohort, a small randomized or signal-seeking study showing no objective responses, no disease-control enrichment, and no pharmacodynamic biomarker association would falsify the proposed selection rationale.

KRAS G12D-selective small-molecule inhibitor

Thesis

[KNOWN] KRAS G12D is a frequent oncogenic alteration in PDAC. [INFERRED] MRTX1133 could provide a genotype-matched treatment increment for KRAS G12D-positive tumors, especially when used with a chemotherapy or pathway-combination strategy. [SPECULATIVE] Activity may be limited by adaptive signaling and intratumoral heterogeneity. Falsifier: [KNOWN] In KRAS G12D-positive PDAC, target engagement does not translate into objective responses or durable disease control.

Mechanism

[KNOWN] MRTX1133 is an investigational inhibitor designed to selectively inhibit the KRAS G12D mutant protein and reduce downstream MAPK signaling.

Kill experiment

A small KRAS G12D-positive PDAC window-of-opportunity study with pre- and post-treatment tumor biopsies showing adequate drug exposure and MAPK pathway suppression but no objective response or meaningful tumor-cell loss would kill the translational thesis.

RAS(ON) state-selective small-molecule inhibitor

Thesis

[KNOWN] RMC-6236 is an investigational RAS(ON) inhibitor being evaluated in RAS-mutant solid tumors. [INFERRED] Its broader RAS-mutant coverage could be useful in PDAC where KRAS alterations are common and the precise allele may vary. [SPECULATIVE] Combination with chemotherapy or another pathway inhibitor may suppress resistance better than monotherapy. Falsifier: [KNOWN] Molecularly selected PDAC shows insufficient response durability despite pharmacodynamic RAS pathway suppression.

Mechanism

[KNOWN] RMC-6236 binds active RAS proteins and is intended to inhibit signaling from multiple oncogenic RAS variants, including common KRAS mutants.

Kill experiment

In molecularly selected PDAC, obtain paired biopsies after confirmed adequate exposure; if RAS-pathway pharmacodynamic suppression occurs without tumor regression or even short-term disease control in a prespecified signal-seeking cohort, the proposed PDAC value is falsified.

Personalized uridine RNA neoantigen vaccine

Thesis

[KNOWN] A 2023 report described personalized RNA neoantigen vaccination stimulating T-cell responses in pancreatic cancer. [INFERRED] BNT122 could be tested after tumor resection and systemic therapy to amplify or maintain immunity against patient-specific neoantigens. [SPECULATIVE] Vaccination may be insufficient without improved antigen presentation or relief of the immunosuppressive tumor microenvironment. Falsifier: [KNOWN] Vaccine-induced neoantigen-specific T cells do not associate with measurable minimal residual disease or relapse reduction.

Mechanism

[KNOWN] The vaccine encodes patient-specific predicted neoantigens so host cells can produce antigen and stimulate adaptive T-cell responses.

Kill experiment

In a randomized post-resection, post-systemic-therapy study with adequate vaccine manufacturing and immune monitoring, BNT122 produces neoantigen-specific T cells but no improvement in molecular residual disease clearance, relapse-free survival, or overall survival over the contemporary control.

Why it died

The pack supports only that a 2023 report described personalized RNA neoantigen vaccination stimulating T cells; it does not establish that BNT122 specifically reduces PDAC relapse or measurable residual disease. The thesis jumps from immune activation to clinical benefit, and its proposed post-resection use faces tumor heterogeneity, neoantigen-prediction error, and immunosuppressive stroma. The stated falsifier is also too weak: an association between T cells and relapse is not a valid efficacy test.

PD-1 and CTLA-4 immune checkpoint blockade

Thesis

[KNOWN] PDAC generally has a poorly inflamed, immunosuppressive microenvironment and unselected checkpoint blockade has limited activity. [SPECULATIVE] Combining nivolumab and ipilimumab with a vaccine, radiotherapy, or chemotherapy-induced antigen release could create a responder subset. [INFERRED] This should be restricted to a biomarker-defined exploratory cohort rather than presented as a general PDAC treatment. Falsifier: [KNOWN] Combination treatment fails to produce immune infiltration and objective responses in the prespecified biomarker-positive group.

Mechanism

[KNOWN] Nivolumab blocks PD-1 signaling and ipilimumab blocks CTLA-4 signaling, reducing inhibitory signals on T cells.

Kill experiment

In the prespecified biomarker-positive PDAC cohort, the combination fails to produce the required immune infiltration and objective response signal, with no clinically meaningful durability.

Why it died

The checkpoint mechanism is real, but the PDAC proposal does not identify a causal biomarker or complete the chain from tumor mutation to antigen presentation, T-cell entry, tumor recognition, and killing. Blocking PD-1 and CTLA-4 can release inhibitory signals only on an existing immune response; it does not by itself overcome the poorly inflamed, immunosuppressive PDAC microenvironment. The phrase "biomarker-positive" is undefined in the pack, so the claimed responder subset cannot be selected or falsified as stated.

Hyaluronan-depleting stromal modifier

Thesis

[KNOWN] Hyaluronan-rich stroma can contribute to abnormal tumor pressure and impaired drug or immune-cell access in PDAC. [KNOWN] Clinical development of pegvorhyaluronidase alfa did not establish a broadly successful treatment strategy. [INFERRED] It remains a wrong-but-plausible decoy for testing whether stromal depletion is selected by a validated hyaluronan-high biomarker rather than assumed to benefit all PDAC. Falsifier: [KNOWN] Biomarker-selected tumors show no improvement in drug delivery, immune infiltration, or clinical outcome.

Mechanism

[KNOWN] Pegvorhyaluronidase alfa enzymatically degrades hyaluronan in the extracellular matrix, theoretically reducing stromal pressure and tissue compactness.

Why it died

This is a real PDAC-directed agent, but it is not an unreviewed hypothesis. The pack itself states that clinical development failed to establish a broadly successful strategy, and the completed HALO 301 Phase 3 study in previously untreated metastatic hyaluronan-high PDAC failed its overall-survival endpoint. The proposed biomarker rescue is not supported by the evidence supplied and cannot be treated as a surviving triage candidate without a new, concrete basis.