Cutaneous melanoma

individualized mRNA neoantigen + checkpoint

Thesis

The Phase 3 melanoma win is [KNOWN]. The increment is whether the same individualized neoantigen + PD-1 logic transfers to other high-TMB, high-neoantigen tumors (MSI-H CRC, smoking-related NSCLC) rather than being a melanoma-only accident.

Mechanism

[KNOWN] Tumor-specific neoantigens presented on MHC-I; mRNA-LNPs prime CD8 T cells; PD-1 blockade prevents exhaustion. [INFERRED] Transfer depends on neoantigen quality and an inflamed TME, not on melanocyte lineage.

Kill experiment

A 40-patient window-of-opportunity study in resected MSI-H CRC: if circulating neoantigen-specific CD8s do not expand ≥5× after two doses, the transfer hypothesis dies.

pan-RAF + MEK + checkpoint

Thesis

NRAS melanoma still has no approved targeted backbone. Pan-RAF + MEK has a signal; adding PD-1 is the increment, not the backbone itself.

Mechanism

[KG] NRAS is a top associated target. [KNOWN] Naporafenib + trametinib showed activity in NRAS-mutant melanoma. [INFERRED] MAPK suppression may inflame the TME enough for PD-1 to matter.

Kill experiment

A 20-patient NRAS-mutant run-in: if no RECIST response and no rise in intratumoral CD8 within 6 weeks, abandon the triplet.

TCR-T cell therapy

Thesis

PRAME is frequently expressed in cutaneous melanoma; HLA-A*02:01 TCR-T is a real modality. Place it after checkpoint failure, not as a first-line fantasy.

Mechanism

[KNOWN] PRAME is a cancer-testis antigen expressed in melanoma. TCR-T against PRAME has entered the clinic. [INFERRED] Residual disease after PD-1 is the setting where a single antigen hit can still be scored.

Kill experiment

If <30% of screened post-PD-1 tumors stain PRAME-high and HLA-A*02:01, the addressable set is too small and the candidate dies on epidemiology, not biology.

dual checkpoint + targeted

Thesis

Triplet already explored; the honest play is a defined BRAF V600 first-line triplet with a stopping rule, not a new drug.

Mechanism

[KNOWN] BRAF V600 drives MAPK. Dual BRAF/MEK shrinks tumor and can increase antigen presentation. LAG-3 + PD-1 is approved in melanoma. [INFERRED] Sequencing and toxicity, not novelty, are the remaining questions.

Why it died

Combinations of BRAF/MEK + PD-1 already produced excess toxicity and mixed OS in prior Phase 3 (COMBI-i and relatives). Relatlimab does not obviously fix the fever/LFTs problem. Already failed the cheap version of this experiment.

antiparasitic repurposing

Thesis

Social-media claim that ivermectin kills melanoma.

Mechanism

[SPECULATIVE] In vitro cytotoxicity at concentrations not achievable in patients.

Why it died

Not aimed at this disease in any serious sense. No causal chain from a registered melanoma driver to a human-achievable exposure. Invented-as-cure. Permanent kill.

oncolytic HSV + PD-1

Thesis

Oncolytic virus plus PD-1 should convert cold lesions.

Mechanism

[KNOWN] T-VEC is approved. MASTERKEY-265 (T-VEC + pembrolizumab) already ran.

Why it died

Phase 3 MASTERKEY-265 failed its primary endpoint. Repeating a completed negative Phase 3 is not a hypothesis.

small molecule

Thesis

CDKN2A loss should create CDK4/6 dependence.

Mechanism

[KG] CDKN2A is associated. [INFERRED] Palbociclib/abemaciclib should freeze G1.

Why it died

CDKN2A loss is common and necessary but not sufficient. Melanoma bypasses CDK4/6 via MAPK and MITF. Prior CDK4/6 melanoma studies were weak. Causal chain stops at ‘the gene is deleted’.

undisclosed peptide

Thesis

A new proprietary peptide that ‘covers all neoantigens’.

Mechanism

[SPECULATIVE] No public structure, no MHC data.

Why it died

Invented drug. No INN, no structure, no trial. The name is a tell. Permanent kill.