── Diligence workbench ──

divarasib· Hoffmann-La Roche

Second-generation competitor example. Divarasib (Roche) is a covalent second-gen KRAS G12C inhibitor in a Phase 3 head-to-head (NCT06497556) vs sotorasib/adagrasib. A clean test of the framework on a well-capitalized sponsor in a competitive, target-validated field.

Asset ▸phase 3 · oncology · small molecule · well capitalized · 0 designations · 3 competitors[ edit ]

Asset name

Sponsor

Phase

Therapeutic area

Modality

Mechanism

Supply constraint (manufacturing / raw material)

Target validated by prior approval

Biomarker-enriched trial

Direct competitors in this indication: 3

Regulatory designations

── Risk & limits ─────────────────────────

Q: what could break this thesis?

Known limitations

Rare disease (small N)
BIO/Informa rare-disease base rate (~16% from P1) is built on a small population of approvals. Cumulative LOA estimates are noisier for indications with <20 historical approvals.
Platform companies
Single-asset LOA doesn't capture portfolio effects. A platform with 6 correlated programs has a different risk profile than 6 independent shots-on-goal. v2 portfolio view addresses this.
China-origin assets
BIO base rates reflect FDA/EMA cohorts. NMPA development paths differ; cross-jurisdictional risk profiles are not yet modeled. Treat outputs as US/EU-only.
Pre-2018 capital-position data
Reflexivity multipliers are calibrated to post-2018 trial-design and biomarker-enrichment norms. Older comparables may understate the gap between well-capitalized and constrained sponsors.

Game-Theory Adversary · agent

Waiting for PoS / rNPV / scorecard / comparables…