DD Signal Report
12 RAS-axis tickers cross-referenced against primary sources. Three externally verified catches: LLY dual-tier ASCO sponsorship, NVS opnurasib registry still ACTIVE_NOT_RECRUITING 22 months after Novartis discontinued the program, AMGN sotorasib sales narrative. Eleven self-audit corrections — including a 39 percent understatement caught in bigbio's own LLY R&D ledger ($7.9B captured → $13.0B SEC XBRL primary). Pre-publication QC works. Revolution Medicines' daraxonrasib thesis holds across two stress-test cycles.
Three external catches plus eleven self-audit corrections against bigbio's own pre-publication ledger. The $5.1B LLY R&D delta is internal-regression discipline, not a Street miss. RVMD Plenary 5/31 (LBA5) is where the daraxonrasib thesis pays — held intact across two stress-test cycles.
The audit track record
Three externally verified catches and eleven self-audit corrections — all pre-published. External catches are gaps in how the Street reads the registry; self-audit corrections are internal QC discipline caught before ship. Both shown; neither marketed as the other.
LLY cash and equivalents $5,282M as of 2026-03-31 (Q1 2026 10-Q).
“us-gaap:CashAndCashEquivalentsAtCarryingValue = 5,282,000,000 USD (end=2026-03-31, fy=2026, fp=Q1)”
Implication Pre-flight fix to numeric_audit.md before Cut B. Trivial $-amount delta, material for an audit-grade report.
LLY R&D expense TTM $13,045M (sum of last four quarters: $3,510M Q1'26 + $3,466M Q3'25 + $3,336M Q2'25 + $2,734M Q1'25).
“us-gaap:ResearchAndDevelopmentExpenseExcludingAcquiredInProcessCost — quarters: 2026-03-31: 3,510,000,000; 2025-09-30: 3,465,700,000; 2025-06-30: 3,336,100,000; 2025-03-31: 2,733,700,000”
Implication Material correction needed before Cut B ship. LLY R&D is ~$13B/yr, not ~$8B. Honest framing: pre-publication QC caught an internal aggregation regression against bigbio's own SEC XBRL primary capture.
FDA granted Breakthrough Therapy Designation to olomorasib + pembrolizumab on 2025-09-04 for 1L unresectable advanced or metastatic NSCLC with KRAS G12C mutation and PD-L1 >=50%.
“On September 4, 2025, Eli Lilly and Company announced that the U.S. Food and Drug Administration (FDA) has granted Breakthrough Therapy Designation to olomorasib (LY3537982) in combination with pembrolizumab (KEYTRUDA) as a potential first-line treatment for patients with unresectable or metastatic non-small cell lung cancer (NSCLC) harboring a KRAS G12C mutation and PD-L1 expression >=50%.”
Implication Date precision matters for an audit-grade report. A 7-month discrepancy in BTD timing distorts catalyst-window framing (BTD is ~20 months old, not 6 weeks old). Correct framing: BTD validated > 8 months ago; the 2026-05/06 catalyst is the ASCO 2026 abstract, NOT a fresh BTD.
CONTRADICTED by primary source: canonical.yaml lists opnurasib (JDQ443) Phase 3 NSCLC as an active Novartis asset, but Novartis publicly discontinued opnurasib MONOTHERAPY development across all studies in July 2024, including halting the Phase 3 KontRASt-02 trial. Combo dose-...
“Novartis has decided to discontinue development of its experimental KRAS G12C inhibitor opnurasib (JDQ443) across all studies following a strategic review, including halting the Phase III KontRASt-02 trial evaluating the drug in non-small-cell lung cancer (NSCLC). 'This decision was made in light of the increasing options available to patients with KRAS G12C-driven cancers,' Novartis told FirstWord in a statement. 'It was not driven by clinical data and no new safety signals have been observe...”
Implication Highest-signal finding. Cut B must restate NVS opnurasib row from Ph3 to Ph1/2 combo and apply combo-PoS. Cluster correlation matrix needs to drop the implied KRAS G12C late-stage NVS competitor — this UNDERSTATES RVMD's competitive moat slightly (one fewer Ph3 G12C competitor...
Genentech terminated its 2021 SHP2-inhibitor collaboration with Relay Therapeutics on 2024-07-11 (notice received; effective 180 days later, approximately 2025-01-07), returning rights to migoprotafib / GDC-1971 / RLY-1971. Termination was 'without cause.' Relay had received a...
“The notice of termination was received by the company on Monday, July 11, 2024, and will take effect 180 days from that date ... Under the terms of the Agreement, the Company received aggregate consideration of $121.8 million from Genentech, including $75.0 million in an upfront payment, $45.0 million in milestone payments as well as reimbursement of certain research and development costs.”
Cut A canonical.yaml and prior workstream referenced the Phase 3 trial as 'ASPyRas'.
“CT.gov v2 API JSON has no 'acronym' field populated for NCT07409272; press release wording is 'first patient has been successfully dosed in the Phase 3 registrational study' with no trial-name acronym.”
IDYA close on 2026-05-15 = $27.88 (Stooq T2)
“15 May 2026 | Open 28.59 | High 29.18 | Low 27.88 | Close 28.21”
Implication Market cap recomputation: 87,856,154 sh × $28.21 = $2.478B (vs erroneous $2.449B at $27.88). EV recomputation: $2.478B - $972.9M net cash = $1.505B (vs $1.476B). | IDYA out-of-cluster scope (mechanism set excludes PKC/Werner/Pol-theta/MAT2A; preserved for traceability).
Market cap on 2026-05-15 ≈ $2.45B (using $27.88 × 87.86M)
“Derived from C-IDYA-MKT-01 (CONTRADICTED) × C-IDYA-MKT-02 (VERIFIED)”
Implication IDYA out-of-cluster scope (mechanism set excludes PKC/Werner/Pol-theta/MAT2A; preserved for traceability).
Enterprise value (mcap minus net cash) ≈ $1.48B as of 2026-05-15
“Derived from C-IDYA-MKT-03 (CONTRADICTED) − C-IDYA-FIN-01 (VERIFIED)”
Implication IDYA out-of-cluster scope (mechanism set excludes PKC/Werner/Pol-theta/MAT2A; preserved for traceability).
IDE275 = Werner Helicase inhibitor; IDEAYA plans to DISCONTINUE development after GSK terminated June-2020 Collaboration Agreement (eff. March 9, 2026); strategic options under evaluation
“IDE275 is a potential first-in-class, oral small molecule inhibitor of the helicase domain of the Werner protein, a RecQ enzyme involved in the maintenance of genome integrity. In December 2025, GSK notified us of its intention to terminate the GSK Collaboration Agreement, which became effective on March 9, 2026. We plan to discontinue development of IDE275 and are currently evaluating strategic options for this asset.”
Implication IDYA out-of-cluster scope (mechanism set excludes PKC/Werner/Pol-theta/MAT2A; preserved for traceability).
$7,926M→$13,045.5M
Self-audit: 39 percent understatement ($5.1B delta) in bigbio's W4 _cluster_summary.json
The $7,926M figure existed only in bigbio's own W4 _cluster_summary.json (internal aggregation error). The Street never published this number. W5 corrected to SEC XBRL primary. Self-audit catch, not Street catch.
2026-04-04→2025-09-04
7-month overstatement
BTD granted 7 months earlier than corpus claimed.
Silver only→Silver + Visionary (dual)
Dual-tier undercount
asco.org/annual-meeting/donors verbatim: 'Sponsorship Donors (Silver): Gilead and Kite; HELSINN; Lilly. Annual Meeting Track Donors (Visionary): AbbVie, Inc.; AstraZeneca; BioNTech; Lilly; Novartis.'
4 active Ph3 trials→5 active Ph3 trials
+1 Ph3 (NCT07554339 KANDLELIT-015, calderasib + durvalumab)
Pairings: pembro/Qlex x3 (KANDLELIT-004/-007/-013), cetuximab + mFOLFOX6 x1 (KANDLELIT-012, NOT pembro), durvalumab x1 (KANDLELIT-015). Reframes calderasib from Keytruda-only to combination-portfolio asset.
$896M→$1,345M
+$449M per program (programs count 3 -> 2)
naporafenib discontinued; ERAS-007/601 deprioritized. Denominator correction reframes ERAS valuation.
Resolution used ACTIVE_NOT_RECRUITING follow-up only + ABSENT from current Novartis pipeline (conservative middle; phase Ph3 preserved per registry; status_note carries narrative truth)
ACTIVE_NOT_RECRUITING (enrollmentActual 95 of planned 360, statusVerifiedDate 2026-04, lastUpdatePostDate 2026-04-21, whyStopped empty)
ABSENT from Novartis pipeline (106 disclosed projects, retrieved 2026-05-18; opnurasib + JDQ443 + KontRASt + TNO155 + batoprotafib all missing)
Novartis has decided to discontinue development of its experimental KRAS G12C inhibitor opnurasib (JDQ443) across all studies following a strategic review, including halting the Phase III KontRASt-02 trial (ApexOnco, 2024-07-18; FirstWord quote attributed to Novartis).
Resolution used TERMINATED (corporate disclosure post-dates CT.gov by 2 days and is authoritative on program status). Degrader thesis narrows to setidegrasib (G12D, VHL) + ASP5834 (pan-KRAS).
ACTIVE_NOT_RECRUITING (statusVerifiedDate 2026-01, lastUpdatePostDate 2026-02-02)
Astellas has decided to discontinue ASP4396, which used a cereblon E3 ligase, after determining that its data were no better than those it reported for setidegrasib. (Astellas Q3 FY2025 earnings call 2026-02-04, ApexOnco quote)
Cluster framing shifts
What sell-side carried into 2026 versus what the audit shows. One reframe per ticker.
opnurasib JDQ443 active Ph3 KRAS G12C NSCLC lead asset
Zombie status: KontRASt-02 ACTIVE_NOT_RECRUITING follow-up at n=95/360; opnurasib + JDQ443 + KontRASt + TNO155 + batoprotafib ALL ABSENT from Novartis pipeline (106 disclosed projects, retrieved 2026-05-18).
Reframe page rather than delete; NVS no active KRAS G12C Ph3 disclosed; downgrade in cluster re-rank.
calderasib = Keytruda lifecycle (4 Ph3 trials, pembro-paired)
Combination-portfolio asset spanning pembro + cetuximab + durvalumab across 5 Ph3s (KANDLELIT-012 = cetuximab + mFOLFOX6 NOT pembro; KANDLELIT-015 = calderasib + durvalumab/AstraZeneca, NOT_YET_RECRUITING, PCD 2032-12-15).
Strong asset, weaker Keytruda-only thesis; broader combo optionality than sell-side assumes.
ASP4396 cereblon E3 degrader on H2 2026 Ph1 readout tape; ASPyRas Ph3 setidegrasib
ASP4396 TERMINATED (Astellas Q3 FY2025 earnings call 2026-02-04: 'data were no better than those it reported for setidegrasib'). Degrader thesis narrows to setidegrasib (G12D, VHL) + ASP5834 (pan-KRAS). ASPyRas trial name does not exist; setidegrasib Ph3 = NCT07409272.
Drop H2 2026 ASP4396 Ph1 readout from forward catalyst tape; sell-side likely still on the old pipeline.
Silver ASCO 2026 sponsor
Dual sponsorship: Silver Sponsorship Donor AND Visionary Annual Meeting Track Donor (asco.org/donors verbatim).
Cluster valuation matrix should reflect breadth of ASCO presence as a non-financial signal; LLY undercount, not error.
Krascendo 2 = NCT06793219
Krascendo 2 = NCT06793215 (divarasib + pembrolizumab Ph3 1L KRAS G12C NSCLC, RECRUITING, last update 2026-05-11). NCT06793219 returns CT.gov 404.
Single-digit transcription error; minor but live on the site. Corrected in Cut A.1.
GSK collaboration + dual Werner/Pol-theta degrader pipeline (IDE275 + IDE705 Phase 1)
IDE275 + IDE705 BOTH DISCONTINUED (Q1 2026 10-Q accession 0001193125-26-206357). GSK Collaboration Agreement effective termination 2026-03-09. Story moves to: Servier ex-US mid-teens to low-twenties royalty (8-K 2025-08-29) + darovasertib monotherapy (OptimUM-02 437n, 159 events at Jan 23 cut-off) + degrader-narrowed pipeline.
Out-of-cluster scope (mechanism set excludes PKC/Werner/Pol-theta/MAT2A); preserved for traceability only. Does not propagate cluster narrative.
BMY adagrasib + MountainTAP framing absent from prior corpus
MountainTAP-29/-30/-9 pivotals + navlimetostat INN + ARLO-cel + CELMoD strategy surfaced as novel cluster intelligence (W4 12 novel claims).
Cluster framing expands BMY beyond adagrasib lifecycle to MTAP synthetic-lethal axis (BMS-986504 collaboration with RVMD).
Sotorasib LUMAKRAS lead-and-only KRAS G12C asset
Class pressure inbound: LLY olomorasib BTD class pressure + VSTM RAMP 203 combo + BI SOS1 combo. AMGN sotorasib competitive set thickens.
Sotorasib 2L+ NSCLC franchise faces sotorasib-alternative pressure; cluster competitive baseline shifts.
RAS(ON) multi-selective (daraxonrasib) lead franchise
RAS(ON) multi-selective + ivonescimab combo (NCT07397338 Ph1/2) + Q61H + G13C selectives newly disclosed (AACR LB337 daraxonrasib 1L mPDAC ORR 47%) — RAS(ON) pipeline broader than 10-Q discloses.
RVMD optionality undercounted; ivonescimab combo + Q61H + G13C selectives expand cluster-defining asset surface.
GDC-1971 / RLY-1971 SHP2 Roche/Relay collaboration progressing
Roche/Genentech TERMINATED Relay SHP2 collaboration 2024-07-11 (effective ~2025-01-07); migoprotafib rights returned to Relay; no Ph2 initiated. RIT1 M90I divarasib-resistance + migoprotafib reversal surfaced as novel mechanistic intelligence.
Roche SHP2 axis collapses to monotherapy-only divarasib Ph3 reads; SHP2 combo optionality off the table.
Named trade application
RVMD NDA submission via FDA Commissioner's National Priority Voucher (per 2026-04-13 press release + RVMD Q1 2026 10-Q Section 21 Royalty Pharma agreement).
Anchor event
RASolute 302 Ph3 topline 2026-04-13: median OS 13.2 vs 6.7 mo, HR 0.40, p<0.0001; confirmed via RVMD press release.
zoldonrasib (RMC-9805)
Ph1 KRAS G12D NSCLC
confirmed ORR 52% (n=27), median PFS 11.1 mo, 12-mo PFS 48%
RMC-6236 (daraxonrasib)
AACR LB337 1L mPDAC
ORR 47% (novel claim surfaced in W4)
| Ticker | Abstract / LBA | Session | Date / time | Asset | Note |
|---|---|---|---|---|---|
| IDYA | LBA9503 | Melanoma/Skin oral | 2026-06-01 08:00-11:00 CDT | darovasertib (OptimUM-02) | Out-of-cluster mechanism but same catalyst window. |
| BMY | Adagrasib MountainTAP pivotals + BMS-986504 MTAP-deletion combo with daraxonrasib (NCT07492680, RVMD collaboration). | ||||
| AMGN | Sotorasib LUMAKRAS/LUMYKRAS competitive baseline read against RVMD daraxonrasib. | ||||
| LLY | Olomorasib LY3537982 Ph3 KRAS G12C NSCLC; ASCO breadth as non-financial signal. | ||||
| MRK | Calderasib MK-1084 5-trial Ph3 portfolio (pembro x3 + cetuximab x1 + durvalumab x1). | ||||
| ROCH | Divarasib GDC-6036 Krascendo 1/2/3 reads; SHP2 axis collapsed post-Relay termination. | ||||
| NVS | Opnurasib status UNRESOLVED — read as ABSENT from pipeline per Novartis own disclosure. |
Positioning
Plenary LBA5 detail-reveal is the highest-density read in the six-month window. Spot $144.835 (2026-05-15) vs TipRanks 20-analyst median PT $181.30 = ~20 percent implied upside; HC Wainwright moved $73 -> $169 post-RASolute 302 topline. Trade structure: cluster-screen asymmetry favors RVMD on RAS(ON) franchise breadth (daraxonrasib + elironrasib + zoldonrasib + 5552 + ivonescimab combo) vs single-asset KRAS G12C peers.
Pre-embargo caveat
ASCO 2026 full-body abstracts for non-RVMD cluster members lift 2026-05-21 17:00 ET; re-audit BMY, AMGN, LLY, MRK, ROCH, NVS, ASLN, ERAS after embargo lift for satellite symposia + poster reads.
Per-ticker reads
12 nodes — primary source + audit-verdict pill + 1-sentence cluster note.
Revolution Medicines, Inc.
Lead asset
RMC-6236 daraxonrasib (RAS(ON) multi-selective)
V:38 · NC:4 · NCP:10
Benchmark node. Plenary LBA5 2026-05-31 daraxonrasib 1L mPDAC.
Amgen
Lead asset
sotorasib (LUMAKRAS/LUMYKRAS)
V:27 · NC:1 · U:2 · NCP:6
Approved KRAS G12C standard; class pressure from olomorasib + divarasib + calderasib.
Astellas Pharma
Lead asset
setidegrasib (ASP3082, KRAS G12D degrader)
V:17 · C:1 · NC:4 · NCP:3
Degrader thesis narrows post-ASP4396 termination; ASP5834 pan-KRAS carries optionality.
Boehringer Ingelheim
Lead asset
BI 1701963 + BI 1823911 (SOS1 inhibitors)
NC:2 · V:8 · NCP:2
Private equity — no SEC filings; SOS1 combo strategy via partnerships.
Bristol-Myers Squibb
Lead asset
adagrasib (MRTX849)
V:22 · NC:1 · NCP:12
Adagrasib NSCLC/CRC + BMS-986504 MTAP-deletion combo with daraxonrasib (NCT07492680, RVMD collaboration).
Erasca
Lead asset
ERAS-0015 (RAS-MAPK suite)
V:24 · C:2 · NC:3 · U:1
Programs count corrected 3 -> 2 (naporafenib discontinued, ERAS-007/601 deprioritized); EV/program +$449M.
IDEAYA Biosciences
Lead asset
darovasertib (PKC, OptimUM-02 437n)
C:3 · V:36 · D:2 · U:2 · NC:2
OUT-OF-CLUSTER (mechanism excludes PKC/Werner/Pol-theta/MAT2A). IDE275 + IDE705 DISCONTINUED; Servier ex-US deal mid-teens to low-twenties royalty.
Eli Lilly and Company
Lead asset
olomorasib (LY3537982, KRAS G12C)
V:21 · C:3 · NC:1 · NCP:5
Ph3 NSCLC; BTD granted 2025-09-04 (7-month corpus error corrected). Dual ASCO sponsor (Silver + Visionary).
Merck & Co. (MSD)
Lead asset
calderasib (MK-1084, KRAS G12C)
V:26 · NC:3 · NCP:7
5 Ph3 trials post-W5 (pembro x3 + cetuximab x1 + durvalumab x1); combination-portfolio asset.
Novartis
Lead asset
opnurasib (JDQ443, KRAS G12C) — UNRESOLVED
V:13 · C:1 · NC:1 · OS:1 · NCP:5
KontRASt-02 ACTIVE_NOT_RECRUITING follow-up at n=95/360; ABSENT from Novartis pipeline page (106 projects).
Roche / Genentech
Lead asset
divarasib (GDC-6036, KRAS G12C)
V:16 · NC:1 · C:1 · NCP:10
Krascendo 1/2/3 Ph3 reads; SHP2 axis collapsed post-Relay collaboration termination.
Verastem Oncology
Lead asset
avutometinib + defactinib (RAF/MEK clamp + FAK)
V:36 · NCP:2 · NC:1 · OS:1 · C:1
RAMP 203/301 combo strategies; no ASCO 2026 booth listed 2026-05-15 (borderline cluster member).
Methodology
D+C composition — Discipline (verbatim primary-source audit) applied to Catalyst (ASCO 2026 window, 2026-05-29 through 2026-06-03).
| Tier | Definition |
|---|---|
| T1 | Primary public-record source: SEC EDGAR filings (10-Q, 10-K, 20-F, 6-K, 8-K), ClinicalTrials.gov v2 API, FDA approval letters, ASCO official (asco.org donors page, meetings.asco.org abstract program), peer-reviewed publications (PubMed/J Clin Oncol/Annals of Oncology/NEJM/JTO), company IR pages, USPTO. |
| T2 | Specialist trade press that re-quotes named primary sources: ApexOnco/OncologyPipeline, OncLive, FirstWord Pharma, Synapse/PatSnap, market data terminals (Stooq, TipRanks). |
| T3 | Internal corpus and derived computations: prior dd-bigbio corpus, math worksheets, network analyses, cluster-internal cross-references. |
IDYA pipeline (darovasertib PKC inhibitor, IDE275 Werner helicase, IDE705 Pol-theta helicase, IDE397 MAT2A, IDE892, IDE161, IDE849, IDE034, IDE574) is outside the cluster mechanism inclusion criteria (KRAS_G12C, KRAS_G12D, KRAS_G12D_degrader, RAS_ON_multi, pan_RAS, pan_KRAS, SHP2, SOS1, MEK, RAF_MEK_clamp, FAK, ERK, mTOR_adjacent). IDYA audit findings preserved here for traceability — IDE275 + IDE705 DISCONTINUED, Servier ex-US deal 2025-08-29, OptimUM-02 437n — but IDYA does NOT propagate cluster narrative and does NOT count as a true cluster node.
/Users/davis/code/dayjobs_v2/dd-reports/rvmd-cluster-asco-2026_2026-05-15/canonical.yaml/Users/davis/code/dayjobs_v2/dd-reports/rvmd-cluster-asco-2026_2026-05-15/per_company/claims/*.json/Users/davis/code/dayjobs_v2/dd-reports/rvmd-cluster-asco-2026_2026-05-15/numeric_audit.md/Users/davis/code/dayjobs_v2/dd-reports/rvmd-cluster-asco-2026_2026-05-15/W5_reconciliation.md/Users/davis/code/dayjobs_v2/dd-reports/rvmd-cluster-asco-2026_2026-05-15/W6_design_spec.mdBibliography
Every claim above traces to one of these. 195 entries.