Tovorafenib (Ojemda) — Public-Domain Intelligence & Comparison Framework vs Plixorafenib

Competitive-intelligence dossier. Prepared for the Fore team. Data cut noted per section; tovorafenib figures reflect the 2024 registration cut unless a SNO 2025 three-year update is explicitly labeled.

⚠️ SCOPE & PROVENANCE CAVEAT — READ FIRST

This dossier is asymmetric by design. The tovorafenib (Ojemda) content in every section is drawn entirely from public, cited sources (Kilburn et al., Nature Medicine 2024; the OJEMDA U.S. Prescribing Information; the FDA Approval Summary, PMC11996598; EMA EPAR; Day One/Servier/Ipsen SEC and press disclosures). Each figure carries N, denominator, and source.

The plixorafenib content exists in one place only — the head-to-head comparison framework in Section 8 — and reproduces ONLY the figures on the Shubin slide, labeled "Fore-internal (per slide) — verify." These are Fore-internal, embargoed, and UNVERIFIED in this dossier. No plixorafenib efficacy or safety data was researched, inferred, or fabricated. Every plixorafenib efficacy/safety cell is a template placeholder for the Fore team to complete from cleared internal data (FORTE Sub A/Sub B, PLX120-03).

[FLAG: DATA-CUT / TEMPORAL — verify before external use] The tovorafenib efficacy column reflects the 2024 registration data cut. As of mid-2026 a three-year FIREFLY-1 update was presented at SNO 2025 with higher response estimates (RAPNO-LGG ORR ~53%, RANO-HGG ORR ~71%) plus updated DoR/PFS and treatment-free-observation data. A mid-2026 competitive dossier must either (a) label every tovorafenib figure with the registration cut date, or (b) refresh to the SNO 2025 cut. Do not present 2024 numbers as "current."

1. Executive Summary & Key Takeaways

Tovorafenib (Ojemda, Day One Biopharmaceuticals; ex-U.S. rights licensed to Ipsen; Day One acquired by Servier, closed 23 Apr 2026) is an oral, once-weekly, CNS-penetrant type II pan-RAF (RAF-dimer) inhibitor. It holds FDA accelerated approval (23 Apr 2024) and EU conditional marketing authorization (20 Apr 2026, EMA/EC) for relapsed/refractory (R/R) pediatric low-grade glioma (pLGG) harboring a BRAF fusion/rearrangement or BRAF V600 mutation, age ≥6 months, after ≥1 prior systemic therapy. It is the first FDA-approved targeted therapy for the KIAA1549::BRAF-fusion pLGG population — a segment that type I RAF inhibitors (dabrafenib) structurally cannot serve.

The single most important competitive nuance: tovorafenib's headline ORR is criterion-dependent. The FIREFLY-1 Arm 1 dataset yields ORR 67% by RANO-HGG (the prespecified primary endpoint) but only ~51–53% by RANO-LGG / RAPNO-LGG, and the FDA label deliberately adopted the lower RAPNO-LGG figure (51%). Any cross-drug comparison must fix the criterion first; a bare "67%" against a competitor's RANO-LGG number is not like-for-like.

Safety is a manageable pediatric profile with no boxed warning, dominated by on-target class effects (hair depigmentation, rash, CPK elevation, transaminase elevation) that are high-frequency but largely low-grade and reversible. Its chief differentiating liabilities are major/tumoral hemorrhage (one fatal) and reversible growth-velocity suppression (without growth-plate closure).

Commercially the launch is ramping fast — FY2025 U.S. net product revenue $155.4M (+172% YoY), 4,635 prescriptions (+181%), 2026 guidance $225–250M — a trajectory Servier's ~$2.5B acquisition validates. The key future inflection is FIREFLY-2 (frontline Phase 3 vs chemotherapy), fully enrolled 8 May 2026 (~140 sites, ~400 patients), topline expected mid-2027; a positive readout would move tovorafenib to first line and is required to convert accelerated approval to traditional approval (not yet occurred).

Real-world evidence is genuinely immature (~2 years marketed). Nearly all durability/discontinuation/rebound data are trial-derived (FIREFLY-1 extension), not independent RWE; the only true independent RWE located is a single 8-patient off-label adult case series (Mayo Clinic, SNO 2025).

Key Takeaways

1. First-in-class for the fusion majority. As a type II dimer inhibitor, tovorafenib owns the BRAF-fusion (KIAA1549::BRAF) majority of pLGG that type I inhibitors cannot address without paradoxical MAPK activation — its core commercial and mechanistic moat.
2. ORR is criterion-dependent; fix the criterion before comparing. 67% RANO-HGG (primary) vs 51% RAPNO-LGG (label) on the same patients — the ~16-point gap is a criteria artifact, not a difference in tumor control. Criterion labels are applied inconsistently across Day One's own materials (see Section 3 contradiction flag).
3. Manageable pediatric safety, no boxed warning. Grade ≥3 TRAEs 42%, discontinuation only 7%, no confirmed DILI, no meaningful QT signal; differentiating risks are hemorrhage (one fatal) and reversible growth suppression.
4. Fast commercial ramp under new ownership. FY2025 $155.4M (+172%), 2026 guidance $225–250M; Servier acquired Day One (~$2.5B, closed 23 Apr 2026); Ipsen holds ex-U.S. rights.
5. FIREFLY-2 is the inflection. Frontline Phase 3 fully enrolled (8 May 2026), topline mid-2027; required to convert accelerated → traditional approval and to contest chemo / dabrafenib+trametinib in first line.
6. RWE is still essentially trial-derived. Durability, treatment-free-interval, rebound (31%), retreatment (−38.3%), and growth-recovery data all come from FIREFLY-1 extension, not independent cohorts — the central interpretive risk for a competitive dossier.
7. Plixorafenib column is a template only. All plixorafenib cells in Section 8 are per-slide, embargoed, unverified placeholders for the Fore team to complete from cleared internal data.

2. Mechanism, PK/PD & Formulation

Tovorafenib only. Plixorafenib not referenced.

2.1 Mechanism of action — type II pan-RAF (dimer) inhibitor

Oral, CNS-penetrant, type II RAF inhibitor that inhibits both RAF monomers and RAF dimers, with enzymatic activity against BRAF V600E–mutant, wild-type BRAF, and wild-type CRAF (FDA label §12.1; Singh et al., Clin Cancer Res 2025;31(8):1383–1389).

• Type II / DFG-out binding. X-ray structures show binding in the classic type II mode — kinase in a DFG-out, αC-helix-in conformation, occupying the back hydrophobic pocket (J Biol Chem 2023;299(5), PMC10149214).
• RAF-isoform potency (enzymatic IC₅₀): CRAF ≈ 94.2 nM; wild-type BRAF ≈ 633 nM; ARAF > 3000 nM (incomplete inhibition at 10 µM); BRAF V600E ≈ 495 nM for reference — markedly weaker ARAF activity, a class property of type II RAF inhibitors.
• Cooperative dimer inhibition. Against BRAF wild-type dimers, strong positive cooperativity (Hill slope ≈ −2.9); CRAF wild-type dimers show milder cooperativity (Hill slope ≈ −1.6); no cooperativity with monomeric BRAF V600E or ARAF. [FLAG: original text stated "Hill slopes ≈ −2.6 to −3.2" for both BRAF and CRAF dimers; the source Table 1 reports CRAF WT ≈ −1.55 ± 0.43 (not in that range) and BRAF WT ≈ −2.89 ± 0.19. Human should confirm exact Hill-slope values and any additional dimer assays against Table 1 before use.]
• Alteration-class agnostic. By blocking dimers, addresses BRAF alterations signaling independently of upstream RAS — notably the KIAA1549::BRAF fusion (constitutively dimerized, the leading pLGG driver) — plus V600 monomer disease. Antitumor activity shown in murine intracranial models expressing either BRAF V600E or KIAA1549::BRAF.
• Avoidance of paradoxical MAPK activation. By inhibiting the dimer itself, tovorafenib is designed to avoid the paradoxical pERK rise that limits type I / "type 1.5" inhibitors (vemurafenib, dabrafenib) in RAS-active/wild-type-BRAF contexts. In NF1-loss preclinical models a bell-shaped pERK curve was still observed and single-agent antiproliferative activity was minimal (EC₅₀ up to >10 µmol/L). [FLAG: original stated "EC₅₀ >7.5 µmol/L in most lines"; source Table 1 reports a range of ~0.99 to >10 µmol/L (e.g., MeWo 7.5, NCI-H1838 >10), so "most lines >7.5" slightly overstates — some lines were ~1 µmol/L. Confirm the exact per-line distribution.] [FLAG: NF1-loss paradoxical-pERK finding is preclinical only; limited relevance to the approved BRAF-fusion/V600 indication — retained from original.]

2.2 Pharmacokinetics (FDA PI §12.3; %CV in parentheses)

• Weekly-dosing rationale: ~56 h half-life underpins once-weekly dosing; steady state by ~12 days without meaningful accumulation.
• Food effect: high-fat meal caused no clinically significant change in Cmax/AUC (Tmax delayed ~3 h → ~6.5 h); may be taken with or without food.
• Metabolism: primarily aldehyde oxidase and CYP2C8; CYP3A/2C9/2C19 minor.
• Elimination: ~65% feces (8.6% unchanged), ~27% urine (0.2% unchanged) — predominantly metabolic.

2.3 Pharmacodynamics / exposure-response

• Cardiac: at 380 mg/m² weekly (max 600 mg), mean QT increase >20 ms was not observed (§12.2).
• Efficacy E-R: across 290–476 mg/m², no clinically significant exposure-response for ORR.
• Safety E-R: greater exposure associated with increased Grade ≥2 AEs (rash, transaminase elevation, CPK). Approved 380 mg/m² is a modest reduction from the initially intended 420 mg/m²; PK/PD modeling indicated improved benefit-risk, including for growth-velocity reductions in children. [FLAG: growth-velocity effect is a pediatric-specific driver of dose selection.]

2.4 Drug–drug interactions (FDA label §7)

• CYP2C8 (substrate): avoid strong/moderate CYP2C8 inhibitors (↑ exposure→toxicity) and inducers (↓ exposure→loss of efficacy).
• Effect on other drugs: midazolam (CYP3A4 substrate) Cmax/AUC predicted to ↓ ≥20% (net CYP3A induction); hormonal contraceptives may lose efficacy — use non-hormonal contraception.
• Transporters: not a substrate of BCRP, P-gp, OATP1B1/1B3; is a BCRP inhibitor at clinical concentrations.
• Special populations: no adjustment for mild hepatic or mild–moderate renal impairment (eGFR ≥30); not studied in moderate–severe hepatic or severe renal.

2.5 Formulations & dosing (FDA label §2, §3)

• Dosage forms: film-coated tablets 100 mg; powder for oral suspension 25 mg/mL (reconstituted).
• Recommended dosage: 380 mg/m² orally once weekly, max 600 mg/week; with or without food; same scheduled time weekly; no loading dose.

CNS-penetration caveat: quantitative brain-to-plasma / Kp,uu ratio not found in public sources; CNS penetration is supported qualitatively and via intracranial-model efficacy only.

3. Clinical Efficacy

3.1 Interpretive summary (read before the tables)

Tovorafenib's headline ORR depends entirely on the response criterion applied to the same patients. FIREFLY-1 Arm 1 yields ORR 67% by RANO-HGG (prespecified primary) but only ~51–53% by RANO-LGG/RAPNO-LGG (FDA label adopted the lower RAPNO-LGG figure). [FLAG: criterion-dependent ORR — the ~16-point gap is a criteria artifact, not a difference in tumor control.] [FLAG: criterion LABELS are applied inconsistently across Day One's own materials — see the RAPNO/RANO-LGG contradiction flagged below; do not present a clean "51%=RAPNO / 53%=RANO" split as settled.]

3.2 FIREFLY-1 (PNOC026 / NCT04775485) — pivotal, Arm 1

Phase 2, single-arm, open-label, multicenter. Patients 6 months–25 years with R/R BRAF-altered pLGG progressing after ≥1 prior line. Dose 380 mg/m² PO once weekly (max 600). Primary source: Kilburn LB, et al. Nat Med 2024;30:207–217.

Primary endpoint — ORR by independent review, RANO-HGG (prespecified primary): RANO-HGG evaluable population = 69. [VERIFIED: N=69 denominator confirmed against Nat Med 2024.]

Post-hoc exploratory — ORR by RANO-LGG:

FDA label basis — RAPNO-LGG (US label figures): Efficacy population N=76, ORR 51% (95% CI 40–63); 0 CR, 28 PR (37%), 11 minor responses (14%); median DoR 13.8 mo (11.3–NE); among 39 responders 85% had DoR ≥6 mo, 23% ≥12 mo; median TTR 5.3 mo (1.6–11.2). Approved 23 Apr 2024. [VERIFIED: label ORR 51%, DoR 13.8 mo, TTR 5.3 mo, 37% PR + 14% MR, approval date 23 Apr 2024 all confirmed.] [FLAG — CONTRADICTION, criterion labeling: Day One's SNO 2025 press release attributes the 53% (40/76) figure to RAPNO-LGG criteria, whereas the FDA label attributes 51% (39/76) to RAPNO-LGG. Same criterion name, different responder count (40 vs 39) and ORR (53% vs 51%). The workstream's assignment of 53%→RANO-LGG and 51%→RAPNO-LGG is therefore NOT how Day One itself labels these numbers. A competitive dossier must reconcile this one-responder / criterion-label discrepancy against the primary paper's supplementary tables before asserting which criterion each number belongs to.]

ORR by BRAF alteration class:

ORR by age not reported as a powered subgroup; efficacy described as consistent across 2–21 yr. [FLAG: age-subgroup ORR not published with N/CI.]

3.3 FIREFLY-1 baseline characteristics — Arm 1 (N=77)

Heavily pretreated, high-risk population (median 3 prior lines; 60% prior MAPK inhibitor; predominantly optic-pathway/unresectable) — critical context when comparing raw ORR to any frontline or less-pretreated dataset.

Arm 2 (N=60): extension/access cohort for recurrent/progressive RAF-altered pLGG; primary endpoint was safety, not efficacy. Arms 1+2 = 137-patient safety population. [VERIFIED: 137-patient two-arm safety population confirmed; Arm 1 n=77 used for efficacy.] [FLAG: Arm 2 baseline demographic breakdown not fully published in accessible sources; only cohort definition and N=60 confirmed.]

3.4 FIREFLY-1 durability update — 3-year follow-up (SNO 2025)

Day One press release (24 Nov 2025; cutoff 6 Jun 2025) + abstract CTP-17; Neuro-Oncology Pediatrics 2025;2(2):wuag022. Arm 1 efficacy-evaluable N=76:

Update emphasized clinical stability off treatment: 77% of patients who entered the treatment-free observation period remained off therapy ≥12 months (median study duration 40.6 mo); responses can persist after cessation, and retreatment is feasible — a distinctive commercial claim. [FLAG: the 13.8→19.4 mo DoR-lengthening claim inherits the unconfirmed 19.4-mo DoR above; verify before using as a comparative claim.]

3.5 FIREFLY-2 / LOGGIC-FIREFLY-2 (NCT05566795) — frontline, registrational

Phase 3, randomized, open-label, 2-arm, global (~140 sites) with SIOPe LOGGIC Consortium; Day One + Servier. Source: Bouffet E, et al. BMC Cancer 2024 (PMC10826080); ASCO 2023 TPS10067; Servier/Day One release (8 May 2026).

This is the study that would move tovorafenib from R/R accelerated approval to a frontline, chemo-controlled label — the key future competitive inflection. No efficacy readout yet. [FLAG: no additional pediatric-LGG efficacy datasets located in public sources beyond FIREFLY-1 (R/R) and FIREFLY-2 (pending).]

4. Safety & Tolerability

Tovorafenib only. Phase 2 FIREFLY-1 (NCT04775485) + FDA OJEMDA label.

4.1 Safety populations (denominators matter)

• N=137 — R/R pLGG, FIREFLY-1 Arms 1+2. Primary label safety population for the Adverse Reactions table and dose-modification rates.
• N=172 — pooled safety population (pLGG + advanced solid tumors); 86% exposed ≥6 months, 49% ≥1 year. Denominator for Warnings & Precautions rates and lab abnormalities.

[FLAG: label version] Figures here are drawn from the initial 2024 OJEMDA PI. A 2025 label revision exists (accessdata …/2025/218033s002s003lbl.pdf) — a human must confirm no Warnings-&-Precautions or Adverse-Reactions rates were updated before this enters the dossier.

[FLAG: source-dependent denominators — VERIFIED as real] Warnings-&-Precautions rates (37% hemorrhage, 42% ALT, etc.) are quoted against the N=172 pooled population in the label narrative; the Adverse Reactions table percentages are against N=137 pLGG. Confirmed against PMC11996598: rash 77% (table, N=137) vs 67% (§5.2, pooled) and hemorrhage 42% (table) vs 37% (pooled §5.1) are both genuine label figures, differing only by population.

4.2 Common adverse reactions (all-grade), FIREFLY-1 pLGG, N=137 (≥30%)

TRAE view (Kilburn et al., N=137): hair color changes 76% (104/137), elevated CPK 56% (77/137), anemia 49% (67/137), fatigue 44% (60/137), maculopapular rash 41% (56/137).

4.3 Grade ≥3 events

• Grade ≥3 TRAEs: 42% (58/137). Most common: elevated CPK 12%, anemia 10%, maculopapular rash 8%.
• Label Grade ≥3 adverse reactions (≥5%): decreased growth velocity, anemia, blood CPK increased, maculopapular rash, ALT increased.

4.4 Boxed warnings

None. OJEMDA carries no boxed warning.

4.5 Warnings & Precautions (§5), pooled N=172 unless noted

1. Hemorrhage (§5.1): 37% all-grade; serious/major 5%; Grade 5 fatal tumor hemorrhage in 1 patient — reported by FDA as "1 patient (1%)". [FLAG: the FDA Approval Summary/label state the fatal hemorrhage as "1 patient (1%)"; the original workstream's "0.6%" silently used the larger pooled denominator. A human must decide which denominator the dossier presents and align it — recommend quoting FDA's "1 patient (1%)" verbatim.] Epistaxis 26%; intratumoral/intracranial hemorrhage 9%; gingival bleeding 5%. Permanent discontinuation for hemorrhage ~2%.
2. Cutaneous / photosensitivity (§5.2): Rash 67% (Grade 3 12%); dermatitis acneiform 26% (Grade 3 0.6%); photosensitivity 12% (Grade 3 0.6%, N=1). Sun-protection counseling required.
3. Hepatotoxicity (§5.3): ALT increased 42% (Grade 3 4%); AST increased 74% (Grade 3 2%); bilirubin increased 23% (Grade 3 0.6%). LiverTox pooled: ALT >5×ULN 5%, AST >5×ULN 2%; median onset 14 days (3–280); dose interruption for LFTs ~5%; no jaundice, no life-threatening/fatal DILI, zero published clinically apparent liver-injury case reports. DILI likelihood score E*. [FLAG: LiverTox ALT>5×ULN 5%, AST>5×ULN 2%, onset 14d and DILI score E* not independently re-verified in this pass — low risk, but confirm against NBK613920.]
4. Reduction in growth velocity (§5.4): In FIREFLY-1 patients ≤18 yr (N=133), growth-velocity reduction all-grade 46%, Grade ≥3 35%. Height percentile: median change −14 at 12 mo (N=107), −20 at 18 mo (N=95); z-score −0.6 and −0.9. Recovery after ≥90 days off-treatment: annualized growth velocity ~4.2 cm/yr (N=17). Dose interruption 5%, reduction 2.3%, discontinuation 3%. Growth-plate: among evaluable with hand radiographs (N=17 of 81), no premature epiphyseal growth-plate closure and no bone-age advancement. [FLAG: the −14/−20 height-percentile, z-score, 4.2 cm/yr recovery, and 17-of-81 growth-plate figures are cited to §5.4 but not independently re-verified in this pass — confirm against the label before dossier use.]
5. Embryo-fetal toxicity (§5.5): Embryo-lethal in rats at ~0.8× human exposure; contraception advised.
6. NF1-associated tumors (§5.6): Warning based on nonclinical data that RAF inhibition may promote growth of tumors with NF1 loss; no clinical incidence quantified.

4.6 Cardiac — LVEF / QT

• QT: at recommended dose, mean QT increase >20 ms not observed; no clinically meaningful QT signal reported. [FLAG: QT language attributed to the label was not surfaced in the FDA Approval Summary (which does not report QT); confirm the ">20 ms not observed" wording exists in the PI/Clinical Pharmacology section before quoting.]
• LVEF: no LVEF/ejection-fraction toxicity signal reported. [FLAG: absence of reported data, not proven absence of effect.]

4.7 Discontinuation / dose-modification (FIREFLY-1 pLGG, N=137)

• Permanent discontinuation due to AEs: 7% (9/137) — most common tumor hemorrhage (3) and decreased growth velocity (2).
• Dose reduction: 24% (33/137) — most common rash, fatigue.
• Dose interruption: label 57% (all-cause); primary paper 37% (50/137) (TRAE-specific, median 14 days). Difference = all-cause vs treatment-related coding.
• Serious adverse reactions: 45%. No on-trial treatment-related deaths (the single fatal tumor hemorrhage was categorized as a Grade 5 hemorrhage event).
• 3-year follow-up (N=137, SNO 2025): no new safety signals identified.

4.8 FDA FAERS post-marketing signals (since April 2024)

openFDA FAERS returns a very sparse, low-count dataset. Top terms: "off-label use" (6), "disease progression" (2), then single-report (n=1) terms including death, intracranial tumor hemorrhage, reversible cerebral vasoconstriction syndrome (RCVS), rhabdomyolysis, meningeal metastases, hepatic cytolysis, haematotoxicity, rash, dermatitis acneiform, thunderclap headache, somnolence, confusional state. [FLAG: the specific per-term FAERS counts above were not re-run in this verification pass; a human must re-query the openFDA endpoint at dossier-freeze date and confirm counts, since FAERS is dynamic and these single-report terms will change.] - Potential emerging signal: RCVS / thunderclap headache is not an established labeled AE. [FLAG: unvalidated — FAERS single-case reports carry no causality/denominator; per FDA disclaimer these are not confirmed signals and require a disproportionality analysis to interpret.]

4.9 Safety bottom line for CI

Profile dominated by on-target class effects (hair depigmentation, rash, CPK elevation, transaminase elevation) — high-frequency but largely low-grade and reversible — plus three management-defining risks: major/tumoral hemorrhage (one fatal), reversible growth-velocity reduction without growth-plate closure, and photosensitivity. Grade ≥3 TRAEs 42%, discontinuation only 7%, no boxed warning, no meaningful QT signal, no confirmed DILI — a manageable pediatric profile whose chief differentiating liabilities are hemorrhage and growth suppression.

5. Real-World Evidence (RWE)

Scope / maturity caveat. FDA accelerated approval 23 Apr 2024; EU conditional 2025 — post-approval RWE is genuinely immature (~2 years as of mid-2026). Published "durability," "discontinuation," and "rebound" data are overwhelmingly trial-derived (FIREFLY-1 extension), not independent real-world cohorts. Truly independent RWE = a single small off-label adult case series plus commercial uptake/access signals. [FLAG: absence of evidence, not evidence of absence — searched, no tovorafenib-specific FAERS disproportionality study found as of Jul 2026.]

5.1 Independent real-world / off-label use (only true RWE series found)

Adult, off-label CNS use — Mayo Clinic single-institution series (SNO 2025). Williams et al., abstract PATH-94, Neuro-Oncology 2025;27(Suppl 5):v263. - 8 adult patients (range 21–66, median 42; 5M/3F), Mar 2024–Jun 2025 — starting almost immediately after the pediatric-only approval. - Tumor types: pilocytic astrocytoma (n=2), HGAP (n=2), Erdheim–Chester disease (n=2), pleomorphic xanthoastrocytoma (n=1), unclassified LGG (n=1). BRAF: KIAA1549 fusion (n=4), V600E (n=3), GAB2 fusion (n=1). - Radiographic response in 5/8: 3 stability (1 later progressed), 2 tumor-size decreases (both KIAA1549-fusion). [FLAG: response criteria (RANO vs RAPNO) not stated in abstract — do not label as RANO/RAPNO in the dossier without the full abstract text.] - Tolerability: fatigue (n=2), plus headache, myalgias, diplopia, cytopenias, transaminitis. 3/8 paused or stopped for toxicity. - Interpretation: early, hypothesis-generating; N=8, single center, uncontrolled — not generalizable. [FLAG: gap — pediatric RWE remains entirely FIREFLY-1-based; no published pediatric real-world outcomes cohort found.]

5.2 Discontinuation, treatment-free interval, rebound growth (trial-derived, FIREFLY-1 extension)

SNO 2025 3-year update; Neuro-Oncology Pediatrics 2025;2(2):wuag022; median follow-up 40.6 months. - After a planned 26 cycles (median treatment ~24.6 mo), patients could stop and enter observation with retreatment allowed at progression. - 39 patients entered treatment-free observation; median treatment-free interval not reached; 30/39 (77%) treatment-free ≥12 months. - Rebound growth: 12/39 (31%) met rebound criteria (≥25% increase); minimal in first 6 months off therapy; 10/12 (83%) of rebounders still below pre-treatment baseline — rebound generally did not erase benefit. - Retreatment (n=8): all 8 remained on therapy at cutoff; median tumor-size change −38.3%; median retreatment duration 9.0 months — early re-sensitization signal. - Median time to next treatment 42.6 months. [FLAG: 42.6-month TTNT figure not independently confirmed in this pass — verify against the SNO 2025 abstract/slides before citing.]

Growth-velocity recovery on stopping (ASCO 2025, J Clin Oncol 2025;43(16_suppl):10029; Neuro-Onc Peds 2025;2(2):wuag021): off-treatment AGV recovered time-dependently — median ~4.3 cm/y at 3 mo, ~10.2 cm/y at 6 mo, ~7.7 cm/y at 12 mo — with 34/38 (89%) recovering AGV and 28 (74%) increasing height Z-score toward baseline, without advanced bone age or premature epiphyseal closure. [CORRECTED: the original "average on-treatment ~1.1 cm/y rising to ~8 cm/y post-treatment" oversimplified a time-dependent recovery curve; replaced with the published 3/6/12-month AGV values and recovery/catch-up rates.] [FLAG: the on-treatment baseline AGV (originally stated ~1.1 cm/y) was not independently confirmed — verify the on-treatment velocity figure before citing a single number.]

[FLAG: optimal duration and stop/re-treat strategy not yet answered by RWE.] The 26-cycle stop rule is a trial construct; no real-world dataset yet validates stop-and-observe outside FIREFLY-1.

5.3 Regrowth-pattern definitions (infrastructure for future RWE)

International Pediatric LGG Coalition modified-Delphi consensus distinguishing resistance vs rebound vs recurrence (Neuro-Oncology 2024;26(8):1357), standardizing endpoints so future real-world series can be compared.

5.4 Access, uptake, commercial real-world signals (Day One SEC/press)

Access improvement: Q4 2024 CMS granted Ojemda an Exclusively-Pediatric designation, lowering its Medicaid/340B minimum rebate from 23.1% to 17.1%. [CORRECTED: "Dec 2024 CMS agreed" reframed as the Q4-2024 Exclusively-Pediatric designation, the mechanism actually disclosed.]

Caveat: revenue/prescription growth reflects commercial adoption (including off-label use), not real-world efficacy or safety. [FLAG: TEMPORAL — as of mid-July 2026, Day One's Q1 2026 (quarter ended 31 Mar 2026) results are typically reported by early/mid-May and are therefore likely now available; the dossier should incorporate Q1 2026 actuals rather than treating them as not-yet-reported.]

5.5 FAERS / real-world safety

• No tovorafenib-specific FAERS disproportionality study located. [FLAG: gap — post-marketing pharmacovigilance analysis not yet published; likely below the report-volume threshold used for such analyses.]
• Only marketed-period real-world tolerability data are the Mayo adult series (§5.1): 3/8 toxicity-driven interruptions/stops — consistent with label signals but too small to constitute a real-world signal.

5.6 RWE bottom line and explicit gaps

• Real-world data that exist: rapid commercial uptake ($155.4M revenue, 4,635 scripts in 2025, broad prescriber base), a favorable CMS access change (23.1%→17.1%), and one 8-patient off-label adult case series (esp. KIAA1549-fusion responses).
• Still trial-only (not RWE): all durability, treatment-free-interval, rebound (31%), retreatment (−38.3%), and growth-recovery data come from FIREFLY-1 extension.
• Key gaps: (1) no independent pediatric real-world outcomes cohort/registry; (2) no published FAERS pharmacovigilance study; (3) optimal duration and stop-and-re-treat strategy unresolved outside the trial; (4) long-term (>3 yr) real-world durability/late-toxicity data absent by definition; (5) adult/off-label use evidenced only anecdotally.

6. Regulatory & Label Status by Region

6.1 Corporate / sponsor context (affects regulatory holder identity)

• U.S. holder: Day One Biopharmaceuticals — original NDA holder. On 23 April 2026 Servier completed its acquisition of Day One (tender offer $21.50/share, ~$2.5B equity value); U.S. rights now sit under Servier, though the NDA holder of record is Day One.
• Ex-U.S. holder: Ipsen licensed ex-U.S. rights from Day One in 2024 and is the EU marketing authorisation holder. [FLAG: $461M is total incl. milestones, not upfront — confirm upfront vs. contingent split before citing as a single number.]
• Confirmatory-trial partner: FIREFLY-2 run by Day One "and Servier" with the SIOPe LOGGIC Consortium.

6.2 United States — FDA (OJEMDA)

• Action / date: Accelerated approval, 23 April 2024. Priority review, breakthrough therapy, orphan drug designation.
• Exact indication: "Treatment of patients 6 months of age and older with relapsed or refractory pediatric low-grade glioma (pLGG) harboring a BRAF fusion or rearrangement, or BRAF V600 mutation."
• First-in-class note: first FDA-approved systemic therapy for pLGG with BRAF rearrangements/fusions.
• Basis (FIREFLY-1): single-arm Phase 2, N=76 efficacy; ORR 51% (95% CI 40–63) by RAPNO (blinded independent central review); median DOR 13.8 mo (11.3–NE). Response by RAPNO, not RECIST.
• Dosing: 380 mg/m² orally once weekly, max 600 mg; with or without food.
• Formulations: 100 mg film-coated tablets; powder for oral suspension (25 mg/mL). [FLAG: only a 100 mg tablet strength was confirmed in sources; if a later 2025 label added additional tablet strengths I could not verify it — the accessdata label PDFs returned 404.]
• Companion diagnostic: no CDx co-approved at launch; postmarketing commitment to validate a CDx; later OncLive reporting references a subsequent CDx approval. [FLAG: confirm exact CDx name and approval date from the OncLive/FDA source before citing as a completed milestone.]

Confirmatory-trial obligation & conversion (US): - Continued approval contingent on FIREFLY-2 / LOGGIC (NCT05566795) — Phase 3 randomized (1:1) tovorafenib vs investigator's-choice SoC chemo in newly diagnosed/treatment-naïve pLGG. - Trial footprint: ~140 sites; ~400 participants. [CORRECTED: earlier "~100 global sites" reflected the 2024 BMC protocol paper; completion release reports ~140 sites / ~400 participants.] - Enrollment completed 8 May 2026; topline expected mid-2027. - Conversion to full/traditional approval: Not yet occurred. [FLAG: no public FDA action converting accelerated→traditional approval found; status is "confirmatory trial fully enrolled, awaiting data/verification."]

6.3 European Union — EMA / European Commission (Ojemda)

• Action / date: Conditional marketing authorisation, EU-wide, granted 20 April 2026 (one-year conditional MA). [FLAG: 2-day date discrepancy between EMA/EC "20 April 2026" and Ipsen's announced EC decision "22 April 2026" — treat EMA's 20 April 2026 as authoritative.]
• Preceded by: positive CHMP opinion (Ipsen).
• Exact indication (EMA EPAR): "Ojemda is indicated as monotherapy for the treatment of patients 6 months of age and older with paediatric low-grade glioma (LGG) harbouring a BRAF fusion or rearrangement, or BRAF V600 mutation, who have progressed after one or more prior systemic therapies."
• Formulations: tablets + powder for oral suspension; oral, once weekly.
• Geographic scope: all 27 EU member states + Iceland, Liechtenstein, Norway.
• HTA: subject of the EU Joint Clinical Assessment (JCA) — the first published JCA under the new EU HTA framework; German G-BA orphan-drug (AMNOG) benefit assessment initiated post-approval.

[FLAG — indication-scope discrepancy: Ipsen's press-release headline says approved "regardless of BRAF alteration," but the binding EMA indication is restricted to tumours harbouring a BRAF fusion/rearrangement or V600 mutation and to prior-progression (R/R) patients. The headline appears to mean "regardless of the type of BRAF alteration," not a BRAF-agnostic indication. Rely on EMA EPAR wording.]

6.4 Japan — PMDA

• Status: Orphan drug designation documented; no PMDA marketing approval or formal filing found. [FLAG: Japan approval status unverified — only orphan designation confirmed; no PMDA approval date located. Ipsen holds ex-U.S. rights and would be the likely applicant.]

6.5 Other regions

• No confirmed approvals beyond U.S. (accelerated) and EU (conditional). [FLAG: UK MHRA, Swissmedic, Health Canada, Australia TGA statuses not verified — no public approval records surfaced.]

6.6 Quick-reference matrix

Common label attributes (US + EU): monotherapy; age floor 6 months; once-weekly oral dosing (US 380 mg/m², max 600 mg); tablet + oral-suspension formulations; response by RAPNO (FIREFLY-1) / RANO (FIREFLY-2), not RECIST.

7. Commercial, Competitive Position & IP

7.1 Developer, ownership & commercial structure

• Originator / U.S. marketer: Day One Biopharmaceuticals (Nasdaq: DAWN). Originated as MLN2480 / TAK-580 (Takeda lineage), developed by Day One as DAY101.
• Ex-U.S. rights (Ipsen): 23-Jul-2024, exclusive ex-U.S. commercialization rights. Terms: ~$111M upfront (~$71M cash + ~$40M equity), up to ~$350M in milestones, and tiered double-digit royalties starting mid-teens %. [FLAG: the "royalties run ≥10 years from first commercial sale per country" detail was in the original draft but is NOT stated in the public Ipsen release — confirm against the licensed agreement exhibit (Day One 10-Q dawn-ex10_1, redacted) before external use; removed from the sentence above pending confirmation.]
• Acquisition by Servier (closed): agreed 6-Mar-2026 to acquire Day One for $21.50/share cash (~$2.5B equity value) — reported ~68% premium to the 5-Mar-2026 close. Tender offer completed 23-Apr-2026. OJEMDA now sits inside Servier's oncology franchise alongside Voranigo (vorasidenib), giving Servier a paired pediatric/adult glioma portfolio. [FLAG: ~68% premium figure not re-verified against a primary source in this pass — confirm against the offer document. Post-close consolidated OJEMDA reporting under Servier not yet public as of this dossier.]
• Priority Review Voucher monetization: rare-pediatric-disease PRV sold for $108M on 30-May-2024 ($8.1M to Viracta to satisfy a license obligation) — a non-dilutive cash event distinct from product sales.

7.2 Pricing (U.S.)

• WAC (2024 launch): $35,272.64 per package for the 100 mg tablet (16-/20-/24-count bottles); ~$8,818/bottle for the oral suspension (7-day supply). [FLAG: original draft stated "~$8,479" for the oral suspension — that figure does not match the disclosed 2024 WAC of $8,818.16/bottle; corrected. Verify formulation-per-bottle wording ("300 mg/12 mL") against the DailyMed label.]
• WAC (current, mid-2026): Per the Colorado prescriber filing effective 01-Jan-2026, the 100 mg tablet package is ~$38,702.90 and the oral suspension ~$9,675.73/bottle. [FLAG: TEMPORAL — for a mid-2026 dossier use the current (Jan-2026) WAC, not the 2024 launch price; the $35,272.64 figure is stale as a "current price."]
• Annual cost: not published as a single headline; weight-based dosing makes per-patient annualized cost six-figure. [FLAG: no official annual-cost figure disclosed; do not state a single dollar amount without weight/dose assumptions and source.]

7.3 Launch uptake & net sales trajectory (U.S.)

[FLAG: the "+11% QoQ" revenue figure for Q1'25 in the original draft was not directly verified (Day One's disclosed Q1'25 QoQ metric was for prescriptions, +16%); confirm the revenue-QoQ figure against the Q1'25 8-K or drop it.] Revenue figures are U.S. net product revenue; ex-U.S. reaches Day One as Ipsen royalties, not product revenue.

7.4 Competitive positioning in pediatric BRAF-altered LGG

The clinically decisive axis is BRAF alteration type (fusion vs V600 point mutation): - BRAF fusions/rearrangements (e.g., KIAA1549::BRAF) — the most common BRAF alteration in pLGG. Type I RAF inhibitors (dabrafenib) are ineffective and can paradoxically activate MAPK in fusion-driven tumors, so dabrafenib+trametinib is not indicated here. Tovorafenib (type II) blocks fusion-kinase activity without paradoxical MAPK activation — its core differentiator; the fusion segment is effectively its open field. - BRAF V600E mutation: head-to-head with dabrafenib+trametinib, which holds first-line FDA approval (2023) for V600E pediatric LGG (Bouffet et al., NEJM 2023). Tovorafenib's label is narrower here — R/R, accelerated approval. [FLAG: verify the exact NEJM citation — the Bouffet dabrafenib+trametinib pediatric-glioma paper is NEJM 2023;389 but the start page is likely 1108, not 1118 as originally cited; confirm volume/page.] - MEK inhibitors (selumetinib / Koselugo): not FDA-approved for LGG (LGG use investigational; approved indication = NF1 plexiform neurofibromas); BRAF-agnostic. [FLAG: selumetinib LGG data is trial-stage, not label.] - Differentiation levers: (i) once-weekly oral dosing vs twice-daily multi-drug regimens; (ii) coverage of the fusion majority; (iii) greater CNS penetration than dabrafenib and reportedly less severe dermatologic/cardiac/ophthalmologic toxicity vs type I RAF / MEK inhibitors (per FDA summary). - Efficacy anchor (FIREFLY-1): ORR 67% (46/69) RANO-HGG; 51% (39/76) RAPNO (28 PR + 11 MR); DoR 16.6 mo (RANO-HGG) / 13.8 mo (RAPNO); Grade ≥3 TRAEs 42%. CNS response by RANO/RAPNO, not RECIST. - EU status: EC conditional MA (20 Apr 2026 per EMA / announced 22 Apr 2026 by Ipsen; all 27 EU + Iceland/Liechtenstein/Norway); positioned by Ipsen as the first targeted therapy in R/R pediatric LGG "regardless of BRAF alteration." [FLAG: "regardless of BRAF alteration" is Ipsen's framing (it means the label spans both fusion and V600, unlike dabrafenib+trametinib's V600-only label) — verify exact SmPC indication wording before external use.]

7.5 Intellectual property & exclusivity timeline

• Composition-of-matter: per Day One SEC filings, the tovorafenib COM patent protects the U.S. franchise to ~mid-2036 with PTE (est. ~5 years). Broader estate covers compositions, formulations, synthesis/intermediates, methods of treatment, combinations, potentially reaching into the 2040s. [FLAG: confirm the specific Orange Book patent numbers and any granted PTE date once OJEMDA appears in the Orange Book — the "mid-2036" figure is the company's disclosed estimate, not an adjudicated PTE grant.]
• In-licensed IP: non-exclusive license of one patent family from Takeda (catalyst usable in preparation); separate Viracta obligation (satisfied via PRV-sale carve-out).
• Regulatory exclusivities: FDA orphan (malignant glioma, Sept 2020) → 7-year orphan exclusivity from 23-Apr-2024 (~2031); EU orphan → up to 10-year EU market exclusivity; Rare Pediatric Disease Designation (July 2021) → PRV (monetized $108M); Breakthrough + Priority Review. [FLAG: verify the specific Sept-2020 / July-2021 designation dates against FDA designation records; not re-verified in this pass.]
• LOE read: effective U.S. exclusivity gated by the COM patent (~2036, PTE-dependent); formulation/method/dosing-regimen patents provide secondary barriers into the 2040s but are more vulnerable to design-around. Net LOE anchor ≈ mid-2030s (U.S.). [FLAG: any generic/505(b)(2) Paragraph-IV timeline is speculative until Orange Book listing is public.]

7.6 Bottom line for competitive planning

Tovorafenib owns the BRAF-fusion majority of pediatric LGG that type I inhibitors structurally cannot serve, backed by once-weekly oral convenience and a durable FIREFLY-1 signal; FY2025 U.S. net sales of $155.4M (+172%) and 2026 guidance of $225–250M show a fast-ramping launch that Servier's $2.5B acquisition validates. Key catalysts/risks: (i) FIREFLY-2 front-line readout (upside first-line label; risk chemo non-superiority), (ii) accelerated-approval confirmation obligations, (iii) PTE adjudication setting true U.S. LOE, (iv) integration/commercial execution under Servier.

8. Head-to-Head Comparison Framework (Shubin-slide axes)

SCOPE & GUARDRAILS FOR THIS SECTION. The tovorafenib column is built entirely from public sources (Kilburn et al., Nature Medicine 2024; OJEMDA U.S. PI 2024; FDA Approval Summary PMC11996598). The plixorafenib column reproduces ONLY the figures on Shubin's slide, labeled "Fore-internal (per slide) — verify." No plixorafenib efficacy/safety numbers are supplied; those cells are template placeholders for the Fore team to complete from cleared internal data. Every CNS response is RANO-HGG or RAPNO-LGG (never RECIST), with N and denominator.

[FLAG: DATA-CUT / TEMPORAL — verify before external use] The tovorafenib column reflects the 2024 registration data cut. As of mid-2026 a three-year FIREFLY-1 update was presented at SNO 2025 with higher response estimates (RAPNO-LGG ORR ~53%, RANO-HGG ORR ~71%) plus updated DoR/PFS and treatment-free-observation data. A mid-2026 competitive dossier must either (a) label every tovorafenib figure below with the registration cut date, or (b) refresh to the SNO 2025 cut. Do not present 2024 numbers as "current."

Block 1 — Baseline characteristics (LGG population)

Cohorts per slide: FIREFLY-1 Arm 1 (n=77) and Arm 2 (n=60) = tovorafenib (public); FORTE Sub A (n=30), Sub B (n=12), and PLX120-03 (n=7; 9 by BSA) = plixorafenib (Fore-internal/embargoed, per slide only).

Reconciliation note (tovorafenib, verify vs slide): The slide's Arm 1 values match the public record closely. FDA/Kilburn efficacy population (n=76, ≈Arm 1) reports optic-pathway 39 (51%), KIAA1549::BRAF 56 (74%), V600E 12 (16%), prior MEK 42 (55%) ³ — vs the slide's Arm 1 56 (73%), 13 (17%), 43 (56%). Small deltas reflect Arm 1 n=77 vs efficacy denominator n=76 and are within rounding. [FLAG: the slide's "≥3 lines 39 (51%)" for Arm 1 is NOT independently confirmed and is internally suspect — see the ≥3 row. The arm-2 full baseline table and the fine-grained location / prior-line distributions are not broken out in the primary Kilburn 2024 publication; combined-cohort figures (median age 9, range 1-24; optic 50%; KIAA1549 74%; V600E 16%; ≥1 prior BRAF/MEK 61%) are the published cross-check.]

Block 2 — Efficacy, safety, mechanism, dosing, label

All tovorafenib efficacy figures are the 2024 registration cut unless noted — see the data-cut flag above.

Unverifiable / flagged items (this block): chiefly the fine-grained Arm-2 baseline table and the per-arm tumor-location, prior-line, prior-radiation, and prior-surgery distributions, not itemized by arm in Kilburn 2024. New: (a) the "≥3 prior lines 39 (51%)" Arm-1 cell is internally contradictory and likely a transcription duplication of the optic value — do not use as cited; (b) the RANO-HGG DoR (16.6) vs PFS (19.4) assignment may be swapped, and RAPNO PFS is listed identically to RAPNO DoR (13.8 mo) — reconcile against the Kilburn 2024 tables; (c) the 67% RANO-HGG ORR assessor label ("investigator") should be corrected to blinded independent central review pending source check. The 380 vs 420 mg/m² dose distinction (approved vs studied) is verified and reconciled. All plixorafenib cells are template placeholders per embargo — no plixorafenib efficacy/safety was researched, inferred, or fabricated.

End of dossier. Every tovorafenib figure is public and cited to the consolidated references below; every plixorafenib cell is a Fore-internal, per-slide, unverified template for the Fore team to complete.

9. Development Trajectory Beyond FIREFLY

9.1 What "CONNECT TarGeT-B" actually is

The study circulating in competitive briefings as a "Day One high-grade glioma program" is real and registered, but its sponsorship structure — not its age window — is the informative signal. The registry acronym is literally CONNECT TarGeT-B, registered as NCT07206849, official title "A Phase 2 Study of Tovorafenib in Pediatric and Young Adult Patients Newly Diagnosed With High-Grade Glioma (HGG), Including Diffuse Intrinsic Pontine Glioma (DIPG), Which Harbor Alterations in the Mitogen-Activated Protein Kinase (MAPK) Pathway." Status is NOT_YET_RECRUITING; estimated start May 2026; primary completion May 2030; study completion May 2037; estimated enrollment N=79. First submitted 18 Sep 2025, last update posted 8 Apr 2026 (source: ClinicalTrials.gov registry record NCT07206849, CTG API v2, last update posted 2026-04-08).

The lead sponsor is Nationwide Children's Hospital. Day One Biopharmaceuticals is listed only as an INDUSTRY collaborator. The record carries secondary ID R01FD008167 (FDA) — the R01FD prefix is the FDA Orphan Products Grants Program series (source: NCT07206849, sponsor/collaborator and secondary-ID modules). This is an investigator-initiated consortium study running on federal orphan-products grant money with an industry collaborator attached, not a company-sponsored registrational program.

[FLAG — CONTRADICTION 42: The received framing describes this as "a Day One or Servier study." Both halves are wrong. The registry shows lead sponsor Nationwide Children's Hospital with Day One as collaborator only, and shows no Servier involvement with tovorafenib anywhere in the registry. The ex-US partner sponsoring new tovorafenib trials is Ipsen (NCT07441707, Japan), not Servier. Note this is distinct from the corporate fact in Section 13 that Servier acquired Day One on 23 Apr 2026 — the trial-sponsorship record and the corporate-ownership record are different objects and must not be merged. — Confirm by re-querying ClinicalTrials.gov for sponsor="Servier" AND intervention="tovorafenib" (expected: zero records) and reading the Ipsen license announcement for the ex-US development scope.]

9.2 Design: newly diagnosed, post-radiotherapy, three strata, one primary

The setting is newly diagnosed post-radiotherapy maintenance, not relapsed/refractory. Detailed description: "patients with newly-diagnosed HGG and DIPG harboring alterations in the MAPK pathway will be treated with tovorafenib following initial standard-of-care treatment with radiotherapy (RT)." Prior therapy language: "Surgery, radiation (RT), and/or dexamethasone are permissible… No other prior anticancer therapy for HGG will be allowed." RT dosing is mandated: 54 Gy DIPG; 54–59.4 Gy other HGG; 45–54 Gy spinal; 36–39.6 Gy craniospinal with boost for metastatic disease (source: NCT07206849 detailed description and eligibility criteria).

The genomic net is materially wider than BRAF (source: NCT07206849 arms/eligibility modules):

• Stratum A — localized intracranial, non-pontine, non-thalamic HGG with BRAF V600 mutation. This is the only BRAF V600-restricted stratum, and it is the only stratum carrying the primary endpoint.
• Stratum B — DIPG/DMG, or localized HGG with MAPK alterations excluding BRAF V600, or primary spinal tumours.
• Stratum C — metastatic HGG with any MAPK pathway alteration.
• Strata B and C additionally accept BRAF V600, KIAA1549:BRAF fusion, KRAS/NRAS, CRAF/RAF1, other RAF mutations, or FGFR alterations.

The primary endpoint is single-arm PFS in Stratum A only, against historical controls — externally controlled, no randomised comparator. Per-stratum sample sizes are not published; on a three-stratum split of N=79 with the primary confined to the narrowest genotype, Stratum A is estimated at 20–35 patients. Method: design inference from the N=79 total, the three-stratum structure, and the observation that BRAF V600 is the rarest of the eligible alterations in HGG; this is a range, not a sourced figure. A historical-control PFS comparison in newly diagnosed HGG after RT is confounded by pseudoprogression, and the registry record names no response criteria by which pseudoprogression would be adjudicated.

9.3 The age window, and why it is not an adult signal

TarGeT-B eligibility is "≥12 months and ≤39 years of age at the time of enrollment." Against the two FIREFLY trials: FIREFLY-1 (NCT04775485) minimum 6 Months, maximum 25 Years; FIREFLY-2/LOGGIC (NCT05566795) maximum "Less than 25 years of age" (source: NCT07206849, NCT04775485, NCT05566795 eligibility modules). The ceiling therefore extends 14 years beyond both FIREFLY trials — while the floor rises, from 6 months to 12 months. The expansion is entirely upward into the AYA band, with simultaneous contraction at the infant end, which is the opposite of what a company widening a paediatric label would do.

The 39-year ceiling should be discounted as an adult-market signal. It mirrors the envelope already used in PNOC029 (craniopharyngioma), the standard AYA convention of paediatric neuro-oncology consortia. No tovorafenib study anywhere in the registry enrols adults above 39 or targets adult glioma; the only uncapped-age tovorafenib studies are terminated melanoma trials and an NCI hairy cell leukaemia study. TarGeT-B is the age-widest CNS study in the programme and it stops at 39. It is also the only tovorafenib HGG/DIPG study in existence — all other active CNS work is LGG or craniopharyngioma.

[FLAG — CONTRADICTION 43: FIREFLY-2 is commonly described as enrolling from 6 months. The registry record for NCT05566795 returns a NULL minimum-age field, with only "Less than 25 years of age" as a stated bound. The 6-month floor likely derives from protocol or publication text. Do not cite 6 months for FIREFLY-2 to a registry source. — Confirm from the LOGGIC/FIREFLY-2 protocol publication or the EU CTIS record for the same study.]

[FLAG — CONTRADICTION 44: Response-criteria discipline is inconsistent across the Day One-associated portfolio. FIREFLY-2 explicitly specifies RAPNO (IRC-assessed primary ORR) with RANO for measurable-lesion definition. TarGeT-B — an HGG study, where RANO-HGG would be the expected standard — specifies no named criteria at all: not RANO-HGG, not RAPNO, not iRANO, not RECIST. — Confirm from the protocol synopsis; see Flag 45.]

9.4 The competitive gap this defines

Three structural features of TarGeT-B bound what tovorafenib can claim in HGG. First, it is newly-diagnosed post-RT maintenance, so it says nothing about activity in recurrent disease and leaves the relapsed/refractory HGG question entirely open. Second, the primary rests on ~20–35 patients (Stratum A, method above) against historical controls, capping the evidentiary weight of any readout. Third, primary completion is May 2030 and study completion 2037: there is no tovorafenib HGG evidence of any kind before 2030.

The occupied and unoccupied territory therefore separates cleanly, and the head-to-head table in Section 8 should carry these four rows:

• Adult HGG (age >39) — tovorafenib: no trial, no registry footprint, uncontested through at least 2030. Plixorafenib: Fore-internal — to be completed from cleared data.
• Recurrent / refractory HGG — tovorafenib: not studied; TarGeT-B is newly-diagnosed only. Plixorafenib: Fore-internal — to be completed from cleared data.
• Prior-BRAF-inhibitor-exposed patients — tovorafenib: no dedicated stratum or subgroup in any registered study. Plixorafenib: Fore-internal — to be completed from cleared data.
• CNS-penetrant activity under RANO-HGG in measurable enhancing disease — tovorafenib: no HGG-criteria-assessed dataset exists or is scheduled before 2030. Plixorafenib: Fore-internal — to be completed from cleared data (ORR, assessment criteria, and CNS PK).

[FLAG 45: The formal response-assessment criteria for NCT07206849 (RANO-HGG vs RAPNO-HGG vs iRANO) are not stated in the registry record — confirm by requesting the protocol synopsis or statistical analysis plan from the CONNECT consortium coordinating centre at Nationwide Children's (listed contact Kelsey H Troyer PhD, 614-722-3284), or by watching for a protocol-design abstract at SNO 2026 / ISPNO.]

[FLAG 46: Per-stratum sample sizes for Strata A/B/C within the N=79 total are unpublished, so the 20–35 estimate for Stratum A is design-inferred, not sourced — confirm from the protocol synopsis or the FDA Orphan Products Grant R01FD008167 award abstract, searchable in NIH RePORTER and the FDA OOPD grants listing.]

[FLAG 47: The site list for NCT07206849 is empty in the registry (not yet recruiting), so geographic footprint and the number of participating CONNECT member institutions cannot be established — confirm after the record posts locations at activation, or from CONNECT consortium membership listings.]

[FLAG 48: Day One's financial and operational commitment to TarGeT-B (drug supply only vs funded study) is not disclosed in the registry — confirm from Day One 10-K/10-Q R&D pipeline discussion or collaboration disclosures in the most recent SEC annual filing, which should indicate whether HGG appears as a company-sponsored programme.]

[FLAG 49: FIREFLY-2's 6-month minimum age is not confirmable from the registry record — confirm from the LOGGIC/FIREFLY-2 protocol publication or the EU CTIS record.]

[FLAG 50: Whether the approved label text restricts to paediatric LGG with a specific age floor, versus what these trials enrol, is addressed in Section 10 but was not established in the trajectory pass — LABEL vs TRIAL vs OFF-LABEL must be kept separate throughout. Confirm by reading the current OJEMDA US PI Sections 1 and 8.4 and the EMA product information directly.]

Bottom line for Fore. Read the sponsorship structure before the age window. Day One's own capital remains entirely in LGG — FIREFLY-1 and the 418-patient FIREFLY-2 — and Ipsen is funding a Japanese LGG bridging study, not HGG. HGG is being explored on someone else's federal grant, which is option-buying rather than label-building, and option-buying is cheap to abandon. The practical consequence is that Fore's HGG and adult positioning is not contested by data before 2030, and Fore should say so with the registry citation rather than hedging. The second, sharper consequence is methodological: any cross-trial ORR comparison between a RAPNO-LGG-assessed tovorafenib figure and a RANO-HGG-assessed plixorafenib figure is invalid on its face, and TarGeT-B's silence on criteria means even the eventual 2030 HGG comparison may not be like-for-like. That gives Fore a documented, principled reason to decline the naive cross-trial comparison a reviewer will inevitably request — a refusal that reads as rigour rather than evasion only if the criteria argument is made first, unprompted, and in writing.

10. Label Scope, Age Boundaries and Off-Label Reality

10.1 The label has no upper age limit — in either geography

The widely repeated claim that OJEMDA is "limited to patients up to 25" is factually wrong and should be removed from Fore materials. The US indication reads verbatim: "OJEMDA is indicated for the treatment of patients 6 months of age and older with relapsed or refractory pediatric low-grade glioma (LGG) harboring a BRAF fusion or rearrangement, or BRAF V600 mutation. This indication is approved under accelerated approval based on response rate and duration of response… Continued approval for this indication may be contingent upon verification and description of clinical benefit in a confirmatory trial(s)." There is no age ceiling and no "pediatric patients" qualifier on the word "patients" (source: OJEMDA US Prescribing Information, NDA 218033, DailyMed SetID ea3a9631-3a66-6a7c-e053-2995a90ae2ad, label version 2 September 2025, Section 1). The EU indication likewise carries a lower bound only.

Three further structural facts reinforce that the ceiling question was treated as inapplicable rather than as restricted: neither label carries a Geriatric Use section; dosing is body-surface-area-based with a flat cap rather than age-banded, so the dosing algorithm functions unchanged in an adult; and the US label itself contains adult exposure data — 32 adults dosed at flat 600 mg — meaning an adult regimen is a characterised regimen already inside the approved document, not an off-label inference (source: OJEMDA US PI, Sections 2, 8 and 12).

10.2 Where 25 actually comes from

Twenty-five years is trial eligibility, and it appears in both the pivotal and the confirmatory trial, which is why it migrates into secondary literature as though it were label text. FIREFLY-1 (NCT04775485): minimumAge "6 Months", maximumAge "25 Years". LOGGIC/FIREFLY-2 (NCT05566795, Phase 3 frontline, active not recruiting): maximumAge "25 Years" (source: ClinicalTrials.gov, protocolSection eligibilityModule, records retrieved 20 July 2026). The number "25" appears nowhere in the US indication and appears in EU SmPC 5.1 only as a description of who was enrolled: "Eligible patients (n=76), from 6 months to 25 years of age…"

[FLAG — CONTRADICTION 51: Response criteria conflict across four primary sources for the same N=76 dataset. (a) FDA PI Section 14: ORR "by blinded independent central review based on RAPNO-LGG… criteria", with RANO-LGG (2011) also named elsewhere in the label. (b) EU SmPC 5.1: "independent review based on RANO-LGG (2011) criteria." (c) ClinicalTrials.gov NCT04775485 Arm 1 primary outcome names RANO-HGG — a high-grade criterion — for a low-grade trial. (d) NCT05566795 names RAPNO. This is not a transcription nit: RAPNO-LGG admits a minor response category (MR, ≥25% to <50% reduction) that RANO-HGG does not, and the FDA ORR definition explicitly includes MR ("complete response (CR), partial response (PR), or minor response (MR)"). Any ORR comparison against a figure computed without an MR category is not like-for-like. — Confirm by reading PI Section 14, SmPC 5.1 and both registry outcome modules side by side and tabulating criterion against reported figure; never report a tovorafenib ORR without naming the criterion that generated it.]

[FLAG — CONTRADICTION 52: The US and EU indications diverge on line of therapy. US: "relapsed or refractory pediatric low-grade glioma" — no explicit prior-therapy count. EU: patients "who have progressed after one or more prior systemic therapies" — an explicit ≥1 prior-systemic-therapy gate. The EU label is the narrower. A single global positioning slide treating the two as one will misstate the EU line of therapy. — Confirm by reading US PI Section 1 against Ojemda EU SmPC Section 4.1 verbatim.]

[FLAG — CONTRADICTION 53: Eligibility ceiling was 25 years, but the actual efficacy-population range was 2–21 years, with only 6/76 (7.9%) patients older than 16 and zero over 21 (source: SmPC 5.1 / PI Section 14). Sources citing "6 months to 25 years" as the studied population are quoting the protocol, not the data. — Confirm from PI Section 14 demographics table.]

10.3 "Pediatric low-grade glioma" is a disease term, not an age term

The indication conditions on three things: (i) age ≥6 months, (ii) the disease entity "pediatric low-grade glioma", (iii) a molecular biomarker. Age appears exactly once, as a floor; the other two clauses are nosologic and molecular. The molecular breakdown of the N=76 efficacy set is KIAA1549:BRAF fusion 74% (56/76), BRAF V600E 16% (12/76), other BRAF duplication/rearrangement 11% (8/76) (source: OJEMDA US PI Section 14). KIAA1549:BRAF-fused pilocytic astrocytoma is classified in WHO CNS5 under circumscribed astrocytic gliomas, an entity that is not age-defined and does occur in adults (source: WHO Classification of Tumours of the Central Nervous System, 5th ed., 2021). Method: textual analysis of the indication's three conditioning clauses — the absence of any ceiling term combined with the presence of a nosologic term supports reading the restriction as disease-entity rather than age; WHO CNS5 is cited only to establish that the histology is not age-exclusive. This is an inferred, not a sourced, reading.

10.4 Practice: adults are already being treated, at documented N=8

The only peer-reviewed adult tovorafenib CNS experience located is a Mayo Clinic series of 8 adults, ages 21–66. Its methodology is weak in a way that matters. The abstract's methods section reads, in full, "METHODS: Chart review." The outcome phrasing is "Radiographic response to therapy was seen in 5/8 patients with 3 cases of radiographic stability… and 2 cases of decrease in tumor size" — so the 5/8 conflates stability with shrinkage, and true tumour-size reduction is 2/8 (25%), with one of the three "stable" patients later progressing. Three of eight required pause or discontinuation for toxicity. No response criterion is named — not RANO-HGG, not RAPNO-LGG, not iRANO, not RECIST — and assessment is investigator chart review. Two of the eight patients had Erdheim-Chester disease, a histiocytosis, not a glioma at all (source: Williams et al., Abstract PATH-94, Neuro-Oncology 2025;27(Suppl 5):v263). Beyond this series, no additional published adult tovorafenib CNS cohort was located; true off-label adult exposure is certainly larger but is unquantified in the public record.

Section 12 documents a second, independent adult series (Johns Hopkins, N=7) with a severe safety signal; the two series together constitute the entire published adult experience at N=15.

[FLAG 54: Whether US payer compendia — NCCN Drugs & Biologics Compendium, Lexi-Drugs, AHFS-DI, Micromedex DrugDex — list tovorafenib with an age restriction or with any adult off-label recognition. Compendia listing, not the label, determines CMS and commercial coverage for off-label use under SSA §1861(t)(2)(B). — Confirm by pulling the current NCCN Drugs & Biologics Compendium entry (subscription, NCCN.org) and the NCCN Central Nervous System Cancers guideline current version, checking whether tovorafenib appears in any adult algorithm page (e.g. adult pilocytic astrocytoma / circumscribed glioma) and what Category of Evidence is assigned; cross-check the Micromedex DrugDex off-label rating for adult BRAF-altered LGG.]

[FLAG 55: Whether any payer has issued a written medical policy with an explicit age edit (e.g. "age ≤25") on tovorafenib, versus a diagnosis edit only — confirm by retrieving published commercial medical policies from UnitedHealthcare, Aetna Clinical Policy Bulletins, Cigna, and 3–5 state Medicaid PDLs, all of which post policy PDFs publicly, and reading the coverage-criteria section for age language versus ICD-10 diagnosis language.]

[FLAG 56: Whether any adult-inclusive tovorafenib trial exists or is planned — load-bearing for how durable the adult white space is. Confirm by querying ClinicalTrials.gov and EU CTIS for all interventional studies of tovorafenib / DAY101 / TAK-580 / MLN2480 with no maximumAge or maximumAge >25, and reading Day One's most recent Form 10-K Item 1 and 10-Q pipeline tables plus the Ipsen ex-US license announcement for stated adult development intent.]

[FLAG 57: Whether the SNO 2025 Mayo abstract (PATH-94) has since been published as a full peer-reviewed manuscript with named response criteria and per-patient volumetrics, which would materially change its evidentiary weight — confirm by a PubMed author search on Williams A / Sener U / Ruff M with "tovorafenib" filtered 2025–2026, and by checking the Mayo Clinic neuro-oncology publication list.]

[FLAG 58: Whether tovorafenib's EU conditional marketing authorisation carries a Paediatric Investigation Plan or Annex II specific obligation that functionally constrains use to the paediatric setting even though SmPC 4.1 sets no ceiling — confirm by reading Annex II of the Ojemda EPAR Product Information (specific obligations / conditions of the marketing authorisation) and the CHMP assessment report's benefit-risk section.]

[FLAG 59: Plixorafenib comparator cells on this axis — Fore's trial age eligibility, enrolled adult age range and N, and any adult CNS efficacy by RANO-HGG/iRANO. Fore-internal — to be completed from cleared data.]

Bottom line for Fore. Stop asserting the 25-year ceiling; it will not survive a medical lead opening the label, and asserting it costs credibility for everything adjacent to it in this dossier. The defensible framing is three-layer and is stronger than the false one: the LABEL permits adults with no ceiling in either geography; the EVIDENCE stops at 21 years with 6/76 patients over 16 and zero over 21; PRACTICE is already treating adults off-label at a documented N=8 with 2/8 shrinkage and 3/8 toxicity-driven interruption. That structure points two ways at once, and both matter. Defensively, the real barrier to adult tovorafenib is not the label — it is the absence of adult evidence and the resulting compendia and payer friction, a barrier Servier/Ipsen can dissolve with a single adult study they have every incentive to run. Fore should not model the adult segment as structurally protected. Offensively, the gap between an unbounded label and an evidence base that stops at 21 is precisely the space a rigorous adult CNS programme occupies, and the two published off-label series are unpaid market validation that clinicians already want to prescribe into it. The sharpest wedge is methodological rather than clinical: tovorafenib's approval is a single-arm accelerated approval whose ORR includes an MR category, named inconsistently across the FDA label, the EU SmPC and the registry. Fore's operative decision from this section is a drafting one — pursue indication text scoped to a BRAF-altered tumour entity rather than an age band, because that single choice is worth more addressable population than any downstream guideline footnote.

11. Endpoint Strategy — What an ORR Primary Buys and Costs

11.1 The central regulatory fact: FDA declined the trial's own primary endpoint

FIREFLY-1 prespecified ORR by RANO-HGG and met its bar at 67% (Arm 1, n=77), with median DoR 16.6 months and median time to response 3.0 months. The identical patients scored 51% by RAPNO (including minor responses), median DoR 13.8 months, median TTR 5.3 months; safety population n=137 (source: Kilburn LB et al., Nat Med 2024 Jan;30(1):207-217, PMID 37978284, DOI 10.1038/s41591-023-02668-y). FDA then declined to carry the 67% figure into the OJEMDA label at all and re-designated the trial's secondary endpoint (RAPNO) as the "major efficacy outcome measure" for regulatory purposes. This is the single most consequential regulatory fact on this axis.

Two refinements matter. First, within the label itself the cross-criterion gap nearly vanishes: RAPNO-LGG and RANO-LGG give essentially the same total answer on the same 76 patients — 53% vs 54% — but redistribute depth (29 PR / 11 MR vs 23 PR / 18 MR). The 16-point swing is specific to applying an adult high-grade criterion to a low-grade paediatric population. Second, the MR category is the operative lever: it exists in RAPNO-LGG and RANO-LGG, is absent from RANO-HGG and RECIST, and accounted for 11 of 40 responders — 14 of the 53 labeled percentage points, over a quarter of the labeled ORR.

[FLAG — CONTRADICTION 60: The commonly circulated formulation "ORR 67% by RANO-HGG versus 51% by RAPNO-LGG per the FDA label" is partly wrong and must not enter this dossier in that form. The 67%/51% pair is correct and is same-patient, but it comes from Kilburn Nat Med 2024 (n=77, Arm 1), not from the FDA label. The label (rev 8/2025) never prints 67%; it prints 53% (RAPNO-LGG) and 54% (RANO-LGG) on n=76. Attributing 67% to the label is a citable error a Day One / Servier medical-affairs reviewer could exploit. — Confirm by text-searching the PI PDF for "67".]

[FLAG — CONTRADICTION 61: ORR 51% vs 53% for the same trial. FDA's approval summary (Clin Cancer Res 2025;31(8):1383-89) reports 51% (95% CI 40–63) with median DoR 13.8 months; the current label (rev 8/2025) reports 53% (95% CI 41, 64) with median DoR 18 months. Both cite an efficacy population of 76. The most probable reconciliation is a later data cutoff carried into the 8/2025 revision, but neither document states this. Do not present 51% and 53%, or 13.8 and 18 months, as interchangeable. — Confirm by comparing the data-cutoff dates stated in the approval summary methods and in PI Section 14.]

[FLAG — CONTRADICTION 62: Denominator conflict in the dabrafenib control arm. NEJM 2023 reports 110 randomised 2:1 as 73 vs 37, ORR 11% (4/37). The TAFINLAR label text retrieved states 73 versus 33 for carboplatin+vincristine. Likely a randomised-vs-efficacy-population distinction, but unresolved from the documents read. — Confirm from TAFINLAR PI Section 14.7 (see Flag 68).]

[FLAG — CONTRADICTION 64: Kilburn reports FIREFLY-1 Arm 1 as n=77 while the FDA efficacy population is n=76, the label footnote restricting to patients with ≥1 measurable lesion by RAPNO-LGG at baseline. Do not use 77 and 76 interchangeably. — Confirm from PI Section 14 footnote.]

11.2 Day One has since conceded the criterion fight

The January 2024 FIREFLY-2 design paper (BMC Cancer 2024;24:147, PMID 38291372) states the primary endpoint is ORR per RANO-LGG. The current ClinicalTrials.gov record for NCT05566795 (last updated 2026-05-12) states the primary is ORR per RAPNO, with RANO-HGG and RANO-LGG demoted to secondary. The switch tracks the April 2024 accelerated approval: Day One realigned its confirmatory primary to the criterion FDA actually labels on. Note also that the original stated rationale for RANO-LGG was explicitly to copy the dabrafenib registrational trial — the endpoint choice was precedent-driven rather than science-driven, and was therefore cheap to abandon. Registry enrollment reads 418 actual against a design-paper target of ~400; cite 418.

[FLAG — CONTRADICTION 63: BMC Cancer 2024 (RANO-LGG) and ClinicalTrials.gov (RAPNO) cannot both describe the same protocol version — one supersedes the other. Treat the registry as current and the design paper as historical, and treat the change itself as the finding. — Confirm via the endpoint-switch date, Flag 66.]

11.3 The precedent that matters: ORR alone did not do the work

The only paediatric glioma precedent producing a traditional (not accelerated) approval on a randomised ORR primary is dabrafenib+trametinib. Its paediatric LGG indication carries no accelerated-approval statement — "in combination with trametinib, for the treatment of pediatric patients 1 year of age and older with low-grade glioma (LGG) with a BRAF V600E mutation who require systemic therapy" — whereas the same label's tumour-agnostic solid-tumour indication (1.6) is explicitly "approved under accelerated approval based on overall response rate (ORR) and duration of response (DoR)" (source: TAFINLAR US PI, Sections 1 and 14.7, DailyMed setid fee1e6b1-e1a5-4254-9f2e-a70e0f8dbdea). Critically, that approval carried ORR 47% (34/73) vs 11%, and PFS 20.1 vs 7.4 months, HR 0.31 (NEJM 2023;389(12):1108-1120). The ORR delta did not carry the conversion by itself; a large, concordant, independently-assessed time-to-event effect did. That is the evidentiary template.

The competitive consequence of the label/trial distinction is that in BRAF V600E pLGG, dabrafenib+trametinib already holds the frontline traditional-approval label ("who require systemic therapy", no prior-therapy requirement, age ≥1 year), while OJEMDA is confined to relapsed/refractory under accelerated approval with thin V600 evidence (ORR 50% on n=12). FIREFLY-2 is therefore an expansion play into occupied frontline territory, not merely a confirmation exercise.

11.4 Cross-trial ORR benchmarking in pLGG is indefensible — which is the point

Published MAPK-pathway ORRs in pLGG span roughly 7% to 67% depending on agent, line, genotype and criterion: selumetinib 7.1%–26.9% across PBTC strata with 2-year PFS 57.1%–74.8% (Fangusaro J et al., Neuro Oncol 2025;27(9):2415-2428, PMID 40241281); BRAF-inhibitor monotherapy in V600E pLGG at a single centre 40% ORR with 90% 2-year PFS (McThenia SS et al., Front Oncol 2024;14:1505951, PMID 39839763); dabrafenib+trametinib 47% (34/73) by RANO; tovorafenib 67% RANO-HGG / 51–53% RAPNO (n=76–77). Anyone constructing a league table across these has produced an artefact of criterion selection.

11.5 What ORR cannot buy: durability, rebound, and function

ORR is measured on-treatment; rebound occurs off-treatment. Day One concedes the durability problem in its own design paper — "responses in pLGGs are often only durable for as long as the drug can be administered" — and the International pLGG Coalition has formally defined the rebound phenomenon. A chemotherapy-controlled ORR delta cannot address it. Meanwhile OS is structurally uninterpretable in FIREFLY-2 for two independent reasons: an indolent natural history (~90% 20-year OS) and protocol-mandated crossover. That is why OS sits fourth among roughly twenty secondaries. What the field accepts instead is a functional battery — FIREFLY-2 elevates Vineland-III adaptive behaviour, visual acuity, visual PFS and PROMIS HRQoL to formal secondaries. These are instrumented, not powered.

Three publicly defensible lines of attack on an ORR-primary FIREFLY-2 result follow, and none of them is "ORR is the wrong endpoint" — regulators have accepted ORR in this disease. They are: (a) the primary criterion, RAPNO-LGG, is one its own authors state "will need to be validated in prospective clinical trials," and which Day One's own design paper originally slotted as a secondary for that validation purpose before promoting it to primary — Day One is now measuring its pivotal effect with an instrument it described as still under validation; (b) reader/measurement variability under RAPNO is documented by the RAPNO authors themselves and is worst in the non-enhancing disease that dominates pLGG, which is exactly the regime where a 15-point delta is being measured; and (c) the trial is frontline and RAF-broad while the accelerated approval it must confirm is relapsed/refractory and BRAF-narrow — a confirmatory trial in a different line of therapy and a wider molecular population does not cleanly verify the benefit that was granted.

Every axis above requiring plixorafenib data — observed ORR by any criterion, DoR, PFS, off-treatment rebound rate, functional outcomes, and the CNS-penetration comparison — is Fore-internal, to be completed from cleared data. Nothing in this section infers or estimates a plixorafenib figure.

[FLAG 65: FDA's actual internal reasoning for setting aside RANO-HGG as the labeling criterion is unverified — the approval summary states what criterion was used but not why the prespecified primary was displaced. Confirm by retrieving the FDA Multi-Discipline Review / Clinical Review for NDA 218033 from Drugs@FDA, which contains the review division's endpoint discussion; the Drugs@FDA overview page 404'd on direct fetch and needs manual navigation.]

[FLAG 66: The exact date and stated reason for the FIREFLY-2 primary-endpoint switch from RANO-LGG to RAPNO is unverified — the ClinicalTrials.gov version-history API endpoint returned 404. Confirm by opening the NCT05566795 "Study Record Versions" tab and diffing the primary outcome field across versions to bracket the change against the 23 April 2024 accelerated approval date.]

[FLAG 67: Whether FDA has formally agreed that a frontline, RAF-broad trial can verify clinical benefit for a relapsed/refractory, BRAF-narrow accelerated approval is unverified — confirm via the postmarketing requirement text itself, obtainable from the FDA approval letter for NDA 218033 or FDA's Postmarketing Requirements and Commitments database, which states the PMR's exact wording and due date.]

[FLAG 68: The labeled (as opposed to published) efficacy figures for dabrafenib+trametinib Study CDRB436G2201 could not be retrieved — the DailyMed fetch returned Section 14.7's heading but not its table. Confirm by downloading the TAFINLAR PI PDF from DailyMed setid fee1e6b1-e1a5-4254-9f2e-a70e0f8dbdea and reading Section 14.7 directly; label-vs-publication divergence is exactly the pattern documented for tovorafenib.]

[FLAG 69: No SIOPe LOGGIC consortium position paper on response-endpoint selection was located, and no EMA-side document (EPAR, PIP, or CHMP scientific advice) on tovorafenib endpoints was retrieved — the research session exhausted its search budget, limiting discovery to direct fetches of known URLs. Confirm by searching the EMA medicines database for the OJEMDA EPAR, the SIOPe Brain Tumour Group publications page, and the RANO 2009–2025 progress report (Neuro Oncol 2025;27(9):2209-2224, PMID 40317212) full text for its pLGG endpoint section.]

[FLAG 70: Whether FDA has ever converted an accelerated approval to traditional approval on a randomised ORR primary alone, absent a supportive time-to-event endpoint, is unverified — the FDA guidance "Clinical Trial Considerations to Support Accelerated Approval of Oncology Therapeutics" 404'd at both attempted URLs. Confirm by retrieving the final guidance from FDA's guidance-document search and cross-checking FDA's Project Confirm accelerated-approval conversion database for oncology precedents.]

Bottom line for Fore. Choose the criterion you will be labeled on, not the one that maximises the headline — that is the entire lesson of this file. Day One prespecified RANO-HGG, hit 67%, got a press release, and then watched FDA label 51% by RAPNO and never print 67% anywhere; it has since abandoned RANO-LGG for RAPNO in its own confirmatory trial. Fore should prespecify the settled criterion for its target population from the start — RAPNO-LGG, BICR-adjudicated, for low-grade or paediatric-adjacent disease; RANO-HGG or RANO 2.0 for high-grade or adult — and under no circumstances score a low-grade population with a high-grade criterion. Second, prespecify and publish the cross-criterion concordance analysis, and decide the MR question explicitly and up front; an MR inclusion discovered later by an adversarial reader reads as gaming, and note that ORR is criterion-robust while PR-depth and CR+PR-only definitions are not. Third, do not let ORR carry the conversion alone: allocate alpha to a properly powered PFS/EFS endpoint and prospectively instrument off-treatment follow-up and re-treatment outcomes, because rebound is where an ORR-primary competitor is structurally undefended. Fourth and most opportunistic, function is unclaimed high ground — Day One instruments vision and neurodevelopment but does not power them, and in a disease whose harm is morbidity rather than death, a powered functional endpoint is the one place a guideline committee will readily accept that a competitor's ORR advantage did not translate. Finally, plan against a successful FIREFLY-2 primary: it is powered against a 30% control-ORR assumption versus an 11% randomised benchmark, roughly 3x conservative, so it is very likely to read out positive. Compete on durability, function and genotype breadth — not on the ORR number.

12. Guideline and Compendia Positioning

12.1 The expected story is not supported by the public record

The anticipated narrative for this axis — "NCCN listing lets tovorafenib reach adults beyond its paediatric label" — has no verifiable support in the peer-reviewed or publicly retrievable record. NCCN Pediatric CNS Cancers, Version 2.2025 — the current published version — covers only diffuse high-grade glioma and medulloblastoma. It does not cover low-grade glioma in its published scope, so there is no verifiable NCCN tovorafenib algorithm, recommendation wording, or Category of Evidence and Consensus in the public record. The adult NCCN CNS Cancers guideline likewise carries no verifiable BRAF-altered low-grade glioma or tovorafenib content. No SIOPe, SNO or EANO guideline addressing tovorafenib or BRAF-altered paediatric-type low-grade glioma was retrievable.

The absence is the finding. Tovorafenib's route into adult patients today runs through permissive label text — "6 months of age and older", no upper bound, with "pediatric" attaching to the tumour entity (Section 10) — not through guideline or compendium expansion. Day One did not need NCCN to open that door; FDA left it open.

12.2 The criteria fragility, restated for guideline committees

Three regulatory-grade documents name three different response criteria for one drug. The ClinicalTrials.gov record for NCT04775485 names RANO-HGG as the primary outcome for the low-grade glioma arms; the FDA label and approval summary report 51% (95% CI 40–63), N=76, by RAPNO (Clin Cancer Res 2025;31(8):1383-1389, PMID 39808502); the Phase 3 LOGGIC/FIREFLY-2 protocol paper specifies RANO-LGG (BMC Cancer 2024;24:147, PMID 38291372). Within a single FIREFLY-1 cohort of 35 patients, the criteria choice alone moves ORR across 64% / 55% / 50% (RANO-HGG / RANO-LGG / RAPNO) — approximately 14 percentage points in the same patients. Using a high-grade criteria set as the registered primary endpoint for a low-grade population is internally inconsistent on its face and is a legitimate point to put to a guideline committee.

Note the discipline this imposes on Fore in both directions. The widely circulated 67% is the RANO-HGG figure and is not the label figure; press and investor materials citing 67% are not citing the label, and Fore should not let it enter a comparison table unchallenged. Equally, Section 13 establishes that the label itself is criteria-appropriate — RAPNO-LGG with RANO-LGG as secondary — so an attack framed as "tovorafenib used the wrong criteria in its label" is wrong on the facts. The correct attack is narrower and survives scrutiny: the registered primary endpoint used a high-grade criterion, the resulting number is the one in circulation, and it is not the number the regulator accepted.

12.3 Adult off-label use is published — and the first safety signal is severe

A Johns Hopkins report of adult off-label tovorafenib describes 2 of 7 adults with grade 4 intratumoral haemorrhage, with limited HGG efficacy and a median duration of approximately 8 weeks (source: Neuro-Oncol Pract 2025;12(6):1051-1057, PMID 41458920). This is the first published adult safety signal for the asset, and it sits alongside the Mayo series (N=8, Section 10). Together the entire published adult experience is N=15. The tovorafenib evidence base contains zero patients over age 25 from either FIREFLY trial — a hard gap between the label's open-ended age text and the data supporting it.

[FLAG 71: Whether the full online NCCN Pediatric CNS Cancers guideline (current version) contains a low-grade glioma section listing tovorafenib, and if so its exact recommendation wording, line of therapy, and Category of Evidence and Consensus (2A vs 2B) — nccn.org returns HTTP 403 to all automated access. A human should register a free NCCN account at nccn.org/guidelines, open "Pediatric Central Nervous System Cancers" (recording the version number and date on the title page, as these update several times a year), navigate to the PEDCNS low-grade glioma pages, and text-search "tovorafenib". Repeat for the adult "CNS Cancers" guideline.]

[FLAG 72: Whether tovorafenib has an NCCN Drugs & Biologics Compendium listing, and whether any listed use extends beyond the paediatric label — the single item that most directly drives US payer coverage of off-label adult use. Requires a paid subscription; confirm at nccn.org/compendia-templates/compendia/nccn-drugs-biologics-compendium by searching the tovorafenib monograph and reading the Disease/Setting column and its Category.]

[FLAG 73: Whether Day One or Ipsen has publicly claimed NCCN listing in an SEC filing or investor deck — EDGAR full-text search returned HTTP 403 to automated access because it requires a declared User-Agent header. A human should run EDGAR full-text search for "tovorafenib" AND "NCCN" restricted to 10-K and 10-Q, and read the Commercial/Competition section of Day One's most recent 10-K directly.]

[FLAG 74: Whether SIOPe has issued a European paediatric LGG treatment recommendation addressing BRAF-targeted therapy — not indexed in PubMed under the searches run. Confirm via siope.eu publications and the SIOPe Brain Tumour Group LGG working-group pages, and check the EANO guideline index at eano.eu for a paediatric-type glioma document.]

[FLAG 75: The total N and full denominator of the Johns Hopkins adult off-label series could not be extended beyond 7 patients, and the response criteria used are not stated in the abstract. Obtain the full text of Neuro-Oncol Pract 2025;12(6):1051-1057 (PMID 41458920) and read the Methods for the criteria set and the per-patient response table.]

Bottom line for Fore. The guideline lever Fore expected to find is not there, and the substitution is more actionable than the original. Tovorafenib's adult reach is a drafting outcome, not a guideline outcome — which means Fore's equivalent play is not "get into NCCN" but secure indication text scoped to a BRAF-altered tumour entity rather than an age band, with compendium listing pursued later as a payer-coverage exercise once Flags 71–73 are closed by someone with account access. In the interim, the least-defended ground in the competitor's position is adult tolerability: 2 of 7 adults with grade 4 intratumoral haemorrhage and ~8-week median duration is a thin, alarming dataset defending a label that permits every adult in the country. If Fore has adult safety and durability data — Fore-internal, to be completed from cleared data — that is where it does the most work per slide. Pre-committing publicly to RAPNO-LGG or RANO-LGG, and saying so before anyone asks, converts what is otherwise a methodological chore into a credibility asset in exactly the forum (guideline and compendium committees) where methodological credibility is the currency.

13. Commercial Infrastructure and Field Presence

13.1 The field force is eighteen people, and it did not grow

Day One discloses its US OJEMDA field size directly, requiring no triangulation: "In the United States, pLGG is treated by a focused group of specialists, which has enabled us to build a very efficient sales organization with 18 representatives calling on the approximately 200 accounts that treat over 90% of pLGG patients" (source: Day One Biopharmaceuticals FY2025 Form 10-K, filed 2026-02-24, Item 1 Business (Commercialization), p.16; SEC CIK 0001845337, accession 0001193125-26-066956). That is roughly 11 target accounts per territory.

The same sentence, with the same three numbers, appears verbatim in the FY2024 10-K filed 2025-02-25. Over that interval net product revenue rose from $57.2M to $155.4M and prescriptions from >1,600 (FY2024, 8 months post-April launch) to 4,635 (FY2025) (sources: FY2024 and FY2025 Forms 10-K; Day One 8-K 2026-02-24 Ex-99.1; 8-K 2025-02-25 Ex-99.1). Repeated boilerplate would normally raise a staleness concern, but the SG&A line independently corroborates a flat force: personnel-related SG&A grew only $0.8M year over year.

Total company headcount as of 2025-12-31 was 178 FTE, of whom 90 were in R&D and technical operations, capping the entire non-R&D organisation — commercial, medical affairs and G&A combined — at 88 people; 43 hold PhD/PharmD/MD, and approximately 20% are HQ-based in Brisbane CA with the remainder remote, consistent with a field-deployed commercial org (source: FY2025 10-K, "Employees and Human Capital Resources", p.35; prior year 181 FTE / 57 R&D per FY2024 10-K p.34). Triangulating MSLs and field reimbursement against industry ratios gives a total customer-facing field force of 27–36 people, of which 18 are the disclosed representatives. Method: disclosed 18 reps plus an assumed MSL:rep ratio of 0.4–0.7 and 3–5 field reimbursement managers, bounded above by the 88-person non-R&D cap; the MSL and FRM components are ratio estimates, not disclosures (Flag 81).

Full national coverage parity is therefore an 18–20 rep problem costing on the order of $6–10M/year fully loaded — method: 18–20 FTE at $300–500K fully loaded per rep including car, samples, and management overhead, a standard rare-disease oncology range — not a $100M commercial build. Revenue and volume productivity per representative run roughly 2–4x typical rare-disease oncology benchmarks, which is precisely why a flat 18-person force could carry tripling revenue. Field footprint is not the barrier to entry here.

13.2 Where the moat actually is

Three elements, each attackable differently:

• Distribution concentration. Two customers carry 97.6% of net product revenue. Extreme concentration is a two-way street: the same two specialty pharmacy/distributor relationships that lock in OJEMDA would, once contracted, deliver a competitor the entire market in a single negotiation.
• Persistence economics. Reported volumes are TRx — total prescriptions including refills — so revenue compounds off a refill base while new patient starts grew only 19% quarter over quarter in Q3 2025. Growth is substantially persistence-driven rather than acquisition-driven. The defensible asset is the installed patient base, not the prescriber relationship. A competitor cannot take that base cheaply, but it competes for new starts on near-equal footing — and new starts are the smaller, more contestable number.
• Institutional position. Formulary and P&T standing across ~200 centres, now reinforced by Servier's neuro-oncology presence. This is the most durable of the three.

13.3 The Servier transaction, read narrowly

Servier completed the acquisition of Day One on 23 April 2026 via all-cash tender offer at $21.50/share with no contingent value right — an unusually clean structure signalling conviction in the commercial asset (source: Day One Schedule 14D-9 filed 2026-03-26 and subsequent close disclosures). Day One's own board cited inability to scale sales, marketing and distribution as a stated risk of remaining independent — direct evidence that 18 reps was viewed internally as a constraint, not merely an efficiency. Assume the 18 does not stay 18.

The synergy is narrower than the deal narrative implies. VORANIGO (vorasidenib) covers ages ≥12 in adult-type diffuse glioma; OJEMDA covers ≥6 months in pLGG. The genuine shared call point is the adolescent/young-adult segment at academic centres — real, but partial. Below age 12 the buyer is a children's hospital paediatric neuro-oncology service, a channel Servier did not previously hold at scale. This is a channel addition Servier must build into, not a free overlay onto an existing force. That distinction is the single item most likely to be gotten wrong in a competitive briefing, and it should be stated explicitly to Fore's medical lead. Note also that ex-US OJEMDA rights sit with Ipsen, not Servier: Servier bought a US-only commercial franchise plus pipeline and must coordinate with a licensee it does not control.

13.4 The open window

Every Day One stock option was accelerated and cashed out at close — a textbook post-merger attrition trigger arriving simultaneously across the entire 18-person field force. Two separately branded patient-support hubs ("EveryDay Support From Day One" and ServierONE) are still running three months post-close, an integration cost and a defection window. The next 12–18 months is therefore simultaneously the peak attrition window for the only 18 people in the US who have carried a pLGG bag, and the period of maximum hub and access disruption — a recruiting opportunity and the window in which account relationships are most re-openable.

13.5 What not to attack

OJEMDA's approval rests on RAPNO-LGG, not RECIST, and the label is criteria-appropriate for CNS disease. Label verbatim: "The major efficacy outcome measure was overall response rate (ORR), defined as the proportion of patients with complete response (CR), partial response (PR), or minor response (MR) by blinded independent central review based on RAPNO-LGG (Response Assessment in Pediatric Neuro-Oncology) criteria." ORR 53% (40/76), comprising 29 PR and 11 MR; RANO-LGG (2011) as secondary yielded 54%; FIREFLY-1 = NCT04775485; accelerated approval based on response rate and duration of response (source: OJEMDA US PI Section 14 via DailyMed setid ea3a9631-3a66-6a7c-e053-2995a90ae2ad; accelerated-approval status corroborated in the Schedule 14D-9 filed 2026-03-26). Attacking the label on response-criteria grounds would be wrong on the facts and would cost credibility. The real fragility is the accelerated approval itself — Day One's board listed FDA revocation as a standalone risk factor. Compete on the confirmatory-data timeline and on new-start capture.

All plixorafenib comparison cells on this axis remain Fore-internal — to be completed from cleared data.

[FLAG — CONTRADICTION 76: Account denominators differ and must not be conflated. Briefing materials cite "~110 accounts treating at least one patient through Dec 2024"; the 10-K cites "~200 accounts that treat over 90% of pLGG patients." These are not contradictory — one is productive accounts, the other the target universe — but together they imply ~55% account penetration (110/200) as of Dec 2024, a materially different story from "200 accounts covered." — Confirm per Flag 79.]

[FLAG — CONTRADICTION 77: Growth-rate framing. The FY2025 press release uses 181% for prescription growth and 172% for net revenue growth ($57.2M → $155.4M = 171.7%). Both are correct for their own metric; citing 181% against revenue is an error. — Confirm from 8-K 2026-02-24 Ex-99.1.]

[FLAG — CONTRADICTION 78: Literal subtraction implies non-R&D headcount fell from 124 to 88 (−36) year over year, which conflicts with 172% revenue growth and reaffirmed 2026 guidance of $225–250M. The Mersana acquisition and technical-operations scale-up are the likely reconciliation via reclassification, but Day One never discloses commercial-only headcount, so the conflict cannot be closed from public filings. This should not be read as a 36-person commercial layoff. — Confirm via the FY2025 10-K segment/headcount footnotes and the Mersana transaction disclosures.]

[FLAG 79: The ~110 prescriber-account figure is not in any Day One 10-K, 10-Q or 8-K press release retrieved — pull the Q4 2024 earnings call transcript (call held 2025-02-25) or the January 2025 JPMorgan Healthcare Conference deck from Day One's IR site, where launch-metric account counts were typically presented.]

[FLAG 80: Whether the 18-rep field force has been expanded, held or absorbed since the 2026-04-23 close is unverifiable from filings — Day One's SEC reporting obligations terminated at close and Servier is privately held with no US segment disclosure. Confirm via LinkedIn Sales Navigator headcount trend on "Day One Biopharmaceuticals" and "Servier Pharmaceuticals" filtered to Sales/Business Development, plus current OJEMDA territory-manager postings on servier.us careers and LinkedIn Jobs.]

[FLAG 81: MSL and field-reimbursement headcount are estimated from industry ratios and are never disclosed by Day One. Confirm by counting distinct individuals with Day One or Servier MSL / Field Reimbursement / Key Account titles in LinkedIn, and by checking CMS Open Payments 2024–2025 for the count of unique Day One-affiliated payers of consulting/travel to paediatric neuro-oncologists, which approximates field medical reach.]

[FLAG 82: Whether Servier will consolidate the "EveryDay Support From Day One" hub into ServierONE, and on what timeline, is not public. Confirm by monitoring ojemda.com for a hub rebrand and by checking whether the OJEMDA co-pay program terms and enrollment fax/phone numbers migrate to ServierONE infrastructure.]

[FLAG 83: Servier US total employee count and oncology field-force size are not disclosed anywhere on servier.us. Confirm via Servier Group's annual financial report (servier.com, published each December for the FY ending 30 September), which reports employees by geography, or via a paid LinkedIn/ZoomInfo firmographic pull on Servier Pharmaceuticals LLC. Note also that Servier US markets additional oncology products commonly cited in trade coverage which were not verified from primary sources and are deliberately not asserted here.]

[FLAG 84: The degree of genuine prescriber overlap between VORANIGO (age ≥12, adult-type diffuse glioma) and OJEMDA (age ≥6 months, pLGG) is asserted on label logic, not measured. Confirm by cross-tabulating CMS Open Payments 2025 recipient NPIs receiving payments from both Servier Pharmaceuticals LLC and Day One Biopharmaceuticals — the intersection size is a direct public measure of shared call point.]

[FLAG 85: Post-close attrition from the 18-rep field force is inferred from the option-acceleration mechanism, not observed. Confirm by tracking LinkedIn profile changes among individuals listing Day One territory-manager roles as of Q1 2026, over the 2026-07 to 2027-04 window.]

[FLAG 86: The FY2025 confirmatory-trial status for OJEMDA's accelerated approval bears directly on how long this commercial infrastructure has an asset to sell. Confirm via clinicaltrials.gov NCT05566795 status and enrollment, and any FDA post-marketing requirement milestones in the Drugs@FDA approval package for NDA 218033.]

Bottom line for Fore. The usual competitive read inverts here. A Servier-backed asset presents as commercially unassailable; the filings say the opposite about field scale specifically. Any Fore plan premised on out-hiring Servier is solving the wrong problem — coverage parity is an $6–10M/year problem, and the disclosed SG&A line materially overstates what a competitor would need to replicate because field cost is a small minority of it. What Fore must actually plan against is different: an installed refill base it cannot take cheaply, two distributor relationships that are a single negotiation away from being neutral, and an institutional formulary position that is the genuinely durable asset. The operative conclusion is one of timing. The next 12–18 months stacks three things at once — post-merger option-cashout attrition across the entire experienced US pLGG field force, dual-hub access disruption, and a channel Servier must build rather than inherit below age 12. That is the window in which accounts are re-openable and in which the only people who have ever sold into this call point are recruitable. Fore should compete for new patient starts (the smaller, contestable number) rather than for the installed base, and should attack the confirmatory-data timeline rather than the response criteria — the criteria attack is factually wrong against the label and would forfeit exactly the credibility the rest of this dossier is built on.

14. KOL Perception and Social-Listening Design

14.1 The independent discourse is narrow — and already skeptical on the right axes

Tovorafenib's non-investigator KOL discourse exists but is thin, and where it exists it is skeptical on precisely the axes a differentiated RAF inhibitor would target: durability, rebound on discontinuation, and growth-velocity toxicity in children. Critically, these arguments have been built by the field's own referees without any competitor prompting.

The stop-and-observe question already has formal consensus vocabulary authored predominantly by non-FIREFLY-1 investigators: O'Hare 2024 (26 of 32 authors non-FIREFLY-1) defines "rapid rebound growth within 3 months of discontinuing targeted therapy" as a named entity. Separately, the paediatric AE-management consensus reached agreement on cutaneous toxicity but explicitly failed to reach consensus on many rarer adverse events: 50 of 129 drafted statements reached ≥75% consensus among 82 global experts, with consensus concentrated on general AE management and cutaneous conditions and many rare AEs lacking sufficient evidence for a recommendation (source: Hargrave D, Bowers DC, Goldman S, Liu GT, Huang JT, van Santen HM, Robison NJ, Zapotocky M, Bouffet E. Neurooncol Pract 2026 Feb 6;13(3):620-633, PMID 42131759; 8 of 9 authors are non-FIREFLY-1 authors, Hargrave being the sole overlap).

The only dedicated non-investigator editorial on tovorafenib in paediatric LGG is authored by Eric Bouffet, who is not a FIREFLY-1 primary-report author — making him the single highest-value independent voice on this asset. PubMed indexes no abstract for it, so his actual positions are not recoverable from the free record (Flag 89).

14.2 Growth-velocity toxicity is labeled, quantified, and under-reported by the primary publication

OJEMDA label §5.4 "Effect on Growth" reports reductions in growth velocity, with growth effects in 46% of 133 FIREFLY-1 patients aged ≤18 years, median height percentile change −14 (z-score −0.6) at 12 months, median on-treatment growth velocity 0.86–1.8 cm/year against post-treatment median annualized growth velocity 4.2 cm/year. The label directs routine growth monitoring and notes improvement after interruption (source: OJEMDA US PI §5.4 via DailyMed, retrieved 2026-07-20; single-source — see Flag 92).

[FLAG — CONTRADICTION 87: Toxicity salience diverges sharply between publication and label. The FIREFLY-1 abstract's most-common-AE list is hair colour change (76%), CPK elevation (56%), anaemia (49%) — with no mention of growth toxicity at all — whereas the label carries growth as a dedicated Warnings and Precautions section affecting 46% of paediatric patients. The primary publication's framing materially under-weights the toxicity that matters most to a paediatric prescriber. — Confirm by reading the Nat Med 2024 abstract AE list against PI §5.4 and §6.1 side by side.]

[FLAG — CONTRADICTION 88: The DailyMed retrieval reported "Grade 3+ severity" in 35% for growth effects. A Grade 3 growth-velocity rate of that magnitude is implausible on its face and is more likely a mis-association with a different labeled AE. Reported here rather than silently corrected or dropped. — Confirm from the authoritative Drugs@FDA PDF §5.4 and §6.1 verbatim (Flag 92).]

Note also the denominator discipline required across this asset: efficacy is reported at n=77 (Nat Med Arm 1) and n=76 (FDA efficacy population), Grade ≥3 AEs at 42% and AE discontinuation at 7% (9 patients) against an implied larger safety set (n=137), and growth effects at 46% of 133 patients aged ≤18. Any cross-rate comparison must state which denominator is in use.

14.3 Plixorafenib's problem is absence of perception, not adverse perception

Plixorafenib is effectively absent from neuro-oncology KOL discourse: zero editorials, zero congress commentary in the indexed literature, and an indexed corpus of 32 publications that is a preclinical PLX8394 mechanism literature rather than a clinical conversation — the clinical footprint being a single Langerhans-cell-histiocytosis case report. This is materially different from being viewed skeptically, and it changes what a blinded study should measure: unprompted awareness, not preference.

The one genuine third-party signal is NCT06610682, a Johns Hopkins-sponsored (not Fore-sponsored) CSF-ctDNA trial in BRAF-altered glioma. That investigator is the closest thing to an existing academic champion and should be treated as a named asset in the KOL plan, not merely as a trial site.

14.4 The instructive negative case

Samuel Abbou has 76 indexed publications, a Gustave Roussy platform, deep paediatric neuro-oncology output, and an endocrine-late-effects co-authorship network placing him squarely on the growth-toxicity axis — yet has no public position on tovorafenib, plixorafenib, or RAF inhibition at all, while publishing actively on ONC201 and BIOMEDE. His silence is a citing-competitors signal, not an absence of platform. He is a high-priority blinded-interview target precisely because his network is endocrine late effects rather than RAF pharmacology: he can be credible on growth toxicity without being a RAF partisan. No position may be attributed to him (Flag 94).

The citing-competitors-instead pattern is detectable programmatically and should be run across the full ~40-name pLGG KOL list before the blinded study is fielded, converting an expensive question ("who is unengaged?") into a free one.

14.5 Social-listening protocol

Channel inventory. Only four public channel classes carry genuine neuro-oncology practitioner discussion at usable density — the indexed literature (editorials, consensus statements, commentary), congress abstract books and discussant sessions, practitioner-interview media (OncLive, Targeted Oncology), and clinical-trial registry activity as a proxy for investigator engagement. Everything else is pharma-amplified echo and should be excluded from the frame.

Search strategy. Run a standing three-query lattice rather than ad-hoc searching: (A) direct-mention queries across both assets and all synonyms; (B) a citing-competitors query — named KOLs publishing on adjacent MAPK/glioma targets while not citing either asset, which surfaces unengaged platforms; (C) a criterion-choice query tracking which ORR figure (67% RANO-HGG vs 51–53% RAPNO) each commentator elects to quote, which is itself a measurable perception signal.

What it can and cannot answer. Public listening answers framing and salience questions — which toxicity is discussed, which number is quoted, whose platform is silent. It cannot answer prescribing intent, switching thresholds, or line-of-therapy placement. Reserve the paid blinded study strictly for the latter.

Sample-size and bias limits. Public listening is a census of a small, structurally biased population — the publishing subset of an already small prescriber base — not a sample of prescribers. It must never be quantified as prevalence. The prescriber-base denominator used to bound this is an order-of-magnitude figure, not a counted one (Flag 96).

Whether plixorafenib has anything to say on the durability-on-stop or paediatric-growth axes is Fore-internal — to be completed from cleared data. No plixorafenib comparator number appears anywhere in this section, by design.

[FLAG 89: The actual stated positions of Yaman & Bouffet on type II vs type I RAF inhibition, fusion-vs-V600E, and durability are unverified — PubMed carries no abstract for PMID 38293894. Confirm by obtaining the publisher full text via DOI 10.1080/14728214.2024.2312817 (Taylor & Francis/Atypon) through an institutional subscription or document delivery; it is a 3-page piece, so a single PDF settles it.]

[FLAG 90: OncLive and Targeted Oncology practitioner-interview content is entirely unassessed — onclive.com/search returned HTTP 403 to automated fetch. Confirm by browsing onclive.com and targetedonc.com manually (or via a media-monitoring subscription such as Meltwater/Cision) for "tovorafenib" 2024–2026, capturing the named physician, institution, and FIREFLY affiliation for each interview.]

[FLAG 91: Congress discussant remarks at SNO, SIOP, ASPHO and ASCO 2024–2026 are entirely unassessed — the richest source of unfiltered comparative KOL opinion. Confirm by pulling the Neuro-Oncology and Pediatric Blood & Cancer supplement abstract books for those meetings, and obtaining discussant slide decks or session recordings via the societies' on-demand portals (typically member login required).]

[FLAG 92: The OJEMDA §5.4 growth figures rest on a single retrieval route (DailyMed) and the 35% Grade 3+ value is internally implausible. Confirm by downloading the authoritative PDF from Drugs@FDA under NDA 218033 — the constructed 218033s000lbl.pdf URL returned 404, so navigate from the Drugs@FDA search page — and reading §5.4 and §6.1 verbatim.]

[FLAG 93: Whether the near-total absence of plixorafenib KOL commentary reflects genuine low awareness versus embargo/disclosure constraints on people who do know the asset is unverified. Confirm via unprompted-awareness questioning in the blinded study — precisely the question public listening structurally cannot answer.]

[FLAG 94: Samuel Abbou's position is characterised only as documented public silence on RAF inhibition; no position may be attributed to him. Confirm by naming him as a blinded-interview target, with the growth/endocrine-late-effects angle as the highest-yield line of questioning given his co-authorship network.]

[FLAG 95: The plixorafenib registry list retrieved (NCT06385119, NCT06610682) is likely incomplete — the ClinicalTrials.gov API returned only two records for the query term. Confirm by re-running the registry search across all synonyms (plixorafenib, FORE8394, PLX8394) with no field restriction, and cross-checking against Fore Biotherapeutics' own pipeline disclosures and any SEC filings.]

[FLAG 96: The paediatric neuro-oncology prescribing-base denominator used in the bias-limits estimate is an order-of-magnitude figure, not a counted one. Confirm via SNO and SIOP membership counts by subspecialty, or a claims-based prescriber count for selumetinib / dabrafenib-trametinib in the pLGG indication.]

Bottom line for Fore. The competitive argument Fore needs is already being made by tovorafenib's own referees, which means Fore's job is not to make it but to supply the answer. Independent voices — 26 of 32 non-FIREFLY-1 authors on the rebound consensus, 8 of 9 on the AE consensus, Bouffet on the sole editorial — have built formal vocabulary for exactly the failure modes a differentiated RAF inhibitor would target, and the label itself carries a growth-velocity warning affecting 46% of 133 paediatric patients that the primary publication's abstract does not mention at all. Those axes are open, pre-legitimised, and currently unanswered; that is a rare configuration and it will not persist. Second, Fore should recognise that plixorafenib's KOL problem is invisibility rather than skepticism, which reprices the research plan: run the public lattice to exhaustion first — it costs days and this single pass already answered five field-mapping questions outright — then field the blinded study scoped only to unprompted awareness, switching thresholds, and line-of-therapy placement, the three things public listening structurally cannot reach. Third, treat the Johns Hopkins CSF-ctDNA investigator as a named champion asset and Abbou as the highest-yield unengaged platform, because the citing-competitors query converts KOL mapping from a paid exercise into a free one across the full ~40-name list. One caveat must travel with every use of this material: quote the RAPNO figure in comparative contexts, both because it is the criterion appropriate to low-grade disease and because it is the one the regulator used — quoting the flattering-to-competitor 67% is an own goal, and quoting nothing reads as evasion.

Appendix — New Open Flags

Flags 42–96, continuing the dossier's existing 41-flag sequence. Contradiction flags are marked; all others are verification gaps.

Section 9 — Development Trajectory Beyond FIREFLY

• 42 [CONTRADICTION] — Sponsor and partner misattribution: NCT07206849 is sponsored by Nationwide Children's Hospital with Day One as collaborator; no Servier trial involvement exists; the ex-US partner is Ipsen. Confirm: re-query CTG for sponsor="Servier" + intervention="tovorafenib"; read the Ipsen license announcement.
• 43 [CONTRADICTION] — FIREFLY-2 minimum age is NULL in the registry; the 6-month floor is not registry-sourced. Confirm: LOGGIC/FIREFLY-2 protocol publication or EU CTIS record.
• 44 [CONTRADICTION] — TarGeT-B names no response criteria while sibling FIREFLY-2 explicitly specifies RAPNO. Confirm: see Flag 45.
• 45 — Formal response-assessment criteria for NCT07206849 unknown. Confirm: protocol synopsis/SAP from CONNECT coordinating centre at Nationwide Children's (Kelsey H Troyer PhD, 614-722-3284), or a design abstract at SNO 2026 / ISPNO.
• 46 — Per-stratum sample sizes within N=79 unpublished; Stratum A 20–35 is design-inferred. Confirm: protocol synopsis or FDA Orphan Products Grant R01FD008167 abstract via NIH RePORTER / FDA OOPD listing.
• 47 — Site list empty; geographic footprint and participating CONNECT institutions unknown. Confirm: registry after activation, or CONNECT membership listings.
• 48 — Day One's financial/operational commitment to TarGeT-B (drug supply vs funded study) undisclosed. Confirm: Day One 10-K/10-Q R&D pipeline and collaboration disclosures.
• 49 — FIREFLY-2 6-month minimum age unconfirmable from registry. Confirm: protocol publication or EU CTIS.
• 50 — Label age floor vs trial enrolment not established in the trajectory pass. Confirm: OJEMDA US PI Sections 1 and 8.4; EMA product information.

Section 10 — Label Scope, Age Boundaries and Off-Label Reality

• 51 [CONTRADICTION] — Four primary sources name at least three different response criteria for the same N=76 dataset (FDA PI: RAPNO-LGG; EU SmPC: RANO-LGG 2011; NCT04775485: RANO-HGG; NCT05566795: RAPNO); MR inclusion makes cross-comparison non-equivalent. Confirm: read PI §14, SmPC 5.1, and both registry outcome modules side by side; tabulate criterion against figure.
• 52 [CONTRADICTION] — US indication has no prior-therapy count; EU requires progression after ≥1 prior systemic therapy. Confirm: US PI §1 vs EU SmPC §4.1 verbatim.
• 53 [CONTRADICTION] — Eligibility ceiling 25 years vs enrolled efficacy range 2–21 years, 6/76 (7.9%) over 16, zero over 21. Confirm: PI §14 demographics table.
• 54 — Whether US compendia (NCCN D&B, Lexi-Drugs, AHFS-DI, Micromedex DrugDex) carry an age restriction or adult off-label recognition; compendia drive coverage under SSA §1861(t)(2)(B). Confirm: NCCN Compendium tovorafenib monograph; NCCN CNS Cancers adult algorithm pages; Micromedex DrugDex off-label rating.
• 55 — Whether any payer medical policy carries an explicit age edit vs a diagnosis edit. Confirm: published policies from UnitedHealthcare, Aetna CPBs, Cigna, and 3–5 state Medicaid PDLs.
• 56 — Whether any adult-inclusive tovorafenib trial exists or is planned. Confirm: CTG and EU CTIS queries for tovorafenib/DAY101/TAK-580/MLN2480 with maximumAge absent or >25; Day One 10-K Item 1 and 10-Q pipeline tables; Ipsen license announcement.
• 57 — Whether Mayo abstract PATH-94 has become a full manuscript with named criteria and volumetrics. Confirm: PubMed author search Williams A / Sener U / Ruff M + "tovorafenib", 2025–2026; Mayo neuro-oncology publication list.
• 58 — Whether the EU conditional MA carries a PIP or Annex II obligation functionally constraining paediatric use. Confirm: Annex II of the Ojemda EPAR Product Information; CHMP assessment report benefit-risk section.
• 59 — Plixorafenib comparator cells: trial age eligibility, enrolled adult age range and N, adult CNS efficacy by RANO-HGG/iRANO. Fore-internal — to be completed from cleared data.

Section 11 — Endpoint Strategy

• 60 [CONTRADICTION] — The 67%/51% pair is from Kilburn Nat Med 2024 (n=77), not the FDA label, which prints 53%/54% on n=76. Confirm: text-search the PI PDF for "67".
• 61 [CONTRADICTION] — ORR 51% (95% CI 40–63), DoR 13.8 mo in the FDA approval summary vs 53% (95% CI 41, 64), DoR 18 mo in the 8/2025 label, both on n=76. Confirm: compare stated data-cutoff dates in approval summary methods vs PI §14.
• 62 [CONTRADICTION] — Dabrafenib control-arm denominator: 37 (NEJM 2023) vs 33 (TAFINLAR label text). Confirm: TAFINLAR PI §14.7.
• 63 [CONTRADICTION] — FIREFLY-2 primary criterion RANO-LGG (BMC Cancer 2024) vs RAPNO (registry, updated 2026-05-12). Confirm: registry version history, Flag 66.
• 64 [CONTRADICTION] — n=77 (Kilburn Arm 1) vs n=76 (FDA efficacy population, ≥1 measurable lesion by RAPNO-LGG at baseline). Confirm: PI §14 footnote.
• 65 — FDA's internal reasoning for displacing RANO-HGG as the labeling criterion. Confirm: Multi-Discipline Review / Clinical Review for NDA 218033 from Drugs@FDA (overview page 404'd; manual navigation required).
• 66 — Date and stated reason for the FIREFLY-2 endpoint switch. Confirm: NCT05566795 "Study Record Versions" tab; diff the primary outcome field against the 23 April 2024 approval date.
• 67 — Whether FDA agreed a frontline, RAF-broad trial can verify a relapsed/refractory, BRAF-narrow accelerated approval. Confirm: PMR text from the NDA 218033 approval letter or FDA's Postmarketing Requirements and Commitments database.
• 68 — Labeled (vs published) efficacy figures for dabrafenib+trametinib Study CDRB436G2201. Confirm: TAFINLAR PI PDF, DailyMed setid fee1e6b1-e1a5-4254-9f2e-a70e0f8dbdea, §14.7.
• 69 — No SIOPe LOGGIC endpoint position paper and no EMA-side endpoint document retrieved. Confirm: EMA medicines database OJEMDA EPAR; SIOPe Brain Tumour Group publications; RANO 2009–2025 progress report (Neuro Oncol 2025;27(9):2209-2224, PMID 40317212) pLGG endpoint section.
• 70 — Whether any accelerated-to-traditional conversion has ever rested on a randomised ORR primary alone. Confirm: FDA guidance "Clinical Trial Considerations to Support Accelerated Approval of Oncology Therapeutics" (404'd at two URLs — use FDA guidance search); FDA Project Confirm conversion database.

Section 12 — Guideline and Compendia Positioning

• 71 — Whether the current online NCCN Pediatric CNS Cancers guideline contains an LGG section listing tovorafenib, with wording, line of therapy and Category. Confirm: free NCCN account at nccn.org/guidelines; open "Pediatric Central Nervous System Cancers", record version and date, text-search "tovorafenib"; repeat for adult CNS Cancers. (nccn.org returns HTTP 403 to automated access.)
• 72 — Whether tovorafenib has an NCCN Drugs & Biologics Compendium listing extending beyond the paediatric label — the item most directly driving US payer coverage of adult off-label use. Confirm: paid Compendium subscription; read the Disease/Setting column and Category.
• 73 — Whether Day One or Ipsen has publicly claimed NCCN listing in an SEC filing or investor deck. Confirm: EDGAR full-text search "tovorafenib" AND "NCCN" limited to 10-K/10-Q (requires declared User-Agent); read the Competition section of the most recent 10-K.
• 74 — Whether SIOPe has issued a European paediatric LGG recommendation addressing BRAF-targeted therapy. Confirm: siope.eu publications; SIOPe Brain Tumour Group LGG working-group pages; EANO guideline index at eano.eu.
• 75 — Total N, denominator and response criteria of the Johns Hopkins adult off-label series. Confirm: full text of Neuro-Oncol Pract 2025;12(6):1051-1057, PMID 41458920 — read Methods and the per-patient response table.

Section 13 — Commercial Infrastructure and Field Presence

• 76 [CONTRADICTION] — ~110 productive accounts vs ~200 target accounts implies ~55% penetration as of Dec 2024; the two denominators must not be conflated. Confirm: see Flag 79.
• 77 [CONTRADICTION] — 181% is prescription growth; net revenue growth is 172% ($57.2M → $155.4M). Confirm: 8-K 2026-02-24 Ex-99.1.
• 78 [CONTRADICTION] — Non-R&D headcount arithmetic implies 124 → 88 (−36) against 172% revenue growth and $225–250M 2026 guidance; likely reclassification plus Mersana, unclosable from public filings. Confirm: FY2025 10-K headcount footnotes; Mersana transaction disclosures.
• 79 — Source of the ~110 prescriber-account figure (absent from all 10-K/10-Q/8-K retrieved). Confirm: Q4 2024 earnings call transcript (2025-02-25) or January 2025 JPM deck from Day One IR.
• 80 — Post-close status of the 18-rep field force (Day One reporting terminated at close; Servier is private). Confirm: LinkedIn Sales Navigator headcount trend for Day One and Servier Pharmaceuticals filtered to Sales/BD; OJEMDA territory-manager postings on servier.us and LinkedIn Jobs.
• 81 — MSL and field-reimbursement headcount are ratio estimates, never disclosed. Confirm: LinkedIn title counts for MSL / Field Reimbursement / Key Account; CMS Open Payments 2024–2025 unique Day One-affiliated payers to paediatric neuro-oncologists.
• 82 — Whether and when the Day One hub consolidates into ServierONE. Confirm: monitor ojemda.com for rebrand; track migration of co-pay terms and enrollment fax/phone to ServierONE.
• 83 — Servier US total headcount and oncology field-force size (not on servier.us); also, unverified additional Servier US oncology products deliberately not asserted. Confirm: Servier Group annual financial report (servier.com, published each December for FY ending 30 September); or paid LinkedIn/ZoomInfo firmographic pull on Servier Pharmaceuticals LLC.
• 84 — VORANIGO/OJEMDA prescriber overlap is asserted on label logic, not measured. Confirm: cross-tabulate CMS Open Payments 2025 recipient NPIs receiving payments from both Servier Pharmaceuticals LLC and Day One Biopharmaceuticals.
• 85 — Post-close field attrition is inferred from option acceleration, not observed. Confirm: track LinkedIn profile changes among Q1 2026 Day One territory managers over 2026-07 to 2027-04.
• 86 — Confirmatory-trial status and PMR milestones bearing on how long the asset survives. Confirm: NCT05566795 status and enrollment; PMR milestones in the Drugs@FDA package for NDA 218033.

Section 14 — KOL Perception and Social-Listening Design

• 87 [CONTRADICTION] — FIREFLY-1 abstract lists hair colour change 76%, CPK 56%, anaemia 49% with no growth toxicity, while label §5.4 carries growth effects in 46% of 133 paediatric patients. Confirm: read Nat Med 2024 abstract AE list against PI §5.4 and §6.1 side by side.
• 88 [CONTRADICTION] — Reported 35% "Grade 3+" for growth effects is implausible and likely mis-associated with another AE. Confirm: Drugs@FDA PDF §5.4 and §6.1 verbatim (Flag 92).
• 89 — Yaman & Bouffet's stated positions on type II vs type I RAF inhibition, fusion vs V600E, and durability (no PubMed abstract for PMID 38293894). Confirm: publisher full text via DOI 10.1080/14728214.2024.2312817 through institutional subscription or document delivery — a 3-page piece.
• 90 — OncLive / Targeted Oncology practitioner-interview content entirely unassessed (HTTP 403). Confirm: manual browse or Meltwater/Cision for "tovorafenib" 2024–2026; capture named physician, institution and FIREFLY affiliation.
• 91 — Congress discussant remarks (SNO, SIOP, ASPHO, ASCO 2024–2026) entirely unassessed. Confirm: Neuro-Oncology and Pediatric Blood & Cancer supplement abstract books; discussant decks/recordings via society on-demand portals.
• 92 — §5.4 growth figures rest on a single DailyMed retrieval. Confirm: authoritative PDF from Drugs@FDA under NDA 218033 (constructed 218033s000lbl.pdf URL 404'd — navigate from the search page); read §5.4 and §6.1 verbatim.
• 93 — Whether plixorafenib's discourse absence reflects low awareness or embargo/disclosure constraints. Confirm: unprompted-awareness questioning in the blinded study.
• 94 — Abbou is characterised only by documented public silence; no position may be attributed. Confirm: blinded interview, with growth/endocrine late effects as the highest-yield line.
• 95 — Plixorafenib registry list (NCT06385119, NCT06610682) likely incomplete. Confirm: re-run CTG search across plixorafenib / FORE8394 / PLX8394 with no field restriction; cross-check Fore pipeline disclosures and SEC filings.
• 96 — Paediatric neuro-oncology prescriber denominator is order-of-magnitude, not counted. Confirm: SNO and SIOP membership counts by subspecialty, or a claims-based prescriber count for selumetinib / dabrafenib-trametinib in pLGG.

Consolidated References

1. FDA OJEMDA (tovorafenib) US Prescribing Information, 2024 — https://www.accessdata.fda.gov/drugsatfda_docs/label/2024/217700s000lbl.pdf
2. FDA OJEMDA label 2025 revision (218033s002s003) — https://www.accessdata.fda.gov/drugsatfda_docs/label/2025/218033s002s003lbl.pdf [verify for updated safety language]
3. OJEMDA US Prescribing Information (218033s000lbl), Day One Biopharmaceuticals, 2024 — https://www.accessdata.fda.gov/drugsatfda_docs/label/2024/218033s000lbl.pdf
4. DailyMed OJEMDA label (setid ea3a9631-3a66-6a7c-e053-2995a90ae2ad) — https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?setid=ea3a9631-3a66-6a7c-e053-2995a90ae2ad
5. Kilburn LB, et al. The type II RAF inhibitor tovorafenib in relapsed/refractory pediatric low-grade glioma: the phase 2 FIREFLY-1 trial. Nat Med 2024;30:207-217 — https://pmc.ncbi.nlm.nih.gov/articles/PMC10803270/ (nature.com/articles/s41591-023-02668-y); Author Correction: nature.com/articles/s41591-024-02910-1 (PMC11108768)
6. Singh S, et al. FDA Approval Summary: Tovorafenib for Relapsed or Refractory BRAF-altered Pediatric Low-Grade Glioma. Clin Cancer Res 2025;31(8):1383-1389 — https://pmc.ncbi.nlm.nih.gov/articles/PMC11996598/
7. LiverTox: Tovorafenib. NCBI Bookshelf NBK613920 — https://www.ncbi.nlm.nih.gov/books/NBK613920/
8. openFDA FAERS drug/event endpoint (tovorafenib), queried Jul 2026 — https://api.fda.gov/drug/event.json?search=patient.drug.medicinalproduct:tovorafenib
9. Day One Biopharmaceuticals press release, 3-Year FIREFLY-1 follow-up, SNO Annual Meeting Nov 2025 — https://ir.dayonebio.com/news-releases/news-release-details/day-one-announces-three-year-follow-data-ojemdatm-tovorafenib/
10. FDA D.I.S.C.O. Burst Edition: Ojemda (tovorafenib) approval — https://www.fda.gov/drugs/resources-information-approved-drugs/fda-disco-burst-edition-fda-approval-ojemda-tovorafenib-relapsed-or-refractory-pediatric-low-grade
11. FDA grants accelerated approval to tovorafenib, 23 Apr 2024 — https://www.fda.gov/drugs/resources-information-approved-drugs/fda-grants-accelerated-approval-tovorafenib-patients-relapsed-or-refractory-braf-altered-pediatric
12. ClinicalTrials.gov NCT04775485 (FIREFLY-1) — https://clinicaltrials.gov/study/NCT04775485
13. Structure and RAF family kinase isoform selectivity of type II RAF inhibitors tovorafenib and naporafenib. J Biol Chem 2023;299(5) — https://pmc.ncbi.nlm.nih.gov/articles/PMC10149214/
14. Preclinical Activity of the Type II RAF Inhibitor Tovorafenib in Tumor Models Harboring Either a BRAF Fusion or an NF1 Loss-of-Function Mutation. Cancer Res Commun 2025;5(4):668-679 (doi:10.1158/2767-9764.CRC-24-0451) — https://pmc.ncbi.nlm.nih.gov/articles/PMC12015663/
15. Bouffet E, et al. LOGGIC/FIREFLY-2 (NCT05566795) trial design. BMC Cancer 2024 — https://pmc.ncbi.nlm.nih.gov/articles/PMC10826080/
16. ASCO 2023 abstract TPS10067 (FIREFLY-2 design) — https://ascopubs.org/doi/abs/10.1200/JCO.2023.41.16_suppl.TPS10067
17. Day One and Servier complete enrollment in pivotal Phase 3 FIREFLY-2 trial, PRNewswire/Servier, 8 May 2026 — https://www.prnewswire.com/news-releases/day-one-and-servier-complete-enrollment-in-pivotal-phase-3-firefly-2-trial-of-tovorafenib-as-a-front-line-treatment-for-pediatric-low-grade-glioma-plgg-302766277.html
18. Williams et al. Off-label tovorafenib in adult CNS tumors (Mayo Clinic series). SNO 2025 abstract PATH-94, Neuro-Oncology 2025;27(Suppl 5):v263 — https://academic.oup.com/neuro-oncology/article/27/Supplement_5/v263/8319150
19. FIREFLY-1 3-year follow-up. Neuro-Oncology Pediatrics 2025;2(2):wuag022; abstract CTP-17, Neuro-Oncology 2025;27(Suppl 5):v149
20. ASCO 2025 growth-recovery analysis. J Clin Oncol 2025;43(16_suppl):10029; Neuro-Oncology Pediatrics 2025;2(2):wuag021
21. International Pediatric Low-Grade Glioma Coalition consensus regrowth definitions. Neuro-Oncology 2024;26(8):1357
22. Day One Biopharmaceuticals SEC filings (10-K FY2024, 8-Ks, CIK 1845337); FY2025 preliminary revenue + 2026 guidance, GlobeNewswire 11-Jan-2026
23. Servier acquisition of Day One — SEC 8-K / SC TO-T & SC TO-T/A filings (CIK 1845337), Mar-Apr 2026 — https://www.sec.gov/Archives/edgar/data/1845337/000119312526095174/d108979dex991.htm ; FiercePharma, Mar 2026
24. Ipsen gains ex-US rights to tovorafenib (23-Jul-2024) — Ipsen press release; BioWorld — https://www.bioworld.com/articles/711043-from-day-one-ipsen-gains-ex-us-rights-to-tovorafenib-for-461m
25. Ipsen EU press release: Ojemda approved in the European Union — https://www.ipsen.com/press-release/ojemda-approved-in-the-european-union-as-the-first-targeted-therapy-in-relapsed-or-refractory-pediatric-low-grade-glioma-regardless-of-braf-alteration-3278647/ ; EMA Ojemda EPAR — https://www.ema.europa.eu/en/medicines/human/EPAR/ojemda
26. Ipsen CHMP positive opinion release — https://www.ipsen.com/press-release/ipsen-receives-positive-chmp-opinion-for-ojemda-... ; ESMO — https://www.esmo.org/oncology-news/ema-recommends-granting-a-conditional-marketing-authorisation-for-tovorafenib
27. EU Joint Clinical Assessment report for tovorafenib (Ojemda), EC — https://health.ec.europa.eu/publications/joint-clinical-assessment-report-tovorafenib-ojemda_en ; Ipsen JCA update
28. NCI Cancer Currents, Apr 2024: tovorafenib for pediatric low-grade glioma — https://www.cancer.gov/news-events/cancer-currents-blog/2024/pediatric-low-grade-glioma-tovorafenib-braf
29. Bouffet E, et al. Dabrafenib + trametinib in pediatric BRAF V600 low-grade glioma. NEJM 2023;389 [FLAG: confirm start page — likely 1108, not 1118]
30. Day One state price-transparency filings (Connecticut / North Dakota 2024 WAC; Colorado prescriber filing effective 01-Jan-2026 WAC)
31. Day One Biopharmaceuticals PRV sale press release, GlobeNewswire, 30-May-2024 ($108M)
32. Federal Register 2024-10583, 15-May-2024 (orphan drug designation record)
33. OncLive: FDA approves companion diagnostic for tovorafenib in BRAF-altered pediatric low-grade glioma — https://www.onclive.com/view/fda-approves-companion-diagnostic-for-tovorafenib-in-braf-altered-pediatric-low-grade-glioma [verify CDx name/date]
34. PharmaCompass tovorafenib patent-expiry listing (secondary; verify against Orange Book)
35. Drugs.com / Ojemda USPI — https://www.drugs.com/pro/ojemda.html ; FDA label media/180679 — https://www.fda.gov/media/180679/download
36. [TO ADD if refreshing to current cut] SNO 2025 three-year FIREFLY-1 update (Day One IR, Nov 2025) — updated ORR (RAPNO ~53% / RANO-HGG ~71%), DoR/PFS, treatment-free observation — https://ir.dayonebio.com/news-releases/

Appendix — Open Verification Flags (41)

Competitive-intelligence dossier; every item below needs human/primary-source sign-off before external or regulatory use. Tovorafenib data is public+cited; the plixorafenib comparison column carries ONLY Shubin-slide figures (Fore-internal, unverified) and is a template for the team to complete.

• [DATA-CUT/TEMPORAL] Tovorafenib efficacy column uses the 2024 registration cut; a 3-year FIREFLY-1 update (SNO 2025) reports higher estimates (RAPNO-LGG ~53%, RANO-HGG ~71%) plus updated DoR/PFS and treatment-free-observation data. Label every figure with its cut date or refresh to SNO 2025.
• [EFFICACY — CRITERION CONTRADICTION, high] Day One's SNO 2025 release calls 53% (40/76) 'RAPNO-LGG' while the FDA label calls 51% (39/76) 'RAPNO-LGG'. Same criterion name, different responder count (40 vs 39). The 53%->RANO-LGG / 51%->RAPNO-LGG mapping is NOT how Day One labels these. Reconcile against Nat Med 2024 supplementary response-criteria tables before any 'X% by criterion Y' claim.
• [EFFICACY] Median PFS 19.4 mo (RANO-HGG, Arm 1) could not be independently confirmed; DoR 16.6 mo is confirmed and a secondary source reports 16.6-mo PFS, raising a possible DoR/PFS conflation. Verify against Nat Med 2024 results table. RANO-LGG PFS 13.9 mo likewise unverified.
• [EFFICACY] SNO 2025 median DoR 19.4 mo (13.8-27.2) asserted but not confirmed in located press release (which headlines TTNT >3.5 yr and treatment-free interval); confirm against abstract CTP-17 before the '13.8 -> 19.4 mo DoR lengthening' claim.
• [EFFICACY] Age-subgroup ORR not published with N/CI; FIREFLY-1 Arm 2 (N=60) baseline demographics not fully published. Do not infer subgroup ORRs.
• [SAFETY] Fatal tumor hemorrhage: FDA publishes '1 patient (1%)'; workstream's 0.6% used the pooled N=172 denominator. Quote FDA verbatim ('1 patient (1%)') or state both denominators; do not present 0.6% alone.
• [SAFETY] All figures sourced to the initial 2024 OJEMDA PI; a 2025 label revision (218033s002s003) exists. Pull the current 2025 label and confirm no W&P / Adverse-Reactions rates changed before finalizing.
• [SAFETY] QT '>20 ms not observed' attributed to the label but not surfaced in the FDA Approval Summary (which does not report QT). Confirm the exact PI Clinical Pharmacology wording; else soften to 'no QT signal reported in FIREFLY-1 secondary endpoints'.
• [SAFETY] LVEF: no LVEF toxicity data reported — absence of reported data, not demonstrated cardiac safety.
• [SAFETY] LiverTox specifics (ALT/AST >5xULN, 14-day onset, DILI score E*) and the §5.4 height-percentile/z-score/growth-plate figures cited but not independently re-verified this pass; spot-check against NBK613920 and label §5.4.
• [SAFETY] FAERS per-term counts (off-label use 6, disease progression 2, single-report RCVS/thunderclap headache/etc.) are point-in-time and dynamic; re-query openFDA at dossier-freeze date. RCVS/thunderclap headache is an unvalidated single-case signal, not a labeled AE.
• [RWE] Mayo adult series (5/8 radiographic responses) does not state RANO vs RAPNO criteria; keep N=8 explicit and do not attach a response criterion.
• [RWE] Nearly all durability/rebound/retreatment/growth-recovery data are FIREFLY-1 trial-derived, not independent RWE; retain trial-vs-real-world provenance labels; do not compress into a generic 'real-world durability' claim.
• [RWE] '110 total accounts treating >=1 patient (through Dec 2024)' not confirmed in FY2024 8-K/press (which reported prescription volume, not account count); verify against Q4/FY2024 earnings-call transcript or drop.
• [RWE] Median TTNT 42.6 mo and Q1 2025 prescription detail (>900, +16% QoQ) not independently re-verified; cross-check against SNO 2025 abstract and the Q1 2025 8-K.
• [RWE] On-treatment baseline AGV (~1.1 cm/y in original) not confirmed; use the corrected 3/6/12-month off-treatment AGV curve (~4.3/10.2/7.7 cm/y) and 89%/74% recovery rates.
• [RWE — TEMPORAL] Q1 2026 (quarter ended 31 Mar 2026) results are likely now reported (early/mid-May); incorporate Q1 2026 actuals so the commercial section is current.
• [RWE] No tovorafenib-specific FAERS disproportionality study located; post-marketing pharmacovigilance analysis not yet published.
• [REGULATORY] Ipsen ex-US deal '~$461M' is total incl. milestones (upfront ~$111M / ~$71M cash + ~$40M equity / up to ~$350M milestones); confirm the upfront-vs-milestone split before quoting a single number.
• [REGULATORY] US accelerated->traditional conversion has NOT occurred; FIREFLY-2 topline expected ~mid-2027. Keep status as 'accelerated approval, confirmatory trial fully enrolled, data pending'; do not imply full approval.
• [REGULATORY] EMA/EC date (20 Apr 2026) vs Ipsen-announced EC decision (22 Apr 2026): 2-day discrepancy; treat EMA's 20 Apr 2026 as authoritative and footnote the Ipsen variance.
• [REGULATORY — INDICATION SCOPE, high] Ipsen headline 'regardless of BRAF alteration' contradicts the restrictive binding EMA indication (BRAF fusion/rearrangement or V600 only, R/R). Always use the EMA EPAR / SmPC wording; never reproduce the promotional headline as the indication.
• [REGULATORY] Only a 100 mg tablet strength confirmed; possible additional 2025-label strengths unverifiable (accessdata PDFs 404). Pull current USPI from DailyMed to confirm all strengths.
• [REGULATORY] Companion-diagnostic 'later CDx approval' referenced without confirmed name/date; verify the CDx assay name and approval date from OncLive/FDA before presenting as a completed milestone.
• [REGULATORY] Japan PMDA (orphan designation only, no approval located) and UK MHRA / Swissmedic / Health Canada / Australia TGA statuses unverified; mark as 'not verified / no public approval record'.
• [MECHANISM_PK] Dimer-cooperativity Hill-slope range '-2.6 to -3.2' was misapplied to both isoforms; JBC 2023 Table 1 shows BRAF WT ~ -2.89 and CRAF WT ~ -1.55. Confirm exact per-isoform Hill slopes against Table 1 (PMC10149214).
• [MECHANISM_PK] NF1-loss single-agent EC50 range is ~0.99 to >10 umol/L (some lines ~1 umol/L), not '>7.5 umol/L in most lines'; preclinical, outside the approved indication. Confirm per-line values if a specific number is needed.
• [MECHANISM_PK] CNS-penetrant claim is qualitative only; no quantitative Kp,uu / brain-to-plasma ratio in public sources — do not attach a numeric penetration figure.
• [MECHANISM_PK] Approval is ACCELERATED (single-arm ORR/DoR), not full/regular; 'first FDA-approved targeted therapy for the KIAA1549::BRAF-fusion pLGG population' is the accurate framing (FDA says 'first targeted agent', not 'first systemic therapy').
• [COMPETITIVE_IP — pricing, high/temporal] 2024 launch WAC ($35,272.64 tablet / $8,818.16 suspension) is stale for mid-2026; use current Jan-2026 Colorado-filing WAC (~$38,702.90 tablet / ~$9,675.73 suspension). Oral-suspension figure corrected from an erroneous ~$8,479.
• [COMPETITIVE_IP] Q1'25 '+11% QoQ' revenue growth and the ~68% Servier acquisition premium were not re-verified against primary sources; Day One's disclosed Q1'25 QoQ metric was +16% for prescriptions (not revenue). Confirm or drop.
• [COMPETITIVE_IP] Ipsen 'royalties run >=10 years from first commercial sale per country' not present in the public Ipsen release; confirm against the redacted license exhibit (Day One 10-Q) or remove.
• [COMPETITIVE_IP] COM patent 'mid-2036 with PTE (~5 yr)' is Day One's disclosed estimate, not an adjudicated PTE grant; confirm Orange Book patent numbers and any granted PTE date once OJEMDA is listed. Any generic/505(b)(2) Paragraph-IV timeline is speculative until Orange Book listing is public.
• [COMPETITIVE_IP] Verify Sept-2020 (orphan) / July-2021 (rare pediatric) designation dates against FDA records (not re-verified this pass).
• [COMPETITIVE_IP] Confirm Bouffet et al. NEJM 2023;389 start page (likely 1108, not 1118).
• [SLIDE FRAMEWORK — CONTRADICTION/DENOMINATOR, high] FIREFLY-1 Arm-1 '>=3 prior lines = 39 (51%)' is internally suspect: the 1/2/>=3 distribution is marked not-published-by-arm and the value duplicates the optic-pathway 39 (51%) exactly. FDA reports only median 3 prior lines (range 1-9). Remove the cited value and replace with a verify-flag until traced to the Day One CSR.
• [SLIDE FRAMEWORK — DoR/PFS SWAP, high] RANO-HGG DoR (16.6 mo) and PFS (19.4 mo) may be transposed (public reporting surfaces DoR ~19.4 / PFS ~16.6); RAPNO PFS listed as 13.8 mo is identical to RAPNO DoR (likely copy error). Reconcile all four values against Kilburn 2024 tables.
• [SLIDE FRAMEWORK — ASSESSOR ATTRIBUTION] The 67% RANO-HGG and 51% RAPNO ORRs were labeled 'investigator' in a prior draft, but the FIREFLY-1 primary endpoint was assessed by blinded independent central review (BICR). Correct the assessor attribution.
• [SLIDE FRAMEWORK] 'DoR >=6 mo in 85%' cited to FDA summary but unconfirmed (FDA reports ~23% with DoR >=12 mo); verify or drop.
• [SLIDE FRAMEWORK — SCOPE, confirmed clean] All plixorafenib cells (FORTE Sub A/Sub B, PLX120-03) remain per-slide, Fore-internal, embargoed template placeholders; no plixorafenib efficacy/safety was researched, inferred, or fabricated. Fine-grained Arm-2 baseline table and per-arm tumor-location/prior-line/prior-radiation/prior-surgery distributions are not itemized by arm in Kilburn 2024.
• [FIREFLY-2 site count] '~100 sites' (2024 protocol) superseded by ~140 sites / ~400 participants (May 2026 completion release); confirm exact count against NCT05566795 if needed.