Ojemda (tovorafenib) — what has changed

The hub’s Tier 1 row already carries the corporate and trial picture as of its 3 September 2026 refresh — Servier ownership, Ipsen ex-US, FIREFLY-2 topline mid-2027, enrolment completed May 2026, EU decision to verify. This update adds what that refresh does not have: the regulatory and label record. An efficacy supplement that did not convert Ojemda to full approval, a new growth-velocity warning aimed at its own paediatric population, and a label that warns of increased tumour growth in BRAF-wild-type disease — the paradox question, answered on the label.

1 · The ownership change, dated

Already tracked. The hub’s Tier 1 row records “Day One / Servier · Ipsen ex-US” as of 3 September 2026. What follows adds the dates and the mechanism, not the fact.

Servier Pharmaceuticals acquired Day One Biopharmaceuticals by tender offer: first communication 23 March 2026, offer commenced 26 March, completed 23 April, and the company deregistered on 4 May 2026 (SEC Form 15-12G). EDGAR now returns an empty ticker array for CIK 0001845337.

The consolidation worth naming. The hub separately tracks BDTX-4933 / S241656 as a Servier asset, licensed from Black Diamond in March 2025. With Ojemda, Servier now holds two RAF-directed programmes — an approved paediatric pan-RAF and a licensed follow-on — alongside vorasidenib in IDH-mutant glioma. Neither row states this on its own; it is only visible reading them together. Fore is not facing a single-asset biotech and a separate licensor, but one company assembling a RAF and CNS-tumour position.

2 · The label moved in August 2025 — and not toward full approval

An efficacy supplement was approved on 27 August 2025, yet the indication still reads “approved under accelerated approval based on response rate and duration of response… continued approval may be contingent upon verification… in a confirmatory trial.” The supplement did not buy full approval. The accelerated-approval obligation is still outstanding, more than two years after first approval.

3 · The label warns of increased tumour growth in BRAF-wild-type disease

This is paradoxical MAPK activation described in clinical-label language: in BRAF-wild-type context, the drug may accelerate rather than restrain tumour growth. It sits on the approved label — the strongest source tier available, and considerably stronger than the conference and company material the dossier has been weighing on this question.

The dossier’s open flag asks whether a post-marketing paradox signal exists. Two separate answers, and they point different ways:

• Cutaneous malignancy: nothing. FAERS holds 292 Ojemda reports across 100 distinct reaction terms. Searched directly, cutaneous SCC, keratoacanthoma, malignant melanoma, basal cell carcinoma, skin papilloma and SCC all return zero. What is present is melanocytic naevus (n=15) with hyper- and hypopigmentation — benign RAF class effects.
• Tumour-growth paradox: acknowledged on the label. Section 5.6 exists because the nonclinical data required it.

So the paradox concern for this molecule has been formally recognised, but in the tumour compartment rather than the skin. A comparison that only looks for cuSCC will conclude there is no signal, and will be looking in the wrong place.

4 · Safety profile as labelled

• 5.1 Haemorrhage — haemorrhagic events in 37%; epistaxis 26%; intratumoral haemorrhage 9%; serious bleeding 5%; one Grade 5 tumour haemorrhage (0.6%).
• 5.2 Skin toxicity including photosensitivity · 5.3 Hepatotoxicity · 5.4 Effect on growth (new) · 5.5 Embryo-fetal toxicity · 5.6 NF1 tumour growth.
• FAERS most-reported: off-label use (26), rash (26), disease progression (24), fatigue (21), accidental underdose (19), hair colour change (18).

Intratumoral haemorrhage at 9% in a CNS population is the number worth carrying into any comparison. Off-label use tying for most-reported term, and accidental underdose at 19, together suggest prescribing beyond the label and real-world dosing difficulty.

5 · Where it is going

• The confirmatory trial reads out mid-2027. NCT05566795, 418 patients, active and no longer recruiting, primary completion June 2027. Full approval cannot realistically precede that.
• Expansion beyond pLGG is underway — craniopharyngioma (recruiting) and high-grade glioma / DIPG (opening). Both are investigator- or institution-sponsored, not company pivots, but they widen the clinical footprint.
• Ipsen is running its own Japan study — the ex-US partner building its own dataset.
• Two Day One trials terminated (tovorafenib ± pimasertib, and a monotherapy study). Combination development with a MEK inhibitor stopped.

6 · What this changes for Fore

• Servier is a two-asset RAF competitor, not an acquirer of one. Ojemda plus BDTX-4933/S241656, plus vorasidenib in the same clinics. Positioning that treats Day One and Servier as separate entries understates the position.
• The monotherapy axis just got more interesting. Ojemda’s MEK combination programme was terminated, while its label now carries a wild-type tumour-growth warning.
• Paediatric growth effects are a differentiator to examine, not to assert. Whether plixorafenib carries comparable exposure-duration risk is a question for the team, not a claim to make from this document.
• There is a dated window. Ojemda remains on accelerated approval until at least mid-2027.

7 · Watch list

• Sponsor name changing on the FDA label from Day One to Servier — the grounding-tier confirmation of the transition.
• Any further supplement to NDA 217700 / 218033, especially one converting accelerated to full approval.
• NCT05566795 status moving to Completed, or its primary completion date slipping past June 2027.
• FAERS acquiring any cutaneous-malignancy term, which would change the paradox picture.
• EU: CHMP opinions and HTA decisions — not covered by the sources below.