Tolebrutinib

Tolebrutinib is Sanofi's oral, brain-penetrant BTK inhibitor - the most clinically advanced of its class in multiple sclerosis. In September 2024, HERCULES (non-relapsing secondary progressive MS) hit its primary endpoint, cutting 6-month confirmed disability progression by 31% versus placebo - the first positive Phase 3 ever in this MS subtype [1][2]. The same week, GEMINI 1 and 2 missed their primary endpoint of annualized relapse rate against teriflunomide in relapsing MS, though disability progression signals were positive on secondaries [1][2]. In December 2025, PERSEUS missed its primary endpoint in primary progressive MS and Sanofi announced it will not pursue PPMS registration [3][14]. Days later, the FDA issued a Complete Response Letter (CRL) for the non-relapsing SPMS NDA, citing a fatal drug-induced liver injury case and an unfavorable benefit-risk profile across patient subpopulations [15]. The EMA went the other way: CHMP adopted a positive opinion for tolebrutinib (brand name Cenrifki) in nrSPMS on April 24, 2026 [16]. The asset is now a transatlantic split decision rather than the breakout Sanofi paid $3.68B for Principia to acquire in 2020.

Novel compound, not yet approved anywhere as of June 2026. Phase 3 program spans three MS subtypes via four key studies: GEMINI 1 (NCT04410978, relapsing MS vs teriflunomide, n=974, completed - missed primary on ARR) [4], GEMINI 2 (relapsing MS, completed - missed primary on ARR) [1], HERCULES (non-relapsing SPMS vs placebo, n~1,130, completed - met primary on 6-month cCDP) [1], and PERSEUS (NCT04458051, PPMS vs placebo, n=767, completed Dec 2025 - missed primary on 6-month cCDP; brain volume loss signal on secondaries) [3][14][17]. A long-term safety extension (NCT06372145) is enrolling ~2,500 participants across studies [5]. Regulatory status: FDA placed tolebrutinib on partial clinical hold in June 2022 after four drug-induced liver injury (DILI) cases; the hold was lifted with enhanced ALT/AST monitoring and tighter screening criteria [6]. Sanofi submitted an NDA to the FDA for nrSPMS based on HERCULES (PDUFA originally Sept 28, 2025; extended to Dec 28, 2025 after a major amendment); on Dec 23, 2025, the FDA issued a CRL citing a fatal DILI case and that 'a favorable benefit-risk profile could not be established for any patient subpopulation' [15][18]. Sanofi expected further regulatory dialogue by end of Q1 2026 [15]. In the EU, the CHMP issued a positive opinion for Cenrifki (tolebrutinib) in nrSPMS on April 24, 2026; EC decision expected to follow within ~60 days [16]. Japan PMDA status not publicly disclosed. No FDA breakthrough or fast track designation. Sanofi will not file for PPMS [14]. The next near-term catalyst is FDA Type A/B meeting feedback and the formal EC decision on the EU approval.

BTK (Bruton's tyrosine kinase) is a signaling protein that sits inside two cell types relevant to MS: B cells (the immune cells that mature into antibody factories) and microglia (the brain's resident immune cells) [8]. In MS, B cells drive the peripheral inflammation that causes relapses, while activated microglia drive the slow, smoldering tissue damage that causes disability accumulation in progressive MS. Existing high-efficacy MS drugs - ocrelizumab, ofatumumab - deplete B cells in the blood but can't cross the blood-brain barrier in meaningful amounts, leaving the microglial fire burning. Tolebrutinib is small and lipid-soluble enough to enter the brain at therapeutic concentrations and covalently bind BTK in both compartments [8][9]. The genetic validation is solid: humans born with broken BTK get X-linked agammaglobulinemia (no functional B cells), and ibrutinib's success in CLL and MCL proves BTK inhibition is druggable [10]. What HERCULES established is that the CNS arm of the mechanism actually matters clinically - disability progression slowed in a non-relapsing population. PERSEUS complicates that story: a similar primary endpoint in a mechanistically adjacent population (PPMS) missed, suggesting either disease-stage heterogeneity, dosing/exposure limitations, or that microglial-driven progression in PPMS is less BTK-dependent than the nrSPMS data implied [14][17].

HERCULES enrolled ~1,130 non-relapsing SPMS patients randomized 2:1 to tolebrutinib 60mg daily vs placebo; primary endpoint was 6-month confirmed disability progression (cCDP), met with a 31% relative risk reduction [1][2]. PERSEUS (NCT04458051) ran the same 2:1 design in PPMS, n=767, primary endpoint 6-month cCDP - missed in December 2025, with a positive but non-significant trend; brain volume loss was reduced on a key secondary [3][14][17]. GEMINI 1 (NCT04410978, n=974) and GEMINI 2 used annualized relapse rate vs teriflunomide as primary - an aggressive bet, given teriflunomide is a moderately effective oral standard [4]. Both GEMINI trials missed that endpoint but showed numerical advantage on disability progression, hinting that ARR was the wrong yardstick for this drug [2]. Combined key enrollment across the four trials was ~3,800 participants (note: the 2,500 number commonly cited refers to the long-term safety extension NCT06372145, not the keys) [5]. The placebo-controlled progressive MS designs were appropriate - there's no approved therapy for non-relapsing SPMS, and ocrelizumab is the only PPMS option. The active-comparator design in GEMINI raised the bar in relapsing disease; in retrospect a non-inferiority frame or a head-to-head against an anti-CD20 might have produced a cleaner regulatory story. With PERSEUS missed and GEMINI's ARR endpoint missed, the entire commercial case now rests on nrSPMS - and FDA has rejected that filing pending additional data.

Our model estimates a 65% chance this drug is eventually approved. That figure starts from the historical approval rate for Phase 3 drugs in this area, which is about 51%. It then adjusts based on ten facts about the trial and sponsor - the biggest boosts come from strong earlier-phase results, larger-than-typical enrollment, a non-randomized design, and light or open-label blinding. The remaining facts are close to average for this stage, so they leave the estimate near where the base rate set it.

Liver toxicity is the dominant safety issue and the proximate cause of FDA rejection. The 2022 FDA partial clinical hold followed four DILI cases [6]; the hold was lifted with stricter ALT/AST monitoring and screening exclusions. Subsequent late-stage development included at least one fatal liver injury case, which the FDA cited in its December 2025 CRL alongside an unfavorable benefit-risk determination 'for any patient subpopulation' [15][18]. ALT/AST elevations >3x ULN have been reported in roughly 4% of treated patients in Phase 2/3 studies per the published BTK-in-MS systematic review literature [13][9]. Any future EU label and any potential US resubmission will carry hepatic monitoring requirements that will constrain real-world uptake. The class precedent is ugly: evobrutinib had its own liver signal and missed efficacy in evolutionRMS 1 and 2 (announced December 5, 2023), prompting Merck KGaA to discontinue the program [12][19]. Fenebrutinib (Roche, non-covalent/reversible) and remibrutinib (Novartis) are still in Phase 3 and may have differentiated safety profiles. Efficacy risk is now realized in PPMS (PERSEUS miss) and relapsing MS (GEMINI miss). Commercial risk: nrSPMS is a narrow population (~10-15% of MS patients), and even with EU approval, payers will scrutinize a drug requiring routine liver monitoring against ocrelizumab. With the FDA path uncertain, US peak sales scenarios that previously assumed approval in 2026 are now meaningfully lower.

The risk-reward has shifted decisively to the downside since the September 2024 HERCULES win. Three signals to watch: 1. **EC decision on Cenrifki (EU)**: Following the April 24, 2026 CHMP positive opinion, the European Commission decision typically lands within ~60 days [16]. This would make tolebrutinib the first approved therapy globally for non-relapsing SPMS - meaningful but commercially modest given EU pricing. 2. **FDA dialogue on CRL path**: Sanofi expected substantive regulatory engagement by end of Q1 2026 [15]. The CRL's language ('benefit-risk could not be established for any patient subpopulation') is more severe than a typical confirmatory-trial CRL and suggests either an additional Phase 3 or substantial restricted-label negotiation. Watch for any disclosure of a Type A/B meeting outcome. 3. **Competitive class evolution**: fenebrutinib (Roche) is the most differentiated competitor because it's non-covalent and reversible, which may improve liver safety [9]. If fenebrutinib reads out cleanly in progressive MS, tolebrutinib's window narrows further. The Principia $3.68B deal looked vindicated in September 2024 and looks much worse now: one missed Phase 3 (PPMS), one missed primary in relapsing MS (GEMINI), and a US rejection. Sanofi's 2024 net sales were ~€41.1B [20]; even a successful nrSPMS launch in EU alone won't move the needle materially. The asset is a defensible specialty MS franchise rather than the platform play the deal economics implied. Peak sales analyst consensus is currently in flux post-CRL and post-PERSEUS; we are not citing a credible figure here pending updated sell-side models. Composition-of-matter patent expiry not verified in this writeup.