Command Palette
Search for a command to run...
Brimarafenib plus mirdametinib: a vertical MAPK doublet stopped for tolerability
A Phase 1/2a study of the RAF dimer inhibitor brimarafenib with the MEK inhibitor mirdametinib in advanced solid tumors was terminated after 23 patients for combination tolerability, a therapeutic-index failure on a heavily validated pathway.
Mechanism Risk Score
| Component | Points |
|---|---|
| Phase-weighted failure burden | 4.7 / 40 |
| Archetype severity | 8.5 / 25 |
| Temporal recency | 4.3 / 15 |
| Genetic evidence deficit | 2.7 / 15 |
| Programmatic saturation | 2.5 / 5 |
For BRAF in BRAF/NRAS-mutant melanoma and advanced solid tumors, the Mechanism Risk Score is 23/100 (green band). The score is a failure-burden index derived from Claidex post-mortems on this target–disease pair, not a probability of approval.
MRS 23/100 (GREEN). 1 programs across BRAF have been documented for BRAF in BRAF/NRAS-mutant melanoma and advanced solid tumors: 0 Phase 3, 0 Phase 2, 0 Phase 1 — of which 0 were efficacy failures, 1 safety, 0 biomarker, and 0 operational (enrollment, sponsor, or funding). The most informative failure on file is Brimarafenib plus mirdametinib: a vertical MAPK doublet stopped for tolerability. This score quantifies the documented failure burden; the Open Targets association score of 0.82 provides strong human-genetic anchoring that partially offsets the documented failure record. The MRS is not a prediction of future trial outcomes — it is a structured summary of the empirical record, recomputed live from the Claidex claims table, and intended to flag mechanisms where any new program must explicitly resolve each prior failure mode before pursuit is justified.
This score does not predict whether the next trial will succeed. It flags how heavy the documented mechanistic failure record is before a new program is justified.
What was tried
SpringWorks Therapeutics ran a Phase 1/2a open-label, dose-escalation and expansion study (NCT05580770) that paired brimarafenib (BGB-3245), an oral next-generation RAF dimer inhibitor, with mirdametinib, an oral MEK1/2 inhibitor, in adults with advanced solid tumors [CT.gov NCT05580770]. The study enrolled 23 patients and was then terminated [CT.gov NCT05580770 results]. The posted reason for stopping was lack of tolerability of brimarafenib when given in combination with mirdametinib [CT.gov NCT05580770]. The design was non-randomized and sequential with no masking, and the Part 1 primary outcomes were treatment-emergent adverse events, maximum tolerated dose, and the recommended Phase 2 dose [CT.gov NCT05580770]. Brimarafenib came out of MapKure, a BeiGene and SpringWorks joint venture, and mirdametinib (PD-0325901, ChEMBL507361) is the MEK1/2 inhibitor cleared in 2025 for neurofibromatosis type 1 plexiform neurofibromas [ChEMBL CHEMBL507361].
The biological hypothesis
BRAF sits in one of the best-validated oncology pathways. The Open Targets association score for BRAF in melanoma is 0.819, with a genetic-association component of 0.699 and a somatic-mutation component of 0.796 [Open Targets, EFO_0000756]. First-generation BRAF inhibitors only block RAF monomers and can paradoxically activate the MAPK pathway in RAS-mutant or RAF-dimer-driven tumors. Brimarafenib was built as a paradox breaker that inhibits RAF dimers, which extends the rationale to NRAS-mutant and class II/III BRAF tumors [Pigment Cell Melanoma Res 2025, 10.1111/pcmr.70062]. Adding a MEK inhibitor downstream was meant to clamp pathway reactivation, the same vertical logic that made BRAF plus MEK doublets standard in BRAF V600 melanoma. Preclinically, brimarafenib combined with mirdametinib suppressed NRAS-mutant melanoma growth [Pigment Cell Melanoma Res 2025, 10.1111/pcmr.70062].
What actually happened
Of 23 enrolled patients, 16 (70%) had serious adverse events and 3 (13%) died during the reporting window [CT.gov NCT05580770 results]. The most common treatment-emergent adverse events across the 23-patient population were fatigue (52%), diarrhoea (52%), nausea (48%), blood creatine phosphokinase increased (48%), dermatitis acneiform (35%), and vomiting (30%) [CT.gov NCT05580770 results]. Serious events included pyrexia in 4 patients and a case of rhabdomyolysis [CT.gov NCT05580770 results]. The creatine phosphokinase elevation in nearly half of patients, together with the rhabdomyolysis case, tracks the muscle toxicity that is characteristic of MEK inhibition, while the acneiform dermatitis and diarrhoea are on-target MAPK class effects. No combined dose that paired active exposures of both drugs with acceptable tolerability was established, so the study closed on tolerability rather than on a defined efficacy readout [CT.gov NCT05580770].
Failure mechanism, best guess
The most parsimonious read is a therapeutic-index collapse from additive on-target MAPK toxicity, not target invalidity. Both agents act on the same pathway at adjacent nodes, so their class toxicities (skin, gastrointestinal, muscle and creatine phosphokinase, ocular) stack and force dose reductions below the exposure needed for durable dual suppression. BRAF remains heavily validated (Open Targets melanoma association 0.819), which argues against a biology problem [Open Targets]. The constraint was the joint-dose window. Mirdametinib post-marketing reports in openFDA FAERS are led by dermatitis acneiform, diarrhoea, nausea, and dehydration, the same toxicities that dominated the combination [openFDA FAERS, mirdametinib]. A combination that suppresses NRAS-mutant melanoma in mice ran into a human tolerability ceiling once both nodes were inhibited continuously.
How to prevent this next time
Two quantitative tools would have changed the odds. First, a Bayesian dose-finding design that estimates the posterior probability of a dose that is both tolerable and pharmacodynamically active, updated after every cohort. The posterior probability of success can be written as
and a pre-set stopping rule (for example, halt if the posterior probability of an active-and-tolerable dose falls below 0.10) converts a vague tolerability signal into a decision. Second, historical base-rate adjustment from prior vertical MAPK doublets, where BRAF plus MEK and pan-RAF plus MEK combinations have repeatedly shown overlapping skin, gastrointestinal, and creatine phosphokinase toxicity, should have set a low prior on a continuous-daily combined dose clearing the bar. Translational validation sequencing of intermittent or pulsatile schedules, tested against the same NRAS-mutant models before the human study, would have given the design a toxicity-sparing schedule to escalate. The single highest leverage change would have been to run an adaptive Bayesian dose-escalation that prioritized intermittent or pulsatile dosing of brimarafenib and mirdametinib before committing to continuous daily exposure of both drugs.
What this means for similar programs
Vertical same-pathway doublets carry a predictable overlapping-toxicity tax. Any RAF dimer plus MEK, pan-RAF plus MEK, or analogous two-node combination should pre-register the combined-dose tolerability hypothesis as the primary risk, not the efficacy hypothesis. Schedule design is the lever. Intermittent dosing, node-alternating schedules, and tumor-selective delivery are the realistic ways to recover a window when both drugs share class toxicities. The Claidex Mechanism Risk Score for BRAF is 23 (band green), with a low genetic-evidence deficit (2.72 of 15) and most of the burden coming from the combination-tolerability archetype (8.52 of 25). The single early-phase failure clusters around delivery and tolerability rather than biology.
Open questions
Would an intermittent schedule have opened a tolerable-and-active window for the doublet, as it has for some BRAF plus MEK regimens. Was there any antitumor signal in the 23 treated patients, given that objective response rate was a Part 2 endpoint that was never reached [CT.gov NCT05580770]. Does brimarafenib retain a monotherapy path in NRAS-mutant or class II/III BRAF tumors, where the dimer-breaker rationale is strongest. And does the muscle toxicity signal reflect mirdametinib exposure specifically, which would favor a different MEK partner.
Sources
- ClinicalTrials.gov, NCT05580770, study record and posted results: enrollment 23, termination reason, serious adverse events 16 of 23, deaths 3 of 23, adverse-event frequencies (fatigue 52%, diarrhoea 52%, nausea 48%, creatine phosphokinase increased 48%, dermatitis acneiform 35%, vomiting 30%), pyrexia and rhabdomyolysis serious events. https://clinicaltrials.gov/study/NCT05580770 - Open Targets Platform, BRAF (ENSG00000157764) and melanoma (EFO_0000756): overall association 0.819, genetic 0.699, somatic mutation 0.796. https://platform.opentargets.org/target/ENSG00000157764 - ChEMBL, mirdametinib CHEMBL507361 (PD-0325901), MEK1/2 inhibitor. https://www.ebi.ac.uk/chembl/explore/compound/CHEMBL507361 - openFDA FAERS, mirdametinib reaction counts (dermatitis acneiform, diarrhoea, nausea, dehydration). https://api.fda.gov/drug/event.json - Mukhopadhyay et al., Pigment Cell and Melanoma Research, 2025. Combination of the novel RAF dimer inhibitor brimarafenib with the MEK inhibitor mirdametinib is effective against NRAS mutant melanoma. https://- Karoulia et al., Cancer Cell review on RAF dimer biology and paradoxical activation, supporting the paradox-breaker rationale. https://- Mechanism Risk Score for BRAF computed from the Claidex failure graph: total 23 (phase 4.7, archetype 8.52, recency 4.25, genetic 2.72, saturation 2.5). Saturation uses the Claidex failure-graph program count of 2. For external context, ChEMBL lists 14 distinct BRAF programs (CHEMBL5145).