Worksheet — Top 6 BRAF Fusion Partners Responsive to Plixorafenib

What this is: a patient-selection / enrichment worksheet. A BRAF fusion is a gene rearrangement in the tumor — the partner is the gene joined to BRAF (a patient genotype), not a drug combination. Plixorafenib is a single drug (monotherapy paradox-breaker). We are ranking which fusion genotype is most likely to respond.

The core principle (drives every score): response tracks structure, not partner identity — a fusion that (1) retains the intact BRAF kinase domain, (2) loses the N-terminal RBD/autoinhibitory region, and (3) gains a partner-contributed dimerization (coiled-coil) domain = a RAS-independent Class-2 dimer a paradox-breaker can disrupt. With a dimer-capable inhibitor, response is largely partner-agnostic once the BRAF kinase domain is retained — so KIAA1549 leads mainly on prevalence + validation, and partners #2–6 are a close, biology-driven cluster with thin partner-specific data.

Scoring rubric — score each 0–3; edit any cell, totals update live

A · Prevalence / access

How many patients carry it (population size to enrich)

B · BRAF kinase domain retained

True BRAF-domain fusion (not RAF1/CRAF)

C · Dimerization-competent partner

Partner supplies a coiled-coil/oligomerization domain

D · Preclinical response evidence

Shown to respond to a paradox-breaker / dimer inhibitor

E · Clinical / class validation

Response in patients (tovorafenib class, or plixorafenib)

Scale: 0 = none/no · 1 = weak/limited · 2 = moderate · 3 = strong/yes | Max total = 15

#	Fusion partner (genotype)	Total
1	KIAA1549–BRAF 66–90% of pediatric LGG / pilocytic astrocytoma; #1 fusion pan-tumor (25%). Strong coiled-coil dimer. PLX8394 disrupts it preclinically; tovorafenib FIREFLY-1 cohort was 74% KIAA1549 (67% ORR). Strong	15 / 15
2	TRIM24–BRAF Recurrent, ~#3 partner pan-tumor. TRIM24 supplies a coiled-coil → oligomerizes. Validated only as part of the BRAF-fusion group; no partner-specific clinical data. Moderate	9 / 15
3	AGK–BRAF 2nd most frequent partner pan-tumor; pediatric thyroid ~10%, melanoma. Dimerization caveat: AGK–BRAF did NOT drive paradoxical activation in Botton 2019 (may dimerize less) — confirm before prioritizing. Moderate	8 / 15
4	SND1–BRAF Recurrent (solid tumors, glioma, gastric). SND1 Tudor-SN provides an oligomerization domain. Class-validated only. Moderate	8 / 15
5	FAM131B–BRAF Early-described recurrent glioma fusion (pediatric LGG). Partner contributes dimerization. Class-validated only. Moderate	8 / 15
6	AKAP9–BRAF Papillary thyroid carcinoma (enriched in radiation-induced). AKAP9 is a large coiled-coil scaffold. Class-validated only; niche population. Moderate	8 / 15

De-prioritize (not in the top 6): RAF1 / CRAF fusions (e.g., SRGAP3–RAF1) — these fuse a different kinase (CRAF, not BRAF), dimerize differently, and are the notable exception likely to respond less to a BRAF-targeted paradox-breaker. Score B = 0 for these.

Honest gap: there is no published plixorafenib clinical response in a confirmed fusion patient yet — every "clinical" score above rests on the class (tovorafenib FIREFLY-1) plus plixorafenib preclinical data. FORTE sub-protocol D is the readout that would upgrade column D/E for plixorafenib specifically.

How to use

Edit any score to reflect your own read (totals recompute live), set an Enroll? decision per genotype, and add Team notes. Print/PDF for the enrichment discussion. Suggested action: prioritize BRAF-kinase-domain fusions with dimerization-competent partners (KIAA1549 first), confirm the AGK dimerization caveat, exclude RAF1/CRAF fusions, and differentiate on monotherapy (no MEK combo) vs belvarafenib/naporafenib.

Sources & references

What this sheet is built from. Unlike the CRC duration benchmark (which is registry-sourced), every score here is literature-derived. The six publications below are the complete source set. Each was resolved to its PubMed record and verified on 2026-07-09 via NCBI E-utilities — titles, journals, years and PMIDs come from the registry, not transcribed by hand. The claim → source map that follows shows what each citation actually supports, including the claims that currently have no citation at all.

Sievert AJ, Gerrish K, Hazlin K, et al. Paradoxical activation and RAF inhibitor resistance of BRAF protein kinase fusions characterizing pediatric astrocytomas. Proc Natl Acad Sci U S A. 2013;110(15):5957–5962. PMID 23533272
Foundational for this sheet's core principle: KIAA1549–BRAF signals as a RAS-independent constitutive dimer, and first-generation RAF inhibitors paradoxically activate it. Newly added — the principle was previously uncited.
Ross JS, Wang K, Chmielecki J, et al. The distribution of BRAF gene fusions in solid tumors and response to targeted therapy. Int J Cancer. 2016;138(4):881–890. PMID 26314551
Source for pan-tumor partner prevalence (column A): KIAA1549 as the #1 fusion, AGK as the second most frequent.
Jain P, Fierst TM, Han HJ, et al. Overcoming resistance to single-agent therapy for oncogenic BRAF gene fusions via combinatorial targeting of MAPK and PI3K/mTOR signaling pathways. Oncotarget. 2017;8(49):84697–84713. PMID 29156677
Cited on this page as “PLX8394 vs KIAA1549-BRAF (2017)”. Read the title carefully — see the claim map below.
Yao Z, Gao Y, Su W, et al. RAF inhibitor PLX8394 selectively disrupts BRAF dimers and RAS-independent BRAF-mutant-driven signaling. Nat Med. 2019;25(2):284–291. PMID 30559419
The dimer-disruption mechanism the whole sheet rests on; also the basis for de-prioritising RAF1/CRAF fusions (BRAF-selective, spares CRAF/ARAF). This page labels it “Yao 2018” — published online Dec 2018, print Feb 2019.
Botton T, Talevich E, Mishra VK, et al. Genetic Heterogeneity of BRAF Fusion Kinases in Melanoma Affects Drug Responses. Cell Rep. 2019;29(3):573–588.e7. PMID 31618628
Source of the AGK–BRAF dimerization caveat (column C). Note the cohort is melanoma, not glioma.
Kilburn LB, Khuong-Quang DA, Hansford JR, 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(1):207–217. PMID 37978284
Class validation (column E): the 74% KIAA1549 cohort and the 67% ORR. Criterion not stated on this page — see the claim map.

Claim → source map

Claim on this page	Source	Status	Note
Response tracks structure, not partner identity — the governing principle	Sievert 2013 · Yao 2019	✓ supported	Mechanistic inference from the dimer biology. Sound, but note that no plixorafenib–fusion co-structure exists.
KIAA1549–BRAF is the #1 fusion pan-tumor (25%)	Ross 2016	✓ supported	Directly reported.
66–90% of pediatric LGG / pilocytic astrocytoma	—	✗ uncited	No citation on this page, and flagged as overstated in the Mechanism & Biomarker Verification Record: the 66–90% band applies to pilocytic astrocytoma, not to all pLGG (~30–60%). Ross 2016 does not support it. Human re-scope pending; figure left unchanged here.
FIREFLY-1 cohort was 74% KIAA1549, 67% ORR	Kilburn 2024	⚠ partial	Both figures are correct, but the response criterion is not named. ORR is 67% by RANO-HGG and 51% by RAPNO. Specify which, wherever it is quoted.
AGK–BRAF did not drive paradoxical activation (may dimerize less)	Botton 2019	✓ supported	Reported in melanoma models. The sheet's own “confirm before prioritizing” caveat is the right posture.
PLX8394 disrupts KIAA1549–BRAF preclinically	Jain 2017	⚠ partial	Jain 2017's actual finding is that single-agent PLX8394 meets resistance in BRAF-fusion models and requires combinatorial MAPK + PI3K/mTOR targeting. That sits in tension with this sheet's suggested action to “differentiate on monotherapy (no MEK combo).”
A RAS-independent “Class-2” dimer	Sievert 2013	⚠ partial	The biology is right; the “Class-2” label was removed from the BRAF·Plixorafenib 3D page as a house-style call and remains here. Settle the nomenclature across both.
RAF1/CRAF fusions de-prioritized (score B = 0)	Yao 2019	✓ supported	PLX8394 disrupts BRAF-containing dimers while sparing CRAF/ARAF; a CRAF fusion lacks the kinase domain the drug engages.

Three open items, carried from the Mechanism & Biomarker Verification Record (2026-07-08), still awaiting human sign-off. (1) The 66–90% prevalence figure is uncited and overstated — it must be scoped to pilocytic astrocytoma. (2) The 67% ORR needs its criterion named (RANO-HGG 67% vs RAPNO 51%). (3) “Class-2” nomenclature is inconsistent with the 3D page. None of these changes the top-6 ranking — KIAA1549 leads on prevalence and class validation either way — but all three must be resolved before external use. Figures on this page are left unchanged pending that review.

Honest gap, restated. There is no published plixorafenib clinical response in a confirmed BRAF-fusion patient. Every column E score rests on tovorafenib class-validation (Kilburn 2024) plus plixorafenib preclinical data (Jain 2017, Yao 2019). No plixorafenib–fusion co-structure has been deposited; pocket identity between V600E and fusions is an inference from kinase-domain sequence conservation. FORTE sub-protocol D is the readout that would upgrade columns D/E for plixorafenib specifically.