STRC h01 Fenamic Scaffold Tox Audit 2026-04-23
tl;dr — verdict
Parent fenamates (niflumic / flufenamic / meclofenamic / mefenamic / tolfenamic) are NOT developable as-is for an 8-y-o DFNB16 patient. Ion-channel promiscuity (TRPM4, Cx50, BK, KCNQ-like, TMEM16A) at concentrations (2–60 µM) overlapping our pocket Kd window (~30 µM) creates unacceptable cochlear off-target risk. Mefenamic adds a pediatric seizure signal.
The scaffold itself is developable via tafamidis-style bioisosteric optimization — same path Pfizer took from flufenamic → tafamidis for TTR. Anthranilic N–H + carboxylate retained, distal aryl halides replaced with polar small groups, COX-essential carboxylate moved off-plane.
Action for Phase 4h: use Phase 3c v2 top hits (niflumic, flufenamic, sulfasalazine) as chemical probes for MD-based pocket validation only. Do NOT commit wet-lab triage to parent compounds. After probe confirms pocket geometry, hand pharmacophore to medchem for a tafamidis-style bioisosteric replacement.
Compound-by-compound safety table
| Compound | Regulatory status | Verdict | Decisive flag |
|---|---|---|---|
| Mefenamic acid (Ponstel) | FDA-generic-only, brand discontinued | CAUTION (KILL for peds) | Seizures in 38% of overdoses ≥2.5 g (Kamour 2017) |
| Meclofenamic acid (Meclomen) | Essentially off-market since 2010s | KILL as drug, keep as probe | Kv2.1 IC₅₀ 56 µM, KCNQ2/3 opener — pan-neuronal channel hit |
| Niflumic acid (Nifluril) | National approvals EU (FR/BE/IT/ES/PT/AT); never FDA | CAUTION | TMEM16A/CaCC IC₅₀ 12 µM (stria vascularis); UGT1A9 DDI 2.5 µM |
| Flufenamic acid (Arlef) | Oral discontinued, topical OTC-EU; never FDA-oral | CAUTION (richest probe) | Most promiscuous ion-channel modulator in class; TTR pharmacochaperone precedent (PMID 10465408) |
| Tolfenamic acid (Clotam) | UK/EU migraine indication; never FDA | CAUTION | LiverTox cross-reactivity warning with mefenamic |
Hepatotoxicity
- Mefenamic: LiverTox likelihood D, rare idiosyncratic hypersensitivity (NBK548029 / PMID 31643361). Severe cases documented; no chronic injury. One pediatric case: 4-y-o, ALT 1920 U/L at 2 wks (herb-confounded).
- Class cross-reactivity explicitly noted in LiverTox: “avoid other anthranilic-acid derivatives after mefenamic-induced injury.”
- No fenamate has been formally withdrawn for hepatotoxicity (c.f. bromfenac/ibufenac/benoxaprofen — PMID 16456879). Liver signal is real but not first-tier.
Pediatric safety — decisive block
- Mefenamic CNS overdose profile uniquely severe: seizures in 38% of overdoses, onset 4.4 h, threshold ≈ 2.5 g, 4-h plasma ≥ 21 mg/L convulsed a 13-y-o girl (Kamour 2017 BJCP, PMID 28066875 / PMC5346865). Hypothesis: PG/TXA depletion + GABA-A modulation lowers seizure threshold.
- No Reye-syndrome signal for fenamates (Reye is aspirin-specific; Springer Pediatr Drugs 2022, doi 10.1007/s40272-022-00514-1).
- Mefenamic not licensed for children (UK BNF, Ponstel label). Tolfenamic/niflumic have had historical pediatric use in EU.
Cochlear / ototoxicity — critical for DFNB16
No case reports link any fenamate to sensorineural HL, tinnitus, or vestibular injury (Kyle/Wang/Shin 2015 OHNS 152(3):393-409, PMID 25560405 / PMC4465569 — reviewed 23 NSAIDs, zero fenamates surfaced).
Class-level NSAID ototoxicity weakly positive: Curhan 2010 (men, PMID 20193831) HR 1.21 any NSAID; Curhan 2016 (women, PMID 27974293) aspirin NOT associated — suggests reversible-inhibition mechanism is tolerable, irreversible NSAIDs (indomethacin/piroxicam/naproxen) drive risk. Mechanism: COX inhibition → reduced cochlear blood flow → strial degeneration — directly antagonistic to Misha’s therapeutic goal.
Fenamate-specific cochlear off-target list (from meta-review PMC12530847, DOI 10.1002/cmdc.202500301, PMID 40600901):
| Target | Best IC₅₀/EC₅₀ in class | Cochlear relevance |
|---|---|---|
| TRPM4 | 2.8 µM (FFA), 3.4 µM (MCFA) | Cardiac/neuronal but hair-cell Ca²⁺ signaling possible |
| Cx50 (gap junction) | 3 µM (FFA) | Cochlear gap junctions — stria/OHC coupling |
| TMEM16A/ANO1 | 12 µM (NFA) | Stria vascularis expressed |
| BK channels | 25–300 µM | OHC critical |
| Kv2.1 | 56 µM (MCFA) | Spiral ganglion |
| KCNQ2/3 opener | MCFA | KCNQ4 mutations = DFNA2 HL — direct paralogy |
| KATP | 25 µM (MFA) | Hair cell metabolic sensing |
At Kd ≈ 30 µM (our Phase 3c v2 hits) the on-target concentration overlaps TRPM4/Cx50/BK at or below effective off-target concentrations. On-target window is crowded.
CYP / UGT inhibition
| Enzyme | Mefenamic | Flufenamic | Meclofenamic | Niflumic |
|---|---|---|---|---|
| CYP2C9 substrate | Major (3-OH) | ✓ | ✓ | ✓ |
| CYP2C9 inhibition | Moderate (warfarin INR ↑) | Inhibitor | Inhibitor | Lower |
| UGT1A9 | — | — | — | 2.5 µM selective |
| UGT (mycophenolic) | IC₅₀ 63 ± 8 µM | IC₅₀ 19 ± 9 µM | — | IC₅₀ 8 ± 1 µM (PMID 11214772) |
DDI implication for 8-y-o chronic dosing: warfarin, methotrexate, phenytoin, valproate would need monitoring. Manageable but not trivial.
Fenamate → pharmacochaperone precedent (the good news)
Flufenamic acid was the lead scaffold for tafamidis discovery (Ottonello 1999 BMC 10465408 → Pfizer clinical program → FDA-approved Vyndaqel 2019, transthyretin amyloid cardiomyopathy). Iododiflunisal is the cleaner diphenyl-ether analog in the same mechanism class. This is the direct precedent for h01: fenamic-family hit → medchem optimization → pediatric-acceptable chaperone drug.
Tafamidis-specific lessons:
- Anthranilic N–H and carboxylate retained as pharmacophore.
- Benzoxazole bioisostere replaced the COX-essential geometry → killed the COX off-target.
- 3-position substituent on the anthranilic ring tuned for selectivity.
- Pediatric use cleared safety.
Recent (2020–2026) scaffold activity
- Fenamate → NLRP3 inflammasome inhibition → AD/glioblastoma clinical trials (Daniels lineage, reviewed PMC12530847 2025).
- Pentafluorosulfanyl-FFA analogs (PMC4599580) — channel/COX dual.
- Niflumic → TEAD inhibitors 2024 (doi 10.1021/acs.jmedchem.4c…, ResearchGate 385518489).
Design constraints for h01 Phase 4h→Phase 7 medchem
Two non-negotiables if we carry the fenamic scaffold forward:
- Kill COX: swap the planar 2-arylamino-benzoic-acid motif for a tafamidis-style benzoxazole or benzothiazole. Preserves the H-bond acceptors, breaks the COX pharmacophore.
- Kill the cochlear ion-channel panel: replace distal aryl halides (3-CF₃ / 2,6-diCl / 3-Cl-2-Me) with polar small groups (CN, OMe, F). TRPM4/Cx50/BK affinity tracks with lipophilic halogenated distal ring. Tolerable tradeoff on Vina ΔG if Phase 3c v3 covalent branch or Phase 3c v4 fragment-growing finds alternative anchoring.
Optional bioisosteres to test in Phase 3c v4:
- Anthranilic carboxylate → tetrazole / acylsulfonamide (already in Phase 3c v3 library).
- 2-arylamino → 2-aryloxy (diphenyl-ether = fenoprofen class). Loses N–H donor — only acceptable if Phase 5b/5d docking poses do not require N–H H-bond to K1141 backbone.
- Benzoxazole N replacing anthranilic N-H (tafamidis move).
Impact on h01 next steps
- Phase 4h validation — proceed on the three Phase 3c v2 hits as chemical probes, not wet-lab candidates. MD + MM-GBSA on niflumic / flufenamic / sulfasalazine to confirm pose geometry and H-bond pattern.
- Phase 4i new — pharmacophore extraction from validated probe pose + medchem briefing document for bioisosteric replacement (tafamidis-style).
- Phase 8 wet-lab SOP — design for tafamidis-analog class, not fenamate class. Include iododiflunisal as benchmark positive control; consider tafamidis analog synthesis path.
- Phase 6c selectivity audit — expand from proteome-wide Lys pockets to cochlear ion-channel panel (TRPM4, Cx50, BK, KCNQ4, TMEM16A). Use CheMBL binding data + AF3 structures where available.
Primary sources
- LiverTox: Mefenamic Acid — NBK548029 / PMID 31643361
- Kamour 2017 BJCP, mefenamic CNS overdose: PMID 28066875 / PMC5346865
- Kyle 2015 OHNS, NSAID ototoxicity systematic review: PMID 25560405 / PMC4465569
- PMC12530847, fenamate channel-pharmacology review 2025: DOI 10.1002/cmdc.202500301 / PMID 40600901
- Curhan 2010 (men): PMID 20193831 / PMC2831770; Curhan 2016 (women): PMID 27974293 / PMC5209586
- Ottonello 1999 BMC, FFA→TTR pharmacochaperone: PMID 10465408
- Peretz 2005 Mol Pharmacol, MCFA→KCNQ2/3: PMID 15598972
- Peretz 1999, MCFA→Kv2.1: PMID 10493112
- Tafamidis FDA approval history (Vyndaqel, 2019) — Pfizer
Ranking delta
Scaffold-level audit, no compute. A-tier hypothesis intact. Impact on h01 workstream:
- h01 NORMAL-monotherapy path: already RED at chemistry (Phase 5b/3c v2). Tox audit adds scaffold-kill risk for parent fenamates → MILD-MODERATE adjunct path also needs medchem optimization, not just Kd improvement.
- Framing update: h01 now requires (a) pocket-validated probe (niflumic as MD-tested chemical probe), (b) tafamidis-style bioisosteric optimization (new work), (c) cochlear ion-channel selectivity panel (Phase 6c expansion), (d) wet-lab SOP designed for optimized analog class not fenamate parents.
- All hypothesis scores unchanged (no mechanism change). Risk profile sharpened.
Connections
[part-of]h01 hub[see-also]STRC h01 Phase 3c v2 Expanded Screen 2026-04-23[see-also]STRC Pharmacochaperone Phase 4 Plan[about]Misha- STRC h01 Phase 8 Wet-Lab Triage SOP