Animal data (clean profile)

• No overt toxicity at therapeutic doses (0.25-2.5 mg/kg) in mouse models or non-human primates.
• No pancreatic toxicity (Halliday 2015 — direct comparison vs. PERK inhibitor).
• One Alzheimer's model: 5 mg/kg mortality signal — therapeutic window has an upper bound.

Human data (limited to fosigotifator analog, not ISRIB itself)

From AbbVie/Calico Phase 1 fosigotifator trials (n≈100+ across studies, up to 14 days oral):

• Mild-to-moderate AEs only, similar to placebo
• No QTc signal in 72-subject cardiac safety study
• No food-effect on PK
• No CYP-mediated DDI with rosuvastatin or digoxin at tested doses
• HEALEY ALS trial (n=300, longer dosing): treatment-emergent AEs similar between active and placebo arms

Important caveat: this is fosigotifator (ABBV-CLS-7262), a structurally optimized analog. ISRIB itself has zero formal human safety data.

Theoretical / emerging concerns

• Ubiquitin-proteasome system impairment (Nature Communications Biology 2024). Chemical inhibition of the ISR via ISRIB impaired the UPS in cell models. Implication: chronic ISR suppression may exacerbate misfolded-protein accumulation — paradoxically promoting the kind of pathology ISR activation is supposed to defend against. This is the strongest published-evidence concern against long-term ISRIB use, especially in older brains predisposed to aggregation pathology. Worth taking seriously, not theoretical.
• Specificity / off-target activity: Bentham Medicinal Chemistry review (2024, "Maximizing ISRIB Potential…") flagged that ISRIB's specificity profile is incompletely characterized; long-term safety data is non-existent in humans; disease-specific dosing windows likely vary.
• Pre-existing protein-aggregation pathology: ALS, AD, PD — paradoxical worsening risk. ISR is partially protective in these states; suppressing it long-term may shift the balance unfavorably. (Counter: in animal prion / DS / VWM models, ISRIB was protective. The story is not settled.)
• Vendor identity / purity risk (separate from molecular safety): research-chem ISRIB is sold by vendors with variable QC. Trans- vs. cis- isomer matters (only trans- is biologically active at the canonical site). COA verification, third-party HPLC + mass-spec confirmation of identity is the bare minimum due diligence.

Specific watch periods for users in this archetype if he runs a research-chem trial

• First 7 days: any GI symptoms (nausea, diarrhea), unusual fatigue, infection-like symptoms (theoretical UPS / immune-modulation concern). Mood / sleep tracking.
• Days 7-28: subjective cognition log, training quality, recovery, sleep architecture.
• Stop triggers: any new neurologic symptom (paresthesia, vision change, cognition worse than baseline), GI illness lasting >48 h, mood change.
• Hard limit on cumulative exposure: 4 weeks per cycle, no more than 2 cycles in any 6-month window pending better long-term safety data.

Contraindications (constructed; no formal label exists)

• Any active neurodegenerative diagnosis (AD, ALS, PD, prion) — paradox risk
• Pregnancy / lactation — no data
• Active infection — UPS concern + immune modulation theoretical
• Concurrent investigational ISR-axis drugs

For the user: none of the above apply. He's the lowest-risk plausible user.