OncoRef 1.24.0: TCR-T clinical evidence and serious events

The human selection remains 522 CTA proteins (558 member gene loci), 18,031 Unsafe, 20 Neither and 1,290 unassigned. There are no category changes from the pinned 1.23 selection. This release adds a reviewed clinical-evidence layer and shows how that evidence changes individual RNA ceilings.

Full curation report and funnel · TCR-T serious-event review · Nine-page clinical PDF · Logic slides · Rules and API

What the clinical review covers

Eight TCR-T products, nine treated cohorts and eleven sentinel events are linked to their source papers, exact epitopes, HLA restrictions and protein origins. Fifty-five peptide/isoform links were checked against the pinned Ensembl 115 protein sequences. Twelve original paper/protocol artifacts use the existing checksum-pinned reference-data API and cache layout.

This is a targeted review, not an exhaustive list of trials or all SAEs. SAE seriousness and CTCAE severity are recorded separately. The report distinguishes treatment-related injury from evidence that a receptor recognized healthy tissue:

• The MAGE-A3 a3a cardiac deaths implicated unintended recognition of titin.
• The separate NCI MAGE-A3/A9 receptor neurological cases implicated the cross-recognized MAGE-A12 peptide.
• c259's late cardiac death and Guillain–Barré cases remain explicit caveats without an established peptide-specific injury mechanism.
• TBI-1301's grade 3 lung injury retains unresolved specificity and blocks credit for that product.

How the RNA screen changes

A reviewed treated cohort without an identified grade 4–5 specificity-related signal can support one existing RNA ladder step, capped at Enhanced-negative. Trials do not stack. Missing or unresolved attribution does not count as a favorable result. Protein and clinical-injury exclusions remain independent.

Tissue tier	Baseline RNA ceilings	With qualifying clinical credit
Vital	4 / 8 / 16 / 32	8 / 16 / 32 / 32
Protected	8 / 16 / 32 / 64	16 / 32 / 64 / 64
Manageable	16 / 32 / 64 / 128	32 / 64 / 128 / 128

Each ladder follows missing/unusable, Approved-negative, Supported-negative, Enhanced-negative IHC. Values are nominal TPM/nTPM research parameters, not measured clinical toxicity limits; corneal CP10K keeps its separate /100 scale. The credit does not invent or upgrade IHC evidence, waive coverage, or establish cancer or Reproductive expression.

Five products currently qualify: c259/lete-cel, IMA203, afami-cel, IMA202 and ADP-A2M10. Their evidence covers six protein profiles and seven member gene loci. Every member of a collapsed protein must be supported; there is no family, alias-prefix or cross-species transfer. Exact peptide/product limits remain visible.

CTAG2 was already CTA. HPA v23 heart RNA is 5.1 nTPM with Enhanced-negative IHC and a 32 nTPM ceiling. The new credit cannot raise an already capped ceiling. c259 recognizes SLLMWITQC shared by NY-ESO-1 and LAGE-1A; this does not establish safety for every CTAG2 isoform, epitope or receptor.

API and reproducibility

from oncoref import tcr_trials

tcr_trials.products()  # product-level credit, exact targets and limitations
tcr_trials.cohorts()   # treated denominators and reviewed attribution
tcr_trials.events()    # sentinel events; unknown grades remain missing
tcr_trials.protein_credit(["ENSG00000126890"])  # CTAG2

oncoref tcr-trials --table all exports the complete ledger offline. Existing selection explanations now show baseline and effective RNA ceilings and the supporting product IDs. Pin package 1.24.0, policy oncoref.cta-selection.2026-10-29.4, and the returned source snapshot.

The clinical report includes the baseline snapshot, exact membership comparison, context-level threshold effects, source pins and reproduction scripts. These are research classifications with epitope-specific clinical evidence, not clinical clearance of whole proteins.