GI Oncology Daily Digest

Single-arm Phase 2 (n=30) treating HCC patients with major Vp3 or Vp4 portal vein tumor thrombus — a population routinely excluded from registrational trials — with durvalumab plus hepatic arterial infusion FOLFOX, followed by maintenance durvalumab. The pre-specified primary endpoint was met: 1-year OS rate 63.3% and median OS 13.9 months (95% CI 10.7-NR), roughly double historical benchmarks of ≤8 months for sorafenib monotherapy in this anatomic subgroup. Median PFS 6.0 months and 80% disease control rate further support a meaningful clinical signal. Importantly, single-nucleus RNA-seq identified a MECOM+ resistant hepatocyte subcluster in non-responders and expanded cytotoxic T-cell programs in responders, offering both a candidate predictive biomarker and a mechanistic target for combination strategies. NCT04945720 establishes a foundation for randomized confirmation in PVTT-restricted HCC.

Yamaguchi, Manning, and colleagues at MD Anderson establish MUC13 as a theranostic target across colorectal cancer. Across 72 primary CRC and 100 liver metastasis tissues, MUC13 was highly expressed in approximately 70% of mCRC samples and was inversely correlated with overall survival (P<0.01). A MUC13-targeted radiopharmaceutical therapy using terbium-161 (¹⁶¹Tb), an Auger-electron emitter, delivered significant survival benefit in preclinical mCRC xenograft models, while paired ⁸⁹Zr-PET imaging quantified target density in vivo and supported patient selection. A DNA damage response (DDR) gene signature was identified as a candidate sensitivity predictor for RPT response. With ~70% biomarker prevalence in metastatic disease, broad expression conservation between primary tumors and liver metastases, and a complete diagnostic-therapeutic pair, MUC13-RPT establishes a credible first-in-human pathway for radioligand therapy in colorectal cancer — a tumor type where theranostics has long been underdeveloped.

Intrahepatic cholangiocarcinoma (iCCA) carries 5-year OS under 10% and, beyond FGFR2 fusions and IDH1 mutations, has almost no actionable mechanistic drivers. Lu, Xu, and colleagues identify the NAT10-mediated N4-acetylcytidine (ac4C) RNA modification as a central oncogenic regulator: NAT10 ac4C-modifies and stabilizes CHAF1A mRNA, and CHAF1A then epigenetically silences the HERV9NC transposable element via chromatin remodeling. Loss of HERV9NC-derived double-stranded RNA muffles innate antiviral signaling, drops cytotoxic T-cell infiltration, and accelerates iCCA proliferation. Pharmacologic NAT10 inhibition (Remodelin) restored HERV9NC expression, reignited interferon signaling, increased T-cell infiltration, and slowed tumor growth in both PDX and orthotopic ICC models. The work establishes NAT10-ac4C as a credible druggable axis for iCCA — and an early example of using transposable-element reactivation as an immunotherapy-sensitizing strategy in cold biliary tract tumors.

Khan, Ateeq, and colleagues identify DKC1 (dyskerin pseudouridine synthase 1) as a clinically actionable hub for FOLFOX resistance in colorectal cancer. DKC1 sits in a canonical WNT positive feedback loop driving CRC proliferation, while concurrently dysregulating sphingolipid biosynthesis — downregulating SGPP2 and accumulating very-long-chain ceramides — which collectively confer FOLFOX chemoresistance. Critically, FOLFOX-resistant patient-derived organoids regained chemosensitivity when treated with DKC1 inhibitors or WNT pathway inhibitors, demonstrating direct therapeutic actionability rather than correlative association. DKC1 elevation associates with worse overall survival and with the WNT-enriched CMS2 consensus molecular subtype, making DKC1 simultaneously a prognostic biomarker and a precision therapeutic target in a CRC subset (CMS2) that has historically been hard to drug. The findings position DKC1 and the sphingolipid axis as drug-development priorities for refractory mCRC.

MEND-IT II launches as the first Phase III trial directly comparing FOLFOXIRI-based total neoadjuvant therapy with CAPOX/FOLFOX-based TNT in high-risk locally advanced rectal cancer. The 394-patient Dutch multicenter design (NCT07472868) defines high-risk by any of: mesorectal fascia invasion, EMVI grade IV, tumor deposits, or ≥2 enlarged lateral pelvic lymph nodes (≥7 mm short axis). Both arms include chemoradiotherapy, and the primary endpoint is a composite of pathologic complete response or sustained clinical complete response (cCR) at 1 year, with watch-and-wait offered to cCR patients. The trial directly addresses one of the most consequential open questions in rectal organ preservation: does the triplet add enough cCR / pCR advantage in the highest-risk patients to justify its incremental toxicity, or is doublet-based TNT sufficient? Results remain several years off, but this is precisely the model trial the field has been waiting for in high-risk LARC.