Top 5 Papers
#1
Source: Journal of Clinical Oncology (May 5, 2026) | Authors: Gögenur I, Justesen TF, Tarpgaard LS, Pfeiffer P, Qvortrup C, et al. — Danish Colorectal Cancer Group, Zealand University Hospital + Odense + Aarhus + Aalborg + Vejle + Herlev + Bispebjerg + Hvidovre + Horsens + Randers + Rigshospitalet | Published: May 5, 2026
Score: 15/20 — Base 8 (JCO) + 2 (Phase II) + 3 (potential new standard of care for localized dMMR colon cancer / paves nonoperative management pathway) + 1 (biomarker-guided / dMMR) + 1 (Colleague Engagement: @DraMartinezLago surfaced via X)
Multicenter investigator-initiated Phase II from the Danish Colorectal Cancer Group (NCT05662527), enrolling 85 patients with resectable Stage I-III dMMR colon cancer (median age 74) between Feb 2023 and March 2024. All patients received a SINGLE cycle of pembrolizumab (4 mg/kg, max 400 mg, every-6-weeks dosing), followed by preoperative endoscopy with biopsies and surgery 3-5 weeks later. Primary endpoint pathological complete response (pCR) was achieved in 44% (37/84; 95% CI 33-55), and major pathological response (MPR) in 57% (48/84; 95% CI 46-68) — comparable to multi-cycle regimens at a fraction of the cost and toxicity. Stage-stratified pCR was 61% (Stage I-II) vs 33% (Stage III; p=0.02). Grade 3 AEs occurred in only 11% (9/85), with three treatment-related events. Tragically, two patients (2%) died of post-operative complications within 30 days. At median 18.4 mo follow-up, just one patient had a recurrence — overall survival 98%, disease-free survival 96%. Endoscopy + biopsy after a single cycle had 86% accuracy for predicting pCR — a critical signal for organ-preserving nonoperative management in dMMR colon cancer, mirroring the watch-and-wait paradigm in rectal cancer. Bottom line: a single dose of pembrolizumab can produce pCR in nearly half of dMMR colon cancer patients within weeks. This trial reframes the entire treatment algorithm for the ~15% of localized colon cancers that are dMMR.
Post angle: 🚨 RESET-C in JCO: ONE cycle of neoadjuvant pembrolizumab in Stage I-III dMMR colon cancer (n=85) → pCR 44%, MPR 57%, 1-yr DFS 96%, only 11% G3 AEs. Endoscopy 86% accurate for pCR — paves nonoperative management pathway. NICHE-2 territory but with a single dose. #CRC #ColorectalCancer #dMMR #MSI #GIOnc #Immunotherapy
#2
Source: AACR 2026 late-breaking mini-symposium / OncLive coverage / Revolution Medicines (April 21, 2026; reported through May 2026) | Authors: Wolpin BM (Dana-Farber Cancer Institute), Revolution Medicines GI-102 investigators | Published: Late-breaking AACR April 21, 2026; updated coverage May 2026
Score: 10/20 — Base 7 (AACR oral / late-breaking mini-symposium) + 2 (Phase 2) + 1 (biomarker-guided / RAS-mutant)
The GI-102 platform Phase 1/2 trial, presented as a late-breaking mini-symposium at AACR 2026 by Brian Wolpin (Dana-Farber), enrolled 40 patients with previously untreated RAS-mutant metastatic pancreatic ductal adenocarcinoma to receive daraxonrasib (RMC-6236, RAS(ON) multi-selective inhibitor) 200 mg PO daily plus gemcitabine + nab-paclitaxel on D1/D15 of 28-day cycles. Confirmed ORR was 58% (95% CI 41–73%) — roughly double the historical 23–32% ORR with gemcitabine + nab-paclitaxel alone. Six-month landmark Kaplan-Meier estimates were 84% (95% CI 68–93) for PFS and 90% (95% CI 76–96) for OS — both unprecedented in 1L PDAC. Median PFS and OS were not yet mature at the December 1, 2025 data cutoff. Most common Grade ≥3 TRAEs were anemia (33%), neutropenia (20%), and fatigue (18%) — no new safety signals beyond the known profiles of either drug. Combined with the recent Phase 3 RASolute 302 OS benefit in 2L PDAC (mOS 13.2 vs 6.7 mo) and the FDA's April 30 expanded access authorization, daraxonrasib is rapidly emerging as a potential paradigm-shifting RAS(ON) inhibitor for the ~90% of PDAC patients with RAS-mutant disease.
Post angle: 🔥 GI-102 Phase 1/2 (AACR 2026): Daraxonrasib + GnP 1L mPDAC — ORR 58%, 6-mo PFS 84%, 6-mo OS 90%. Roughly 2x the historical ORR with GnP alone in RAS-mutant PDAC (~90% of patients). Pan-RAS(ON) inhibition + chemo backbone may finally crack 1L PDAC. #PDAC #GIOnc #KRAS #PancreaticCancer
#3
Source: NPJ Precision Oncology (May 5, 2026) | Authors: Zheng-Lin B, Shibuki T, Strickler JH, Yoshino T, Borad MJ, Bekaii-Saab T, et al. | Published: May 5, 2026
Score: 10/20 — Base 6 (NPJ Precision Oncology, Nature-family) + 1 (real-world/retrospective) + 2 (OS benefit) + 1 (biomarker-guided)
International collaborative real-world cohort from Mayo Clinic, Duke Cancer Institute, National Cancer Center East (Japan), and the SCRUM-Japan GOZILA/MONSTAR-SCREEN registry — 1049 patients with advanced biliary tract cancer (BTC) molecularly profiled and stratified into three groups: non-actionable (n=691), actionable + matched targeted therapy (n=160), and actionable but unmatched (n=198). Among 358 patients (34.1%) with actionable alterations, only 44.7% received matched therapy. The matched-therapy group had a strikingly longer median overall survival of 23.3 mo (95% CI 19.5–30.4) compared to unmatched (14.7 mo, 95% CI 12.8–17.9) and non-actionable (17.1 mo, 95% CI 15.5–18.7) groups (HR 0.57, 95% CI 0.46–0.71; p<0.0001). In multivariable analysis restricted to ≥2L therapy (n=700), matched therapy remained independently associated with improved survival (HR 0.62, 95% CI 0.43–0.90). Critically, patients with actionable tumors who did NOT receive matched therapy fared worse than those with non-actionable disease — an equity argument for universal molecular profiling and matched-therapy access in advanced BTC. With FGFR2 fusions, IDH1, ERBB2 amp, and MSI-H all carrying FDA-approved targeted options, this is the largest real-world dataset to date supporting molecular profiling as standard of care in advanced BTC.
Post angle: 📊 Largest BTC precision oncology cohort to date: n=1049 across Mayo + Duke + NCC East + SCRUM-Japan. Matched targeted therapy doubled OS gap: 23.3 vs 14.7 mo, HR 0.57, p<0.0001. Patients with actionable alterations who DON'T get matched therapy do WORSE than non-actionable. Universal molecular profiling is no longer optional in BTC. #Cholangiocarcinoma #BTC #PrecisionOncology #GIOnc
#4
Source: BMC Cancer (May 5, 2026) | Authors: Vendrell JA, Dalmon I, Cabello-Aguilar S, Colinge J, Assenat E, Solassol J — Montpellier University Hospital | Published: May 5, 2026
Score: 7/20 — Base 5 (BMC Cancer) + 1 (retrospective) + 1 (biomarker-guided / precision)
Single-center prospective integrated DNA + RNA profiling cohort of 62 patients with intra/extrahepatic cholangiocarcinoma at CHU Montpellier. Targeted DNA sequencing detected most frequent alterations TP53 (43.1%), KRAS (29.3%), IDH1 (10.3%, intrahepatic-only). Actionable alterations (IDH1, FGFR2 fusions, ERBB2 amp, MSI-H) were detected in 25% and associated with substantially longer mOS (67.5 vs 20.8 mo; p=0.0004) — consistent with matched-therapy benefit rather than intrinsic biology, validating the Zheng-Lin global cohort findings. Extended RNA sequencing identified five transcriptomic subtypes: Immune (21%), Proliferative (18%), Mesenchymal (42%), Immune-Proliferative (8%), and Unclassified (11%). KRAS-TP53 co-mutation emerged as the strongest adverse prognostic factor (38.1 vs 13.2 mo; p=0.0004) — independent of actionability. Mesenchymal transcriptomic subtype was associated with significantly shorter PFS (5.9 vs 17.9 mo; p=0.045) on chemo ± immunotherapy, but did not significantly affect OS. KRAS-TP53 co-mutation and the Mesenchymal subtype are both detectable on routine FFPE — practical biomarkers for trial enrichment.
Post angle: 🧪 BMC Cancer (n=62): DNA + RNA profiling in cholangiocarcinoma identifies 2 prognostic features beyond standard actionability — KRAS-TP53 co-mutation (mOS 13.2 vs 38.1 mo, p=0.0004) and Mesenchymal transcriptomic subtype (mPFS 5.9 vs 17.9 mo). Both detectable on FFPE → trial enrichment + counseling. #BTC #Cholangiocarcinoma #PrecisionMedicine
#5
Source: Communications Biology (May 5, 2026) | Authors: Jia K, Ding Y, Hu Y, Chen Z et al., Zhejiang University 2nd Affiliated Hospital | Published: May 5, 2026
Score: 7/20 — Base 5 (Communications Biology, Nature-family) + 1 (retrospective + functional) + 1 (biomarker-guided + drug target)
Translational study integrating clinical observation (77 CRC patients undergoing neoadjuvant chemotherapy), single-cell transcriptomics, and in vivo models. Favorable responders to neoadjuvant chemo had well-defined vascular lumens; poor responders were enriched for senescent endothelial cells (senEndo). Mechanistically, chemotherapy-induced senescence reduced GPX4 ubiquitylation in senEndo, leading to GPX4 accumulation and packaging into extracellular vesicles. These EVs are taken up by adjacent CRC cells, where GPX4 inhibits ferroptosis and confers 5-FU/oxaliplatin resistance. Targeting GPX4 with the ferroptosis activator RSL3 restored chemosensitivity in vitro and in vivo. This is a non-tumor-cell-autonomous mechanism of chemoresistance — the stromal vasculature, not the tumor itself, drives the resistance phenotype via paracrine GPX4 transfer. Therapeutic implication: combining standard FOLFOX with senolytics or ferroptosis inducers (e.g., RSL3, sulfasalazine) could improve neoadjuvant response rates. Aligns with the broader trend of stromal-targeted therapy in CRC.
Post angle: 🔬 Communications Biology: Senescent endothelial cells in colorectal tumors traffic GPX4 via extracellular vesicles to CRC cells → ferroptosis blockade → 5-FU/oxaliplatin resistance. n=77 patients + functional models. RSL3 ferroptosis activator restores response. Stromal-targeted therapy: not just CAFs anymore — senescent vasculature too. #CRC #GIOnc #ColorectalCancer #Ferroptosis
Additional Papers of Interest
-
J ImmunoTherapy of Cancer (May 4) — Patient-derived organotypic tumor spheroids (n=30) preserve tumor + TILs, 80% concordance with PDX, 'Organoid Killing Index' transcriptomic signature predicts atezo+bev response with r=0.829, p<0.001 — functional precision oncology for HCC.
-
Cell Death & Disease (May 4) — CXCR4 mediates glutamine metabolism reprogramming via PI3K-Akt-SMAD4, polarizing TAMs to M2 and exhausting CD8 T cells in CRC; CXCR4 knockdown reverses immunosuppression and sensitizes to anti-PD-1.
-
Gut (May 4) — Cubero & Macias commentary highlighting tumor-cell-intrinsic IL-6 autocrine signaling as a metabolic vulnerability in iCCA — potential combination target with FGFR/IDH1 inhibitors.
-
BMJ Open Gastroenterology (May 5) — Sun Yat-sen Cancer Center review framing EMERALD-1 + LEAP-012 + EMERALD-3 readouts and proposing multidimensional patient stratification (tumor burden, MVI, liver function, AFP, ctDNA) for TACE + IO selection.
-
Cancer Metastasis Reviews (May 5) — Daniel Principe (Northwestern) review establishing metastatic PDAC as a distinct immunologic entity: liver mets are profoundly immunosuppressive while lung mets retain higher immune activity — site-specific therapy implications.
-
Cell Death Discovery (May 5) — PBX1 paradoxically inhibits proliferation/migration but suppresses apoptosis via direct BCL2L1 transcription. BCL2L1 knockdown + 5-FU sensitizes CRC cells — Bcl-xL inhibitor + chemo as a translational lead.
Back to all digests