GI Oncology Daily Digest

May 14, 2026 — BRAF Paradox Breaker Lands at JCO + Oligoprogressive HCC Strategy + Why MSS-CRC Resists IO
Curated by Dr. Allan Pereira — Moffitt Cancer Center

Top 5 Papers

#1
Mosperafenib — First-in-class paradox-breaker BRAF inhibitor delivers single-agent activity in BRAFᵛ⁶⁰⁰-mutant CRC, including BRAFi-exposed patients (JCO Phase Ia/b)
Source: Journal of Clinical Oncology (JCO)  |  Authors: Vieito, Fontana, Han, Castanon, Pinato, Bechter et al. — multicenter European/NZ consortium; sponsor Roche  |  Published: 2026-05-10
Score: 9/20 — Base 8 (JCO Tier-1) + biomarker-guided/precision (+1) = 9. Phase I excluded from study-type bonus.
Eighty patients with BRAFᵛ⁶⁰⁰-mutant advanced solid tumors — 63 CRC (79%), 13 melanoma, 4 other — received oral mosperafenib up to 3,600 mg QD across 28-day cycles. Sixty percent were BRAFi-exposed. MTD was not reached; only 2 DLTs (both rash). Grade 3–4 TRAEs 16.3%, no grade 5 events. Linear PK, sustained pERK inhibition ≥90%, and an exposure–response relationship were observed. Overall ORR was 24.2% (2 CRs, 14 PRs). Median PFS reached 6.4 months in CRC (across 200 mg QD–1200 mg TID) and 3.5 months in melanoma. Critically — no palmar-plantar erythrodysesthesia and no keratoacanthoma, the hallmark off-target paradox toxicities of first-generation BRAF inhibitors. Brain penetration was documented in preclinical models, opening a path for CNS-directed BRAF therapy. This is the first paradox-breaker BRAF inhibitor to clear Phase I in a CRC-dominated cohort.
Post angle: A second-generation BRAF inhibitor that doesn't activate wild-type BRAF, doesn't cause PPE or keratoacanthoma, and gives 6.4-mo PFS in BRAFᵛ⁶⁰⁰ CRC — including patients already failed by encorafenib? Combos coming with anti-EGFR. #GIOnc #CRC #BRAFV600 #PrecisionMedicine
#2
Continuation of first-line systemic therapy with progression-directed radiotherapy in oligoprogressive HCC — Multicenter Phase II (10 cancer centers)
Source: Clinical Cancer Research  |  Authors: Shi F, Wang H, Yang Y, Ding L, Yue J et al. — 10 Chinese cancer centers (Shandong, Jilin, Sun Yat-sen, Tianjin, Zhejiang)  |  Published: 2026-05-13
Score: 8/20 — Base 6 (CCR Tier-2) + Phase II (+2) = 8.
Thirty-six patients with oligoprogressive HCC on first-line systemic therapy received progression-directed radiotherapy (PDRT, BED ≥60 Gy) while continuing FLST. At a median follow-up of 10.9 months, median PFS was 7.0 months (95% CI 4.9–9.7) with 3-/6-/9-month PFS rates of 73.7%/64%/38.8%. ORR 64.7%, DCR 98.0%. Median OS and DoR not yet reached; 1-year OS 86.4%. RT-related toxicities (mostly grade 1–2) in 44.4%, grade ≥3 in 11.1%. Type of FLST and baseline ALBI grade independently predicted PFS. The signal supports a "continue and ablate" strategy for HCC patients whose disease relapses at a limited number of sites on systemic therapy — instead of switching lines and burning a sequencing option.
Post angle: Oligoprogression on first-line systemic HCC therapy? Don't switch — ablate. Phase II across 10 centers: keep the FLST, add PDRT, get 7-mo PFS and 86% 1-yr OS. Practical, preserves sequencing, low grade-3 toxicity. #HCC #LiverCancer #RadiationOncology
#3
MSS colorectal cancer evades immune surveillance via an intrinsic, secretome-driven program — independent of neoantigen load
Source: Cancer Discovery  |  Authors: Cattaneo CM, Scardellato S, Mauri G, Marsoni S, Germano G, Bardelli A et al. — IFOM/IRCCS Humanitas/Niguarda/Netherlands Cancer Institute consortium  |  Published: 2026-05-13
Score: 7/20 — Base 6 (Cancer Discovery Tier-2) + biomarker-guided/precision (+1) = 7.
The Bardelli group engineered an experimental system where MSI and MSS CRC cells present identical levels of a defined antigen recognized by TCR-engineered T cells. Despite matched antigen presentation, MSS tumors still resisted T-cell killing — impaired activation, reduced cytotoxicity, and blunted synapse formation. The block was traced to the MSS tumor secretome itself, which suppressed responses even in immunogenic MSI cells. Surfaceome mass-spectrometry identified glycosylation-dependent surface alterations as the proximal mechanism impairing immune recognition. Bottom line: MSS-CRC immune evasion isn't purely a story of low TMB and weak neoantigens — there's an intrinsic, antigen-independent suppressive program that anti-PD-1 alone won't fix. Targeting tumor glycosylation pathways becomes a credible co-strategy for re-sensitizing MSS-CRC to immunotherapy.
Post angle: Why does MSS-CRC ignore checkpoint inhibitors even when you force antigen presentation? Bardelli lab: it's not just neoantigen load — the MSS secretome itself, via glycosylation-dependent surface changes, sabotages the immune synapse. New combo strategy emerging. #CRC #GIOnc #Immunotherapy
#4
Robotic vs Laparoscopic Low Anterior Resection for Rectal Cancer — Updated meta-analysis of 33 studies and 82,149 patients
Source: Journal of Robotic Surgery  |  Authors: Kossenas K, Konstantinidis M, Sakarellos P, Machairas N, Schizas D et al. — National & Kapodistrian University of Athens / Aristotle University of Thessaloniki  |  Published: 2026-05-13
Score: 7/20 — Base 5 (Other journal) + Meta-analysis (+2) = 7.
PRISMA-compliant systematic review and meta-analysis of 33 comparative studies (RoB 2 + ROBINS-I + GRADE) restricted to low anterior resection — a deliberate move to escape the heterogeneity of prior "any rectal procedure" reviews. R-LAR (n=56,290) vs L-LAR (n=25,859) showed significantly lower conversion (OR 0.45, p<0.00001, I²=28%), lower 30-day reoperation (OR 0.86, p=0.01), lower 30-day mortality (OR 0.65, p=0.002), and higher complete-TME rate (OR 2.71, p=0.01). Operative time was longer by ~30 minutes. No significant difference in major complications, anastomotic leak, ileus, LOS, readmission, CRM positivity, lymph-node yield, local recurrence, DFS, or OS. The signal is consistent with surgical-quality advantages of the robotic platform — better TME and lower short-term mortality — even though long-term oncologic equipoise persists pending high-quality RCTs.
Post angle: 82,149 patients, 33 studies, only LAR: robotic gives lower conversion, lower 30-day mortality, and better complete TME than laparoscopic — but operative time is ~30 min longer and DFS/OS are similar. Surgical advantage real; oncologic advantage still TBD. #ColorectalSurgery #RectalCancer
#5
Multi-phase plasma metabolomics framework yields a 12-metabolite signature for early gastric cancer detection — AUC 0.951 in validation
Source: Nature Communications  |  Authors: Bai L, Hu F, Xie Y, Zhang S, Min L et al. — Beijing Friendship Hospital, Tongji Medical College, Renmin Hospital of Wuhan University, National Cancer Center of China  |  Published: 2026-05-13
Score: 7/20 — Base 6 (Nat Commun) + biomarker-guided/precision (+1) = 7.
Multi-center study of 1,706 plasma samples integrating untargeted, relative-quantitative targeted, and absolute-quantitative targeted metabolomics. An interpretability-driven biomarker-selection algorithm identified 84 enriched metabolites — concentrated in caffeine metabolism and primary bile-acid biosynthesis — and distilled them into a 12-metabolite panel covering multiple functional modules. Machine-learning models built on the absolute-quantitative signature achieved AUC 0.951 in the validation cohort. Beyond the diagnostic number, the framework returns mechanistic readouts (bile-acid handling, caffeine metabolism) rather than the typical opaque black-box panel — a step toward interpretable, biology-grounded gastric-cancer screening blood tests in regions where endoscopy access is limited.
Post angle: 12 plasma metabolites, AUC 0.951 for gastric cancer — and the panel actually maps to biology (bile-acid + caffeine metabolism) rather than a black box. Built across 1,706 multicenter samples. Endoscopy-light screening for high-prevalence regions is getting closer. #GastricCancer #GIOnc #LiquidBiopsy

Additional Papers of Interest

  1. Are age, comorbidity and frailty associated with complications after minimally invasive esophagectomy? — ICAN trial post-hoc
    Diseases of the Esophagus — In 245 patients from the randomized ICAN trial, age ≥75, ASA ≥3, CCI ≥2 and CIRS-G were not independently associated with severe complications, overall complications, readmission, or hospital/ICU LOS — supporting MIE as a safe option in older, comorbid patients selected by current clinical practice.
  2. Targeted degradation vs inhibition of USP7 in solid cancer cells (incl. PDAC) — Distinct phenotypes between deubiquitinase PROTACs and inhibitors
    Nature Communications — Side-by-side proteomic and cellular comparison in melanoma and pancreatic cancer cells shows that selective USP7 PROTAC degradation modulates a different protein set than prolonged inhibitor exposure (which causes USP7-independent metabolic dysregulation) — a methodologic caution for hydroxypiperidine-based USP7 inhibitors and a chemical biology toolbox for cancer-specific DUB studies.
  3. Long-term colonoscopy screening effects on CRC incidence and mortality — Multicountry RCT (Lancet, May 9)
    Lancet — Recap from the May 12 screening edition: the 13-year NORDICC update reaffirms CRC incidence reduction with one-time invitation but with attenuated mortality benefit; complete the file by re-reading the screening digest if you missed it.
  4. Daraxonrasib full NEJM publication of Phase 1/2 (RMC-6236-001) in RAS-mutant PDAC — Posted for reference; primary coverage May 13
    New England Journal of Medicine — Full text now online for the RAS multi-selective inhibitor that drove last week's RASolute 302 OS readout; relevant for anyone discussing the path to ASCO Plenary May 31.
  5. X KOL sweep status — Grok-mediated check of 6 curated handles (May 12–14, 2026 window)
    Run accountability — 6/6 handles checked via Grok (@DrRishabhOnco, @VJOncology, @OncBrothers, @DaisukeKotani, @OncoAlert, @DraMartinezLago). Only qualifying post was @OncBrothers sharing the FDA tazemetostat (Tazverik) withdrawal — heme/sarcoma indication, not GI. No Colleague Engagement bonus applied this run.
Back to all digests