Top 5 Papers
#1
Source: New England Journal of Medicine | Authors: Della-Torre E, Baker MC, Stone JH, et al. (international, INDIGO investigators) | Published: June 2, 2026
Score: 13/20 — NEJM base (10) + Phase 3 RCT (+3) = 13. No additional SoC bonus applied because IgG4-RD has multi-organ scope (pancreatobiliary dominant ~60% of cases, but kidney, salivary, lung, lymph nodes also affected) — included in Top 5 because the GI/pancreatobiliary footprint of IgG4-RD is the dominant clinical manifestation, and AIP and IgG4 sclerosing cholangitis have lacked a steroid-sparing maintenance Phase 3 until now.
Obexelimab is a bifunctional monoclonal antibody that coengages CD19 and FcγRIIb to suppress B-cell activity without depleting the B-cell pool — distinct from rituximab. INDIGO randomized 194 patients with active IgG4-related disease (the systemic immunopathology behind type-1 autoimmune pancreatitis, IgG4 sclerosing cholangitis, and a constellation of pancreatobiliary, mesenteric, and salivary manifestations) to weekly subcutaneous obexelimab 250 mg vs placebo for 52 weeks, with a standardized glucocorticoid taper to discontinuation at week 8. Results across the entire primary/secondary endpoint hierarchy: time to first disease flare requiring rescue therapy — HR 0.44 (95% CI 0.28–0.71, p<0.001), with flares in 26.8% vs 54.6%. Complete remission at 52 weeks — 37.1% vs 19.6% (p=0.005). Cumulative glucocorticoid rescue dose — 329.5 mg vs 929.8 mg (p=0.004), a 65% reduction. Serious adverse events were actually lower with obexelimab (10.3% vs 18.6%). For GI oncology and pancreatobiliary medicine specifically, the AIP and IgG4-SC populations have lived under chronic steroid burden and recurrent flares for two decades — this is the first Phase 3 that delivers a steroid-sparing maintenance therapy in this disease. Expect rapid uptake in dedicated AIP/IgG4-SC referral centers and a re-examination of which IgG4-RD patients should be managed maintenance-first vs flare-first.
Post angle: INDIGO (NEJM Phase 3): obexelimab in IgG4-related disease — the systemic disease behind type-1 AIP and IgG4 sclerosing cholangitis. Flares HR 0.44, complete remission 37% vs 20%, and a 65% drop in cumulative glucocorticoid rescue (n=194). A first-in-class steroid-sparing backbone for the GI/pancreatobiliary face of IgG4-RD. This changes management of a disease most of us treat suboptimally. #GIOnc #Pancreatology #IgG4 #AIP #Cholangitis
#2
Source: Nature Medicine | Authors: Bando H, Yoshino T, Kato T, et al. (CIRCULATE-Japan platform, multicenter) | Published: June 8, 2026
Score: 12/20 — Nat Med base (9) + Phase 3 RCT (+3) = 12. Primary endpoint not met, so no SoC/survival-benefit bonus. Landmark negative trial that defines the field: the first randomized data testing whether MRD-positive patients benefit from a non-fluoropyrimidine post-adjuvant intervention. The answer this round is no — and that itself is a clinically pivotal data point.
ALTAIR is the long-awaited first Phase 3 RCT of MRD-guided post-adjuvant therapy in resected colorectal cancer. From the CIRCULATE-Japan platform, 243 patients with stage 0–IV resected CRC who remained or became ctDNA-positive after standard adjuvant care were randomized to trifluridine/tipiracil (FTD/TPI) vs placebo for 6 months. Primary endpoint DFS: mDFS 9.30 mo with FTD/TPI vs 5.55 mo with placebo, HR 0.79 (95% CI 0.60–1.05), p=0.107 — NOT statistically significant. Grade ≥3 hematologic AEs were dramatic: 73% in the FTD/TPI arm vs 3.3% in placebo. This is the negative trial the MRD-guided community has been bracing for — proof that ctDNA positivity identifies the right patients to worry about (placebo arm mDFS only 5.55 mo confirms the prognostic signal), but the wrong drug at the wrong intensity does not rescue them. The next-generation MRD-guided trials (BESPOKE CRC arms with IO, CIRCULATE-US escalation) now have a calibration point: passive escalation with FTD/TPI doesn't work, so the field shifts to immune-active or biology-matched escalation. Read alongside CIRCULATE Phase 3 (positive in stage II MSS) from the June 2 digest: ctDNA-guided strategy works when the escalation is appropriate to the disease state, not when you just throw more fluoropyrimidine at it.
Post angle: ALTAIR is the negative MRD trial we needed. Post-adjuvant FTD/TPI in ctDNA+ resected CRC did NOT improve DFS (9.30 vs 5.55 mo, HR 0.79, p=0.107, n=243). ctDNA correctly flagged the high-risk patients — placebo mDFS only 5.55 mo — but FTD/TPI was the wrong drug. The MRD field doesn't lose; it pivots to immune-active and biology-matched escalation. #GIOnc #CRC #ctDNA #MRD
#3
Source: Nature Medicine | Authors: Kaplan LM, le Roux CW, Sanyal AJ, et al. (multicenter US + Spain, Boehringer-sponsored) | Published: June 7, 2026
Score: 11/20 — Nat Med base (9) + Phase 3 RCT (+3) − 1 (not a primary oncology trial — endpoint is liver fat / weight, not cancer) = 11. Promoted to Top 5 because MASLD/MASH is now the fastest-rising etiology of HCC in the US, and the Phase 3 efficacy on liver-fat and weight is large enough that this is the most consequential upstream HCC-prevention read in years.
SYNCHRONIZE-MASLD is a Phase 3 RCT of survodutide, a once-weekly subcutaneous dual glucagon-receptor/GLP-1-receptor agonist, in 216 adults with obesity (BMI ≥30 or ≥27 with complications) plus at-risk MASLD (defined by noninvasive tests showing inflammation/fibrosis or biopsy-confirmed MASH). 2:1 randomization to survodutide 6 mg weekly vs placebo for 48 weeks. Both co-primary endpoints — ≥30% reduction in MRI-PDFF liver fat content and percentage body-weight change — were met decisively. 84.2% of survodutide-treated patients achieved ≥30% liver fat reduction vs 24.3% placebo (p<0.0001); mean body-weight change −12.2% vs −1.0% (p<0.0001). Adverse events were predominantly GI and dose-titration-related. Not an oncology trial — but consequential for GI oncology because MASLD/MASH is the fastest-rising etiology of HCC in the US, and a drug that can deliver this magnitude of liver-fat regression and weight loss in obese MASLD has the realistic potential to bend the HCC incidence curve upstream of cirrhosis. For hepatologists and GI oncologists co-managing pre-cirrhotic MASH patients, this is the Phase 3 evidence the field has been waiting for to push hepatic-steatosis-active pharmacotherapy into routine practice. Expect FDA submissions and rapid integration into AASLD MASH guidance over the next 12 months.
Post angle: SYNCHRONIZE-MASLD (Nat Med Phase 3): survodutide (glucagon/GLP-1 dual agonist) drops MRI-PDFF liver fat ≥30% in 84% vs 24% placebo and body weight 12% vs 1% at 48 weeks in obese MASLD (n=216). Not an oncology trial — but MASLD is the fastest-rising etiology of HCC in the US. This is the most consequential upstream HCC-prevention read in years. #GIOnc #HCC #MASLD #MASH #Prevention
#4
Source: Gastroenterology | Authors: Dominitz JA, Gawron AJ, Kaltenbach T, et al. (Veterans Health Administration, multicenter) | Published: June 5, 2026
Score: 10/20 — Gastroenterology base (7) + cluster RCT (+2) + measurable impact on a CRC-mortality-linked metric (+1) = 10. ADR is inversely associated with post-colonoscopy CRC death; the previously mixed real-world evidence on CADe gets a strong VHA-scale corroboration here.
Computer-aided detection (CADe) for colonoscopy has had a curious evidence base — randomized academic trials consistently showed ADR gains, but real-world deployments have repeatedly failed to demonstrate sustained benefit, raising concerns about Hawthorne effects and selection bias. The Veterans Health Administration ran a pragmatic cluster randomized rollout: 42 facilities (269 endoscopists, 71,594 pre- and 35,399 post-deployment colonoscopies) received CADe between Oct 2022 and Feb 2023; 97 control facilities (547 endoscopists, 151,792 pre- and 75,415 post-deployment colonoscopies) were assigned random deployment dates. Result: at CADe sites, ADR rose 4.2 percentage points (50.7% → 54.9%, 95% CI 3.48–4.74); at control sites, ADR drifted slightly down (51.8% → 51.1%). In mixed-effects logistic regression, CADe availability independently increased the odds of adenoma detection (OR 1.22, 95% CI 1.15–1.28, p<0.001). The effect was independent of baseline endoscopist ADR quintile, and withdrawal time did not change — arguing against simple Hawthorne or behavioral effects. Two open questions the authors flag honestly: (1) does the ADR gain translate to a measurable reduction in post-colonoscopy CRC incidence/mortality (the field's most important downstream outcome); (2) does long-term CADe reliance produce endoscopist deskilling. NCT05888623. For health systems debating CADe procurement, this is the largest real-world cluster RCT to date and the cleanest signal that the academic-trial ADR benefit transfers.
Post angle: VHA cluster RCT in Gastroenterology: AI-assisted colonoscopy across 269 endoscopists and 334,200 procedures lifts ADR 50.7% → 54.9% while control sites stayed flat (OR 1.22). The largest real-world CADe deployment yet — finally addresses why prior real-world studies underwhelmed. Open questions: does this translate to CRC mortality, and is there an endoscopist deskilling cost? #GIOnc #CRC #AI #Colonoscopy
#5
Source: Gut | Authors: Van den Bossche JL, Houbracken I, Baldan J, Rooman I, et al. (VUB + ULB BruPaCT, multicenter EU) | Published: June 4, 2026
Score: 8/20 — Gut base (6) + biomarker/subtype definition (+1) + clinical-translation lever for PDAC stratification and the ASCP-vs-PDAC distinction (+1) = 8. Translational mechanism paper with a real impact-on-classification angle for the basal-like PDAC subtype debate.
Pancreatic cancer subtyping has bifurcated along a 'classical' vs 'basal-like' axis since the Moffitt and Bailey landmark transcriptomic papers — basal-like tumors have worse survival and a distinct chemotherapy responsiveness. What was missing was a clean architectural anchor in the native pancreatic duct from which these subtypes descend. This Gut paper combines new spatial transcriptomics datasets with integrated public single-cell RNAseq, validated by multiplex immunofluorescence. Findings: groups of Keratin-5+ cells in the larger native pancreatic ducts harbor a stem/(supra)basal gene signature. At single-cell resolution this comprises ΔNp63+ basal cells (BAS) and ΔNp63− supra-basal luminal-B cells (LUM-B), the latter expressing previously unreported MUC4 and MUC16 — distinct from the conventional luminal lineage. In cancer, BAS and LUM-B signatures associate with basal-like PDAC and correlate with worse survival. The architectural surprise: adenosquamous carcinoma of the pancreas (ASCP) preserves LUM-B and BAS in a spatially unmixed pattern, while PDAC exhibits a random spatial mixing and fragmented native expression programmes. ΔNp63 itself drives the cell-plasticity to BAS, conserved from native tissue to cancer. Two implications: (1) basal-like PDAC has a native-duct origin signature that can be tracked at single-cell and spatial resolution, opening up MUC16/MUC4-based diagnostic and therapeutic strategies for this subtype; (2) ASCP and basal-like PDAC, often lumped, may warrant separate consideration in research and trial design.
Post angle: Pancreatic luminal-basal spatial map (Gut, VUB/ULB BruPaCT): native pancreatic duct harbors ΔNp63+ BAS and ΔNp63− LUM-B (MUC4/MUC16+) supra-basal cells. In ASCP these are spatially preserved; in PDAC they're fragmented and mixed. BAS/LUM-B signatures track basal-like PDAC + worse survival. Argues ASCP and basal-like PDAC deserve separate trial design — and surfaces MUC16/MUC4 as potential basal-like PDAC targets. #GIOnc #PDAC #PrecisionMedicine
Additional Papers of Interest
-
Chinese Medical Journal — Multicenter randomized double-blind trial of pronase (a proteolytic enzyme used as a mucolytic flush) vs saline during colonoscopy. n=1,942 across 10 hospitals. Visibility improvement rate 97.2% vs 86.7% (p<0.0001) and higher endoscopist satisfaction, but no difference in adenoma, polyp, or precancerous-lesion detection rates. A nuanced read on the gap between operator perception and downstream cancer prevention metrics.
-
Gastroenterology — Mayo Precision Medicine for Obesity Program, n=483 adults with obesity, identifies a 27% subphenotype with fast gastric emptying, low postprandial GLP-1, reduced mucosal GCG/PYY mRNA, and elevated SCFAs — and significantly enhanced response to tirzepatide (21.5% vs 11.7% weight loss at 6 months). Precision-medicine framework for the GLP-1/GIP receptor agonist class. Adjacent to oncology because the obesity → GI cancer axis is plausibly modifiable through targeted obesity pharmacotherapy.
-
Cell Communication and Signaling — Mechanism paper from Universidad de Chile linking H. pylori infection to gastric carcinogenesis via a previously unmapped HIF-1α / ALS2 (guanine exchange factor) / Rab5 (endosomal trafficking GTPase) / β-catenin axis. Validated in MKN74 and AGS gastric cancer cells and in patient-derived tumor organoids. Adds a druggable node to the H. pylori → gastric cancer pathway that has been mechanistically opaque between bacterial infection and WNT activation.
Back to all digests