GI Oncology Weekly — Issue 4

This week’s calls — 5 readouts · 2 to act on · ~8 min read. Tap any row to jump.

CheckMate 648 (5-year follow-up) First-line nivolumab combinations hold their survival advantage in advanced ESCC at five years Practice-shaping SOLARIS High-dose vitamin D3 does not improve progression-free survival in first-line metastatic colorectal cancer Practice-shaping Mountain-02 Tumour-specific MHC class II status separates responders from non-responders to perioperative tislelizumab Worth watching DENEB A 39-microRNA blood test detects four in five advanced adenomas, the endpoint liquid screening keeps missing Worth watching Lynch syndrome surveillance cohort (English NHS) 4,732 MMR carriers: lower mortality with ≤3-year colonoscopy, but no fall in incidence and a rise at ≤2 years On the radar

This week · 5 readouts

The Verdicts

Every trial that matters this week, distilled to a single call — act on it, watch it, or keep it on the radar.

Practice-shaping Oesophagus · advanced squamous cell carcinoma

CheckMate 648 (5-year follow-up)

The durability question answered. Both first-line nivolumab regimens kept their overall survival advantage over chemotherapy at a minimum of five years, with no new safety signals. The nuance that belongs in clinic is the chemotherapy-free arm: it buys overall survival without buying progression-free time, which makes it a conversation with a patient rather than a default.

· PD-L1≥1% OS HR 0.62 (0.48–0.79) nivolumab + chemotherapy· PD-L1≥1% OS HR 0.62 (0.48–0.80) nivolumab + ipilimumab· All randomised OS HR 0.77 both arms· PD-L1≥1% PFS HR 0.67 nivo+chemo, HR 1.03 nivo+ipi· n=970

Go deeper

CheckMate 648 randomised 970 patients with previously untreated advanced oesophageal squamous cell carcinoma to first-line nivolumab plus chemotherapy, nivolumab plus ipilimumab, or chemotherapy alone. At a minimum of five years' follow-up, both experimental arms retained their overall survival advantage in the PD-L1≥1% population — OS HR 0.62 (95% CI 0.48–0.79) for nivolumab plus chemotherapy and OS HR 0.62 (95% CI 0.48–0.80) for nivolumab plus ipilimumab — with an OS HR of 0.77 for both arms across all randomised patients. Progression-free survival tells a more divided story: nivolumab plus chemotherapy improved PFS in the PD-L1≥1% group (HR 0.67), while the chemotherapy-free nivolumab plus ipilimumab arm did not (PFS HR 1.03). No new safety signals emerged with the additional follow-up. Full daily entry →

Annals of Oncology — Kato K, Chau I and the multicentre CheckMate 648 investigators

Practice-shaping Colorectal · first-line metastatic disease

SOLARIS

A decade-old hypothesis, properly powered and properly retired. High-dose vitamin D3 added nothing to first-line chemotherapy plus bevacizumab on progression-free survival, response rate or overall survival. Correct deficiency as you would in any patient; stop treating 4000 IU as an anticancer drug. The authorship matters — the group that generated the original phase 2 signal ran the trial that closed it.

· Median PFS 11.8 mo (10.3–13.3) vs 10.3 mo (9.4–12.2), 1-sided log-rank p=0.25, NS· ORR 51% (44–58) vs 44% (37–50), p=0.12· Median OS 25.6 vs 27.0 mo, p=0.66· n=455· median follow-up 20 mo

Δ +1.5 mo median · p=0.25

Survival over time · schematic — PFS*

High-dose vitamin D3 (4000 IU)Standard-dose vitamin D3 (400 IU)

Go deeper

SOLARIS (Alliance A021703) is the properly powered phase III test of an idea that has circled colorectal oncology for a decade — that high-dose vitamin D3 improves outcomes in metastatic disease. 455 patients with previously untreated mCRC were randomized double-blind across the US National Clinical Trials Network to mFOLFOX6 or FOLFIRI plus bevacizumab with either high-dose D3 (8000 IU daily for 14 days as a loading dose, then 4000 IU daily) or standard-dose D3 (400 IU daily), continued until progression or intolerance. Median PFS was 11.8 months (95% CI 10.3–13.3) with high-dose versus 10.3 months (9.4–12.2) with standard dose — a 1.5-month numerical gap that did not approach significance (1-sided log-rank p=0.25). The secondary endpoints were equally flat: objective response rate 51% (44–58%) versus 44% (37–50%), p=0.12, and median overall survival 25.6 versus 27.0 months, p=0.66. There were no clinically meaningful differences in grade ≥3 adverse events (neutropenia 32% vs 30%, hypertension 20% vs 23%) and no excess vitamin D–associated toxicity. Median follow-up was 20 months; database freeze July 2024. Full daily entry →

JAMA — Kimmie Ng, Fang-Shu Ou, Jeffrey A. Meyerhardt, Alliance for Clinical Trials in Oncology (A021703)

Worth watching Stomach · resectable gastric / GEJ adenocarcinoma

Mountain-02

A rare and welcome design: the biomarker was built into the randomisation rather than mined for afterwards. In tsMHC-II-positive tumours adding tislelizumab more than doubled the major pathological response rate; in tsMHC-II-negative tumours there was no significant benefit. Not yet a clinic-ready assay, but a template for how perioperative immunotherapy trials in gastric cancer should be built.

· tsMHC-II-positive major pathological response 61.8% vs 26.5% (p=0.003), primary endpoint met· pCR 32.4% tsMHC-II-positive vs 5.9% tsMHC-II-negative (p=0.006)· no significant benefit in tsMHC-II-negative· n=136

Major pathological response, tsMHC-II-positive tumours · 0–100% scale

61.8%Perioperative tislelizumab + chemotherapy

26.5%Chemotherapy alone

050100%

Go deeper

Mountain-02 (NCT06374901) randomised 136 patients with operable cT3–4aN+M0 gastric or gastro-oesophageal junction cancer to perioperative tislelizumab plus chemotherapy or chemotherapy alone, prospectively stratified by tumour-specific MHC class II (tsMHC-II) expression. In tsMHC-II-positive tumours, adding tislelizumab raised the major pathological response rate to 61.8% from 26.5% (p=0.003), meeting the primary endpoint; pathological complete response was 32.4% in tsMHC-II-positive versus 5.9% in tsMHC-II-negative tumours (p=0.006). No significant benefit was seen in the tsMHC-II-negative subgroup. The design is the point: rather than running an all-comers perioperative immunotherapy trial and mining for a biomarker afterwards, the investigators built the stratification into randomisation. Full daily entry →

Cancer Cell — Xu Z, Cheng X and colleagues, five Chinese cancer centres

Worth watching Colorectal · screening and early detection

DENEB

Blood-based colorectal screening has consistently been strong on established cancer and weak on precancer — backwards for a test whose purpose is prevention. DENEB is the first assay to post an advanced adenoma sensitivity that looks like it might fix that. The design is outcome-enriched case-control, so spectrum bias is the obvious caveat and a screening-population cohort is the necessary next step.

· AUC 95% (92–98) for CRC + advanced adenoma vs controls· CRC sensitivity 91% any stage, 92% stage I–III· Advanced adenoma sensitivity 81%· Specificity 85% for advanced neoplasia, 83% against negative colonoscopy

Assay sensitivity by lesion type · 0–100% scale

91%Colorectal cancer (any stage)

81%Advanced adenoma

050100%

Go deeper

DENEB (NCT06342440) is a multicentre diagnostic accuracy study of a blood-based assay combining 19 cell-free and 20 exosomal microRNAs by RT-qPCR, with a locked machine-learning model. In the testing cohort the assay reached an AUC of 95% (95% CI 92–98) for colorectal cancer plus advanced adenoma versus controls, with 91% sensitivity for colorectal cancer at any stage and 92% for stage I–III disease. Advanced adenoma sensitivity was 81%; specificity was 85% for advanced neoplasia and 83% against a negative colonoscopy. The advanced adenoma figure is the headline, because blood-based colorectal screening has consistently performed well on established cancer and poorly on precancer — which is backwards for a test whose purpose is prevention. The outcome-enriched case-control design means spectrum bias is the obvious caveat, and a true screening-population cohort is the necessary next step. Full daily entry →

The Lancet Gastroenterology & Hepatology — Mannucci A, Goel A and colleagues, multicentre

On the radar Colorectal · Lynch syndrome surveillance

Lynch syndrome surveillance cohort (English NHS)

Uncomfortable and worth sitting with. Surveillance at ≤3-year intervals tracked with lower colorectal cancer-specific and all-cause mortality — but with no reduction in total incidence and no clear stage shift, which is mechanistically awkward. At ≤2-year intervals total incidence was higher, driven by early-stage disease with no fall in late-stage: the shape overdiagnosis makes. Non-randomised, so selection bias is equally live. Not an argument for scoping Lynch carriers less; an argument against assuming a shorter interval is automatically a better one.

· n=4,732 MMR carriers, English Lynch registry linked to NHS records 2010–2022· ≤3-year mean interval (n=3,028): lower CRC-specific and all-cause mortality, no change in total CRC incidence, no strong stage shift· ≤2-year mean interval (n=1,569): higher total CRC incidence, driven by early-stage disease

Go deeper

This national observational study linked an English Lynch syndrome registry to NHS digital records to follow 4,732 mismatch repair gene carriers between 2010 and 2022, asking what colonoscopic surveillance actually delivers. Surveillance at a mean interval of three years or less (n=3,028) was associated with lower colorectal cancer-specific and all-cause mortality after multivariable adjustment — but with no accompanying reduction in total colorectal cancer incidence and no strong evidence of stage shift. More provocatively, surveillance at a mean interval of two years or less (n=1,569) was associated with a higher total colorectal cancer incidence, driven by more early-stage cancers with no corresponding fall in late-stage disease. The authors are appropriately careful: short follow-up, spectrum bias, overdiagnosis and selection bias are all live explanations in a non-randomised design. A mortality benefit without a stage shift is mechanistically awkward, and an incidence rise at the tightest interval is exactly the shape overdiagnosis makes. Full daily entry →

Gut — Huntley C, Turnbull C and colleagues, Institute of Cancer Research and NHS England National Disease Registration Service

* Survival curves are illustrative — modeled from each trial’s reported median and hazard ratio (assuming exponential survival), not the published Kaplan–Meier curves.

The take

The take — a result has to earn its place

The strongest work this week was not the most positive work. Three of these five papers move the field by subtraction — by closing a question, qualifying a benefit, or refusing to let a number mean more than it does.

SOLARIS is the cleanest example. High-dose vitamin D3 in metastatic colorectal cancer has been appealing for a decade on the strength of a phase 2 signal; the properly powered phase 3 found a PFS gap that did not approach significance and no survival difference. What makes it exemplary is that the group who generated the original signal ran the trial that retired it. That is how the process is supposed to work, and it happens less often than it should.

The Lynch cohort is harder to sit with, because it does not resolve cleanly. Lower mortality at ≤3-year intervals, no fall in incidence, no stage shift — and at ≤2 years, more cancers found rather than fewer. Each finding has an innocent explanation in a non-randomised design. Together they should still unsettle the reflex that a tighter interval is a safer one.

Against those, CheckMate 648 shows what a positive result looks like once time has tested it. Five years is where immunotherapy claims either hold or quietly deflate; these held. And Mountain-02 is the design lesson — stratify on the biomarker at randomisation, rather than searching the wreckage of an all-comers trial for a subgroup afterwards.

The common thread:

A benefit that survives five years is worth more than one measured at twelve months.
A biomarker specified before randomisation is worth more than a subgroup found after.
A hypothesis retired by its own authors is worth more than one left politely open.
A surveillance interval should demonstrate what it buys, not merely feel safer.

GI Oncology Weekly · curated by Dr. Allan Pereira, Moffitt Cancer Center. Educational summaries for practising oncologists — not medical advice; verify against primary sources before acting.

All weekly issues