Top 5 Papers
#1
Source: Lancet Oncology | Authors: Sebag-Montefiore D, Adams R, Gollins S, … Hackshaw A, Lopes A — ARISTOTLE Trial Management Group, 75 UK hospital sites (Cancer Research UK) | Published: 2026-08-01
Score: 12/20 — Base 9 (Lancet Oncology) + Phase III RCT (+3) = 12. Definitively negative trial — no survival-benefit or standard-of-care bonus applied. Ranks above TORCH on the same total by base-score tiebreak (Lancet Oncology 9 vs JAMA Oncology 7) and by trial scale: 589 patients across 75 sites with 78-month median follow-up.
ARISTOTLE (EudraCT 2008-005782-59) randomized 589 patients with MRI-defined locally advanced rectal cancer threatening or involving the resection margin to preoperative 45 Gy in 25 fractions with either capecitabine alone or reduced-dose capecitabine plus weekly intravenous irinotecan 60 mg/m² in weeks 1–4; 564 were analysed in the modified intention-to-treat population. After a median follow-up of 78 months the trial is unambiguously negative: 36-month disease-free survival was 68% (95% CI 63–73) with irinotecan versus 67% (61–72) with standard of care, hazard ratio 0.91 (95% CI 0.68–1.23, p=0.54), and deaths were near-identical at 31% versus 32%. The cost side is where the trial speaks loudest. Grade 3 or worse events occurred in 78% of the irinotecan group versus 52% of controls, with diarrhoea at 14% versus 4%, lymphopenia at 66% versus 35%, and neutropenia at 10% versus 1%. That toxicity fed straight back into deliverability — only 75% of irinotecan patients received the full 45 Gy versus 89% of controls, and only 68% received at least 90% of planned capecitabine versus 89%. Five deaths were attributed to protocol treatment, three of them in the irinotecan arm. The authors' conclusion is appropriately blunt: irinotecan should not be combined with radiotherapy plus capecitabine in this setting.
Post angle: The cleanest lesson here is a deliverability lesson, not a drug lesson. Intensification that patients can't complete isn't intensification — 78% grade ≥3 toxicity meant a quarter of the irinotecan arm never finished radiotherapy. Pair this with the ARISTOTLE tumour-cell-density post hoc we covered July 30: the trial is negative overall, but the biomarker signal suggests the failure may be one of selection rather than of biology. #GIOnc #RectalCancer #CRC #Chemoradiotherapy
#2
Source: JAMA Oncology | Authors: Lyu N, Zhao M, et al. — Sun Yat-sen University Cancer Center, multicenter (China) | Published: 2026-07-30
Score: 12/20 — Base 7 (JAMA Oncology) + Phase III RCT (+3) + OS and PFS survival benefit (+2) = 12.
This open-label phase 3 randomized trial enrolled 241 patients with BCLC stage B unresectable hepatocellular carcinoma across two Chinese centres and compared transarterial chemoembolization followed by selective radiofrequency ablation of residual viable tumour against TACE alone. Median progression-free survival was 17.7 versus 7.3 months (hazard ratio 0.47, 95% CI 0.34–0.65, p<0.001), and median overall survival was 88.6 versus 35.1 months (hazard ratio 0.50, 95% CI 0.34–0.73, p<0.001). Grade 3–4 treatment-related adverse events were comparable at 23.2% versus 18.3%, so the benefit does not appear to be bought with meaningful added toxicity. Two caveats deserve stating plainly. This is a two-centre trial in a highly selected BCLC-B population treated by operators skilled in both modalities, and the 88.6-month median overall survival in the experimental arm is exceptionally long for intermediate-stage disease — a number that will need reproduction in a Western, more aetiologically mixed cohort before it anchors practice. The direction of effect, though, is consistent with the principle that residual viable tumour after chemoembolization is the thing that kills these patients.
Post angle: TACE has always had a completeness problem: you embolize, and viable tumour survives at the rim. TORCH says go back and ablate what's left. The PFS effect is large and the toxicity cost is essentially nil — but hold the 88.6-month OS number loosely until it's reproduced outside two expert Chinese centres. #GIOnc #HCC #LiverCancer #InterventionalOncology
#3
Source: ESMO Open | Authors: Oh D-Y, … Meric-Bernstam F — MD Anderson and international multicenter (AstraZeneca / Daiichi Sankyo) | Published: 2026-07-30
Score: 9/20 — Base 5 (ESMO Open) + Phase II (+2) + biomarker-guided / HER2 IHC selection (+1) + colleague engagement on X (+1) = 9.
This is the biliary tract and pancreatic subgroup analysis of DESTINY-PanTumor02 Part 1, in which trastuzumab deruxtecan 5.4 mg/kg was given to patients with HER2-expressing (IHC 3+ or 2+) solid tumours after prior systemic therapy. In the biliary tract cohort (n=41), objective response rate was 22.0% by investigator assessment and 26.8% by independent central review, rising to 56.3% among centrally confirmed IHC 3+ tumours — a genuinely meaningful signal in a disease where second-line options remain thin. The pancreatic cohort (n=25) was a different story: ORR was 4.0% by investigator and 12.0% by central review, and median overall survival was 5.0 months versus 7.0 months in the biliary cohort. The safety asymmetry is as important as the efficacy one. Interstitial lung disease or pneumonitis occurred in 17.1% of the biliary cohort versus 4.0% of the pancreatic cohort — a rate high enough that HER2 IHC 3+ selection, not merely HER2 expression, should govern who is offered this drug. Both cohorts are small, and the investigator-versus-central-review gap in the pancreatic arm (4.0% vs 12.0%) is wide enough on n=25 to caution against over-reading either number.
Post angle: Two lessons in one subgroup analysis. First, HER2 in biliary tract cancer is a real target — 56.3% ORR in centrally confirmed IHC 3+ is not noise. Second, HER2 expression is not one biomarker across GI: the same drug at the same dose gave 22.0% in biliary and 4.0% in pancreatic. And with 17.1% ILD in the biliary cohort, selection has to be IHC 3+, not merely HER2-positive. #GIOnc #Cholangiocarcinoma #BiliaryTractCancer #HER2 #PrecisionMedicine
#4
Source: Annals of Oncology | Authors: Chen X, Qiu MY, … Zhang L, Liu ZY — Guangdong Provincial People's Hospital, Alibaba DAMO Academy, and international multicenter (China, Czech Republic) | Published: 2026-08-01
Score: 8/20 — Base 7 (Annals of Oncology) + large multicenter model development with independent external and real-world validation (+1) = 8. No study-type bonus — retrospective diagnostic-accuracy design, not a randomized trial.
COCA (COlorectal Cancer detection with AI) was developed on 1,321 colorectal cancer patients and 1,357 controls from two centres using a joint lesion segmentation and classification architecture with mixed-supervised learning, then validated across six centres and two large real-world consecutive cohorts. In multicentre international validation of 2,053 patients, area under the curve ranged from 0.967 to 0.996. Against radiologists, COCA improved sensitivity by 20.4 percentage points and specificity by 5.4. The real-world numbers are the point of interest: in 9,014 consecutive patients sensitivity was 88.2% with specificity 99.5%, and in a second cohort of 18,419 consecutive patients sensitivity was 86.6% with specificity 99.8% and positive predictive value 63.4%. Performance held across physical exams, emergency, outpatient, and inpatient settings. The framing that matters clinically is opportunistic screening — this is not a replacement for colonoscopy but a way of extracting colorectal cancer signal from the enormous volume of noncontrast abdominal CT already being performed for unrelated indications, where cancers are currently missed. The design is retrospective throughout, and prospective validation with clinical follow-through remains the necessary next step.
Post angle: The interesting claim isn't that AI beats radiologists on a curated set — it's that 86–88% sensitivity held up across 27,433 consecutive unselected scans done for other reasons. Every one of those is a CT a patient already had. Opportunistic screening is where imaging AI earns its keep. #GIOnc #CRC #AI #CancerScreening
Additional Papers of Interest
-
Annals of Oncology — the peer-reviewed paper behind the ASCO LBA3500 topline this digest covered on June 2; ctDNA-positive stage II colon cancer randomized to adjuvant chemotherapy versus observation, now with full methods and follow-up in print
-
Annals of Oncology — Tarazona and Parikh's accompanying editorial on CIRCULATE, arguing that a positive ctDNA result establishes risk but not yet the optimal regimen or the right timing
-
Lancet Oncology — Fokas and Rödel's companion editorial to this digest's lead paper, on why a decade of chemoradiotherapy intensification attempts in rectal cancer keeps failing
-
Clinical Cancer Research — phase I dose escalation in 7 patients, ORR 28.6% and DCR 71.4%, all cytokine release syndrome grade 1 with no dose-limiting toxicities; single-cell sequencing found baseline NK enrichment in responders
-
Journal of Hepatology — mechanistic counterpart to this digest's TORCH paper, arguing the problem after TACE is a dysfunctional vascular-immune niche rather than residual tumour bulk alone
-
Journal of Gastrointestinal Cancer — 11 cohort studies; delaying adjuvant chemotherapy raised the risk of death (HR 1.32, 95% CI 1.20–1.45), with the penalty concentrated beyond an 8-week threshold (HR 1.47, 95% CI 1.14–1.91)
Back to all digests