


GI Medical Oncologist & Associate Attending @MSKCancerCenter. Views my own. #Meded


🧬 Can ctDNA guide adjuvant immunotherapy in resected dMMR tumors? 🔬 Phase II: 23/184 MRD+ ⏳ 6/19 relapsed before treatment 💊 Pembrolizumab (n=13): ✅ ctDNA clearance: 85% ⚠️ Recurrence: 38% 💡 Clearance ≠ cure. Clinical benefit remains unproven. @OncoAlert

OK, last NEN article post of the night (thanks, Delta, for excellent in-air wifi and for keeping me hypercaffeinated). This is a really cool paper, an @ENETS_ORG consensus paper led by @JHernando3, an expert panel, on unresolved clinical scenarios and emerging approaches with somatostatin analogues. Yes, I read it already! It is great, clinically helpful, hits many of the classic SSA controversies, and if you see patients with NETs, you should read it too... onlinelibrary.wiley.com/doi/…

Are you like me, feeling intimidated by and unsure about the management of patient with mixed neuroendocrine/non-neuroendocrine neoplasms (MiNENs) of gastroenteropancreatic origin? If so, this is a review article you will want to read. Excellent @ENETS_ORG position statement, led by @FRANCES88189587 and a panel of folks who know their stuff. Check it out! onlinelibrary.wiley.com/doi/…

Real-world anti-EGFR rechallenge in ctDNA-profiled mCRC: the REalCHALLENGE study European Journal of Cancer 👉ctDNA WT: ORR 29% vs 0%, DCR 68% vs 42% mPFS 4.7 vs 2.9, mOS 21.9 vs 11.1 mo 🧐support value of molecular profiling to optimize selection for anti-EGFR rechallenge @myESMO @ASCO

We still do not know if LV adds a benefit without bolus 5-FU and we could not optimize real world data without significant confounding factors / no funding for the pragmatic clinical trial.
💊 Leucovorin shortage dosing in mCRC (JCO Global Oncology) Leucovorin shortages are recurrent and global. The guidelines say nothing. Riyadh tested what centres were already quietly doing. 👥 99 patients 2020–2025: 50 at leucovorin 200 mg/m² vs 49 at 400 mg/m² ➖ mPFS 8.0 vs 8.1 months, p=0.943 ➖ mOS 71 months vs not reached, p=0.336 📉 Grade 1 diarrhoea 18.4% vs 51%, p=0.001 🧮 1,180 vials spared ⚠️ Retrospective, n=99, not powered for non-inferiority - reassurance, not equivalence. REMEMBER: Omitting the 5-FU bolus is well supported. A 2026 systematic review and meta-analysis (7 studies, 12,698 patients with GI cancers) found no difference in PFS (HR 0.94, 95% CI 0.83–1.07) or OS (HR 0.96, 95% CI 0.89–1.03) between bolus-containing and bolus-free regimens, while the bolus arms had significantly more grade 3-4 AE. If the 5-FU bolus is not needed, is leucovorin still necessary? #CRC #ColorectalCancer #GIOnc ascopubs.org/doi/10.1200/GO-…

🧪 # — SIERRA, Phase IIIb | Journal of Hepatology Patients with poor prognostic features - Child-Pugh B liver function, ECOG 2, or main-trunk portal vein thrombosis (Vp4) - are routinely excluded from phase III HCC trials. 🎯 This study gave STRIDE (tremelimumab + durvalumab) to 👥 111 patients: Child-Pugh B7/B8 44 · ECOG 2 47 · Vp4 PVTT 20 ⚠️ Grade 3/4 related AEs at 6 mo: 19.8% overall, 25.0% in Vp4 📊 ORR 14.4% overall · ECOG 2 21.3% · Vp4 15.0% 📉 Child-Pugh B ORR only 6.8% This is the patient in clinic every week who was never in HIMALAYA. Toxicity looks manageable across all three cohorts, but the Child-Pugh B response rate of 6.8% is the sobering line - liver reserve still limits what immunotherapy can do. #GIOnc #HCC #LiverCancer pubmed.ncbi.nlm.nih.gov/4283…

Hard to compete with Tesla’s new Roadster, but check this out. 🔥 NEONIPIGA, 4-year follow-up | JAMA Oncology 🎯 Resectable dMMR/MSI-H gastric + GEJ adenocarcinoma 💊 Neoadjuvant nivo + ipi → surgery → adjuvant nivo 👥 n=32, median follow-up 48.3 mo ⏱️ 4-yr OS 84.1%, 4-yr EFS 83.5% 🔬 pCR 59% (17/29) 🔁 1 relapse in the whole cohort ✅ All 3 non-operated patients: clinical complete response ⚠️ Single-arm, no comparator Four years out, NEONIPIGA has one relapse in 32 patients with resectable MSI-H gastric/GEJ cancer, and every patient who skipped surgery is in clinical complete response. The authors won't claim a survival win without a control arm. The question that matters now is who still needs a gastrectomy. #GIOnc #GastricCancer #MSI doi.org/10.1001/jamaoncol.20…

🎯 Observation vs maintenance PD-1 after cCR in dMMR/MSI-H CRC, non-operative (JITC) The day after an FDA label made PD-1 monotherapy a first treatment for non-metastatic dMMR rectal cancer, here is an interesting clinical question: once the response is complete, do you keep dosing? Of 318 patients with dMMR/MSI-H colon and rectal cancer treated with PD-1 inhibitors at five Chinese hospitals (2018–2025). 173 (54.4%) achieved a clinical complete response --> (cCR) and --> 129 entered non-operative management: 👥66 stopped immunotherapy (observation) and 👥63 continued for at least two more cycles (maintenance). ⏱️ Median follow-up 32.2 mo from cCR ➖ 3-yr DFS 96.2% vs 98.3%, p=0.508 🔁 Only 2 regrowths, both colon, both salvaged 🚫 No distant metastases ⚠️ New toxicity after cCR: 27.0% of the maintenance group vs 1.5% of the observation group 🔬 Colon cCR → pCR at surgery in 40/41 One day after the dostarlimab label for rectal cancer: once it's complete, stopping looks reasonable. Note that vast majority of pts in this multicenter retrospective study were *non metastatic* cases. For Stage IV cases, current guidance rests on a guideline-endorsed 2-year ceiling supported entirely by retrospective cohorts. #CRC #dMMR #Immunotherapy jitc.bmj.com/content/14/10/e…

Look at these curves - 🔥 NICHE-2 at five years - Phase 2 (The Lancet) 🎯 Stage II–III dMMR colon cancer, before surgery 💊 Ipilimumab ×1 + nivolumab ×2, then resection 👥 111 per-protocol patients, six Dutch hospitals ✅ 3-year DFS 100% at a median 58.7 mo ✅ 5-year OS 100% (exploratory) 🩸 ctDNA cleared in 80/97 before surgery, all negative after ⚠️ Single-arm, no comparator The pathological response rates from NICHE-2 were already striking; this is the survival follow-up that tells us they held. Zero recurrences across 111 patients, after only roughly a month of treatment w/o chemo before surgery. The comparison that now matters is with *ATOMIC*, which added atezolizumab to adjuvant FOLFOX after surgery: two routes to curing dMMR colon cancer, one of them with six months of chemotherapy and one without. NICHE-2 is single-arm, so one can say that a randomized comparison is still owed before it replaces surgery-first. #CRC #dMMR thelancet.com/journals/lance…

🔥 AZUR-1 Phase 2, Annals of Oncology (ESMO LBA87) AZUR-1 is a global, single-arm, Phase 2 trial of 💊 Dostarlimab 500 mg q3w for 9 cycles in adults with 🎯 Untreated stage II/III dMMR/MSI-H rectal cancer 👥 154 enrolled, 144 with ≥12 mo follow-up ✅ Sustained cCR at 12 mo: 68% (investigator) ✅ cCR 1y after end of treatment + pCR: 76% 🔁 98% of cCRs held ≥12 mo, no recurrences in responders ⚠️ Single arm, central review less mature This is the full dataset behind the accelerated approval we covered on October 20, and it holds up. 2 in 3 pts keep their rectum with a PD-1 antibody alone, and no one in complete response has recurred so far. Two caveats: it is single-arm, and the investigator read runs ahead of central review, which the authors say is less mature. With a median follow-up of 26 months, late local regrowth is the number to watch. Everyone must test MMR on every rectal biopsy before anyone starts total neoadjuvant therapy. #CRC #RectalCancer #dMMR #ESMO26 #GIOnc annalsofoncology.org/article…

☢️ MASTERPLAN - randomised Phase 2 (Lancet Oncology) This is the clearest randomized evidence so far that ablative-range neoadjuvant SBRT after modern chemo decreases locoregional recurrence. 🎯 High-risk localised PDAC: >4 cm, node-positive, borderline or LAPC 💊 mFOLFIRINOX or gem–nab-paclitaxel ± SBRT 40 Gy/5 fx 👥 74 patients, closed early (120 planned) 📉 12-mo locoregional recurrence 20.2% vs 52.0%, sHR 0.44 ➖ Grade ≥3 AEs 70% vs 64% ⚠️ No survival endpoint reported Why did it close early ? slow accrual. Powered for PFS and OS: no. Radiotherapy in pancreatic cancer has a long history of trials that improved local control without moving survival. (when I say long I mean at least: PREOPANC-1 and 2, Alliance A021501, ESPAC-5, PANDAS-PRODIGE 44, CONKO-007, LAP07, SABER... ) MASTERPLAN shows modern SBRT, layered onto mFOLFIRINOX or gem–nab-paclitaxel, more than halves 12m locoregional recurrence without a clear toxicity penalty. For borderline and locally advanced patients where local failure drives symptoms, this is a strong argument for SBRT in the multidisciplinary plan, but not yet proof it changes survival. #PDAC #PancreaticCancer #SBRT #RadOnc #GIOnc thelancet.com/journals/lanon…

🔥 CLARITY-Gastric01 - Phase 3 (AstraZeneca release, ESMO LBA7) 🎯 Previously treated (≥2L) CLDN18.2+ gastric, GEJ or oesophageal adenocarcinoma - How positive, you may ask? Not the usual 75%, but rather 25% expression. 💊 Sonesitatug vedotin (CLDN18.2 ADC) vs investigator's choice 👥 532 patients ✅ 3L+ OS 10.0 vs 5.0 mo, HR 0.47 ✅ ITT OS 10.6 vs 6.4 mo, HR 0.58 ➖ PFS co-primary missed, HR 0.77, p=0.053 ⚠️ 2L-only OS HR 0.88, CI crosses 1 ⚠️ Prior zolbetuximab exposure, if any, not reported In 3L+, mOS doubles from 5 to 10 months, in a population defined at 25% CLDN18.2 expression, far broader than the zolbetuximab cut-off. Now read the second line carefully: PFS missed, and the second-line OS HR was 0.88 with a CI crossing 1, so the ITT result leans on later lines. Let’s please approve this already, at least for 3L. 🙏 But wait!! What can be said about patients with prior exposure to zolbetuximab ??? Also, G-HOPE-001 (see below), selected patients at a much stricter threshold (2+ in 75% or more), so do not compare the hazard ratios head to head. The open question for both ADCs is activity after first-line zolbetuximab. #GastricCancer #CLDN182 #ESMO26 #GIOnc astrazeneca.com/media-centre…

📊 HARMONi-GI1 - Phase 3 (Akeso release, ESMO LBA8) 🎯 First-line advanced biliary tract cancer, China 💊 Ivonescimab (PD-1/VEGF) + gem-cis vs durvalumab + gem-cis 👥 682 patients, interim analysis ✅ OS 14.9 vs 12.1 mo, HR 0.75 ✅ PFS 8.4 vs 5.8 mo, HR 0.61 📈 ORR 53.8% vs 32.5% 🩸 Grade ≥3 bleeding 1.5% vs 1.8% ⚠️ China only, interim analysis A head-to-head win over TOPAZ-1. The US ECOG-ACRIN trial will decide it for Western patients. Potential new SOC? Yes, but: ⚠️ This is an interim analysis with immature follow-up. Median follow-up was 14.3 months, and the 18-month OS difference rests on few patients still at risk. TOPAZ-1's benefit grew with longer follow-up (HR 0.80 → 0.76 → 0.74 → 0.75). Durability needs to be confirmed in both directions. ⚠️ Post-progression treatment was imbalanced. Only 31.8% of the ivonescimab arm received further therapy vs 45.9% of controls. This would normally dilute an OS benefit, so the HR may be conservative. However, the reasons for the gap (clinical decline, toxicity, how the denominator was defined) need to be in the full report. ⚠️ Subsequent therapy differs from US practice. In the US, second-line care for BTC increasingly involves targeted agents for FGFR2, IDH1, HER2, and others. Whether a China-only OS result carries over depends partly on that downstream care. FDA has stressed that subsequent therapies must match US practice when OS is the endpoint #BTC #Cholangiocarcinoma prnewswire.com/news-releases…

I didn't want to post this one: 📊 # PRESET-RCT (JAMA Network Open) a single-blinded, two-arm superiority trial in gastroesophageal adenocarcinoma planned for neoadjuvant treatment and curative surgery. 🔀 202 patients: twice-weekly exercise from diagnosis to surgery vs usual care ➖ No curative surgery: 9.9% vs 7.9%, OR 1.28, P=.62 ➖ No change in chemo discontinuation or complications 📉 3-yr OS 51.0% exercise vs 65.0% control, HR 1.58 (1.02–2.45) ⚠️ Secondary endpoint, one centre Uncomfortable findings. Prehabilitation is widely recommended (myself included), yet in a properly randomized trial it did not get more patients to surgery. 3yOS was numerically lower with exercise (51% vs 65%, HR 1.58, not statistically significant). That is a secondary endpoint, from one center, without a mechanism, so it should not be read as exercise harming patients. It is also not CHALLENGE: that trial showed a DFS and OS benefit from structured exercise after adjuvant chemotherapy for colon cancer, a different setting and timing. The honest message is that preoperative exercise during neoadjuvant chemotherapy for gastroesophageal cancer is unproven, and the survival signal needs replication before anyone treats it as safe or unsafe. #GastricCancer #Prehab #GIOnc doi.org/10.1001/jamanetworko…

🦄 UNICORN (Nature Medicine) - Phase 2 Short-course ('NICHE-2 style') preoperative botensilimab ± balstilimab 👥 56 resectable locally advanced colon cancers: 28 pMMR + 28 dMMR ❌ pMMR, botensilimab alone: pathological major response (pMR) 0% ✅ pMMR, combination: pMR 36% ✅ dMMR, combination: pMR 100% ⚠️ irAEs 29% vs 54%; cohorts of 14, no survival data 🩸 ctDNA clearance before surgery tracked response Two neoadjuvant BOT-BAL studies now report major pathological responses in roughly a third or more of pMMR colon cancers: UNICORN at 36% here, NEST at 41% . Note: ⚠️Anti-CTLA-4 alone did nothing in pMMR. ⚠️Small sequential cohorts and no survival data yet #CRC #ColonCancer #Immunotherapy nature.com/articles/s41591-0…