This week in Gastroenterology
Nov 17, 2026
Efficacy and safety of dabigatran in cirrhosis with chronic portal vein thrombosis: A randomized, placebo-controlled double-blind trial (DAB-PVT).
In 84 patients with Child-Pugh A or B cirrhosis and chronic portal vein thrombosis, dabigatran achieved complete recanalization in about 64 percent versus 21 percent on placebo, with fewer variceal bleeds.
American Journal of Gastroenterology · 2026 · PubMed
This week’s highlights
Read the weekly briefing
This AudioScholar briefing is generated by artificial intelligence from PubMed abstracts for healthcare professionals and is not medical advice.
This week: a blood thinner reopens clotted portal veins in cirrhosis, national data question direct oral anticoagulants for venous clots in liver disease, and robotic pancreatic surgery matches open surgery on complications.
Welcome to This Week in Gastroenterology. We're covering 10 papers this week: 3 in depth, 4 in brief, 3 quick signals.
The American Journal of Gastroenterology published a randomized, placebo-controlled, double-blind trial from Meena and colleagues testing dabigatran for chronic portal vein thrombosis in cirrhosis [1]. Eighty-four patients, mostly Child-Pugh A or B, received dabigatran 150 mg twice daily or placebo for twelve months. These were not mild patients: more than half of them had medium-to-large varices, and close to half of them had bled from varices before. Complete recanalization of the portal vein at twelve months occurred in about 64 percent of patients on dabigatran against about 21 percent on placebo, roughly a threefold difference. Variceal bleeding episodes were less frequent with dabigatran, three episodes against twelve. Child-Turcotte-Pugh and ALBI scores improved more with dabigatran, while MELD trajectories did not differ between groups. The main limitation is size: one trial of eighty-four patients cannot settle hard outcomes, and the authors themselves describe the differences in clinical events and severity scores as exploratory, needing larger comparative studies. What it adds is the first placebo-controlled evidence that a direct oral anticoagulant restores portal flow in this population, without an excess of bleeding over the trial period.
A different picture emerges from a nationwide cohort in Clinical Gastroenterology and Hepatology, led by Tanaka [2]. Using Epic Cosmos records from 2019 to 2024, the investigators compared propensity-matched patients with cirrhosis starting a direct oral anticoagulant or warfarin, across more than ninety-three thousand initiators. In atrial fibrillation, direct oral anticoagulants were associated with similar mortality and transplant rates but less major bleeding, including intracranial hemorrhage. In venous thrombosis the direction reversed. For portal vein thrombosis, direct oral anticoagulant initiators had roughly 43 percent higher mortality, transplant rates about three quarters lower, and more gastrointestinal bleeding than warfarin initiators. In a smaller transplant registry cohort, warfarin-driven inflation of the INR, and therefore of MELD, explained just under a third of the transplant gap. This is observational data with the usual risk of residual confounding, and the comparator is warfarin, whereas the dabigatran trial used placebo. The two papers address the same clinical question with different comparators, and they leave the choice of agent for portal vein thrombosis unsettled.
The Lancet Gastroenterology and Hepatology published an individual patient data meta-analysis led by Menso comparing robotic and open pancreatoduodenectomy [3]. It pooled four randomized trials covering 745 patients at 25 experienced, high-volume centres in seven countries. Major complications occurred in about 34 percent of patients after robotic surgery and 32 percent after open surgery, a difference that was not statistically significant. Functional recovery was about two days faster with the robotic approach, a median of nine days against twelve. Exploratory subgroup analyses by age, obesity and pancreatic fistula risk found no sign that the effect varied between groups. Three of the four trials carried some concerns about bias, and every centre was high volume, so the findings describe expert practice rather than general surgical practice. What it adds is pooled randomized evidence large enough to look across subgroups, which the individual trials could not do, supporting the robotic approach as a safe alternative in experienced hands.
Now, four in brief.
In The Lancet Gastroenterology and Hepatology, Noureddin and colleagues report 48-week results of IMPACT, a phase 2b trial of weekly pemvidutide in 212 patients with biopsy-confirmed steatohepatitis and F2 or F3 fibrosis across 83 sites in the United States and Australia [4]. At the higher 1.8 mg dose, liver fat fell by about 55 percent against about 8 percent on placebo, and corrected T1, ALT and the enhanced liver fibrosis score all improved significantly. Nausea affected about two in five patients on the higher dose. No biopsies were taken at 48 weeks, so these are non-invasive surrogate markers, and the authors call for a phase 3 trial.
Also in Clinical Gastroenterology and Hepatology, a prospective multicenter cohort led by Tapper followed 230 patients with cirrhosis, two in five of whom were on a sedating medication at baseline [5]. Each additional sedating drug was associated with about a third higher hazard of falls and a forty to fifty percent higher hazard of hospitalization. Taking two or more of these drugs roughly doubled both. The link with overt hepatic encephalopathy reached significance only in some adjustment models. Nonselective beta-blockers and diuretics showed no association with falls or hospitalization. The cohort is modest and observational, and the authors call for deprescribing trials.
In Alimentary Pharmacology and Therapeutics, a narrative review led by Rabiee examines cirrhosis recompensation as defined by Baveno seven and eight [6]. The authors estimate that about 35 percent of patients with decompensated cirrhosis recompensate, with lower liver-related mortality, though rates vary by cause. Liver cancer risk and, in some patients, portal hypertension persist. The definition remains unsettled for causes without a curative therapy. As a trial endpoint, the authors judge recompensation meaningful but dependent on careful handling of competing risks. This is a synthesis of existing evidence, not new data.
Clinical Gastroenterology and Hepatology also carries LIFT-HCC-1, a global retrospective study by Rezaee-Zavareh and colleagues of 408 patients with liver cancer who received immune checkpoint inhibitors before transplant [7]. Over a median of about 21 months, roughly one in five patients had rejection, about a quarter had recurrence, and about a fifth died. A washout longer than 90 days was associated with about half the rejection risk, while non-hepatitis B causes and immune-related adverse events raised it. Downstaged patients did as well as those always within Milan criteria. There was no control group, and the authors call for prospective validation.
Three quick signals.
In Gastrointestinal Endoscopy, Magahis and colleagues report the first international Delphi consensus on endoscopic ultrasound-guided coil embolization of gastric varices [8]. Thirty expert endoscopists took part, and almost all statements reached consensus, favouring careful variceal classification and coils combined with adjunctive therapy. This is expert opinion rather than trial evidence.
In Gastroenterology, Crockett and colleagues review serrated neoplasia, which accounts for up to 30 percent of colorectal cancers and a disproportionate share of post-colonoscopy cancers [9]. The review covers the updated World Health Organization histology and the sessile serrated lesion detection rate as a quality benchmark. It is a review, not new data.
The Lancet carries a review of Lynch syndrome by Ryan and colleagues [10]. It affects about one in 350 to 450 people, and most carriers remain undiagnosed despite universal tumour testing. The review covers surveillance, aspirin, risk-reducing surgery and checkpoint inhibitors, and it is a synthesis of existing evidence.
Stepping back across this week's papers, the evidence on anticoagulation in cirrhosis shows how a small placebo-controlled trial and a very large observational cohort can answer neighbouring questions and still pull in different directions, because they use different comparators and different endpoints. Elsewhere, much of the week rests on surrogate markers, expert consensus, uncontrolled cohorts or narrative synthesis, which describe direction well but leave hard outcomes to future trials.
If you only have time for one paper this week, make it the dabigatran trial in portal vein thrombosis [1]. It provides the first placebo-controlled evidence on a direct oral anticoagulant in cirrhosis, and it reopens the question of which agent and which patients, given the conflicting nationwide data.
That's your roundup for This Week in Gastroenterology. The full transcript and references are available on the episode page. See you next week.
If this weekly briefing is useful, follow the show in your podcast app so new episodes arrive automatically. And think of one colleague, in any specialty, who never has time to keep up with the literature. Tell them about AudioScholar: a ten-minute weekly for every specialty, to listen to in any podcast app, or to read at audioscholar dot C C.
This is an automated summary generated by artificial intelligence, which can make mistakes. Use it to choose which papers to read in full.
Get Gastroenterology every week
By email, Spotify, Apple or any podcast app.
Spot something worth flagging?