DAY 3: Late-Breaking Clinical Trial Findings

Dr. Abdullah Sultany, from Guthrie Robert Packer Hospital in Sayre, PA, presented results from a real-world study of glucagon-like peptide-1 receptor agonist (GLP-1 RA) effects on mortality and healthcare utilization in patients with cirrhosis. The retrospective cohort included 3222 patients with cirrhosis seen between January 2022 and April 2026, of whom 542 received a GLP-1 RA and 2680 did not. The primary outcome was all-cause mortality after 3 months of GLP-1 RA exposure. Secondary outcomes were hospitalizations, emergency department (ED) visits, and hospital days. Compared with unexposed patients, the GLP-1 RA group included more women, had a higher body mass index and a greater prevalence of diabetes, and had less ascites. In multivariable Cox regression adjusted for disease severity (MELD 3.0 score), age, sex, ascites, heart failure, gastrointestinal bleeding, diabetes, chronic kidney disease, hypertension, and varices, GLP-1 RA exposure was associated with significantly lower all-cause mortality (adjusted hazard ratio [aHR], 0.28; 95% CI, 0.20–0.38). All-cause and liver-related hospital admissions and hospital days were also significantly lower in the exposed group, with no difference in ED visits. A sensitivity analysis in a propensity-matched cohort of 1124 patients confirmed the mortality benefit (aHR, 0.43; 95% CI, 0.26–0.71), which was independent of cirrhosis severity. Dr. Sultany acknowledged that the study was limited by its single-center, retrospective design and residual confounding from factors such as metabolic status or alcohol use. He suggested that if the data are confirmed in prospective randomized trials, GLP-1 RA treatment could represent a paradigm shift in the management of cirrhosis.

Dr. Aasma Shaukat, from NYU Langone Health in New York, NY, presented results from an updated version of the SimpleScreen CRC blood test, which was revised to improve detection of early-stage colorectal cancer (CRC) and precancerous lesions. The current version of the test, evaluated in the PREEMPT CRC study in 2025, received FDA approval in July 2026, but its sensitivity for early-stage disease is limited. To improve performance, the assay and the algorithm were updated and evaluated using untested samples from PREEMPT CRC and a subset of samples from the PREEMPT validation study. From 37,653 available samples, the evaluable cohort of 9001 patients included 96 with CRC, 1142 with advanced precancerous lesions (APLs), and 157 with high-grade dysplasia (HGD). The mean age was 60 years, with an even split between men and women; 10% of participants were Hispanic and 10% were Black. The primary endpoints were sensitivity for CRC and APLs, and the secondary endpoint was sensitivity for HGD. Specificity was locked at 90%, and sensitivity was weighted by age and sex. The updated test showed a sensitivity of 80.4% for CRC, 18.2% for APLs, and 41.9% for HGD. CRC sensitivity was similar to that of the previous version of the test across stages, except for stage III disease, which increased by 16.6 percentage points to 97.3%. Sensitivity remained lowest for stage I cancers, although detection of stage I T2 tumors improved 6.7 percentage points to 81.2%. Compared with the previous version, sensitivity rose by 4.6 percentage points to 18.2% for APLs and by 11 percentage points to 41.9% for HGD. In a modeling analysis assuming screening every 3 years from age 45 to 75 years, these gains translated into roughly 10% fewer CRC cases, 10% fewer CRC deaths, and 8% more life-years gained. Dr. Shaukat concluded that further assay and algorithm refinements may continue to improve sensitivity for APLs and HGD. For now, colonoscopy and stool-based tests remain the preferred screening options, and blood-based testing can be offered to patients who decline them, because some screening is better than none.

Dr. Marla C. Dubinsky, from the Icahn School of Medicine at Mount Sinai in New York, NY, presented 12-week results from SHINE-2, a phase 3 trial of the anti–interleukin-23 antibody mirikizumab in children and adolescents with moderately to severely active ulcerative colitis (UC). Treatment options for pediatric UC remain limited, and an earlier phase 2 study had shown that mirikizumab was effective and safe in patients aged 2 to 17 years with moderately to severely active disease. For the SHINE-2 study, patients had to have been diagnosed at least 3 months before enrollment and failed prior therapy. Concomitant medications were allowed. Induction doses were chosen by body weight based on pharmacokinetic modeling: patients weighing more than 40 kg received the adult dose of 300 mg IV, and lighter patients received 5 mg/kg IV, every 4 weeks for 12 weeks. Maintenance doses were administered subcutaneously (SC) at 200 mg every 4 weeks for the highest weight group, 100 mg for patients >20 to ≤40 kg, and 50 mg for patients <40 kg. Eligibility criteria and outcomes were similar to those of the adult trials. The primary endpoint was clinical remission by modified Mayo score at week 52 among patients who responded at week 12. Week 12 endpoints evaluated at the end of induction included clinical remission and response, endoscopic remission and improvement, and Pediatric Ulcerative Colitis Activity Index (PUCAI) remission and response. Because the trial had no placebo arm, the investigators used a Bayesian decision-theoretic approach: efficacy was established if the lower bound of the estimate for mirikizumab exceeded the upper bound of the placebo response in the adult LUCENT trials. Of the 63 patients enrolled, 54% had refractory disease or had failed advanced therapy, and enrolling very young children was difficult (only 4 patients were in the lowest weight group). At week 12, mirikizumab outperformed the adult placebo benchmarks in LUCENT-1 for clinical remission (41.3% vs. 14.6%), clinical response (81.0% vs. 43.4%), endoscopic remission (25.4% vs. 4.5%), and endoscopic improvement (50.8% vs. 21.0%). PUCAI outcomes were consistent across weight groups. Compared with adults, children had greater reductions in fecal calprotectin and faster, deeper symptomatic remission. Pharmacokinetic analyses confirmed appropriate weight-based dosing, and the safety profile matched that seen in adults. Dr. Dubinsky concluded that early treatment with mirikizumab led to meaningful clinical improvements in pediatric patients with UC. The 52-week SHINE-2 results will be presented next week.

Dr. Dubinsky also presented data from a comparative efficacy analysis of subcutaneous (SC) versus intravenous (IV) infliximab in patients with ileum-involved Crohn’s disease (CD). Rates of surgery for ileal CD have not fallen the way they have for UC since the introduction of advanced therapies, and infliximab may be a suitable option for these patients. A post hoc analysis of the phase 3 LIBERTY-CD trial had shown that SC infliximab was effective across disease locations, but no study has directly compared SC and IV infliximab in ileal CD. The investigators pooled patient-level data from four infliximab registration studies (two phase 1 trials and the phase 3 PLANET-CD and LIBERTY-CD trials) and focused on patients with ileal (L1) or ileocolonic (L3) disease. All patients in the pooled analysis received IV induction and then maintenance with either IV infliximab 5 mg/kg every 8 weeks or SC infliximab 120 mg every 2 weeks. Of 200 patients in the pooled dataset, 140 were included in a propensity score–matched cohort. Patients had a mean age of 35 years, most were White, and immunomodulator use was more common in the IV group. In the unmatched cohort, SC maintenance led to significantly higher rates of clinical remission, endoscopic response, and endoscopic remission at week 54 than IV maintenance. SC maintenance was also superior across all endpoints in the matched cohort. Among responders, drug concentrations were higher in the SC group. In a subgroup analysis limited to ileal (L1) disease, endoscopic response remained better with SC than with IV infliximab, although patient numbers were small (23 SC vs. 22 IV). Dr. Dubinsky suggested the benefit may reflect the more constant drug exposure with SC dosing, which may also make anti-drug antibodies less important. She noted that concerns about adherence with SC therapy are unproven, and freeing patients from infusion visits could improve adherence and quality of life. While it can be hard to change practice for patients who are doing well on IV infliximab, she emphasizes to patients that moving to SC infliximab does not mean switching drugs, only changing how the same drug is given.

Dr. Evan S. Dellon, from the University of North Carolina at Chapel Hill, NC, presented efficacy and safety results from CROSSING, a phase 3 trial of tezepelumab in adults and adolescents with eosinophilic esophagitis (EoE). Current standard-of-care treatments do not work well for many patients with EoE. Thymic stromal lymphopoietin (TSLP) is an epithelial alarmin that drives downstream type 2 inflammation at epithelial barrier surfaces. TSLP expression is elevated in EoE, and polymorphisms in the TSLP gene have been linked to the disease. Tezepelumab is a human monoclonal antibody against TSLP that is already approved for severe asthma and chronic rhinosinusitis. In this 52-week trial, 368 patients were randomized to one of two tezepelumab doses or placebo. Nearly all patients remained in the study through weeks 24 and 52. The co-primary endpoints were histologic remission and change in dysphagia symptoms. Baseline characteristics were well balanced across groups: patients had been diagnosed 5 years earlier on average, were highly symptomatic, and had severe disease. Prior treatments were also balanced, and 4% of patients had previously received a biologic. Response rates for the co-primary endpoint were 65% to 69% with tezepelumab compared with 3.5% in the placebo group. Symptom improvement began as early as week 4 and continued through week 52. At week 24, tezepelumab also significantly improved the secondary endpoints of endoscopic severity and histologic severity (both grade and stage). Adverse events were similar across groups. The most common were nasopharyngitis, headache, and diarrhea, which occurred at similar rates in the placebo group, and two serious adverse events were considered related to the study drug. The 52-week results will be presented at UEG Week in Barcelona.