ObesityWeek 2025 – Machine-Learning Gene Risk Score Predicts Semaglutide Response After Bariatric Surgery
A predictive tool developed at Mayo Clinic showed potential to identify semaglutide responders among post-bariatric patients.
Medical Affairs
3 min read
November 4, 2025
Metabolic and bariatric surgery remains the most effective treatment for obesity, yet weight recurrence continues to be a common challenge over time. Semaglutide has demonstrated substantial efficacy in promoting weight loss, though with high interindividual variability, highlighting the need for reliable predictors of treatment response and a more personalized approach to obesity management.
Researchers from the Mayo Clinic retrospectively analyzed data from 68 post-bariatric patients included in the Mayo Clinic Biobank (n=52,619 genotyped individuals) who had undergone Roux-en-Y gastric bypass (75%) or sleeve gastrectomy (25%) and were treated with semaglutide (0.25–2.4 mg). The team developed a machine-learning–based genetic score, the Calories to Satiation Gene Risk Score (CTSGRS), derived from 41 gut–brain axis genes, to evaluate predictors of treatment response.
Participants were stratified as HG (+) (CTSGRS <0.50) and HG (–) (CTSGRS ≥0.50). After six months of treatment, total body weight loss (TBWL%) was –10.0 ± 6.6 in the HG (+) group and –5.2 ± 6.6 in the HG (–) group (p = 0.05). The model achieved an AUC of 0.63 in predicting semaglutide response, defined as TBWL ≥ 10%.
These findings suggest that the CTSGRS may serve as a potential biomarker to identify semaglutide responders among post-bariatric patients, reducing response heterogeneity and advancing precision medicine in obesity care. Prospective studies are still required to validate the clinical applicability of this genetic score.
Editorial note: This content was developed with the support of artificial intelligence technologies to optimize clarity and structure. All material was carefully reviewed and validated by medical experts prior to publication to ensure scientific accuracy and adherence to editorial best practices.
References:
- Villamarin J, Espinosa MA, Fredrick TW, et al. Performance of a Machine-Learning Gene Risk Score on Semaglutide Response in Post-Bariatric Patients. Presented at: ObesityWeek 2025; Abstract Oral-003.
