Novel Gene Risk Score Predicts Outcomes After RYGB Surgery

A novel gene risk score, informed by machine learning, predicted weight-loss outcomes after Roux-en-Y gastric bypass (RYGB) surgery, a new analysis showed.

The findings suggested that the MyPhenome test (Phenomix Sciences) can help clinicians identify the patients most likely to benefit from bariatric procedures and at a greater risk for long-term weight regain after surgery.

“Patients with both a high genetic risk score and rare mutations in the leptin-melanocortin pathway (LMP) had significantly worse outcomes, maintaining only 4.9% total body weight loss [TBWL] over 15 years compared to up to 24.8% in other genetic groups,” Phenomix Sciences Co-founder Andres Acosta, MD, PhD, told Medscape Medical News.

The study included details on the score’s development and predictive capability.

‘More Precise Bariatric Care’

The researchers recently developed a machine learning-assisted gene risk score for calories to satiation (CTSGRS), which mainly involves genes in the LMP. To assess the role of the score with or without LMP gene variants on weight loss and weight recurrence after RYGB, they identified 707 patients with a history of bariatric procedures from the Mayo Clinic Biobank. Patients with duodenal switch, revisional procedures, or who used antiobesity medications or became pregnant during follow-up were excluded.

To make predictions for 442 of the patients, the team first collected anthropometric data up to 15 years after RYGB. Then they used a two-step approach: Assessing for monogenic variants in the LMP and defining participants as carriers (LMP+) or noncarriers (LMP-). Then they defined the gene risk score (CTSGRS+ or CTSGRS-).

The result was four groups: LMP+/CTSGRS+, LMP+/CTSGRS-, LMP-/CTSGRS+, and LMP-/CTSGRS-. Multiple regression analysis was used to analyze TBWL percentage (TBWL%) between the groups at different timepoints, adjusting for baseline weight, age, and gender.

At the 10-year follow-up, the LMP+/CTSGRS+ group demonstrated a significantly higher weight recurrence (regain) of TBW% compared to the other groups.

At 15 years post-RYGB, the mean TBWL% for LMP+/CTSGRS+ was -4.9 vs -20.3 for LMP+/CTSGRS-, -18.0 for LMP-/CTSGRS+, and -24.8 for LMP-/CTSGRS-.

Further analyses showed that the LMP+/CTSGRS+ group had significantly less weight loss than LMP+/CTSGRS- and LMP-/CTSGRS- groups.

Based on the findings, the authors wrote, “Genotyping patients could improve the implementation of individualized weight-loss interventions, enhance weight-loss outcomes, and/or may explain one of the etiological factors associated with weight recurrence after RYGB.”

Acosta noted, “We’re actively expanding our research to include more diverse populations by age, sex, and race. This includes ongoing analysis to understand whether certain demographic or physiological characteristics affect how the test performs, particularly in the context of bariatric surgery.”

The team also is investigating the benefits of phenotyping for obesity comorbidities such as heart disease and diabetes, he said, and exploring whether early interventions in high-risk patients can prevent long-term weight regain and improve outcomes.

In addition, Acosta said, the team recently launched “the first prospective, placebo-controlled clinical trial using the MyPhenome test to predict response to semaglutide.” That study is based on earlier findings showing that patients identified with a Hungry Gut phenotype lost nearly twice as much weight on semaglutide compared with those who tested negative.

Overall, he concluded, “These findings open the door to more precise bariatric care. When we understand a patient’s biological drivers of obesity, we can make better decisions about the right procedure, follow-up, and long-term support. This moves us away from a one-size-fits-all model to care rooted in each patient’s unique biology.”

Potentially Paradigm-Shifting

Onur Kutlu, MD, associate professor of surgery and director of the Metabolic Surgery and Metabolic Health Program at the Miller School of Medicine, University of Miami, Miami, commented on the study for Medscape Medical News. “By integrating polygenic risk scores into predictive models, the authors offer an innovative method for identifying patients at elevated risk for weight regain following RYGB.”

“Their findings support the hypothesis that genetic predisposition — particularly involving energy homeostasis pathways — may underlie differential postoperative trajectories,” he said. “This approach has the potential to shift the paradigm from reactive to proactive management of weight recurrence.”

Because current options for treat weight regain are “suboptimal,” he said, “prevention becomes paramount. Preoperative identification of high-risk individuals could inform surgical decision-making, enable earlier interventions, and facilitate personalized postoperative monitoring and support.”

“If validated in larger, prospective cohorts, genetic risk stratification could enhance the precision of bariatric care and improve long-term outcomes,” he added. “Future studies should aim to validate these genetic models across diverse populations and explore how integration of behavioral, psychological, and genetic data may further refine patient selection and care pathways.”

The study presented at Digestive Disease Week (DDW) 2025 was funded by Mayo Clinic and Phenomix Sciences. Gila Therapeutics and Phenomix Sciences licensed Acosta’s research technologies from the University of Florida and Mayo Clinic. Acosta declared receiving consultant fees in the past 5 years from Rhythm Pharmaceuticals, Gila Therapeutics, Amgen, General Mills, BI, Currax, Nestle, Phenomix Sciences, Bausch Health, and RareDiseases, as well as funding support from the National Institutes of Health, Vivus Pharmaceuticals, Novo Nordisk, Apollo Endosurgery, Satiogen Pharmaceuticals, Spatz Medical, and Rhythm Pharmaceuticals. Kutlu declared having no conflicts of interest.

Marilynn Larkin, MA, is an award-winning medical writer and editor whose work has appeared in numerous publications, including Medscape Medical News and its sister publication MDedge, The Lancet (where she was a contributing editor), and Reuters Health.