METABOLIC & BARIATRIC SURGERY

Vol. 2, Issue 14   |   October 13, 2026

This issue covers GLP-1 therapy for weight recurrence, patient satisfaction 12 years out, the modeled risk of staged SADI-S, phase 3 olatorepatide data, and early work on AI-guided trocar entry and magnetic anastomosis.

A — Metabolic Surgery Research

Tarik Yuce, Rotana M. Radwan, Yao An Lee, et al. | Obesity, Aug. 25, 2026 | 10.1002/oby.70280
Suggested by Phil Schauer

Investigators used the OneFlorida+ electronic-record network to compare adults with at least 10% regain from nadir weight who started a GLP-1 receptor agonist against matched nonusers. After inverse probability weighting, the 12-month analysis included about 72 users and 139 nonusers. Users lost 9.6% of their regained weight at 6 months versus 4.3%, and 14.8% versus 8.1% at 12 months; the measure is relative to the regained portion, not total body weight. Semaglutide was the most common agent (52%), with tirzepatide in 9%.

Benefit separated earlier in patients with type 2 diabetes (13.6% versus 4.6% at 6 months) and by 12 months in those without it (14.6% versus 7.1%). The composite adverse-event risk did not differ (hazard ratio 0.80, 95% CI 0.61 to 1.05 at 12 months), and abdominal pain was less frequent among users (15.3% versus 25.2%). Two randomized liraglutide trials had shown benefit in this setting; this cohort adds mixed agents and routine-care weights. Treatment was not randomized, so residual confounding is the main limit.

KEY FINDING

Starting a GLP-1 agonist was associated with 14.8% loss of regained weight at 12 months versus 8.1% in matched nonusers, without a significant difference in the composite of gastrointestinal and post-surgical adverse events (hazard ratio 0.80, 95% CI 0.61 to 1.05).

Alexandra Burtea, Jaime Almandoz, Marielle Nagele, et al. | Obesity Surgery, Oct. 8, 2026 | 10.1007/s11695-026-08984-9
Suggested by John D. Scott, MD; Phil Schauer

The Hindsight 20/20 study invited 15,991 adults with prior bariatric surgery at a large academic center and linked the 2,385 responses (15%) to health records. Respondents were a mean 12.0 years from surgery; 42% had a sleeve, 31% a bypass and 23% a band. Overall, 74.2% would choose surgery again, 7.3% were unsure and 18.5% would not. Among those willing, 89% cited overall health benefit and 88% weight loss.

Among respondents who were unsure or unwilling, 56% cited weight recurrence, 49% nonsurgical complications such as hypoglycemia and nutrient deficiency, and 41% insufficient weight loss; 24% said they would have chosen a different operation. In the exploratory adjusted model, each added year since surgery (odds ratio 0.95) and each point of current BMI (0.94) lowered the odds of choosing surgery again, while greater maximum weight loss raised them (1.03). Respondents resembled the full cohort in age and sex, but a 15% response rate still leaves the percentages open to nonresponse bias.

KEY FINDING

A mean 12 years after surgery, 74.2% of respondents would choose it again; weight recurrence (56%), nonsurgical complications (49%) and insufficient loss (41%) were the leading reasons among those who were unsure or unwilling.

Akio Kozato, Suhani S. Patel, Eduardo Somoza, et al. | Surgery for Obesity and Related Diseases, Oct. 7, 2026 | 10.1016/j.soard.2026.09.025
Suggested by John D. Scott, MD

Using MBSAQIP data from 2020 to 2024, the investigators identified 749 SADI-S operations at BMI 60 to 70, 22,503 sleeve gastrectomies at the same BMI and 1,071 sleeve-to-SADI conversions at BMI 42 to 49, the range expected after about 30% weight loss. After inverse probability weighting, primary SADI-S carried more 30-day complications than sleeve alone (6.7% versus 4.5%). The authors combined sleeve and conversion risks from these separate cohorts to model the staged pathway, which reached 11.5%.

The modeled two-stage pathway also had higher readmission (7.3% versus 4.3%), reoperation (3.4% versus 1.3%), non-operative intervention (2.1% versus 0.8%) and emergency visits (17.5% versus 11%); deaths, leaks and bleeding were comparable. The conversion operation alone had more 30-day reoperations than primary SADI-S (2.8% versus 1.3%), which the authors suggest may reflect re-sleeving, a detail the registry could not identify. The two-stage figure is a product of unlinked cohorts, so it estimates the pathway and does not follow the same patients through both operations.

KEY FINDING

Modeled two-stage SADI-S carried 11.5% 30-day complications against 6.7% for one stage, with readmission 7.3% versus 4.3% and reoperation 3.4% versus 1.3%; death, leak and bleeding rates were similar.

ChatGPT is insanely powerful.

But most people waste 90% of its potential by using it like Google.

These 1,000+ proven ChatGPT prompts fix that and help you work 10X faster.

Sign up for Superhuman AI and get:

Claim your free prompts

B — Science of Obesity Metabolism

Yu-Yang Lee, Young-Jen Lin, Wei-Shiung Yang, et al. | Surgical Endoscopy, Oct. 9, 2026 | 10.1007/s00464-026-13491-0
Suggested by John D. Scott, MD

In a prospective single-center cohort from Taiwan, 105 patients (52 with type 2 diabetes) had circulating GDF15 measured before sleeve gastrectomy and at 6 and 12 months. Baseline levels were higher with diabetes (844 versus 558 pg/mL). Levels rose at 6 months in both groups, to 1,021 and 681 pg/mL, then fell back toward baseline at one year with diabetes (831 pg/mL) but stayed above it without diabetes (650 pg/mL). Mean one-year total weight loss was 32.1%.

Higher baseline GDF15 was associated with less one-year weight loss (r = -0.30), and the association held in stepwise regression alongside baseline BMI, age and hsCRP. GDF15 also tracks age (r = 0.42) and kidney function (r = -0.50 with eGFR), so its link to weight loss sits among correlated variables. Within each diabetes subgroup, one-year GDF15 and its change did not correlate with weight loss, and the paper offers no cutoff that separates faster from slower responders.

KEY FINDING

GDF15 followed different one-year courses by diabetes status, and higher baseline levels went with lower total weight loss (r = -0.30, n = 105).

L. Ji, L. Zhu, L. Gao, et al. | EASD 2026 late-breaking oral presentation LBA 27, Oct. 1, 2026 | Source ↗ | Hansoh release, October 2
Suggested by John D. Scott, MD

Hansoh's October 2 release reports LIGHTEN, a randomized, double-blind, placebo-controlled phase 3 trial in 604 Chinese adults with overweight or obesity, presented at EASD 2026. Weekly olatorepatide at 5, 10 and 15 mg produced mean weight loss of 12.5%, 17.5% and 19.3% at 48 weeks (efficacy estimand) against 2.7% with placebo; treatment-regimen estimates were 11.8%, 17.0% and 18.7%. Loss of at least 15% was reached by 33%, 60% and 72% of patients at the three doses and 3% on placebo.

The 10 and 15 mg arms were close at the 5% and 10% thresholds (97% versus 95%; 89% versus 89%) and separated at 15%. Pooled nausea was 7.8% and vomiting 4.9% against 2.6% and 2.0% with placebo, with no treatment-related discontinuations; the sponsor calls this lower than published rates for approved agents, a cross-trial comparison. China's NMPA accepted a weight-management application in June 2026, and Regeneron holds overseas rights. The results are sponsor-reported and not yet peer-reviewed.

KEY FINDING

Sponsor-reported weight loss at 48 weeks was 19.3% at 15 mg and 12.5% at 5 mg versus 2.7% with placebo; the 15 mg arm separated from 10 mg mainly at the 15% loss threshold (72% versus 60%).

Yu-Fang Ling, M.-Z. Liang, Y.-W. Liu, et al. | EASD 2026 late-breaking presentation LBA 57, Sept. 30, 2026 | Source ↗ | Science Media Centre report
Suggested by John D. Scott, MD

A conference abstract at EASD 2026 reported a subgroup analysis from two phase 2 trials of CBL-514, an injectable fat-reducing agent. MRI at the L2 to L3 level showed lower abdominal visceral-fat volume than placebo, with follow-up extending 8 to 12 weeks after the final dose. In the Science Media Centre's expert reaction, Marie Spreckley of Cambridge noted mild to moderate transient injection-site reactions and no serious adverse events. Separate rat experiments found that adding CBL-514 to tirzepatide reduced weight regain after tirzepatide was stopped.

Justin Rochford of Aberdeen cautioned that fat reduction is not automatically beneficial, since loss from healthy stores can shift fat to the liver and worsen lipids and glucose, which makes liver fat, circulating lipids and glycemic control the outcomes that will decide the drug's value. He noted that the agent also reduces subcutaneous fat, a healthy fat store, that the abstract gave no participant numbers, and that the placebo group gained weight and fat volume, which widens the apparent effect.

CONTEXT

The human data are an MRI visceral-fat reduction in a small phase 2 subgroup; the maintenance effect after incretin withdrawal has been tested only in rats.

Iuliia Karavaeva, et al. | Science, Oct. 1, 2026 | 10.1126/science.adz4797

Karavaeva and colleagues used mouse brown fat and cultured adipocytes to identify what integrates circadian, temperature and dietary signals. The mitochondrial carrier SLC25A34 followed a daily rhythm: REV-ERB repressors switched it off during sleep, PPARα and PPARγ restored it after waking, and cold exposure or a lipid-rich meal overrode the clock to raise it further. Inducible deletion in mouse brown adipocytes lowered norepinephrine-induced respiration.

The proposed mechanism has SLC25A34 import oxaloacetate into mitochondria, sustaining the TCA cycle while adipocytes break down and resynthesize lipids at the same time. Knockdown in primary brown adipocytes from four human donors reduced cAMP-inducible respiration in three, and a meta-analysis of 24 studies linked higher subcutaneous-fat expression to lower BMI, waist circumference, triglycerides and HOMA-IR, with no association in omental fat. Direct oxaloacetate transport has not yet been shown in reconstituted systems.

CONTEXT

SLC25A34 emerges as a convergence point for clock, cold and diet signals in thermogenic fat, supported by mouse genetics and human adipocyte data; its role as an oxaloacetate importer remains a proposed mechanism.

C — Metabolic Innovation and Technology

Mohammad Javad Farzadmanesh, Ali Jaliliyan, Parisa Tavana, et al. | Surgical Endoscopy, Oct. 9, 2026 | 10.1007/s00464-026-13404-1
Suggested by John D. Scott, MD

Investigators annotated 951 video frames from 10 patients undergoing direct optical trocar insertion (five with prior abdominal surgery) into rectus muscle, rectus sheath, peritoneum and background, then trained a DeepLabV3+ network with patient-level five-fold cross-validation. Mean Dice was 0.883 and mean intersection over union 0.794. By layer, Dice was 0.923 for muscle, 0.903 for sheath and 0.824 for peritoneum, and confusion between tissue classes was 0.0 to 0.3%, with most errors at tissue-background boundaries.

The peritoneum is the layer crossed last at entry and the one the model labeled least well, so the final transition is where an overlay has the least margin. The authors cite entry as the phase of nearly half of major laparoscopic complications, but this study measured agreement with expert labels and not injury rates. Each fold held out about two of the 10 patients, so performance across a wider range of body habitus and difficult views is unmeasured.

KEY FINDING

A DeepLabV3+ model labeled abdominal wall layers in 10 patients with mean Dice 0.883, highest for muscle (0.923) and lowest for peritoneum (0.824).

Yung Lee, Jerry Dang, Benjamin Clapp, et al. | Surgery for Obesity and Related Diseases, Sept. 29, 2026 | 10.1016/j.soard.2026.08.032
Suggested by Jerry Dang, MD, PhD

Disclosure: Board member Jerry Dang is a co-author and suggested this article; he recused himself from voting on it.

The ASMBS Emerging Technology and Procedures Committee synthesizes the evidence for AI across the perioperative continuum: preoperative risk stratification, intraoperative computer vision and robotics, ambient clinic documentation and postoperative monitoring. Its eight recommendations cover the role of AI as clinical decision support, MBS-specific validation, equity audits, documentation, governance, technical risks, workflow integration and learning health systems. The committee describes AI development in the field as resting mostly on retrospective, single-center data.

Several recommendations are specific. The committee asks programs to audit performance across BMI strata, age, sex, race and ethnicity, rural versus urban referral and insurance status, to record AI use in the health record and consent forms along with how disagreements between AI output and clinician judgment are resolved, and to decide whether each tool falls under clinical, research or quality-improvement oversight. It names dataset shift, model drift, poor explainability and false reassurance as the main technical risks.

KEY FINDING

The committee frames AI as clinician-led decision support and ties adoption to MBS-specific validation, subgroup audits and post-deployment monitoring.

Ricardo Nassar, Felipe Giron, Mateo Visbal-Zuluaga, et al. | Obesity Surgery, Oct. 5, 2026 | 10.1007/s11695-026-08973-y

A Bogotá series describes three consecutive patients, aged 41, 50 and 67, who underwent laparoscopic magnetic duodeno-ileostomy (GT Metabolic MagDI) after sleeve gastrectomy for insufficient response or weight regain, with the magnets docked 300 cm proximal to the ileocecal valve. Operative times were 113, 72 and 70 minutes, and all three went home the same day. Contrast study and endoscopy at three to four months showed a patent anastomosis in each, without stricture or retained device. Six-month total weight loss from the conversion date was 7.6%, 5.3% and 20.2%.

The longest operation needed an enterotomy 20 cm distal to the 300 cm mark to seat the distal magnet. One patient had persistent meal-related symptoms compatible with early dumping, managed with diet, and there were no leaks, bleeding, readmissions or reinterventions through six months. The authors state that three patients cannot support conclusions about safety or efficacy. The institution bought the devices at commercial cost, and first author Ricardo Nassar has been a speaker for GT Metabolic.

KEY FINDING

Magnetic conversion after sleeve produced a patent anastomosis in all three patients, with six-month weight loss of 5.3% to 20.2% and one case of persistent dumping-type symptoms.

D — Metabolic Marketplace

Ascensia Diabetes Care | Ascensia Diabetes Care / Business Wire, Oct. 9, 2026 | Source ↗

Ascensia Diabetes Care, a subsidiary of PHC Holdings, announced on October 9 that it led Sava's Series B financing, made a strategic equity investment and obtained exclusive commercialization and distribution rights, beginning in Europe. Sava's wearable sits just beneath the skin and is designed to measure biomarkers and molecules, starting with glucose, for non-adjunctive use. The company plans a pivotal study this year and targets CE approval and launch in 18 to 24 months. The release gives no investment amount and no accuracy data.

Non-adjunctive use means labeling for treatment decisions without fingerstick confirmation, which puts accuracy in the low glucose range at the center of any role in post-bariatric hypoglycemia monitoring. The pivotal study is where those figures will first appear. Ascensia sells in more than 90 countries and describes planned availability through retail channels under established reimbursement structures; the timeline is a company target pending regulatory approval.

CONTEXT

A financing and distribution agreement for a sensor still in development, with a pivotal study planned for 2026 and CE approval and launch targeted within 18 to 24 months.

Closing the gaps in knowledge...

Production Editor John D. Scott MD FACS | [email protected]

All content voted on by the Editorial Board
MBS Digest | October 13, 2026 | Vol. 2, Issue 14
For educational purposes only.