Roughly one patient in five never reaches the expected weight loss after MBS, and long-term regain after SG is well documented. These patients are identifiable at the 12-month visit. Today they are watched, coached, and treated only when failure is established.
Whether the drug works in this group is no longer the question. BARI-STEP, published this year, randomized 70 patients with less than 20% weight loss after gastric bypass (including one-anastomosis) or sleeve gastrectomy, a mean of seven years after surgery, to semaglutide 2.4 mg or placebo: 18.0% loss versus 0.4% gain at 68 weeks, an adjusted difference of 19.1 percentage points. Its population was mixed, 55.7% with recurrent weight gain and 44.3% with suboptimal initial response, and its exploratory comparison found no difference in drug response between the two presentations. BARI-OPTIMISE showed the same direction with liraglutide, as did a randomized trial after gastric banding. Efficacy is settled.
Timing is not. BARI-STEP recruited patients a mean of seven years past surgery, after failure had hardened. No trial has tested treating at first recognition, the 12-month visit, against the current standard of structured waiting. The field's own reviews name timing and duration as the open questions. The 2024 IFSO international consensus reached 100% agreement that medication after MBS should generally be withheld until the weight plateau, and 100% agreement that adding it for a suboptimal response can then improve outcomes. That is exactly our immediate arm: treat at the plateau, at first recognition. The triggered arm is what clinics actually do: wait for established failure. Both positions rest on expert agreement, not on a trial. This is the randomized test of the consensus pathway. A registered open-label pilot starts tirzepatide at 12 or 18 months but is small, unblinded, and not a strategy comparison.
The 2026 consensus draws the line for us, and this protocol respects it. The Delphi panel defined surgical nonresponse as less than 10% TWL at 12 months, defined recurrent weight gain as regain of more than 25% of lost weight from nadir, and treated a 10% swing in %EWL from nadir as normal physiologic fluctuation. It gave no 12-month label at all to the patient with 10 to 19.9% TWL, and the IFSO Delphi consensus expects the weight nadir up to 2 years after surgery (Obes Surg 2023). So a patient at 14% TWL at one year may still be on the way down, and calling that failure would be wrong. The protocol therefore splits eligibility into two tiers. Below 10% TWL at 12 months: consensus-defined nonresponse, randomized at once, waiting adds nothing. 10 to below 20% TWL: randomized only after a documented plateau, defined in Section 4, with a hard cap at month 18 so waiting cannot drift into neglect. Randomization is stratified by tier, and results are reported by tier.