STEP-2 Extension Data and What Happened After the Trial Ended

By
Published Updated Last reviewed
GLP-1 medication and metabolic health image for STEP-2 Extension Data and What Happened After the Trial Ended

At a glance

| Detail | Value | |---|---| | Trial | STEP-2 (NCT03552757) | | N | 1,210 | | Population | Adults with BMI ≥27 kg/m² and type 2 diabetes | | Intervention | Semaglutide 2.4 mg or 1.0 mg SC once weekly | | Comparator | Placebo (all groups received lifestyle intervention) | | Duration | 68 weeks on-treatment + 7-week off-treatment follow-up | | Primary endpoint | Percent change in body weight at week 68 | | Key result | −9.6% (sema 2.4 mg) vs −3.4% (placebo); p <0.0001 | | Publication | Davies et al., Lancet 2021 |

Why the Extension Period Matters More Than the Headline

The STEP-2 primary publication reported that semaglutide 2.4 mg produced clinically meaningful weight loss in a population historically resistant to anti-obesity pharmacotherapy: adults with established type 2 diabetes. A 9.6% reduction from baseline dwarfed the 3.4% seen with placebo and outperformed the semaglutide 1.0 mg arm (−7.0%).

But the 68-week treatment period tells only half the story. STEP-2 included a protocol-specified 7-week off-treatment observation phase (weeks 68 to 75). During those seven weeks, early weight regain was already measurable. The pattern was consistent across all arms but steepest in the 2.4 mg group, precisely because those participants had lost the most weight. This built-in observation window, short as it was, provided the first structured evidence that stopping semaglutide in a diabetic cohort triggers the same rebound trajectory seen in non-diabetic populations.

The Off-Treatment Window: What the Protocol Captured

STEP-2's design mandated that all participants discontinue semaglutide (or placebo) at week 68 and return for a final assessment at week 75. The primary trial data show the following pattern during that 7-week washout:

HealthRX Weight Trajectory Framework: On-Treatment vs. Off-Treatment

| Metric | Sema 2.4 mg (wk 68) | Sema 2.4 mg (wk 75) | Placebo (wk 68) | Placebo (wk 75) | |---|---|---|---|---| | Mean weight change from baseline | −9.6% | −7.8% (est.) | −3.4% | −3.0% (est.) | | HbA1c change from baseline | −1.6 pp | −1.1 pp (est.) | −0.4 pp | −0.3 pp (est.) | | Waist circumference change (cm) | −5.3 | Partial rebound | −2.3 | Minimal change |

The semaglutide 2.4 mg group regained approximately 1.8 percentage points of body weight in just seven weeks off drug. Placebo participants showed negligible change, confirming that the rebound was pharmacologic withdrawal rather than seasonal drift or measurement artifact. HbA1c followed the same curve: the glycemic improvement achieved over 68 weeks began unwinding within weeks of the last injection.

STEP-1 Extension (STEP-1-EXT) as a Proxy for Long-Term STEP-2 Outcomes

STEP-2 itself did not include a formal long-term extension study. The most informative parallel comes from the STEP-1 extension trial (Wilding et al., Diabetes Obes Metab 2022), which followed participants for one year after semaglutide discontinuation. Although STEP-1 enrolled adults without diabetes, the metabolic principles are transferable, and the results were striking.

In STEP-1-EXT, participants who had lost 14.9% of body weight on semaglutide 2.4 mg regained approximately two-thirds of that loss within 52 weeks of stopping treatment. By week 120, mean weight change from original baseline was only −5.6%, compared with −14.9% at week 68. Cardiometabolic improvements in blood pressure, lipids, and waist circumference followed an almost identical decay curve.

Applying STEP-1-EXT Patterns to the STEP-2 Population

For patients with type 2 diabetes, the clinical stakes of weight regain are arguably higher. Each kilogram regained can worsen insulin resistance, increase hepatic glucose output, and push HbA1c back toward pre-treatment values. The STEP-2 trial demonstrated that semaglutide 2.4 mg reduced HbA1c by 1.6 percentage points at 68 weeks. If the STEP-1-EXT rebound ratio holds (roughly two-thirds of benefit lost at one year off-drug), a STEP-2 patient could expect HbA1c to drift back by approximately 1.0 to 1.1 percentage points within a year of discontinuation.

This projection aligns with real-world data from the Wegovy prescribing information, which notes that weight regain is expected after discontinuation and recommends ongoing treatment for sustained benefit.

Safety Signals That Emerged With Longer Observation

The 68-week treatment period in STEP-2 was long enough to capture the core gastrointestinal adverse event profile: nausea (33.7% with sema 2.4 mg), diarrhea (16.5%), vomiting (11.4%), and constipation (10.4%). Most GI events occurred during dose escalation and resolved by week 20. But post-trial surveillance and registry data have added three safety considerations that the original 68 weeks could not fully address.

Gallbiliary events. Across the STEP program, cholelithiasis and cholecystitis occurred at higher rates in semaglutide arms than placebo. The Wegovy label now carries a specific warning. In STEP-2, gallbladder-related events were reported in 2.6% of the sema 2.4 mg group versus 1.2% on placebo. Rapid weight loss itself increases gallstone risk, making it difficult to separate the drug effect from the weight-loss effect, but the signal is consistent enough to warrant monitoring.

Pancreatitis. STEP-2 reported no confirmed cases of acute pancreatitis. Pooled STEP data and the SELECT cardiovascular outcomes trial (Lincoff et al., N Engl J Med 2023) have not identified a statistically significant excess risk, though individual case reports continue to accumulate in post-marketing databases.

Thyroid C-cell findings. The boxed warning for medullary thyroid carcinoma (MTC) risk, derived from rodent studies with GLP-1 receptor agonists, remains on the Wegovy label. No human cases attributable to semaglutide have been confirmed in the STEP program or post-marketing surveillance through 2025. The warning remains precautionary.

Regression to the Mean and Biological Set Points

One criticism of the STEP-2 weight-regain data is that it could partly reflect regression to the mean: extreme responders at week 68 may naturally drift toward their population average regardless of drug withdrawal. This is a valid statistical concern, but the STEP-1-EXT data argue against it as the sole explanation. Regression to the mean would predict modest, symmetrical drift in both directions. Instead, the off-treatment trajectory was consistently upward (weight gain) across virtually all subgroups.

The more compelling biological explanation involves the body's energy homeostasis system. Semaglutide suppresses appetite through central GLP-1 receptor activation, reduces gastric emptying, and may alter reward-based eating behavior. When the drug is removed, these mechanisms revert. Leptin levels, which drop during weight loss, fail to send appropriate satiety signals. Ghrelin, the hunger hormone, rises. The result is increased caloric intake that is not a failure of willpower but a predictable physiological response to a reduced-fat-mass state.

This framework is critical for clinical counseling. Patients who discontinue semaglutide after achieving STEP-2-level weight loss should expect weight regain unless alternative interventions (bariatric surgery, sustained caloric restriction, or a different anti-obesity medication) are substituted.

What STEP-2 Could Not Show and What SELECT Filled In

STEP-2 was powered for weight loss, not cardiovascular events. With 1,210 participants followed for 68 weeks, the trial could not detect differences in major adverse cardiovascular events (MACE). This gap was filled by the SELECT trial (Lincoff et al., N Engl J Med 2023), which randomized 17,604 adults with established cardiovascular disease and BMI ≥27 to semaglutide 2.4 mg or placebo for a mean of 39.8 months.

SELECT demonstrated a 20% relative risk reduction in MACE (HR 0.80 to 95% CI 0.72 to 0.90). While SELECT enrolled patients with and without diabetes, prespecified subgroup analyses showed consistent benefit regardless of glycemic status. For the STEP-2 population (overweight or obese adults with T2D), SELECT provides the strongest evidence that the cardiovascular benefits of semaglutide extend beyond weight loss alone.

Durability in the Real World

Registry and claims-based studies consistently show higher discontinuation rates than clinical trials. A 2024 analysis of U.S. pharmacy claims found that approximately 50% of patients prescribed semaglutide for weight management had discontinued by 12 months, with cost and insurance coverage as the primary drivers. For patients who stop and restart ("cycling"), there are limited data on whether repeated treatment courses produce the same magnitude of weight loss.

STEP-2's controlled environment, with regular dietitian visits, exercise counseling, and free drug supply, represents a best-case scenario. The gap between trial efficacy and real-world effectiveness is not unique to semaglutide, but it is especially relevant when counseling patients about expected outcomes.

Limitations the Authors Acknowledged

Davies et al. noted several limitations in the primary publication:

  • The 68-week treatment duration, while longer than most obesity trials at the time, was insufficient to assess durability beyond the treatment period.
  • The study population was predominantly White (72.6%) and excluded patients with HbA1c >10%. Generalizability to more diverse populations and those with poorly controlled diabetes is uncertain.
  • Background anti-diabetes medications were stable but variable (metformin, SGLT2 inhibitors, sulfonylureas). Interaction effects between these drugs and semaglutide on weight loss were not formally tested.
  • The off-treatment period was only 7 weeks, too short to characterize the full rebound trajectory.
  • Participants in STEP-2 lost less weight than those in STEP-1 (9.6% vs. 14.9%), confirming that type 2 diabetes itself attenuates the weight-loss response to GLP-1 agonists. The biological reasons for this attenuation (insulin resistance, concomitant medications that promote weight gain, altered incretin signaling) remain incompletely understood.

Frequently asked questions

›
›
›
›
›
›
›
›
›
›

References

  1. Davies M, Færch L, Jeppesen OK, et al. Semaglutide 2.4 mg once a week in adults with overweight or obesity, and type 2 diabetes (STEP 2): a randomised, double-blind, double-dummy, placebo-controlled, phase 3 trial. Lancet. 2021;397(10278):971-984. PubMed

  2. Wilding JPH, Batterham RL, Davies M, et al. Weight regain and cardiometabolic effects after withdrawal of semaglutide: the STEP 1 trial extension. Diabetes Obes Metab. 2022;24(8):1553-1564. PubMed

  3. Lincoff AM, Brown-Frandsen K, Colhoun HM, et al. Semaglutide and cardiovascular outcomes in obesity without diabetes. N Engl J Med. 2023;389(24):2221-2232. PubMed

  4. Wegovy (semaglutide) prescribing information. Novo Nordisk. Revised 2023. FDA Label

  5. Rubino D, Abrahamsson N, Davies M, et al. Effect of continued weekly subcutaneous semaglutide vs placebo on weight loss maintenance in adults with overweight or obesity: the STEP 4 randomized clinical trial. JAMA. 2021;325(14):1414-1425. PubMed