Presentation (CTI)

OBJECTIVES: Semaglutide and tirzepatide are widely used obesity management medications. Recent approvals of newer semaglutide formulations and doses (oral 25 mg/day and higher-dose subcutaneous 7.2 mg) have further expanded the treatment landscape. Existing network meta-analyses often combine trials conducted in adults with obesity with/without type 2 diabetes (T2D), which may confound comparative estimates because T2D is associated with smaller percentage total body weight loss (%TBWL). A subgroup-effect/meta-regression framework, as described in NICE DSU-TSD3, provides a transparent approach to adjust for effect modification and generate population-relevant estimates. We evaluated the impact of comorbid T2D on %TBWL and generated population-specific comparative estimates for semaglutide and tirzepatide formulations in adults with obesity.
METHODS: Randomized controlled trials from an existing systematic literature review (upto January 2026) were analyzed. A Bayesian network meta-regression was fitted with T2D population status (T2D-only vs with/without T2D) as an effect modifier, with random-effects to account for between-study heterogeneity. The outcome was placebo-adjusted %TBWL at endpoint. Predicted %TBWL was estimated for semaglutide (2.4 mg subcutaneous, 7.2 mg subcutaneous, 25 mg oral), and tirzepatide (10-15 mg subcutaneous) in each population.
RESULTS: Comorbid T2D was a statistically significant effect modifier (p<0.05), with lower %TBWL in T2D-only population. Placebo-subtracted %TBWL for T2D status (T2D-only vs with/without T2D) was 6.8% vs 11.3% for semaglutide 2.4 mg, 9.5% vs 14.1% for 7.2 mg, 8.4% vs 10.8% for 25 mg, and 11.6% vs 16.1% for tirzepatide 10-15 mg. Tirzepatide was the only intervention achieving >10% placebo-adjusted %TBWL in both populations. Although tirzepatide showed numerically greater %TBWL, overlapping credible intervals indicated no clear evidence of a difference vs higher-dose semaglutide (7.2 mg and 25 mg oral).
CONCLUSIONS: Comorbid T2D attenuates %TBWL and can bias indirect comparisons when trials are pooled across mixed populations. Subgroup-effect adjusted approach provides population-relevant comparative estimates for semaglutide and tirzepatide, supporting more appropriate interpretation for clinical decision-making.