Presentation (CTI)

OBJECTIVES: In patients with metastatic non-small cell lung cancer (mNSCLC), sugemalimab was evaluated in the GEMSTONE-302 trial conducted exclusively in an Asian population, whereas the comparator trials enrolled predominantly non-Asian patients. Differences in ethnic composition across trials may therefore affect the validity of unadjusted indirect comparisons. Available evidence suggests that ethnicity acts primarily as a prognostic factor rather than a treatment-effect modifier, although this has not been conclusively established. An ethnicity-adjusted ITC was therefore conducted to address this potential bias. The objective was to evaluate the relative efficacy of sugemalimab plus chemotherapy versus pembrolizumab regimens, nivolumab in combination with ipilimumab and chemotherapy, and atezolizumab monotherapy in terms of overall survival (OS) in patients with mNSCLC, using an ethnicity-adjusted indirect comparison.
METHODS: A targeted literature search (PubMed and Google) was conducted in April 2026 to identify studies reporting OS hazard ratios (HR) by ethnicity subgroup (Asian vs. non-Asian) for PD-1/PD-L1 inhibitors in advanced NSCLC. Where Asian versus non-Asian ratio of hazard ratios (RHR) were not directly reported, it was derived from subgroup-specific HRs, assuming independence between subgroups. Across identified studies, pooled RHR estimates were consistent, ranging from approximately 0.90 to 1.01, with none reaching statistical significance, suggesting no conclusive evidence of ethnicity as a treatment-effect modifier. The prioritized source (Peng 2020a; 1st-line NSCLC) yielded a pooled RHR of 0.90 (95% CI: 0.68-1.19), which was used to derive an adjusted HR for sugemalimab on the logarithmic scale, weighted by the difference in Asian patient proportions between studies, prior to indirect comparison.
RESULTS: All ethnicity-adjusted indirect comparisons of OS were statistically non-significant across histological subtypes and PD-L1 expression levels.
CONCLUSIONS: Sugemalimab plus chemotherapy showed no statistically significant OS differences versus other 1L PD-1/PD-L1-based regimens in mNSCLC without oncogenic drivers, supporting its comparability within this therapeutic class.