Early physiological phenotypes and survival outcomes in advanced non-small cell lung cancer: a longitudinal study of dynamic muscle loss and functional decline.

OBJECTIVE: To characterize early changes in skeletal muscle and physical function among patients with advanced non-small cell lung cancer (NSCLC) receiving penpulimab monotherapy or combination therapy, and to identify distinct treatment-associated physiological phenotypes with potential clinical relevance. METHODS: We retrospectively analyzed 381 patients with advanced NSCLC (monotherapy, n = 153; combination therapy, n = 228) treated between September 2021 and March 2023. Inverse probability of treatment weighting (IPTW) was applied to balance baseline characteristics between treatment groups. Skeletal muscle index (SMI), muscle attenuation, and functional measures were assessed at baseline and 12 weeks. Unsupervised clustering was performed to identify early physiological phenotypes based on longitudinal changes in muscle and functional parameters. Exploratory survival analyses were conducted using 12-week landmark models to evaluate the association between physiological phenotypes and overall survival (OS). RESULTS: After IPTW adjustment, combination therapy was associated with greater early physiological deterioration, including larger declines in muscle mass and physical performance, increased incidence of new-onset sarcopenia, and poorer treatment tolerability compared with monotherapy. Unsupervised clustering identified three distinct physiological phenotypes: Stable (P1), Wasting (P2), and Collapse (P3). Combination therapy was associated with a higher likelihood of transition to vulnerable phenotypes, particularly P2 and P3. These phenotypes demonstrated distinct clinical characteristics and survival patterns, with the Collapse phenotype exhibiting the poorest OS (median OS, 18.0 months), whereas OS was not reached in the Stable phenotype. Incorporation of physiological phenotypes improved the discrimination of landmark survival models beyond baseline clinical characteristics and radiographic response assessments. Exploratory pathway analyses suggested potential associations between early physiological deterioration patterns and survival differences; however, these findings require prospective validation. CONCLUSIONS: Early treatment-associated physiological deterioration in advanced NSCLC is heterogeneous and can be characterized by reproducible dynamic phenotypes. Combination therapy is associated with greater early declines in muscle and physical function, while longitudinal physiological phenotyping provides complementary information for identifying patients at increased risk of adverse outcomes beyond conventional tumor response assessment. Prospective studies are warranted to validate these findings and determine whether interventions targeting physiological decline can improve clinical outcomes.