Purpose: CT-based Body-composition (BC) parameters correlate with the patients’ outcome in metastatic cancer patients treated with chemotherapies or targeted therapies. Our aim was to investigate similar associations regarding immune checkpoint inhibitor (CPI).

Methods: Patients were consecutively included as they were treated with CPI at our institution for a metastatic solid cancer with baseline CT-scan (CT0) and early evaluation CT-scan (CT1, 2 months later). At each evaluation, the areas corresponding to psoas muscles alone, skeletal muscle, subcutaneous, visceral and total adipose tissues at L3 vertebral level were extracted and weighted by height², providing PMI, SMI, SATI, VATI and TATI, respectively, and their changes (Δt-) from the first day of treatment to CT1. Correlations between continuous BC-parameters and progression free survival (PFS) were evaluated in men, women and whole population with univariate Cox regressions. After dichotomizing the BC-parameters per whole-population and sex-specific tertiles, uni- and multivariate Cox models were built to identify independent predictors of the PFS.

Results: Between December 2013 and December 2016, 117 patients were included (55 women, mean age: 62.4) and 78 showed a progression (median PFS = 125 days, 95 %CI = 87-117). Changes in BC-parameters did not depend on sex. None of the baseline BC-parameters correlated with PFS while Δt-PMI and Δt-SATI did (multivariate HR = 2.41, p = 0.0008 and HR = 2.82, p = 0.0004, respectively).

Conclusions: The occurrence of subcutaneous adipopenia and sarcopenia after beginning CPI treatment, estimated through Δt-SATI and Δt-PMI, correlated with higher risk of progression.