Measurement of psoas major muscle area
The psoas major muscle area was calculated at the L3 position on computed tomography. The PMMA was calculated before chemotherapy and before pembrolizumab infusion. We were able to measure ∆PMMA in only 112 patients of the whole cohort, since the rest were staged only with a CT scan of the thorax and upper abdomen and the area of the patient’s psoas major muscle at the L3 position was not available. The PPMA was measured in the region of interest by tracing an outline using the image viewer software “DICOM”. The following formula was used: % change of PMMA= ([1-PMMA before P/PMMA before CT]*100. Patients with a change in PMMA≥10% were considered to have sarcopenia [16].

Endpoints

The tumor response was assessed according to the Response Evaluation Criteria in Solid Tumors (ver. 1.1) (RECIST 1.1), and clinical tumor response was assessed every 3 months or at clinical deterioration. Hyperprogression was defined if at least three of the following existed: 1. time to treatment failure<3 months; 2. increase ≥50% in the sum of target lesion major diameters between baseline and first radiological evaluation; 3. appearance of at least two new lesions in an organ already involved between baseline and first radiological evaluation; 4. spread of the disease to a new organ between baseline and first radiological evaluation; and 5. clinical deterioration with a decrease in ECOG PS ≥2 during the first three months of treatment. Pseudoprogression was defined as initial progression followed either by partial response or stable disease lasting at least 6 months. OS was defined as the interval between diagnosis of the disease and death or the date of the last follow-up evaluation.

Statistical design and analysis

Data were managed and analyzed using SPSS software ver. 23. The demographic characteristics were expressed as frequencies and percentages for categorical variables and as medians and means with standard deviations for quantitative variables. The Mann–Whitney U test, Spearman correlation and χ2 test were used to compare and evaluate the correlations between the biomarkers and the clinicopathological characteristics of the patients, such as age, sex, the NLR, and the PLR. To assess the correlations between test results, rho values were interpreted as follows: <0.39, weak correlation; 0.40-0.59, moderate correlation; 0.60-0.79, strong correlation; and ≥0.80, very strong correlation. The Kruskal-Wallis one-way analysis of variance with was used to compare the levels of hematological biomarkers, ΔPMMA and response to pembrolizumab at the first CT scan. The Wilcoxon and McNemar tests were used to compare quantitative and categorical biomarker values and their derivations. The diagnostic accuracy of biomarkers was determined by obtaining the largest possible area under the curve (AUC) in receiver operating characteristic curve (ROC) analysis. AUC values ≥0.9 were considered “excellent”, ≥0.80 were considered “good”, ≥0.7 were considered “fair” and <0.70 were considered “poor”. Survival curves according to the response on the first CT scan were estimated using the Kaplan-Meier method, and differences were assessed using the log-rank test. We also performed multinomial logistic regression to estimate the effects of hematological biomarkers and ∆PMMA on the response to treatment. Two-tailed p-values <0.05 were considered significant.