LAUSR

e20046 Background: Sarcopenia is a loss of skeletal muscle mass that has been studied as prognostic factor in several cancers. Retrospective studies have suggested that sarcopenia is associated with poorer survival outcomes and with an increase of major chemotherapy toxicities resulting in dose reduction and delay. This study examined the value of sarcopenia in patients with stage III non-small cell lung cancer (NSCLC). Methods: This retrospective analysis includes 68 patients affected by stage III NSCLC treated with induction chemotherapy followed by surgery or radical radiation therapy in our cancer center. Weight and height were obtained from medical records at diagnosis. Skeletal muscle index (SMI) was measured by the analysis of electronically stored computed tomography images obtained before the start of chemotherapy; sarcopenia was defined by international consensus as a SMI≤39 cm2/m2 for women and ≤55 cm2/m2 for men. Kaplan-Meier method and Log-Rank test were used to determine the impact of sarcopenia on overall survival (OS) and progression-free survival (PFS). Exact Fisher test and Chi-squared test were used to establish the association between the presence of sarcopenia and other variables. Results: A total of 68 patients (stage 3A = 39; stage 3B = 29) with performance status 0-1 and median age 67 yrs were analyzed. Forty-five patients (66%) were sarcopenic: 100% of underweight patients (BMI ≤18.5), 83% of patients with normal weight (BMI 18.5-24.9), 56% of overweight patients (BMI 25-29.9) and 30% of obese (BMI≥30). Sarcopenia was not associated with age≥70 yrs (p = 0.67), Charlson Comorbidity Index (p = 1.00), stage (p = 0.53), response rate to chemotherapy (p = 0.78) or toxicities of grade≥3 (p = 0.83). Median OS in sarcopenic patients was 18.2 months compared with 33.2 months in nonsarcopenic patients (p = 0.03); the difference in terms of PFS was not statistically significant (10.7 vs 14.9 months; p = 0.19). Conclusions: Sarcopenia is associated with shorter OS in patients with locally advanced NSCLC but it seems not related with worse response to induction chemotherapy or higher toxicities. These data should be validated in larger prospective clinical studies.