Negative prognostic impact of sarcopenia before and after neoadjuvant chemotherapy for pancreatic cancer.

OBJECTIVES: To elucidate the prognostic impact of sarcopenia before and after neoadjuvant chemotherapy (NAC) for pancreatic cancer (PC). METHODS: We retrospectively studied 75 consecutive PC patients who underwent neoadjuvant gemcitabine plus S-1 combination therapy followed by pancreatectomy between 2008 and 2016. According to the skeletal muscle volume index (SMI), the patients were divided into the muscle attenuation group (MAG) and normal group (NG) before or after NAC. Prognostic factors for overall survival (OS) were analyzed by Cox proportional hazards models. RESULTS: The MAG showed significantly poorer OS than the NG before and after NAC. Pre-NAC, median OS was 20.0 months in the MAG versus 49.0 months in the NG (p = 0.006). Post-NAC, median OS was 21.3 months in the MAG versus 48.8 months in the NG (p = 0.014). Multivariate analysis, excluding muscle attenuation after NAC because of confounding factors and lower hazard ratio (2.08, 95% confidence interval: 1.14-3.78, p = 0.016) than that before NAC (2.14, 1.23-3.70, p = 0.007) by univariate analysis, revealed the following independent prognostic factors: muscle attenuation pre-NAC (2.25, 1.26-4.05, p = 0.007); borderline resectability (1.96, 1.04-3.69, p = 0.038); operative blood loss (2.60, 1.38-4.88, p = 0.003); and distant metastasis (3.31, 1.40-7.82, p = 0.006). CONCLUSIONS: Sarcopenia before and after NAC for PC is suggested to be a poor prognostic factor, with a stronger impact before than after NAC.