Predictive impact of sarcopenia in solid cancers treated with immune checkpoint inhibitors: a meta-analysis

Sarcopenia, which is characterized by a decrease in muscle quantity or quality, is commonly observed in patients with cancer. Recent research has reported contradictory results on the association between sarcopenia and the efficacy of immune checkpoint inhibitors (ICIs). We conducted a systematic review and meta-analysis to investigate this discrepancy. We systematically searched three electronic databases to identify articles reporting on the association between sarcopenia and treatment outcomes in patients with solid cancers who received ICIs. The outcomes assessed were hazard ratios (HRs) for overall survival (OS) and progression-free survival (PFS), and odds ratios (ORs) for objective response rate (ORR), disease control rate (DCR), and toxicity. Pooled estimates and their 95% confidence intervals (CIs) were calculated. A total of 2501 patients from 26 studies were analysed. Sarcopenia was observed in 44.7% (95% CI: 38.2-51.3) of the patients and was significantly associated with poor survival (HR = 1.55, 95% CI = 1.32-1.82 for OS and HR = 1.61, 95% CI = 1.35 to 1.93 for PFS). The HRs (95% CIs) for OS according to the diagnostic measures used were 1.97 (0.88-4.41) for psoas muscle index (PMI), 1.41 (0.87-2.28) for skeletal muscle density (SMD), and 1.43 (1.23-1.67) for skeletal mass index (SMI). The HRs (95% CIs) for PFS were 1.86 (1.08-3.21) for PMI, 1.27 (0.94-1.71) for SMD, and 1.38 (1.11-1.71) for SMI. Poor radiological response to ICI therapy was observed in patients with sarcopenia (OR = 0.52, 95% CI = 0.34-0.80 for ORR and OR = 0.45, 95% CI = 0.30-0.67 for DCR). The ORs for ORR (95% CIs) were 0.56 (0.15-2.05) for PMI and 0.78 (0.56-1.09) for SMI. The oncologic outcomes associated with melanoma and non-small cell lung cancer (NSCLC) were comparable with those observed overall (HR for OS = 2.02, 95% CI = 1.26-3.24 for melanoma and HR for OS = 1.61, 95% CI = 1.19-2.18 for NSCLC). In contrast, the occurrence of severe toxicity was not associated with sarcopenia (OR = 1.13, 95% CI = 0.51-2.52). Poor survival and poor response in patients with sarcopenia indicate a negative association between sarcopenia and efficacy of ICIs. Sarcopenia's predictive ability is consistent across various tumour types. For the selection of patients who may respond to ICIs pre-therapeutically, the presence of sarcopenia should be assessed in clinical practice.