Integrating chest CT-derived skeletal muscle index with SARC-F for sarcopenia screening in patients with lung cancer.

AIMS: To evaluate and compare the diagnostic performance of thoracic computed tomography (CT)-derived skeletal muscle indexes (SMIs), used alone and combined with the SARC-F questionnaire, for identifying sarcopenia in lung cancer patients.

METHODS

In this prospective cross-sectional study, SMIs were measured at the 10th (T10SMI), 11th (T11SMI), and 12th (T12SMI) thoracic vertebral levels from routine CT imaging. Sarcopenia was defined using Asian Working Group for Sarcopenia (AWGS) criteria, incorporating appendicular skeletal muscle mass index (ASMI), handgrip strength (HGS), and the five-times sit-to-stand test (5STS).

RESULTS

Of the 208 lung cancer patients enrolled, 38 (18.3%) were diagnosed with sarcopenia. T10SMI, T11SMI, and T12SMI showed high consistency with ASMI (0.852, 0.852, and 0.857, respectively), 5STS (0.855, 0.852, and 0.857, respectively), and HGS (0.996, 0.996, and 0.995, respectively). The SARC-F score showed moderate consistency with ASMI (0.733) and HGS (0.875), and high consistency with 5STS (0.909). When used individually, T10SMI, T11SMI, T12SMI, and SARC-F demonstrated accuracy of 0.827-0.846, sensitivity of 0.184-0.395, F1 scores of 0.286-0.484, AUCs of 0.752-0.879, specificity of 0.947-0.977, PPVs of 0.563-0.636, NPVs of 0.843-0.875, Cohen's κ of 0.222-0.399, and Youden indices of 0.161-0.342. Combining thoracic SMIs with SARC-F improved diagnostic performance, yielding accuracy of 0.851-0.894, sensitivity of 0.421-0.500, F1 scores of 0.508-0.633, AUCs of 0.864-0.900, specificity of 0.947-0.982, PPVs of 0.640-0.864, NPVs of 0.880-0.898, Cohen's κ of 0.425-0.577, and Youden indices of 0.368-0.482. T12SMI consistently demonstrated the best diagnostic performance, significantly outperforming T10SMI and T11SMI in both the overall cohort (Z =  - 2.130 to - 2.870) and the male cohort (Z =  - 2.385 to - 2.412).

CONCLUSIONS

Integrating thoracic CT-derived SMIs with SARC-F screening enhances sarcopenia detection in patients with lung cancer. This combined approach capitalizes on routinely acquired imaging data and a simple patient-reported instrument, offering a clinically feasible and comprehensive strategy for early sarcopenia identification.