CT-derived body composition and clinical outcomes in ovarian cancer patients: A scoping review.

INTRODUCTION

Epithelial ovarian cancer (EOC) is associated with poor prognosis. Computed tomography (CT)-derived body composition, encompassing muscle and adipose tissue metrics, has emerged as a potential biomarker for prognosis and treatment guidance. This scoping review maps current evidence on associations between CT-derived body composition and clinical outcomes in EOC, with attention to methodological variation and research gaps.

METHODS

The review protocol was prospectively registered in the Open Science Framework on December 11, 2024: https://osf.io/4p5u6. A comprehensive literature search was conducted in Medline (via PubMed), Embase (via Ovid), and Web of Science, following JBI methodology for scoping reviews and PRISMA-ScR guidelines. Eligible studies assessed CT-derived body composition in relation to clinical outcomes among EOC patients.

RESULTS

50 studies were included, 47 retrospective. Muscle and fat quantities were measured as cross-sectional areas or volumetric estimates, while muscle quality was assessed as skeletal muscle radiodensity in Hounsfield Units (HU). Considerable heterogeneity was observed in measurement techniques, anatomical landmarks, and cut-off definitions for sarcopenia and adiposity. Outcomes investigated included survival, chemotherapy-related, and surgical endpoints. Muscle quality, joint fat-muscle measures, and treatment-related muscle loss were more consistently predictive of survival than muscle mass or adipose tissue alone. Associations with chemotherapy-related and surgical outcomes were less consistent.

CONCLUSION

CT-derived body composition may have prognostic relevance in EOC; however, associations remain inconsistent and is largely based on retrospective studies. Findings should therefore be considered hypothesis-generating. Methodological inconsistencies limit clinical applicability and highlight the need for standardized measurement techniques, consensus cut-offs, and prospective validation.