Association between sarcopenia and survival in patients with gynecologic cancer: A systematic review and meta-analysis
Wen‐Li Lin, Thi‐Hoang‐Yen Nguyen, Cheng-Yao Lin, Li‐Min Wu, Wen‐Tsung Huang, How‐Ran Guo
Abstract
Background Despite prior attempts to evaluate the effects of sarcopenia on survival among patients with gynecologic cancer, the results of these studies have not been consistent. The present study evaluated the association between sarcopenia and survival among patients with gynecologic cancer by aggregating multiple studies. Methods We performed a literature search using computerized databases and identified additional studies included in the bibliographies of retrieved articles. The quality of each study was evaluated using the Newcastle–Ottawa Scale, and meta-analyses were performed to evaluate overall survival (OS) and progression-free survival (PFS). We constructed a forest plot for each outcome and assessed publication bias using Begg’s test. Heterogeneity was assessed using I 2 statistics. Results From the 5,933 initially identified articles, 16 studies describing 2,031 participants with a mean age of 60.34 years were included in the meta-analysis. We found that compared with patients with gynecologic cancer but without sarcopenia, patients with sarcopenia had worse OS, with a pooled hazard ratio (HR) of 2.61 (95% confidence interval [CI]:1.52–4.46), and worse PFS (HR: 1.37, 95% CI: 1.09–1.73). The quality of studies was generally good, and no publication bias was detected among studies for either OS or PFS. Although 4 of 12 studies were of fair quality, we conducted a sensitivity analysis excluding studies or fair quality and obtained similar results. Conclusions These meta-analysis results suggest that sarcopenia is associated with worse OS and PFS among patients with gynecologic cancer. The use of different case definitions appeared to be a major source of heterogeneity among the studies. Further studies remain necessary to confirm our findings, especially those examining OS and PFS, because publication bias was identified.