Sarcopenia and Short-Term Outcomes After Esophagectomy: A Meta-analysis

2020 
This meta-analysis aimed to investigate the value of preoperative sarcopenia in predicting complications after esophagectomy. Clinicopathologic characteristics of sarcopenia patients, which may support sarcopenia management, also were studied. This study searched for articles describing an association between sarcopenia and short-term outcomes after esophagectomy using PubMed, EMBASE, and the Cochrane Library. Mantel–Haenszel and inverse variance models were used for the meta-analyses of end points. The meta-analysis included 14 studies comprising a total of 2387 patients. Sarcopenia was significantly associated with advanced age (weighted mean difference [WMD], 3.48; 95% confidence interval [CI], 2.22–4.74), lower body mass index (WMD − 2.22; 95% CI − 2.65 to − 1.79), squamous cell carcinoma (odds ratio [OR], 2.78; 95% CI 1.72–4.47), advanced clinical tumor stage (OR 1.65; 95% CI 1.28–2.15), and neoadjuvant therapy (OR 1.87; 95% CI 1.38–2.53). The sarcopenia patients showed lower preoperative albumin levels (WMD − 0.11; 95% CI − 0.19 to − 0.04) than the nonsarcopenia patients. Sarcopenia was significantly predictive of pneumonia (OR 2.58; 95% CI 1.75–3.81) and overall complications (OR 1.52; 95% CI 1.07–2.15) after esophagectomy. The sarcopenia patients also showed nonsignificant increases in the risks of anastomotic leakage (OR 1.29; 95% CI 0.99–1.67), vocal cord palsy (OR 2.03; 95% CI 0.89–4.64), and major complications (≥ Clavien-Dindo grade III; OR 1.30; 95% CI 0.95–1.79) but not increased operation time, blood loss, or mortality. Preoperative sarcopenia assessment showed considerable potential for predicting postoperative complications for esophageal cancer patients. To realize this potential, more effective diagnostic criteria and severity classifications for sarcopenia are warranted.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    28
    References
    11
    Citations
    NaN
    KQI
    []