A systematic review and meta-analysis of mediastinoscopy-assisted transhiatal esophagectomy (MATHE).

BACKGROUND:Transhiatal esophagectomy (THE) avoids thoracotomy but sacrifices mediastinal lymphadenectomy. Mediastinoscopy-assisted transhiatal esophagectomy (MATHE) allows for visualisation and en-bloc dissection of mediastinal lymph nodes while retaining the benefits of THE. However, given its novel inception, there is a paucity of literature. This study aimed to conduct the first meta-analysis to explore the efficacy of MATHE and clarify its role in the future of esophagectomy. METHODS:Four databases (PubMed, EMBASE, Scopus, and Cochrane Library) were searched from inception to May 1, 2023. Studies were included if they reported outcomes for patients with esophageal cancer who underwent MATHE. Meta-analyses of proportions and pooled means were performed for the outcomes of intraoperative blood loss, lymph node (LN) harvest, mean hospital length of stay (LOS), mean operative time, R0 resection, conversion rates, 30-day mortality rate, 5-year OS, and surgical complications (anastomotic leak, cardiovascular [CVS] and pulmonary complications, chyle leak and recurrent laryngeal nerve palsy [RLN]). Sensitivity analyses were performed for outcomes with substantial statistical heterogeneity. RESULTS:The search yielded 223 articles; 28 studies and 1128 patients were included in our analysis. Meta-analyses of proportions yielded proportion rates: 30-day mortality (0 %, 95 %CI 0-0), 5-year OS (60.5 %, 95 %CI 47.6-72.7), R0 resection (100 %, 95 %CI 99.3-100), conversion rate (0.1 %, 95 %CI 0-1.2). Among surgical complications, RLN palsy (14.6 %, 95 %CI 9.5-20.4) were most observed, followed by pulmonary complications (11.3 %, 95 %CI 7-16.2), anastomotic leak (9.7 %, 95 %CI 6.8-12.8), CVS complications (2.3 %, 95 %CI 0.9-4.1) and chyle leak (0.02 %, 95 %CI 0-0.8). Meta-analysis of pooled means yielded means: LN harvest (18.6, 95 %CI 14.3-22.9), intraoperative blood loss (247.1 ml, 95 %CI 173.6-320.6), hospital LOS (18.1 days, 95 %CI 14.4-21.8), and operative time (301.5 min, 95 %CI 238.4-364.6). There was moderate-to-high statistical heterogeneity. Findings were robust to sensitivity analyses. CONCLUSION:MATHE is associated with encouraging post-operative mortality and complication rates, while allowing for radical mediastinal lymphadenectomy with reasonable lymph node harvest.