Impact of Upper Mediastinal Anatomy and Anastomotic Position on Anastomotic Leakage After Transcervical-Hiatal Subtotal Esophagectomy With Posterior Mediastinal Reconstruction.

INTRODUCTION: Anastomotic leakage (AL) after esophagectomy increases morbidity and mortality. Although mediastinal anatomy has been implicated in AL, no prior study has evaluated this issue in transcervical-hiatal subtotal esophagectomy (closed mediastinum). We investigated whether the scaled upper mediastinal anteroposterior diameter (sAPD) and anastomotic height (AH) influence AL. MATERIALS AND METHODS: We retrospectively studied 224 patients who underwent transcervical-hiatal subtotal esophagectomy with mediastinoscopy-assisted mediastinal lymphadenectomy and posterior mediastinal reconstruction using a subtotal gastric conduit; the pleura was preserved (closed mediastinum). sAPD was measured on preoperative CT and AH on the first postoperative CT obtained 3-6 months postoperatively. Associations with AL were assessed by logistic regression, and receiver operating characteristic (ROC)-derived cutoffs defined four sAPD × AH risk groups. Two comparison cohorts (transthoracic, n = 91; narrow gastric conduit, n = 20) were also analyzed. RESULTS: AL occurred in 31/224 (13.8%). Narrower sAPD and higher AH were independently associated with AL (odds ratio [OR] 0.86, p = 0.002; OR 1.15, p = 0.049); risk was highest in the narrow-sAPD/high-AH group (35.9%; OR 12.32). Airway-gastric fistula occurred in five patients (2.2%), all with a narrow sAPD. AH alone was associated with AL in the transthoracic cohort; AL was 5.0% in the narrow gastric conduit cohort. CONCLUSIONS: sAPD and AH are practical, imaging-based predictors of AL after transcervical-hiatal subtotal esophagectomy. The association between narrow sAPD and airway-gastric fistula supports conduit-space mismatch as a mechanism. Combined assessment stratifies risk and may guide conduit design and anastomotic level.