Can Proximal Gastrectomy with Double-Tract Reconstruction Replace Total Gastrectomy? A Propensity Score Matching Analysis

Background This retrospective cohort study compared proximal gastrectomy (PG) with double-tract reconstruction (DTR) versus total gastrectomy (TG) with Roux-en-Y reconstruction in terms of clinical outcomes. Methods All consecutive patients with upper early gastric cancer (EGC) who underwent PG-DTR or TG in 2008-2016 were selected. TG patients who matched PG-DTR patients in age, sex, body mass index, clinical stage, and ASA score were selected by propensity score matching. Groups were compared in terms of clinicopathological characteristics, clinical outcomes, early (<= 30 days), late (> 30 days), and severe (Clavien-Dindo grade >= III) postoperative complications, 1-year reflux morbidity, recurrence, and mortality. Results Of 322 patients, 52 underwent PG-DTR. A matching TG group of 52 patients was selected. The PG-DTR group had smaller tumors (p = 0.02), smaller proximal and distal resection margins (p = 0.01, p < 0.01), and fewer retrieved lymph nodes (p < 0.01). PG-DTR associated with shorter times to diet and hospital stay (both p = 0.02). Groups did not differ in early (11.3 vs. 19.2%, p = 0.19), late (1.9 vs. 5.7%, p = 0.31), or severe complication rates (7.7 vs. 13.5%, p = 0.34). At 1 year, the groups did not differ in reflux symptoms (Visick score) or endoscopic esophagitis (Los Angeles Classification). There were no recurrences. Five-year overall survival rates were 100 and 81.6% (p = 0.02), respectively. Conclusion PG-DTR associated with better clinical outcomes and survival. Complication and reflux rates were similar. PG-DTR may be suitable for upper EGC.