Long-term outcomes of proximal gastrectomy versus total gastrectomy for clinically advanced upper-third gastric cancer: an inverse probability weighting analysis

BACKGROUND: Proximal gastrectomy is commonly performed for early upper-third gastric cancers; however, in no studies have the long-term outcomes of total and proximal gastrectomy been compared for clinically advanced upper-third gastric cancer. We assessed the outcomes of both techniques in this context. METHODS: Patients who underwent proximal or total gastrectomy for clinically advanced upper-third gastric cancer (tumor diameter ≤ 50 mm) between June 1994 and July 2021 at Toranomon Hospital, Japan, were included. We compared overall and relapse-free survival between the total gastrectomy (TG) and proximal gastrectomy (PG) groups using inverse probability of treatment weighting, analyzed whether the surgical technique was an independent risk factor for death or recurrence using a Cox proportional hazards model, and evaluated the therapeutic effect using the therapeutic value index for the lymph nodes dissected in TG but undissected in PG. RESULTS: We investigated 45 and 35 patients in the TG and PG groups, respectively. The 5-year overall and relapse-free survival of the TG and PG groups were 59.4% vs. 61.2%, P = 0.921; and 56.5% vs. 57.1%, P = 0.984; respectively. Surgical technique was not an independent risk factor for death or recurrence. The therapeutic value indexes of the relevant lymph nodes (#3b/4d/5/6/12a) were all zero. CONCLUSIONS: Proximal gastrectomy for relatively small clinically advanced upper-third gastric cancer may be acceptable because its 5-year overall and relapse-free survival do not differ significantly from that of total gastrectomy, it is not an independent risk factor for death or recurrence, and no therapeutic effect for the omitted lymph nodes was revealed.