Effect of long-limb Roux-en-Y reconstruction on early glycemic control compared with Billroth II after subtotal gastrectomy in patients with type 2 diabetes and gastric cancer: 6-Month interim analysis of the STARDOM trial.

초록

Background: Gastrectomy can induce improvement in type 2 diabetes (T2DM), yet the metabolic impact may vary according to reconstruction method. Whether reconstruction strategy influences diabetes outcomes in gastric cancer remains unclear. The STARDOM trial was designed to determine the differential metabolic effects of Billroth II (BII), conventional Roux-en-Y (CRY), and long-limb Roux-en-Y (LRY) reconstruction after subtotal gastrectomy. Methods: STARDOM is a prospective, multicenter, phase III randomized controlled trial enrolling patients with stage I–II distal gastric adenocarcinoma, T2DM, and BMI $23 kg/m². Patients were randomized 1:1:1 to BII, CRY, or LRY reconstruction. The primary endpoint is biochemical improvement of diabetes at 12 months, defined as HbA1c ,6.5% without antidiabetic medication. We report prespecified 6-month interim outcomes. Results: Sixty patients were randomized (BII n=20, CRY n=19, LRY n=21) with balanced baseline characteristics. At 6 months, HbA1c ,6.5% was achieved in 28.6% of BII, 53.8% of CRY, and 70.6% of LRY patients. LRY demonstrated a significantly higher response rate compared with BII (p=0.032), representing more than a twofold increase in diabetes control. Mean HbA1c at 6 months was lowest in LRY (6.17%) compared with CRY (6.55%) and BII (6.50%). Weight loss was comparable across groups, suggesting a weight-independent metabolic effect. Surgical morbidity did not differ significantly. Conclusions: Long-limb Roux-en-Y reconstruction produced superior early glycemic control compared with Billroth II after gastrectomy for gastric cancer, without increased morbidity. These findings support reconstruction-dependent metabolic benefit and may influence surgical strategy in patients with gastric cancer and T2DM. (ClinicalTrial.gov ID: NCT04284943.) Clinical trial information: NCT04284943. Research Sponsor: None.