Gastrectomy circumstances that influences early postoperative outcome

Background/Aims: Despite decreasing mortality, gastric resection is still a procedure of significant morbidity. Methodology: Factors predicting post-gastrectomy outcome over 3 years in a tertiary care cancer center, single-surgeon experience were analyzed. Results: Thirty-four patients who underwent total or partial gastrectomy at the City of Hope Cancer Center between 11/1996 and 11/1999 were analyzed. There were 21 males and 13 females, with a median age of 61 years (range: 36-97). Diagnoses included gastric malignancy (n = 28), hemorrhage from diffuse gastritis (n = 4), gastric necrosis with perforation (n = 1), and an aortogastric fistula (n = 1). The operative intent was curative in 22, and palliative in 6 cancer patients. Procedures included total (n = 14), subtotal (n = 9), distal (n = 8), and proximal gastrectomy (n = 3). Reconstruction techniques were Roux-Y (n = 25), BII (n = 5), primary esophagogastric anastomosis (n-3), and primary gastric closure (n = 1). Twenty patients had prior abdominal operations (59%); 10 underwent resection of additional organs (29%), including 2 splenectomies. The median lymph node count was 24, and 20 cancer patients had a RO resection (71%). Postoperative complications occured in 14 patients (41%; major: 26%), with 3 in-hospital deaths and one 90-day fatality (90-day mortality: 12%). Predictors of complications were benign diagnosis (p = 0.01), emergency procedure (p = 0.01), and splenectomy (prior or concurrent) (p = 0.02). Cancer diagnosis (vs. benign) and nonemergent gastrectomy (vs. emergency) were each associated with lower mortality (4 vs. 50%, p = 0.01), median length of stay (12 vs. 19 d, p = 0.02), and tube feed duration (7 vs. 194 d, p = 0.04). Gastrectomies for cancer with curative intent (vs. palliative or therapeutic) had no mortality (p = 0.004), a major complication rate of 14% (p = 0.02), and a median stay of 12 days (p = n.s). For patients with gastric cancer, pathologic stage was the only multivariate predictor of survival (p = 0.04) at a median follow-up of 9 months (15 for survivors); a median survival for patients with potentially curable disease (stage IA-IIIB) has not yet been reached. Conclusions: Gastrectomies for cancer, especially when done electively with curative intent, can lead to excellent postoperative recovery. Palliative gastrectomies or emergency procedures for "benign" conditions have significantly more complicated outcomes.