Correlation of surgical volume and outcomes with gastrectomy.

Abstract

272 Background: Gastric cancer(GC) is associated with significant mortality worldwide. Radical gastrectomy with lymphadenectomy is considered the only curative option for GC. Traditionally, these operations are associated with significant morbidity. However, a significant proportion of gastric resections are performed at low volume centers. We sought to compare oncologic outcomes of gastrectomy by surgical volume. Methods: Utilizing the National Cancer Database we identified patients with gastric cancer who underwent gastrectomy. We then stratified based upon volume of institution. Low volume(LV) < 10 gastric resections per year, and high volume(HV) > 10 gastric resections per year. Mann-Whitney U and Kruskal were used to compare continuous variables and Pearson’s Chi-square test was used to compare categorical variables. Unadjusted survival analyses were performed using the Kaplan-Meier method. Multivariate analysis (MVA) was performed to identify predictors of survival. All statistical tests were two-sided and p < 0.05 was considered significant. Results: We identified 29,216 patients who underwent gastrectomy with a median age of 66.5 (18-90) years. There were 21,166 (72%) gastric resections performed at LV centers and 8,050 at HV centers, p < 0.001. Neoadjuvant therapy was most often utilized in the HV centers at 41.7% vs 24.1% in the LV centers, p < 0.001. The median number of lymph nodes removed were 19 (13-28) in the HV and 14 (8-21) in the LV groups respectively, p < 0.001. The R0 resection rates were 91.1% in high volume centers and 84.6% in the low volume centers p < 0.001. The length of hospitalization did not differ between groups (median 8 (6-12) days), p = 0.33. The 30-day readmission rates however, were higher in the low volume centers at 6.5% compared to 5.8% in the high volume centers. The 30 and 90-day mortality was 1.7% and 4.1% in the HV and 4.9% and 9.2% in the LV groups, p < 0.001 and p < 0.001 respectively. The median and overall 5-year survival was 83.8mo and 56% in the HV and 49.4mo and 46% in the LV cohorts, p = 0.001. Multivariate analysis revealed that age, sex, Charlson/Deyo score, tumor location, histology, grade, stage, facility volume, and neoadjuvant therapy were all predictors of survival. Conclusions: While the majority of gastric resections are performed at LV centers, oncologic outcomes, survival and mortality rates are worse compared to patients who undergo gastric resection at high volume centers. Gastric resections should be regionalized to high volume centers to ensure the best surgical and oncologic outcomes.