MULTIVISCERAL RESECTION COMPARED TO STANDARD GASTRECTOMY FOR GASTRIC ADENOCARCINOMA: A PROPENSITY SCORE-MATCHING ANALYSIS
André R. Dias, Marina A. Pereira, Marcus F. Ramos, Ulysses Ribeiro*, Bruno Zilberstein, Sergio Carlos Nahas
Universidade de Sao Paulo Instituto do Cancer do Estado de Sao Paulo, Sao Paulo, São Paulo, Brazil
Introduction: Gastric cancer (GC) with invasion of adjacent organs (T4b) should be managed with en bloc gastrectomy combined with the invaded organs (multivisceral resection - MVR), when free margins can be obtained. MVR in GC is associated with high morbidity and mortality and, apparently, worse results compared to standard gastrectomy (SG). However, studies contain bias and many confounding variables, especially related to patient`s selection. Thus, the real role of MVR on GC remains controversial. On the other hand, the Propensity Score Matching (PSM) allows balancing 2 distinct groups, whose initial clinicopathological characteristics may different.
Objective: This study aimed to compare the outcomes of MVR and SG using PSM.
Methods: We reviewed all gastric adenocarcinoma patients who underwent curative gastrectomy. PSM analysis including 9 variables (sex, age, comorbidities, ASA, hemoglobin, preoperative chemotherapy, type of gastrectomy, pT and pN) was conducted to reduce patient selection bias. A 1:1 case-control match was used. Main outcome was 90-day mortality.
Results: A total of 685 GC patients were eligible for inclusion (621 SG and 64 MVR). In MVR group, R0 rate was 95.3%, mean post-operative stay was 12.5 days, and 53.1% were pT4b. After PSM, 57 patients were matched in each group. All variables assigned in the score were well matched. Statistically different outcomes between groups became equivalent: major postoperative complications (pre-PSM: 14.5 vs 26.6%, p 0.011 / post-PSM: 21.1 vs 26.3%, p=0.509); 90-day mortality (pre-PSM: 6.6 vs 10.5%, p=0.020 / post-PSM: 10.5 vs 17.5%, p=0.281). Overall survival was also similar between both groups after PSM. Age >65 years old (OR: 3.39) and resection of 2 or more organs, beside the stomach (OR: 4.62), were factors associated with 90-day mortality. R1 resection and surgery alone (no chemotherapy) were independent prognostic factors for poor OS (HR: 3.61 and 2.81, respectively).
Conclusions: After PSM, the difference in morbidity, mortality and survival of MVR compared to SG was no longer statistically significant, revealing that MVR is an acceptable therapeutic strategy to these patients with advanced tumors. Old age and resection of 2 or more organs, beside the stomach, were the main factors associated with 90-day mortality in resectable GC.
Back to 2022 Abstracts