Chemotherapy after surgery for gastric cancer: a case-control study

Background: Despite its continued decline over the past 30 years, gastric carcinoma (GC) remains a ommon cancer and a major health problem worldwide. Several studies have investigated the possible role of chemotherapy (CHT) after surgery for GC. In fact, biased chiefly by a low number of patients (pts) recruited, lack of uniformity in surgical and oncological treatment, inadequate homogeneity (sometimes absence) of the control group, they failed to show a clear indication. Aim: To assess the impact on survival of CHT in a subset of pts operated on for GC. Materials and Methods: Between January 1993 and December 2001, 275 out of 311 pts underwent surgery for GC (resectability rate: 88.4%). Surgical resection included a total/subtotal gastrectomy and a D2 lymphadenectomy. Because of medical, ethical or logistical reasons, only 53 pts (26.9%) out of 197 in the whole series for whom the treatment was indicated (cases without both nodal and serosal invasion were excluded) received CHT (group A). Pharmacological schedule included ELF (30 pts), PELF (17), EPI (6). Pts as close as possible to main prognostic variables but different because they did not undergo CHT were selected to compose group B. Clinicopathological characteristics and survival of these two groups were compared. finally a statistical analysis of survival was performed on the whole series (group A+ group B), by means of uni and multivariate analysis. Results: The rate of pts submitted to CHT did not increase during the period of the study and pts included in group A and group B are homogeneously distributed within this same period. The 2 groups are similar according to main clinicopathological variables, except for age (median: 60 years in group A and 72 in group B, respectively). The 5-yr survival was 39.1% (±4.5, CB 95%: 30.4-47.7%) for the whole series, with a median survival of 25.5 months (mths) for all pts, 45.2 mths and 11.7 mths for survivors and pts who died, respectively. Wall invasion, lymph node involvement and presence of residual disease turned out to be significant according to univariate analysis. Difference in 5-yr survival between group B and group A (41.2 vs 38.0%, respectively) was not significant. Multivariate analysis showed only wall invasion and lymph node involvement had independent prognostic value. Discussion: Our study, although retrospective, was performed on a typical western series of GC pts, operated on in a single institution by 2 major dedicated surgeons (C.C, P.B.) within a limited period of time. Survival analysis, nevertheless based on actuarial data, appeared really quite reliable. The use of different schedules seemed of little value, as till now no regimen was proved to have more efficacy than others. In our series we highlighted, as others did, a little but not significant survival benefit for pts receiving CHT. So our data argue against a routine use of CHT in pts operated on for GC. Chemotherapy after surgery for gastric cancer: a case-control study.