P-0082 - Routine Use of Perioperative Chemotherapy for Resectable Gastroesophageal Cancer

Abstract

Methods: Patients diagnosed with locally advanced gastric cancer were treated with perioperative chemotherapy and surgery in our institution. Perioperative chemotherapy schedules were selected by the oncologists according to each patient’s baseline characteristics and comorbidities. We obtained available clinical, radiological, surgical and pathological data for all patients analyzed. The primary end point was the complete resection (R0) rate. Secondary end points were disease-free survival (DFS), overall survival (OS), toxicity, radiological response rate, pathological response rate and downstaging rate. We also looked for prognostic and predictive factors for DFS, overall survival (OS), pathological complete response rate and the R0 rate. The study got the approval of the local Research Ethics Committee. Kaplan-Meier method and long-rank test were apply to univariate survival analysis. Multivariate analysis was performed. Cox regression and multiple linear regression analysis was applied when required. Results: Forty patients were found eligible for this retrospective analysis. The median follow-up period was 19 months (range 5.7-54.8). At diagnosis, 52.5% of patients were classified as stage II and 47.5% were stage III. The most common chemotherapy schedule used was a combination of epirubicin, cisplatin and capecitabine. Forty percent of patients completed three preoperative cycles and three postoperative cycles. A tolerable toxicity related to chemotherapy was found without any chemotherapy toxicity–related deaths. Thirty-nine patients underwent surgery: 80% reached a complete resection (R0), down-staging was detected in 57.5% and 17.5% had a pathologically complete response. The median time of disease-free survival was 34.05 months (IC95% 25.6-42.4) and the median time of overall survival was 39.01 months (IC95% 30.8-47.1). Pathological response, radiological response, perineural invasion and resection type were related to better disease-free survival (p0.001, p < 0.001, p0.001 and p < 0.001 respectively) and better overall survival (p0.001, p0.024, p < 0.001 and p0.002 respectively). However, the statistical significance was lost in the multivariate analysis. Predictive factors for pathological complete response in the univariate analysis were: vomiting at diagnosis (p0.0005), the presence of comorbidities (p0.0016) and radiological response (p0.004). Only the presence of comorbidities maintained statistical significance after the multiple linear regression analysis (p0.024). The resection type could be predicted by the chemotherapy schedule used (p0.014) and the clinical response (p0.048); these factors maintained significance after the multiple linear regression with p values of 0.029 and 0.025, respectively. Conclusion: Our results support that perioperative chemotherapy for locally advanced gastric cancer can be safely delivered in daily clinical practice, obtaining an improvement of the pathologic response and the complete resection of gastric cancer. We found that the presence of comorbidities were independent predictive factors for the pathologic response, while the chemotherapy schedule and the clinical response could independently predict a complete resection.