RESUMO
PURPOSE: The aim is to clarify the use of perioperative chemotherapy in resectable goblet cell carcinoma (GCC). METHODS: A retrospective study was carried out based on the Surveillance, Epidemiology, and End Results study. The population was divided: into patients who received only radical surgery (group A) and those who received radical surgery plus chemotherapy (group B). An entropy balancing was carried out to correct the imbalance between the two groups. Two models were generated. Model 1 contained only high-risk patients: group B and a "virtual" group A with similar characteristics. Model 2 included only low-risk patients: group A and "virtual" group B with identical attributes. The efficacy of entropy balancing was evaluated with the d value. The overall survival was compared and reported with Hazard Ratio (HR) within a confidence interval of 95% (95 CI). RESULTS: The groups A and B were imbalanced for tumor size (d = 0.392), T (d = 1.128), N (d = 1.340), M (d = 1.456), mean number of positive lymph nodes (d = 0.907), and LNR (d = 0.889). Before the balancing, the risk of death was higher in group B than in A (4.3; 2.5 to 7.4). After reweighting, all large differences were eliminated (d < 0.200). In high-risk patients, the risk of death was higher in patients who underwent surgery alone than those who received perioperative chemotherapy (HR 0.5; 0.2 to 1.3) without statistical significance (p = 0.187). In low-risk patients, the risk of death was similar (HR 1.1; 0.3 to 3.3). CONCLUSION: Perioperative chemotherapy could provide some marginal advantages to high-risk patients.
