Factors affecting survival in laparoscopic vs. open surgery among stage III colon cancer patients.

Abstract

e15095 Background: The role of laparoscopic surgery (LS) in the cure of advanced non-metastatic colon cancer is controversial. The study aims to define long-term survival of LS and open surgery (OS) of stage III colon cancer as well as to examine factors that influence the choice in surgical approaches. Methods: We abstracted 22,821 patients from the National Cancer Database who were diagnosed with stage III colon cancer from 2010 to 2012 who had undergone surgery of their primary site by either LS or OS; we only included patients who had also received adjuvant chemotherapy. The probability of undergoing a LS or OS was estimated using multivariable marginal logistic regression model, whereas the between-procedure survival difference was estimated with the Kaplan-Meier method with associated log-rank test and a multivariable marginal Cox regression model. Results: Holding the patient- and facility-level characteristics constant, significantly greater odds of receiving LS was associated with private insurance, patients living in areas with higher socio-economic status, patients receiving care at academic facilities, and patients living in the northeast United States. Kaplan-Meier results indicated that patients undergoing LS had significantly longer survival compared to patients undergoing OS (χ21 = 111.6, p < .001). Table 1 shows three-year survival estimates. After adjusting for the patient- and facility-level covariates, patients who underwent LS had 26.7% lower risk of death compared to those undergoing OS (95% CI: 21.7% to 31.4%, p < .001). In general, lower adjusted risk of death was observed in patients who were younger, female, as well as patients with fewer comorbid conditions and those with private insurance (all p< .05). Conclusions: In this largest, most recent analysis of surgical treatment of stage III colon cancer, our data suggest that long-term survival after LS is superior to conventional OS. Identification of associated demographic factors may prove useful for future allocation of treatment resources. [Table: see text]