Surgery in pancreatic cancer (PC): Identifying potential barriers.

Abstract

290 Background: Due to lack of effective systemic therapy, surgical resection remains the mainstay of treatment of PC and the only potential for cure. Even though mortality of PC surgery has decreased in recent years, surgical rates remain lower than for other cancer types. We performed an analysis of early-stage PC to determine factors that may prevent patients from undergoing surgery. Methods: PC cases diagnosed between 1996 and 2008 were identified from the Veteran's Affairs Central Cancer Registry. Patients with early stage disease, defined as AJCC clinical stage group 1-3, were included. Data were analyzed using biostatistical software SPSS. Chi-square analysis was performed for categorical variables comparing those who underwent surgery (SUR) and those in which surgery was not performed (SNP). Kaplan-Meier analysis was used for estimates of overall survival. Results: Of 1,306 cases identified, 97.7% were male, 78% where white. Surgery was performed in 251 (19.2%) cases. Surgery was not recommended in 739 (56.6%) and contraindicated due to patient related factors in 198 (15.2%) cases. Surgery was recommended but not performed due to unknown reasons in 59 (4.5%); patient refusal in 43 (3.3%) and “other reasons” in 16 (1.2%) cases. Mean age in SUR vs. SNP was 64.1 vs. 67.5 years. Median survival was 11.1 months in SUR vs. 5.5 months SNP (p<0.001). On univariate analysis, patients with diabetes had less surgery (10.5%) compared to other comorbidities (p<0.001). Tumors in the body of the pancreas had less surgery (4.8%) compared to the head or tail (19.9%, 23.6%, p<0.001). Sex, race, alcohol, and tobacco use did not significantly impact surgery rates (p>0.05). On multivariable analysis including race, sex, primary site, comorbidity, chemotherapy, radiation, alcohol, and tobacco use, only comorbidity was independently associated with less surgery. Conclusions: In this registry-based, retrospective analysis of early stage PC, although SUR patients had better survival than SNP, surgery was not recommended as a treatment option for most patients by their providers. Whereas comorbid conditions do seem to influence this decision, race and patient refusal was not a major factor. No significant financial relationships to disclose.