Real-world impact of initial treatment modality on survival outcomes in patients with stage I-III pancreatic ductal adenocarcinoma.

Abstract

684 Background: Management of non-metastatic pancreas cancer has been extensively studied in clinical trials over the past two decades, with a growing body of non-randomized data supporting the use of neoadjuvant chemotherapy in combination with surgery. Guidelines now recommend use of this treatment paradigm, but real-world follow-up is critical in understanding how these recommendations are translated into the clinic. Using real-world data, we aimed to evaluate the treatment modalities and subsequent survival outcomes for patients diagnosed with pancreatic adenocarcinoma within the MHealth/Fairview medical system. Methods: We identified 844 patients diagnosed with stage I-III pancreatic adenocarcinoma within the MHealth/Fairview system (Minneapolis, MN) from 2001-2021 with EHR data enriched by cancer registry abstracted data. Diagnoses were confirmed with ICD-10 coding and clinical staging. Patients were stratified by initial treatment with either surgery (N = 345) or neoadjuvant-intent chemotherapy (n = 499). Correlation between initial treatment modality, demographic identifiers, and time to treatment was calculated using both Pearson and Spearman method. Effect of initial treatment modality on overall survival was calculated using a Cox proportional Hazard (CPH) model controlling for age, sex, clinical stage, and time to treatment initiation (TTI). A sub-analysis with race was performed on samples where these data were available. Results: Relative to neoadjuvant chemotherapy, up-front surgery was associated with a significantly improved overall survival (Hazard Ratio (HR) 0.55 95% Confidence Interval (CI): 0.37 – 0.76; p < 0.001), when controlling for age, sex, clinical stage, and TTI. Surgical patients also had shorter mean TTI values (21 days; p < 0.001) relative to chemotherapy (28 days), and a slightly younger patient population (mean age 63, vs 66 for chemotherapy, p = < 0.001; rho = -0.14). We observed no significant correlation between age (p = 0.027) or TTI (p = 0.12) and survival within this population. We noted no significant correlation between race, sex, or clinical stage and initial treatment modality or overall survival. Conclusions: We observed that MHealth/Fairview system patients who received upfront surgery for stage I-III pancreas cancers demonstrated improved overall survival relative to other treatment modalities, even after controlling for the differences in age and TTI found between these patients and those who undergo initial chemotherapy. This is likely due to a multitude of factors not readily captured within this data set, including surgical patient selection and tumor biology. Further work is needed to understand why these findings diverge from study norms, and to optimize care for all patients with pancreatic adenocarcinoma.