Pathological complete response to chemotherapy in pancreatic cancer: frequency, predictors and impact on overall survival

Abstract

Background: While Pathological complete response (pCR) after neoadjuvant chemotherapy (NCT) in pancreatic cancer has been associated with improved overall survival (OS), the rate of pCR and its predictors have not been elucidated yet. Our aim was to study factors predicting rate of pCR in a large data set and the impact on OS. Methods: Patients listed in the National Cancer Data Base from 1998 to 2011 were studied. We included patients with NCT followed by surgery. Demographics, cancer characteristics, and treatment modalities including NCT to surgery (NCT-S) interval were studied. Results: A total of 2093 patients were included. Mean age was 62, 51% were male, and 71% had neoadjuvant radiotherapy (NRT). NCT-S was divided into time quintiles in weeks: 8–11, 12–14, 15–19, 20–29 and >29 weeks; as well as a continuous variable). Median follow up was 74 months. Rate of pCR was 2.1% (44/2093). OS in pCR vs non-pCR: 61 vs. 31 months, HR 0.3, 95% CI 0.1–0.9, p 0.03). There was a significant increase in pCR with longer NCT-S interval (quintiles: (5/510 = 1.0%, 9/551 = 1.6%, 8/462 = 1.7%, 12/403 = 3.0% and 10/167 = 6.0%, p < 0.001). In logistic regression: NRT (OR 2.5, 95% CI: 1.1–6.1, p = 0.03) and NCT-S interval per week increase (OR 1.1, 95% CI 1.03–1.09, p < 0.001) as well as NCT-S >29 weeks (OR 6.1, 95% CI 2.02–18.50, p < 0.001) were predictive of increased pCR rate. Multivariate Cox regression showed that age, pCR and NCT-S >29 weeks were significantly predictive of OS, while gender, race, facility type, tumor grade, cT, cN stage, single vs. multiple chemotherapy agents and NRT were not.. Conclusion: NCT-S interval >29 weeks and pCR were independent predictors of better OS in pancreatic cancer patients. NRT predicted increased pCR but not OS in the multivariate analysis. Further studies are needed to optimize multimodality regimens that best improve OS.