Competing Risk Analysis of Outcomes of Unresectable Pancreatic Cancer Patients Undergoing Definitive Radiotherapy.

Purpose: We investigated potential factors, including clinicopathological features, treatment modalities, neutrophil-to-lymphocyte ratio (NLR), carbohydrate antigen (CA) 19-9 level, tumor responses correlating with overall survival (OS), local progression (LP), and distant metastases (DMs), in patients with locally advanced pancreatic cancer (LAPC) who received definitive radiotherapy (RT).

Methods: We retrospectively analyzed demographic characteristics; biologically effective doses (BED, calculated with an α/β of 10) of RT; and clinical outcomes of 57 unresectable LAPC (all pancreatic adenocarcinoma) patients receiving definitive RT using modern techniques with and without systemic therapy between January 2009 and March 2019 at our institution. We used Response Evaluation Criteria in Solid Tumors (RECIST) version 1.1 to evaluate the radiographic tumor response after RT. The association between prognostic factors and OS was assessed using the Kaplan-Meier analysis and a Cox regression model, whereas baseline characteristics and treatment details were collected for competing-risk regression of the association with LP and DM using the Fine-Gray model.

Results: A median BED of 67.1 Gy resulted in a disease control rate of 87.7%, and the median OS was 11.8 months after a median follow-up of 32.1 months. The 1-year OS rate, cumulative incidences of LP, and DM were 49.2%, 38.5%, and 62.9%, respectively. Multivariate analyses showed that pre-RT NLR ≥3.5 (adjusted hazard ratio [HR] = 8.245, < 0.001), CA19-9 reduction rate ≥50% (adjusted HR = 0.261, = 0.005), RT without concurrent chemoradiotherapy (adjusted HR = 5.903, = 0.004), and administration of chemotherapy after RT (adjusted HR = 0.207, = 0.03) were independent prognostic factors for OS. Positive lymph nodal metastases (adjusted subdistribution HR [sHR] = 3.712, = 0.003) and higher tumor reduction after RT (adjusted sHR = 0.922, < 0.001) were significant prognostic factors for LP, whereas BED ≥ 67.1 Gy (adjusted sHR = 0.297, = 0.002), CA19-9 reduction rate ≥50% (adjusted sHR = 0.334, = 0.023), and RT alone (adjusted sHR = 2.633, = 0.047) were significant prognostic factors for DM.

Conclusion: Our results indicate that pre-RT NLR and post-RT monitoring of CA19-9 and tumor size reduction can help identify whether patients belong to the good or poor prognostic group of LAPC. The incorporation of new systemic treatments during and after a higher BED RT dose for LAPC patients is warranted.