Regional variation for tumour-directed treatment and survival in patients with pancreatic adenocarcinoma: a nationwide population-based analysis

Background: In the Netherlands, cancer care is increasingly organised within oncology networks involving multiple hospitals. This nationwide population-based study aims to assess variation between oncology networks for pancreatic adenocarcinoma (PAC). Method: Patients with PAC (2015–2020) were included from the Netherlands Cancer Registry and assigned to eight oncology networks based on first hospital visit. Multilevel multivariable logistic and survival regression models were used. Results: Among 16,130 patients with PAC, tumour-directed treatment was applied in 40 % (range 35 %–44 % across regions), resection in 15 % (12 %–19 %), and chemotherapy in 33 % (29 %–37 %) of patients. Casemix also varied significantly between regions. Compared to the grand mean, the probability to undergo resection was higher in one region (odds ratio [OR] = 1.25, 95 % confidence interval [CI] 1.07–1.45) and lower in another (OR = 0.70, 95%CI: 0.56–0.87). Chemotherapy use followed similar patterns (OR = 1.21, 95%CI 1.06–1.38, and OR = 0.81, 95%CI 0.72–0.92). Median overall survival was 3.7 months (range 3.4–4.3; log-rank p < 0.001), multivariable multilevel analysis revealed no significant survival differences (p = 0.245). Conclusion: Although the likelihood of receiving chemotherapy and resection for patients with PAC varied between Dutch oncology networks, no clinical meaningful survival differences were found after case-mix adjustment. Concerningly, the majority of patients with PAC do not receive any tumour-directed treatment.