Noninvasive Models Screen New-Onset Diabetics at Low Risk of Early-Onset Pancreatic Cancer

Objectives Several noninvasive models have been developed to identify new-onset diabetics at higher risk of developing pancreatic ductal adenocarcinoma (PDAC). However, they need external validation before implementation. Methods This study validated one such model (Boursi model) among a cohort of new-onset diabetics. A bivariate analysis of the model's components was done between patients who developed PDAC and type 2 diabetics. The model performance was assessed through receiver-operative characteristic curve analysis. Results Patients with PDAC had significantly lower total cholesterol and alkaline phosphatase at diagnosis of diabetes ( P < 0.01). They were observed losing body mass index (BMI) preceding diagnosis (ΔBMI = −0.42 kg/m 2 , P < 0.01). The model's area under the curve was 0.83 (95% confidence interval, 0.79–0.88). The cutoff that maximized the Youden index was at 0.8%. At this cutoff, the sensitivity was 75%, specificity was 80%, and the prevalence of pancreatic cancer increased from 0.19% at baseline to 0.69%. Conclusions Boursi model enriches the prevalence of PDAC among new-onset diabetics.