Risk factors analysis and nomogram development for idiopathic chronic pancreatitis with common bile duct stricture / 中华胰腺病杂志

ABSTRACT

Objective:To identify the risk factors and develop nomogram for idiopathic chronic pancreatitis (ICP) patients with common bile duct stricture (CBDS).Methods:The clinical data of 1 633 ICP patients admitted to the Department of Gastroenterology of First Affiliated Hospital of Naval Medical University from January 2000 to December 2013 were collected retrospectively and prospectively. The patients were classified into CBDS group ( n=259) and non-CBDS group ( n=1 374) according to whether CBDS occurred. The cumulative incidence of CBDS after the onset and diagnosis of ICP were calculated by Kaplan-Meier method. After excluding patients who had developed CBDS before/or at the diagnosis of ICP, the remaining patients were randomly divided into the training set and the validation set. The univariate and multivariate Cox proportional hazards regression analysis were used to establish a risk predicting nomogram for CBDS after ICP onset. Its clinical application value was evaluated through the consistency index (C index). Results:15.9%(259/1 633) of patients developed CBDS after the onset of ICP. The cumulative incidence of CBDS at 3, 5, and 10 years after the onset of ICP was 9.6% (95% CI 0.082-0.111), 11.2% (95% CI 0.097-0.129) and 16.2% (95% CI 0.142-0.184), respectively. 9.4%(143/1 517) of patients developed CBDS after the diagnosis of ICP. The cumulative incidence of CBDS at 3, 5, and 10 years after the diagnosis of ICP was 8.3% (95% CI 0.069-0.099), 8.9% (95% CI 0.074-0.105) and 13.3% (95% CI 0.110-0.162), respectively. Univariate analysis found that factors including gender, age at onset of ICP, age at diagnosis of ICP, being adolescents at onset of ICP, smoking history, alcohol intake, initial manifestations, pancreatic duct stones, fatty steatorrhea, main pancreatic duct (MPD) morphology and pain type were significantly different between CBDS group and non-CBDS group. Multivariate analysis showed that male ( HR 2.134, 95% CI 1.336-3.408), age at diagnosis of ICP ( HR 1.038, 95% CI 1.024-1.052), first manifestation (pancreatic abdominal pain) and main duct morphology (complex lesion) were identified as independent risk factors for CBDS in ICP patients. A nomogram for predicting CBDS after ICP diagnosis was established based on the above four variables. The nomogram had a C-index of 0.740 (95% CI 0.700-0.790) for internal validation in the training set and 0.650 (95% CI 0.570-0.730) for external validation in the validation set. Conclusions:The nomogram established in this study can evaluate the risk of developing CBDS in ICP patients, benefit the early diagnosis and timely intervention of CBDS in clinical practice, and prevent potential related complications.