Predicting the clinical manifestations in necrotizing acute pancreatitis patients with splanchnic vein thrombosis.

Background: Splanchnic venous thrombosis (SVT) is a relatively rare but important complication of necrotizing acute pancreatitis (NAP). Clinical manifestations and severity of this complication in different patients vary greatly, ranging from mild abdominal discomfort even asymptomatic to lethal gastrorrhagia or hepatic failure. The aim of the present study was to develop a model to predict the clinical manifestations of SVT in NAP patients.

Methods: This retrospective cohort study was conducted in the surgical intensive care unit (SICU) of Jinling Hospital. Patients with the presence of both pancreatic necrosis and SVT were selected for possible inclusion. Both univariate and multivariate logistic regression analyses were applied using 12 indices including age, gender, Acute Physiology and Chronic Health Evaluation II scores (APACHE II), CRP(C - reactive protein) levels, etc to assess potential predictors for symptomatic pancreatic splanchnic venous thrombosis (PSVT) in this cohort. A prognostic nomogram was also applied to develop an easy-to-use prediction model.

Results: A total of 104 patients with necrotizing acute pancreatitis (NAP) and splanchnic vein thrombosis (SVT) from January 2012 to December 2013 were enrolled for analysis. A quarter of study subjects (26 of 104, 25%) developed variable symptomatic manifestations including variceal bleeding, persistent ascites and enteral nutrition (EN) intolerance during the disease course. In the multivariable regression model, the following factors were found to be associated with the occurrence of symptomatic SVT: Balthazar's computed tomography (CT) score (OR = 1.818; 95% CI: 1.251-2.641; P = 0.002), intra-abdominal pressure (IAP) (OR = 1.172; 95% CI: 1.001-1.251; P = 0.043 and presence of SMVT (OR = 6.946; 95% CI: 2.290-21.074; P = 0.001). A prediction model incorporating these factors demonstrated an area under the receiver operating characteristic curve of 0.842.

Conclusions: Balthazar's CT score, IAP and SMVT are predictors of symptomatic SVT in NAP patients. The nomogram we conducted can be used as an easy-to-use risk stratification tool in either clinical practice or future studies.