| Background and Objective:Pancreatic portal hypertension(PPH)is a regional syndrome of portal hypertension caused by the obstruction of splenic outflow due to various pancreatic diseases and complications.PPH is the only curable type of portal hypertension with a low morbidity,insignificant symptoms,and combined complications at the time of detection.The incidence of pancreatic diseases increases as the standard of living improves.In addition,with the progress of medical technology and clinical understanding of PPH,the discovery rate of PPH has increased.The primary complication of PPH is the rupture and bleeding of isolated gastric fundal varices,which can be fatal.Thus,early diagnosis and treatment are crucial.As one of the most common causes of PPH,chronic pancreatitis accounts for 31.8% to 53.8% of all aetiologies.Chronic pancreatitis induces splenic vascular endothelial damage through a systemic inflammatory response that alters the coagulation system and forms splenic vein thrombosis,causing splenic blood return system blockage.In this context,this study aimed to investigate the risk factors for the development of PPH and the risk factors for the development of splenic vein thrombosis in chronic pancreatitis,respectively,and to construct corresponding risk prediction models in order to provide guidance and help for clinical diagnosis and treatment.Methods:1.The clinical data and laboratory indexes of 185 patients with pancreatic diseases admitted from January 2004 to January 2022 to the First Affiliated Hospital of the Army Medical University were retrospectively analyzed.The patients were divided into a study group(n = 75)and a control group(n = 110)according to the presence or absence of portal hypertension.Univariate regression,Spearman correlation,multivariable regression,and linear regression analyses were used to screen for PPH risk factors.A nomogram model was established based on these variables.The model’s predictive capability was demonstrated using receiver operating characteristic(ROC)curves.Calibration and decision curve analyses were performed to validate the calibration and utility.2.The clinical data and biochemical indexes of 226 patients with chronic pancreatitis admitted from January 2015 to October 2020 to the First Affiliated Hospital of the Army Medical University were retrospectively analysed.The patients were divided into a study group(n = 100)and a control group(n = 116)according to the presence or absence of spleen vein thrombosis.Univariate and LASSO logistic regression analyses were used to screen risk factors for splenic venous thrombosis in chronic pancreatitis.Subsequently,a nomogram model comprising the potential indicators was created.The ROC curve,calibration curve,and decision curve analyses were drawn to demonstrate the prognostic capability,calibration,and efficacy of the model.3.To validate the predictive accuracy of the established model,at least 100 patients with chronic pancreatitis admitted from January 2010 to December 2014 to the First Affiliated Hospital of the Army Medical University were enrolled as an external validation cohort.Results:1.Risk factors for PPH(1)The proportions of smoking history,diabetes mellitus,and gastrointestinal hemorrhagic anemia were significantly higher in the study group than in the control group,while the differences in erythrocytes,platelets,total bilirubin,alanine aminotransferase,aspartate aminotransferase,alkaline phosphatase,glutamyl transferase,prothrombin time,and the international normalised ratio were statistically significant between the two groups(P<0.05);(2)A correlation analysis indicated that smoking,diabetes,gastrointestinal bleeding,prothrombin time,and the international normalised ratio were positively correlated with PPH incidence.On the other hand,erythrocytes,platelets,total bilirubin,alanine aminotransferase,aspartate aminotransferase,alkaline phosphatase,and glutamyl transferase were negatively correlated with morbidity(P < 0.05);(3)A logistic regression analysis indicated that smoking and diabetes were risk factors for PPH(OR = 5.008,3.913,3.690),whilst erythrocyte count was a protective factor(OR = 0.362);(4)Based on the results of variable screening,the Nomogram model for PPH was constructed,and the area under the curve(AUC)of the model was 0.789,and the calibration curve had a good fit,indicating that the model had a good predictive effect.The decision curve analysis showed that a significant net benefit value was obtained when the risk threshold was between 15%-85%,indicating the high clinical applicability of the model.2.Risk factors for splenic vein thrombosis in chronic pancreatitis(1)The proportions of the male gender,smoking,alcohol consumption,and pancreatic pseudocysts were significantly higher in the study group than in the control group.A significant difference(P < 0.05)was observed between the two groups in laboratory findings such as erythrocytes,haemoglobin,serum calcium,prothrombin time,and D-dimer;(2)A LASSO logistic regression analysis identified Ca2+ as a protective factor(OR = 0.065),whilst smoking,pancreatic pseudocyst,and prolonged PT were risk factors(OR = 2.854,4.558,1.971);(3)A nomogram model was established for splenic vein thrombosis in chronic pancreatitis based on the variable selection results.The AUC of the model was 0.794,which showed a strong alignment with the calibration curve,indicating that the model had viable predictive capability.A decision curve analysis showed a positive net benefit with a risk threshold between 15% and 85%,indicating that the model showed high clinical applicability.3.Validation of the model for splenic vein thrombosis in chronic pancreatitisThe validation group data were included in the model.Then,when the ROC curve was plotted,an AUR of 0.789 was elicited,which displayed a trend similar to the calibration curve.A decision curve analysis suggested a positive net benefit when the risk threshold was between 20% and 90%,indicating that the model provided viable predictive capability.Conclusion:1.Smoking,aetiology,and diabetes were risk factors for the occurrence of PPH.Smoking and diabetes increased the probability of pancreatic portal hypertension by 5.008 and 3.913 times,respectively.However,RBC was a protective factor against pancreatic portal hypertension.The incidence of pancreatic portal hypertension in patients with increased red blood cell counts was 0.362 lower than in patients with decreased red blood cell counts.The risk factor model for PPH constructed in this study showed viable predictive capability.2.Smoking,pancreatic pseudocysts,and prolonged PT values were identified as risk factors for splenic vein thrombosis in chronic pancreatitis.The probabilities of splenic vein thrombosis in chronic pancreatitis due to smoking,pseudocysts,and prolonged PT values increased by 2.854,4.558,and 1.971 times,respectively.On the other hand,serum Ca2+ concentration was a protective factor.The probability of splenic vein thrombosis in patients with increased serum Ca2+ concentrations was reduced by 0.065 compared to those with lower serum Ca2+ levels.In this study,the constructed risk factor model for the progression of splenic vein thrombosis in chronic pancreatitis showed viable predictive capability,as confirmed via external validation. |
PDF Full Text Request