| PART? Prevalence of incidental pancreatic cystic lesions in physical examination population undergoing chest CTObjective: The prevalence of incident pancreatic cystic lesions(PCLs)in general population has not clarified.Based on the chest CT imaging examination and clinical data analysis of the physical examination population in Jiangsu Provincial Physical Examination Center from January 2018 to January 2019,the characteristics and prevalence of incident PCLs in general population were calculated.Materials and Methods: Retrospective analysis was performed on the imaging data,biochemical data,and clinical data of 9826 patients underwent physical examination in Jiangsu Provincial Hospital of Traditional Chinese Medicine from 2018 to 2019.The chest CT of all patients was reviewed one by one to detect the types and distribution of pancreatic lesions,and the prevalence of PCLs and CT characteristics of the lesions were calculated.Pancreatic cystic lesions were divided into two groups based on the size of the lesions,greater than 20 mm and t less than 20 mm.The age,gender,and number,location,growth pattern,calcification,and dilatation of the main pancreatic duct of the two groups of lesions were compared.The prevalence of PCLs in each sub-age group(18-29 y,30-39 y,40-49 y,50-59 y,60-69 y,70-79 y,?80y)were evaluated.In addition,patients in each group were divided into by gender and the prevalence of males and females in different age groups were calculated and analyzed.Finally,the prevalence of PCLs in different pancreatic sites were also analyzed..The difference of continuous variables between the two groups was compared with the independent sample t test or Mann-Whitney U test,and the difference of qualitative variables between the two groups was compared by ?2 test or Fisher exact test method.Result: Four types of pancreatic lesions were found in our population: pancreatic calcifications,pancreatic fatty lesions,pancreatic duct dilatation,and PCLs.The prevalence rates were 2.38%(234 cases),0.51%(50 cases),0.41%(40 cases),and 1.75%(172 cases),respectively.A total of 18 cases(10.5%)of incidental PCLs with a lesion diameter greater than 20 mm were found.The sizes of pancreatic cystic lesions were related to the number of lesions and the growth pattern of the lesions(both p <0.001).There were no significant differences in patient age,gender,lesion location,calcification,and main pancreatic duct dilatation(p values were 0.382,0.502,0.467,0.086,0.713,respectively)between large PCLs and small ones.The prevalence of PCLs is significantly related to age.With the increase of age,the prevalence of PCLs increases significantly.There was no significant difference in the prevalence of PCLs between male and female patients in each age group(p values were 1.00,0.362,0.359,0.193,0.900,0.366,0.793,0.737,respectively).The PCLs was commonly occurred in the pancreatic body(0.69%)than other sites.Conclusion: 1.Low dose chest CT screening for lung cancer has a good ability to detect pancreatic lesions.Occasional pancreatic lesions mainly include pancreatic calcification,pancreatic fatty lesions,pancreatic duct dilatation and pancreatic cystic lesions.The prevalence rates are 2.38%,0.51%,and 0.41 %,1.75%,respectively.2.The diameter of incidental pancreatic cystic lesions is significantly related to the number of lesions and growth patterns.3.The prevalence of pancreatic cystic lesions is significantly related to the patient's age,but not to the patient's gender.PCLs commonly occurred in the pancreatic body.PART II Risk factors for incidental pancreatic cystic lesions detected during physical examinationObjective: The risk factors of incidental PCLs were not clear.In this part,we indentified the possible risk factors related with PCLs.Materials and Methods: The 9826 patients included in the first part were divided into two groups: the pancreatic cystic disease group(172 cases)and the non-cystic disease group(9654 cases).The gender,age,ratio of liver / spleen CT value,ratio of pancreas / spleen CT value,liver cyst,Renal cyst,splenic cyst,white / globulin ratio,albumin,total protein,total bilirubin,aspartate aminotransferase,serum alanine aminotransferase,urea,alkaline phosphatase,creatinine,blood glucose,total Cholesterol,low-density lipoprotein cholesterol,high-density lipoprotein cholesterol,and triglycerides were compared between PCLs and non-PCLs.The difference of continuous variables between the two groups was compared with the independent sample t test or Mann-Whitney U test,and the difference of qualitative variables between the two groups was compared by ?2 test or Fisher exact test.The clinical data,biochemical indicators,and CT characteristics of the general populationwere included in the multivariate binary logistic regression analysis.The independent risk of each factor was obtained by the input method.A joint prediction model was built based on CT features and clinical data,and the receiver operating characteristic(ROC)curve was used to calculate the predictive effectiveness of the model.Moreover,the occasional PCLs(172 cases)detected in the first part were divided into high-risk group and low-risk group according to sizes or pancreatic duct dilatation(whether the diameter was greater than 30 mm or the main pancreatic duct dilatation was greater than 5mm)..The clinical data,biochemical indicators,and features of CT werecompared between the two groups..The independent risk factors were identified using logistic regression analysis..A joint prediction model based on CT features and clinical data was developed,and the ROC curve was used to calculate the predictive effectiveness of the model.Result: A total of 9826 medical examination patients were included,including 172 patients with incidental PCLs and 9,654 patients without cystic lesions.Univariate analysis revealed that the age,ratio of liver / spleen CT value,ratio of pancreas / spleen CT value,liver cyst,kidney cyst between the two groups.There were statistically significant differences in acid aminotransferase,alkaline phosphatase,and blood glucose(p values were <0.001,0.004,0.009,0.009,0.01,0.038,0.004,<0.001)between the two groups..Multivariate regression analysis of the total population found that age and ratio of liver / spleen CT value were independent risk factors for pancreatic cystic disease [AUC = 0.721,95% CI(0.682-0.760)],and the sensitivity of the model was 78.4%,which was specific.In the subgroup analysis,it was found that there were significant differences in the risk factors for pancreatic cystic disease men and women.In the male population,age,liver / spleen,white / globulin,globulin and albumin are high-risk factors for PCls.Age and blood glucose were independent risk factors for the female population.Univariate analysis of high-risk PCLs and low-risk PCLs showed that patients with pancreatic / spleen values,lesion location,growth pattern,urea,and creatinine were statistically different between the two groups(p values were 0.036,0.005,0.015,0.002,0.048).Multivariate regression analysis showed that urea,lesion location,and growth mode were independent risk factors for predicting the high risk of pancreatic cystic disease [AUC = 0.799,95% CI(0.681-0.918)],and the sensitivity of the model was 92.9%,with a specificity of 53.2%.Conclusion: 1.Age,liver / spleen are risk factors for PCLs,while the risk factors for pancreatic cystic disease are different in male and female populations.2.High-risk pancreatic cystic lesions are mainly located in the head and neck of the pancreas,and they tend to grow beyond the outline of the pancreas,and associated with urea abnormalities.PART III Progression prediction of incidental pancreatic cystic lesions detected in physical examination populationObjective: 172 cases of pancreatic cystic lesions were followed up for 12 months to explore the associations between baseline characteristics of PCLs(initial examination characteristics)and progression of PCLs.Material method: A retrospective analysis was performed on the follow-up results of 172 patients,and a total of 107 patients were found to have imaging and clinical data at follow-up.The 107 patients were grouped in accordance with the PCLs progression criteria in the ACR guidelines(progressive group / stable group).The baseline data(CT characteristics,clinical data and biochemical characteristics)of the two groups were compared.The independent sample t test or Mann-Whitney U test was used to compare the differences between the continuous variables between the two groups.The ?2 test or Fisher exact test was used to compare the differences between the qualitative variables between the two groups.The parameters with statistical differenceswereincluded in the multivariate binary logistic regression.The independent risk factors predicting the progression of pancreatic PCLs were screened out by the stepwise LR method,and a joint prediction model based on CT features and clinical data and biochemical features was constructed.The ROC curve was used to calculate the model's Forecast performance.Result:Finally 14 cases(13.1%)compared with the initial inspection has been progressed at follow-up.Univariate analysis showed that the age,lesion location,growth pattern,presence or absence of main pancreatic duct dilatation,and urea levels were statistically different between the two groups(p values were 0.028,0.011,0.039,0.027,0.007,respectively).The average age of patients in the progressive group was older than that in the stable PCLs group.In addition,progressive PCLs usually located in the head and neck,showed exogenous growth,had dilated main pancreatic ducts,and had high elevated urea levels.Multivariate logistic regression showed that the lesion location and growth pattern were independent risk factors for predicting the progressive PCLs.The specificity is 93.3% and the specificity is 58.7%.Another model combined the urea and CT imaging findings had AUC value of 0.862(95% CI:0.758-0.939),and the sensitivity was 99.3%,and the specificity was 66.3%.Conclusion: 1.CT imaging features,including the location of the lesion,lesion grows pattern,and the dilatation of the pancreatic duct can help to predict the progression of pancreatic cystic lesions.2.Multivariate logistic regression showed that lesion location and growth pattern are independent risk factors for predicting the progressive of PCLs.3.A combined model including head location,exogenous growth and elevated blood urea levels showed acceptable performance in predicting the progression of PCLs. |
PDF Full Text Request