Secretin-stimulated MR cholangio-pancreatography and pathological correlative study in a swine obstructive chronic pancreatitis model
Jianhua Wang,Jian Zhang,Gaofeng Sun,Yutao Wang,Juanli Mao,Guixia Pan,Ye Peng,Jianming Zheng,Changjing Zuo
DOI: https://doi.org/10.3760/cma.j.issn.1005-1201.2015.09.013
2015-01-01
Abstract:Objective To investigate the correlativity between secretin-stimulated magnetic resonance cholangiopancreatography (sMRCP) findings and pathological severity in a swine chronic <br> pancreatitis (CP) model. Methods Thirty-nine swine were divided randomly into control group (n=12) and experimental group (n= 27). In experimental group, the main pancreatic duct (MPD) was incompletely ligated to establish the model of obstructive CP. In control group, laparotomy was performed but without ligating the MPD. At the 4th, 8th and 12th week after modeling, one third swine of each group were undergone a series of dynamic sMRCP scans before (0 min) and at 1, 3, 5, 7, 9, 11 min after administration of secretin (0.6 μg/kg). And the MPD diameter and duodenum filling (DF) degree were measured. All survivals were sacrificed to pathological examination including HE and Van Gieson staining for histopathological grading. According to pathological severity, swine were divided into normal group, mild CP group and moderate to severe CP group. MRI features and indexes, including baselined diameter (BD), end diameter (ED), maximum diameter (MD), the largest expansion rate (LER), time to peak (Tpeak) and end change rate of pancreatic duct and duodenal filling (DF) scores were measured. The relationships between pathological grading and sMPCP indexes were analysed. The comparison of sMRCP data among the 3 groups were used ariance analysis, χ2 test and U test. Correlations between sMPCP indexes and pathological severity were tested using Spearman rank correlation coefficients. The diagnostic efficiency of sMRCP indexes were evaluated by ROC method. Results (1) In experimental group, 22 CP models were established and 19 CP swine (mild CP, n= 8; moderate and severe CP, n=11) were performed sMRCP successfully. Eleven swine in normal group were obtained satisfactory MRCP images. (2) sMRCP results:BD of 3 groups were (1.56 ± 0.46),(2.95 ± 1.17),(7.41 ± 1.91) mm, respectively. ED were (1.49 ± 0.31),(2.96 ± 1.17) and (7.37±1.90) mm, respectively. MD were (2.39±0.43),(3.91±1.27) and (7.86±1.87)mm, respectively. The median of LER were 42.10%, 34.85% and 6.58%, respectively. The median of DF scores were 3, 3, 2, respectively. The differences of above indexes have statistically significance (P values were all<0.01). There were correlation between sMRCP indexes (BD, ED, MD, LER and DFscores) and pathological severity (r values were 0.89, 0.92, 0.90,-0.85,-0.66, respectively and P values were all<0.01). Tpeak and end change rate of pancreatic duct had no significant differences (P values were>0.05),and no correlation with pathological severity(P values were all>0.05).For differential diagnosis between normal and mild CP, the area under ROC of BD, ED, MD, LER and DFscores were 0.915, 0.977, 0.926, 0.778 and 0.472, respectively and differential diagnosis between mild CP and moderate to severe CP group, the area under ROC were 0.966,0.966,0.960,1.000 and 0.915, respectively. Conclusions sMRCP findings of CP have characteristics and could be used for in vivo evaluation on the CP pathologic grades.