| Objective: Pancreatic ductal adenocarcinoma(PDAC)accounts for 85-95% of pancreatic cancer.It is one of the most lethal tumors in the world.The 5-year overall survival rate is less than 8%.MR elastography(MRE)can be used to noninvasively measure the elasticity of normal pancreas and pancreatic lesions.However,no MRE based biomarkers have been used to predict the prognosis of PDAC.The purpose of this study was to investigate the prognostic value of MRE in patients with PDAC and its correlation with histopathological features.Methods: Methods: 231 patients who had received pancreatectomy according to the plan were prospectively collected.Before the operation,routine MRI and MRE sequences were used to evaluate tumor elasticity,signal and enhancement features,and tumor location,tumor size,pathological TNM stage and other information were extracted from the postoperative reports.The disease-free survival rate(DFS)and total survival rate(OS)were followed up.The tumor sections were stained with hematoxylin eosin(H&E)and Sirius red,and the mean ratio of tumor stroma was calculated.The multivariate lasso penalty Cox proportional risk model was used to determine the role of clinicopathological information combined with MR imaging features in predicting DFS and OS,and the relationship between tumor hardness and pathology was analyzed.Results: 231 patients with PDAC were divided into high elasticity group(> 3k Pa)and low elasticity group(< 3k Pa).The high elasticity was correlated with high TNM stage(P = 0.007)and large PDAC diameter(P < 0.001).The incidence of high elastic value PDAC in the tail of pancreas(P = 0.003)was significantly higher than that in the head.In single factor Cox regression analysis,tumor location,higher histopathologic T,N stage,R1 surgical stage,higher histological grade,larger tumor diameter and volume,ring enhancement and tumor elasticity were closely related to the poor prognosis of patients.In multivariate Cox regression analysis,elastic value,histological grade and tumor volume were significant independent predictors of DFS;elastic value,histological grade,tumor volume and T stage were significant independent predictors of OS.Among them,pancreatic elasticity was the most predictive independent factor for DFS and OS(hazard ratio was 1.922;P < 0.001 and 1.532;P < 0.001).The Cox model of image + clinical + pathology was significantly better than that of clinical + pathology only(DFS(c-index value: 0.773 [95% CI: 0.738,0.808] and 0.668 [95% CI: 0.629,0.708],P < 0.001);OS(c-index value: 0.736 [95% CI: 0.698,0.774] and 0.682 [95% CI: 0.638,0.725],P = 0.015)).The median OS of patients with high elastic value was 18.0 months,that of patients with low elastic value was 33.0 months,that of patients with high elastic value was 11.0 months,and that of patients with low elastic value was 22.0 months.The proportion of stroma was 41.1 ± 12.5% in tumor sections,and the proportion of interstitial fibrosis in PDAC was weakly and moderately correlated with elastic value(r = 0.322,P < 0.001).In addition,there was correlation between elastic value and dense fibrous stroma,histological grade,tumor cell density,tumor necrosis and the presence of residual acini(P < 0.001).Conclusion: 1.Our study showed that the elastic value and tumor volume of MR images before operation,as well as pathological T stage and tumor grade were independent factors influencing the prognosis of DFS or OS,among which the elastic value measured by MRE was the most predictive independent factor.The risk of poor prognosis of the patients with hard PDAC was significantly higher than that of the patients with soft PDAC,and the OS and DFS prediction model of imaging combined with clinicopathological information was significantly better than that of clinicopathological information.2.The mean interstitial ratio of PDAC was only weakly correlated with the elasticity measured by MRE.The elastic value of tumor is closely related to tumor grade,tumor cell density,fiber content,dense fibrosis,tumor necrosis and the presence of residual acini. |
PDF Full Text Request