Follow-up Value Of Multi-modality Imaging In The MRI Assessment Of Recurrence After Transcatheter Arterial Chemoembolization Combined Radiofrequency Ablation Of Nodular Hepatocellular Carcinoma

Objective:To study the follow-up value of Multi-modality imaging in the MRI assessment of recurrence after transcatheter arterial chemoembolization(TACE)combined radiofrequency ablation(RFA) of nodular hepatocellular carcinoma(HCC).Materials and Methods: We retrospectively reviewed electronically stored imaging interpretation records during a 5-year period(from September 2009 to September 2014) and identified consecutive patients with HCCs that were comfirmed by clinical diagnosis or biopsy and pathology. The inclusion criteria for HCCs wasthe ministry of health of the People’s Republic of China 2011 edition. We enrolled 149 cases that were treated with TACE combined RFAand then followed-up by multi-modality imaging of MRI after 3-5 weeks. Following-up time was 3-4 months as regular intervals, no other treatments were done until the target lesion tumor recurrences. We enrolled 105 cases include 87 cases of male and 18 cases of female, age 46-83 yearfinally,the median age was 63 years old. Hepatic MRI examination consisted of conventional T1 WI, T2 WI, DWI and LAVA sequences.The cases were divided into three groups: recurrence group, suspicious and postoperative recurrence group.All the MRI datum analysis according to the following aspects:1. The recurrent focal area signal’s change and morphological signs:Combining with the focal area tumor recurrence by positioning observation in the recurrence period in each of sequence(T1WI, T2 WI, DWI, LAVA) can analysis all the signal’s change and morphological signs; 2.datum analysis:(1) we measured the focal involved area in three recurrence period in DWI by ADC values and took the average of the three measurements.(2) the enhancement index:we measured separately the signal value in focal recurrence of involved areas and the normal liver tissue in the planning scanning and the arterial enhancement, then we would take all the values into the formula and calculate the enhancement index.(SI nodule/ SI liver) aeterial phase /(SI nodule / SI liver)unenhanced, where SI nodule represents the SI of the tumor foci tumor recurrence, and SI liver represents the SI of the liver parenchyma.3. The statistical analysis:(1) signal, morphological information: By using Chi-square test we compared the sequence signal and shape change between the three groups. While the difference was statistically significant, we compared two groups between the three groups respectively. According to the recurrence group,we numbered the percentage of signal and shape of each sequence. Comparing the recurrence group- postoperative a group to the sequence in the diagnosis of threshold by using ROC curve. According to the threshold, we divided suspicious and postoperative sets of signals scoreand shape into two groups(with or without). Logistics regression was used to calculate the sequences,meanwhile, signal and the morphological characteristics were used to diagnosis the sensitivity, specificity in suspicious group. To find the best unite in the biggest Youden index, we unite the sequences by two joints, three joints and four joints.(2) the quantitative information: we used nonparametric the Mann- Whitney U test and the Bonferroni method corrected the test level among the recurrence, suspicious and postoperative groups. Using ROC curve in ADC values and enhancement ratio of suspected group, we calculated the area under curve(AUC), sensitivity and specificity. Logistics regression is used to calculate ADC values and enhancement ratio of suspected group and got the sensitivity, specificity and Youden index to find the best quantitative diagnostic method and the best sequence for diagnosis of suspected unite.Result:1. Signal and morphological analysis results: The involvement of three phase Signal, recurrence of area morphology of chi square test, P< 0.05, with significant.the results of Logistics regression, showed the signal characteristics of suspicious period : T1 WI signal is priority to low and mixed; T2 WI signal is high and mixed; DWI signal is high signal; LAVA signal is significantly enhanced. Morphological characteristics: most of each sequence with half moon. 2. The diagnosis effectiveness:(1)diagnostic effectiveness alone: the sensitivity, specificity of the suspicious DWI were respectively: 76.2%,93.3%.It was better than the other groups.(2)diagnosis of the two joints: the joint of T2 WI and DWI can get the biggest Youden index(0.762) and the best diagnosis effectiveness.(3)diagnosis of the three joints: the joint of T2 WI, T1 WI and DWI can get the biggest Youden index(0.754);(4)The diagnostic sensitivity, specificity and the Youden index of T1 WIcombined with T2 WI, DWI and LAVA were respectively 85.7 % and 94.3%and 0.800. 3. The quantitative data was used in rank test and if the P value less that 0. 05, the difference was statistically significant. The sensitivity and specificity in combination of three groups with ADC values and enhancement ratio index in the diagnosis of suspicious lesions wereas followed respectively: 98.1% and 96.2%.4. Joint to morphology and quantitative value can get the sensitivity, specificity and Youden index respectively 97.1%, 96.29% and 0.848, the positive predictive value and negative predictive value were respectively 93.11% and 90.76%.Conclusion: Multi-modality imaging of MRI has the potential to assess the recurrence after TACE combined RF of nodular HCCs.