| Part I Treatment for early stage hepatocellular carcinoma: prognostic comparison of different therapies by Bayesian network-meta analysisObjective: Hepatocellular carcinoma(HCC) is the fifth malignant tumor in the world, hepatic resection(HR) remains the first-line treatment for early stage patients; Radiofrequency ablation(RFA) is the first line ablation technique, in addition, it was reported that TACE combined with radiofrequency ablation may enhanced ablation efficacy. The purpose of this study was to compare prognosis of three treatments: liver resection, radiofrequency ablation, TACE combined with radiofrequency ablation for HCC patients within Milan criteria.Methods: Database including Pubmed, Embase, Web of Science database, OVID SP, Cochrane databases and Google scholar were searched to collect randomized controlled trial, as well as prospective and retrospective cohort study about the prognostic comparison of HR and RFA, HR and TACE plus RFA, RFA and TACE plus RFA for HCC patients within Milan criteria from inception to Feb 2016. This study only included published papers in English. Primary study endpoints is the 3-year, 5-year overall survival(OS) rate and 3-year, 5-year disease-free survival(DFS) rates; secondary endpoint was the rate of major complications. Statistical methods was meta-analysis and Bayesian network-meta analysis.Result: A total of 43 studies containing 7 RCTs were included, the total number of patients involved is 19271.The meta-analysis of prognostic comparison between HR and RFA found HR was associated with a higher 3- year, 5-year OS rate and 3- year, 5-year DFS rate compared with RFA, but RFA had a lower major complications rate compared with HR. These differences were statistically significant.The meta-analysis of prognostic comparison between HR and TACE plus RFA found HR was associated with a higher 5-year OS rate and 3- year, 5-year DFS rate compared with TACE plus RFA, but TACE plus RFA had a lower major complications rate compared with HR. These differences were statistically significant.The meta-analysis of prognostic comparison between RFA and TACE plus RFA found TACE plus RFA was associated with a higher 3- year, 5-year DFS rate compared with RFA, these differences were statistically significant.Network meta-analysis of 3-year OS rate comparison found that HR is better than RFA(OR=1.66; 95% CI 1.28-2.51), the difference was statistically significant. TACE+RFA ranks first, the second is HR, and third place goes to RFA.Network meta-analysis of 5-year OS rate comparison found that HR is better than RFA(OR=1.64; 95% CI 1.38-1.95), the difference was statistically significant. HR ranks first, the second is TACE+RFA, and third place goes to RFA.Network meta-analysis of 3-year DFS rate comparison found that HR is better than RFA(OR=2.05; 95% CI 1.51-2.79), the difference was statistically significant. HR ranks first, the second is TACE+RFA, and third place goes to RFA.Network meta-analysis of 5-year DFS rate comparison found that HR is better than RFA(OR=2.16; 95% CI 1.58-2.93), the difference was statistically significant. HR ranks first, the second is TACE+RFA, and third place goes to RFA.Network meta-analysis of major complications rate comparison found that RFA had a lower complication rate than HR(OR=0.22; 95% CI 0.14-0.37), the difference was statistically significant. RFA ranks first, the second is TACE+RFA, and third place goes to HR.Network meta-analysis(only RCT studies included) of 3-year OS rate comparison found no statistical difference between the prognosis of HR, RFA and TACE+RFA. HR ranks first, the second is TACE+RFA, and third place goes to RFA.Network meta-analysis(only RCT studies included) of 5-year OS rate comparison found no statistical difference between the prognosis of HR, RFA and TACE+RFA. HR ranks first, the second is TACE+RFA, and third place goes to RFA.Network meta-analysis(only RCT studies included) of 3-year DFS rate comparison found no statistical difference between the prognosis of HR, RFA and TACE+RFA. RFA ranks first, the second is HR, and third place goes to TACE+RFA.Network meta-analysis(only RCT studies included) of 5-year DFS rate comparison found no statistical difference between the prognosis of HR, RFA and TACE+RFA. HR ranks first, the second is TACE+RFA, and third place goes to RFA.Network meta-analysis(only RCT studies included) of major complications rate comparison found that HR had a higher complication rate than RFA(OR=13.17; 95% CI 3.18-56.11) and TACE+RFA(OR=4.49; 95% CI 1.10-40.22), the difference was statistically significant. RFA ranks first, the second is TACE+RFA, and third place goes to HR.Conclusions: Network meta-analysis(only RCT studies included) found there is no prognostic difference between HR, RFA and TACE+RFA treatment for early-stage HCC patients within Milan Criteria. Besides, network meta-analysis(only RCT studies included) of major complications rate comparison found that HR had a higher complication rate than RFA and TACE+RFA.Traditional meta-analysis and the network meta-analysis found HR had a better prognosis than RFA, and HR ranks first, so for early-stage HCC patients within Milan Criteria, HR may still be the first line treatment; TACE+RFA always rank superior to RFA, so for RFA patients, combined TACE may be considered. All the conclusions of this study need more studies for further verifying.Part II Treatment for early stage hepatocellular carcinoma: appropriate population selection for radiofrequency ablation based on risk evaluation of microvascular invasionObjective: Microvascular invasion(MVI) has been validated as an important predictive factor for prognosis of early-stage hepatocellular carcinoma(HCC) patients. However, MVI positive or negative can only be found by pathology, this narrows the application of MVI. By assessing the MVI risk of HBV related early-stage HCC patients treated by radiofrequency ablation(RFA), an individualized RFA treatment may be established.Patients and Methods: 5850 patients underwent hepetectomy from Eastern Hepatobiliary Surgery Hospital were retrospectively collected between Jan 2008 and Dec 2010, finally 1344 patients constructed the hepetectomy cohort; 1367 patients underwent radiofrequency ablation from the same hospital between Jan 2008 and Dec 2012 were retrospectively collected, finally 430 patients constructed the ablation cohort. A scoring system for predicting MVI was constructed using clinical data of the hepetectomy cohort. Patients in the ablation cohort were scored and the prognosis were compared with MVI low risk and high risk. For all MVI high risk patients in the study, prognosis comparison between hepatectomy and RFA was done, and the same in the MVI low risk group.Results: 360(26.8%) patients in the hepatectomy cohort(1344 patients) were MVI positive. The MVI scoring system was constructed by the following independent risk factors of MVI after logistic regression: AFP ≥ 20μg/L, HBV DNA>104 IU/m L, incomplete tumor capsule, tumor diameter larger than 3cm and PLT<100*109/L. The MVI score system ranging from 0-6 is MVI low risk, and 7-12 is MVI high risk. Sensitivity of the MVI score system was 74.17%(95%CI 69.3%- 78.6%), specifity was 71.75%(95%CI 68.8%- 74.5%), positive likelihood ratio was 2.63(95%CI 2.4- 2.8), negative likelihood ratio was 0.36(95%CI 0.3- 0.4). In the ablation group, overall survival, time to recurrence and disease free survival in MVI low risk patients were significantly better than the MVI high risk patients; for the MVI high risk patients, hepatectomy had a significantly better prognosis than RFA; but in the MVI low risk patients, no statistical difference was found between the two treatments.Conclusion: MVI was an important factor for RFA patients’ prognosis. The proposed MVI scoring system can differentiate the prognosis of HBV related early-stage HCC patient treated by RFA. For individualized RFA treatment, TACE plus RFA, larger ablation zone and longer ablation time may be proposed for MVI high risk patients, and for MVI low risk patients, RFA treatment alone may provide an optimal prognosis. |
PDF Full Text Request