| ObjectiveReview clinical benefits,including the effective rate,disease control rate,survival rate,quality of life,adverse events rate of anticancer drug,etc.by ginsenosides combined with TACE/chemotherapy in advanced liver cancer and do Meta analysis MethodsIn accordance with the principles of the ?Cochrane Handbook for Systematic Reviews of Interventions? and ?Preferred reporting items for systematic reviews and meta-analyses: The PRISMA statement ?,to develop appropriate research strategies and approaches,make full use of medical research tools and software as much as possible,carry out document retrieval,literature screening,inclusion studies,risk assessment,extraction data,statistical analysis,and evaluation of statistical results.Firstly,Medline,Embase,EBM Reviews,Chinese National Knowledge Infrastructure(CNKI)and Wan Fang data were searched from inception to October2016,to collect randomized controlled trials(RCTs)about treating liver cancer with ginsenosides and select qualified literature.Secondly,the bias risk of the included studies was assessed by using the RevMan software.A set of outcome indicators for assessing clinical efficacy was extracted from the included studies and meta-analysis was performed on each outcome index of the included studies.The statistical methods were chosen according to the data type of outcome.If the data type is continuous outcome,mean difference(MD)and its 95% confidence intervals(CI)will be used for the comparison between group difference;if the data type is dichotomous outcome,relative risk(RR)and its 95% confidence intervals(CI)will be usedinstead.The statistical model was selected according to the heterogeneity test.If the heterogeneity is not statistically significant,a fixed effect model will be used;if the heterogeneity is statistically significant,a random effect model will be used.The primary outcome indicators are effective rate,disease control rate,survival rate,elevate rate of life quality,effective rate of life quality in the study.The secondary outcome indicators are side effects of anti-cancer drug,improvement,stabilization and reduction index of Child-Pugh's rating score,ALT index,AFP index and immune indexes(CD3,CD4,CD4/CD8 ratio and NKC).Finally,the results(evidence)of the statistical analysis of the primary outcome indicators are evaluated from three aspects.The publication bias is evaluated by using the funnel plot,Begg's test and Egger's tests.The sample size is evaluated by using the TSA software.The level of evidence is evaluated by applying GradePro software.ResultsAccording to the inclusion and exclusion criteria,18 randomized controlled trials(RCTs)were collected.A total of 1118 patients including the treatment group of 557 cases and the control group of 561 cases.All the included studies were from mainland China and were published in Chinese medicine journals in mainland China.The risk of bias summary of 18 included studies showed that 13 included studies determined as unclear,indicating the unknown risk of bias;5 included studies determined as low risk,indicating the low risk of bias.The risk of bias graph of the included study showed that the risk assessment of each item of the 18 included studies were as follows: random sequence generation: 8 low risk accounting for 44%,10 unclear accounting for 56%;allocation concealment:5 low risk accounting for 28%,13 unclear accounting for 72%;blinding of outcome assessment:18 unclear accounting for 100%;blinding of participants and personnel: 5 low risk accounting for 28 %,1 high risk accounting for 5.6%,12 unknown accounting for 67%;incomplete outcome: 16 low risk accounting for89%,2 high risk accounting for 11%;selective reporting: 18 unclear accounting for 100%;other Bias: 18 low risk accounting for 100%.The heterogeneity of the outcome indicators showed that there were not statistically heterogeneous in terms of effective rate,survival rate,elevate rate of life quality,effective rate of life quality,gastrointestinal toxicity rate of anti-cancer drug,renal function damage toxicity rate of anti-cancer drug,ALT index,improvement,stability and reduction of Child-Pugh's grading score,CD3 before treatment,CD4,CD4/CD8 ratio,NKC of immune indexes,there were no heterogeneity in terms of disease control rate,blood system toxicity rate of anti-cancer drug,AFP indicators,CD3 after treatment.Meta-analysis of the primary outcome indicators showed that effective rate(RR=1.30,95% CI=1.13,1.50,Z=3.73,P=0.0002);disease control rate(RR=1.11,95% CI=1.02,1.21,Z=2.29,P=0.02);Overall survival rate(RR=1.24,95%CI=1.11,1.39,Z=3.86,P=0.0001);elevate rate of life quality(RR=1.44,95%CI=1.19,1.74,Z=3.73,P=0.0001);effective rate of life quality(RR=1.24,95% CI=1.14,1.35,Z=4.98,P<0.00001).All the outcomes of the above two groups were compared,the difference was statistically significant indicating the clinical efficacy.Meta-analysis of secondary outcome indicators showed that the gastrointestinal toxicity rate of anticancer drugs(RR=0.6,95%CI=0.47,0.77,Z=4.15,P=<0.0001)and the blood system toxicity rate of anti-cancer drug(RR=0.49,95%CI=0.34,0.70,Z=3.9,P=<0.0001),the difference between these two groups was statistically significant,indicating that ginsenosides can significantly reduce the gastrointestinal toxicity rate and the blood system toxicity rate,however the renal function damage toxicity rate of anti-cancer drug(RR=0.81,95%CI=0.63,1.05,Z=1.62,P=0.11)was no statistically significant difference between these two groups,indicating that ginsenosides did not improve the side effects of renal injury caused by anticancer drugs.The improvement and stabilization index of Child-Pugh's grading score(respectively: RR=1.59,95%CI=1.08,2.34,Z=2.36,P=0.02 and RR=1.20,95%CI=1.03,1.14,Z=2.26,P<0.02)and the lower index of Child-Pugh's grading score(RR=0.43,95%=0.27,0.68,Z=3.57,P=0.0 004),the difference between these two groups was statistically significant,indicating that ginsenosides contribute to Child-Pugh's grading assessment,improve and stabilize liver function and prevent liver function deterioration.ALT quantitative analysis showed that ALT(MD=-6.48,95%CI=-10.17,-2.78,Z=3.44,P=0.0006)before treatment and ALT(MD=-30.79,95%=-36.12,-25.47,Z=11.33,P<0.00001)after treatment,the results showed that although the difference between these two groups was statisticallysignificant,but they were from different baseline before treatment,thus ALT could not be correctly assessed.The result of ALT qualitative analysis(RR=0.76,95% CI=0.58,0.99,Z=2.03,P=0.04)showed that the decrease rate of ALT in ginsenoside group was higher than in control group,the difference was statistically significant,indicating that ginsenosides could reduced ALT and improve liver cell damage.AFP analysis results(RR=2.43,95% CI=0.87,6.78,Z=1.70,P=0.09)showed the difference between the two groups was not statistically significant,indicating that ginsenosides did not reduce AFP indicators.Sensitivity analysis,however,shows that the heterogeneity included in the study is extremely high,indicating that the result is not stable and has low credibility.Meta-analysis of immune indexes(CD3,CD4,CD3/CD4 ratio and NKC)showed that there were no significant difference between these two groups before treatment,but the difference between these two groups after treatment were statistically significant indicating that ginsenosides improved the immune index.The results(evidence)of the statistical analysis of the primary outcome indicators are evaluated from three aspects as follows:(1)The qualitative analysis of the funnel plot,the results of the Begg's test and the Egger's test showed that the publication bias was not detected in the results of the statistical analysis of the two outcome indicators(effective rate,elevate rate of life quality)and the publication bias was detected in the results of the statistical analysis of the three primary outcomes(disease control rate,overall survival rate,effective rate of life quality).(2)Trial Sequential Analysis(TSA)showed that although the cumulative Z-curve(Z-curve)results of the five primary outcome indicators(effective rate,disease control rate,overall survival rate,elevate rate of life quality,effective rate of life quality)through the statistical Z value(the traditional boundary value),indicating that the statistics Z greater than 1.96,and P <0.05 were statistically significant,but the Z-curve(Z-curve)did not crossed the TSA boundary value and the cumulative amount of information did not reach RIS,indicating that the results of the meta-analysis may have a false positive and require more follow-up randomized controlled trials to further validate the index.(3)GRADE analysis showed that although the five primary outcomes were statistically significant,however the statistical outcomes(evidence)of the two outcome indicators(effective rate,elevate rate of life quality)were at Moderate quality level,the statistical outcomes(evidence)of three outcome indicators(disease control rate,overall survival rate,effective rate of life quality)were at low quality level.ConclusionCurrent evidence have shown the effectiveness in treating liver cancer.However due to limitation of quantitative and qualitative included studies,the reliability of the meta-analysis will requit more high quality studies,and more sample size for further improvement. |
PDF Full Text Request