| Pulmonary tuberculosis which is caused by Mycobacterium tuberculosis is a chronic,infectious disease of the lung. As morbidity and mortality of tuberculosis rapidly increasing in recent years,the prevention and the treatment of tuberculosis have become serious challenges in the medical field of the century. Since 1994,when WHO announced Rifapentine(RPT) and Rifampicin(RFP) as the most potential medicines for the treatment of pulmonary tuberculosis at the world congress of lung cancer healthy,discussion has been focused on both of their efficacy and safety. In 2007,these two medicines were listed as first-line drugs for treatment of tuberculosis by the Medical Letter of America. As the researches going on, more and more related studies about the efficacy and safety of RPT and RFP for the treatment of pulmonary tuberculosis have been done all over the world,although not all of them have the same conclusion. Some of them did not provide enough testing efficiency due to small size of sample, and there are no clinical studies yet that aims to resolve this issue using meta-analysis method and get systematic estimated summing-up,furthermore they still need support of evidence- based medicine.ObjectiveIn this paper,the author analyzed the efficacy and safety of RPT and RFP for the treatment of Pulmonary Tuberculosis by Meta-analysis, the core method of evidence-based medicine. Reliable and comprehensive conclusions were draw by enhancing sample size and testing effectiveness, such as whether RPT really had the advantages of high curative and long-term effects, low toxicities and side effects, and whether it was fit for the intermittent drug using and easy to be monitored. The author aimed to provide a scientific evidence for RPT drug using in TB treatment, which could be better used for medical decision making and used as a therapy guide.Methodsthe author searthed the related papers which analyzed the efficacy and safety of RPT and RFP for the treatment of pulmonary Tuberculosis by randomized clinical controlled trials and clinical controlled trials in the openly published journals both in China and abroad from 1999 to 2009 among Ovid-medline, CBMdisc,Pubmed,CNKI,Wanfang and Weipu databases. At the same time the author inspected the relevant PTB magazines, the proceedings of academic meeting and the references of related papers and reviews by hand, and finally searched the ignored papers by Google and digital library of Chinese hospitals. According to the strict including criteria and excluding criteria, and quality assessments were made according to the evaluation criteria of Jadad scale, and the data was extracted by self-designed data-extracting tables. RevManManage 5.0 provided by the Cochrane Collaboration was used to analysis the data. Combined OR was used as the variables in the study were the binary variables. Heterogeneity test-P value showed by the forest chart was used to judge the research heterogeneity. The combined statistic was calculated by M-H fixed effect model for the index with homogeneity,and by M-H random effect model for the index with heterogeneity. Z value and P value showed by the forest chart for the combined statistical data were used to find out whether it existed significantly difference. The author divided efficacy outcomes into the newly diagnosed and the relapse bacillary pulmonary tuberculosis and added subgroup analysis because the two groups of patients had much similarity in some respects such as subject of treatment,intervention and outcome statement. The sensitivity analysis was performed by fixed effect model and random effect model,respectively,either by OR and RR. Publication bias was analyzed by the degree of asymmetry showed in the funnel chart.Results①According to the strict including criteria and excluding criteria, 24 papers were finally selected including 22 Chinese papers and 2 English papers , 18 RCTs and 6 CCT s. According to the evaluation criteria of Jadad Quality Scale,only one paper mentioned single blind method , but did not mentioned subject of treatment of the 24 articles, and 15 papers given detailed description about default and following-up, 8 papers were mentioned by intention-to-treat analysis, 7 papers by per protocol analysis . Among 24 articles, 2 papers scored 3 (high-quality research), 22 papers scored 0～2 (low-quality research).②The combined analysis of the therapeutic effect showed that RPT and RFP were homogeneous in every index of efficacy (P>0.05). The combined results showed as follows: 1) In the aspect of bacteriology conversion of sputum. 2-month sputum negative conversion rates of the primary treated TB cases were significantly different between the treatments of using RPT and the treatment of using RFP (Z=2.73, P=0.006); but for the relapse TB cases,it did not show significant differences (Z=1.80, P=0.07); at the end of the period of the treatment, there were no significant differences between using RPT and using RFP either in new TB case group or relapse TB case group (Z=0.31, P=0.75; Z=1.45, P=0.15). 2) In the aspect of the foci absorption. There were no significant differences between the group of using RPT and the group of using RFP at the end of the period of treatment on the indexes of obvious effective rate,effective rate of bilateral disease,curable rate,effective (curable/reduction) rate of cavity of the first visit cases (P>0.05). In contrast,for the relapse TB cases, all these indexes were significant differences between the two groups (P<0.05). 3) In the aspect of recurrence rate within the 2 or 3 years after the period of the treatment. There were significantly differences both in the new cases and relapse cases between RPT using group and RFP using group ( for new cases,95%CI: 1.74(1.26,2.39)>1,overall effect Z=3.39,P<0.001; for relapse cases, 95%CI: 0.20(0.06,0.75)<1,overall effect Z=2.39,P=0.02). The results of sensitivity analysis showed that there were no significantly differences in all indexes of efficacy between two groups, such as combined OR, combined RR and their 95% confidence interval, respectively, which prompt stable assessment of efficacy of the two drugs. The publication bias analysis showed that there was no publication bias except sputum negative conversion rate and ineffective (failure/relapse) rate of primary and re-treated positive bacteria TB cases in forest plots.③As for the evaluation of safety,since the side effects of anti-tuberculosis drug were not influenced by whether new or relapse TB cases,side effects caused by RPT or RFP during the treatment of the new cases and the relapse cases were combined to calculated. Expect for the indexes of serum bilirubin elevating,reduction in white blood-cell counts,and drug fever,which were homogeneous,others were all heterogeneous,such as total drug side-effects,abnormal liver function,ALT elevating, gastrointestinal adverse reactions,skin rash,and discontinuation of medication (P<0.05). The combined analysis results showed that expect for skin rash and drug fever showed no significantly differences (Z=1.54,P=0.12;Z=0.80,P=0.43),other indexes were all showed significantly differences between the two groups (P<0.05)。The publication bias analysis showed that expect for the incidence of gastrointestinal adverse reactions,serum bilirubin elevating,and reduction in white blood-cell counts,most of the incidences of the side effects were asymmetric in the funnel chart,which meant there maybe exist publication bias.④In the study,the author, based on the related work, innovated in the part of safety analysis. The author analyzed the incidence of discontinuation of medication in which situation patients of non-tolerance or non-compliance with drugs dosing were calculated as a variable to assess the safety by Meta-analysis. The combined result showed that there was significantly difference between the two groups (Z=3.99, P<0.001). The result of sensitivity analysis was stable, but the forest chart suggested potential publication bias. The novel method was designed for the patients who acquired rifamycin resistance if they were broken in medicines use,and it needed further research to prove its feasibility.ConclusionFrom the results above,RPT was more effective than RFP when it is used to treat TB during the intensive phase of primary treated TB cases and focus absorption in X-ray appearance of re-treated TB cases. Recurrence rate of TB in the following 2 or 3 years after a period of treatment were comparable between RPT (twice a week) and RFP (once a day) in the primary treated TB cases, but lower in the re-treated TB cases. And both were similar in other aspects of efficacy. As for safety,the side effects of RPT were less than that of RFP,especially for gastrointestinal adverse reactions and serum bilirubin elevating. The result of comparison of the incidence of discontinuation of medication showed that patients possibly had better tolerance and adherence to RPT. However,definite conclusions could not be draw because of the small sample size,low quality of the individual trails,potential publication bias existed in some of the evaluation indexes,miscellaneous reasons existed in some of the studies,and there were also lack of foreign researches in the study of the relapse TB cases. So a series of large-scale,multi-center,prospective,double- blind and better designed randomized controlled clinical trials are still expected to provide sufficient evidences for RPT on treating TB for clinical practice.
|
PDF Full Text Request