Efficacy and Safety of Linezolid Combined with Routine Anti-tuberculosis Drugs for Tuberculous Meningitis ： A Meta-analysis / 中国药房

ABSTRACT

OBJECTIVE：To systematically evaluate the efficacy and safety of linezolid （LZD）combined with routine anti- tuberculosis drugs in the treatment of tuberculous meningitis （TBM），so as to provide evidence-based reference for clinical medi- cation. METHODS ：Retireved from PubMed ，Cochrane Library ，Embase，CNKI and Wanfang database ，randomized controlled trials（RCT）of LZD combined with routine anti-tuberculosis drugs （trial group ）versus routine anti-tuberculosis drugs （control group）were collected from the inception to Jan. 2020. After literature screening and data extraction ， the quality of the included literature were evaluated with bias risk assessment tool recommended by Cochrane system evaluator handbook 5.2. Meta-analysis was conducted by using Rev Man 5.3 software，and sensitivity analysis and publication bias analysis were performed. RESULTS ： Totally 9 RCTs involving 602 patients were included. Meta-analysis showed that total response rate [OR ＝4.05，95％CI（2.26，7.26）， P＜0.000 01]， changes of protein content of cerebrospinal fluid [MD ＝0.48，95％CI（0.20，0.77），P＝0.000 8]，changes of white blood cells count of cerebrospinal fluid [MD ＝44.43，95％CI（20.06，68.81），P＝0.000 4]，changes of cerebrospinal fluid glucose/ synchronous blood glucose [MD ＝0.09，95％CI（0.05，0.14），P＜0.000 1] of trial group were significantly higher than those of control group. There was no statistical significance in the changes of chloride content of cerebrospinal fluid [MD ＝8.08，95％CI（－0.64， 16.80），P＝0.07] and the incidence of ADR [OR ＝1.34，95％CI（0.57，3.11），P＝0.50] between 2 groups. The results of sensitivity analysis showed that there were significant differences comparison with before exclusion when the change of protein content in cerebrospinal fluid and the change of glucose/synchronous blood glucose in cerebrospinal fluid were taken as indexes ，and there was no significant difference comparison with before exclusion when the changes of white blood cell count and chloride content in cerebrospinal fluid were taken as indexes. The results of publication bias analysis showed that there was a certain publication bias in this study. CONCLUSIONS ：LZD combined with conventional anti-tuber culosis drugs is effective and safe for TBM. Because the inconsistent results of sensitivity analysis and publication bias exists in publication bias analysis ，the conclusions need to be further confirmed by more large sample and multi-center studies.