多黏菌素B和黏菌素体外药敏试验系统评估

Assessment of in vitro AST system for polymyxin B and colistin

目的:评估国内临床实验室常用的7种多黏菌素药物敏感性检测系统,为临床实验室对多黏菌素体外药物敏感性检测提供支持。方法:选取全中国多中心2017-2018年抗菌药物耐药趋势监测研究(SMART)项目中的512株革兰阴性菌,采用肉汤微量稀释(BMD)法检测菌株对多黏菌素B和黏菌素的最低抑菌浓度(MIC),选取MIC为0.25~64 mg/L的102株菌株,其中大肠埃希菌60株,肺炎克雷伯菌37株,不动杆菌属5株。采用BMD-A法、纸片扩散(KB)法、E-test法(Etest-A和Etest-B)和自动化药敏检测系统(NMIC/ID4、NMIC-413及DL-120,简称DL-120药敏系统)的4种方法,检测大肠埃希菌、肺炎克雷伯菌和不动杆菌属,对7种多黏菌素药物敏感性试剂及系统进行评估。以BMD法为参考方法分析4种方法与参考方法的基本一致率(EA)、分类一致率(CA)、非常严重错误(VME)及重大错误(ME)。结果:BMD-A试剂检测多黏菌素B的MIC的CA、EA均最高,分别为98%和85%,且VME最低,KB法检测多黏菌素B与黏菌素的抑菌圈直径和MIC存在相关关系(r=-0.579,r=-0.410;P<0.001)。在2种E-test法中,Etest-A法检测的CA、EA、VME和ME分别为93.1%、78.2%、13.3%和3.5%,而Etest-B的MIC偏低,VME为57.8%,相比于Etest-B,Etest-A的差异(MD)为26.5%。NMIC/ID-4及NMIC-413一致性较好,CA、VME均为95.1%和10.6%。DL-120药敏系统CA最低,为71.6%,且VME也略有欠缺,为61.7%。结论:BMD-A法试剂检测的菌株对多黏菌素B的MIC性能较好,NMIC/ID4与NMIC-413两种系统一致性较好,而Etest-B法对临床菌株检测有较高的VME,不应在临床用于多黏菌素B和黏菌素的体外敏感性检测。

Objective: To assess 7 kinds of common antimicrobial susceptibility test(AST) systems for polymyxin in domestically clinical laboratories, so as to provide support for in vitro AST of clinical laboratory for polymyxin.Methods: 512 strains of gram-negative bacterium were selected from the multi central study for monitoring antimicrobial resistance trends(SMART) in China during 2017-2018. The broth micro dilution(BMD) was adopted to detect the minimal inhibitory concentration(MIC) of bacterial strain on polymyxin B and colistin, and 102 bacterial strains which MIC was between 0.25 mg/L and 64 mg/L were selected. Among of them, 60 strains were Escherichia coli, and 37 strains were Klebsiella pneumonia, and 5 strains were Acinetobacter bacteria. BMD-A method, KirbyBauer, E-test method(Etest-A and Etest-B) and automatically AST system(NMIC/ID4, NMIC-413 and DL-120which abbreviation was DL-120 AST system) were adopted to detect Escherichia coli, Klebsiella pneumonia and Acinetobacter bacteria. The above 7 kinds of AST agents and systems were assessed for polymyxin. The BMD method was used as reference method to analyze the essential agreement(EA), category agreement(CA), very major error(VME) and major error(ME) among 4 kinds of methods and reference method. Results: The CA and EA of MIC of BMD-A were the highest in detecting polymyxin B, and they were 98% and 85%, respectively, and VME of that was the lowest. There were correlations between the diameters of inhibition zones of KB method and MIC for polymyxin B and colistin(r=-0.579, r=-0.410, P<0.001). In two kinds of E-test methods, the CA, EA, VME and ME of Etest-A method were 93.1%, 78.2%, 13.3% and 3.5%, respectively, while the MIC of Etest-B method was lower, and the VME of that were 57.8%. Compared with Etest-B, the major discrepancy(MD) of Etest-A was 26.5%. The consistency between NMIC/ID-4 and NMIC-413 was favorable, and CA and VME were 95.1% and 10.6%, respectively. In addition,the CA of DL-120 AST system was lowest, which was 71.6%, and VME of that also was deficiency, which was 61.7%.Conclusion: MIC performance of BMD-A method is better in detecting bacterial strain for polymyxin B, and the consistency between NMIC/ID-4 and NMIC-413 is favorable. Etest-B has higher VME in detecting clinical bacterial strain, which should not be used in in vitro AST for polymyxin B and colistin in clinical application.