Livermore David M, Mushtaq Shazad, Warner Marina, James Dorothy, Woodford Neil
Antimicrobial Resistance and Healthcare Associated Infections Reference Unit, Public Health England, London, UK Norwich Medical School, University of East Anglia, Norwich, Norfolk, UK
Antimicrobial Resistance and Healthcare Associated Infections Reference Unit, Public Health England, London, UK.
J Antimicrob Chemother. 2015 Dec;70(12):3259-66. doi: 10.1093/jac/dkv265. Epub 2015 Aug 27.
A 1 mg/L susceptibility breakpoint for ceftaroline and staphylococci is universally agreed; EUCAST counts MIC >1 mg/L as resistant whereas CLSI and FDA count 2 mg/L as intermediate and >2 mg/L as resistant. We investigated whether routine diagnostic tests reliably distinguish MICs of 1 versus 2 mg/L.
Thirty-five UK laboratories collected Staphylococcus aureus isolates and performed tests with 5 μg (as EUCAST) or 30 μg (as CLSI) discs and either confluent growth on Mueller-Hinton agar (as EUCAST and CLSI) or semi-confluent growth on Iso-Sensitest agar (as BSAC). They also ran Etests for MRSA. Reference MICs were determined centrally by CLSI and BSAC agar dilution.
We obtained paired local disc and central MIC results for 1607 S. aureus (33% MRSA). EUCAST's zone breakpoint recognized 56% of isolates found resistant in MIC tests, but the positive predictive value (PPV) for resistance was 11.0%; corresponding proportions by CLSI testing were 28.0% and 13.4%. The BSAC disc method detected 25% of resistant isolates, with a PPV of 18.2%. Essential agreement, ±1 dilution, of local Etests and central agar MICs was >95%, but only 20% of the isolates found non-susceptible by agar dilution were found non-susceptible by Etest and vice versa. Review for isolates with the modal MIC (0.25 mg/L) indicated that the same laboratories reported large or small zones irrespective of disc and method, implying systematic bias.
MRSA with ceftaroline MICs of 1 and 2 mg/L were poorly discriminated by routine methods. Solutions lie in greater standardization, automation or dosages justifying a higher breakpoint.
头孢洛林对葡萄球菌的药敏折点为1mg/L已得到普遍认可；欧洲抗菌药物敏感性试验委员会（EUCAST）将MIC>1mg/L计为耐药，而美国临床和实验室标准协会（CLSI）及美国食品药品监督管理局（FDA）将2mg/L计为中介，>2mg/L计为耐药。我们调查了常规诊断试验能否可靠区分1mg/L和2mg/L的MIC。
35家英国实验室收集金黄色葡萄球菌分离株，用5μg（按照EUCAST）或30μg（按照CLSI）的纸片进行试验，并在Mueller-Hinton琼脂上进行融合生长试验（按照EUCAST和CLSI）或在Iso-Sensitest琼脂上进行半融合生长试验（按照英国抗菌化疗学会（BSAC））。他们还对耐甲氧西林金黄色葡萄球菌（MRSA）进行了Etest试验。参考MIC由CLSI和BSAC琼脂稀释法集中测定。
我们获得了1607株金黄色葡萄球菌（33%为MRSA）的配对局部纸片和中心MIC结果。EUCAST的抑菌圈折点识别出MIC试验中56%的耐药菌株，但耐药的阳性预测值（PPV）为11.0%；CLSI试验的相应比例分别为28.0%和13.4%。BSAC纸片法检测出25%的耐药菌株，PPV为18.2%。局部Etest试验与中心琼脂MIC试验在±1个稀释度的基本一致性>95%，但琼脂稀释法检测为非敏感的菌株中，只有20%通过Etest试验检测为非敏感，反之亦然。对MIC值为众数（0.25mg/L）的菌株进行审查表明，相同的实验室无论纸片和方法如何，报告的抑菌圈大小差异很大，这意味着存在系统偏差。
常规方法难以区分头孢洛林MIC为1mg/L和2mg/L的MRSA。解决办法在于提高标准化、自动化程度或采用更高折点的合理剂量。
