Systematic Review and Meta-Analysis of <i>In Vitro</i> Synergy of Polymyxins and Carbapenems

<jats:title>ABSTRACT</jats:title> <jats:p> Our objective was to examine the evidence of <jats:italic>in vitro</jats:italic> synergy of polymyxin-carbapenem combination therapy against Gram-negative bacteria (GNB). A systematic review and meta-analysis were performed. All studies examining <jats:italic>in vitro</jats:italic> interactions of antibiotic combinations consisting of any carbapenem with colistin or polymyxin B against any GNB were used. A broad search was conducted with no language, date, or publication status restrictions. Synergy rates, defined as a fractional inhibitory concentration index of ≤0.5 or a >2-log reduction in CFU, were pooled separately for time-kill, checkerboard, and Etest methods in a mixed-effect meta-analysis of rates. We examined whether the synergy rate depended on the testing method, type of antibiotic, bacteria, and resistance to carbapenems. Pooled rates with 95% confidence intervals (CI) are shown. Thirty-nine published studies and 15 conference proceeding were included, reporting on 246 different tests on 1,054 bacterial isolates. In time-kill studies, combination therapy showed synergy rates of 77% (95% CI, 64 to 87%) for <jats:named-content xmlns:xlink="http://www.w3.org/1999/xlink" content-type="genus-species" xlink:type="simple">Acinetobacter baumannii</jats:named-content> , 44% (95% CI, 30 to 59%) for <jats:named-content xmlns:xlink="http://www.w3.org/1999/xlink" content-type="genus-species" xlink:type="simple">Klebsiella pneumoniae</jats:named-content> , and 50% (95% CI, 30 to 69%) for <jats:named-content xmlns:xlink="http://www.w3.org/1999/xlink" content-type="genus-species" xlink:type="simple">Pseudomonas aeruginosa</jats:named-content> , with low antagonism rates for all. Doripenem showed high synergy rates for all three bacteria. For <jats:named-content xmlns:xlink="http://www.w3.org/1999/xlink" content-type="genus-species" xlink:type="simple">A. baumannii</jats:named-content> , meropenem was more synergistic than imipenem, whereas for <jats:named-content xmlns:xlink="http://www.w3.org/1999/xlink" content-type="genus-species" xlink:type="simple">P. aeruginosa</jats:named-content> the opposite was true. Checkerboard and Etest studies generally reported lower synergy rates than time-kill studies. The use of combination therapy led to less resistance development <jats:italic>in vitro</jats:italic> . The combination of a carbapenem with a polymyxin against GNB, especially <jats:named-content xmlns:xlink="http://www.w3.org/1999/xlink" content-type="genus-species" xlink:type="simple">A. baumannii</jats:named-content> , is supported <jats:italic>in vitro</jats:italic> by high synergy rates, with low antagonism and less resistance development. These findings should be examined in clinical studies. </jats:p>