A comparison of MIC-based screening tests for betalactamase-negative ampicillin-resistant Haemophilus influenzae

Most B-lactamase negative ampicillin (AMP) resistant H. influenzae (BLNAR) have an N526K substitution in penicillin binding protein 3 (PBP3). Their detection is problematic because MICs cluster near the breakpoints (BPs), there is no consensus on BPs and there is poor correlation between MIC and disc diffusion (DD) zones. By strict CLSI criteria, BLNAR strains have AMP MICs >=4 mg/L although most use the non-susceptible BP of >=2, consistent with the EUCAST resistant BP of >1. CLSI DD uses a 10ug disc even though Karpanoja (2004) showed poor correlation of zones sizes and MIC near the BP with these discs and recommended using a 2ug disc with better correlation, which is the disc strength used by EUCAST. Many of these problems are compounded in strains with both altered PBP3 and B-lactamase. Given that strains with altered PBP3 have reduced susceptibility to AMP, amoxicillin-clavulanate (AMC) and cephalosporins, a paradigm shift away from AMP MIC as a basis for detection is warranted. Here we compare the use of a cefotaxime (CTX) screen against current methods for the detection of strains with altered PBP3.