Effects of Frozen Storage on Detection of Intermediate Vancomycin Susceptibility and Heteroresistance in Methicillin-Resistant Staphylococcus aureus Blood Isolates

LETTER

Screening for heterogeneous vancomycin-intermediate Streptococcus aureus (hVISA) and VISA phenotypes among clinical methicillin-resistant S. aureus (MRSA) isolates has been performed utilizing Etest-based and vancomycin- or teicoplanin-supplemented agar methods. Screening for hVISA and VISA on brain heart infusion (BHI) agar supplemented with vancomycin concentrations of 3 mg/liter (BHI-V3) and 4 mg/liter (BHI-V4) correlates much better with population analysis profile-area under the curve (PAP-AUC) results (1). These tests are sometimes done on frozen isolates saved in the laboratory for various durations. Frozen-storage conditions had been reported to alter the organism's susceptibility results (2, 3). The effect on hVISA and VISA screening results is unknown. We prospectively monitored blood cultures (BCs) (1 December 2010 to 30 November 2012), selected all MRSA isolates from inpatients (≥18 years old), and screened them on manufacturer-made BHI agar biplates containing BHI-V3 and BHI-V4 (Hardy Diagnostics, Santa Maria, CA) prestorage. BCs were subcultured weekly onto Trypticase soy agar (TSA) with 5% sheep blood. Duplicate 10-μl droplets of a 0.5 McFarland standard suspension in saline were inoculated onto the biplates (1). Duplicate growth of ≥2 CFU/droplet (200 CFU/ml) on BHI-V3 and BHI-V4 after 48 h of incubation at 35°C was considered a positive hVISA and VISA screen, respectively. A portion of the original subculture was frozen in skim milk at −80°C. They were thawed 3 to 24 months later for molecular typing, and a sample of the new cultures was frozen. All prestorage screen-positive samples were rescreened 1 to 3 years poststorage.

Prestorage screening was done on 474 BCs from 197 patients within 0 to 82 days of isolation (median = 7 days). We detected 56 (11.8%) BHI-V3 and 21 (4.4%) BHI-V4 prestorage screen-positive samples. Poststorage screenings showed that 21/56 (37.5%) BHI-V3 screen-positive samples were poststorage screen negative and that 8/21 (38.1%) BHI-V4 screen-positive samples were poststorage BHI-V4 screen negative but remained BHI-V3 screen positive. The number of CFU in poststorage BHI-V3 screening samples was lower in 30/56 (53.6%), identical in 22 (39.3%), and higher in 4 (7.1%) instances, with similar results in poststorage BHI-V4 screenings. The screening differences were 2 to 100 CFU/droplet, which were enough to alter the result interpretation of 21 (37.5%) BHI-V3 and 8 (38.1%) BHI-V4 screen-negative samples. Screen interpretation was also changed poststorage in one additional sample which showed no growth in the prestorage BHI-V4 screen and then showed 3 and 4 CFU on poststorage screening. Our findings illustrate that frozen storage may affect hVISA and VISA screening results and lead to an underestimation of hVISA's and VISA's true prevalence. This effect was more apparent in hVISA samples than in more-stable VISA isolates, probably due to their heterogeneous expression of susceptibility.

The reasons for these changes are uncertain, but the instability of resistant subpopulations, the loss of genetic determinants of reduced susceptibility, revertant mutations after passages on nonselective media, and a selective advantage for susceptible progenies have been proposed as possible causes of an overrepresentation of the susceptible populations after storage (4). Whatever the reasons are, the differences in storage conditions may explain the conflicting reports of the relevance of hVISA and isolates with high vancomycin MIC values. Therefore, frozen-storage conditions and duration should be disclosed and accounted for in hVISA and VISA prevalence and relevance studies.