Decreasing in Vitro Susceptibility of Clinical Staphylococcus aureus Isolates to Vancomycin at the New York Hospital: Quantitative Testing Redux

Recent descriptions of Staphylococcus aureus isolates with intermediate- and high-level vancomycin resistance [1, 2] have been accompanied by reports of the slow bactericidal activity of vancomycin and of vancomycin treatment failures in persons with infection due to methicillin-resistant S. aureus (MRSA) [1–3]. However, vancomycin-resistant isolates of S. aureus are exceedingly rare, and most reported cases of vancomycin treatment failure involve organisms that are fully susceptible to vancomycin [2, 3]. One recent study of vancomycin-susceptible MRSA isolates correlated the clinical efficacy of vancomycin with both the MIC for the target organism and the magnitude of bactericidal activity exerted by vancomycin. Among MRSA isolates with MICs ⩽2 μg/mL, a range of >7 log10 µg/mL in the bactericidal activity of vancomycin was observed [3]. These results emphasize the potential importance of the bactericidal activity of vancomycin and the relative imprecision of the MIC measurement as a quantitative surrogate of vancomycin susceptibility among S. aureus. We reviewed the vancomycin susceptibility patterns of methicillin-susceptible S. aureus and MRSA isolates tested in the clinical microbiology laboratory at Weill Medical College of Cornell University (New York, NY) during 1994–1999. Over this period, 171 isolates obtained from patients with invasive S. aureus infections (mainly endocarditis or osteomyelitis) were tested for the vancomycin MIC and for the minimum bactericidal concentration (MBC) of vancomycin required to kill 99.9% of bacteria in a starting inoculum of 1 × 105 bacteria/mL over a 24-h period using a single, macrotube-based technique [4]. At the beginning of this period, similar proportions (∼15%) of isolates with MICs and MBCs ⩾1.6 µg/mL were observed. By the end of this period, however, a ∼3-fold (∼50%) increase in the proportion of isolates with MBCs ⩾1.6 µg/mL was identified by a χ2 test-for-trend analysis (P < .001; figure 1). No trend was detected when tracking either the proportion of isolates with MICs ⩾1.6 µg/mL (figure 1) or MBCs ⩾3.2 µg/mL. Analysis of 52 clinical isolates tested during 2002–2004 by means of a microtiter plate-based assay [5] similarly demonstrated an increasing proportion of isolates with MBCs ⩾2 µg/mL, whereas the percentage of isolates with MICs ⩾2 µg/mL (∼15%) remained small and unchanged. Vancomycin MIC and minimum bactericidal concentration (MBC) for 171 isolates obtained from patients with invasive Staphylococcus aureus infection during 1994–1999, New York. Together, these results support a subtle but significant decrease over time in vancomycin susceptibility among S. aureus isolates. Given the concurrence of this in vitro phenomenon with increasing reports of clinical vancomycin failure [3], we believe that the epidemiologic and potential clinical utility of more-quantitative susceptibility testing for the detection of evolving intermediate-level vancomycin resistance and for the clinical management of MRSA-associated infections merits reevaluation. Potential conflicts of interest K.Y.R. has served as a consultant to Cubist Pharmaceuticals. D.F.G. and M.C.: no conflicts.