|Home | About | Journals | Submit | Contact Us | Français|
In a recent landmark trial of bacteremia caused by methicillin-resistant Staphylococcus aureus (MRSA) isolates, vancomycin MICs were ≥1 μg/ml for only 16% of the isolates, and accessory gene regulator (agr) function as measured by delta-hemolysin activity was absent or reduced in only 28.1% of the isolates. This clinical study did not capture a population of MRSA isolates predictive of vancomycin treatment failure.
Methicillin-resistant Staphylococcus aureus (MRSA) isolates with higher vancomycin MICs (>1 μg/ml), reduced vancomycin bactericidal activity in vitro (5, 13), and attenuated agr function (1) have been associated with reduced vancomycin efficacy in bacteremia. Clinical studies targeting bacteremia caused by MRSA with these properties would be important to define the practical role of newer antimicrobials. We evaluated the properties of MRSA isolates from a recent study demonstrating the noninferiority of daptomycin compared to standard therapy in cases of S. aureus bacteremia and endocarditis (2).
Eighty-nine baseline MRSA isolates from the previously published study comparing standard therapy (vancomycin plus low-dose gentamicin for the first 4 days) to daptomycin in MRSA bacteremia and endocarditis were analyzed (87 of 89 isolates were analyzed genotypically) (2).
Vancomycin susceptibility testing was confirmed by CLSI broth microdilution methods. Susceptibility to other antimicrobials was performed by Cubist Pharmaceuticals as part of the prior study (2). agr function was determined semiquantitatively using delta-hemolysin activity as previously described (12). The genotype and staphylococcal cassette chromosome mec (SCCmec) type was determined by previously published PCR methods (3, 6, 11). spa X-repeat polymorphisms were determined by nucleotide sequencing as described previously (7, 11). spa types were compared with results from previous studies (11) and with the Ridom spa server (4) to provisionally assign isolates to multilocus sequence typing-defined clonal complexes (CCs). Clonal complex 1 (CC1) equaled spa types 131 and 950; CC5 equaled spa types 2, 12, 23, 47, 65, 203, 302, 385, 402, and 437; CC8 equaled spa types 1, 4, 7, 363, and 954; CC30 equaled spa types 16 and 19; CC45 equaled spa types 15, 756, 951, and 953; CC59 spa equaled types 17 and 206. Note that four new spa types are included above as follows: spa types 950 (UJJJFFKBPE), 951 (XE3BMBKB), 953 (XE3BBKB), and 954 (YHGCO). Clinical outcomes were defined as described in the initial published study (2).
The vancomycin MICs for the 89 MRSA isolates were as follows: 0.25 μg/ml (n = 1 [1%]), 0.5 μg/ml (n = 74 [83%]), and 1 μg/ml (n = 14 [16%]). There was 100% concordance (defined as within 1 dilution tube) between the MICs performed for the purposes of this study and the Cubist Central Microbiology Laboratory for the original clinical study. Semiqualitative assessment of agr function demonstrated that 28% of isolates had a reduced or absent delta-hemolysin phenotype. The distribution of agr function between the two treatment arms is shown in Table Table11.
MRSA genotyping results between the two treatment arms are shown in Table Table1.1. Interestingly, over half of the isolates were SCCmec type IV, which is associated with community-acquired MRSA. This result was investigated further by analyzing the susceptibility data of other antibiotics and stratifying the results by SCCmec type. MRSA isolates with SCCmec type IV were more likely to be susceptible to clindamycin and levofloxacin and less likely to be susceptible to tetracycline than strains with SCCmec type II. Of the MRSA isolates with SCCmec II, 7.7% had agr group I genotype, 89.7% had agr group II genotype, and 2.6% had agr group III genotype; in contrast, MRSA isolates with SCCmec IV, 87.2% had agr group I genotype, 4.3% had agr group II genotype, and 8.5% had agr group III genotype. This association is consistent with prior data (10).
Clinical response to therapy with either daptomycin or vancomycin evaluated by SCCmec type, agr genotype, and agr function is listed in Table Table22.
At this time, the epidemiology of MRSA is in flux, with “health care-associated” MRSA marked by increases in vancomycin MICs and the agr group II genotype that have been linked with decreased vancomycin efficacy on the one hand, and “community-associated” MRSA with lower vancomycin MICs but linked with rapidly invasive necrotizing infections and a high severity of illness on the other hand (10, 16), with the boundaries between these becoming less discriminatory. The MRSA isolates from this study were heavily weighted (84%) toward organisms with vancomycin MICs of <1 μg/ml, leaving too small a subset with higher vancomycin MICs to evaluate. However, this observation brought to light the fact that the study comparing the efficacy of daptomycin to vancomycin plus an initial 4 days of low-dose gentamicin did not capture the organisms required to address the pressing problem of optimal antimicrobial therapy for MRSA bacteremia caused by organisms with vancomycin of 2 μg/ml. Only 25 of 89 (28%) of the isolates had significantly reduced or absent delta-hemolysin activity, in contrast to prior evaluations of MRSA bacteremia isolates where 75% of 81 isolates had this phenotype (14). This finding, in addition to the results of the vancomycin MICs, suggested that “community-acquired” MRSA isolates represented a significant proportion of this group of isolates. Such strains represent an increasing cause of S. aureus bacteremia and endocarditis.
In this trial, 44 of 87 (51%) of MRSA isolates belonged to agr group I and 38 of 87 (44%) belonged to agr group II, in contrast to prior data showing that greater than 60% of MRSA bacteremia isolates collected from multiple centers around the United States belonged to agr group II (9). Extrapolation from spa typing revealed that 33 (38%) of the isolates belonged to CC8, the genetic background of the well-recognized USA300 community MRSA pulsotype (14). Consistent with this was the finding that 87% of agr group I strains harbored SCCmec type IV, another feature of community MRSA, while 90% of agr group II strains harbored SCCmec type II. “Community-acquired SCCmec type IV MRSA” as causes of hospital-acquired infections increased from 17% in 1999 to 56% in 2003 at one center, suggesting that this epidemiological shift in MRSA had been well under way during the enrollment period of this trial (8).
The shift in MRSA epidemiology toward “community-acquired” clones may render the evaluation of infections caused by strains with microbiological features that are currently predictive of vancomycin treatment failure (agr group II genotype, reduced agr function, and vancomycin MIC of >1 μg/ml) more difficult. Such investigations would need to enrich their samples for isolates with higher vancomycin MICs or enroll a higher number of patients in order to capture subsets large enough to get statistically significant data, which may be prohibitively expensive and lengthy. It is important to appreciate that features predictive of treatment failure are dynamic and therefore may change temporally and geographically with shifts in molecular epidemiology and differences in antibiotic selection pressure over time and in different hospital centers. If the population studied in this trial is representative of the current predominant S. aureus strains, additional studies are warranted to determine the microbiological features associated with treatment failure in bacteremia caused by community-acquired S. aureus. In this study, our attempts to show superiority of daptomycin over vancomycin for MRSA bacteremia caused by organisms within specific agr group genotype subsets did not achieve statistical significance.
In summary, MRSA isolates with SCCmec type IV have emerged to become significantly represented in MRSA bacteremia, resulting in the trial comparing daptomycin with vancomycin plus gentamicin not capturing isolates with vancomycin MICs of 2 mg/ml, recently associated with higher mortality when vancomycin was used empirically for bacteremia (15). Additional studies would be needed to define optimal therapy in cases of MRSA bacteremia where the vancomycin MIC is >1 mg/ml, the genotype is agr group II, and agr function is reduced or absent. Given the poor performance of vancomycin in MRSA bacteremia in such cases, the need for the timely completion of such investigations to keep up with the rapid evolution of MRSA is of utmost importance. Until this happens, clinicians are encouraged to quantitate vancomycin MICs in their local hospitals and for individual patients with serious MRSA infections.
This study was funded through a research grant from Cubist Pharmaceuticals (Lexington, MA). D.A.R. is supported in part by grants from the American Heart Association and the National Institute of General Medical Sciences (GM080602). J.S., S.A.L., and P.A.M. are employed by Cubist Pharmaceuticals. G.S. has received research funding from Cubist and Pfizer Pharmaceuticals and honoraria from Cubist, Pfizer, and Wyeth Pharmaceuticals.
Published ahead of print on 7 July 2008.