Hospital acquired pneumonia (HAP) is the second-most common nosocomial infection in hospitals and the most common in the intensive care unit1. Mortality for HAP exceeds 20%2. Although mechanical ventilation is the greatest risk factor and increases the risk of dying from HAP, the organisms responsible for causing pneumonia in a given ICU are the same regardless of whether patients are intubated3.
Staphylococcus aureus has become the leading pathogen responsible for hospital-acquired pneumonia (HAP) and is responsible for about 20% of cases4. Staphylococcus aureus may be either sensitive or resistant to oxacillin/methicillin. Oxacillin resistance rates are increasing nationwide and the average rate of resistance to oxacillin currently exceeds 52% nationwide5. While these "MRSA" isolates at UK represent only 47% of all staph aureus isolates, rates of methicillin resistant staphylococcus aureus (MRSA) may approach 80% in some institutions (unpublished observation).
Presence or absence of risk factors for MRSA determines whether or not MRSA drug therapy is used7. If a patient is thought to have pneumonia and they possess risk factors for MRSA, then antimicrobial therapy active against MRSA will be used empirically and "de-escalated" if objective culture data does not demonstrate MRSA. Empiric or culture-based therapy for MRSA HAP includes one of the following: 1) the glycopeptide vancomycin or 2) the oxazoladinone linezolid (Zyvox®). Antimicrobials with activity against MRSA but who do not currently have licensed indication for use in HAP include: 3) the glycylcycline, tigecycline (Tygacil®) and 4) the streptogramin, quinupristin-dalfopristin (Synercid®).
Vancomycin is the primary antimicrobial choice for MRSA HAP worldwide. Vancomycin is relatively inexpensive in per-dose pricing, generally effective and has a favorable safety profile. However, staph aureus isolates which are incompletely sensitive or even resistant to vancomycin have been reported8. These vancomycin-intermediately sensitive staph aureus (VISA) and vancomycin-resistant staph aureus (VRSA) isolates are rare but increasing in prevalence9. Since the first report of VISA in the United States in 1999, other reports have documented the existence of staph aureus isolates with incomplete sensitivity to vancomycin, termed as vancomycin "heteroresistance"9.
Heteroresistant vancomycin-intermediate sensitive staph aureus (hetero-VISA, hVISA, hGISA) isolates are those in which the prevalent colonies are vancomycin susceptible but in which some of the isolates demonstrate phenotypes which are intermediate or resistant to vancomycin9. Hetero-VISA has been proposed as an etiology for vancomycin treatment failure in some patients. These hetero-VISA isolates may exhibit vancomycin-susceptibility when analyzed by automated testing because automated systems are incapable of identifying rare hetero-VISA colony forming units (CFU's) within single large inoculums10. Without plating and more detailed testing, the true incidence of hetero-VISA isolates may be underestimated. No large direct comparison of automated and non-automated MRSA isolate sensitivities to vancomycin has been reported. This investigation seeks to determine the prevalence of hetero-VISA in sputum isolates of MRSA in a major tertiary referral center and define the sensitivity of automated testing.
Adequate empiric antibiotic therapy improves outcome in patients with pneumonia11-13. Conversely, inadequate empiric therapy worsens outcome compared to adequate therapy. Clinicians should therefore seek to deliver adequate empiric antibiotics at the time of clinical diagnosis of pneumonia and not wait until definitive cultures are obtained14. This type of antibiotic approach which starts broad and then reduces therapy based upon objective culture data is termed "de-escalation" therapy. De-escalation has led to reduced antibiotic duration, less antibiotic administered, fewer occurrences of second pneumonias, and less mortality15. The goal of pneumonia therapy is to start adequately and to narrow objectively and as soon as appropriate.
Since adequate empiric therapy determines outcome, it is imperative that clinicians treating patients at risk for MRSA with antimicrobials certain to eradicate MRSA. Traditional therapy was almost exclusively the glycopeptide vancomycin but emerging evidence suggests that in some cases vancomycin is not as effective as other available drugs. New antimicrobials have emerged which may exhibit superior efficacy against MRSA. In 2004, Marin Kollef et al reported improved clinical cure and survival in patients with MRSA ventilator-associated pneumonia (VAP) who were treated with linezolid compared to those treated with vancomycin (survival odds ratio 4.6 for linezolid vs vancomycin)16.
While the prevalence of MRSA among all Staph aureus isolates is increasing, in vitro evidence of growing resistance of MRSA isolates to vancomycin is also increasing. In vitro susceptibility testing for any organism and antimicrobial is expressed as minimal inhibitory concentration (MIC). MIC is the lowest concentration of an antimicrobial that will inhibit growth of an organism. That is, the higher the MIC for a given antimicrobial and organism, the greater concentration of drug required to produce killing. Vancomycin MIC's for MRSA are rising, a phenomenon referred to as "MIC creep"18. Wang et al reported mean vancomycin MIC's for MRSA at UCLA to be ? 0.5ug/ml in 2000 and 1mg/dl in 2004 and the prevalence of isolates with MIC=1ug/ml to be 70% in 2004 vs 20% in 2000. MIC's for a given isolate in a given patient may change over time as well. Webster et al reported a MRSA isolate with an MIC changing from ? 1 to 4ug/ml over several months, demonstrating the ability of MRSA to acquire or alter its resistance patterns19. S. aureus isolates for which vancomycin MICs are 2-8 ?g/mL are classified as vancomycin-intermediate, and isolates for which vancomycin MICs are ?16 ?g/mL are classified as vancomycin-resistant by the Centers for Disease Control20. Clinical outcome with vancomycin therapy now appears to correlate inversely with MIC's above 0.5ug/ml, so that as MIC's increase, clinical outcome worsens, suggesting that although MIC's ? 2ug/ml are technically reported as "Susceptible", the clinical efficacy of vancomycin may vary between isolates with MIC's of 0.5ug/ml and isolates with MIC's of 1 or 2ug/ml.
A confounding factor in the determination of MIC's for vancomycin or other antimicrobials is the variability with which MIC's are measured. MIC's may be obtained by automated equipment or by direct visual assessment of killing by antimicrobial disks or strips on a culture plate. Automated equipment historically does not report specific vancomycin MIC's below or equal to 2ug/ml. The limitation with such equipment is that as vancomycin MIC's for MRSA have crept upward over time, we now have automated equipment that reports all MRSA isolates with vancomycin MIC's ? 2 as susceptible-with no distinction of whether the MIC is 0.5, 1 or 2ug/ml. Automated equipment generally reports MIC's in doubling of antimicrobial concentration. Furthermore, while the majority of the sample may contain isolates with vancomycin MIC's in the susceptible range, we now know that susceptibility may vary within a given sample. Some solitary organisms or colony-forming units (CFU's) may demonstrate a phenotype in which vancomycin MIC's are higher than the majority of the sample but their numbers are so small as to be unrecognized. This variability in MRSA resistance to vancomycin is termed heteroresistance and heteroresistance may contribute to treatment failure. Moreover, ideal methods to detect and address vancomycin heteroresistance in S. aureus (hVISA) in every institution or laboratory have yet to be established.
An alternative to automated testing is the use of E-test strips (AB Biodisk). E-tests are strips with increasing concentration of the particular antimicrobial agent from one pole to another which when applied to a plate of bacterial isolates produces a tear-drop shaped area of killing, with greater killing toward the end with the greater concentration (Figure). Though more cumbersome and expensive than automated testing, an accurate assessment of the presence or absence of heteroresistance and the more precise determination of MIC's are possible with E-tests compared to automated testing. Determination of prevalence of MRSA heteroresistance to vancomycin and relationship between E-test and automated vancomycin MIC's for MRSA are primary objectives of this investigation.
When combined with clinical cure data above, this in vitro evidence strongly questions the prudence of using vancomycin as the empiric antimicrobial agent in patients known to be at risk for MRSA, who may be critically ill with VAP and for whom a wealth of studies establish that failure to adequately treat empirically increases mortality. Determining the prevalence of MRSA isolates with marginal susceptibility to vancomycin or heteroresistance is the primary aim of this investigation. An accurate measure of MIC reproducibility between techniques and the current prevalence of MIC's which are only marginally susceptible will allow our institution and the medical community to determine the most appropriate role for vancomycin in the empiric antimicrobial regimen for patients thought to have VAP who are at risk for MRSA.