Appropriate Antimicrobial Therapy for Community-Acquired Methicillin-Resistant Staphylococcus aureus Carrying the Panton-Valentine Leukocidin Genes

SIR—We have reviewed with great interest the article by Francis et al. [1], which describes 4 cases of pneumonia due to community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) carrying the Panton-Valentine leukocidin (PVL) genes. Although the discovery of this virulence factor is of interest from a research perspective, it is uncertain what impact the discovery may have on the determiniation of appropriate antibiotic therapy for CA-MRSA. These questions remain: should we routinely test for the presence of the PVL genes in severe cases of CA-MRSA, and, if so, would the presence of this gene alter decisions about antimicrobial management of the infection? In an attempt to answer the questions posed, we searched the literature to identify and evaluate other case reports about pneumonia due to PVL-secreting CA-MRSA (table 1) [2–4]. of the 25 cases reviewed, we determined, on the basis of the susceptibility profile of the S. aureus isolates that were cultured, that patients in 20 of the cases received appropriate antimicrobial therapy. Despite receipt of “appropriate” therapy, 14 of the 20 patients died within days after admission to a hospital. It is interesting that, of the patients who died while receiving appropriate therapy, not one had any underlying disorders, indicating that this organism is, in fact, a highly virulent pathogen. The question remains whether vancomycin, because of its pharmacodynamic properties, would serve as appropriate empirical therapy for this patient population. Characteristics of case reports of Panton-Valentine leukocidin–secreting community acquired methicillin-resistant Staphylococcus aureus pneumonia. It is well documented that the penetration of vancomycin in lung tissue is poor because the drug is a highly polar compound with increased molecular weight [5]. Therefore, vancomycin at a standard dosage of 1000 mg intravenous every 12 h cannot maintain a concentration above the MIC90 of MRSA (i.e., 4 µg/mL) in the lung epithelium of a patient with normal renal function [5]. Additionally, <20% of the plasma concentration is actually found in the epithelial lining fluid (ELF), leading to findings that at least 40% of patients who are treated with vancomycin do not achieve adequate concentrations [6]. We postulate that patients who received appropriate therapy with vancomycin for the treatment of CA-MRSA pneumonia may not have received a high enough dose to penetrate their already necrotic lung tissue and to overcome the large inocula of highly virulent MRSA present. When determining the dosages for the majority of patients who receive vancomycin, the pharmacokinetics service at our institution (Huntsville Hospital, Huntsville, AL) uses the Matzke nomogram [7], which adjusts dosages to 19 mg per kilogram of actual body weight given at intervals of 2 times the elimination half-life. Furthermore, if warranted in the case of severe pneumonia, a loading dose of 25 mg per kilogram of actual body weight may be administered. The goal of using this nomogram is to achieve a serum trough level of 15–20 mg/dL, rather than 5–10 mg/dL, to achieve an plasma concentration of ∼3–4 mg/dL in the ELF of the lungs. Additionally, a study by Zimmermann et al. [8] that included 273 patients showed that nephrotoxicity did not occur if trough levels were maintained at <20 mg/L. Moreover, patients with vancomycin trough concentrations of >10mg/L were more likely to become afebrile and to achieve a normalized WBC count within 72 h. In contrast to vancomycin, the oxazolidinone derivative linezolid concentrates in lung ELF, maintaining a concentration greater than the MIC90 of MRSA throughout the dosing interval [9]. Furthermore, retrospective analyses have shown linezolid to be significantly more efficacious against MRSA pneumonia than is vancomycin [10]. Therefore, when faced with a severe case of pneumonia caused by PVL-secreting MRSA, linezolid may be a more appropriate antimicrobial agent. The decision to use clindamycin in the treatment of PVL-secreting MRSA pneumonia also warrants discussion. In the case report by Francis et al. [1], patient 4 received clindamycin in addition to vancomycin therapy on the basis of the susceptibility profile, and the treatment was later changed to vancomycin, linezolid, and rifampin. The susceptibility profile of all isolates tested, including those from patient 4, demonstrated erythromycin resistance and clindamycin susceptibility, leading to the question of inducible resistance. A report by Siberry et al. [11] revealed that 51% of MRSA isolates that were from adults and that were shown to be erythromycin resistant and clindamycin susceptible had inducible macrolide-lincosamide-streptogramin B (MLSB) resistance because of the presence of erythromycin ribosomal methylase (erm) genes. A later report by Ellis et al. [12] indicated that 20% of the isolates studied had inducible MLSB resistance. Therefore, when a susceptibility profile such as that described above is reported, it is recommended to conduct the double-disk diffusion test to lessen the chances of a therapeutic failure. It has yet to be decided whether all cases of CA-MRSA pneumonia should be tested for the presence of PVL genes; however, it is imperative to recognize that PVL-secreting MRSA is a highly virulent strain of MRSA that results in significant morbidity and mortality. Early and appropriate empirical therapy should be initiated in an attempt to thwart the fatal process that may ensue. Unfortunately, the risk factors for the presence of this pathogenic organism have yet to be determined. Therefore, starting empirical treatment with broad-spectrum antimicrobials and de-escalating the therapy after the pathogen is identified and susceptibility profiles are available are justified when there is a suspicion of S. aureus carrying the PVL genes. Potential conflicts of interest. K.A.W. and E.H.E: no conflicts.