Methicillin-resistant Staphylococcus aureus meningitis:: Has the time come for an alternative to vancomycin?

We read with great interest the paper by Chang et al. [1] which compared clinical features and outcome of meningitis due to methicillin-resistant (MRSA) and methicillin-sensitive Staphylococcus aureus (MSSA) in adults. The authors highlighted the progressively increasing incidence of MRSA in adult staphylococcal meningitis, together with the higher morbidity and mortality rates among patients with such an infection, compared to those with MSSA meningitis. In addition, the authors suggest that these different characteristics reflect both different underlying diseases (i.e. postneurosurgical status in MRSA cases and medical disorders in MSSA cases) and different sources of infection: hematogenous, community-acquired in MSSA cases and nosocomial in MRSA ones. In our opinion, there are sonic issues of the paper that should be discussed: extensive surveillance programs have outlined the rapid spread of MRSA worldwide, especially in nosocomial settings [2]. This threat is of particular concern among neurosurgical patients with meningitis who have undergone insertion of intracranial devices [3]. Also, data on the use of vancomycin in the treatment of CNS infections due to MRSA are lacking [1]. Previous experience indeed indicates that it can have an unpredictable CNS penetration when delivered intravenously, despite adequate plasma levels [4], especially in patients under concomitant dexamethasone medication [5]. In addition, potential toxicities and untoward effects, such as seizures, can limit its intrathecal instillation [4]. Moreover, it has been demonstrated that vancomycin ceases to be bactericidal in the presence of a high bacterial inoculum [6]. The results reported by Chang et al. [1] seem to confirm this insufficient therapeutic performance of vancomycin in the MRSA cases; indeed, 46% of the patients died and most of the survivors suffered from sequelae. At the beginning of the 21st century, these numbers should be clinically unacceptable. In this respect, new classes of antimicrobials, such as streptogramins and oxazolidinones, might offer effective alternatives against multidrug-resistant gram-positive cocci. Intravenous quinupristin/dalfopristin has recently been demonstrated ineffective in such CNS infections [7] and should be delivered intrathecally [8]. in contrast, preliminary studies with linezolid demonstrated a potential usefulness of this antimicrobial in such a difficult setting [9, 10]. As part of an expanded access program, we also experienced the promising results of linezolid in five patients with postneurosurgical CNS infections due to gram-positive strains [11]. In all five patients this oxazolidinone effected cure without substantial treatment-related side effects. Moreover, concomitant pharmacodynamic investigations showed a very good CNS penetration of the drug, with CSF concentrations always higher than plasma concentrations [12], thus also allowing a switch from intravenous to oral administration within a few days. Therefore, we think that although a severe antibiotic policy should always be kept in mind [6,13], it could also be the time to rethink the cost-benefit ratio of different therapeutic approaches, at least in some restricted, nosocomial and often life-threatening situations.