New Antibiotics for the Treatment of Nosocomial Central Nervous System Infections

Abstract

Nosocomial central nervous system (CNS) infections with carbapenem- and colistin-resistant Gram-negative and vancomycin-resistant Gram-positive bacteria are an increasing therapeutic challenge. Here, we review pharmacokinetic and pharmacodynamic data and clinical experiences with new antibiotics administered intravenously for the treatment of CNS infections by multi-resistant bacteria. Cefiderocol, a new siderophore extended-spectrum cephalosporin, pharmacokinetically behaves similar to established cephalosporins and at high doses will probably be a valuable addition in our therapeutic armamentarium for CNS infections. The new glycopeptides dalbavancin, telavancin, and oritavancin are highly bound to plasma proteins. Although effective in animal models of meningitis, it is unlikely that they reach effective cerebrospinal fluid (CSF) concentrations after intravenous administration alone. The β-lactam/β-lactamase inhibitor combinations have the principal problem that both compounds must achieve adequate CSF concentrations. In the commercially available combinations, the dose of the β-lactamase inhibitor tends to be too low to achieve adequate CSF concentrations. The oxazolidinone tedizolid has a broader spectrum but a less suitable pharmacokinetic profile than linezolid. The halogenated tetracycline eravacycline does not reach CSF concentrations sufficient to treat colistin-resistant Gram-negative bacteria with usual intravenous dosing. Generally, treatment of CNS infections should be intravenous, whenever possible, to avoid adverse effects of intraventricular therapy (IVT). An additional IVT can overcome the limited penetration of many new antibiotics into CSF. It should be considered for patients in which the CNS infection responds poorly to systemic antimicrobial therapy alone.

Keywords: avibactam; cefiderocol; cerebrospinal fluid; dalbavancin; eravacycline; meningitis; relebactam; tazobactam; vaborbactam; ventriculitis.

Figure 1

Schematic drawings of concentration–time curves in serum (red) and cerebrospinal fluid (blue). (A) hydrophilic molecule with a molecular weight of approx. 300–600 g/moL (such as most β-lactam antibiotics and β-lactamase inhibitors). This type of concentration–time curve has been shown for aztreonam, ceftazidime, meropenem, piperacillin, and tazobactam [5,32,33,34,35] and is also expected for cefiderocol, ceftolozane, and the new β-lactamase inhibitors with an AUC_CSF/AUC_S ratio of 0.02–0.2 depending on molecular weight, hydrophilicity, plasma protein binding, and active transport at the blood–CSF and blood–brain barriers [5]. Since they are smaller than the respective β-lactam, the AUC_CSF/AUC_S ratios of β-lactamase inhibitors are expected to be slightly greater than the AUC_CSF/AUC_S ratios of the respective β-lactams (as shown for piperacillin/tazobactam) [34]. Please note the lagging of the concentration–time curve in CSF behind the respective curve in serum. Because retrograde diffusion across the blood–CSF and blood–brain barriers is small for hydrophilic antibiotics, elimination depends on CSF bulk flow (and to a smaller extent on efflux pumps), leading to long elimination half-lives. (B) Moderately lipophilic compound with a molecular weight of approx. 300 g/moL. This type of concentration–time curve represents ideal penetration into CSF (AUC_CSF/AUC_S ratio > 0.8). Linezolid is a prominent example of this relatively rare behavior [31]. Here, the concentration–time curves in serum and CSF almost run in parallel, and the elimination half-lives in serum and CSF are approx. equal. Since tedizolid is highly protein-bound and a ligand of several outward transport systems at the blood–CSF and blood–brain barriers, it is less suitable for the treatment of CNS infections than linezolid.