Cerebrospinal fluid penetration of fosfomycin in patients with ventriculitis: an observational study

Study design and population

This analysis was reported to the Ethics Committee of the Hamburg Chamber of Physicians (Reference: WF-028/20, February 11, 2020). Due to the non-interventional nature of this study and anonymous recording of data, written informed consent was waived.

Patients included were diagnosed with ventriculitis and treated with fosfomycin by continuous infusion as combination therapy with meropenem and vancomycin as initial empiric therapy. TDM measurements were performed in regular intervals during EVD diagnostics from serum and CSF. Antibiotic dosages were adjusted according to TDM results from serum and CSF samples. Moreover, antibiotic regimens were deescalated, adjusted, or discontinued after species identification or at the discretion of the treating physician.

Diagnosis of ventriculitis was generally applied according to the CDC/NHSN surveillance definition [12]. Since clinical criteria like meningeal or cranial nerve signs could not be obtained in some patients due to sedation, impaired consciousness or interfering neurological deficits, a suspicion for ventriculitis and subsequent treatment indication was solely based on pathological CSF parameters such as increased leucocytes, elevated protein, decreased absolute glucose or decreased CSF/serum glucose ratio also if no growth was seen in the CSF culture in these cases.

Data collection

Demographic and clinical data were obtained from the patients’ electronic records (Integrated Care Manager ICM, version 10.1, Drägerwerk AG, Lübeck, Germany, and Soarian Clinicals 4.01 SP08, Cerner Health Services, Idstein, Germany). We recorded data on antibiotic serum and CSF concentrations. Renal function was determined by the estimated glomerular filtration rate (eGFR) calculated according to the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) [13]. Moreover, clinical routine data such as CSF parameters as well as microbiological results from CSF and blood cultures were obtained. As fosfomycin is available as a disodium salt, serum sodium levels were collected from the blood gas analysis (BGA), as well. The Simplified Acute Physiology Score II (SAPS II) [14] was recorded as a measure of disease severity. Outcome was assessed at discharge according to the Glasgow Outcome Scale (GOS), which consists of the five categories death (1), persistent vegetative state (2), severe (3), moderate (4) and low disability (5) [15].

Drug administration

Fosfomycin (Infectopharm, Heppenheim, Germany) was reconstituted with water and used with a final concentration of 100 mg/mL. It was administered by continuous intravenous infusion (CI) with an initial dose of 1 g/h. Considering a potential disequilibrium effect in CSF two times the minimal inhibitory concentration (MIC) was targeted. Regarding the general European Committee on Antimicrobial Susceptibility Testing (EUCAST) breakpoint for susceptibility (32 mg/L) 64 mg/L was targeted. For Pseudomonas spp. and Enterococcus spp., a target concentration above 128 mg/L was aimed for. This represented a pragmatic approach adapted from the published wild-type MIC distributions although neither an epidemiological cut-off value (ECOFF) nor a breakpoint has been defined in these cases by the EUCAST, yet [16]. Fosfomycin exposure in serum and CSF was optimized by adapting the infusion rate taking into account a penetration rate of approximately 30% [17].

Bioanalytical method

Fosfomycin serum samples were collected and centrifuged (5000 rpm, 10 min, 20 °C), subsequently. The serum supernatant and CSF samples were stored at dry ice until being further processed. Serum samples were analyzed by LC–MS/MS according to the procedure described by Martens-Lobenhoffer et al. [18]. CSF samples were prepared and analyzed in a similar manner. Calibration ranges were 15–750 mg/L in serum as well as in CSF. Coefficients of variation for serum samples were 6.0% at 15 mg/L and 4.0% at 750 mg/L, respectively, with accuracies of 10.9% to − 0.2%. The corresponding coefficients of variation in CSF were 5.6–3.6%, with accuracies of − 13.9–9.7%. The lower limits of quantification were set to the lower end of the calibration range of 15 mg/L for both matrices. Samples with concentrations above the upper limit of quantification (750 mg/L) were diluted with blank serum (or water in case of CSF samples) and were re-analyzed.

Statistics

Data management, non- compartmental calculations for fosfomycin clearance (CL) and area under the curve (AUC) were performed by using Microsoft Excel 365 (Microsoft Corp., Redmond, WA, USA). CL (Eq. 1) and AUC (Eq. 2) were calculated as follows:

$$CL \left[ } \right. \kern-0pt} h}} \right] = \frac}\,}\,\left[ \mathord }} \right. \kern-0pt} }} \right]}}}\,}\,}\,}\,\left[ } \mathord} }}} \right. \kern-0pt} }}} \right]\, * \,24\,}}}\,$$

(1)

$$AUC\left[ $} \!\mathord }}\right.\kern-0pt} \!\lower0.7ex\hbox$}}} \right] = \frac}\,}\left[ }} \mathord}} }}}} \right. \kern-0pt} }}}} \right]}}}\,\left[ } \mathord} }}} \right. \kern-0pt} }}} \right]}}$$

(2)

Visualization, statistical evaluation as well as determination of coefficients of correlation (R2) were either calculated by linear or non-linear regression methods included in Prism 9 (GraphPad Software, San Diego, CA, USA).

留言 (0)

沒有登入
gif