Sir,
Elizabethkingia meningoseptica, formerly Chryseobacterium meningosepticum, is a non-fermentative Gram-negative bacterium and rare cause of nosocomial meningitis in adults.1,2 Selection of appropriate therapy is difficult due to inherent resistance. We report a case of E. meningoseptica meningitis and bacteraemia in which therapeutic drug monitoring (TDM) allowed optimization of pharmacokinetic and pharmacodynamic targets and microbiological cure.
A 73-year-old patient, previously treated with craniofacial irradiation for nasopharyngeal carcinoma, had a progressive decline in cognitive function and magnetic resonance imaging showed radiation-induced encephalomalacia with a new large right frontal cyst with mass effect in the right frontal lobe. The patient was admitted for open fenestration of the intracranial cyst. Dexamethasone was initiated post-craniotomy as standard of care for reduction of swelling.
On post-operative day (POD) 3 the patient developed lethargy, fever and neck stiffness. The patient underwent a lumbar puncture, which showed cerebral spinal fluid (CSF) pleocytosis (1572 cells/mm3) with neutrophilic predominance (87%), increased protein (253 mg/dL) and slight hypoglycorrhachia (54 mg/dL). Gram staining of CSF and blood revealed Gram-negative bacteria. The patient developed septic shock and empirical meropenem was initiated. The use of norepinephrine was required intermittently for 48 h for blood pressure support but the patient never suffered any renal or hepatic impairment.
On POD 6 the organism was identified as E. meningoseptica. Because of clinical deterioration, antibiotics were empirically switched to piperacillin/tazobactam, rifampicin and vancomycin. Blood cultures from POD 6 also grew E. meningoseptica. The organism was susceptible to piperacillin/tazobactam, ciprofloxacin and levofloxacin and resistant to ampicillin, ceftazidime, gentamicin, meropenem and trimethoprim/sulfamethoxazole. On POD 7 ciprofloxacin was added to the regimen and blood cultures were negative. Repeat CSF cultures on POD 8 grew E. meningoseptica.
On POD 9, CSF cultures remained positive and an external ventricular drain (EVD) was placed for elevated intracranial pressures and administration of intraventricular vancomycin. TDM was suggested due to failure to sterilize the CSF and concern regarding antimicrobial penetration. Assays were available in the laboratory of one author (W. H. F.) as part of an NIH-funded study on antimicrobial pharmacodynamics, but they were neither FDA-approved nor performed according to Good Laboratory Practice protocols for clinical use. Informed consent for measuring antimicrobial concentrations in serum and CSF was obtained by one author (W. H. F.). HPLC methods were implemented on an Agilent 1200 series HPLC.3
Piperacillin/tazobactam levels were drawn on POD 10. CSF levels were low and the dose was increased and administered by continuous infusion on POD 11. On POD 13, CSF cultures continued to grow E. meningoseptica and repeat piperacillin/tazobactam levels were again low. To evaluate the intraventricular vancomycin, CSF troughs were measured. The dose was increased on POD 14 to maximize the inhibitory quotient. The first documented negative CSF cultures were obtained on POD 14 and the patient received intraventricular vancomycin for 14 days. On POD 15, ciprofloxacin levels were determined. Due to low CSF levels and inability to meet pharmacokinetic/pharmacodynamic targets, the patient was switched to levofloxacin on POD 17. Levofloxacin levels were determined on POD 20 and higher CSF levels were achieved. TDM results are shown in Table 1.
Table 1.
Pharmacokinetic data
| Free drug concentration (mg/L) |
|||||
|---|---|---|---|---|---|
| Drug (dose) | CSF | serum | CSF:serum ratio, % | MICa (mg/L) | PK–PD parameter |
| Piperacillin/tazobactam (3.375 g iv every 4 h) | 8 | ||||
| piperacillin (3 g every 4 h) | |||||
| 1 h | 1.36 | 96.10 | 1 | T>MIC | |
| 2 h | 1.47 | 63.20 | 2 | 100% serum | |
| 4 hb | 1.24 | 35.48 | 3 | 0% CSF | |
| tazobactam (375 mg every 4 h) | |||||
| 1 h | ND | 4.93 | 0 | ||
| 2 h | ND | 2.50 | 0 | ||
| 4 hb | ND | 2.45 | 0 | ||
| Piperacillin/tazobactam (22.5 g continuous infusion) | 8 | ||||
| piperacillin (20 g continuous infusion) | T > MIC | ||||
| steady state | 2.296 | 99.389 | 2 | 100% serum | |
| tazobactam (2.5 g continuous infusion) | 0% CSF | ||||
| steady state | 0.485 | 9.677 | 0.5 | ||
| Vancomycin (10 mg IVT every 12 h) | |||||
| 11 h | 119.0 | — | — | 16 | IQ 7.4 |
| Vancomycin (15 mg IVT every 12 h) | |||||
| 13 h | 280.4 | — | — | 16 | IQ 17.5 |
| Ciprofloxacin (400 mg iv every 8 h) | ≤0.5 | ||||
| 2 h | 0.318 | 2.314 | 14 | Cmax:MIC | |
| 4 h | 0.339 | 1.345 | 25 | serum 4.6 | |
| 6 h | 0.270 | 1.043 | 25 | CSF 0.80 | |
| 8 h | 0.232 | 0.794 | 29 | ||
| Levofloxacin (500 mg iv every 12 h) | 0.25 | ||||
| 2 h | 3.589 | 6.245 | 57 | Cmax:MIC | |
| 4 h | 3.801 | 5.204 | 73 | serum 25 | |
| 6 h | 3.543 | 4.111 | 86 | CSF 15.2 | |
| 12 hb | 3.448 | 3.252 | 106 | ||
IQ, inhibitory quotient; iv, intravenous; IVT, intraventricular; ND, not detectable; PK–PD parameter, pharmacokinetic–pharmacodynamic parameter; T > MIC, time above MIC.
aMICs of piperacillin/tazobactam and ciprofloxacin were determined by Vitek 2® and MICs of vancomycin and levofloxacin were determined by Etest®.
bTrough level from previous dose.
The patient responded to therapy with slow neurological progress. The EVD was removed on POD 24, the catheter tip was not cultured and CSF cultures from a repeat lumbar puncture on POD 28 demonstrated no growth. The patient was discharged to a long-term care facility on POD 49. The question of a potential abscess in the right frontal lobe complicated treatment and required a prolonged course of levofloxacin (total of 18 weeks). Six months post-hospitalization the patient was home and off antibiotics with significant neurological and radiographic progress.
Given the limited treatment options, the paucity of data from the literature and our patient's lack of response to therapy, we pursued TDM. The use of multiple timepoints to characterize CSF pharmacokinetic/pharmacodynamic parameters provided additional information over traditional serum:CSF ratios.
Despite using a high daily dose of piperacillin/tazobactam, CSF levels in our patient were lower than previously reported.4 Dexamethasone usage or a higher volume of distribution due to septic shock could have contributed. Plasma concentrations exceeded the MIC for the entire dosing interval and cultures sterilized; however, this pharmacokinetic/pharmacodynamic parameter was not achieved, despite dose increase and continuous infusion, in the CSF and cultures remained positive. More data are needed to elucidate piperacillin/tazobactam CSF pharmacokinetics and determine whether they are adequate to treat bacterial meningitis.
Conflicting data exist regarding vancomycin for the treatment of E. meningoseptica meningitis. Case reports have described success, but more recently this treatment has been questioned because of poor in vitro activity.5 Testing from the SENTRY database found that only 4% of isolates were susceptible.6 In our patient, intraventricular therapy was required to reach therapeutic CSF concentrations. TDM was performed to calculate the inhibitory quotient, which was below the goal with the initial dose. Attainment of the goal inhibitory quotient (10–20) led to sterilization of CSF cultures, consistent with the literature.7 To our knowledge this is the first report of intraventricular vancomycin use for E. meningoseptica meningitis in an adult.
The serum:CSF ratio for both fluoroquinolones varied over the dosing interval and CSF peak occurred after the serum peaked. The target Cmax:MIC of >10–12 was achieved only with levofloxacin in both serum and CSF; this may have been due to the presence of transporter proteins.8,9 Although AUC:MIC was not calculated, extrapolating levofloxacin Cmax:MIC in our patient should have achieved the target for this pharmacokinetic/pharmacodynamic parameter.
In conclusion, we report a case of post-neurosurgical E. meningoseptica meningitis in an adult where the use of TDM to optimize pharmacokinetics/pharmacodynamics led to a clinical and microbiological cure.
Funding
This study was carried out as part of our routine work. NIH grant 1R21DK088045 (to W. H. F.) funded the assays used in this paper.
Transparency declarations
None to declare.
References
- 1.Chan KH, Chau PY, Wang RY, et al. Meningitis caused by Flavobacterium meningosepticum after transsphenoidal hypophysectomy with recovery. Surg Neurol. 1983;20:294–6. doi: 10.1016/0090-3019(83)90082-4. [DOI] [PubMed] [Google Scholar]
- 2.Harrington SP, Perlino CA. Flavobacterium meningosepticum sepsis: disease due to bacteria with unusual antibiotic susceptibility. South Med J. 1981;74:764–6. [PubMed] [Google Scholar]
- 3.Connor MJ, Jr, Salem C, Bauer SR, et al. Therapeutic drug monitoring of piperacillin-tazobactam using spent dialysate effluent in patients receiving continuous venovenous hemodialysis. Antimicrob Agents Chemother. 2011;55:557–60. doi: 10.1128/AAC.00548-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Nau R, Kinzig-Schippers M, Sorgel F, et al. Kinetics of piperacillin and tazobactam in ventricular cerebrospinal fluid of hydrocephalic patients. Antimicrob Agents Chemother. 1997;41:987–91. doi: 10.1128/aac.41.5.987. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ozkalay N, Anil M, Agus N, et al. Community-acquired meningitis and sepsis caused by Chryseobacterium meningosepticum in a patient diagnosed with thalassemia major. J Clin Microbiol. 2006;44:3037–9. doi: 10.1128/JCM.00588-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kirby JT, Sader HS, Walsh TR, et al. Antimicrobial susceptibility and epidemiology of a worldwide collection of Chryseobacterium spp.: report from the SENTRY Antimicrobial Surveillance Program (1997–2001) J Clin Microbiol. 2004;42:445–8. doi: 10.1128/JCM.42.1.445-448.2004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Tunkel AR, Hartman BJ, Kaplan SL, et al. Practice guidelines for the management of bacterial meningitis. Clin Infect Dis. 2004;39:1267–84. doi: 10.1086/425368. [DOI] [PubMed] [Google Scholar]
- 8.Maeda T, Takahashi K, Ohtsu N, et al. Identification of influx transporter for the quinolone antibacterial agent levofloxacin. Mol Pharm. 2007;4:85–94. doi: 10.1021/mp060082j. [DOI] [PubMed] [Google Scholar]
- 9.de Lange EC. Potential role of ABC transporters as a detoxification system at the blood-CSF barrier. Adv Drug Deliv Rev. 2004;56:1793–809. doi: 10.1016/j.addr.2004.07.009. [DOI] [PubMed] [Google Scholar]