Management of Meningitis Caused by Multi Drug-Resistant Acinetobacter Baumannii: Clinical, Microbiological and Pharmacokinetic Results in a Patient Treated with Colistin Methanesulfonate

This paper reports on a 71- year-old Caucasian male who underwent neurosurgery for an oligodendroglioma, followed by a cranial-sinus fistula and cerebrospinal fluid rhinorrhea. The clinical course was complicated due to an extensively drug-resistant Acinetobacter baumannii meningitis. The patient was treated with colistin methanesulfonate, intrathecal for 24 days and intravenous for 46 days. In addition, the patient received meropenem and teicoplanin to treat a urinary tract infection and a bacterial aspiration pneumonia. Cerebrospinal fluid trough colistin levels resulted above the MIC of A. baumannii. Colistin cerebrospinal fluid concentration did not increase over the treatment period. Meningitis was cured and A. baumannii eradicated. No side effects from the antimicrobial therapy were observed. In conclusion, this case highlights the issues in treating infections caused by resistant Gram negative bacteria and supports previous findings on the efficacy, pharmacokinetic and tolerability of intravenous and intrathecal colistin treatments.


Introduction. Over the last decade, extensively drugresistant
Gram-negative bacteria, including Acinetobacter baumannii, have become a serious cause of hospital-acquired infections. A. baumannii has also emerged as a cause of central nervous system (CNS) infections, which are often associated with the use of external cerebrospinal fluid (CSF) catheters. [1][2][3] The treatment of these infections can often be extremely complex due to antimicrobial resistance and the inadequate antimicrobial concentration at the site of infection. [1][2][3][4][5][6] The increased rate of infections due to multi drug-resistant Gram-negative bacteria has been reported to lead to a revival in the use of "forgotten" antibiotics, such as colistin. 4,5 Colistin, a polymyxin antibiotic that is administered intravenously (IV) as colistin methanesulfonate, is a prodrug that is converted in vivo and in vitro into its active form colistin. 4,6 Colistin methanesulfonate and colistin poorly cross the brain blood barrier, 4,6 thus in order to treat CNS infections, colistin methanesulfonate needs to be administered either intrathecally (IT) or intraventricularly (IVT). 3,6 The authors report on a case of meningitis caused by multi-drug resistant A. baumannii treated with IV and IT colistin.
Case Report. A 71-year-old Caucasian male was admitted to our hospital with cerebrospinal fluid rhinorrhea one month after having undergone neurosurgery for an oligodendroglioma. Ten days after admission (Table 1), the patient manifested acute meningitis caused by methicillin-resistant Staphylococcus aureus (MRSA) and Corynebacterium striatum, treated with vancomycin IV 750 mg TID and imipenem IV 500 mg QD. At the same time, the cranial-sinus fistula was repaired. The clinical course was complicated by pneumonia and acute respiratory insufficiency requiring assisted mechanical ventilation ( Table 1). A week later, the patient was extubated and re-admitted to the floor. The following day, the patient manifested a low-grade fever and blood tests evidenced increased leukocyte and neutrophil counts; whereas the C-reactive protein (C-RP) and erythrocyte sedimentation rate (ESR) values were 15.4 mg/dL (normal <0.5mg/dL) and 8 mm 1st h (normal 1-30), respectively. Due to a persistent drowsiness and a suspected hydrocephalus, an external CSF lumbar catheter was positioned. The CSF from the catheter resulted having normal cell and glucose values, and the microbiological investigations were negative. Additionally, Enterobacter cloacae urinary infection was treated with meropenem ( Table 1). Five days later, the patient's temperature rose to 38.8°C, the leukocyte, neutrophil, C-RP and ERS values also increased, the patient manifested a more depressed level of consciousness and the patient complained of neck stiffness. Simultaneous CSF findings from the lumbar catheter were consistent with acute Gram-negative bacterial meningitis. 7 In addition, the SeptiFast realtime PCR (SF) (Roche Diagnostics, Monza, Italy) 8 performed on the CSF sample from the lumbar catheter resulted positive for A. baumannii and K. pneumoniae, while the CSF mass spectrometry by matrix-assisted laser desorption/ionization time-of-light (MALDI-TOF) (Bruker Daltonics, Bremen, Germany) 9 was negative. CSF culture yielded A. baumannii and a few colonies of K. pneumoniae, both susceptible only to colistin. Antimicrobial therapy was administrated : intravenous colistin methanesulfonate 4.500.000 International Unit (IU) (equal to 150 mg of colistin based activity) BID (infused over 30'), meropenem 2 g TID, rifampin 600 mg OD and teicoplanin 600 mg OD after the loading dose plus colistin methanesulfonate IT 125.000 IU (equal to 4.16 mg of colistin based activity) a day. 3,10 Rifampin had to be discontinued soon after due to an allergic reaction. Two days later, after three doses of IV colistin and a single dose of IT colistin, a repeated culture of CSF, from both the lumbar catheter and rachicentesis, evidenced A. baumannii. Whenever IT colistin was administered (range of time ± 4h), the catheter was kept closed for 3 hours after. CSF samples for laboratory investigations and concentrations were collected from the lumbar catheter before colistin was administered. Colistin concentrations were evaluated on samples (stored at -20°C until testing) using an HPLC method having fluorimetric detection and netilmicin as an internal standard. Linear calibration curves were obtained by the concentrations of colistin sulfate from 0.30 to 5.0 mg/L in plasma. 11 On day four of therapy, the patient was without fever, CSF cell count was decreased, and the culture resulted negative. After a total of 24 days of therapy, the lumbar catheter was removed, while a lumbar-peritoneal catheter was positioned to treat a hydrocephalus that had developed. Results of CSF findings are reported in Table 2. IT colistin was discontinued while IV colistin, meropenem, and teicoplanin were continued for a further 22 days, followed by meropenem 3 g and oral doxycycline 200 mg per day for another 11 days. During this period, the patient was without fever but multiple episodes of acute respiratory insufficiency occurred, along with alternatively reduced or increased neutrophils values, C-RP values, and lung infiltrates. Repeated bronchoscopic aspirations were performed, and a percutaneous endoscopic gastrostomy (PEG) was positioned ( Table 1). Repeated respiratory secretion cultures evidenced MRSA and K. pneumoniae resistant to colistin, but fosfomycin susceptible ( Table 1). Despite fosfomycin therapy, the patient had a fatal episode of acute respiratory insufficiency leading to his death.
Discussion. Over the last decade, the frequency of CNS infections caused by Gram-negative bacteria has increased from 12-27% of cases, 1-3 as well as meningitis caused by A. baumannii.
In our patient, clinical and microbiological findings supported a diagnosis of hospital acquired A. baumannii meningitis. 7 In fact, 1) A. baumannii was detected by culture and SF in CSF samples obtained from both rachicentesis and the lumbar catheter on the third day of treatment; 2) A. baumannii DNA was detected by SF (data not shown) in the CSF from day 12 of treatment; 3) airways were colonized/infected with K. pneumoniae, leading us to deduce that the CSF could have been contaminated with this microorganism during collection. 3,7 To this regard, it is important to report that K. pneumoniae was cultured with A. baumannii from the CSF taken on both day one when the patient had acute bacterial meningitis and on day 17 of treatment when there was clinical improvement and CFS laboratory parameters resulted normalized. Pseudomonas aeruginosa detected from the lumbar   catheter was considered of no clinical relevance, given an absence of symptoms and CFS abnormalities. 7 Culture results, SF, and MALDI-TOF tests were performed. Microbial culture is still considered the reference method for infection diagnosis. SF test is a molecular based method used to detect bacteria from blood, but it has also been applied to samples different from blood. 8 One of the advantages of SF over culture is that the results of this test can be available in less than 6 h, allowing for a prompt and more accurate empiric therapy. Moreover, SF has a high sensitivity for identifying microbial DNA in patients receiving antimicrobial therapy. 8 However, the clinical significance of blood microbial DNA, even when patients are septic in the absence of microorganism growth, is not well defined. 12 As well, there are limited data on the reliability of performing SF on other biological samples. 8 MALDI-TOF is considered a reliable, rapid method for identifying bacterial strains from colonies on solid culture media 9 and has also been employed to analyze clinical specimens such as urine and CSF for direct bacterial identification. Nevertheless, for our patient, the best results were obtained when the bacterial concentration in the sample was ≥105CFU/mL. 9 Colistin is defined as a concentration dependent bactericidal antibiotic, therefore, according to Hara GH et al., peak levels seem to be more predictive of clinical efficacy. 4 In vitro and in vivo animal studies suggest that the area under the curve AUC/MIC and Cmax/MIC ratio is the best predictor of antibacterial activity. However, the pharmacodynamic parameters that best predict efficacy are not well defined. 5,6 Overall, it has been suggested to maintain steady state levels ≥2 mg/L for effective therapy. 5,6 Considering the in vitro antimicrobial susceptibility of A. baumannii isolate and the poor capacity of colistin methanesulfonate to cross the blood brain barrier, colistin methanesulfonate was administered both intravenously and intrathecally without a loading dose. 3,10 Overall, the treatment resulted being both effective and well tolerated.
In our patient, only trough CSF values were obtained, and on 3 different days values below 2 mg/L were observed. The ratio between CSF concentration and A. baumannii MIC ranged between 2 and 70. This broad range could have been due to the different collection times of CSF and/or CSF efflux fluctuations through the external drainage. 6 Overall, the CSF colistin concentration did not increase over time, mirroring results by Imberti et. al. 6 Regarding the Colistin blood levels without a loading dose, values above ≥2 mg/L were registered on day 5 of therapy and a lower colistin concentration in the respiratory secretions most likely favored the selection of Colistin hetero-resistant K. pneumoniae isolates.
Considering in vitro susceptibility results of K. pneumoniae isolates to fosfomycin, it is plausible that the variable results reported from our laboratory were due to MIC being close to the susceptibility break point. When these K. pneumoniae isolates were evaluated at a reference laboratory they were reported as susceptible to fosfomycin, suggesting that laboratory fluctuation could have induced variable susceptibility results. Furthermore, PFGE analysis of these isolates showed an analogous pattern which was similar to the the international blaKPC-3-positive ST258b hybrid clone (data not shown).
It has been suggested that CSF catheters need to be removed in order to achieve recovery from a CNS infection. However, the exact time of CSF catheter removal has yet to be clearly defined. 3,10 In our case, the infection was controlled, and CSF cultures were negative after 4 days of treatment, thus, we decided to keep the external lumbar catheter in place. It was removed only after the meningitis was cured, and a lumbar-peritoneal derivation could be placed without a high risk of relapse.
In conclusion, this case highlights the issues involved in treating infections caused by drug-resistant Gram-negative bacteria and supports previous findings on the efficacy, pharmacokinetics and tolerability of intravenous and intrathecal Colistin treatments.