Neurosurgery Department, University General Hospital of Alicante, Foundation for the Promotion of Health and Biomedical Research in the Valencian Region (FISABIO), Alicante, Spain
Even though meningitis caused by Acinetobacter baumannii is relatively rare, it is associated with high mortality rates especially in neurosurgery patients and represents a serious therapeutic problem due to the limited penetration of effective antibiotics into the cerebrospinal fluid.
In a case report Brunetti et al. identified A. baumanni by MALDI-TOF technique directly from the CSF drawn from the external ventricular drainage of a patient with severe confusional state and signs of meningism. Simultaneously the antibiotic susceptibility test was performed by automated method from the pellet of the broth-enriched sample. The MALDI-TOF technique allowed microbial identification in less than 30 minutes, and the susceptibility test result was available in eight hours, thus allowing a fast diagnosis ready for prompt and targeted antimicrobial therapy 2).
In 2013 a review of the available literature regarding intraventricular (IVT) or intrathecal (ITH) administration of colistin in multidrug-resistant (MDR) and extensively drug-resistant (XDR) A. baumannii ventriculitis/meningitis was conducted and a total of 83 episodes in 81 patients were identified (71 cases in adults and 10 in children and neonates). Colistin was administered via the IVT and ITH route in 52 and 22 cases, respectively, whilst in 7 cases the exact route was not identified. The median dose of local colistin was 125000 IU (10mg) with a range of 20000 IU (1.6 mg) to 500000 IU (40 mg) in adults, whilst a dose of 2000 IU/kg (0.16 mg/kg) up to 125000 IU (10mg) was used in the paediatric population. The median duration of treatment of IVT/ITH polymyxin E was 18.5 days, whilst the median time to achieve sterilisation of cerebrospinal fluid was 4 days. The rate of successful outcome was 89%, and toxicity related to treatment mainly manifested as reversible chemical ventriculitis/meningitis was reported in nine cases (11%). Nowadays, IVT and ITH colistin represents the last resort treatment of MDR and XDR A. baumannii ventriculitis/meningitis, offering a unique, rather safe and successful mode of therapy 3).
Thirty-four patients presented nosocomial meningitis/ventriculitis; 11 (32.5 %) were included in the intravenous colistin (IVC group) and 23 (67.6 %) in the intrathecal/intraventricular colistin (ITC group). The most frequent isolated bacteria were Acinetobacter baumannii. The mean dose was 170,000 (±400) IU and the duration of intraventricular treatment was 16.0 (±8.3) days. The duration of intravenous treatment was 16.0 (±8.3) days in the ITC group and 15.3 ± 7.6 days in IVC group. Hospital mortality was significantly lower in the ITC group compared with the IVC group (13 vs. 72.7 %, p = 0.001).
The combination of intravenous plus intraventricular (IV-IVT) colistin therapy may improve outcomes in patients attending with meningitis/ventriculitis due to multi-drug resistance infections 4).
In an 11-year period, information on 18 consecutive patients with extensively drug-resistant A. baumannii ventriculomeningitis was collected. Infection was defined on the basis of (i) isolation of A. baumannii from the cerebrospinal fluid (CSF); (ii) laboratory evidence of CSF infection; (iii) signs/symptoms of central nervous system (CNS) infection. Patients were divided into group 1 (nine patients, IV colistin alone) and group 2 (nine patients, IV plus IVT colistin).
Cerebrospinal fluid sterilization was documented for 12 of 18 patients (66.6%). The CSF sterilization rate was 33.3% in group 1 and 100% in group 2 (P = 0.009). The mean time to CSF sterilization was 21 days (range 8-48). Five patients died due to A. baumannii CNS infection (all in group 1), and five deaths were unrelated to A. baumannii ventriculomeningitis. Intensive care unit mean length of stay was shorter in group 2 (20.7 vs. 41.6 days, P = 0.046). Crude relative risk ratio of cumulative incidence of persistent CNS infection in group 1 versus group 2 was 13. No cases of chemical meningitis due to intrathecal colistin administration were encountered 5).
Treatment results of six post-neurosurgical ventriculitis and meningitis cases caused by extensively drug-resistant Acinetobacter baumannii after application of an intraventricular loading dose of 500000 IU (40 mg) of colistin followed by a dose of 125000-250000 IU (10-20 mg) every 24-48 h plus parenteral colistin are reported. Simultaneous bacteraemia with an identical Acinetobacter strain was observed in three patients. The mean duration of treatment was 17.2 days (range 15-21 days) and the median time of sterilisation of cerebrospinal fluid was 2.5 days (range 1-5 days). All patients were cured, however one patient presented with chemical meningitis and one with chemical ventriculitis, conditions that clinically and biochemically resemble bacterial meningitis 6).
In a case series of seven Thai patients and 17 patients identified in the literature, clinical and microbiological cure rates with IT/IVT colistin therapy were 83% and 92%, respectively. Three patients (13%) developed chemical ventriculitis and one (4%) experienced treatment-associated seizures. Death was associated with delayed IT/IVT colistin therapy compared to survival (mean time from diagnosis to IT/IVT colistin, 7 vs. 2 days; p 0.01). The only independent predictor of mortality was the severity of illness (APACHE II score > 19, adjusted odds ratio 49.5; 95% CI 1.7-1428.6; p 0.02). This case series suggests that administration of primary or adjunctive IT/IVT colistin therapy was effective for drug-resistant A. baumannii CNS infection 7).
Two patients with multiresistant Acinetobacter baumannii central nervous system infection, successfully treated with either intravenous and/or intraventricular colistin are presented. Unresolved issues such as dose and duration of intraventricular colistin are discussed 9).
Five patients, all were admissions to the neurosurgical ICU and all were cured of their CNS infections. Three cases were complicated by drug-induced aseptic meningitis or ventriculitis.
This largest case series till 2006 shows that direct instillation of colistin into the CNS may cause chemical meningitis or ventriculitis but it is an effective treatment option for MRAB CNS infection. Further study of dosing regimens is needed 10).
Shrestha et al report a case of MDR Acinetobacter ventriculitis treated with intravenous and intraventricular colistin together with intravenous tigecycline. The patient developed nephrotoxicity and poor neurological outcome despite microbiological cure. Careful implementation of bundle of measures to minimize EVD-associated ventriculitis is valuable 11).
Full remission in a patient with catheter-associated ventriculitis due to Acinetobacter baumannii treated with intrathecal and intravenous colistin besides coinfections with other multidrug-resistant bacteria 12).
A case of meningitis due to extensively drug-resistant A baumannii in an Austrian patient who had undergone neurosurgery in northern Italy. The case illustrates the limits of therapeutic options in central nervous system infections caused by extensively drug-resistant pathogens 13).
A Baby PR was delivered vaginally in a district hospital of central India at 32 weeks of gestation following the premature rupture of membranes for more than 72 hours and spontaneous onset of labor. His mother received two doses of betamethasone as well as antibiotics before delivery. He had an Apgar score of 6 at the 1st minute and 9 at the 5th minute and weighed 1.5 kilograms at birth. His initial neonatal course was relatively straightforward with administration of one dose of surfactant, ventilation for 12 hours and subsequent continuous positive airway pressure (CPAP) for one week, and IV amoxicillin and gentamicin for 48 hours, which were stopped as blood cultures were negative. The baby did not require total parenteral nutrition or umbilical catheterizations. His initial head scan on day 2 was normal. Beyond the first week of life, the baby was stable with full enteral feeding and no respiratory support.
The baby had a septic deterioration at the end of the second week with recurrent apneas requiring ventilation. Blood tests revealed an increase in C-reactive protein (CRP: 80 mg/L; normal: <5 mg/L) and white blood cell count (23,000/mm3; normal: 4000–11000/mm3). All of the sepsis screening was performed and treatment was commenced with IV cefotaxime and gentamicin for late-onset sepsis. Blood culture obtained on that day showed the growth of MDRAB. The organism was sensitive to ceftazidime, polymyxin B, trimethoprim, colistin, netilmicin, and amikacin, but was resistant to all other 3rd and 4th generation cephalosporins, other aminoglycosides, quinolones, carbapenems, modified and enhanced penicillins (also with sulbactam), aztreonam, and chloramphenicol. Lumbar CSF examination revealed severe pyogenic meningitis. There was elevated CSF protein level - 1014 mg/dl (normal: 65-150 mg/dl), low CSF glucose – 20 mg/dl (normal: 24-63 mg/dl), and high white blood cell count - 500 cells/mm3 with 100% polymorphs (normal: 0-29 lymphocytes/mm3).
However, CSF obtained was insufficient for culture. The patient was referred to our tertiary neonatal unit on the 18th day of life because of a further increase in CRP (220 mg/L) and the positive blood culture results.
Antibiotic treatment was changed to IV ceftazidime and amikacin as per the blood culture results. The patient showed a clinical improvement over the next 48 hours with successful extubation and decline in his CRP.
Unfortunately, there was further deterioration on the 22nd day of life with bulging anterior fontanel and increasing head circumference. Cranial ultrasonography was suggestive of hydrocephalus and ventriculitis. Ventricular CSF examination on that day showed elevated proteins - 863 mg/dl, low sugar - 13 mg/dl, and a high white cell count - 2,360/mm3 with 90% polymorphs. CSF culture showed the growth of PDRAB that was sensitive only to polymyxin B and netilmicin, but resistant to all other antibiotics as listed above.
Further advice was sought at this point from the microbiology and neurosurgical teams. IV netilmicin (5 mg/kg 12 hourly for 6 weeks) and polymyxin B (20,000 units 12 hourly for 6 weeks) were started on day 26 of life. IVT polymyxin B (40,000 units per dose) was given by alternate day ventricular puncture for four weeks (14 doses) as the family did not consent to the insertion of a ventricular reservoir.
Ventricular CSF examination after four weeks of therapy showed improvement: protein – 124 mg/dl, sugar – 35 mg/dl, and cell counts – 48/ mm3 (mainly lymphocytes). CSF culture was negative. IV therapy was continued for another two weeks.
Computed tomography of the brain performed after the treatment showed massive communicating hydrocephalus but no evidence of ventriculitis. Ventriculoperitoneal shunting was performed four weeks after stopping treatment, and the baby was discharged at the chronologic age of 4.5 months.
On subsequent follow-up at the corrected (for prematurity) age of two years, his head was growing consistently along the 25th percentile on the World Health Organization (WHO) growth chart and weight and height were just below the 10th percentile. His automated brain evoked audiometry was normal. He had a Bayley’s Motor Development Index (MDI) of 80 and a Pervasive Development Index (PDI) of 65.
Bayley Scales of Infant Development-II (1993) assess the attainment of key developmental milestones in children from 1 to 42 months in two main domains: MDI and PDI as above. It is a tester-observed score only and although parental reports can be recorded, they do not contribute to the final scores. Raw scores are adjusted for the chronological age and index scores are obtained. Despite a lack of standardization in premature babies, this test is widely accepted as a reliable measure of development 14).
A 38-year-old, 84-kg Caucasian woman with a recent history of craniotomy was admitted with nausea, fever, headache, photophobia, and drainage from her craniotomy incision. She underwent a repeat craniotomy on hospital day 4 with abscess debridement and repair of a cerebrospinal fluid leak. Cultures grew MDR A. baumannii, coagulase-negative Staphylococcus species, and methicillin-resistant Staphylococcus aureus. Based on the limited published pharmacokinetic and pharmacodynamic data for colistin, we determined a favorable outcome with i.v. colistin monotherapy was unlikely and decided to treat the patient with simultaneous i.v. and intraventricular colistin, as well as intraventricular tobramycin and i.v. rifampin. She was treated with a total of 36 days of intraventricular colistin, 40 days of intraventricular tobramycin, 51 days of i.v. colistin and rifampin, and 56 days i.v. vancomycin for infection that persisted despite multiple debridements. The patient had subsequent improvement in clinical manifestations and eradication of infection. She was subsequently discharged to an acute rehabilitation facility on hospital day 77 with posttreatment sequelae including mental impairment and renal failure requiring hemodialysis. Follow-up visits revealed significant improvement in her mental status, speech, and strength on the side not affected by the stroke.
Prolonged combination therapy with intraventricular colistin and tobramycin plus i.v. colistin, rifampin, and vancomycin led to the resolution of a persistent central nervous system infection caused by MDR A. baumannii 15).
A case of a 42-year-old male patient affected by low-grade ependymoma who developed AB-MDR post-neurosurgical ventriculitis. Initially, because of in vitro susceptibility, De Pascale et al used a combination of intravenous colistin and tigecycline. This treatment resulted in the improvement of the patient's initial condition. However, soon after, the infection relapsed; tigecycline was stopped and treatment with intrathecal colistin was initiated. Cure was achieved by continuing this treatment for approximately three weeks, without adverse effects 16).
A 2-month-old girl with ventriculitis caused by MDRAB is reported. Despite therapy with intravenous (IV) colistin ventricular fluid, cultures remained positive for MDRAB. Institution of combination therapy with IV and intraventricular colistin resulted in a successful clinical and microbiological outcome. Intraventricular/intrathecal and IV colistin might be the best therapeutic option in the treatment of central nervous system infection caused by MDRAB. Further studies are required to evaluate pharmacokinetic and pharmacodynamic parameters of combined IV and intraventricular/intrathecal colistin administration, especially in children 17).
A case of a 40-year old man was admitted to the intensive care unit due to subarachnoid haemorrhage. The patient developed a ventriculitis due to A.baumannii treated successfully with sulbactam IV and intrathecal amikacin 18).