Despite the fact that the first IVT administration

Transcription

1 J Neurosurg 119: , 2013 AANS, 2013 Intraventricular and lumbar intrathecal administration of antibiotics in postneurosurgical patients with meningitis and/or ventriculitis in a serious clinical state Clinical article František Remeš, M.D., 1 Robert Tomáš, M.D., Ph.D., 1 Vlastimil Jindrák, M.D., 2 Václav Vaniš, M.D., 2 and Michal Šetlík, M.D. 1 Departments of 1 Neurosurgery and 2 Microbiology, Na Homolce Hospital, Prague, Czech Republic Object. To date, reports on the clinical efficacy of intraventricularly and intrathecally administered antibiotics for the treatment of neurosurgical ventriculitis and meningitis in adults are limited. The authors aimed to evaluate the efficacy and safety of the intraventricular (IVT) and lumbar intrathecal (IT) administration of antibiotics in critically ill neurosurgical patients. Methods. Thirty-four postneurosurgical patients with meningitis and ventriculitis were studied. Intraventricular/ lumbar intrathecal antibiotics were administered due to positive CSF cultures persisting despite the use of intravenous antibiotics. The time period until CSF sterilization, changes in clinical state, and efficacy of different routes of antibiotic administration were evaluated. Results. The mean time necessary to obtain CSF sterilization was 2.9 ± 2.7 days (range 1 12 days). The CSF cultures became negative within 24 hours after the administration of antibiotics in 17 patients (50%) and up to 48 hours in a further 6 patients (18%). The clinical outcome of patients assessed by the modified Rankin Scale improved in 17 patients (50%), stayed unchanged in 10 patients (29%), and was impaired in 1 patient (3%). Six patients (18%) died; however, 2 of them died due to reasons not directly related to meningitis or ventriculitis, so the overall mortality rate for meningitis and/or ventriculitis was 11.8% in this group of patients. All patients with ventriculitis (n = 4) were treated by antibiotics administered via the IVT route. The average time to CSF sterilization was 6.5 days in the patients with ventriculitis. Thirty patients had clinical signs of meningitis without ventriculitis. Despite the higher ratio of unfavorable Gram-negative meningitis in the subgroup of patients treated via lumbar drainage, the mean duration of CSF sterilization was 2.2 days compared with 2.6 days in the subgroup treated via external ventricular drainage, a difference that was not statistically significant (p = 0.3). Adverse effects of antibiotics appeared in 3 of 34 patients and were of low clinical significance. Conclusions. Intraventricular/lumbar intrathecal antibiotics can lead to very quick CSF sterilization in postneurosurgical patients with meningitis and ventriculitis. The relapse rate of meningitis and/or ventriculitis is also very low among patients treated by antibiotics. Intraventricular/lumbar intrathecal administration of antibiotics appears to be an effective and safe treatment for infections of the CNS caused by multidrug-resistant organisms. In patients with signs of ventriculitis, the authors prefer the IVT route of antibiotics. This study did not prove a lower efficacy of administration of antibiotics via lumbar drainage compared with the ventricular route in patients with meningitis. ( Key Words antibiotic intraventricular intrathecal meningitis ventriculitis brain abscess infection Despite the fact that the first IVT administration of antibiotics was performed more than 50 years ago, 3 the indication for the use of IVT or IT antibiotics remains controversial. There is no clinical trial based evidence in the adult population and no current recommended or approved therapy. The Cochrane review of Abbreviations used in this paper: EVD = external ventricular drainage; IT = intrathecal; IVT = intraventricular; mrs = modified Rankin Scale. IVT antibiotics for Gram-negative meningitis in infants found only 1 study that assessed the effectiveness and safety of IVT antibiotics (with or without intravenous antibiotics) in neonates with meningitis (with or without ventriculitis) as compared to treatment with intravenous antibiotics alone. 8,12 The conclusion of this 1 multicenter study was that the use of IVT antibiotics in addition to intravenous antibiotics in neonates resulted in a 3-fold increased relative risk of death compared with standard treatment with intravenous antibiotics alone. Based on 1596 J Neurosurg / Volume 119 / December 2013

2 Intraventricular and lumbar intrathecal antibiotics this result, IVT antibiotics as tested in this trial should be avoided. The results of that study, however, are not applicable to the neurointensive care of adult patients. Most patients in that study were children up to 30 days old, and gentamicin was administered by repeated IVT punctures. The statement that taps of fluid-filled spaces may cause harm, as the needle has to penetrate the brain tissue, is not in accordance with observations made in the adult population. Insertion of the ventricular catheter into the lateral ventricle of the brain in adults is a safe procedure with a complication rate of cerebral bleeding during the procedure of up to 1%. 11 The overall risk of disabling and persistent symptoms following a lumbar puncture has been estimated to be 0.1% 0.5%. 15 Moreover, half of the children in the IVT group who died received only 1 dose of IVT gentamicin, leading to speculation about the cause of death in the experimental group. Among other explanations for this negative trial were the delay in starting treatment and endotoxin release from Gram-negative organisms associated with increased cerebral edema. 16 Bacterial ventriculitis and meningitis is rarely noted in neurosurgical patients but is a serious complication. Clinically, the diagnosis is often difficult to establish because of its sometimes insidious symptoms. The underlying trauma or neurosurgery may result in a meningeal inflammatory response that will consequently affect CSF parameters. 13 Postneurosurgical ventriculomeningitis is typically caused by Staphylococcus aureus or coagulasenegative staphylococci. 4,14 A Gram-negative origin is associated with a severe underlying disease and a worse prognosis. 4 The overall mortality rate among patients with neurosurgical Gram-negative bacillary ventriculitis or meningitis has been reported to be 8% 70%, with the highest rates reported before the introduction of thirdgeneration cephalosporins. 13 Intraventricular and IT antibiotics have been studied in severe meningitis (with or without associated ventriculitis), IVT rupture of purulent brain abscess, and CSF shunt or drainage infection. 16 Delivery of antibiotics to the CNS in selected, seriously ill patients requires a combination of intravenous and IVT or IT administration. 6,16 We retrospectively investigated the results of our group of 34 neurosurgical patients in whom IVT or IT antibiotics were used. In reviewing their clinical records, we focused on the rate of CSF sterilization, the difference between the causative organisms and clinical outcome, and the difference between the lumbar and IVT route of administration on CSF sterilization. Methods Study Population Overall, 23,258 surgical procedures were performed between the years 2000 and 2010 at the Department of Neurosurgery, Na Homolce Hospital, Prague. The incidence of infectious complications in that period was 1.14%. Thirty-four neurosurgical patients operated on between 2000 and 2010 were treated by IVT or IT antibiotics for meningitis and/or ventriculitis (Table 1). The patients with meningitis and/or ventriculitis were chosen J Neurosurg / Volume 119 / December 2013 for antibiotics only in cases in which intravenous antibiotics were not efficient and laboratory and clinical signs of meningitis and/or ventriculitis persisted for 5 8 days (average 7.2 days) despite treatment by intravenous antibiotics. The majority of these patients were in a serious clinical state. Another 185 patients were treated for postoperative meningitis and/or ventriculitis between the years 2000 and 2010 solely using intravenous antibiotics. The patients ranged in age from 17 to 83 years old (mean 55.8 ± 15.8 years). There were 20 men and 14 women. Eighteen patients in the group were admitted to surgery due to a brain tumor; 5 because of an aneurysm after a subarachnoid hemorrhage; 3 for a ventriculoperitoneal shunt for hydrocephalus; 3 for an intracerebral hematoma operation; 3 for treatment of the intracranial trauma; and 2 for a primary brain abscess. The data were analyzed retrospectively. Surgical revision of operative fields or infected ventriculoperitoneal shunts was performed in 22 of 34 patients. In 15 patients more than 1 revision was needed (range 2 5 revisions, mean 2.53 revisions). Data Collection In all cases we conducted standard laboratory tests daily, including a complete blood count, C-reactive protein levels, and serum biochemistry. When signs of inflammation were found in CSF, the cultures of CSF obtained via EVD, lumbar drainage, or lumbar puncture were performed every 2 3 days. Once CSF cultures were positive, they were repeated daily until the sterility of CSF was restored. The neurological status of patients was assessed when clinical and laboratory signs of CNS inflammation began and classification was made according to the mrs. The neurological and functional outcome (mrs score) at the time of discharge was compared with the mrs score before the administration of antibiotics began. Antibiotic Administration When CSF inflammatory markers and CSF cultures were found to be positive, intravenous antibiotics were administered. In our group of patients a serious clinical state persisted, despite treatment by intravenous antibiotics, and the application of antibiotics was considered. As a route for the administration of antibiotics we used EVD or lumbar drainage, which had been used previously in patients due to other reasons such as hydrocephalus, hematocephalus, and others. When the need for antibiotics arose and a patient had not previously undergone either EVD or lumbar drainage, the insertion of a lumbar drain was preferred due to the lower risk of complications for the patients. Five antimicrobial agents were used for treatment (gentamicin, vancomycin, colistin, meropenem, and netilmicin). All antibiotics were administered at 24-hour intervals. The administration was stopped within hours from the CSF becoming sterile. The dosages of gentamicin, vancomycin, and colistin reflected recommendations found in the literature. The dosages of meropenem and netilmicin were derived from the minimal inhibition concentration of antibiotics. 1597

3 F. Remeš et al. TABLE 1: Clinical data in the group of 34 patients* Case No. Age (yrs), Sex Admission Diagnosis No. of Surgical Procedures Reason for Causative Organism mrs Score Before Route Used Dosage (g/24 hrs) Time Until First Cultivation of CSF Negativity (days) Days of IVT/ IT Antibiotic Therapy Discharge mrs Score Clinical State vs Before 1 62, F brain tumor 4 meningitis ECFC 5 LD GEN, VAN 0.01, improved 2 82, F brain abscess 3 meningitis STAU 3 EVD GEN impaired 3 48, M aneurysm 5 ventriculitis KLPN 5 EVD COL, GEN 0.1, improved 4 61, M brain abscess 2 meningitis STAU 5 EVD VAN dead 5 38, F hydrocephalus 10 meningitis STEP 3 EVD VAN improved 6 62, M ICH 5 meningitis ATSP, ECFE 5 EVD GEN unchanged 7 70, F ICH 4 ventriculitis SEMA 5 EVD GEN unchanged 8 83, F brain tumor 1 meningitis KLPN 5 LD GEN unchanged 9 44, M aneurysm 4 ventriculitis KLPN, PSAE 5 EVD COL dead 10 71, F brain tumor 9 meningitis ESCO, PSAE 5 LD/EVD GEN unchanged 11 72, F brain tumor 6 meningitis STEP 4 LD/EVD GEN, VAN 0.01, improved 12 49, F aneurysm 6 meningitis STEP, STAU 5 LD/EVD VAN improved 13 53, M brain tumor 5 meningitis STEP, ESCO, ENCL 5 EVD VAN dead 14 66, M brain tumor 3 meningitis MGMO 2 LD GEN improved 15 50, M brain trauma 2 meningitis ECFE 4 LD GEN improved 16 53, M brain tumor 3 meningitis STEP 1 LD VAN improved 17 37, M hydrocephalus 3 meningitis COJK 5 EVD VAN improved 18 55, M brain tumor 7 meningitis STEP 4 LD VAN improved 19 49, M brain tumor 3 meningitis PSAE 5 LD COL improved 20 32, M brain tumor 4 meningitis PSAE 5 LD COL dead 21 17, M brain trauma 2 meningitis STEP 5 EVD VAN improved 22 68, F brain tumor 9 meningitis ATSP, PRMI 5 LD COL, NET 0.125, unchanged 23 32, M brain trauma 2 meningitis KLPN 5 LD GEN unchanged 24 59, M hydrocephalus 6 ventriculitis KLPN 5 EVD MER dead 25 58, F aneurysm 5 meningitis STEP 4 LD VAN improved 26 51, M brain tumor 3 meningitis MRSA 5 LD GEN improved 27 65, M brain tumor 4 meningitis ENCL 5 LD GEN unchanged 28 65, F brain tumor 3 meningitis STHO, ESCO 3 LD VAN improved 29 73, F brain tumor 6 meningitis ECFE, MGMO, PSAE, STHO, ECFC, ESCO, KLPN 5 LD VAN dead (continued) 1598 J Neurosurg / Volume 119 / December 2013

4 Intraventricular and lumbar intrathecal antibiotics TABLE 1: Clinical data in the group of 34 patients* (continued) Clinical State vs Before Discharge mrs Score Days of IVT/ IT Antibiotic Therapy Time Until First Cultivation of CSF Negativity (days) Dosage (g/24 hrs) Used Route mrs Score Before Causative Organism Reason for No. of Surgical Procedures Admission Diagnosis Age (yrs), Sex Case No , F brain tumor 7 meningitis STAU 4 LD GEN unchanged 31 53, M brain tumor 2 meningitis STAU 5 EVD GEN improved 32 20, F brain tumor 6 meningitis STEP 4 LD GEN unchanged 33 56, M aneurysm 2 meningitis STEP 5 EVD VAN improved 34 77, M ICH 1 meningitis ECFC 4 LD VAN unchanged * ATSP = Acinetobacter species; COJK = Corynebacterium jeikeium; COL = colistin; ECFC = Enterococcus faecium; ECFE = Enterococcus faecalis; ENCL = Enterobacter cloacae; ESCO = E. coli; GEN = gentamicin; ICH = intracerebral hematoma; KLPN = Klebsiella pneumoniae; LD = lumbar drainage; MER = meropenem; MGMO = Morganella morganii; MRSA = Methicillin-resistant S. aureus; NET = netilmicin; PRMI = Proteus mirabilis; PSAE = Pseudomonas aeruginosa; SEMA = Serratia marcescens; STAU = S. aureus; STEP = S. epidermidis; STHO = Streptococcus haemolyticus; VAN = vancomycin. Outcome Measures Treatment failure was defined according to criteria proposed by Tängdén et al. 13 and included death attributable to meningitis or relapse within 3 months. Death was not considered to be attributable to meningitis if all of the following criteria were met: 1) there were 2 or more negative CSF cultures before death; 2) inflammatory parameters resolved; 3) clinical signs of meningitis resolved; and 4) a cause other than meningitis was found to be more probable according to the treating physician. Relapse was defined as the isolation of the same organism from CSF after the completion of antibiotic treatment or the development of presumed meningitis with clinical and laboratory signs. 13 Statistical Analysis The group of patients was statistically analyzed using column statistics analysis that described the distribution of values in a column and tested whether the distribution differed significantly from a Gaussian distribution. Nonparametric variables were analyzed using the chi-square test and Fisher exact test. Significance was assumed at p < Statistical analysis was performed using Statistica software (version 8.0, StatSoft). Results All the patients in our group developed clinical and laboratory signs of meningitis and/or ventriculitis. The causative organisms for these conditions are listed in Table 1. Postoperative meningitis or ventriculitis was caused by Gram-positive organisms in 16 cases (47%) and by Gram-negative organisms in 13 cases (38%); cultures were mixed in 5 cases (15%). Antibiotic Administration Intravenous antibiotic therapy was administered in all patients after CSF culture material was obtained; usually a combination of antibiotics was used. As a result of positive CSF cultures despite treatment, therapy was augmented by IVT or IT application of antibiotics. Five antibiotics were chosen for the route of administration: gentamicin was used in 16 patients, vancomycin in 15 patients, colistin in 5 patients, meropenem in 1 patient, and netilmicin in 1 patient. A combination of 2 antibiotics in 1 patient occurred in 4 cases. The period of administration ranged from 1 to 40 days (mean 12.1 ± 10.3 days). The mean time necessary to achieve CSF sterilization was 2.9 ± 2.7 days (range 1 12 days). The CSF cultures became negative within 24 hours after the administration of antibiotics in 17 patients (50%) and up to 48 hours in a further 6 patients (18%; Fig. 1). The average duration from the beginning of intravenous antibiotics until administration was 7.2 days. The IVT route for antibiotics (EVD) was used in 13 cases (38%) and the IT route via lumbar drainage in 18 cases (53%). In 3 cases (9%) the administration of antibiotics via lumbar drainage was changed to the IVT route. All patients with ventriculitis (n = 4) were treated by antibiotics administered via the IVT route. The average J Neurosurg / Volume 119 / December

5 F. Remeš et al. Fig. 2. Clinical outcomes according to the mrs score. Bar graph shows a comparison of patients mrs scores before administration of antibiotics and on discharge. ATB = antibiotic; ITHC = intrathecal. Fig. 1. Graph showing time until CSF sterilization after administration of antibiotics. time to achieve CSF sterilization was 6.5 days in patients with ventriculitis. Thirty patients developed clinical and laboratory signs of meningitis: 9 were treated solely by IVT antibiotics, whereas 18 were treated solely by IT antibiotics. In the subgroup of 9 patients treated solely via EVD there were no cases (0%) with Gram-negative meningitis, 7 cases (78%) with Gram-positive meningitis, and 2 cases (22%) with mixed culture. In the subgroup of 18 patients treated solely via lumbar drainage there were 8 cases (44%) with Gram-negative meningitis, 7 cases (39%) with Gram-positive meningitis, and 3 cases (17%) with mixed culture. Despite the higher ratio of unfavorable Gram-negative meningitis in the subgroup of patients treated via lumbar drainage, the mean duration until CSF sterilization was 2.2 days compared with 2.6 days in the subgroup treated via EVD; this difference was not statistically significant (p = 0.3). Outcome Measures The mean hospitalization of these patients in the Department of Neurosurgery was 41.8 days. The neurological status at the time of antibiotic administration was poor (mrs score > 2) in 32 patients (94%), and in 22 (65%) the mrs score was 5 (Table 1). Only 2 patients (6%) were in a favorable neurological state (mrs score 2). The clinical outcome of patients, assessed by the mrs score, improved in 17 (50%), remained unchanged in 10 (29%), and was impaired in 1 (3%; Fig. 2). Four patients were discharged to home care, 14 patients were transferred to rehabilitation in the neurological department, and 11 patients were transferred to continue intensive care. Six patients (17%) died, but 2 patients died for reasons not attributable to either meningitis or ventriculitis. Thus, the mortality rate for postneurosurgical meningitis or ventriculitis was 11.8% in our group. In the subgroup of patients with ventriculitis (n = 4), 1 patient died due to treatment failure (25%) and the remaining 3 patients (75%) survived with no signs of relapse. The clinical state of patients remained unfavorable, according to the mrs. The outcome of patients with meningitis after IVT/ IT antibiotics was worse in the subgroup with Gram-negative organisms comparing to those with Gram-positive organisms. In the subgroup of patients with Gram-negative meningitis and/or ventriculitis (n = 9), 2 patients (22%) died due to treatment failure and 6 patients (67%) remained in an unfavorable clinical state according to the mrs score at the end of the treatment. In 1 patient there were signs of relapse of meningitis and the patient eventually died due to treatment failure. In the subgroup of patients with Gram-positive organisms, 1 patient died (not due to treatment failure), 75% of patients remained in an unfavorable clinical state, and 19% remained in a favorable clinical state at the time of discharge. Two (40%) of 5 patients died with meningitis caused by mixed Grampositive and Gram-negative organisms, 1 due to treatment failure, and 1 due to reasons not attributable to the meningitis. Two patients (40%) remained in an unfavorable state and 1 patient (20%) improved to a favorable clinical state. In addition to the group of 34 patients treated by IVT/ IT antibiotics, there were 185 patients treated for CNS infection solely using intravenous antibiotics. The mortality rate in this group of patients was 13%. Adverse Effects There were no signs of nephrotoxic or hepatotoxic effects during the administration of antibiotics in our group of patients. We also did not observe any elevation in blood cell count in direct correlation with antibiotics. There was 1 epileptic seizure in a patient after the IVT administration of gentamicin, but the seizure was more probably caused by a coincident intracerebral hematoma. A horizontal nystagmus was observed in a different patient after the IVT administration of gentamicin. The nystagmus disappeared spontaneously within several minutes. One patient suffered from low-back pain and pain in the legs during the IT administration of gentamicin. This pain was managed by analgesics. We found no references in the literature to administration of meropenem and netilmicin. In our patients, no adverse symptoms were observed after the IVT/ IT administration of meropenem and netilmicin. Meropenem, in a dosage of 0.05 g/24 hours, was used in a patient with ventriculitis caused by Klebsiella pneumoniae. It was administered via EVD and led to CSF sterilization 5 days after its administration. Despite the combined therapy by intravenous and IVT antibiotics, the patient died 20 days 1600 J Neurosurg / Volume 119 / December 2013

6 Intraventricular and lumbar intrathecal antibiotics after CSF sterility was achieved. Netilmicin was used in combination with colistin in a case of meningitis caused by Gram-positive and Gram-negative organisms. The combination of antibiotics led to CSF sterilization within 2 days. The patient survived and was transferred to the neurological department for subsequent rehabilitation. Discussion The success rates of CSF sterilization, published in a few previous case reports or small case series, indicate that therapy with is a potentially effective treatment against both Gram-negative and Gram-positive meningitis, as well as fungal meningitis. 16 Tängdén et al. demonstrated a significantly higher cure rate among postneurosurgical patients with Gram-negative bacterial ventriculomeningitis treated with IVT gentamicin. 13 Moreover, no relapse of ventriculomeningitis occurred in that group of patients, whereas 6 of 18 patients treated by intravenous antibiotics alone experienced relapse. 13 There was only 1 case (3%) of ventriculomeningitis relapse among our 34 patients treated by antibiotics. The treatment of nosocomial meningitis is a major challenge due to the complexity of these patients and the emergence of resistance of both Gram-negative and Gram-positive organisms to the antibiotics. In patients with meningitis superimposed on another CNS disease (trauma, tumors, hemorrhage) causing loss of blood-brain barrier integrity, regional differences in CNS drug penetration may be of clinical importance. 16 For example, a cerebral microdialysis study in patients with traumatic brain injury found that concentrations of intravenously administered vancomycin in edematous brains were never above minimum inhibitory concentrations. 2 Moreover, the nonuniform distribution of antibacterial agents in different CNS compartments, many times with the lowest concentration in ventricular CSF, partly explains why ventriculitis is so difficult to eradicate successfully. Patients with meningitis are often postoperative neurosurgical patients, patients with traumatic brain injury, and immunosupressed patients. 16 The mortality rate of meningitis in postneurosurgical patients has been reported to be 3% 33%. 1,7,9,10 The mortality rate attributable to meningitis and/or ventriculitis was 11.8% in our group of patients. However, 94% of patients were in an unfavorable clinical state at the time of administration of antibiotics, and intravenous administration of antibiotics before the antibiotics had not led to CSF sterilization in those patients. Our study did not prove that there was a significantly better effect of antibiotics administered via EVD on CSF sterilization in the group of patients with meningitis compared with the lumbar route. The mean duration until CSF sterilization was 2.2 days in a subgroup of patients treated via lumbar drainage compared with 2.6 days in a subgroup treated via EVD, although the rate of Gramnegative meningitis was 44% in the lumbar drainage subgroup and 0% in the EVD subgroup. This result is contrary to a previously published finding in which injection of antibiotics into the lumbar CSF did not produce sufficient cisternal and ventricular drug J Neurosurg / Volume 119 / December 2013 concentrations. 5 In the study by Kaiser and McGee, 5 the administration of 5 10 mg of aminoglycoside into the lumbar intrathecal space resulted in mg/ml in the lumbar CSF, but mg/ml in the ventricular CSF. The concentrations of aminoglycoside in cisternal CSF after the lumbar administration were up to 15 mg/ml. The technique used to measure the aminoglycoside concentration in cisternal CSF is not described in that paper. In contrast, aminoglycoside administered into the cerebral ventricles produced concentrations in the lumbar CSF of mg/ml and ventricular CSF of mg/ml. 5 With regard to the rate of CSF sterilization, we did not find superiority of ventricular administration of antibiotics compared with lumbar administration. In the past, administration of most antimicrobial agents has been viewed as associated with significant toxicity. Toxicity appears to be dose related, and early reports may have been associated with inappropriate dosing. Moreover, systemic toxicity of the intravenous route of these same agents may be offset by using the CSF route. Reported toxicity from IVT administration of colistin included meningeal irritation, seizure with dosing, anorexia, eosinophilia, sacral and facial edema, leg pain, CSF xanthochromia, and transient albuminuria. Transient hearing loss, seizures, and eosinophilic granulocytosis of CSF have been reported in cases of gentamicin aseptic meningitis, and sensorineural hearing loss was described in IVT administration of vancomycin. Meropenem and netilmicin were not previously used for administration, and so to date no adverse effects were described for either of them. Adverse effects were noted in 3 (9%) of our cases. There were no signs of nephrotoxicity or hepatotoxicity associated with administration of antibiotics. Conclusions Intraventricular/lumbar intrathecal antibiotics can lead to very quick CSF sterilization in postneurosurgical patients with meningitis and ventriculitis. The relapse rate of meningitis and/or ventriculitis is also very low among patients treated by antibiotics. Intraventricular/ lumbar intrathecal administration of antibiotics appears to be an effective and safe treatment for infections of the CNS caused by multidrug-resistant organisms. In cases with signs of ventriculitis, we prefer the IVT route of antibiotics. In patients with meningitis, we did not show a lower efficacy of administration of antibiotics via lumbar drainage compared with the ventricular route. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Remeš, Tomáš, Vaniš, Šetlík. Acquisition of data: Remeš, Tomáš, Jindrák, Vaniš. Analysis and interpretation of data: Tomáš, Jindrák. Drafting the article: Tomáš. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Remeš. Statistical analysis: Tomáš. Study supervision: Šetlík. 1601

7 F. Remeš et al. References 1. Briggs S, Ellis-Pegler R, Raymond N, Thomas M, Wilkinson L: Gram-negative bacillary meningitis after cranial surgery or trauma in adults. Scand J Infect Dis 36: , Caricato A, Pennisi M, Mancino A, Vigna G, Sandroni C, Arcangeli A, et al: Levels of vancomycin in the cerebral interstitial fluid after severe head injury. Intensive Care Med 32: , Clifford HE, Stewart GT: Intraventricular administration of a new derivative of polymyxin B in meningitis due to Ps. pyocyanea. Lancet 2: , Federico G, Tumbarello M, Spanu T, Rosell R, Iacoangeli M, Scerrati M, et al: Risk factors and prognostic indicators of bacterial meningitis in a cohort of 3580 postneurosurgical patients. Scand J Infect Dis 33: , Kaiser AB, McGee ZA: Aminoglycoside therapy of gramnegative bacillary meningitis. N Engl J Med 293: , Kala M, Kolář M, Hájek V: [Local application of antimicrobial preparations in inflammatory intracranial diseases and its significance.] Cesk Slov Neurol Neurochir 59/92: , 1996 (Czech) 7. Mancebo J, Domingo P, Blanch L, Coll P, Net A, Nolla J: Postneurosurgical and spontaneous gram-negative bacillary meningitis in adults. Scand J Infect Dis 18: , McCracken GH Jr, Mize SG, Threlkeld N: Intraventricular gentamicin therapy in gram-negative bacillary meningitis of infancy. Lancet 315: , O Neill E, Humphreys H, Phillips J, Smyth EG: Third-generation cephalosporin resistance among Gram-negative bacilli causing meningitis in neurosurgical patients: significant challenges in ensuring effective antibiotic therapy. J Antimicrob Chemother 57: , Parodi S, Lechner A, Osih R, Vespa P, Pegues D: Nosocomial enterobacter meningitis: risk factors, management, and treatment outcomes. Clin Infect Dis 37: , Roitberg BZ, Khan N, Alp MS, Hersonskey T, Charbel FT, Ausman JI: Bedside external ventricular drain placement for the treatment of acute hydrocephalus. Br J Neurosurg 15: , Shah S, Ohlsson A, Shah V: Intraventricular antibiotics for bacterial meningitis in neonates. Cochrane Database Syst Rev 7: CD004496, Tängdén T, Enblad P, Ullberg M, Sjölin J: Neurosurgical gram-negative bacillary ventriculitis and meningitis: a retrospective study evaluating the efficacy of intraventricular gentamicin therapy in 31 consecutive cases. Clin Infect Dis 52: , Tunkel AR, Hartman BJ, Kaplan SL, Kaufman BA, Roos KL, Scheld WM, et al: Practice guidelines for the management of bacterial meningitis. Clin Infect Dis 39: , Wiesel J, Rose DN, Silver AL, Sacks HS, Bernstein RH: Lumbar puncture in asymptomatic late syphilis. An analysis of the benefits and risks. Arch Intern Med 145: , Ziai WC, Lewin JJ III: Improving the role of intraventricular antimicrobial agents in the management of meningitis. Curr Opin Neurol 22: , 2009 Manuscript submitted November 12, Accepted June 6, Please include this information when citing this paper: published online August 16, 2013; DOI: / JNS Address correspondence to: František Remeš, M.D., Department of Neurosurgery, Na Homolce Hospital, Roentgenova 2, Prague 5, Czech Republic. frantisek.remes@homolka.cz J Neurosurg / Volume 119 / December 2013