Skip Navigation

Evaluation of the Management of Postoperative Aseptic Meningitis

  1. Virginie Zarrouk1,
  2. Isabelle Vassor1,
  3. Frederic Bert2,
  4. Didier Bouccara3,
  5. Michel Kalamarides4,
  6. Noelle Bendersky5,
  7. Aimée Redondo4,
  8. Olivier Sterkers3, and
  9. Bruno Fantin1
  1. 1Department of Internal Medicine, Assistance Publique—Hôpitaux de Paris, Clichy, France
  2. 2Department of Microbiology, Assistance Publique—Hôpitaux de Paris, Clichy, France
  3. 3Department of Ear, Nose, and Throat, Assistance Publique-Hôpitaux de Paris, Clichy, France
  4. 4Department of Neurosurgery, Assistance Publique—Hôpitaux de Paris, Clichy, France
  5. 5Department of Statistics, Beaujon Hospital, Assistance Publique—Hôpitaux de Paris, Clichy, France
  1. Reprints or correspondence: Dr. Bruno Fantin, Service de Médecine Interne, Hôpital Beaujon, 100 Blvd. du Général Leclerc, 92110 Clichy, France (bruno.fantin{at}bjn.aphp.fr).
 
Next Section

Abstract

Background. A consensus conference recommended empirical antibiotic therapy for all patients with postoperative meningitis and treatment withdrawal after 48 or 72 h if cerebrospinal fluid culture results are negative. However, this approach is not universally accepted and has not been assessed in clinical trials.

Methods. We performed a cohort study of all patients who received a diagnosis of postoperative meningitis from January 1998 through May 2005 in a teaching hospital. From January 1998 through September 2003 (control period), guidelines were lacking or were not implemented. From October 2003 through May 2005 (interventional period), all patients received a predefined intravenous antibiotic therapy that was discontinued on the third day if the meningitis was considered aseptic. Clinical outcome and duration of antibiotic therapy were analyzed for each patient.

Results. Seventy-five episodes of postoperative meningitis (21 cases of bacterial meningitis and 54 cases of aseptic meningitis) were investigated. Patients with aseptic meningitis received antibiotic treatment for a mean ± standard deviation duration of 11 ± 5 days during the control period and 3.5 ± 2 days during the intervention period (P = .001). The duration of antibiotic treatment for bacterial meningitis was not significantly different between the 2 periods. All episodes of bacterial and aseptic meningitis were cured, and complications were rare during both periods.

Conclusions. Stopping antibiotic treatment after 3 days is effective and safe for patients with postoperative meningitis whose cerebrospinal fluid culture results are negative.

Meningitis is a rare but life-threatening complication of intracranial surgery [1, 2]. Several risk factors have been identified, such as postoperative CSF leakage, CSF shunts, an infected surgical route, and emergency surgery [1, 3,4–5].

Diagnosis is sometimes difficult, because the clinical manifestations are often mild and nonspecific during the immediate postoperative period. In addition, CSF protein and cell composition are modified by the surgical procedure itself, making it difficult to interpret laboratory results. Finally, direct bacteriological examination results are often negative [6]. Rapid diagnosis and antimicrobial chemotherapy are crucial, however, because the mortality rate may exceed 20% if treatment does not begin immediately after disease onset [4, 7].

The concept of postoperative aseptic meningitis was first described by Cushing in 1925 [8], but the mechanisms remain unclear. Aseptic meningitis occurs more frequently in children and after surgery involving the posterior fossa, and it may occur as a result of a local inflammatory reaction to blood breakdown products or to tumor antigens [9, 10]. Aseptic meningitis comprises 60%–75% of all cases of postoperative meningitis. The clinical manifestations and CSF findings associated with aseptic meningitis are very similar to those associated with postoperative bacterial meningitis, and the distinction between the 2 entities is currently based solely on the results of CSF culture [8, 11, 12]. Clinical outcome is generally favorable, despite a lengthy period of convalescence in most cases; steroid therapy is advocated by some authors [9].

In 2000, a consensus conference organized by the British Society of Antimicrobial Chemotherapy [13] recommended empirical antibiotic therapy (based on the local bacterial ecology) for all patients with signs of postoperative meningitis and treatment withdrawal after 48 or 72 h if CSF culture results are negative. However, this approach is not universally accepted [6, 14] and has not been assessed in clinical trials.

Previous SectionNext Section

Patients and Methods

We studied all patients who received a diagnosis of postoperative meningitis from January 1998 through May 2005 in the ear, nose, and throat (ENT) surgery and neurosurgery departments of our teaching hospital. Patients with external or internal CSF shunts were excluded.

The study was divided into 2 periods. During the first period (January 1998–September 2003), which served as an observational control period, guidelines were lacking or were not implemented. During the second, interventional period (October 2003–May 2005), we implemented a common management strategy in all the relevant departments of our institution. All patients who received a diagnosis of postoperative meningitis received antibiotic therapy, including a combination of ceftazidime (50 mg/kg 4 times per day), ciprofloxacin (400 mg intravenously 3 times per day), and vancomycin (loading dose of 15 mg/kg, then continuous infusion of 40 mg/kg per day). This regimen was chosen on the basis of the bacterial species and resistance pattern usually encountered in our hospital. During the first 3 days of treatment, the patient's situation was reevaluated by the same clinician specializing in infectious diseases. If the meningitis was shown to be of bacterial origin, antimicrobial chemotherapy was adapted to the isolate and was continued for 2 weeks. If the meningitis was considered to be aseptic, antimicrobial chemotherapy was discontinued on the third day.

Patients. During the first period, patients were recruited retrospectively on the basis of CSF analytical data. Patients' computer files from the department of microbiology were screened for data on all CSF samples obtained in the departments of neurosurgery, ENT surgery, and surgical intensive care. Patients were considered to have postoperative meningitis if they had undergone neurosurgery or ENT surgery <3 months previous to the study period, if CSF analytical results met the criteria for meningitis, and if the clinical chart contained information showing that the clinician in charge of the patient had performed lumbar puncture for a clinical suspicion of meningitis.

During the second period, all patients who had undergone neurosurgery or ENT surgery in the previous 3 months, who had a clinical indication for lumbar puncture, and whose CSF analytical results met the criteria for meningitis were prospectively enrolled. Potential indications for lumbar puncture consisted of fever (temperature, >38°C), headache, a meningeal stiffness, altered consciousness, or onset of a new focal neurological sign that was not present in the immediate postoperative period. If these 2 latter situations occurred, cerebral CT was performed before lumbar puncture.

Definition of meningitis. Meningitis was considered to be of bacterial origin if (1) CSF was positive on direct examination and/or by culture or (2) a microbiological sample (blood, cerebrospinal leakage fluid, or sample from the infected surgical wound) obtained during the same episode was positive for meningitis and a CSF sample contained >100 leukocytes/mm3 [11]. Meningitis was considered to be aseptic if a CSF sample contained >100 leukocytes/mm3 and direct examination and culture results of CSF were negative for meningitis after 72 h [8, 13].

Clinical and biological data collection. For each patient, we collected data on (1) demographic characteristics; (2) comorbidities; (3) indications for surgery, the surgical approach, and the duration of surgery; (4) use of antimicrobial chemoprophylaxis; (5) clinical manifestations at diagnosis (CSF leakage and the interval between leakage and meningitis onset); (6) surgical wound aspect; (7) biological findings (CSF cytology findings, protein and glucose levels, and bacteriological findings), blood culture results, CSF leakage culture results; (8) treatment (type and duration of antibiotic therapy); and (9) outcome of treatment (cure, neurological sequelae, hydrocephaly requiring secondary CSF shunting, or death).

Statistical analysis. Student's t test was used to compare means, and a χ2 test was used to compare proportions. To rule out a Hawthorne effect, the comparability of the control group and the intervention group was assessed by a nonparametric analysis of variance for the hypothesis of no overall period effect.

Previous SectionNext Section

Results

Study population. From January 1998 through May 2005, we identified 75 episodes of postoperative meningitis (44 episodes occurred during the control period, and 31 episodes occurred during the intervention period). Clinical symptoms, indication for surgery, and surgical approach were comparable between the 2 study periods (data not shown; P > .5). Twenty-one patients (28%) had bacterial meningitis, and 54 patients (72%) had aseptic meningitis. The demographic, clinical, and surgical characteristics of the study patients are shown in table 1. Fifty-five patients (73%) had undergone surgery of the posterior fossa for vestibular schwannoma. All but 2 patients had received preoperative antimicrobial chemoprophylaxis; most of them had received a second-generation cephalosporin.

Table 1
View larger version:
Table 1

Initial presentation of patients with postoperative meningitis.

Comparison of bacterial and aseptic meningitis. There was no significant difference between patients with bacterial meningitis and patients with aseptic meningitis with regard to comorbidities, the surgical indications, or the surgical approach (tables 1 and 2). The mean operating time was longer in the aseptic meningitis group (P = .015). A previous neurosurgical procedure was more frequently reported among patients with bacterial meningitis (P = .024). The interval between surgery and symptom onset and the clinical manifestations at diagnosis did not differ between the 2 groups. CSF leakage was present in 41 patients (55%), usually in the form of otorrhea (in 17 patients). Two of the patients with bacterial meningitis and none of the patients with aseptic meningitis developed new neurological signs. Drowsiness was experienced by 10 patients (6 patients with bacterial meningitis and 4 patients with aseptic meningitis).

Table 2
View larger version:
Table 2

Biological findings in CSF samples from patients with bacterial or aseptic postoperative meningitis.

CSF characteristics did not differ between the 2 groups. The degree of pleiocytosis was similar. The CSF glucose level was low in both groups (the glucose level was <50% of the blood glucose concentration in >75% of patients). The glucose value was zero in similar proportions of the patients in the 2 groups. The CSF protein level was not different in the 2 groups.

Bacteriology. Bacterial meningitis was diagnosed in 21 patients; the diagnosis of bacterial meninigitis in 15 of these patients was made on the basis of a positive CSF culture result (table 3). In 1 patient, the CSF was positive for bacterial meningitis by direct examination but negative for bacterial meningitis by culture. In 5 patients, the diagnosis was based on the isolation of a significant pathogen from blood cultures or from a contiguous sample (contaminated scar or cerebrospinal leakage fluid).

Table 3
View larger version:
Table 3

Bacterial isolates in 21 cases of postoperative bacterial meningitis.

Impact of new management guidelines. During the control period, all but 1 patient received antibiotic treatment, for a mean duration of 16 days (range, 0–52 days). During this period, 19 (67%) of 30 patients with aseptic meningitis received antibiotic therapy for >3 days (mean duration of antibiotic therapy, 11 days) (table 4).

Table 4
View larger version:
Table 4

Impact of management of postoperative meningitis on clinical outcome during the 2 study periods.

During the intervention period, the overall mean duration of antibiotic treatment was 7 days (range, 3–21 days; P < .001, compared with the control period). The patients with aseptic meningitis received antibiotic treatment for a mean duration of 3.5 days (range, 2–5 days; P = .001, compared with the control period). The duration of antibiotic treatment for bacterial meningitis was not significantly different between the 2 periods.

All episodes of bacterial and aseptic meningitis were resolved. Hydrocephaly occurred in 5 patients between 10 days and 1 month after onset of meningitis; 2 of 21 of these patients had bacterial meningitis, and 3 of 54 of these patients had aseptic meningitis. Four patients required temporary or permanent CSF shunting. Other complications included cerebral thrombophlebitis in a patient with bacterial meningitis and neutropenia (during cefotaxime therapy) and urticaria (during ciprofloxacin therapy) in patients with aseptic meningitis. Three cases of bacterial meningitis and no cases of aseptic meningitis were followed by neurological sequelae.

Previous SectionNext Section

Discussion

Our single-center study of 75 consecutive cases of meningitis is, to our knowledge, the largest reported series of postoperative meningitis to date. This is mainly explained by the fact that aseptic meningitis occurs more frequent after surgery involving the posterior fossa [9, 10]; our hospital is a reference center for vestibular schwannoma surgery, which explains the predominance of subtentorial operations in this series. However, postoperative meningitis can occur after any procedure associated with a breach in the blood-brain barrier, and there is no data in our study or in the literature indicating different characteristics of aseptic meningitis according to the anatomical origin [3, 11, 12]. Nevertheless, the number of cases of aseptic meningitis was probably underestimated in the retrospective part of the study, because some rapidly resolved cases may not have been included. However, the clinical presentation of aseptic meningitis did not differ between the control and intervention periods.

The most likely mechanism of aseptic meningitis is a reaction to heme breakdown products. Blomstedt [2] provoked meningitis in dogs by injecting RBCs intrathecally. Other authors have discussed the possible bacterial origin of some cases of so-called aseptic meningitis. Salord et al. [14] detected the presence of bacterial DNA, using PCR, in 19 of 20 patients with aseptic meningitis, and Druel et al. [6] recommended that postoperative “aseptic” meningitis should be considered and treated as bacterial meningitis.

With regard to clinical presentation, we failed to identify any signs reliably distinguishing between bacterial and aseptic meningitis. The duration between surgery and onset of meningitis tended to be shorter among patients with aseptic meningitis than among those with bacterial meningitis, in accordance with previous reports [8, 9, 11]. One-third of cases of bacterial meningitis occurred after a repeated surgery, which is one of the main risk factors for surgical wound infection [1, 3, 5, 15]. The duration of surgery was significantly longer among patients who developed aseptic meningitis, possibly leading to a higher concentration of irritants at the surgical site.

Thus, bacterial and aseptic meningitis cannot be distinguished by clinical characteristics alone. Unfortunately, none of the CSF characteristics studied allowed us to distinguish between aseptic and bacterial meningitis. Although the mean CSF polymorphonuclear leukocyte count was higher in patients with bacterial meningitis than it was in patients with aseptic meningitis, the overlap between the ranges observed in each patient group makes this parameter irrelevant to the differential diagnosis. The CSF glucose value was zero in several patients with aseptic meningitis. Our findings confirm that direct examination results are usually negative for postoperative bacterial meningitis [11]. The diagnosis of aseptic meningitis is, therefore, based solely on negative results of direct bacteriological examination of the CSF and on culture negativity, provided the samples are taken before antibiotic therapy is started. The efficacy of molecular methods for determining the bacterial origin of meningitis remains to be evaluated. For bacterial meningitis, our bacterial isolates were similar to those recovered from similar patients and reported elsewhere [7, 8, 16].

There is no consensus for the treatment of aseptic meningitis. Some authors recommend routine antibiotic treatment of all cases of meningitis, even without microbiological documentation [6, 14]. In view of the initial difficulties in identifying patients with bacterial meningitis and the morbidity and mortality resulting from delayed antimicrobial chemotherapy in this context, the British Society of Antimicrobial Chemotherapy recommends that empirical antibiotic therapy be used in every case. If the CSF remains sterile, antibiotic treatment can be withdrawn after 2 or 3 days [13]. In this setting, the choice of empirical antimicrobial chemotherapy has not been standardized and depends on the local bacterial ecology. Our empirical antibiotic regimen included a combination of ciprofloxacin and vancomycin and ceftazidime. This strategy appeared to be effective, because none of the bacteria isolated during the second part of our study were resistant to these drugs. Moreover, the outcome of bacterial meningitis was similar during the 2 study periods, with no attributable deaths and no increase in the rate of sequelae. In retrospect, it was not necessary for our antibiotic regimen to be effective against Pseudomonas aeruginosa, because this species was not isolated during the study period.

As reported elsewhere, our patients with aseptic meningitis always recovered without sequelae, even after the change in treatment strategy. This result indicates that none of the cases of aseptic meningitis corresponded to misdiagnosed bacterial meningitis. The new management strategy for postoperative meningitis appeared to be effective, reducing the mean duration of antibiotic treatment from 16 days to 7 days for patients with either form of postoperative meningitis and from 11 days to 3.5 days for patients with aseptic meningitis. Therefore, this approach has the dual advantages of cost reduction and lower antimicrobial selection pressure, without increasing morbidity or mortality.

Previous SectionNext Section

acknowledgments

Potential conflicts of interest. All authors: no conflicts.

  • Received October 3, 2006.
  • Accepted February 21, 2007.
Previous Section
 

references

  1. ↵
    1. Korinek AM
    . Risk factors for neurosurgical site infections after craniotomy: a prospective multicenter study of 2944 patients. Neurosurgery 1997;41:1073-81.
    CrossRefMedlineWeb of Science
  2. ↵
    1. Blomstedt GC
    . Postoperative aseptic meningitis. Acta Neurochir (Wien) 1987;89:112-6.
    CrossRefMedline
  3. ↵
    1. Blomstedt GC
    . Infections in neurosurgery: a retrospective study of 1149 patients and 1517 operations. Acta Neurochir (Wien) 1985;78:81-90.
    CrossRefMedline
  4. ↵
    1. Buckwold F,
    2. Hand R,
    3. Hansebout R
    . Hospital-acquired bacterial meningitis in neurosurgical patients. J Neurosurg 1977;46:494-9.
    CrossRefMedlineWeb of Science
  5. ↵
    1. Mollman HD,
    2. Haines SJ
    . Risk factors for postoperative neurosurgical wound infection. J Neurosurg 1986;64:902-6.
    CrossRefMedlineWeb of Science
  6. ↵
    1. Druel B,
    2. Vandenesch F,
    3. Greenland T,
    4. et al
    . Aseptic meningitis after neurosurgery: a demonstration of bacterial involvement. Clin Microbiol Infect 1996;1:230-4.
    Medline
  7. ↵
    1. Federico G,
    2. Tumbarello M,
    3. Spanu I,
    4. et al
    . Risk factors and prognostic indicators of bacterial meningitis in a cohort of 3580 postneurosurgical patients. Scand J Infect Dis 2001;33:533-7.
    CrossRefMedlineWeb of Science
  8. ↵
    1. Ross D,
    2. Rosegay H,
    3. Pons V
    . Differentiation of aseptic and bacterial meningitis in postoperative neurosurgical patients. J Neurosurg 1988;69:669-74.
    CrossRefMedlineWeb of Science
  9. ↵
    1. Carmel PW,
    2. Greif LK
    . The aseptic meningitis syndrome: a complication of posterior fossa surgery. Pediatr Neurosurg 1993;19:276-80.
    CrossRefMedlineWeb of Science
  10. ↵
    1. Finlayson AI,
    2. Penfield W
    . Acute postoperative aseptic leptomeningitis. Arch Neurol Psychiatr 1941;46:250-76.
  11. ↵
    1. Forgacs P,
    2. Geyer CA,
    3. Freidberg SR
    . Characterization of chemical meningitis after neurological surgery. Clin Infect Dis 2001;32:179-85.
    Abstract/FREE Full Text
  12. ↵
    1. Pönkä A,
    2. Treppo AM,
    3. Webe TH
    . The differential diagnosis of bacterial and aseptic meningitis using cerebrospinal fluid laboratory tests. Infection 1983;11:129-31.
    CrossRefMedlineWeb of Science
  13. ↵
    1. The management of neurosurgical patients with postoperative bacterial or aseptic meningitis or external ventricular drain-associated ventriculis
    . Infection in Neurosurgery Working Party of the British Society for Antimicrobial Therapy. Br J Neurosurg 2000;14:7-12.
    CrossRefMedlineWeb of Science
  14. ↵
    1. Salord F,
    2. Druel B,
    3. Grando J,
    4. et al
    . Méningites aseptiques: mise en évidence dans le LCR d'ADN bactérien par amplification génique. Ann Fr Anesth Réa 1995;14:320-5.
  15. ↵
    1. Patir R,
    2. Mahapatira AK,
    3. Banerji AK
    . Risk factors in postoperative neurosurgical infection: a prospective study. Acta Neurochir (Wien) 1992;119:80-4.
    CrossRefMedline
  16. ↵
    1. De Bels D,
    2. Korinek AM,
    3. Bismuth R,
    4. Trystram D,
    5. Coriat P,
    6. Puybasset L
    . Empirical treatement of adult postneurosurgical nosocomial meningitis. Acta Neurochir (Wien) 2002;144:989-95.
    CrossRefMedline
| Table of Contents

This Article

  1. Clin Infect Dis. (2007) 44 (12): 1555-1559. doi: 10.1086/518169
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 54 (5)
  1. Clinical Infectious Diseases
  1. Alert me to new issues

Most