|Home | About | Journals | Submit | Contact Us | Français|
The term ‘aseptic meningitis’ refers to a meningitis for which an etiological agent is not apparent after Gram stain and bacterial culture of cerebrospinal fluid (CSF) (1). Clinicians who assess children with aseptic meningitis recognize that the majority of cases are caused by viruses, but are often faced with having to exclude partially treated bacterial meningitis in children who had been on oral antibiotic(s). The following case underscores the difficulties.
A two-year-old boy presented to an emergency room in July with a two-day history of fever, nausea and vomiting. There were no prodromal symptoms, no infectious contacts, no pets at home and no recent travels. He was prescribed amoxicillin two days before, and had taken six doses in total. He was immunized previously with four doses of diphtheria- tetanus-acellular pertussis-polio-Haemphilius influenzae b (DTaP-polio-Hib) and a single dose of measles, mumps and rubella (MMR) at 12 months. He was lethargic and had nuchal rigidity. Rectal temperature was 39.5°C, heart rate was 126 beats/min and respiratory rate was 24 breaths/min. Blood pressure was normal. Tympanic membranes and pharynx appeared normal. There were no enlarged cervical nodes, nor signs of hepatosplenomegaly. No rash was evident, and there were no swollen or painful joints.
Initial lumbar puncture revealed cloudy cerebrospinal fluid (CSF), and Gram stain showed moderate neutrophils but no bacteria. CSF white blood cell (WBC) count was 420×106/L, with 80% neutrophils, 15% lymphocytes and 5% monocytes. CSF protein 0.78 g/L (normal 0.15 to 0.4 g/L) and CSF glucose 2.6 mmol/L (simultaneous blood glucose level of 5.3 mmol/L). CSF latex agglutination tests for Streptococcus pneumoniae, Hib, Neisseria meningitidis serogroups A, C, Y and W-135, and N meningitidis serogroup B/Escherichia coli K1 were negative.
The child was admitted to hospital. Parenteral antibiotic therapy was not initiated, and the oral antibiotic was discontinued. Blood culture, and bacterial and viral cultures of CSF were sent for study. Throat swab and stool specimen were also sent for viral culture. Twenty-four hours later, the child had improved clinically. There were no positive blood and CSF bacterial cultures, and it was too early for viral culture results. The decision was made to repeat the lumbar puncture, which now showed a total CSF WBC of 380×106/L, with 70% lymphocytes. The CSF protein was 0.85 g/L and CSF glucose 3.5 mmol/L. This shift from predominantly neutrophilic to lymphocytic pleocytosis was most consistent with viral meningitis, and the child was discharged home.
Three days later, an enterovirus was isolated from both CSF and stool culture, later confirmed to be an echovirus strain. The child was seen two weeks later, and had a normal neurological examination.
On occasion, the medical history may help point to possible etiological agents. This includes information such as a history of recent contact with other children or adults with viral illnesses, the lack of immunization against Hib or mumps, a history of exposure to insects in arbovirus endemic areas or a history of recent cat scratch disease. Nevertheless, the results of a lumbar puncture are ultimately needed to determine whether a case of meningitis is of viral or other etiology. The questions that commonly arise when managing children on oral antibiotic(s) presenting with aseptic meningitis are discussed below.
The major concern regarding the patient described in the case scenario was the possibility of partially treated bacterial meningitis (because of the prior amoxicillin use, the predominance of neutrophils in the initial CSF and the CSF glucose being less than 50% of the serum glucose). However, viral etiology was still most likely for two reasons.
The neutrophil predominance in the CSF is not unusual early in viral meningitis. For example, neutrophilic pleocytosis was found in up to two-thirds of cases of enteroviral meningitis (2). Modest hypoglycorrhachia has also been reported, occuring in up to 18% of enteroviral meningitis (3) and about one-quarter of mumps meningitis. A relatively low WBC count and mildly elevated protein in the CSF are consistent with viral infection (4), but one cannot completely rule out bacterial etiology because there have been bacterial meningitis cases without significant CSF pleocytosis (5).
In partially treated bacterial meningitis, the CSF WBC counts and protein levels are typically higher than was found in the case scenario (5). In one study, the mean CSF WBC count in patients with Hib meningitis previously treated with antibiotics was 5235×106/L (with 84% neutrophils), while the CSF to serum glucose ratio was 0.28 and mean CSF protein 1.37 g/L; CSF Gram stain revealed bacteria in 84% (6). In this study, the CSF total WBC count, percentage of neutrophils, glucose concentration and culture positivity rate did not differ significantly between previously treated and untreated groups (6). However, antibiotic pretreatment led to a significantly lower mean CSF protein concentration (1.37 g/L versus 2.08 g/L) and rate of positive Gram stain (84% versus 92%, respectively) (6). Although fewer patients were studied than for Hib, similar CSF findings were described for pneumococcal meningitis cases, but findings were less consistent in meningococcal meningitis (7–9). Pretreatment with antibiotics diminished the CSF culture positive rate from 95% to 68% in one study (8). While a small group of bacterial meningitis patients have CSF findings that mimic viral meningitis, clinicians should be able to apply the usual criteria in the majority of cases to differentiate between bacterial and viral etiology in those pretreated with antibiotic(s).
Diagnostic work-up of aseptic meningitis is often incomplete, and an etiological agent is identified in only about 10% of all cases. A virology laboratory may not be readily accessible; furthermore, during outbreaks, laboratory investigations are not necessary when there is clear epidemiological link between cases. However, it is helpful to investigate the early cases of an epidemic to determine the agent responsible and the likely prognosis. Laboratory tests that may or may not be useful are discussed below.
A repeat lumbar puncture performed 6 to 12 h after the first, showing a rapid shift in CSF WBC differential from neutrophil to lymphocyte predominance is highly suggestive of a viral infection (10,11).
Viral culture of stool (or rectal swab), throat swab and CSF may reveal the causative viral agent. The cultures take three to seven days and are not immediately useful in patient management. The rate of isolation of enterovirus was highest from stool (86%), followed by throat (57%) and CSF (39%) (5). The rate of enterovirus isolation from the CSF has varied from 40% to 80% (2). Unfortunately, isolation of an enterovirus from throat swab, rectal swab or stool is not always diagnostic because virus may be shed for several weeks after concurrent infection, and has been found in about 5% to 10% of healthy controls during enteroviral outbreaks.
Polymerase chain reaction technology that can identify the genetic material of enteroviruses in both CSF and stool exists as a means to diagnose enteroviral infection (12,13). This test is not yet widely available, but may become an important diagnostic tool in the future.
CSF latex agglutination tests for bacterial pathogens are not routinely useful for diagnosis of bacterial meningitis. Serological tests for enteroviruses are not recommended because there are too many serotypes for which to test. However, serological tests for certain other viruses (mumps, lymphocytic choriomeningitis virus, etc) may be indicated. Consultation with an expert in this area is suggested.
Nonmicrobiological tests, including the C-reactive protein (14), CSF lactate (15), CSF enzyme levels (16) and CSF cytokine levels (17), are not routinely indicated because they lack sufficient sensitivity and specificity in distinguishing bacterial from viral meningitis.
Before discovery of coxsackievirus and echovirus in 1948 and 1949, cases of aseptic meningitis were mainly associated with epidemic poliomyelitis, mumps and lymphocytic choriomeningitis virus. By 1962, the importance of the nonpolio enteroviruses and arthropod-borne viruses in aseptic meningitis was established (2). Subsequent studies of aseptic meningitis cases revealed viral etiological agents in 54% to 72% (18).
Enteroviruses are the most common agents in North America, accounting for 80% to 90% of all cases of viral meningitis (2). Outbreaks caused by specific serotypes of coxsackievirus and echovirus occur, as well as sporadic cases. The predominant serotypes vary from year to year and from one geographical region to the next. Peak incidence is typically in the summer and early fall, but cases do occur throughout the year. The infections are more common in infants and young children. Transmission occurs by the fecal-hand-oral route. In one study of hospitalized infants with enteroviral infection, 51% were found to have associated meningitis (4). The meningitis is usually self-limited, and treatment is supportive.
Circulation of wild polio virus has been interrupted in the Americas, but imported cases can still occur. Rare cases associated with the use of oral poliovirus vaccine have also occurred. However, all provinces in Canada have implemented the sole use of inactivated polio vaccine (in place of the oral vaccine).
Aseptic meningitis occurs with mumps and some mumps vaccine strains (for example the Urabe Am 9 and Leningrad 3 strains). The Urabe Am 9 strain was used in one preparation of MMR vaccine in some parts of Canada in the 1980s and in other countries (19). The current MMR vaccine widely used in Canada has a different (Jeryl-Lynn) strain, which has not been associated with aseptic meningitis.
In adolescents and young adults, aseptic meningitis has been reported in 10% to 30% of those who have symptomatic primary herpes genitalis (1). Herpes simplex virus (HSV) type 2 is more commonly reported than HSV type 1. HSV type 2 has been cultured from the CSF in some patients. Recovery is usually without sequelae. Aseptic meningitis is uncommon in recurrent genital herpes infections.
While aseptic meningitis has been associated with primary human immunodeficiency virus (HIV) infection among young adults (1), it is not known how common it is in HIV-infected children.
Arboviruses tend to cause encephalitis or meningoencephalitis rather than isolated aseptic meningitis (20). The most likely arthropod-borne viruses in North America to cause meningitis are the California encephalitis virus (found in the midwest and eastern part of North America), St Louis encephalitis virus (found throughout North America) and Powassan virus (found in the eastern part of North America) (2). These infections are most common in the summer months, when contact with arthropod vectors are most likely. Although often severe, arboviral infections may be mild enough that they may be confused with enteroviral meningitis. Serological tests for these arboviruses are available in reference laboratories.
Lymphocytic choriomeningitis virus is listed as a relatively important cause of aseptic meningitis (2), acquired from contact with infected rodents and their excreta, but it appears to be uncommon in Canada. It is not routinely looked for in many virology laboratories.
Other viruses that more typically cause encephalitis rather than aseptic meningitis include measles, rubella, Epstein-Barr virus, cytomegalovirus and varicella.
There have been isolated case reports of aseptic meningitis associated with infections caused by adenovirus, parainfluenza virus, influenza virus, human herpes virus 6 and parvovirus B19. More studies are needed to determine how important these agents are in causing aseptic meningitis.
The decision of when to start and to stop empirical antibiotic therapy is a challenging one.
Antimicrobial therapy is generally not effective for the treatment of most forms of viral meningitis. The use of acyclovir for aseptic meningitis associated with primary herpes genitalis appears unjustified because the meningitis is self-limited, but may be indicated for the genital disease. An antiviral agent that appears active against enteroviruses (pleconaril) is under study.
Empirical intravenous antibiotic therapy, appropriate for the age group, is indicated if the clinician suspects bacterial meningitis after reviewing the patient and the initial CSF profile. The decision to use antibiotic(s) is particularly influenced by a history of prior oral antibiotic use, young age and toxic appearance of the child. Peripheral WBC count is not reliable in helping to differentiate viral from bacterial illness.
The decision of when to stop empirical antibiotic therapy is especially difficult when the initial CSF is ambiguous, and blood and CSF cultures do not reveal a bacterial source quickly. If the situation is not clarified within the first few days in hospital, a repeat lumbar puncture may further support a viral rather than bacterial etiology and the antibiotic(s) can be discontinued.
When there is still uncertainty after these steps, it may be advisable to continue antibiotic therapy for a full seven to 10 days, pending viral culture results. In partially treated bacterial meningitis, the child’s clinical response to intravenous antibiotic(s) may suggest bacterial etiology. However, this usually takes several days, and because there is also spontaneous clinical improvement in viral meningitis, the clinician must be cautious in over-attributing clinical improvement to the antimicrobial therapy.
Most children with viral meningitis may be cared for at home. If a child is too ill to be sent home, he or she may be admitted to hospital for care without empirical antibiotic therapy and discharged when well.
From 1980 to 1992, between 250 and 700 cases of viral meningitis were reported yearly to public health authorities (21), an annual incidence of 1.1 to 2.6 cases per 100,000 population. The numbers likely underestimate the true magnitude of viral meningitis because cases are not always reported to public health authorities. To understand the epidemiology of viral meningitis better, both laboratory confirmed and clinically diagnosed viral meningitis cases should be reported to public health authorities.
INFECTIOUS DISEASES AND IMMUNIZATION COMMITTEE
Members: Drs Gilles Delage, Directeur scientifique, Laboratoire de santé publique du Québec, Ste-Anne-de-Bellevue, Québec (chair); François Boucher, Département de pédiatrie, Centre Hospitalier Universitaire de Québec, Pavillon CHUL, Québec; Joanne Embree, The University of Manitoba, Winnipeg, Manitoba; David Speert, Division of Infectious and Immunological Diseases, University of British Columbia, Vancouver, British Columbia; Ben Tan, Division of Infectious Diseases, Royal University Hospital, University of Saskatchewan, Saskatoon, Saskatchewan (principal author).
Consultants: Drs Noni MacDonald, Division of Infectious Diseases, Children’s Hospital of Eastern Ontario, Ottawa, Ontario; Victor Marchessault, Cumberland, Ontario
Liaisons: Drs Neal Halsey, Johns Hopkins University, Baltimore, Maryland (American Academy of Pediatrics); Susan King, Division of Infectious Diseases, The Hospital for Sick Children, Toronto, Ontario (Canadian Paediatric AIDS Research Group); David Scheifele, Division of Infectious Diseases, BC’s Children’s Hospital, Vancouver, British Columbia (Centre for Vaccine Evaluation); Ms Susan Tamblyn, Perth District Health Unit, Stratford, Ontario (Public Health); Dr John Waters, Provincial Health Officer, Alberta Health, Edmonton, Alberta (Epidemiology)
The recommendations in this Practice Point do not indicate an exclusive course of treatment or procedure to be followed. Variations, taking into account individual circumstances, may be appropriate.