Meningitis Workup

The diagnostic challenges in patients with clinical findings of meningitis are as follows:

• Early identification and treatment of patients with acute bacterial meningitis
• Assessing whether a treatable central nervous system (CNS) infection is present in those with suspected subacute or chronic meningitis
• Identifying the causative organism

Bacterial meningitis must be the first and foremost consideration in the differential diagnosis of patients with headache, neck stiffness, fever, and altered mental status. Acute bacterial meningitis is a medical emergency, and delays in instituting effective antimicrobial therapy result in increased morbidity and mortality.

In general, whenever the diagnosis of meningitis is strongly considered, a lumbar puncture should be promptly performed. Examination of the cerebrospinal fluid (CSF) is the cornerstone of the diagnosis. The diagnosis of bacterial meningitis is made by culture of the CSF sample. The opening pressure should be measured, and the fluid sent for cell count (and differential count), chemistry (ie, CSF glucose and protein), and microbiology (ie, Gram stain and cultures). If possible, about 6ml of fluid would be sufficient for routine testing. If tubercular meningitis is suspected, a large volume of CSF (10-20ml) should be sent to lab. Some CSF should be saved for further studies if needed.

Brain and spinal hemorrhage and spinal epidural or subdural hematoma are rare but potentially serious complications of an LP. In the multicenter LP feasibility study, only one of 3558 patients who underwent LP experienced this side effect (leading to death after restarting their oral anticoagulant). Therefore, it is advised to have a recent analysis of the platelet count (that should be >40 × 10⁹/L) and coagulation status (quick >50%; international normalized ratio < 1.5). Coagulopathies and uncorrected bleeding diathesis should be absent. It is advised to discontinue anticoagulant treatments to minimize hemorrhagic risks. The short-acting direct oral anticoagulants (DOACs) have as advantage that they can be discontinued shortly before the LP and that anticoagulation can be restarted within a few hours (6–8 hours) after the procedure. Antiplatelet drugs are only a relative contraindication, and most centers do not interrupt treatment with antiplatelet drugs before LP. Studies on LP complication risks in cases taking combinations of antiplatelet drugs (eg, clopidogrel and acetylsalicylic acid) are lacking, but it is considered safer to temporarily withhold one of both before LP. In case of dual antiplatelet therapy, it is advised to continue acetylsalicylic acid, whereas the intake of thienopyridine derivatives (eg, clopidogrel, ticlopidine) should be temporally withheld (1 or 2 weeks) before LP unless patients are at high thrombotic risk or unless there is an urgent indication to perform an LP, which is the case with meningitis. ^[20]

A concern regarding LP is that the lowering of CSF pressure from withdrawal of CSF could precipitate herniation of the brain. Herniation can sometimes occur in acute bacterial meningitis and other CNS infections as the consequence of severe cerebral edema or acute hydrocephalus. Clinically, this is manifested by an altered state of consciousness, abnormalities in pupil reflexes, and decerebrate or decorticate posturing. The incidence of herniation after LP, even in patients with papilledema, is approximately 1%.

A screening computed tomography (CT) scan of the head may be performed before LP to determine the risk for herniation. A prospective study involving 301 adults with suspected meningitis found that the following baseline patient characteristics were associated with an abnormal finding on head CT ^[25] :

• Age ≥60 years
• Immunocompromise (ie, HIV infection/AIDS, immunosuppressive therapy, or transplantation)
• A history of CNS disease
• A history of seizure within 1 week before presentation
• Any abnormality on neurologic examination

These factors have been included in the Infectious Diseases Society of America guidelines to decide who should undergo CT before LP. ^[26]

Obtaining a CT scan in a patient without any of the IDSA guidelines indications is of no clinical benefit and should be avoided. ^[27]

The decision to obtain a brain CT scan before LP should not delay the institution of antibiotic therapy; such delay can increase mortality. It should be also noted that herniation can occur in patients with bacterial meningitis who have a normal brain CT scan. The most reliable clinical signs that indicate the risk for herniation include deteriorating level of consciousness, brainstem signs, and a very recent seizure.

 Antibiotic therapy for several hours prior to lumbar puncture does not significantly alter the CSF WBC count or glucose concentration. As a general rule, Gram stain and culture of CSF obtained 24 hours after the initiation of antimicrobial therapy should be negative if the organism is sensitive to the antibiotic. The diagnosis of bacterial meningitis then is made on the basis of the abnormalities in CSF WBC count, glucose, and protein concentrations. ^[24]

For a traumatic tap, correction for RBC needs to be made. Leucocyte correction is made by deducting 1 WBC for every 1000 RBC in the CSF. CSF protein is corrected by deducting 0.01g/l for every 1000 RBCs.

Post treatment spinal tap is not needed in most cases. In 1 study in a pediatric population, ^[21]  results from 163 cured patients were provided and the results included the following:

1. Glucose level was less than 50 mg/dL in 36, it was less than 40 mg/dL in 8%. Although the mean CSF/serum glucose ratio after treatment was close to 0.6, which is usually taken to be the normal value in health, eight (13%) were below 0.4. Values for CSF glucose and CSF/serum glucose ratio before treatment did not correlate with those found after treatment.
2. Protein level was over 45mg/dl in 38% of cases, and it was over 100mg/dl in 8% of cases.
3. CSF WBC count were within the "normal" range in only 28% of cases. The total leukocyte count was 40/cu mm or more in 50 cases (32%), and 100/cu mm or more in 23 cases (15%).

Other laboratory tests, which may include blood cultures, are needed to complement the CSF culture. These bacterial cultures are used for identification of the offending bacteria and occasionally its serogroup, as well as for determination of the organism’s susceptibility to antibiotics. Special studies, such as serology and nucleic acid amplification, also may be performed, depending on clinical suspicion of an offending organism. Skin biopsy of a rash sometimes will lead to a diagnosis in patients with a rash and meningitic symptoms.

As many as 50% of patients with pneumococcal meningitis also have evidence of pneumonia on initial chest radiography. This association occurs in fewer than 10% of patients with meningitis caused by H influenzae or N meningitidis and in approximately 20% of patients with meningitis caused by other organisms. (See Bacterial Meningitis Imaging.)

In patients with bacterial meningitis, a complete blood count (CBC) with differential will demonstrate polymorphonuclear leukocytosis with a left shift. Useful elements of the metabolic panel include the following:

• Serum electrolytes, to determine dehydration or syndrome of inappropriate secretion of antidiuretic hormone (SIADH)
• Serum glucose (which is compared with the CSF glucose)
• Blood urea nitrogen (BUN) or creatinine and liver profile, to assess organ function and adjust antibiotic dosing
• HIV testing

The serum glucose level may be low if glycogen stores are depleted, or they may be high in infected patients with diabetes.

A coagulation profile and platelet count are indicated in cases of chronic alcohol use, chronic liver disease, or suspected disseminated intravascular coagulation (DIC). Patients with coagulopathies may require platelets or fresh frozen plasma (FFP) before LP.

Obtaining cultures before instituting antibiotics may be helpful if the diagnosis is uncertain. The utility of cultures is most evident when LP is delayed until head imaging can rule out the risk for brain herniation, in which cases adjunctive dexamethasone and antimicrobial therapy is rightfully initiated before CSF samples can be obtained. These cultures include the following:

• Blood - 50% positive in meningitis caused by H influenzae, S pneumoniae, or N meningitidis
• Nasopharynx
• Respiratory secretions
• Urine
• Skin lesions

Latex agglutination or counterimmunoelectrophoresis (CIE) of blood, urine, and CSF for specific bacterial antigens is occasionally recommended if diagnosis is challenging or in patients with partially treated meningitis. The Binax NOW S pneumoniae antigen test, if done on CSF, has a 99% to 100% sensitivity and specificity and can be positive despite prior antibiotic therapy. ^[28]

The use of nucleic acid amplification (eg, polymerase chain reaction [PCR] testing) has revolutionized the diagnosis of herpes simplex virus (HSV) meningitis. The availability of this technique has confirmed HSV as the cause of the recurrent Mollaret meningitis. This technique also has been applied to the diagnosis of enteroviral infections and the other herpesvirus infections. The PCR assay for enteroviruses has been demonstrated to be substantially more sensitive than culture and is 94% to 100% specific.

A multiplex PCR panel that identifies 14 pathogens (E coli K1, H influenzae, L monocytogenes, N meningitidis, S agalactiae, S pneumoniae, cytomegalovirus, enterovirus, herpes simplex virus 1, herpes simplex virus 2, human herpes virus 6, human parechovirus, varicella zoster, C neoformans/Cryptococcus gattii) in 1 hour with 0.2 mL of CSF is now widely available.

Perform serologic tests to detect syphilis. Screening for syphilis is done with the nontreponemal tests: rapid plasma reagent (RPR) or Venereal Disease Research Laboratory (VDRL). Positive results are confirmed with 1 of the following specific treponemal tests:

• Fluorescent treponemal antibody absorption (FTA-Abs)

• T pallidum hemagglutination (TPHA)

• Microhemagglutination– T pallidum (MHA-TP)

• The newer immune-capture enzyme immunoassay (ICE Syphilis) recombinant antigen test

In patients with syphilis, initial results on nontreponemal tests can serve as a baseline for gauging the success of therapy. Titers decrease and usually revert to negative or undetectable levels after effective treatment.

Increasing data suggest that serum procalcitonin (PCT) levels can be used as a guide to distinguish between bacterial and aseptic meningitis in children. Elevated serum PCT levels predict bacterial meningitis. The results of serum PCT testing, combined with other findings, could be helpful in making clinical decisions. ^[29]

In an analysis of retrospective, multicenter, hospital-based cohort studies, Dubos et al confirmed that measurement of the PCT level is the best biologic marker for differentiating bacterial meningitis from aseptic meningitis in children in the emergency department. With a threshold of 0.5 ng/mL, the sensitivity and specificity of the PCT level in distinguishing between bacterial and aseptic meningitis were 99% and 83%, respectively. ^[29]

Elevated opening pressure correlates with increased risk for morbidity and mortality in bacterial and fungal meningitis. In bacterial meningitis, elevated opening pressure (reference range, 80-200 mm H₂O) suggests increased intracranial pressure (ICP) from cerebral edema. In viral meningitis, the opening pressure usually is within the reference range. The CSF opening pressure may be elevated at times in cryptococcal meningitis, suggesting increased ICP, and it usually is elevated in tuberculous meningitis.

The CSF cell count varies according to the offending pathogen (see Tables 6 and 7 below). It usually is in the few hundreds (100-1000/µL) with a predominance of lymphocytes in patients with viral meningitis. Some cases of echovirus, mumps, and HSV meningitis may produce a neutrophilic picture early in the course of disease. (See Lumbar Puncture.)

Table 6. CSF Findings in Meningitis by Etiologic Agent (Open Table in a new window)

Table 7. Comparison of CSF Findings by Type of Organism (Open Table in a new window)

CFS sample handling

After drawing the CSF sample, do the following with the tubes:

• Tube 1 – Send to the chemistry laboratory for glucose and protein
• Tube 2 – Send to the hematology laboratory for a cell count with differential
• Tube 3 – Send to the microbiology and immunology laboratory
• Tube 4 – Hold for a repeat cell count with differential, if needed (or for other subsequent studies not initially ordered)

Microbiology and immunology studies for tube 3 include the following:

• Gram stain
• Bacterial culture
• Acid-fast bacillus (AFB) stain and tuberculosis cultures
• India ink stain
• Cryptococcal antigen testing
• Fungal cultures, counterimmunoelectrophoresis (CIE), VDRL, and cryptococcal antigen, if indicated

CSF Glucose and Protein

• In bacterial meningitis, the CSF glucose level (reference range, 40-70 mg/dL) is less than 40 mg/dL in 60% of patients. A simultaneous blood glucose determination should be obtained for the purposes of comparison. In patients with elevated blood glucose levels as a result of diabetes mellitus, the CSF-to-blood glucose ratio may not be predictive. The CSF glucose level is usually within the reference range in viral meningitis, but it may be low in some cases of LCM, HSV, mumps virus, or poliovirus infection.
• The CSF protein level (reference range, 20-50 mg/dL) usually is elevated in bacterial meningitis. In viral meningitis, these levels also usually are elevated, though they can be within the reference range. In syphilitic meningitis, abnormal CSF protein levels (elevated) and CSF glucose levels (decreased) may be observed in 10-70% of cases.

CSF Gram stain and acid-fast bacillus stain

Gram staining of the CSF permits rapid identification of the bacterial cause in 60-90% of patients with bacterial meningitis. The presence of bacteria is 100% specific, but the sensitivity of this test for detection is variable. The likelihood of detection is higher in the presence of a higher bacterial concentration and diminishes with prior antibiotic use.

The demonstration of AFB (eg, with auramine-rhodamine stain, Ziehl-Neelsen stain, or Kinyoun stain) in the CSF is difficult and usually requires a large volume of CSF. Meningeal biopsy, with the demonstration of caseating granulomas and AFB on the smear, offers a higher yield than the CSF AFB smear.

CSF culture and antigen testing

CSF bacterial cultures yield the bacterial cause in 70-85% of cases. The yield diminishes by 20% in patients who have received antimicrobial therapy. In these cases, some experts advocate the use of a CSF bacterial antigen assay. This is a latex agglutination technique that can detect the antigens of H influenzae type B (Hib), S pneumoniae, N meningitidis, E coli K1, and S agalactiae (group B streptococcus [GBS]). Its theoretical advantage is the detection of the bacterial antigens even after microbial killing, as is observed after antibacterial therapy.

Another attractive alternative is using the Binax NOW for S pneumoniae in the CSF. This assay has a 99-100% sensitivity and specificity for ruling out the most common cause of bacterial meningitis. ^[30]

Others studies, however, have shown that the CSF bacterial antigen assay may not be better than the Gram stain. Although it is specific (a positive result indicates a diagnosis of bacterial meningitis), a negative finding on the bacterial antigen test does not rule out meningitis (50-95% sensitivity).

Cryptococcal meningitis

C neoformans may be cultured from the CSF in cryptococcal meningitis. Other methods of identification include India ink preparation and the detection of CSF cryptococcal antigen. India ink has a sensitivity of only 50%, but it is highly diagnostic if positive.

Because of the low sensitivity of the India ink preparation, many centers have adapted the use of CSF cryptococcal antigen determination, a test with a sensitivity exceeding 90%. However, the CSF cryptococcal antigen determination is not universally available.

In instances when the India ink results are negative but the degree of clinical suspicion for cryptococcal meningitis is high, the CSF specimen may be sent to reference laboratories that can perform CSF cryptococcal antigen determination to confirm the diagnosis. In addition, the titer of the antigen could serve to monitor the response to treatment. Blood cultures and serum cryptococcal antigen should be obtained to determine whether cryptococcal fungemia is present.

Syphilitic meningitis

In syphilitic meningitis, isolating T pallidum from the CSF is extremely difficult and time-consuming. The spirochete could be demonstrated by using dark-field or phase-contrast microscopy on specimens collected from skin lesions (eg, chancres and other syphilitic lesions).

The diagnosis usually is supported by the CSF VDRL test, which has a sensitivity of 30-70% (a negative result on the CSF VDRL test does not rule out syphilitic meningitis) and a high specificity (a positive test result suggests the disease). Care always must be taken not to contaminate the CSF with blood during spinal fluid collection (eg, traumatic tap).

When there is clinical suspicion of neurosyphilis, CSF also can be sent for CSF TP-PA or CSF FTA -ABS. CSF TP-PA and CSF FTA-ABS demonstrated similar sensitivity and specificity. In 3 studies of patients with definitive neurosyphilis (reactive CSF VDRL), the sensitivity of CSF FTA-ABS was 90.9% to 100%. For CSF TP-PA, 1 study reported a sensitivity of 100% for the CSF TP-PA. The other 3 studies reported a sensitivity of 75.6% to 95.0%. However, diagnostic criteria of the included studies were diverse and included various combinations of signs/symptoms with abnormal white blood cell count/protein and/or reactive CSF VDRL. As T pallidum IgG can cross the intact blood–CSF barrier, reactive treponemal tests in the CSF are not specific for the diagnosis of neurosyphilis. Although the CSF TP-PA and CSF FTA-ABS demonstrated similar sensitivity and specificity, Harding et al found that a negative CSF treponemal test may not rule out neurosyphilis among patients with a high pretest probability (patients with syphilis and neurologic symptoms).67 Therefore, CSF treponemal tests have limitations with both sensitivity and specificity, and results need to be evaluated within the context of the clinical scenario, additional CSF testing (eg, VDRL, cell count, protein), and syphilis prevalence.27

Lyme meningitis

CSF culture for B burgdorferi has a low yield. The CSF Lyme PCR assay, if available, offers a rapid, sensitive, and specific method of diagnosis. This assay is gaining popularity as the method of choice for diagnosing Lyme meningitis.

Cohn et al validated a clinical prediction rule for differentiating Lyme meningitis from aseptic meningitis. Their “rule of 7s” classifies children at low risk for Lyme meningitis when all of the following three criteria are met ^[31] :

• < 7 days of headache
• < 70% CSF mononuclear cells
• Absence of cranial nerve VII palsy or other cranial nerve palsy

Tuberculous meningitis

Culture for Mycobacterium usually takes several weeks and may delay definitive diagnosis. M tuberculosis detection assays involving nucleic acid amplification have become available and have the advantages of rapidity, high sensitivity, and high specificity. There remains a need for mycobacterial growth in cultures because this method offers the advantage of performing drug susceptibility assays.

Viral isolation from CSF

Isolation of viruses from the CSF has a sensitivity of 65% to 70% for enteroviruses. Alternatively, isolation of enteroviruses from throat and stool viral cultures may also indirectly implicate enterovirus as the cause of the meningitis. Culture of mumps virus from the CSF has a low sensitivity (30%-50%). LCM virus may be cultured in blood early in the disease or in urine at a later stage.

CSF characteristics of acute bacterial meningitis

Examination of the CSF in patients with acute bacterial meningitis reveals the characteristic neutrophilic pleocytosis (cell count usually ranging from hundreds to a few thousand, with >80% PMNs). In some (25%-30%) cases of L monocytogenes meningitis, a lymphocytic predominance may occur. A low CSF white blood cell (WBC) count (< 20/µL) in the presence of a high bacterial load suggests a poor prognosis.

According to Seupaul, the following three findings on CSF analysis have clinically useful likelihood ratios for the diagnosis of bacterial meningitis in adults ^[32] :

• CSF glucose−to−blood glucose ratio of 0.4 or lower
• CSF WBC count of 500/µL or higher
• CSF lactate level of 31.53 mg/dL or higher

CSF characteristics of viral meningitis

Viral central nervous system (CNS) infections are typically characterized by cerebrospinal fluid (CSF) pleocytosis with lymphocytic or monocytic predominance. However in one study Enterovirus infections were the cause of 64% of neutrophil-predominant CSF and 33% of lymphocyte-predominant CSF (p <  0.001), while herpes infections were the cause of 46% of lymphocytic pleocytosis and 20% of neutrophilic pleocytosis. ^[33]  Usually the CSF becomes lymphocytic after the first 24 to 48 hrs. In another study a large proportion of patients with serologically confirmed West Nile meningitis were found to have at least 50% PMNs in their initial CSF. ^[34]

In viral meningitis, the opening pressure is 90 mm H₂O to 200 mm H₂O, and the WBC count is 10µL to 300/µL. Although the glucose concentration is typically normal, it can be below normal in meningitis from lymphocytic choriomeningitis virus (LCM), herpes simplex virus (HSV), mumps virus, and poliovirus. The protein concentration tends to be slightly elevated, but it can be within the reference range.

CSF characteristics of fungal meningitis

The diagnosis of cryptococcal meningitis relies on the identification of the pathogen in the CSF. The CSF is characterized by a lymphocytic pleocytosis (10-200/µL), a reduced glucose level, and an elevated protein level. The CSF picture of other fungal meningitides is similar to that of cryptococcal meningitis, usually with lymphocytic pleocytosis. Eosinophilic pleocytosis has been associated with C immitis meningitis.

The definitive diagnosis usually relies on the demonstration of the specific fungal agent (eg, H capsulatum, C immitis, B dermatitidis, or Candida species) from clinical specimens, including the CSF. This could be in the form of fungal culture isolation (eg, C albicans growth from CSF).

More commonly, fungal serology (eg, presence of histoplasma antigen in the CSF) is used in the diagnosis of many cases of fungal meningitis because isolating these organisms from culture has proved difficult. It should be noted, however, that the serology for B dermatitidis is not accurate and a negative serology finding does not rule out the diagnosis.

A test used to detect fungal infection in the blood was successfully used in the diagnosis of fungal meningitis in an outbreak caused by contaminated steroids. ^[35]  This outbreak involved 13,534 US patients who underwent epidural steroid injection and were exposed to methylprednisolone acetate from lots contaminated with environmental fungi; hundreds of these individuals developed serious CNS complications. The test (Fungitell, Beacon Diagnostics Laboratories), which measures levels of b-D-glucan (a glycoprotein found in the fungal cell wall), was used in CSF samples from patients exposed to the contaminated steroids who had negative fungal culture and polymerase chain reaction results. All patients with fungal meningitis had detectable b-D-glucan in their CSF. ^[35]

CSF characteristics of eosinophilic/parasitic meningitis

Primary amebic meningoencephalitis (PAM) caused by N fowleri is characterized by a neutrophilic pleocytosis, low glucose levels, elevated protein levels, and red blood cells (RBCs). Mononuclear pleocytosis may be observed in patients with subacute or chronic forms of PAM. Demonstration of the trophozoites, with the characteristic ameboid movement, on wet preparations of the CSF has been used for diagnosis. Alternatively, the amoeba may be demonstrated in biopsy specimens.

In the presence of exposure, profound peripheral blood eosinophilia, and characteristic eosinophilic pleocytosis, suspicion of meningitis caused by A cantonensis, G spinigerum, or B procyonis should be entertained. Demonstrating the larvae ante mortem is usually difficult, and diagnosis relies on clinical presentation and a compatible epidemiologic history. Serologic tests may aid in the diagnosis. G spinigerum meningitis may mimic cerebrovascular disease in that it may cause cerebral hemorrhage.

CSF characteristics of Lyme meningitis

In patients with Lyme meningitis, the CSF is characterized by low-grade lymphocytic pleocytosis, low glucose levels, and elevated protein levels. Oligoclonal bands reactive to B burgdorferi antigens may be present. Demonstration of the specific antibody to B burgdorferi aids in the diagnosis. Laboratory confirmation of LNB is hampered by the low yield of bacterial culture and the poor sensitivity of polymerase chain reaction (PCR) in the cerebrospinal fluid (CSF). According to the case definition of the European Federation of Neurological Societies (EFNS), intrathecal production of antibodies against Borrelia burgdorferi must be proven to confirm the diagnosis of definite LNB.Therefore determination of Borrelia-specific antibody index (AI) in the CSF has become the traditional diagnostic gold standard, Comparison between the antibody response in the CSF and that in the serum is a helpful diagnostic test. A CSF-to-serum ratio greater than 1 suggests intrathecal antibody production and neuroborreliosis. However up to 20% of patients with proven Borrelia burgdorferi infection have negative AI. In such cases when suspicion of neuroborreliosis persists, a seroconversion in serum has to be demonstrated after 6 weeks to confirm the diagnosis of LNB. Furthermore, early antibiotic therapy can also affect the humoral response and IgM-IgG switch resulting in negative-specific AI and negative serum IgG. ^[28]

Central nervous system (CNS) invasion by Borrelia burgdorferi (Bb) induces local production of the B-cell chemoattractant CXCL13, triggering lymphocyte migration across the blood–brain barrier, in turn leading to intrathecal production of Bb-specific antibodies (ITAb). Intrathecal CXCL13 and IgG production are closely interrelated. In one study, CSF CXCL13 was highly elevated in all patients with untreated acute LNB compared with that in the patients without LNB. At a cutoff of 1229 pg/mL, the sensitivity of CXCL13 was 94.1%, Only seven patients (five with a CNS lymphoma and two with bacterial meningitis) had a CXCL13 level above the cutoff, resulting in a specificity of 96.1%. ^[32]  This high sensitivity and specificity was seen in pediatric patients as well. ^{[28, 36]}

In a recent study, CXCL13 was disproportionately increased in “definite LNB,” defined as having demonstrable Borrelia-specific ITAb, but not “probable LNB,” without ITAb. This disproportionate increase may help identify patients with very early infection or those with active vs treated LNB, or may help to differentiate ITAb-defined active LNB from other neuroinflammatory disorders. However, its reported specificity is closely related to the diagnostic requirement for ITAb. It may add little specificity to the demonstration of a pleocytosis or increased overall or specific IgG production in the CSF. ^[37]

CSF characteristics of Tuberculous Meningitis

In patients with tuberculous meningitis, the CSF is characterized by a predominantly lymphocytic pleocytosis; an elevated protein level, especially if a CSF block is present; and a low glucose level (< 40 mg/dL). CSF sample should be sent for acid-fast smear with the important caveat that a single sample has low sensitivity, on the order of 20% to 40%. Several daily large volume (10–15 mL) lumbar punctures are often needed for a microbiologic diagnosis; sensitivity increases to >85% when four spinal taps are performed. ^[38]  PCR testing can provide a rapid diagnosis, though false-negative results may occur in samples containing very few organisms (< 2 colony-forming units [cfu]/mL. In one study of 72 patients with clinically suspected TBM multiplex PCR showed sensitivity, specificity, positive predictive value, and negative predictive value of 71.4%, 89.6%, 83.3%, and 81.2%, respectively, in the diagnosis of TBM. ^[32]

CSF can be sent for Xpert MTB/RIF Ultra in patients suspected of tubercular meningitis. In one study of HIV infected patients with meningitis Xpert Ultra detected significantly more tuberculous meningitis than did either Xpert or culture. Xpert Ultra had higher sensitivity of 95% than either Xpert (45%) or culture (45%) for definite tuberculous meningitis. Based on the consensus clinical case definition, Xpert Ultra found 70% sensitivity for probable or definite tuberculous meningitis. Testing 6 mL or more of CSF was associated with more frequent detection of tuberculosis than with less than 6 mL (26% vs 7%; p = 0·014). ^[35]

Computed tomography (CT) of the head and magnetic resonance imaging (MRI) of the brain generally do not aid in the diagnosis of meningitis. Imaging may be useful in finding complications of meningitis and in determining parameningeal causes of abnormal CSF . Some patients may show meningeal enhancement, but its absence does not rule out the condition. Routinely obtaining CT scans of the head may lead to unnecessary delay in the performance of diagnostic lumbar puncture and the initiation of antibiotic therapy; the latter may be detrimental to the outcome in these patients.

Cerebral herniation following the lumbar tap procedure is rare in individuals with no focal neurologic deficits and no evidence of increased ICP. If it occurs, it usually does so within 24 hours after the LP; thus, herniation should always be considered in the differential diagnosis if the patient’s neurologic status deteriorates during that time frame.

According to the Infectious Diseases Society of America guidelines, the following are indications for screening head CT before LP in adult patients ^[26] :

• Immunocompromised state
• History of CNS disease (eg, mass lesion, stroke, or focal infection)
• Seizure within 1 week of presentation
• Papilledema
• Abnormal level of consciousness
• Focal neurologic deficit (eg, dilated nonreactive pupil, gaze palsy, or arm or leg drift)

In patients with suspected bacterial meningitis, blood cultures should be obtained and treatment initiated before imaging studies and LP. Neuroimaging may yield normal results or demonstrate small ventricles, effacement of sulci, and contrast enhancement over convexities (see the images below). Late findings include venous infarction and communicating hydrocephalus. Brain abscess, sinus or mastoid infection, skull fracture, and congenital anomalies must be ruled out. (See Imaging in Bacterial Meningitis.)

Acute bacterial meningitis. This axial nonenhanced computed tomography scan shows mild ventriculomegaly and sulcal effacement.

Acute bacterial meningitis. This axial T2-weighted magnetic resonance image shows only mild ventriculomegaly.

Acute bacterial meningitis. This contrast-enhanced, axial T1-weighted magnetic resonance image shows leptomeningeal enhancement (arrows).

Finally, neuroimaging studies are helpful in the detection of CNS complications of bacterial meningitis, such as the following (see the images below):

• Hydrocephalus
• Cerebral infarct
• Brain abscess
• Subdural empyema
• Venous sinus thrombosis

Chronic mastoiditis and epidural empyema in a patient with bacterial meningitis. This axial computed tomography scan shows sclerosis of the temporal bone (chronic mastoiditis), an adjacent epidural empyema with marked dural enhancement (arrow), and the absence of left mastoid air.

Subdural empyema and arterial infarct in a patient with bacterial meningitis. This contrast-enhanced axial computed tomography scan shows left-sided parenchymal hypoattenuation in the middle cerebral artery territory, with marked herniation and a prominent subdural empyema.

Thalamic and basal ganglion involvement is seen in respiratory viruses esp in children Creutzfeldt-Jakob disease, arbovirus, and Mycobacterium tuberculosis.

Magnetization transfer MRI has been proposed as a useful tool in the diagnosis of tuberculous meningitis. Visibility of the meninges on precontrast T1-weighted magnetization transfer images may be considered highly suggestive of tuberculous meningitis. 

In Lyme disease, multifocal nonenhancing patchy lesions on T2 WI can be seen. Neurobrucellosis shows a wide spectrum of imaging findings from normal to nonspecific findings of inflammation of CNS and nerve roots or vascular complications

 The finding of rhombencephalitis may point to listeria monocytogenes as the causative agent. ^[31]

In cryptococcal meningoencephalitis, diffuse meningeal enhancement and also ventriculitis can be seen on MRI. Typical findings are multiple punctuate lesions, often in the basal ganglia. These are characteristic cystic lesions due to cryptococcal invasion of the Virchow-Robin-spaces. They are termed “soap bubble lesions” and allow the quick provisional diagnosis. ^[39]

Typical radiological MRI findings in Herpes Simplex encephalitis are the presence of asymmetrical changes in signal intensities in the mesial temporal lobes, inferior frontal lobes, and insula. ^[39]

It has been proposed that CSF lactate may be a good marker that can differentiate bacterial meningitis (> 6 mmol/l), from partially treated meningitis (4 to 6 mmol/l) and aseptic meningitis (< 2 mmol/l). However, other researchers have suggested that CSF lactate offers no additional clinically useful information over conventional CSF. ^[22]

CSF lactate concentration depends largely on its production from central nervous system (CNS) glycolysis. Its value is independent of blood lactate, probably because lactate in its ionised state crosses the blood– CSF barrier very slowly. It should reach the laboratory promptly following sampling (ideally within 60 min) and should be frozen if analysis is to be delayed for >24 hours, as otherwise the result can be spuriously elevated. ^[40]

Several prospective and retrospective studies, and a well-designed meta-analysis, have shown that a CSF lactate of ≥3.5 mmol/L has a high sensitivity (96%–99%) and specificity (88%–94%) for distinguishing acute bacterial meningitis from acute viral meningitis. Testing for CSF lactate should be done on CSF samples obtained before giving antibiotics, as its sensitivity drops significantly (to less than 50%) after they are started. ^{[22, 25]}

Tumor necrosis factor alpha (TNF-α), interleukin (IL)-1, and other cytokines have received increasing attention as mediators of the inflammatory response during bacterial meningitis. Leist et al reported detecting TNF-α in the CSF of three of three patients with bacterial meningitis, but in zero of seven patients with viral meningitis. Lopez-Cortez et al demonstrated that a TNF-α level higher than 150 pg/mL and an IL-1β level higher than 90 pg/mL showed sensitivities of 74% and 90%, respectively, in discriminating viral from aseptic meningitis.

Mustafa et al demonstrated that IL-1β can be detected in the CSF of 95% of infants and children with bacterial meningitis and that levels higher than 500 pg/mL were correlated with an increased risk of neurologic sequelae. ^[37]

These findings, though requiring both confirmation and amplification, suggest that analysis of TNF and other cytokines, in particular IL-1β, may prove valuable in differentiating acute bacterial meningitis from viral meningitis and possibly in detecting patients at particular risk for an adverse outcome. Their role in guiding adjunctive therapy, such as corticosteroids and nonsteroidal treatment of blood-brain barrier injury, is also under investigation.