CT is not indicated before LP

View The Full Pathway

Lumbar puncture with CSF analysis is the mainstay of diagnosis for meningitis and encephalitis. Lumbar puncture should be performed with the patient in the lateral decubitus position if measurement of opening pressure is desired, but for some patients, it may be performed more easily in an upright position.

Four tubes of CSF should be collected. Tubes 1 and 4 are used for red blood cell (RBC) and white blood cell (WBC) counts. This allows for differentiating traumatic tap from subarachnoid hemorrhage, which will be on the differential diagnosis of many of these patients. Protein, glucose, and lactic acid concentrations as well as polymerase chain reaction (PCR) tests are conducted on fluid from tube 2. Tube 3 is used for Gram stain, specific antigen tests, and cultures. Tube 3 can also be used for additional PCR tests. Collecting larger volumes of CSF increases the sensitivity of several diagnostic CSF tests and ensures an adequate supply in case additional tests are needed at a later time. CSF is produced at a rate of 22 mL/hr, and 15 mL or more can be taken safely.

Normal CSF should contain zero RBCs, <5 WBCs, a CSF to serum glucose ratio >0.67, protein concentration <50 mg/dL, and no visible organisms on Gram stain. Abnormal CSF findings are detected in 88% of cases of all-cause bacterial meningitis, but can also appear deceptively normal. Normal or near-normal findings in some or all of the measured values have been found in 25% of cases of infection with L monocytogenes and 2% of other adult cases.

Opening pressure can be measured if lumbar puncture is performed with the patient in the lateral decubitus position. While opening pressure is usually normal in viral meningitis (<20 mm H2O), it is often elevated in bacterial meningitis, and is >40 cm H2O in 39% of cases.

ABM typically presents with elevated CSF WBC count, often in the range of 10 to 10,000 cells/mcL, usually >100. Cell differential usually shows a predominance of neutrophils, particularly after the first 24 hours of infection, but may be normal in early disease or in immunodeficient states. Lymphocytes may predominate in infections partially treated with antibiotics and in cases caused by L monocytogenes. Importantly, 6% of patients with culture-proven ABM do not have elevated CSF WBC counts, and 10% have lymphocyte-predominant differentials. In viral meningitis, WBC count is generally <250 cells/mcL and typically ranges between 5 cells/mcL and 1000 cells/mcL, with a large proportion of lymphocytes.

A traumatic lumbar puncture can complicate CSF cell count interpretation, as WBCs from outside the CSF will contaminate the sample. Traditionally, corrected cell counts are derived by subtracting 1 WBC for every 500 to 1500 RBCs in the sample. However, multiplying CSF RBCs by the ratio of blood WBCs and blood RBCs yields a more accurate corrected result. The formula for predicting CSF WBCs after traumatic puncture is:

Predicted CSF WBCs = CSF RBCs × (blood WBCs / blood RBCs)

Normal CSF protein concentration is 45 to 50 mg/dL. Nearly all patients with bacterial meningitis have an elevated CSF protein level, with a mean value of 135 mg/dL. Normal to mildly elevated levels are seen in viral meningitis, with a mean of 56 mg/dL.

The ratio of CSF to serum glucose levels has diagnostic value in acute meningitis when samples are measured simultaneously. The normal CSF to serum glucose ratio is >0.67. In bacterial meningitis, this value ranges from normal to significantly decreased, depending on the type of pathogen, the time since infection onset, and presence of dextrose in any intravenous (IV) fluids administered. One study found that using a cutoff point of <0.36 yielded a sensitivity and specificity of 93% for ABM, although the clinical utility of this very low value is unclear. Viral meningitis usually has a normal to mildly decreased ratio.

Gram stain of the CSF has shown a sensitivity of 60% to 99% for bacterial meningitis, with a specificity of 97% to 100%. Sensitivity is negatively impacted if lumbar puncture is performed after antibiotic administration. Thus, a negative Gram stain cannot be used to exclude bacterial meningitis.

CSF culture provides definitive diagnosis and antibiotic sensitivities, but if it is obtained after antibiotics are started, positive yield is negatively impacted. When lumbar puncture is performed >4 hours after antibiotic initiation, culture sensitivity is reduced >30%. Cultures may be sterile within 2 hours of antibiotic administration in infections caused by N meningitidis.

CSF lactate concentration can be helpful in the diagnosis of bacterial meningitis. In bacterial meningitis, this end product of cellular anaerobic metabolism increases due to bacterial production, cerebral edema, vascular inflammation, and cerebral ischemia. Because lactate crosses the blood-brain barrier very slowly, the CSF concentration is independent of the serum level. The normal CSF lactate level is <35 mg/dL. A level >35.1 mg/dL has demonstrated sensitivity of 93% and specificity of 97% for bacterial meningitis when obtained prior to administration of antibiotics. Accuracy of this test is negatively affected by antibiotics. The CSF lactate level is particularly useful in neurosurgery-related infections, as normal postoperative inflammation can alter the findings of many of the conventional markers, but does not affect lactate. CSF lactate may be elevated in fungal meningitis or viral encephalitis, although reliable diagnostic cutoff points have not been reported. CSF lactate can also be elevated due to seizure, ischemia, and hemorrhage.

CSF nucleic acid amplification tests (NAATs) are increasingly being used to detect viral or bacterial DNA and RNA in CSF. In the case of bacterial meningitis, polymerase chain reaction (PCR) testing can detect bacterial DNA even several days after antibiotic therapy has begun, with a sensitivity of 87% to 100% and specificity of 98% to 100%.27 For identification of specific etiologies of viral meningitis, PCR has become the gold standard (although this information is not essential for management of most viral cases).

Other advanced NAATs are increasingly used for rapid diagnosis of causative pathogens in CNS infections, although their value remains unsettled. Regarding cases of tubercular meningitis, traditional smear microscopy has very poor sensitivity (10%-15%). Although culture for tuberculosis has a better sensitivity at 50%, it requires 10 days to 8 weeks to result. Newer NAATs, such as the Xpert® MTB/RIF assay detect M tuberculosis with much greater speed and sensitivity. While tubercular meningitis testing has traditionally been the purview of inpatient medicine, the Xpert® MTB/RIF assay has a 2-hour turnaround time and could potentially be useful in EDs, where available.