You are in: eMedicine Specialties > Neurology > Specialized Neurodiagnostic Tests Lumbar Puncture (CSF Examination)Article Last Updated: Apr 24, 2006AUTHOR AND EDITOR INFORMATIONSection 1 of 8
  Author: Roy Sucholeiki, MD, Medical Director, Comprehensive Seizure and Epilepsy Program, The Neurosciences Institute at Central DuPage Hospital Roy Sucholeiki is a member of the following medical societies: American Academy of Neurology, American Epilepsy Society, and American Neuropsychiatric Association Editors: Ramon Diaz-Arrastia, MD, PhD, Assistant Professor, Department of Neurology, Comprehensive Epilepsy Center, University of Texas Southwestern; Francisco Talavera, PharmD, PhD, Senior Pharmacy Editor, eMedicine; Glenn Lopate, MD, Associate Professor, Department of Neurology, Division of Neuromuscular Diseases, Washington University School of Medicine; Chief of Neurology, St Louis ConnectCare, Consulting Staff, Barnes Jewish Hospital; Matthew J Baker, MD, Consulting Staff, Collier Neurologic Specialists, Naples Community Hospital; Nicholas Y Lorenzo, MD, Chief Editor, eMedicine Neurology; Consulting Staff, Neurology Specialists and Consultants Author and Editor Disclosure Synonyms and related keywords: CSF analysis, cerebrospinal fluid examination, LP, intracranial mass lesions, intracranial pressure, ICP, Lyme disease, multiple sclerosis, Guillain-Barré syndrome, CNS vasculitis, CNS infections, subarachnoid hemorrhage, leptomeningeal carcinomatosis, benign intracranial hypertension, pseudotumor cerebri, xanthochromia, sentinel hemorrhages, INTRODUCTIONSection 2 of 8
  Cerebrospinal fluid (CSF) was first examined in the 19th century using primitive techniques (eg, sharpened bird quills). CSF analysis reached a peak in the 1950s and early 1960s, when almost no workup of a significant central nervous system (CNS) problem was performed without a lumbar puncture (LP). With the advent of sophisticated imaging techniques, particularly computerized tomography (CT) and magnetic resonance imaging (MRI), LP is no longer an important test in the diagnosis of most intracranial mass lesions. This is especially true with the potential risk of brain herniation if intracranial pressure (ICP) is increased markedly. LP remains a critical procedure in the diagnosis of CNS infections and inflammatory diseases. Diagnostic and therapeutic uses The past 2 decades have seen an increase in the sophistication of CSF analysis allowing immunologic confirmation of certain infections (eg, Lyme disease) and fractionation of CSF proteins (eg, myelin basic protein). CSF analysis remains important in the diagnosis of infections (eg, bacterial, mycobacterial, fungal, viral, protozoan) and certain inflammatory diseases (eg, multiple sclerosis, Guillain-Barre syndrome, vasculitis). Moreover, CSF analysis can be helpful in the diagnosis of unusual illnesses, such as pediatric neurotransmitter diseases, as well as particular inborn errors of metabolism (with normal serum and urine analysis) that can cause infantile epilepsy. CSF analysis also is an important diagnostic tool in subarachnoid hemorrhage and leptomeningeal carcinomatosis. Indeed, the extent to which CSF reflects the chemistry of surrounding tissue, and the significant differences in composition from plasma, are reviewed elegantly by Hochwald in his chapter in Clinical Neurology. LP itself can be therapeutic, particularly in benign intracranial hypertension (ie, pseudotumor cerebri), in which serial LPs may be used for treatment. It is used as an access method, most commonly for spinal anesthesia, but also for introduction of radiopaque contrast media (eg, myelography), corticosteroids, antibiotics, and chemotherapeutic agents. However, this article concentrates on the primary diagnostic uses of LP and the examination of CSF. For excellent patient education resources, visit eMedicine's Procedures Center. Also, see eMedicine's patient education article Spinal Tap. LUMBAR PUNCTURE TECHNIQUESection 3 of 8
LP is performed in the interspaces between the lumbar vertebrae, usually at the L4-L5 level. In unusual circumstances, a tap can be done at higher levels. Even at these higher levels the probability of injuring the spinal cord is small. Nevertheless, this should be reserved for situations in which access to the usual sites has been obliterated (eg, by extensive orthopedic fusion) and a specific need exists for the information obtained from the procedure. The spinal cord typically ends at the L1 level in adults (slightly lower in children). If consideration is being given to an LP above the L4-L5 level, it should be performed with subspecialist assistance and with great caution. Fluoroscopic guidance may be helpful in these and other difficult situations. A "clean" specimen obtained fluoroscopically is more helpful than a "traumatic" (ie, blood contaminated) specimen obtained at the bedside. CSF also can be obtained from the cisterna magna by a tap below the external occipital protuberance. However, this technique should rarely, if ever, be required and again should be performed with subspecialist assistance. Preparation The spinal needle used today is disposable. The clinician should inspect the spinal needle for any defects. A 20-gauge needle for adults, or a 22-gauge needle for children, is used typically. Various designs have been marketed, with the aims of increasing ease of use and reducing incidence of post-LP dural leak that could result in spinal headache. Overall, using atraumatic needle (vs cutting type) and smaller needle size is recommended, as these decrease the chance of post–lumbar puncture headache. After carefully explaining the procedure to the patient and/or responsible caregiver, including the risks and benefits of the procedure, the most important determinant in successfully obtaining a CSF specimen is patient positioning. This may require the assistance of a nurse and/or other paramedical personnel. Sedative medication may be required in children or in the confused or combative patient. The bed/gurney should be flat, and the patient should be lying on his/her side (horizontal lateral decubitus position). The patient's body needs to be perfectly perpendicular to the bed/gurney. The patient should assume the fetal position (ie, head/neck, arms, legs flexed as much as possible). The apex of the pelvic bone should be identified and a direct line should be visualized to the spine. The location so identified should be well below the tip of the conus. Two spinous processes in this area (ie, L4 and L5 levels) should be identified by palpation. Procedure Local anesthetic should be infiltrated and then the area should be prepared carefully and draped. The spinal needle then is positioned between the 2 spinous processes already identified and introduced into the skin with the bevel of the needle facing up. The needle should be advanced slowly at a slightly upward angle (ie, toward the patient's head). Accurate placement of the needle is rewarded by a slight "give" and flow of fluid, which normally is clear and colorless. Some resistance may be noted, usually in patients who have some thickening of the dura (eg, elderly patient), but if a feeling of hitting bone is noted, this is probably what is occurring. As with any procedure, experience improves the success rate. One of the primary goals is to prevent the introduction of blood into the CSF sample. A measurement of opening pressure should be attempted, unless the patient is so uncooperative as to invalidate the reading. CSF pressure should be measured with the subject in the horizontal lateral decubitus position (as described previously) and relaxed as much as possible. Normal range is 80-180 mm H2O, with small, visible excursions related to respiration and pulse. In cases of extremely high pressure (eg, 7300 mm H2O) the smallest sample possible (for the required testing) should be removed, followed by consideration of CSF pressure-lowering treatment, with continuous monitoring of the pressure until it decreases significantly. Although radiographic screening identifies patients with intracranial mass lesions (with possible associated increased ICP), ICP increases may be due to cerebral edema secondary to causes such as infection, tumor cell infiltration, and benign intracranial hypertension. Detection of increased CSF pressure (ie, >220 mm H2O) permits treatment (eg, mannitol, corticosteroids, hyperventilation) even before the etiology is identified. Commercially available LP sets contain a manometer and stopcock for this purpose. Following determination of CSF pressure, CSF samples should be obtained. Only a few milliliters are needed for basic studies (eg, protein, glucose). Specialized tests that require concentration of the CSF (eg, cell count, specific antibody studies) require more CSF. In the event of a very low pressure, the question of spinal block may need to be addressed. In 1916, Queckenstedt described a technique for compression of the jugular veins (by an assistant) with concomitant rise in CSF pressure if the subarachnoid space is open. Although this technique is potentially useful, more frequently spinal block is confirmed and its cause determined by standard myelography. Important points If the opening pressure is high, the specimens still should be collected, because any change in intracranial dynamics caused by the LP has occurred already. Premature needle removal without collecting CSF will not change this situation. If the fluid appears to be bloody, several specimens should be collected. If the blood clears in successive tubes then the blood, at least in part, was traumatic in origin. Unfortunately, this sign is neither specific nor sensitive, as in some traumatic taps the amount of blood increases in subsequent tubes. In cases in which the clarity of the CSF is in doubt (eg, a hazy appearance can be produced by either RBCs or WBCs), a tube of CSF should be compared with a tube of water against a fluorescent light bulb or X-ray view box. When sufficient fluid is obtained, the needle is withdrawn and a dry sterile dressing applied to the puncture site. Prolonged compression of the site, or keeping the patient supine for an extended period, has not been proven to reduce the incidence of spinal headache. Spinal headache is characterized by pulsatile head pain, with or without nausea, relieved by lying down and aggravated by standing and Valsalva maneuvers. It is self-limited but may last up to a week (or rarely longer). The placement of an epidural blood patch using the patient's own venous blood often corrects this problem. However, the need for a blood patch is uncommon. The use of intravenous caffeine benzoate (500 mg infusion over 1 h) also has been found to treat post-LP headaches effectively in double-blind, controlled trials. CSF ANALYSISSection 4 of 8
Separate specimens should be sent for microscopic study and for centrifugation. The latter must be done promptly, as RBCs hemolyze within a few hours. Normal CSF may contain as many as 5 lymphocytes per cubic millimeter. A larger-than-usual number of WBCs suggests infection or, more rarely, leukemic infiltration. While bacterial infections traditionally are associated with a preponderance of polymorphonuclear leukocytes (PMNs), many cases of viral meningitis/encephalitis also have a high percentage of PMNs in the acute phase of the illness (when in fact, most LPs are done). In addition, inflammation from any source (eg, CNS vasculitis) can raise the WBC count. A traumatic tap will, of course, introduce both WBCs and RBCs into the CSF. An approximation of 1 WBC per 1000 RBCs can be made, although a repeat tap may be preferable. While no normal value for RBCs in the CSF is known, an occasional RBC may be incident to the tap. Xanthochromia The best way to distinguish RBCs related to intracranial bleeding is examination of the centrifuged supernatant CSF for xanthochromia (yellow color). Although xanthochromia can be confirmed visually, it is identified and quantified more accurately in the laboratory. While xanthochromia can be produced by spillover from a very high serum bilirubin level (ie, >15 mg/dL), patients with severe hyperbilirubinemia usually have been identified prior to the LP (eg, jaundice, known liver disease). With this exception, the presence of xanthochromia in a freshly spun specimen is evidence of preexistent blood in the subarachnoid space. However, note that an extremely high CSF protein level, as seen in LPs below a complete spinal block, also renders the fluid xanthochromic, though without RBCs. Xanthochromia can persist up to several weeks following a subarachnoid hemorrhage (SAH). Thus it has greater diagnostic sensitivity than a CT scan of the head without contrast, especially if the SAH has occurred more than 3-4 days prior to presentation. Patients with aneurysmal leaks (ie, sentinel hemorrhages) may present days after headache onset, increasing the likelihood of a false-negative head CT scan. In some cases, the CSF may be another color that strongly suggests a diagnosis. For example, pseudomonal meningitis may be associated with bright green CSF. Other tests Assuming the CSF has been collected under sterile conditions, microbiologic studies can be performed. Stains, cultures, and immunoglobulin titers can be obtained. The latter are of special importance in diseases in which peripheral manifestations fade while CNS symptoms persist (eg, syphilis, Lyme disease). Assessment of CSF protein level, while nonspecific, can be a clue to otherwise unsuspected neurologic disease. The high protein levels in demyelinating polyneuropathies, or postinfectious states, can be informative. A traumatic tap can introduce protein into the CSF. An approximation of 1 mg of protein per 750 RBCs may be used, although a repeat tap is preferable. CSF glucose level normally approximates 60% of the peripheral blood glucose level at the time of the tap. A simultaneous measurement of blood glucose (especially if the CSF glucose level is likely to be low) is recommended. Low CSF glucose level usually is associated with bacterial infection (probably due to enzymatic inhibition rather that actual bacterial consumption of the glucose). It also is seen in tumor infiltration, and may be one of the hallmarks of meningeal carcinomatosis, even with negative cytologic findings. High CSF glucose level has no specific diagnostic significance and is most often spillover from elevated blood glucose level. Leptomeningeal malignancies: Multiple LP examinations may be required in this situation. At least 3 negative cytologic evaluations (ie, 3 separate samplings) are required to rule out leptomeningeal malignancy (eg, leptomeningeal carcinomatosis). RISKS OF LUMBAR PUNCTURESection 5 of 8
The physician should have a detailed discussion with the patient and/or the caregivers concerning the risks/benefits of the LP procedure. Significant risks include (but are not limited to) the following:
CONCLUSIONSSection 6 of 8
The tests described here constitute the basic CSF analysis. Virtually any CSF analysis should include cell counts, examination for xanthochromia, and protein and glucose studies. Beyond this, culture, serologic tests for syphilis, Lyme titer, electrophoretic pattern, myelin basic protein (for multiple sclerosis), cryptococcal antigen, angiotensin-converting enzyme level (ie, in suspected sarcoidosis), bacterial stains, and cytologic studies should be ordered on an individual basis. Because a CSF specimen is comparatively difficult to obtain, a bit more tolerance for more tests usually is allowed. In summary, LP and CSF examination, while their indications have been reduced, remain indispensable tools in the armamentarium of neurologic diagnosis. ACKNOWLEDGMENTSSection 7 of 8
The authors and editors of eMedicine gratefully acknowledge the contributions of previous author Arthur Waldman, MD to the development and writing of this article. REFERENCESSection 8 of 8
Lumbar Puncture (CSF Examination) excerpt Article Last Updated: Apr 24, 2006 |
