Lumbar Puncture

Key Clinical Questions

What is the role of brain CAT scan imaging prior to lumbar puncture?
What is the best patient positioning to perform lumbar puncture?
What are important factors to consider in preventing post dural puncture headache?

Case 114-1

A 54-year-old morbidly obese man described the “worst headache” of his life and mild photophobia to an examining physician. His medical history did not reveal any risk factors for increased intracranial pressure (age > 60 years, immune-compromised state, presence of CNS disease, new onset of seizure, confusion, or symptoms suggestive of focal neurologic abnormalities). His physical examination was notable for a temperature of 100.5°F, alert and appropriate mental status, supple neck examination, and the absence of focal neurologic findings (hemiparesis, aphasia, visual field cuts, or cranial nerve palsies). His physician ordered immediate empiric antibiotics prior to a head CT scan. A proceduralist performed an ocular ultrasound which demonstrated an optic nerve sheath diameter (ONSD) of 5.8 mm in the left eye and 6.0 mm in the right eye. Because the ONSD was elevated, a CT scan of his brain without contrast was obtained and reported no acute bleed, midline shift, or mass effect.

Using ultrasound guidance and a 22 g × 124 mm Gertie Marx needle, a lumbar puncture was performed to rule out subarachnoid hemorrhage and meningitis. Because of his girth, the patient sat upright during the procedure. After return of fluid was noted, the proceduralist replaced the stylet and gently lowered the patient to the left lateral position. The proceduralist documented an opening pressure of 380 mm Hg with the patient relaxed and breathing comfortably and the presence of respiratory variation in the manometer. He removed 40 cc of colorless CSF for diagnostic evaluation and recorded a closing pressure of 170 mm Hg. The WBC count was 300 with predominant neutrophils and the RBC count was zero. The results of the spinal tap (300 WBCs/hpf with predominant neutrophils and zero RBCs/hpf) were consistent with the diagnosis of acute meningitis and the patient was admitted for continued intravenous antibiotic therapy.

Introduction

Lumbar puncture (LP) is a procedure to sample the cerebrospinal fluid (CSF) surrounding the brain and spinal cord. It can be performed on an inpatient or outpatient basis using local anesthesia. Individuals trained to perform this procedure generally include hospitalists and other internists, proceduralists, emergency physicians, pediatricians, neurologists, and radiologists. It has been estimated that 400,000 diagnostic lumbar punctures are performed in the United States annually with charges estimated between $2000 to $3000 for each uncomplicated procedure.

Physiology

The majority of the CSF is in the subarachnoid space, where the arachnoid membranes bridge the sulci of the brain, in the basal cisterns and around the spinal cord. CSF moves within the ventricles and subarachnoid spaces under the influence of hydrostatic pressure generated by the production of CSF by the choroid plexus of the lateral third and fourth ventricles. The volume of CSF in humans is 140 to 150 ml, constituting 10% to 20% of brain weight. Only 30 to 40 mL is actually in the ventricular system, with a production rate of 21 mL/hr. The turnover rate of total CSF is about 5 hours for an average sized human.

CSF cushions the brain, regulates brain extracellular fluid, allows for distribution of neuroactive substances, and collects the waste products produced by the brain.

Indications

CSF may be collected for both diagnostic and therapeutic purposes. Diagnostically, emergent examination of the CSF provides essential information in many clinical situations:

Suspected central nervous system (CNS) bacterial, fungal, and viral infections. Lumbar puncture is most commonly performed to diagnose or exclude meningitis in patients presenting with fever, altered mental status, and headache with or without meningeal signs. CSF examination is highly sensitive and specific for determining the presence of bacterial and fungal meningitis.
Suspected subarachnoid hemorrhage when brain computerized tomography (CT) is negative. Now that brain CT scanning is more readily available and more technologically sophisticated, there may be a tendency to omit the LP from the workup in the setting of a normal CT scan. It is important to remember that the sensitivity of the CT scan for determining the presence of subarachnoid blood is less than 100% mandating the need in situations with a moderate to high clinical suspicion to perform an LP prior to eliminating this possibility.

A nonemergency LP is indicated to diagnose of the following conditions: pseudotumor cerebri, carcinomatous meningitis, tuberculous meningitis, normal pressure hydrocephalus, neurosyphilis, and vasculitis. LP is also useful but not diagnostic in multiple sclerosis and Guillain-Barré syndrome.

Therapeutically, an LP may be performed to remove excessive fluid or to lower intrathecal pressure as in cryptococcal meningitis or pseudotumor cerebri. It may also be performed to administer medication such as intrathecal chemotherapy, intrathecal antibiotics or for injection of contrast media for myelography and cisternography.

Contraindications

The primary contraindication to performing LP is elevated intracranial pressure (ICP) from a mass lesion or massive cerebral edema. A continued CSF leak leading to fatal herniation can occur as a result; therefore, in general, patients with papilledema or focal neurological deficit must have a CT scan to rule out a mass lesion. If bacterial meningitis is strongly suspected clinically, appropriate antibiotics should be initiated and LP should not be deferred. Relative contraindications may be considered in patients with coagulopathy where caution should be used. The clinician must weigh the risks of the procedure on an individual basis versus the clinical need to obtain CSF. It is our observation that the risks tend to be overestimated and that close attention to needle size, choice of local anesthetic and the experience of the operator can often allow for safe collection of fluid in less than ideal situations.

Indications for Imaging

Many emergency departments and health care professionals practice routine CT scanning prior to LP as the standard of care due to the concern about cerebral herniation. It seems though that the evidence to support this practice is lacking. Routine CT scanning is not necessary in all patients prior to LP and may not be adequate to exclude elevated intracranial pressure in others. Studies suggest that high-risk patients can be identified, allowing the majority of patients to safely undergo LP without screening CT.

In a prospective study of 301 adult patients with suspected meningitis by Hasbun et al (2001), 235 patients had a CT scan before LP and only 5% (11 patients) had a mass effect. The risk of an abnormal CT scan was related to a history of previous CNS disease, seizure within 1 week and abnormal neurological findings such as decreased level of consciousness and focal motor deficit. This study also showed that patients who underwent CT scan before LP experienced an average of 2 hours delay in diagnosis and 1 hour delay in treatment.

Therefore, CT scan performance before an LP in cases of suspected meningitis is not warranted or recommended unless the patient has grossly altered mental status, active and recent seizures, focal neurologic signs, or papilledema. Patients with these findings or clinical risk factors should have a CT scan to identify mass lesions and other causes of increased ICP.

| Print

Practice Point

Given concerns about cost-effective medical practice and emerging risks or radiation exposure, judicious use of CT scanning is prudent.

Other imaging modalities may be used during the actual performance of the procedure. The use of ultrasound has become particularly valuable as the emergence of portable ultrasound machines allows this technology with high-quality image capability to be taken to the bedside. We now routinely use ultrasound to map the anatomic landmarks prior to virtually all LP attempts.

Fluoroscopy-guided LP may be an invaluable adjunct for the extremely challenging patient in which bedside LP has failed. It shows the bony structures of the lumbar spine and provides real-time information about the position of the needle as it is being inserted. To avoid trauma and RBC in CSF, fluoroscopy-guided LP may be best used in cases of suspected subarachnoid hemorrhage with a negative brain CT if an experienced operator is not available. However, fluoroscopy-guided LP requires the availability of interventional radiologist and mobilizing the patient to the radiology suite when a mobile C-arm unit is not available at the bedside. Use of fluoroscopy requires the patient to lie in the prone position which may be uncomfortable and makes the measurement of opening pressures more challenging. Additionally, it results in greater patient charges: $160 for each LP performed and it does require exposure to radiation.

As an adjunct to risk assessment, measurement of optic nerve sheath diameter (ONSD) using ultrasound has been advocated as an additional parameter for the determination of elevated ICP. Using the high frequency linear ultrasound probe, the operator places the probe gently over the closed upper eyelids of the patient until the optic nerve is seen behind the globe (Figure 114-1A,1B). Optic nerve sheath diameter is determined by drawing a vertical line from the infundibulum of the optic nerve to a distance of 3 mm and measuring the diameter of the optic nerve sheath at that point. A value of 5 mm or less is believed to correlate with a normal ICP. This technique, if validated, could be invaluable as a rapid method to determine at the bedside the need for CT scanning. In the original study by Blaivas et al, ONSD was used as an assessment of intracranial injury due to trauma. Furthermore, it could be used to follow patients with anticipated elevations in ICP that would require a repeat LP for therapeutic purposes as in patients with cryptococcal meningitis. At present, confirmatory data is lacking and the procedure is operator dependent but may prove to be useful in the future.