Abstract
Severe meningitis, especially basilar meningitis, can lead to hydrocephalus requiring external ventricular drain (EVD) placement. There are differences in cerebrospinal fluid (CSF) obtained from an EVD compared to a lumbar puncture (LP). Hence, it becomes difficult to compare LP and EVD samples for diagnosis and monitoring of meningitis. Recognizing these differences is important to properly treat and discontinue antibiotics. We report a case series of 6 patients with meningitis comparing EVD and LP CSF study analysis. In all 6 patients, CSF from LP was obtained before EVD placement by 1.7 days on average. Although corrected white blood cell (WBC) counts were elevated in CSF obtained from LP and EVD, the counts were significantly higher in LP CSF. Protein concentration in LP CSF was also significantly higher than EVD CSF. Glucose and red blood cells varied in both LP and EVD samples. Even though EVD CSF was obtained later in the clinical course than LP, slower circulation of CSF in lumbar space as compared to ventricles is likely the reason for a more sterile appearance of EVD CSF for the diagnosis of meningitis. It is important to recognize these differences as EVD CSF analysis for diagnosis of meningitis may lead to a missed diagnosis and false perception of significant improvement when monitoring response to treatment. One can consider repeating LP prior to discontinuation of antibiotics to properly determine the extent of improvement given EVD CSF sample appears more sterile in comparison. Larger studies are needed to confirm the above findings.
Keywords: cerebrospinal fluid, lumbar puncture, external ventricular drain, Meningitis < Central Nervous System Infections
Introduction
Meningitis is an inflammation of the leptomeninges surrounding the brain and the spinal cord. The classic triad for meningitis is nuchal rigidity, fever, and alerted mental status. 1 Diagnosis of meningitis is made with clinical symptoms and cerebrospinal fluid (CSF) analysis. 1 Lumbar puncture (LP) is the usual site of CSF collection for analysis. In cases of severe meningitis, especially basilar meningitis, hydrocephalus may develop requiring placement of external ventricular drain EVD;2,3 Repeat CSF analysis can be indicated in patients with neurologic deterioration. 4 Sampling of CSF from an in place EVD compared to a repeat LP is easier and more comfortable for the patient for further monitoring of meningitis. However, there does seem to be differences in the cell and protein counts of different CSF sampling sites which can mask the worsening or improvement of meningitis.
The differences in LP and EVD CSF have been reported in multiple case reports in different neurological conditions. There seems to be protein gradient on the neural axis, in CSF without any neurological conditions, with increase in protein concentration from ventricular (mean 26mg/dl), to cortical subarachnoid space (mean 53mg/dl), to the lumbar cistern (mean 42 mg/dl).6,7,8 Similar patterns are observed in cases of neurological conditions. A patient with chronic subarchnoiditis and cysticercosis with secondary hydrocephalus was found to have higher cell and protein counts in LP than ventricular sample which also have been observed with bacterial meningitis.5,9 Similarly, cell counts are seen to be higher in LP compared to EVD CSF. 10 However, there is limited data on the differences in LP and EVD CSF analysis in monitoring and treatment of meningitis. These differences are important to recognize as it can falsely lead one to believe improvement or worsening of meningitis depending on the site of CSF collection. Hence, one can consider repeating LP to properly determine improvement, given EVD CSF appears more sterile, prior to discontinuation of antibiotics for treatment of meningitis. Here we present a case series of 6 patients that demonstrate significant differences between LP and EVD samples in patients with meningitis.
Methods
A case series of 6 patients from a single institution over a year with meningitis requiring EVD placement. Collected data included CSF culture, color, cell counts, glucose, and protein from both LP (tube 4) and EVD samples (tube 1). Only one tube was collected from EVD. The data was then analyzed using median and range. 1 WBC was corrected for every 500 RBC and 1.1 mg/dL protein was corrected for every 1000 RBC in both samples.
Description of Cases
Case 1
A 60 year-old male with a past medical history (PMH) of ruptured cerebral aneurysm, residual right spastic hemiparesis status post baclofen pump placement, and epilepsy presented with status epilepticus and fever. Computed tomography (CT) of the head with contrast did not show any abnormal enhancement. Magnetic resonance imaging (MRI) brain was not done due to the presence of his baclofen pump. CSF analysis was concerning for meningitis (Table 1). He was started on prophylactic antibiotic coverage for meningitis prior to the initial LP.
Table 1.
CSF Analysis of LP and EVD Samples.a
| Case | Days between LP and EVD sample (LP-EVD) | Culture | Color (LP/EVD) | LP | EVD | LP | EVD | LP | EVD | LP | EVD |
|---|---|---|---|---|---|---|---|---|---|---|---|
| WBC (ref 0-5/uL) | Glucose (ref 40-75 mg/dL) | Protein (ref 15-45mg/dL) | RBC (ref <1/uL) | ||||||||
| 1 | 1 | few Kocuria species in EVD sample | Clear/Clear | 1310 (1310) | 89 (89) | 47 | 62 | 133 (133) | 39 (39) | 12 | 22 |
| 2 | 4 | Cx neg in both | Yellow/Clear | 11059 (11057) | 33 (33) | 9 | 97 | 247 (246) | 40 (40) | 1120 | 4 |
| 3 | 2 | Coccidioides species in both | Yellow/Clear | 267 (267) | 60 (60) | 3 | 49 | 1830 (1830) | 46 (46) | 31 | 89 |
| 4 | 1 | Cx neg in both | Red/Yellow | 15109 (14879) | 854 (853) | 49 | 95 | 390 (275) | 46 (46) | 114673 | 506 |
| 5 | 1 | Cx neg in both | Clear/Pink | 292 (292) | 114 (111) | 17 1 | 48 | 96 (96) | 57 (56) | 18 | 1675 |
| 6 | 1 | Cryptococcus neoformans in both | Clear/Pink | 231 (231) | <1 (<1) | 79 | 89 | 76 (76) | 10 (9) | 0 | 1527 |
Abbreviation: Neg negative; cx culture; ref reference; RBC red blood cells.
a Table reports values of WBCs and protein prior to corrections for RBCs and “()” reports corrected values for RBCs. Out of 6 patients, 2 were fungal positive. Overall, there was higher WBC and protein in LP than EVD CSF. RBC and glucose varied in both samples.
Case 2
A 55 year-old male with a PMH of obstructive sleep apnea (OSA), paroxysmal atrial fibrillation, dilated cardiomyopathy with an implantable cardioverter-defibrillator, congenital hydrocephalus with ventriculoperitoneal (VP) shunt placed at the age of 5 and removed 1 month prior to this presentation due to cerebral abscess/meningitis, and epilepsy presented with frontal headache and fever. CT head was concerning for small left tempero-occipital infarct versus cerebritis/encephalitis. MRI brain was not completed due to the presence of her defibrillator. CSF analysis was concerning for bacterial meningitis with negative cultures (Table 1). Patient had completed intravenous antibiotics for 3 weeks just prior to this presentation for meningitis.
Case 3
A 41 year-old male with a PMH of intravenous drug use, coccidiomycosis, internal carotid artery dissection, and VP shunt placed 1 month prior presented with headache, generalized weakness, and fever. Given LP on arrival concerning for meningitis, VP shunt was removed and EVD was placed. MRI brain showed extensive leptomeningeal enhancement of the basal cisterns consistent with meningitis. He was found to have coccidioidomycosis meningitis from his CSF cultures (Table 1).
Case 4
A 48 year-old male with PMH of diabetes mellitus type 2 (DM II), hypertension, OSA, end stage renal disease on hemodialysis, metastatic colon cancer to lung, bone and brain, suboccipital decompressive craniectomy with cerebellar mass resection 2 months prior to this admission presented with encephalopathy and fever. MRI brain showed restricted diffusion in lateral ventricle concerning for infective proteinaceous material in subarachnoid spaces/meningitis. He was found to have CSF concerning for bacterial meningitis with negative cultures (Table 1). He was started on antibiotics after collection initial CSF sample.
Case 5
A 59 year-old male with a PMH of human immunodeficiency virus, DM II, cerebellar hemorrhage and obstructive hydrocephalus status post suboccipital craniectomy with EVD and eventual removal, and aseptic meningitis 3 months prior to this admission presented with encephalopathy and fever. CT head without contrast showed stable moderate hydrocephalus and pseudomeningocele compared to prior CT. He was found to have LP results concerning for bacterial meningitis with negative cultures (Table 1). He was started on antibiotics after initial LP CSF collection.
Case 6
A 77 year-old male with a PMH of chronic lymphocytic leukemia, DM II, sick sinus syndrome status post pacemaker, and hypothyroidism presented with nausea, vomiting, encephalopathy, generalized weakness, and fever. CT head was unrevealing. CSF analysis was concerning for meningitis with cultures positive for Cryptococcus neoformans (Table 1).
Results
We compared the CSF studies of all 6 patients with meningitis who had LP and EVD using median given small sample size. Patients age ranged from 41 to 77 years old with a mean and median age of 57. The median of difference between LP and EVD sample collection day(s) was 1 (range of 3) with mean of 1.7. White blood cells (WBCs; per uL) count in LP CSF were higher compared to EVD samples with median of 801 (range of 14648) in LP and 74.5 (range of 853) in EVD after correction for red blood cells (RBCs). Protein (mg/dL) after correction for RBCs was higher in LP CSF than EVD with median of 190 (range of 1754), and 43 (range of 47), respectively. Glucose and RBCs varied in both LP and EVD samples.
Discussion
CSF samples from LP and EVD have differences in cell and protein counts which is important to recognize for monitoring improvement of meningitis and de-escalating treatment. We analyzed samples from 6 patients with clinical meningitis who had LP and EVD placement. We found that all patients had higher WBC counts and protein in LP compared to EVD CSF. Glucose and red blood cells (RBC) varied in both samples (Table 1).
These findings could be attributed to the previously described craniocaudal protein gradient, differences in protein permeability, and flow rate in different locations. There is a higher concentration of protein in the lumbar cistern when compared to the ventricular system likely due to higher protein permeability. The permeability in the ventricular system is 7.6 μg/min while in the lumbar cistern it is 4.8 μg/min. 6 The protein concentration does not change with decreasing CSF flow rate. 11 This difference in permeability likely accounts for the protein concentration differences noted in our cases.
Craniocaudal gradient has also been seen with leukocytes in CSF with higher concentration in LP than EVD sample. 12 There is about 125 ml of CSF in subarachnoid space covering the brain and spinal cord and 25 ml in ventricular system at any given time. 13 However, it could be inferred that majority of the CSF is in the ventricular system than lumbar cistern at any given time, creating dilution effect in the ventricular system. Thus, we speculate that lower cell counts with EVD could be related to dilution effect and more sterile sample. However, exact mechanism by which this gradient is seen is not yet determined.
As seen in our case series glucose and RBCs in both samples could have varied for many reasons. Glucose differences varied in many cases and has not been found to have any consistent gradient pattern.12,14 Likely its value is related to the serum concentration of glucose at the time of the sample collection. Lastly, RBCs could have varied if a traumatic tap occurred or the presence of intraventricular hemorrhage. Presence of RBCs can also increase WBCs and proteins in the sample. 15 It is important to recognize this because if the values are not properly corrected, it can pose a problem in appropriate diagnosis. It can falsely lead one to diagnose meningitis or subarachnoid hemorrhage if significant number of RBCs are present in the sample. In our samples, however, correction for this did not significantly change the uncorrected value.
There are some confounds in this case series including early administration of antibiotics prior to LP and EVD CSF collection and the number of days between LP and EVD placement. These 2 factors could alter the CSF analysis in a variety of ways potentially changing the results. However, a study done by Rogers et al found that administration of antibiotics before CSF collection in patients with meningitis/ventriculitis did not impact CSF analysis of WBC, RBCs, protein, or glucose. 16 It is unlikely for a couple of doses of antibiotics to significantly make EVD CSF sample appear more sterile than LP CSF especially since some patients had received antibiotics prior to LP collection as well.
In addition, some of our patients had recent neurosurgical procedures such as craniectomy for cerebellar tumor, cerebellar hemorrhage, placement and/or removal of VP shunts. These manipulations and diseases can cause inflammatory state possibly affecting CSF analysis.17,18
Our case series brings forth that it is important to recognize these innate differences in leukocyte counts and protein concentration obtain from LP and EVD CSF. EVD samples could appear more sterile due to craniocaudal gradients of cell and protein as well as a diluting effect in the ventricular system given more fluid is in it compared to around the lumbar meninges at any given time. Hence, one should be cautious to directly compare the results of LP CSF to EVD CSF in monitoring improvement of meningitis and perhaps same site sample comparison should be utilized. We suggest perhaps repeating LP and comparing to initial one prior to discontinuation of antibiotics to monitor improvement of meningitis.
Conclusion
Recognizing differences in LP and EVD CSF in diagnosing of meningitis is important as EVD CSF might appear more sterile. Caution should be used comparing CSF analysis of LP and EVD for monitoring CSF improvement for meningitis and repeating LP should be considered prior to discontinuation of antibiotics.
Footnotes
Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iD
Bhavika Kakadia, DO 
https://orcid.org/0000-0003-1240-5884
References
- 1.Bamberger DM. Diagnosis, initial management, and prevention of meningitis. Am Fam Physician. 2010;82(12):1298–1491. [PubMed] [Google Scholar]
- 2.Mactier H, Galea P, McWilliam R. Acute obstructive hydrocephalus complicating bacterial meningitis in childhood. BMJ. 1998;316(7148):1887–1889. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Mangham D, Gerding DN, Peterson LR, Sarosi GA. Fungal meningitis manifesting as hydrocephalus. Arch Intern Med. 1983;143(4):728–731. [PubMed] [Google Scholar]
- 4.van de Beek D, Cabellos C, Dzupova O, et al. ESCMID guideline: diagnosis and treatment of acute bacterial meningitis. Clin Microbiol Infect. 2016;22(suppl 3):S37–S62. [DOI] [PubMed] [Google Scholar]
- 5.Rubalcava MA, Sotelo J. Differences between ventricular and lumbar cerebrospinal fluid in hydrocephalus secondary to cysticercosis. Neurosurgery. 1995;37(4):668–672. [DOI] [PubMed] [Google Scholar]
- 6.Cutler RW, Murray JE, Cornick LR. Variations in protein permeability in different regions of the cerebrospinal fluid. Exp Neurol. 1970;28(2):257–265. [DOI] [PubMed] [Google Scholar]
- 7.Weisner B, Bernhardt W. Protein fractions of lumbar, cisternal, and ventricular cerebrospinal fluid. separate areas of reference. J Neurol Sci. 1978;37(3):205–214. [DOI] [PubMed] [Google Scholar]
- 8.Fishman RA, Ransohoff J, Osserman EF. Factors influencing the concentration gradient in cerebrospinal fluid. J Clin Invest. 1958;37(10):1419–1424. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Cooke RS, Patterson V. Acute obstructive hydrocephalus complicating bacterial meningitis. hydrocephalus was probably non-obstructive. BMJ. 1999;318(7176):124. [PubMed] [Google Scholar]
- 10.Khan SF, Macauley T, Tong SY, et al. When ventricular cerebrospinal fluid assessment misleads: Basilar meningitis and the importance of lumbar puncture sampling. Open Forum Infect Dis. 2019;66(7):ofz324. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Reiber H. Proteins in cerebrospinal fluid and blood: barriers, CSF flow rate and source-related dynamics. Restor Neurol Neurosci. 2003;21(3-4):79–96. [PubMed] [Google Scholar]
- 12.Naija W, Matéo J, Raskine L, et al. Case report: greater meningeal inflammation in lumbar than in ventricular region in human bacterial meningitis. Crit Care. 2004;8(6):R491–R494. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Telano LN, Baker S. Physiology, cerebral spinal fluid (CSF). [Updated 2020 Apr 22]. In: StatPearls [Internet]. StatPearls. Publishing 2020 Jan. Accessed June 6, 2020. https://www.ncbi.nlm.nih.gov/books/NBK519007/ [PubMed] [Google Scholar]
- 14.Torres-Corzo JG, Tapia-Pérez JH, Sánchez-Aguilar M, Della Vecchia RR, Chalita Williams JC, Cerda-Gutiérrez R. Comparison of cerebrospinal fluid obtained by ventricular endoscopy and by lumbar puncture in patients with hydrocephalus secondary to neurocysticercosis. Surg Neurol. 2009;71(3):376–379. [DOI] [PubMed] [Google Scholar]
- 15.Jurado R, Walker HK. Cerebrospinal Fluid. In: Walker HK, Hall WD, Hurst JW, eds. Clinical Methods: The History, Physical, and Laboratory Examinations. 3rd ed. Butterworths; 1990. Chapter 74. [PubMed] [Google Scholar]
- 16.Rogers T, Sok K, Erickson T, et al. Impact of antibiotic therapy in the microbiological yield of healthcare-associated ventriculitis and meningitis. Open Forum Infect Dis. 2019;6(3):ofz050. doi:10.1093/ofid/ofz050 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Hanak BW, Bonow RH, Harris CA, Browd SR. Cerebrospinal Fluid shunting complications in children. Pediatr Neurosurg. 2017;52(6):381–400. doi:10.1159/000452840 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Tschoe C, Bushnell CD, Duncan PW, Alexander-Miller MA, Wolfe SQ. Neuroinflammation after intracerebral hemorrhage and potential therapeutic targets. J Stroke. 2020;22(1):29–46. doi:10.5853/jos.2019.02236 [DOI] [PMC free article] [PubMed] [Google Scholar]