|Year : 2020 | Volume
| Issue : 2 | Page : 135-138
Feasibility and diagnostic benefit of increased cerebrospinal fluid volume and frequency in the diagnosis of tuberculous meningitis
Sheena Evelyn Ebenezer1, Ramya Iyyadurai1, Joy Sarojini Michael2, OC Abraham1, Sowmya Satyendra Sudha Jasmine1, KP P Abhihash1, Jayaprakash Muliyil3, Thambu David Sudarsanam1
1 Department of Medicine, Christian Medical College, Vellore, Tamil Nadu, India
2 Department of Microbiology, Christian Medical College, Vellore, Tamil Nadu, India
3 Clinical Epidemiology Unit, Christian Medical College, Vellore, Tamil Nadu, India
|Date of Submission||30-Mar-2019|
|Date of Decision||21-May-2019|
|Date of Acceptance||03-Jun-2019|
|Date of Web Publication||8-Oct-2020|
Thambu David Sudarsanam
Department of General Medicine, CMC, Vellore, Tamil Nadu
Source of Support: None, Conflict of Interest: None
Background: Definite diagnosis of tuberculous meningitis (TBM) requires demonstration of TB bacilli in the cerebrospinal fluid (CSF) on smear, culture, or nucleic acid amplification. However, the sensitivity of these tests is low. This study was done to see if smear and culture done on a larger volume and repeated samples of CSF increases the diagnostic yield of these tests. Methods: Adult patients with clinical features of meningitis for >5 days were prospectively and consecutively recruited. At admission, the usual 1 ml of CSF was taken for mycobacterial testing; another 4–8 ml was also taken for the same as the comparison. On the 3rd hospital day, 4–8 ml of CSF was taken for mycobacterial testing. Mycobacterial smear and culture were done by Auramine O stains and on modified Lowenstein–Jensen medium, respectively. The composite reference standard for TBM was considered the gold standard for the diagnosis of TBM. Definite and probable diagnosis was taken as positive, while a possible diagnosis and no TB were considered negative for TBM. In a subset, Xpert MTB/Rif assay was also performed on the CSF samples as per the routine diagnostic protocol. Results: 66/80 (82.5%) had the initial CSF examination in the emergency department, 80 (100%) had large-volume CSF on day 1, while 22/80 (27.5%) consented for the second large-volume CSF examination on day 3. There was a marginal increase in sensitivity from 22% to 26% with increasing CSF volume from 1 ml to 4–8 ml and 36% if another large-volume CSF was collected at 72 h. The specificity was 100% at all times. The negative likelihood ratios were 0.78, 0.74, and 0.64, respectively. The sensitivity of Xpert MTB/RI compared to the composite reference standard of TBM was 30.4%. Conclusions: Increasing the volume and frequency of CSF testing for Mycobacterium tuberculosis, marginally improves the sensitivity and negative likelihood ratios, but may not be adequate to rule out TBM with the certainty required to withhold antitubercular therapy. The feasibility of repeat CSF examinations needs to be considered while making guidelines.
Keywords: Cerebrospinal fluid, diagnosis, tuberculous meningitis
|How to cite this article:|
Ebenezer SE, Iyyadurai R, Michael JS, Abraham O C, Sudha Jasmine SS, P Abhihash K P, Muliyil J, Sudarsanam TD. Feasibility and diagnostic benefit of increased cerebrospinal fluid volume and frequency in the diagnosis of tuberculous meningitis. CHRISMED J Health Res 2020;7:135-8
|How to cite this URL:|
Ebenezer SE, Iyyadurai R, Michael JS, Abraham O C, Sudha Jasmine SS, P Abhihash K P, Muliyil J, Sudarsanam TD. Feasibility and diagnostic benefit of increased cerebrospinal fluid volume and frequency in the diagnosis of tuberculous meningitis. CHRISMED J Health Res [serial online] 2020 [cited 2021 Jan 25];7:135-8. Available from: https://www.cjhr.org/text.asp?2020/7/2/135/297578
| Introduction|| |
Tuberculous meningitis (TBM) the most severe form of extrapulmonary tuberculosis, accounts for around 4.5% of extrapulmonary TB in the United States. The estimated mortality of TBM in India is 1.5/100,000 population. The estimated mortality in a large multicountry study was nearly 17%. Early diagnosis and treatment is essential to decrease morbidity and mortality.
Definitive diagnosis of TBM is based on demonstrating TB bacilli in the cerebrospinal fluid (CSF). The diagnostic accuracy of mycobacterial smears and cultures from the CSF are varied. The Xpert MTB/RIF has been advocated as an initial diagnostic step for rapid diagnosis of TBM.
Fallon and Kennedy, in their study, found the sensitivity of CSF smear microscopy to be 37% and of culture to be 52% from an initial lumbar puncture (LP) sample. When the number of LPs were increased to four, the sensitivity of CSF smear microscopy increased to 87% and that of culture increased to 83%.
In most developing countries, CSF volumes sent for mycobacterial testing are 1–2 ml, done only once initially. In a study done in our center over 5 years, only 11% had a mycobacterial smear or culture positivity. Our study hypothesis was that smear and culture done on a larger volume, repeated samples of CSF would increase the sensitivity and specificity of these tests in the diagnosis of TBM.
| Methods|| |
The study was approved by the Institutional Review Board of the Christian Medical College, Vellore (IRB number-7803). It was conducted in the medical wards from May 2012 to June 2013.
All patients with suspected meningitis with symptoms for 5 days or more were prospectively and consecutively recruited. We included patients more than 15 years of age with fever, headache, neck stiffness, and altered sensorium with or without focal neurological deficits.
We excluded patients with a definite contraindication for LP, those with CSF counts <5 cells/cubic mm, those on anti-tubercular therapy (ATT) for more than 1 week, inability to get at least 4 ml of CSF (at least 3 ml is needed for centrifugation), and those who refused informed consent.
In the emergency department, after obtaining informed consent, CSF was collected under aseptic precautions; 0.5 ml was sent for biochemistry tests, 0.5 ml for clinical pathology tests and 1 ml was sent for routine microbiology tests including, acid-fast bacilli (AFB) smear and culture as the standard of care arm. An extra 4–8 ml of CSF was collected later that day in the ward and sent for AFB smear and culture as the day 1 large-volume CSF. On the 3rd hospital day, a second CSF sample of 4–8 ml was collected and sent for mycobacterial smear and culture. However, if a diagnosis of TBM had been made by the first CSF testing or an alternative diagnosis was already made, the day 3 CSF was not collected.
The large-volume (test arm) CSF specimens were centrifuged and the precipitate was stained with Auramine-O and screened for AFB using fluorescent microscope under 20 XS lens. 60 fields were viewed for 5 min by one technician before declaring a slide negative. The precipitate was also cultured on modified Löwenstein–Jensen (LJ) medium as per standard laboratory protocol; these were incubated for 90 days before declaring negative. During the study period, the Xpert MTB/RIF assay was introduced in our institution for routine diagnostics and was done on 44 consecutive patient samples.
We used the validated composite reference standard for the diagnosis of TBM. For the current study, “definitive” and “probable” diagnosis of TBM was considered as positive, while possible and non-TBM were considered negative.
For this diagnostic study, we calculated a sample size of 120 patients, based on an average sensitivity of CSF smear and culture of 55% and specificity 95%, with precision of 10.
| Results|| |
There were 140 patients [STROBE [Figure 1] admitted to the medical wards from May 2012 to May 2013 with a clinical diagnosis of probable central nervous system infection; 40 (28%) had CSF total count <5 cells/cu mm, 12 had raised intracranial pressure and 8 patients refused consent, because of which these patients were excluded. W, therefore, included 80 patients in the study.
66/80 (82.5%) had the initial CSF examination in the emergency department, 80 (100%) had large-volume CSF on day 1, while 22/80 (27.5%) consented for the second large-volume CSF exam on day 3.
Patients with TBM were younger and had a higher rate of HIV coinfection [Table 1]. Overall, 16/80 (20%) had HIV coinfection. 21 of the 42 who were considered non-TBM had alternate diagnosis. These included Scrub typhus meningitis (n = 6.28%), partially treated pyogenic meningitis 3, cryptococcal meningitis 3 (all HIV coinfected), and 2 normal pressure hydrocephalus. One patient each had primary multifocal leukoencephalopathy, septic encephalopathy, CNS lupus, carcinomatous meningitis, cortical vein thrombosis, hyperglycemic seizures, and hypoglycemic seizures.
Only one CSF smear (1/80) was positive for mycobacteria. Six CSF cultures were positive for mycobacteria (6/80), none of whom were smear positive. Two of the six culture positives had HIV coinfection. CSF X-pert PCR was positive in 7, of whom 3 were HIV coinfected.
The sensitivity of AFB culture with 1 ml of CSF on day 1, was 22% (95% confidence interval [CI] 9.2, 40) and the specificity was 100% (95% CI 89.7, 100) [Table 2]. The positive and negative likelihood ratios were infinity and 0.78, respectively.
|Table 2: Positive Mycobacterium tuberculosis smears and/or culture positivity of the cerebrospinal fluid compared with the composite reference standard for tuberculous meningitis for the study patients|
Click here to view
The day 1 large-volume CSF sensitivity and specificity for AFB culture were 26.3% (95% CI 13.4, 43) and 100% (95% CI 91.6, 100), respectively. The positive likelihood and negative likelihood ratios were infinity and 0.74, respectively.
The day 3 repeat large-volume CSF culture had a sensitivity and specificity of 36.4% (95% CI 11, 69) and 100% (95% CI 71.5, 100), respectively. The positive and negative likelihood ratios were infinity and 0.64, respectively.
The sensitivity of Xpert MTB/RIF assay was 30.4% (95% CI 13.2, 53) and the specificity was 100% (95% CI 84,100) [Table 3]. The positive and negative likelihood ratios were infinity and 0.7, respectively.
|Table 3: Mycobacterium tuberculosis on Xpert MTB/RIF in the cerebrospinal fluid sample compared with the composite reference standard for Tuberculous meningitis for the study patients|
Click here to view
We also compared the Xpert MTB/RIF assay with any culture positivity in the CSF as the gold standard. The sensitivity of Xpert MTB/RIF was 45% and specificity was 93%. Two patients, negative by smear and culture, were Xpert MTB/RIF positive (data not shown).
A few patients complained of headache post-LP which spontaneously resolved in all patients. The cost for AFB smear and culture for 1 ml and large volume is the same in our laboratory (INR 490.0), while a LP needle costs INR 75.0.
| Discussion|| |
The feasibility of getting repeat large-volume CSF studies in our tertiary care center was poor (<30%). This has been alluded to in other studies from developing countries. While guidelines suggest up to 3 or 4 repeat large-volume CSF examinations, in actual practice, we found less than a third consent for this procedure.
The increased volume of CSF tested from 1 ml to 4–8 ml marginally increased the sensitivity of LJ medium culture from 22% to 28%, while the specificity remained 100%; on repeating the LJ medium culture on a large volume (4–8 ml) of CSF collected after 3 days the sensitivity increased to 36%. The high positive and not so low negative likelihood ratios suggest that while positive results were useful in ruling in TBM negative results were not useful in ruling out the disease. The sensitivity of CSF culture increased from 72.2% to 81.8% when an automated culture system was used as compared to the LJ medium.
The sensitivity of CSF Xpert MTB/RIF assay was only 30% as compared to the composite reference standard. A later study from our center found the sensitivity of Xpert MTB/Rif assay on CSF to be 13% as compared to the composite reference standard. A meta-analysis found sensitivity of 62.8%. Increasing CSF volumes from 1 ml (uncentrifuged) to 3 ml (centrifuged) increased sensitivity from 47% to 82%. In the current study, the CSF Xpert MTB/Rif assay was done in uncentrifuged small volume samples of CSF. The advanced Xpert Ultrawith a sensitivity of 95% to 100% HIV-positive patients, with large volume (>6 ml) doing better than small volume <6 ml suggesting that in HIV coinfected, and this could be useful <6 ml suggests that in HIV coinfected, and this could be useful test.
The limitations in this study were the initial 1-ml CSF sample was not sent in 14 patients in the emergency department, often due to pressure to move patients to inpatient wards. Many patients did not consent for the repeat large-volume CSF sample.
Variation in the sensitivity of CSF TB tests needs more study. Perhaps, variations in the prevalence of HIV coinfection could be an explanation. Xpert studies on HIV coinfected seem to show better results. Genetic differences leading to variations in cytokine response and lower mycobacterial levels in the CSF could be another. Correlation between patients' genetic factors, CSF cytokines, and mycobacterial diagnostic yield needs further study.
| Conclusions|| |
We found that with increased CSF frequency and volume, there was only a modest increase in sensitivity and mild improvement in negative likelihood ratios. These values are still not good enough to clinically rule out TBM. The sensitivity of Xpert MTB/RIF assay was poor. The feasibility of repeated large-volume CSF tests due to poor patient acceptability is an issue guideline developers need to consider.
Financial support and sponsorship
The financial support for this study was received from the Hospital Internal Research Funds.
Conflicts of interest
There are no conflicts of interest.
| References|| |
Estimating Mortality from Tuberculosis Meningitis in a Community: Use of Available Epidemiological p. Available from: http://www.popline.org/node/254225
. [Last accessed on 2017 May 16].
Erdem H, Ozturk-Engin D, Tireli H, Kilicoglu G, Defres S, Gulsun S, et al.
Hamsi scoring in the prediction of unfavorable outcomes from tuberculous meningitis: Results of Haydarpasa-II study. J Neurol 2015;262:890-8.
Erdem H, Ozturk-Engin D, Elaldi N, Gulsun S, Sengoz G, Crisan A, et al.
The microbiological diagnosis of tuberculous meningitis: Results of haydarpasa-1 study. Clin Microbiol Infect 2014;20:O600-8.
Denkinger CM, Schumacher SG, Boehme CC, Dendukuri N, Pai M, Steingart KR. Xpert MTB/RIF assay for the diagnosis of extrapulmonary tuberculosis: A systematic review and meta-analysis. Eur Respir J 2014;44:435-46.
Kennedy DH, Fallon RJ. Tuberculous meningitis. JAMA 1979;241:264-8.
Singh S, Sankar MM. Diagnostic algorithm for low-volume CSF samples in tuberculous meningitis. Lancet Infect Dis 2017;17:1236-7.
Kurien R, Sudarsanam TD, Samantha S, Thomas K. Tuberculous meningitis: A comparison of scoring systems for diagnosis. Oman Med J 2013;28:163-6.
Marais S, Thwaites G, Schoeman JF, Török ME, Misra UK, Prasad K, et al.
Tuberculous meningitis: A uniform case definition for use in clinical research. Lancet Infect Dis 2010;10:803-12.
Suzana S, Ninan MM, Gowri M, Venkatesh K, Rupali P, Michael JS. Xpert MTB/Rif for the diagnosis of extrapulmonary tuberculosis – An experience from a tertiary care centre in South India. Trop Med Int Health 2016;21:385-92.
Bahr NC, Nuwagira E, Evans EE, Cresswell FV, Bystrom PV, Byamukama A, et al.
Diagnostic accuracy of xpert MTB/RIF ultra for tuberculous meningitis in HIV-infected adults: A prospective cohort study. Lancet Infect Dis 2018;18:68-75.
Hawn TR, Dunstan SJ, Thwaites GE, Simmons CP, Thuong NT, Lan NT, et al.
Apolymorphism in toll-interleukin 1 receptor domain containing adaptor protein is associated with susceptibility to meningeal tuberculosis. J Infect Dis 2006;194:1127-34.
[Table 1], [Table 2], [Table 3]