You are here

Article Text

Article menu

Download PDFPDF

Neuroinfection
Research paper
Cerebral infarction and tuberculoma in central nervous system tuberculosis: frequency and prognostic implications
  1. Mohammad Wasay1,
  2. Salman Farooq2,
  3. Zubair Ali Khowaja3,
  4. Zeeshan Ali Bawa4,
  5. Shehzad Mansoor Ali4,
  6. Safia Awan1,
  7. M Asim Beg5,
  8. Man Mohan Mehndiratta6
  1. 1Department of Medicine, Aga Khan University, Karachi, Pakistan
  2. 2Ziauddin University Hospital, Karachi, Pakistan
  3. 3Nazeer Hussain Hospital, Karachi, Pakistan
  4. 4Ankleseria Hospital, Karachi, Pakistan
  5. 5Pathology and Microbiology, Aga Khan University, Karachi, Pakistan
  6. 6Department of Neurology, Janakpuri Superspeciality Hospital, Janakpuri New Delhi, New Delhi, India
  1. Correspondence to Dr Mohammad Wasay, Section of Neurology, Department of Medicine, Aga Khan University, Stadium Road, Karachi 74800, Pakistan; mohammad.wasay{at}aku.edu; mohammadwasay{at}hotmail.com

Abstract

Background Tuberculoma and cerebral infarctions are serious complications of central nervous system (CNS) tuberculosis. However, there are no studies comparing prognostic value of tuberculoma and infarcts alone and in patients diagnosed with CNS tuberculosis.

Objective The objective of this study was to identify frequency and prognostic value of tuberculoma and cerebral infarcts in a large sample of CNS tuberculosis patients.

Methods Retrospective chart review of patients diagnosed with CNS tuberculosis in a tertiary care hospital in Pakistan over 10-year period was carried out.

Results There were 404 patients included in this study (mean age of 43 years). There were 209 (52%) men and 195 (48%) women. Tuberculoma were present in 202 subjects (50%) while infarcts were present in 25% patients. 147 (36%) had tuberculous meningitis (TBM) without tuberculoma or infarction on CT or MRI, 158 (39%) had TBM with intracranial tuberculomas, 60 (15%) had TBM with cerebral infarction while 39 (10%) had TBM with both tuberculoma and infarction. At discharge, 249 patients (62%) were either normal (Modified Rankin Score (MRS)=0) or mild to moderately disabled (MRS=1–3) while 82 patients (20%) had severe disability (MRS=4–5). 73 (18%) patients died (MRS=6) during hospitalisation. Using logistic regression analysis, significant predictors of poor outcome included old age, high TBM grading, presence of infarction and presence of hydrocephalus.

Conclusions Tuberculomas were present in 50% of patients, while infarcts were present in 25%. Old age, TBM grading, presence of infarction and hydrocephalus were all predictors of poor outcome.

  • TUBERCULOSIS
  • STROKE
  • MRI

https://doi.org/10.1136/jnnp-2013-307178

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Introduction

    Tuberculosis (TB) ranks as one of the leading causes of mortality and morbidity worldwide. In 2010, an estimated 8.8 million incident cases of TB were recorded globally with approximately 1.45 million deaths.1 Pakistan currently ranks as the fifth largest contributor to the global TB burden with 0.33–0.48 million cases recorded in a population of 187.3 million people.1

    In central nervous system (CNS) TB, tuberculous meningitis (TBM) accounts for 5%–10% of all TB cases and is responsible for more than 40% of the deaths due to TB in developing countries.2 TBM is one of the most serious and critical presentations of TB and if undiagnosed and not treated in time, it can advance into complications like space occupying tuberculomas and cerebral infarction.

    The seeding of the meninges by bacilli from a primary pulmonary complex leads to the formation of rich focus in the brain.3 Upon its rupture into the subarachnoid space, three general processes occur which are responsible for subsequent neurological damage.

    Development of tuberculomas, an obliterative vasculitis which causes infarction and inflammatory adhesive exudates, can result in obstructive hydrocephalus and multiple cranial nerve palsies.3 ,4 Invasion of the subarachnoid space may also result from meningeal involvement or by invasion from tuberculous spondylitis.5

    TBM is characterised by varied clinical manifestation and a myriad of complications. Tuberculoma may be present in 16%–40% of patients with CNS TB.6–9 Tuberculoma accounts for 10%–30% of intracranial masses in TB endemic areas.10 Data regarding prognostic value of tuberculoma have been limited by conflicting evidence. Stroke may be present in 15%–57% patients with CNS TB and has been associated with poor outcome.11–14

    To our knowledge, there have been no studies which compare prognostic value of tuberculoma and infarcts alone and with combination among patients with CNS TB. The objective of this study therefore was to identify frequency and prognostic value of tuberculoma and cerebral infarcts in a large sample of CNS TB patients.

    Method

    This was a retrospective study in which we included patients from the hospital record who were diagnosed as TBM or were treated as TBM and had variable response to treatment during the stay. Charts were identified through ICD-9 coding system (Search terms: 01300–01306 (Tuberculosis Meningitis), 01320–01326 (Tuberculosis Brain), 01330–01336 (Tuberculosis Abscess of Brain), 01380–01386 (Other Specified Tuberculosis of CNS), 01390–01396 (Unspecified Tuberculosis of CNS)) of Medical records department at Aga Khan University, Karachi, Pakistan, during a period of 10 years (2002–2011). A total of 435 charts were reviewed out of which 404 were included in the study. In all, 31 subjects were excluded due to incomplete data availability.

    Each patient's hospital record was thoroughly reviewed by a trained reviewer including patient's condition at presentation, clinical features, comorbidities, results of investigations, response to treatment and outcome at discharge. Data were recorded on a prespecified questionnaire.

    Severity of disease at presentation was graded according to the Modified British Research Council clinical criteria for TBM severity:15 Grade-1 (Glasgow Coma scale-15 (GCS 15)) alert and oriented without focal neurological deficit; Grade-2 (GCS 10–14) with or without focal neurological deficit or GCS 15 with focal neurological deficit; Grade-3 GCS less than 10 with or without focal neurological deficit.

    Patients were evaluated for any disability at the time of presentation and at the time of discharge. Severity was graded according to the Modified Rankin Score (MRS).16 ,17

    Patients were categorised according to the TBM diagnostic criteria of CNS TB according to their diagnostic findings (box 1).18 CT/MRI films of the patients were reviewed by an expert radiologist specialised in neuroimaging.

    Box 1

    Diagnostic criteria of CNS tuberculosis

    Confirmed:

    • Culture positive CSF, stain positive CSF

    • Biopsy of intracranial lesion showing organisms or caseous necrosis and granuloma

    Probable:

    • Biopsy or culture proven pulmonary or nodal TB or military TB (on chest X-ray) with abnormal CSF and or enhancing lesions on brain CT/MRI

    • Intracranial enhancing lesions responding to ATT

    Possible:

    • Abnormal CSF and/or abnormal MRI and response to treatment with ATT

    ATT, antituberculous therapy; CNS, central nervous system

    Data were entered and analysed using the statistical package for social sciences 17.0 (SPSS, Inc., Chicago, Illinois, USA). Descriptive statistics were performed. Results were recorded as frequencies, means±SD and p values. A p value of <0.05 was taken as the criteria of significance for all purposes. Continuous variables are presented as mean±SD and categorical variables are presented as a percentage. Continuous variables between groups were compared by using unpaired t test, and categorical variables were compared by using the χ² test. Comparisons of continuous variables among three groups were made by analysis of variance analysis as appropriate.

    Univariable analyses were performed to examine the effect of each variable on the poor outcome. Multivariable models were then constructed, including variables that showed an effect in the prediction of poor outcome in the univariable analyses. All p values were based on two-sided tests and significance was set at a p value less than 0.05. The analyses were performed using SPSS (Statistical Package of Social Sciences) V.19.

    Study protocol was approved by Ethics review committee of Aga Khan Hospital, Karachi (1311-Med-ERC-09).

    Results

    The study included 404 patients with a mean age of 43 years (range 18–84). There were 209 (52%) men. In all, 35 (8.7%) patients were confirmed cases, 231 (58.9%) patients were classified as probable cases and 131 (32.4%) patients were in the possible category.

    History of BCG vaccination was present in 189 subjects (47%). History of HIV/AIDS was present in only one patient. History of recent steroids or immune suppressant therapy was present in 22 subjects (5%). Vascular risk factors included hypertension (59; 15%) patients, diabetes (32; 8%), ischaemic heart disease (25; 6%), dyslipidaemia (11; 3%) and tobacco use (smoking and chewing tobacco) in 44 (11%) patients. Two patients had history of previous stroke and one patient had atrial fibrillation. Tests for HIV were done in 36 patients (9%) out of which only one patient was positive. Previous chest X-rays and X-rays on admission were reviewed and records of 279 patients (69%) were available out of which 128 patients (31.7%) had findings consistent with chest TB and 151 patients had normal X-rays. In all, 17 patients (4%) had lymph node TB and 38 patients (9%) had milliary TB of the chest. Anaemia was present in 72 (18%) patients. Spine X-rays were done in 19 subjects showing Pott's disease (one patient) and collapsed vertebra (one patient).

    Fever was a common presenting symptom found in 317 patients (79%) followed by decreased consciousness in 241 patients (60%), headache in 232 patients (57%) and nausea and vomiting in 212 patients (53%). Neurological symptoms included motor deficit in 116 patients (28%), seizures in 87 patients (22%) and neck stiffness in 140 patients (35%).

    A total of 306 patients had a lumbar puncture (LP) for CSF studies whereas a mean delay in LP and admission was approximately 1–2 days. All patients had elevated white blood cell count (WBCC) (15–2500) out of which 279 patients (91%) had more than 80% lymphocytes while in 27 patients (9%) CSF leucocytes were predominantly polymorphs. The CSF protein concentration was high (>60 mg/dL) in 269 (66.6%) patients and CSF glucose was low (<60 mg/dL) in 223 (52.2%) patients. CSF smear was done in 288 patients (97.2%) with eight patients (2.7%) being positive for tubercular bacilli. CSF culture was performed in 283 patients (70%) and only 28 patients (10%) tested positive. CSF PCR was done in 50 patients and 5 (10%) patients had a positive result. Multi-drug resistant TB was reported in two patients only.

    All patients underwent imaging of brain (CT scan alone=91, MRI alone=251, both CT scan and MRI=62). CT scan was done in 153 patients (38%) and brain MRI was done in 313 patients (78%). Spine MRI was done in 24 patients. Findings are shown in table 1.

    View this table:
    Table 1

    CT scan and MRI findings

    Overall, 147 (36%) had TBM without tuberculoma or infarction on CT or MRI, 158 (39%) had TBM with intracranial tuberculomas, 60 (15%) had TBM with cerebral infarction while 39 (10%) had TBM with both tuberculoma and infarction. Intracranial tuberculoma were present on initial MRI in 197 (49%) patients. 43 (22%) patients had only one tuberculoma, 45 (23%) had two tuberculomas, 24 (12%) had three tuberculomas and 85 (43%) had four or more tuberculomas on MRI. Unilateral (one cerebral hemisphere) tuberculoma was present in 77 (39%), bilateral (both cerebral hemispheres) tuberculoma were present in 106 (54%) while 14 (7%) had brain stem tuberculoma. Overall, 163 (83%) had tuberculomas located only supra tentorially, 14 (7%) had only infratentorial tuberculoma while 20 (10%) had both supra and infratentorial tuberculomas. Follow-up MRI was done in 33 (9%) patients. New tuberculomas appeared in two patients with paradoxical enlargement seen in four patients and 15 patients showed a reduction in size and number of tuberculoma, while in 12 patients MRI remained unchanged. Cerebral infarction was present in 99 (25%) patients on initial MRI. Single infarct was present in 42 (42%) patients while 57 (58%) had multiple infarcts. Infarcts were acute or subacute in 67 (68%) and chronic in 32 (32%). Out of 33 patients who had follow-up MRI scans, 5 (15%) had developed new infarcts.

    Patients were graded for severity of the disease at the time of presentation according to the TBM grading criteria based on the clinical notes found in the patient's record. The Modified British Research Council clinical criteria for TBM severity was grade I in 134 patients (33.2%), grade II in 160 patients (39.6%) and grade III in 110 patients (27.2%). All patients were treated with four-drug antituberculous therapy (ATT) (isoniazid, rifampicin, ethambutol and pyrizinamide), steroids and pyridoxine. In all, 34 patients also received streptomycin. Hydrocephalus was present in 116 patients (28%) out of whom 55 (48%) underwent ventricular drainage or ventriculo peritoneal shunt placement.

    Average length of hospital stay was 12 days (range 2–63) days. At discharge, 249 patients (62%) were either normal (MRS=0) or mild to moderately disabled (MRS=1–3) while 82 patients (20%) had severe disability (MRS=4–5). A total of 73 (18%) patients died (MRS=6) during hospitalisation.

    Patients were categorised into three groups at the time of admission based on their GCS. The group with GCS <9 or TBM grade III had higher mortality (52%) compared with the group with grade I and II combined (p<0.001).

    Patients were divided into four groups based on presence of tuberculoma and infarction on CT/MRI (patients with tuberculoma, infarction, both and the last group with no tuberculoma or infarct). A comparison of baseline features and outcome of these groups is provided in table 2. Infarction alone group was found to have poor outcome (MRS 4–5) and highest mortality (35%) (p<0.001) (table 2). Frequency of cardiovascular risk factors among various groups is shown in table 3. Boxplots of MRS in various groups are provided as figure 1 and boxplots of MRS in various age groups are shown as figure 2. Patients who developed hydrocephalus (25.6%) as a complication of TBM showed poor outcome with a mortality of 42.5% (p 0.004). Using logistic regression analysis, significant factors which predicted poor outcome were age (OR 1.02, 95% CI), TBM grading (OR 3.53, 95% CI), presence of infarction (OR 0.90, 95% CI) and hydrocephalus (OR 1.95, 95% CI).

    View this table:
    Table 2

    Baseline characteristics and outcome of patients groups with TBM

    View this table:
    Table 3

    Cardiovascular risk factor among patients in various groups

    Figure 1
    Figure 1

    Modified Rankin Score (MRS) discharge score in subjects without infarction and tuberculoma, infarction alone, tuberculoma alone and both tuberculoma and infarction. Data are presented as boxplots, Boxes represent the 25th, 50th (median) and 75th centiles.

    Figure 2
    Figure 2

    Comparison of Modified Rankin Score (MRS) discharge score with different age groups. Data are presented as boxplots, Boxes represent the 25th, 50th (median) and 75th centiles.

    Discussion

    Nearly 50% of patients with CNS TB had intracranial mass lesions (tuberculoma) and 25% had cerebral infarcts in this study. Prognosis of tuberculoma patients was good while cerebral infarcts led to poor prognosis. Previous studies have shown presence of tuberculoma in 16%–40% patients with TB meningitis.6–9 Tuberculoma most likely represents an immune response to tuberculous brain parenchymal infections. A previous study of paediatric CNS TB patients had shown that tuberculoma formation was twice as common in non-BCG vaccinated subjects.19 Very high frequency of tuberculoma in our patients could be related to BCG vaccination or previous TB infection. Follow-up MRI data are limited but clearly show dynamic nature of cerebral disease activity. Tuberculoma were increased in six, decreased in 15 and remained unchanged in 12 out of 33 patients with follow-up MRI. Some of these tuberculomas remained unchanged after 2 years of ATT use; however, frequency of paradoxical response is probably higher and may be asymptomatic in a number of patients. Paradoxical response may be a form of immune reconstitution inflammatory syndrome and is more common (up to 30%) in HIV patients receiving highly active antiretroviral therapy.20 ,21 New infarcts were developed in 15% patients. Treatment protocol was same in all these patients. We think diversity of this neuroradiological response is probably related to immune response of host. Future studies looking at immune and inflammatory markers in cerebrospinal fluid (CSF) may provide clues for this variable response.

    Yield of CSF culture and PCR for acid fast bacalli (AFB) was low among these patients. Out of those tested for AFB smear, AFB culture or CSF-PCR, only 3% were AFB smear positive while 10% had positive PCR or AFB culture. Possible contributing factors for this low yield may include overall low threshold for suspicion for CNS TB, improper sampling technique and altered immune response system related to TB in endemic region.

    Only one patient out of 404 was HIV positive on admission and one patient tested positive for HIV out of 36 tested. This frequency is very low but not surprising. Prevalence of HIV is extremely low among TB patients in Pakistan. A large population based study showed only 42 (0.34%) were HIV positive out of more than 12 000 confirmed TB patients.22 This may be a possible reason why HIV status was not evaluated in a majority of TB meningitis patients.

    Outcome of patients with infarction alone or with tuberculoma and infarcts both was worse than patients with no infracts and tuberculoma or with tuberculoma alone. This could be due to older age and higher frequency of cardiovascular risk factors among patients with infracts (tables 2 and 3). Outcome of patients with tuberculoma was not much different from without tuberculoma. These findings are confirmatory to previous reports that tuberculoma is a benign condition with good prognosis.23 ,24 This is in contrast to our previous findings that intracranial tuberculomas are associated with poor prognosis as compared with TBM patients without tuberculoma.25 The data are so far conflicting but there is a growing evidence that prognosis of patients with tuberculoma is good with aggressive and prompt anti-TB therapy. Patients with tuberculoma were more likely to have seizures as compared with other groups in this study.

    Cerebral infarcts were present in 25% of patients, out of which 68% were acute or subacute and 58% were multiple. These infarcts most likely represent vasculitis involving small and medium vessel disease; the infants and children are most severely affected.12 ,13 ,26 An association between poor outcome and cerebral infarcts in TBM patients is well documented.27–29 Findings of this study confirm previous reports. Overall outcome was worse in patients with infarcts as compared with patients with tuberculoma or in patients without tuberculoma or infarcts. Death and disability related to CNS TB could be substantially reduced by preventing cerebral infarctions. The role of aspirin and steroids has been debated in recent literature.27 A study from India showed beneficial effects of aspirin in patients with TBM while another study of paediatric patients with TBM failed to show any advantage of low dose or high dose aspirin.30 ,31 This finding has to be confirmed in large, multi-centre, controlled trials and needs further investigations.

    Hydrocephalus was present in 25% cases out of which 55% underwent shunt placement. Hydrocephalus is known to be an indicator of poor prognosis including this study.32 Mortality was high among this group despite surgical intervention. Spinal involvement was low among these patients with TB meningitis. A previous study showed that a third of patients with spinal cord TB had no TB meningitis.33 Only 24 patients underwent spinal MRI in our study, out of whom 10 had abnormalities involving spinal cord or nerve roots. It is possible that spinal cord involvement in some of these bedridden, comatose patients may not have been diagnosed.

    This study comprises data from a large group of patients with CNS TB. However, these findings have limitations due to the retrospective design of the study, large number of patients with no confirmed diagnosis on histopathology or culture and a lack of long term follow-up. Despite these limitations, this study provides a useful insight into prognostic factors in patients with CNS TB. Management of CNS TB has not changed due to lack of intervention studies and randomised controlled trials. Duration of therapy, high dose steroids and antiplatelet therapies could be important approaches for future studies.

    Acknowledgments

    The authors acknowledge Ms Javeriah Khan (HIMS) at AKU for her support in data collection.

    References

    1. ↵
      World Health Organisation. Global Tuberculosis Control. 2011. http://www.who.int/tb/publications/global_report/en/
    2. ↵
      1. Scheld WM,
      2. Whitley RJ,
      3. Marra CM
      . Infections of the central nervous system. Lippincott Williams & Wilkins, 2004.
    3. ↵
      1. Rich AR,
      2. McCordock HA
      . The pathogenesis of tuberculous meningitis. Bull John Hopkins Hosp 1933;52:5–37.
      OpenUrlWeb of Science
    4. ↵
      1. Dastur DK,
      2. Manghani DK,
      3. Udani PM
      . Pathology and pathogenetic mechanisms in neurotuberculosis. Radiol Clin North Am 1995;33:733–52.
      OpenUrlPubMedWeb of Science
    5. ↵
      1. Jain SK,
      2. Paul-Satyaseela M,
      3. Lamichhane G,
      4. et al
      . Mycobacterium tuberculosis invasion and traversal across an in vitro human blood-brain barrier as a pathogenic mechanism for central nervous system tuberculosis. J Infect Dis. 2006;193:1287–95.
      OpenUrlAbstract/FREE Full Text
    6. ↵
      1. Gupta RK,
      2. Jena A,
      3. Sharma A,
      4. et al
      . MR imaging of intracranial tuberculomas. J Comput Assist Tomogr 1988;12:280–5.
      OpenUrlCrossRefPubMedWeb of Science
    7. ↵
      1. Ravenscroft A,
      2. Schoeman JF,
      3. Donald PR
      . Tuberculous granulomas in childhood tuberculous meningitis: radiological features and course. J Trop Pediatr 2001;47:5–12.
      OpenUrlAbstract/FREE Full Text
    8. ↵
      1. Shah GV
      . Central Nervous system tuberculosis: imaging manifestations. Neuroimaging Clin. B Am 2000;355–74.
    9. ↵
      1. Wasay M,
      2. Kheleani AB,
      3. Moolani M,
      4. et al
      . Brain CT and MRI findings in 100 consecutive patients with intracranial tuberculoma. Journal of Neuroimaging 2003;13:240–7.
      OpenUrlCrossRefPubMed
    10. ↵
      1. Zuger A,
      2. Lowy D
      . Tuberculosis. In: Scheld WM, Whitley RJ. eds. Infections of the central nervous system. 2nd edn. Philadelphia: Lippincott Raven, 1977:417–43.
    11. ↵
      1. Schoeman JF,
      2. Van Zyl LE,
      3. Laubscher JA,
      4. et al
      . Serial CT scanning in childhood tuberculous meningitis: prognostic features in 198 cases. J Child Neurol 1995;10:320–9.
      OpenUrlAbstract/FREE Full Text
    12. ↵
      1. Andronikou S,
      2. Wilmshurst J,
      3. Hatherill M,
      4. et al
      . Distribution of brain infarction in children with tuberculous meningitis and correlation with outcome score at 6 months. Pediatr Radiol 2006;36:1289–94.
      OpenUrlCrossRefPubMed
    13. ↵
      1. Springer P,
      2. Swanevelder S,
      3. van Toorn R,
      4. et al
      . Cerebral infarction and neurodevelopmental outcome in childhood tuberculous meningitis. Eur J Paediatr Neurol 2009;13:343–9.
      OpenUrlCrossRefPubMedWeb of Science
    14. ↵
      1. Taqui AM,
      2. Wasay M
      . Cerebral infarction in tuberculous meningitis. Pakistan Journal of Neurological sciences 2011;6:25–9.
      OpenUrl
    15. ↵
      British Medical and research council. Streptomycin treatment of tuberculous meningitis. BMJ 1948;1:582–97.
      OpenUrlFREE Full Text
    16. ↵
      1. Rankin J
      . Cerebrovascular accidents in patients over the age of 60. Scott Med J 1957;2:200–15.
      OpenUrlFREE Full Text
    17. ↵
      1. Bonita R,
      2. Beaglehole R
      . Modification of Rankin scale: Recovery of motor function after stroke. Stroke 1988;19:1497–500.
      OpenUrlAbstract/FREE Full Text
    18. ↵
      1. Seth R,
      2. Sharma U
      . Diagnostic criteria for tuberculous meningitis. Indian J Pediatr 2002;69:299–303.
      OpenUrlCrossRefPubMed
    19. ↵
      1. Wasay M,
      2. Ajmal S,
      3. Taqui AM,
      4. et al
      . Impact of Bacille Calmette-Guerin vaccination on neuroradiological manifestations of pediatric tuberculous meningitis. J Child Neurol 2010;25:581–6.
    20. ↵
      1. Wasay M
      . Central nervous system tuberculosis and paradoxical response (Editorial). Southern Medical Journal 2006;99:331–2.
      OpenUrlCrossRefPubMed
    21. ↵
      1. Kumar R,
      2. Prakash M,
      3. Jha S
      . Paradoxical response to chemotherapy in neurotuberculosis. Pediatr Neurosurg 2006;42:214–22.
      OpenUrlCrossRefPubMedWeb of Science
    22. ↵
      1. Hasnain J,
      2. Memon GN,
      3. Memon A,
      4. et al
      . Screening for HIV among tuberculosis patients: a cross-sectional study in Sindh, Pakistan. BMJ Open 2012;2:e001677.
      OpenUrlAbstract/FREE Full Text
    23. ↵
      1. Li H,
      2. Liu W,
      3. You C
      . Central nervous system tuberculoma. J Clin Neurosci 2012;19:691–5.
      OpenUrlCrossRefPubMed
    24. ↵
      1. Anuradha HK,
      2. Garg RK,
      3. Sinha MK,
      4. et al
      . Intracranial tuberculomas in patients with tuberculous meningitis: predictors and prognostic significance. Int J Tuberc Lung Dis 2011;15:234–9.
      OpenUrlPubMed
    25. ↵
      1. Wasay M,
      2. Moolani MK,
      3. Zaheer J,
      4. et al
      . Prognostic indicators in patients with intracranial tuberculoma: a review of 102 cases. J Pak Med Assoc 2004;54:83–7.
      OpenUrlPubMed
    26. ↵
      1. Lammie GA,
      2. Hewlett RH,
      3. Schoeman JF,
      4. et al
      . Tuberculous cerebrovascular disease: a review. J Infect 2009;59:156–66.
      OpenUrlCrossRefPubMedWeb of Science
    27. ↵
      1. Thwaites GE,
      2. van Toorn R,
      3. Schoeman J
      . Tuberculous meningitis: more questions, still too few answers. Lancet Neurol. 2013;12:999–1010.
      OpenUrlCrossRefPubMedWeb of Science
    28. ↵
      1. Sheu JJ,
      2. Hsu CY,
      3. Yuan RY,
      4. et al
      . Clinical characteristics and treatment delay of cerebral infarction in tuberculous meningitis. Intern Med J 2012;42:294–300.
      OpenUrlCrossRefPubMed
    29. ↵
      1. Thwaites GE,
      2. Macmullen-Price J,
      3. Tran TH,
      4. et al
      . Serial MRI to determine the effect of dexamethasone on the cerebral pathology of tuberculous meningitis: an observational study. Lancet Neurol 2007;6:230–6.
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Misra UK,
      2. Kalita J,
      3. Nair PP
      . Role of aspirin in tuberculous meningitis: a randomized open label placebo controlled trial. J Neurol Sci 2010;293:12–17.
      OpenUrlCrossRefPubMed
    31. ↵
      1. Schoeman JF,
      2. Janse van Rensburg A,
      3. Laubscher JA,
      4. et al
      . The role of aspirin in childhood tuberculous meningitis. J Child Neurol 2011;26:956–62.
      OpenUrlAbstract/FREE Full Text
    32. ↵
      1. Raut T,
      2. Garg RK,
      3. Jain A,
      4. et al
      . Hydrocephalus in tuberculous meningitis: Incidence, its predictive factors and impact on the prognosis. J Infect 2013;66:330–7.
      OpenUrlCrossRefPubMedWeb of Science
    33. ↵
      1. Wasay M,
      2. Arif H,
      3. Kheleani B,
      4. et al
      . Neuroimaging of Tuberculous myelitis. J Neuroimaging 2006;16:197–205.
      OpenUrlCrossRefPubMed

    Footnotes

    • Contributors MW: study design, proposal writing, data collection, manuscript writing, manuscript review, manuscript submission. SF, ZAK, ZAB and SMA: data collection, manuscript writing, manuscript review. SA, MAB and MMM: data analysis, manuscript writing, manuscript review.

    • Competing interests None.

    • Ethics approval Ethics review committee, Aga Khan University.

    • Provenance and peer review Not commissioned; externally peer reviewed.