Journal of Infection (2013) 66, 330e337

www.elsevierhealth.com/journals/jinf

Hydrocephalus in tuberculous meningitis:Incidence, its predictive factors and impact on theprognosis

Tushar Raut a, Ravindra Kumar Garg a,*, Amita Jain b, Rajesh Verma a,Maneesh Kumar Singh a, Hardeep Singh Malhotra a, Neera Kohli c,Anit Parihar c

aDepartment of Neurology, King George Medical University, Uttar Pradesh, Lucknow 226003, IndiabDepartment of Microbiology, King George Medical University, Uttar Pradesh, Lucknow 226003, IndiacDepartment of Radiodiagnosis, King George Medical University, Uttar Pradesh, Lucknow 226003, India

Accepted 26 December 2012Available online 2 January 2013

KEYWORDSTuberculosis;Meningeal tuberculosis;Tuberculoma;Cerebrospinal fluid;Stroke;Ventriculo-peritonealshunt

* Corresponding author. Tel.: þ91 9E-mail address: garg50@yahoo.com

0163-4453/$36 ª 2012 The British Infhttp://dx.doi.org/10.1016/j.jinf.2012

Summary Background: Hydrocephalus is one of the most common complications of tubercu-lous meningitis. The present study evaluated the incidence, predictive factors and impact ofhydrocephalus on overall prognosis of tuberculous meningitis.Material and methods: In a prospective cohort study, all patients fulfilling the inclusion crite-ria of tuberculous meningitis underwent clinical and cerebrospinal fluid evaluation, togetherwith magnetic resonance imaging of the brain. Patients were treated with antituberculosisdrugs and dexamethasone. Follow up neuroimaging was done after 6 months. Hydrocephaluswas assessed using Evan’s index.Results: Of 80 patients with tuberculous meningitis, 52(65%) had hydrocephalus at presenta-tion. During follow up, 8 new patients developed hydrocephalus. Factors associated with hy-drocephalus included advanced stage of disease, severe disability, duration of illness > 2months, diplopia, seizures, visual impairment, papilledema, cranial nerve palsy, hemiparesis,CSF total cell count > 100/cu.mm, CSF protein > 2.5 g/l. Neuroimaging factors that were sig-nificantly associated with hydrocephalus included basal exudates, tuberculoma and infarcts.Multivariate analysis revealed visual impairment, cranial nerve palsy and the presence of basalexudates as significant predictors of hydrocephalus. In 13 patients, with early tuberculousmeningitis, there was complete resolution of hydrocephalus. Hydrocephalus was significantlyassociated with mortality and poor outcome.

335901790; fax: þ91 522 2258852.(R.K. Garg).

ection Association. Published by Elsevier Ltd. All rights reserved..12.009

Hydrocephalus in tuberculous meningitis 331

Conclusion: Hydrocephalus occurs in approximately two-third of patients with tuberculousmeningitis and has an unfavorable impact on the prognosis. Hydrocephalus in early stages oftuberculous meningitis may resolve completely.ª 2012 The British Infection Association. Published by Elsevier Ltd. All rights reserved.

Introduction

Tuberculous meningitis is a frequent form of central nervoussystem tuberculosis that carries a high mortality and disabil-ity. Tuberculousmeningitis is pathologically characterized bypresence of thick basilar exudates. In the basal region of thebrain, exudates are predominantly present around the circleofWillis, often extending to the ambient, Sylvian and pontinecisterns, and around the optic chiasma.

Hydrocephalus is one of the most common complicationsof tuberculous meningitis and it has been shown toadversely affect the outcome.1e3 Two types (communicat-ing and obstructive) of hydrocephalus are seen in patientswith tuberculous meningitis. The obstructive type of hydro-cephalus develops when the fourth ventricular outlets areblocked by the basal exudates and leptomeningeal inflam-mation, or when there is an obstruction in the aqueduct.4

Communicating hydrocephalus develops when either thereis overproduction of cerebrospinal fluid (CSF) or there is de-fective absorption of CSF in the subarachnoid space. Com-municating hydrocephalus is much more frequent. Mild tomoderate hydrocephalus responds well to medical treat-ment but surgery is required when there are manifestationsof raised intracranial pressure. Ventriculo-peritoneal shuntis preferred surgical treatment. Other surgical options forthe management of hydrocephalus include endoscopic thirdventriculostomy, external ventricular drainage and Om-maya reservoir implantation.5e7

Computed tomography (CT) and magnetic resonance(MR) imaging are commonly used neuroimaging methodsto diagnose hydrocephalus in tuberculous meningitis. MRI ismore sensitive in detecting early changes of hydrocephalus,infarcts, granuloma and exudates. In hydrocephalus, MRI ismore valuable as it demonstrates the ventricular volumeand flow of CSF through the aqueduct of Sylvius andforamen of Monroe.1e3,8,9

In this prospective cohort study, we assessed the in-cidence of hydrocephalus at presentation and during thecourse of the illness. We evaluated the predictive factors ofhydrocephalus in patients with tuberculous meningitis. Wealso evaluated the impact of hydrocephalus on the prog-nosis of the disease.

Material and methods

This study was a prospective cohort study, conducted fromOctober 2010 to august 2012, in the Department ofNeurology, King George’s Medical University, Lucknow,Uttar Pradesh, India. Our university is a large tertiarycare medical Center which caters to more than 100 millionpeople. Our university is situated in a region which is a highendemic zone for tuberculosis. Prior ethical approval wasobtained from the Institutional Ethics Committee. Before

enrollment, a written informed consent was taken, eitherfrom the patients or their legal guardians.

Clinical entry criteria

Patients with a history suggestive of subacute or chronicmeningitis were clinically assessed. Symptoms and signs,those were suggestive of meningitis, included one or moreof the following: headache, vomiting, fever, neck stiffness,convulsions, focal neurological deficits or altered senso-rium. All patients were categorized into definite, probable,or possible tuberculous meningitis groups. The definitecases of tuberculous meningitis either had presence ofacid-fast bacilli in CSF, mycobacteria cultured from CSF orCSF was positive for mycobacterial nucleic acid amplifica-tion test. The probable cases fulfilled the clinical entrycriteria plus a total diagnostic score of 10 or more points(when cerebral imaging is not available) or 12 or morepoints (when cerebral imaging was available) plus exclusionof alternative diagnoses. The possible cases includedclinical entry criteria plus a total diagnostic score of 6e9points (when cerebral imaging is not available) or 6e11points (when cerebral imaging is available) plus exclusion ofalternative diagnoses.10

Disease severity

Patients were classified according to the British MedicalResearch Council (BMRC) staging system. Patients withstage I disease had a Glasgow Coma Scale score of 15with no focal neurologic signs; patients with stage II hadsigns of meningeal irritation with slight or no alteration ofsensorium and minor neurological deficit (like cranial nervepalsies) or no deficit (Glasgow coma scale score 11e14);and patients with stage III had severe alteration ofsensorium, convulsions, focal neurological deficit and in-voluntary movements (Glasgow coma scale score <10).11

Evaluation

All enrolled patients were subjected to a detailed clinicalevaluation. The work up included complete blood count,peripheral blood smear examination, erythrocyte sedimen-tation rate (ESR), blood sugar, blood urea nitrogen andserum creatinine, liver function tests, serum electrolytes,chest X-ray and enzyme-linked immunosorbent assay forhuman immunodeficiency virus (HIV). CSF biochemical andmicroscopic examination, including India-ink preparation,were performed. CSF Sediments were stained and cultured(LowensteineJensen media) by standard methods. CSFspecimens were also tested for mycobacterial polymerasechain reaction (PCR). All patients were also subjected to MRimaging of the brain using Signa Excite 1.5 Tesla instrument(General Electric Medical Systems, Milwaukee, WI, USA).

332 T. Raut et al.

The scans were reviewed by an experienced neuroradiolo-gist, who was unaware of treatment and patient outcome.

Definition of hydrocephalus

Hydrocephalus was defined as ventriculomegaly with Evan’sratio (maximal width of frontal horns/maximal width ofinner skull) more than >0.30; and/or size of one or bothtemporal horns greater than 2 mm. Other imaging changes,like presence of sulcal effacement, periventricular ooze(transependymal flux of CSF into the periventricular paren-chyma, particularly at the tips of the frontal, occipital andtemporal horns), type of hydrocephalus (communicative orobstructive), meningeal enhancement, exudates, tuber-culomas and infarcts, were also recorded. Communicatinghydrocephalus was defined as hydrocephalus with a dilatedfourth ventricle. There was no evidence of an obstructionof the intraventricular CSF pathways, including the fourthventricular outflow and the cerebral aqueduct. Size offourth ventricle remained unchanged in obstructive hydro-cephalus.4,12 (Fig. 1) Mild, moderate, and severe hydro-cephalus were categorized if Evan’s ratios were <0.34,0.35e0.40, and >0.40 respectively.13

Definition of other neuroimaging findings

On gadolinium-enhanced MRI, meningeal inflammation wasdefinedas the enhancement of pia-arachnoid,which extendsinto the subarachnoid spaces of the sulci, basal cisterns andenhancement along the inner table of the skull and in thedural folds of the falx and tentorium. The exudates weredefinedas a thick areaof enhancement in the regions of basalcisterns and Sylvian fissure. Tuberculoma was defined asdiscrete or coalescing lesions which were showing homoge-neous nodular contrast enhancement or showing a ringenhancement. Infarct was defined as an area of abnormalsignal intensity in a vascular distribution without anyevidence of mass effect. Infarcts were hyperintense to

Figure 1 (A) Axial T2 FLAIR sequence shows calculation of Evanhyperintensity indicates periventricular CSF ooze suggestive of asequence shows measurement of temporal horn width.

gray matter on both spin density and T2-weighted images.These lesions were hypointense on T1-weighted images.

Treatment

All included patients were treated with an antituberculosisregimen as recommended by World Health Organization forthe treatment of central nervous system tuberculosis.Patients received 2 months of daily oral isoniazid (5 mg/kg of body weight; maximum, 300 mg), rifampicin (10 mg/kg; maximum, 600 mg), pyrazinamide (25 mg/kg; maxi-mum, 2 g/day) and intramuscular streptomycin (20 mg/kg;maximum 1 gm/day), followed by 7 months administrationof isoniazid and rifampicin.14 Patients also received dexa-methasone for eight weeks. Corticosteroid regimen con-sisted of intravenous dexamethasone for 4 weeks(0.4 mg/kg body weight per day and then tapered off de-creasing 0.1 mg/kg every week) and then oral treatmentfor 4 weeks, starting at a total of 4 mg per day and decreas-ing by 1 mg each week. Antiepileptic drugs were used in thepatients who had seizures. Pyridoxine was given orally20e40 mg/day to all the patients. In patients with clinicalevidence of elevated intracranial pressure, 20% mannitoland acetazolamide were used. Mannitol was used in dosageschedule of 0.25e1.0 g/kg body weight every 6 hourly for2e7 days. Neurosurgical opinion sought in patients with un-satisfactory response to medical treatment.

Follow up and outcome assessment

Disability assessment was done using modified Barthel index(MBI), which is a 20 point scoring system. Assessment ofdisability included degree of dependence for bowel andbladder, grooming, toilet use, transfer, mobility, dressing,feeding, use of stairs and bathing. For each activity, a scoreof 0 indicated a complete dependence, and a score of 2 or 3indicated that the patient was able to perform activitiesindependently. A score of <12 indicated poor functional

s’s ratio (in this case more than 30%). Periventricular T2 FLAIRctive hydrocephalus (marked by arrows). (B) Axial T2 FLAIR

Hydrocephalus in tuberculous meningitis 333

status and a score of >12 indicated good functional status.Assessment of disability as per MBI was done at baseline andat the end of 1st, 3rd and 6th month of follow up. MBI >12,at the end of 6 months, was considered as good outcome.Follow up MR study was done after 6 months.

Statistical analysis

The statistical analysis was performed using the StatisticalPackage for Social Sciences, version 16 for windows (SPSS,Chicago IL, USA) and Microsoft Excel. Both univariate andmultivariate analysis were done to evaluate the predictorsfor hydrocephalus. Univariate analysis was performed byChi-square test for non-parametric data and student’s “t”test for independent variables for parametric data. Rela-tive risks with95% confidence interval were ascertained. Formultivariate analysis, binary logistic regression was per-formed to see the impact of individual predictors ofhydrocephalus. KaplaneMeier analysis was performed toestimate the event free survival for the outcome with orwithout baseline hydrocephalus using the Log Rank test.Statistical significance was defined at p value of <0.05. Sta-tistical analysis was two-tailed.

Results

Baseline characteristics

We enrolled 96 patients of tuberculous meningitis. Sixteenpatients were excluded. Reasons for exclusion of 16patients have been given in Fig. 2. Finally, 80 patientswere evaluated and followed up. Mean age of our cohortwas 30.1 (range 14e68) years. Mean duration of illnesswas 97.7 � 21.2 days. Majority (36; 45%) of the patients,

Figure 2 Flow diag

at the time of inclusion, were in stage- II. In our study,there were 36 (45%) bacteriologically confirmed cases.Four patients were lost to follow up and their last observa-tions were carried forward up to 6 months. None of the pa-tient was HIV positive. In none of the patients drugresistance to antituberculosis drug was demonstrated.Other base-line characteristics have been shown in Table 1.

Incidence of hydrocephalus

At inclusion, hydrocephalus was demonstrated in 52 (65%)patients of tuberculous meningitis. Twenty patients (38.5%)had mild hydrocephalus. Remaining, 32 (61.5%) patients hadmoderate or severe hydrocephalus (Evan’s ratio equal ormore than 0.34). Periventricular ooze was present in 20(38.5%) patients. Five patients had asymmetric dilatation oflateral ventricles. Sulcal effacement was seen in 32 patients(61.5%). Among patients with hydrocephalus, acute infarctswere noted in 22 patients (42.4%), basal exudates in 38(73.1%) and tuberculoma in 27 patients (51.9%). Among 32patients with moderate to severe hydrocephalus, 25 (48%)had a communicating type of hydrocephalus. And 7 (13.4%)patients had possible obstructive type of hydrocephalus.

Predictors of hydrocephalus

On univariate analysis, clinical factors associated withhydrocephalus were advanced stage tuberculous meningitis(p Z 0.0012), baseline MBI�12 (p Z 0.002), duration of ill-ness >2 months (p Z 0.002), diplopia (p Z 0.001), seizures(p Z 0.002), visual impairment (p Z 0.001), papilledema(p Z 0.001, cranial nerve palsy (p Z 0.001) and hemipare-sis (p Z 0.001). Pulmonary tuberculosis (p Z 0.001), CSFtotal cell count >100/cu.mm (p Z 0.009), CSF protein

ram of the study.

Table 1 Baseline characteristics in patients of tuberculous meningitis with and without hydrocephalus (n Z 80).

Characteristics Total N Z 80 Baselinehydrocephalus N Z 52

No baselinehydrocephalus N Z 28

Level ofsignificance

n % n % n % P

Age > 25 years 46 57.5 30 57.7 16 57.1 0.850Male 43 53.8 29 55.8 14 50.0 0.573Female 37 46.3 23 44.2 14 50.0Duration of illness> 2 months

43 53.8 35 67.3 08 28.6 0.002

Pulmonary tuberculosis 18 22.5 14 26.9 04 14.3 0.653MRC stage I 25 31.3 11 21.2 14 50.0 0.001MRC stage II 34 42.5 21 40.4 13 46.4MRC stage III 21 26.3 20 38.5 01 03.6Fever 76 95.9 51 98.1 25 89.3 0.192Headache 73 91.3 48 92.3 25 89.3 0.079Vomiting 59 73.8 41 78.9 18 64.3 0.088Seizures 30 37.5 26 50.0 04 14.3 0.006Altered sensorium 31 38.8 23 44.3 08 28.6 0.117Meningeal signs 71 88.8 48 92.3 23 82.1 0.170Diplopia 51 63.8 41 78.9 10 35.7 0.001Cranial nerveinvolvement

62 77.5 49 94.2 13 46.4 0.000

Papilledema 50 62.5 41 78.9 09 32.1 0.001Optic atrophy 10 12.5 08 15.4 02 07.2 0.001Normal vision 44 55.0 21 40.4 23 82.1Low vision 36 45.0 31 59.6 05 17.9 0.002CSF AFB stain 21 26.3 18 34.6 03 10.7 0.06CSF AFB culture 28 35.0 21 40.4 7 25.0 0.258CSF TB PCR 36 45.0 28 53.8 8 28.6 0.076Definite TBM 36 45.0 28 53.8 8 28.6Probable TBM 25 31.3 15 28.9 10 35.7Possible TBM 19 23.8 11 21.2 8 28.6 0.445Meningeal enhancement 74 92.5 50 96.2 24 85.7 0.091Basal exudates 42 52.5 38 73.1 04 14.3 0.000Tuberculoma 31 38.8 27 51.9 04 14.3 0.001Infarct 26 32.5 22 42.3 04 14.3 0.011Good outcome (MBI > 12) 57 71.3 31 59.6 26 92.9Poor outcome MBI (<12) 23 28.8 21 40.4 02 07.2 0.002

AFB Z acid-fast bacilli; MRC Z Medical Research Council; CSF Z cerebrospinal fluid; PCR Z polymerase chain reaction;TB Z tuberculosis; TBM Z tuberculous meningitis; MBI Z modified barthel index.

334 T. Raut et al.

>2.5 g/l (p Z 0.05) were also associated with hydrocepha-lus. Among the neuroimaging features, presence of basalexudates (p Z 0.001), tuberculomas (p Z 0.001) and in-farcts (p Z 0.011), were associated with occurrence of hy-drocephalus. On binary logistic regression analysis, factorsthat were significantly associated with presence of hydro-cephalus included vision impairment (p Z 0.005), cranialnerve involvement (p Z 0.012) and presence of basal exu-dates (p Z 0.001). These three predictors had a very highsensitivity, specificity (>80%) and a high positive predictivevalue (90.2%) with a fair accuracy (86.3%), for prediction ofhydrocephalus.

Follow up

During the course of study, 10 patients died (median survival50 days, range 25e110 days). All 10 patients, who died, had

hydrocephalus. Of 52 patients with hydrocephalus, 6 weresubjected to surgery (4 later died and 2 had partialimprovement). Among patients with baseline hydrocepha-lus, 31 (59.6%) had a good outcome (MBI > 12). Twenty-onepatients had improved completely after six months. Elevencases (21.2%) had a poor outcome (MBI � 12); clinicalcondition of 5 patients had worsened. In 28 patients withoutbaseline hydrocephalus, 26 (92.8%) had a good outcome(MBI > 12); among these patients 21 (75%) had improvedcompletely. Only 2 patients of later group had a pooroutcome, both experienced deterioration during follow up.

Follow up neuroimaging was done in 54 (32 withhydrocephalus and 22 without hydrocephalus, at inclusion)patients. Among patients with baseline hydrocephalus, 13patients had complete resolution and 12 patients hadpartial resolution. In 3 patients hydrocephalus persistedwhile in another 4 patients, ventricles size further in-creased. Of the 22 patients without baseline

Hydrocephalus in tuberculous meningitis 335

hydrocephalus, 8 patients developed new hydrocephalus.Of these 8 patients, only 2 had clinical deterioration. Mostof the factors, that were significantly associated withcomplete resolution of hydrocephalus, indicated presenceof a less severe form of tuberculous meningitis. (Table 2).

Figure 3 KaplaneMeier survival curves in patients of tuber-culous meningitis with or without hydrocephalus at observedat the time of inclusion in the study.

Prognosis of tuberculous meningitis after 6 months

In our study, poor prognostic factors of tuberculous men-ingitis were seizures (p Z 0.006), visual impairment(p Z 0.001), hemiparesis (p Z 0.001) and paraparesis(p Z 0.001). Baseline MBI � 12 (p Z 0.008) and advancedstage (Stage III) (p Z 0.001) were also associated withpoor outcome. Among the neuroimaging parameters, pres-ence of hydrocephalus (p Z 0.002) and basal exudates(p Z 0.001) were associated with poor outcome. On multi-variate analysis, none of these factors were significantly as-sociated with poor outcome.

Prognosis among patients with baselinehydrocephalus

Factors associated with poor outcome (death or disability),among patients with baseline hydrocephalus, were pres-ence of hemiparesis (p Z 0.029), stage III tuberculous men-ingitis (p Z 0.002), pulmonary tuberculosis (p Z 0.006) andpoor baseline MBI � 12 (p Z 0.03). KaplaneMeier cumula-tive survival curve analysis showed that the patients withbaseline hydrocephalus had significantly higher incidenceof death and severe disability comparable to those withouthydrocephalus (p Z 0.006). (Fig. 3).

Discussion

In our study, hydrocephalus was present in majority (65%)of the patients. During follow up, hydrocephalus developedin 8 new cases. Various studies, in the past, have reported

Table 2 Baseline characteristics predicting complete resolutionsis) (n Z 32).

Characteristic Total Complete resolution (n

n n %

Duration of illness < 2 months 13 7 53.5MRC stage I & II 19 11 84.6

Seizures 13 05 38.5Initial GCS 15/15 18 11 84.6

Normal vision 21 11 84.6Hemiparesis 08 01 07.7Baseline MBI > 12 19 12 92.3

CSF TLC < 100/cu.mm 17 11 84.6

CSF protein < 2.5 g% 17 11 84.6

Mild hydrocephalus 16 12 92.3

Basal exudates 12 03 23.1Infarct 12 03 23.1Tuberculoma 09 02 15.4

CSF Z cerebrospinal fluid; GCS- Z Glasgow coma scale; TLC Z totalValues in bold are statistically significant.

the incidence of hydrocephalus ranging from 17 to 95%.15e19

Earlier, air encephalography was used for the diagnosis ofhydrocephalus. In an air encephalography based study, hy-drocephalus was observed in 62% of the pediatric patientsof tuberculous meningitis.20

Hydrocephalus can develop early in the course of thedisease or can develop paradoxically, following antituber-culosis treatment.21,22 The predictive factors of hydroceph-alus, in patients with tuberculous meningitis, have scantlybeen evaluated. In a study, long duration of illness, ad-vanced stage of disease, focal neurological deficit, andpresence of infarcts were significantly associated with hy-drocephalus.21 In our study as well, hydrocephalus wasmore frequent in patients with advanced disease, severedisability and a long-duration of illness. Presence of

of baseline hydrocephalus after 6 months (Univariate analy-

Z 13) Partial or no resolution (n Z 19) P value

n %

06 31.6 0.28108 42.1 0.028

08 42.1 0.87307 36.8 0.012

10 52.6 0.13607 36.8 0.14607 36.8 0.003

06 31.6 0.005

06 31.6 0.005

04 21.1 0.001

09 47.4 0.30709 47.4 0.30707 36.8 0.355

leucocyte count; MRC Z Medical Research Council.

336 T. Raut et al.

hydrocephalus is, in fact, an indication of severe disease.Most of the complications of tuberculous meningitis, includ-ing hydrocephalus, develop because of thick copious exu-dates, which are dominantly present in theinterpeduncular, suprasellar, and Sylvian cisterns. Opto-chiasmatic arachnoiditis is a devastating form of tubercu-lous meningitis and often associated with profound visionloss and hydrocephalus.2,21 Hydrocephalus is often associ-ated with raised intracranial pressure and several othercomplications of tuberculous meningitis. Increased intra-cranial pressure, if remains untreated, results in alterationin consciousness. In addition to this, elevated intracranialpressure can lead to stretching of already compromisedvessels (like the pericallosal arteries) and may produce in-farcts.23 Optic nerve may be compressed by enlarged di-lated third ventricle resulting in vision loss and opticatrophy. Thus, clinical parameters like vision impairmentand cranial nerve palsy can act as important clinical indica-tors for the presence of hydrocephalus, even before neuro-imaging is obtained.

In earlier studies, hydrocephalus had been found a de-terminant of poor outcome and mortality.1,17,21,24 In ourstudy, as well, hydrocephalus unfavorably affected overallprognosis in patients with tuberculous meningitis. Hydro-cephalus resolution was more frequent in patients withmilder form of disease. Corticosteroids have shown to re-duce the number of deaths and increase the survival in adultpatients, however, role of dexamethasone, in the treatmentof hydrocephalus, is uncertain.25 Thwaites and co-workerscould not observe any significant reduction in the proportionof patients with meningeal enhancement or hydrocephalusat days 60 or 270 of dexamethasone treatment comparedwith placebo.12 Possibly, this is the reason which explainslimited long-term benefit of corticosteroids.26 Mannitol isfrequently used in tuberculous meningitis, but its exactrole needs to be established by well-designed trials. Itmay have a beneficial effect on lowering the intracranialpressure, but not on survival or disability.7,27

In our study, 6 patients were subjected to CSF diversionprocedures. Shunt surgery is, generally, indicated in pa-tients with obstructive hydrocephalus. Shunt surgery is alsoconsidered in patients with either an advanced disease ora deteriorating clinical condition. Currently, two types ofsurgeries are performed; ventriculo-peritoneal shunt orendoscopic third ventriculostomy. Endoscopic third ventri-culostomy creates a communication between third ventri-cle and subarachnoid space bypassing cerebral aqueduct. Itis now considered as a safe and effective treatment optionfor hydrocephalus.7 Unfortunately, not all patients, whoundergo shunt surgery, show a significant improvement. Insome centers shunt surgery is performed when the clinicalcondition of the patient improves following temporary ex-ternal ventricular drainage.4 Studies have shown that tu-berculous meningitis with hydrocephalus requiring CSFdiversion carries significant short-term mortality.4 The pa-tients, who had cerebral infarcts, had an even worse out-come. Mortality in the group with infarcts was 100%,compared to 42% in the group without infarcts.28

We used the Evan’s ratio to diagnose hydrocephalusbecause of its simplicity and its ability to identify ventri-culomegaly.29 Recently, Evan’s ratio has been criticized asits value can vary, depending on the level of the brain

scan image at which the frontal horns and maximal innerskull diameters are measured.30 However, Evan’s ratio isstill frequently used method to diagnose hydrocephalus.31

The distinction between two types of hydrocephalus (com-municating and obstructive type) is often not clear cut. Forexample, if all the ventricles are enlarged, then it is diffi-cult to establish whether the hydrocephalus is due to ob-struction of the fourth ventricular outlets or due toan obstruction to the CSF pathways in the basal subarach-noid spaces. In patients with tuberculous meningitis com-municating hydrocephalus is much more common thanobstructive hydrocephalus.4

In conclusion, hydrocephalus is a common complicationof tuberculous meningitis.

If a patient had neurological deficits, seizures or is in badclinical conditions, cranial imaging should be performed tocheck for hydrocephalus. Majority of patients having mildhydrocephalus had a good outcome and complete resolu-tion with medical management. CSF diversion proceduresmay not bring desired results in rapidly deterioratingpatients.

Appendix A. Supplementary data

Supplementary data related to this article can be found athttp://dx.doi.org/10.1016/j.jinf.2012.12.009.

References

1. Brancusi F, Farrar J, Heemskerk D. Tuberculous meningitis inadults: a review of a decade of developments focusing on prog-nostic factors for outcome. Future Microbiol 2012;7:1101e16.

2. Marx GE, Chan ED. Tuberculous meningitis: diagnosis and treat-ment overview. Tuberc Res Treat 2011;011:798764.

3. Garg RK. Tuberculous meningitis. Acta Neurol Scand 2010;122:75e90.

4. Rajshekhar V. Management of hydrocephalus in patients withtuberculous meningitis. Neurol India 2009;57:368e74.

5. Yadav YR, Parihar V, Agrawal M, Bhatele PR. Endoscopic thirdventriculostomy in tubercular meningitis with hydrocephalus.Neurol India 2011;59:855e60.

6. Tandon V, Mahapatra AK. Management of post-tubercular hy-drocephalus. Childs Nerv Syst 2011;27:1699e707.

7. Figaji AA, Fieggen AG. The neurosurgical and acute care man-agement of tuberculous meningitis: evidence and current prac-tice. Tuberculosis (Edinb) 2010;90:393e400.

8. Schoeman J, Donald P, van Zyl L, Keet M, Wait J. Tuberculoushydrocephalus: comparison of different treatments with re-gard to ICP, ventricular size and clinical outcome. Dev MedChild Neurol 1991;33:396e405.

9. Pienaar M, Andronikou S, van Toorn R. MRI to demonstrate di-agnostic features and complications of TBM not seen with CT.Childs Nerv Syst 2009;25:941e7.

10. Suzaan M, Thwaites GE, Schoeman JF, Torok ME, Misra UK,Donald PR. Tuberculous meningitis: a uniform case definitionfor use in clinical research. Lancet Infect Dis 2010;10:803e12.

11. British Medical Research Council. Streptomycin treatment oftuberculous meningitis. BMJ 1948;1:582e97.

12. Thwaites GE, Macmullen-Price J, Tran TH, Pham PM, Nguyen TD,Simmons CP, et al. Serial MRI to determine the effect of dexa-methasone on the cerebral pathology of tuberculous meningitis:an observational study. Lancet Neurol 2007;6:230e6.

Hydrocephalus in tuberculous meningitis 337

13. Santos de Oliveira R, Barros Juc�a CE, Valera ET, Machado HR.Hydrocephalus in posterior fossa tumors in children. Are therefactors that determine a need for permanent cerebrospinalfluid diversion? Childs Nerv Syst 2008;24:1397e403.

14. WorldHealthOrganization.Treatmentof tuberculosis: guidelines.4th ed. (WHO/HTM/TB/2009.420)World Health Organization,http://whqlibdoc.who.int/publications/2010/9789241547833_eng.pdf; 2010 [accessed 17.10.12].

15. Bhargava S, Gupta AK, Tandon PN. Tuberculous meningitis-A CTscan study. Br J Radiol 1982;55:189e96.

16. Kalita J, Misra UK, Nair PP. Predictors of stroke and its outcome intuberculousmeningitis. J StrokeCerebrovascDis2009;18:251e8.

17. Hsu PC, Yang CC, Ye JJ, Huang PY, Chiang PC, Lee MH. Prognos-tic factors of tuberculous meningitis in adults: a 6-year retro-spective study at a tertiary hospital in northern Taiwan.J Microbiol Immunol Infect 2010;43:111e8.

18. Anderson NE, Somaratne J, Mason DF, Holland D, Thomas MG.Neurological and systemic complications of tuberculous menin-gitis and its treatment at Auckland City Hospital, New Zealand.J Clin Neurosci 2010;17:1114e8.

19. Sharma P, Garg RK, Verma R, Singh MK, Shukla R. Incidence,predictors and prognostic value of cranial nerve involvementin patients with tuberculous meningitis: a retrospective evalu-ation. Eur J Intern Med 2011;22:289e95.

20. Lorber J. Studies of the CSF circulation in tuberculous menin-gitis in children. Part II. A review of 100 pneumoencephalo-grams. Arch Dis Child 1951;26:28e48.

21. Chan KH, Cheung RT, Fong CY, Tsang KL, Mak W, Ho SL. Clinicalrelevance of hydrocephalus as a presenting feature of tubercu-lous meningitis. QJM 2003;96:643e8.

22. Tung YR, Lai MC, Lui CC, Tsai KL, Huang LT, Chang YC, et al. Tu-berculous meningitis in infancy. Pediatr Neurol 2002;27:262e6.

23. Belorgey L, Lalani I, Zakaria A. Ischemic stroke in the setting oftuberculous meningitis. J Neuroimaging 2006;16:364e6.

24. Pehlivanoglu F, Yasar KK, Sengoz G. Tuberculous meningitis inadults: a review of 160 cases. Scientific World J 2012;2012:169028.

25. Thwaites GE, Nguyen DB, Nguyen HD, Hoang TQ, Do TT,Nguyen TC, et al. Dexamethasone for the treatment of tuber-culous meningitis in adolescents and adults. N Engl J Med 2004;351:1741e51.

26. Torok ME, Nguyen DB, Tran TH, Nguyen TB, Thwaites GE,Hoang TQ, et al. Dexamethasone and long-term outcome of tu-berculous meningitis in Vietnamese adults and adolescents.PLoS One 2011;6:e27821.

27. Murthy JM. Management of intracranial pressure in tuberculousmeningitis. Neurocrit Care 2005;2:306e12.

28. Clemente Morgado T, Kinsky M, Carrara H, Rothemeyer S,Semple P. Prognostic value of CT-evident cerebral infarctsin adult patients with tuberculous meningitis and hydrocepha-lus treated with EVD. World Neurosurg 2012 Oct 3;12:01086e8. http://dx.doi.org/10.1016/j.wneu.2012.09.021. pii:S1878e8750[Epub ahead of print].

29. Lee SH, Seol HJ, Kong DS, Nam DH, Park K, Kim JH, et al. Riskfactors and tumor response associated with hydrocephalus af-ter gamma knife radiosurgery for vestibular schwannoma. ActaNeurochir (Wien) 2012;154:1679e84.

30. Toma AK, Holl E, Kitchen ND, Watkins LD. Evans’ index revis-ited: the need for an alternative in normal pressure hydroceph-alus. Neurosurgery 2011;68:939e44.

31. Honeybul S, Ho KM. Incidence and risk factors for post-traumatic hydrocephalus following decompressive craniectomyfor intractable intracranial hypertension and evacuation ofmass lesions. J Neurotrauma 2012;29:1872e8.