Diagnostic accuracy of a uniform research case definition for TBM in children: a prospective study.

BACKGROUND: Bacteriological confirmation of tuberculous meningitis (TBM) is problematic, and rarely guides initial clinical management. A uniform TBM case definition has been proposed for research purposes. METHODS: We prospectively enrolled patients aged 3 months to 13 years with meningitis confirmed using cerebrospinal fluid analysis at Tygerberg Hospital, Cape Town, South Africa. Criteria that differentiated TBM from other causes were explored and the accuracy of a probable TBM score assessed by comparing bacteriologically confirmed cases to 'non-TBM' controls. RESULTS: Of 139 meningitis patients, 79 were diagnosed with TBM (35 bacteriologically confirmed), 10 with bacterial meningitis and 50 with viral meningitis. Among those with bacteriologically confirmed TBM, 15 were Mycobacterium tuberculosis culture-positive and 20 were culture-negative but positive on GenoType(®) MTBDRplus or Xpert(®) MTB/RIF; 18 were positive on only a single commercial nucleic acid amplification test. A probable TBM score provided a sensitivity of 74% (95%CI 57-88) and a specificity of 97% (95%CI 86-99) compared to bacteriologically confirmed TBM. CONCLUSION: A probable TBM score demonstrated excellent specificity compared to bacteriological confirmation. However, 26% of children with TBM would be missed due to the limited accuracy of the case definition. Further prospective testing of an algorithm-based approach to TBM is advisable before recommendation for general clinical practice.

The diagnostic value of cerebrospinal fluid chemistry results in childhood tuberculous meningitis.

PURPOSE: Cerebrospinal fluid (CSF) hypoglycorrhachia and elevated protein is well-described in bacterial meningitis, but evidence for its differential diagnostic value in tuberculous meningitis (TBM) is lacking. We aimed to assess the diagnostic utility of CSF glucose, CSF to serum glucose ratio and CSF protein in children with suspected TBM. METHODS: We describe CSF glucose and protein values as well as CSF to serum glucose ratios in a prospective evaluation of TBM suspects seen at Tygerberg Children's Hospital, Cape Town, South Africa, from January 1985 to January 2014. RESULTS: Of 615 TBM suspects, 88 (14%) had microbiologically confirmed TBM, 381 (62%) 'probable' TBM and 146 (24%) 'non-TBM'. Mean absolute CSF glucose concentration was significantly lower in the microbiologically confirmed (1.87 ± 1.15 mmol/L) and 'probable' TBM (1.82 ± 1.19 mmol/L) groups compared to non-TBM (3.66 ± 0.88 mmol/L). A CSF glucose concentration of <2.2 mmol/L diagnosed TBM with sensitivity 0.68 and specificity 0.96. Sensitivity using a CSF to serum glucose ratio of <0.5 was 0.90. Mean CSF protein was significantly elevated in the microbiologically confirmed TBM (1.91 ± 1.44 g/L) and 'probable' TBM (2.01 ± 1.49 g/L) groups compared to the non-TBM (0.31 ± 0.31 g/L). A CSF protein >1 g/L diagnosed TBM with sensitivity 0.78 and specificity 0.94. CONCLUSION: Absolute CSF glucose values of <2.2 mmol/L and protein values of >1 g/L differentiated between TBM and non-bacterial meningitis with good specificity, although sensitivity was poor. A CSF to serum glucose ratio is more informative than the absolute value.

Chest radiograph findings in children with tuberculous meningitis.

BACKGROUND: Tuberculous meningitis (TBM) is diagnosed based on a combination of clinical, laboratory and radiological findings, including signs suggestive of tuberculosis (TB) on a standard chest X-ray (CXR). METHODS: We describe the radiological features suggestive of intrathoracic TB in children diagnosed with TBM during a prospective evaluation of TBM suspects seen at Tygerberg Children's Hospital, Cape Town, South Africa. RESULTS: Of 84 children treated for TBM, 31 (37%) had 'definite' TBM, 45 (55%) 'probable' TBM and 8 (9%) 'possible' TBM. In total, 37 (44%) TBM patients had CXR findings suggestive of TB, 9 (11%) with disseminated (miliary) TB. Only 1 in 4.39 children aged ≤3 years with TBM had suggestive CXR findings. The presence of complicated intrathoracic lymph node disease was significantly higher in children aged ≤3 years (OR 21.69, 95%CI 2.73-172.67, P < 0.01). Among 6 human immunodeficiency virus infected children, 3 (50%) had intrathoracic lymphadenopathy. CONCLUSION: The majority of the children with TBM, including the very young, did not have signs suggestive of TB on CXR.

Improved diagnosis of childhood tuberculous meningitis using more than one nucleic acid amplification test.

BACKGROUND: Early treatment is critical to reducing tuberculous meningitis (TBM) related morbidity and mortality. Diagnosis based on cerebrospinal fluid (CSF) culture is impractical due to slow turnaround times, while microscopy has poor sensitivity. Enhanced detection methods are essential to guide early treatment initiation, especially in vulnerable young children. METHODS: We assessed the diagnostic accuracy of the GenoType(®) MTBDRplus and Xpert(®) MTB/RIF assays on CSF collected from paediatric meningitis suspects prospectively enrolled at Tygerberg Hospital, Cape Town, South Africa. Fluorescent auramine-O microscopy, liquid culture for Mycobacterium tuberculosis, GenoType and Xpert assays were performed on all CSF samples. RESULTS: Of 101 meningitis suspects, 55 were diagnosed with TBM and 46 served as non-TBM controls. Using a pre-defined TBM case definition as reference standard, sensitivities and specificities were 4% and 100% for fluorescent microscopy, 22% and 100% for culture, 33% and 98% for GenoType, 26% and 100% for Xpert, 22% and 100% for microscopy and culture combined and 49% and 98% for GenoType and Xpert combined. Culture, GenoType and Xpert combined performed best, with 56% sensitivity and 98% specificity. CONCLUSION: Although commercial nucleic-acid amplification tests performed on CSF revealed incrementally improved diagnostic accuracy, providing rapid microbiological confirmation, they cannot serve as a rule-out test.

Tuberculous meningitis.

Tuberculous (TB) meningitis is common in resource-poor communities but also occurs in developed countries where the diagnosis is frequently delayed because of unfamiliarity with the disease. TB meningitis develops whenever a small intracranial tuberculoma (Rich focus) ruptures causing predominantly basal meningitis. This results in hydrocephalus, cranial nerve palsies, and ischemic brain injury secondary to tuberculous vasculitis. The primary source of TB is usually the lung. Early diagnosis is difficult because patients tend to present subacutely with nonspecific symptoms such as fever, cough, vomiting, listlessness, and failure to thrive. Lumbar puncture typically shows clear and colorless CSF with a low, predominantly lymphocytic, leukocytosis and raised protein and low glucose levels. Decision to treat is mostly clinical because of difficulty in demonstrating TB bacilli on routine smear and time needed for culture. A positive TB contact, Mantoux skin test, chest radiograph, CT/MRI brain scan, PCR for tuberculosis on CSF, and demonstration of TB bacilli from extraneural sources are invaluable in supporting a diagnosis of TB meningitis. Current World Health Organization guidelines recommend treatment with a four-drug regimen for 2 months, followed by a two-drug regimen for 10 months, the total duration of treatment being 12 months. Corticosteroids reduce mortality without changing neurological morbidity. Outcome relates closely to age and stage of disease.

Seasonal variation in the incidence rate of tuberculous meningitis is associated with sunshine hours.

Tuberculous meningitis (TBM) is a severe complication of tuberculosis and occurs mainly during early childhood. The incidence rate of TBM varies with season, and serum vitamin D levels, which are dependent on sunlight, might play a role. We studied the association between TBM incidence rate and hours of sunshine in Cape Town, South Africa and found a significant association between the incidence rate of TBM and hours of sunshine 3 months earlier (incidence rate ratio per 100 sunshine hours 0·69, 95% confidence interval 0·54-0·88, P = 0·002). The association supports the hypothesis that vitamin D might play a role in the pathophysiology of TBM. Further prospective studies in which vitamin D status is measured are necessary to determine causality.

Value of different staging systems for predicting neurological outcome in childhood tuberculous meningitis.

BACKGROUND: The stage of tuberculous meningitis (TBM) at presentation is strongly associated with prognosis. OBJECTIVE: To compare different staging systems for TBM in predicting outcome. METHODS: The associations of different staging systems with neurological outcome were assessed using clinical, diagnostic and outcome data of 554 TBM children admitted to Tygerberg Children's Hospital from January 1985 to April 2005. RESULTS: The refined Medical Research Council (MRC) scale after 1 week (84%, 95%CI 81-88) had the highest discriminatory power in predicting neurological morbidity. The Glasgow Coma Score (GCS) on admission, GCS after 1 week, the 'refined' MRC scale on admission and Tygerberg Children's Hospital (TCH) staging method all had excellent discriminatory powers in predicting outcome. Improvement of staging after 1 week occurred in children who did not have signs of raised intracranial pressure (P = 0.004) or brainstem dysfunction on admission (P = 0.030). CONCLUSIONS: The 'refined' MRC scale 1 week after diagnosis showed the best association with neurological outcome after 6 months of treatment. The excellent discriminatory power of the TCH scale and its simplicity of use make it the ideal scale for use in resource-poor settings.

Behaviour profiles after tuberculous meningitis.

The purpose of the study was to investigate child behaviour in children who recovered from tuberculous meningitis (TBM) and to compare behaviour profiles of stage II and stage III patients. The mean age of the cohort of 74 children at the time of evaluation was 10 years and 7 months. At follow-up all patients underwent a thorough neurological examination and a psychometric test battery, which included intellectual assessment and evaluation of behaviour by means of the CBCL/6-18. Results indicated elevated mean scores (T > 60) on CBCL/6-18 scales which measure problems with anxiety, depression, attention, social relationships, disruptive and rule-breaking behaviour. Mean CBCL scores of stage III patients were significantly higher than the mean scores of stage II patients on scales which measure social problems, disruptive and rule-breaking behaviour. In addition, problems with conduct, attention, attention-deficit/hyperactivity problems, affective problems as well as the total problem scores were more pronounced in the patients with stage III TBM. We conclude that general behavioural disinhibitions as well as internalized emotional disorder probably are long-term complications in more than 10% of the survivors of TBM.

Tuberculous meningitis and miliary tuberculosis: the Rich focus revisited.

Tuberculous meningitis (TBM) develops most often when a caseating meningeal or sub-cortical focus, the Rich focus, discharges its contents into the subarachnoid space. It is recognized that TBM is frequently accompanied by miliary tuberculosis, but the relationship between the development of the Rich focus and miliary tuberculosis remains controversial. The original descriptions of Arnold Rich and Howard McCordock are reviewed together with the work of other pathologists and the observations of the natural history of tuberculosis by astute clinicians such as Arvid Wallgren and Edith Lincoln. Rich and McCordock dissociated miliary tuberculosis from a role in the pathogenesis of TBM, and this view continues to appear in reviews and textbooks dealing with TBM. We suggest, particularly in childhood, that miliary tuberculosis is indeed directly involved in the pathogenesis of TBM in as much as that the overwhelming bacillaemia that accompanies miliary tuberculosis serves to increase the likelihood that a meningeal or sub-cortical Rich focus will be established, which may in its turn caseate and give rise to TBM.

Tuberculous meningitis and miliary tuberculosis in young children.

OBJECTIVES: To document the clinical and diagnostic features of tuberculous meningitis (TBM) in young children with and without concomitant miliary tuberculosis (TB). METHODS: A retrospective comparative study. RESULTS: Of 104 children with TBM, 32 (31%), median age 17.0 months, had a miliary appearance on chest radiograph; 72 (69%), median age 30.5 months, had TBM only (P = 0.04). Mediastinal adenopathy was noted in 27 (84%) of the children with miliary TB and 33 (46%) of those with TBM only (P = 0.0005). The mean cerebrospinal fluid (CSF) lymphocyte and polymorphonuclear counts of all children (no significant differences between groups) were 137 x 10(6)/l and 38 x 10(6)/l and the mean protein and glucose concentrations were 1.45 g/l and 0.72 mmol/l, respectively. Polymorphonuclear leukocytes were predominant in the CSF of 17% of children, in 16% the CSF glucose was > 2.2 mmol/l and in 26% the CSF protein was < 0.8 g/l. On Mantoux testing 37 (65%) of 57 children with TBM only and 12 (48%) of 25 children with TBM and miliary TB had an induration of > or = 10 mm (P = 0.23). Ten children (10%) died, five (7%) who had TBM only and five (16%) who had TBM and miliary TB. CONCLUSION: Children with TBM and miliary TB were younger and more likely to have mediastinal adenopathy on chest radiography than those with TBM only. Diagnostic features and investigations in both groups may be misleading at times.

Can CT predict the level of CSF block in tuberculous hydrocephalus?

INTRODUCTION: Treatment of obstructive hydrocephalus in children with tuberculous meningitis (TBM) depends on the level of the cerebrospinal fluid (CSF) block. Air-encephalography is regarded as the gold standard for differentiating communicating and non-communicating hydrocephalus. Since air-encephalography involves a lumbar puncture, it carries the risk of cerebral herniation. AIM. The aim of this study was to determine whether communicating and non-communicating hydrocephalus in TBM can be differentiated by means of cranial computerised tomography (CT). METHODS: A number of CT indices were measured in 50 children with communicating and 34 children with non-communicating hydrocephalus according to air-encephalographic findings. RESULTS: The only CT finding that correlated with the type of hydrocephalus was the shape of the third ventricle. Significantly more children with non-communicating hydrocephalus had a rounded third ventricle than those with communicating hydrocephalus. CONCLUSION: CT is therefore not useful in determining the level of CSF block in TBM. Air-encephalography remains the most reliable way of determining the level of CSF obstruction.

Tuberculosis meningitis and attention deficit hyperactivity disorder in children.

The purpose of this study was to investigate the prevalence of attention deficit hyperactivity disorder (ADHD) in children who recovered from tuberculous meningitis (TBM) as part of an ongoing TBM research project. During this study, each TBM group subject underwent a thorough clinical-neurological examination, and a test battery which included the child behaviour check list (CBCL) Teacher's Report Form and Conners Rating Scale. The parents and teachers of each of the 21 TBM group and 21 control group subjects completed the above-mentioned questionnaires. All 21 TBM group subjects displayed symptoms of ADHD. The TBM group was significantly more hyperactive and unable to sustain attention than the control group. Furthermore, TBM group subjects were perceived as being significantly more unpopular, obsessive, compulsive and aggressive than the control group subjects. With regard to the frequency of externalizing behaviour, the TBM group subjects displayed significantly more externalizing behaviours as well as symptoms of attention deficit and hyperactivity. No significant differences between parents' and teachers' ratings were found. We conclude that ADHD is a common long-term complication of TBM.

The effect of adjuvant steroid treatment on serial cerebrospinal fluid changes in tuberculous meningitis.

Three recent studies found that corticosteroids improve clinical outcome and mortality in tuberculous meningitis (TBM), although the exact mechanism of action of the drug remains speculative. A number of reports on the effect of corticosteroids on cerebrospinal fluid (CSF) findings in TBM have been published, often with conflicting results regarding serial cell counts and protein levels. As part of a controlled, randomized trial on the effect of oral prednisone on outcome in childhood TBM at our institution, CSF was collected and analysed weekly during the 1st month of treatment. We found no significant difference in serial CSF cell counts between the steroid and non-steroid groups in the study. However, the steroid group had significantly lower CSF protein and globulin levels after the 1st month of treatment, and a more steady rise in CSF glucose levels than the non-steroid group. Knowledge of the different CSF responses during the course of anti-tuberculosis therapy is important in clinical decision-making.

Possible role of adjunctive thalidomide therapy in the resolution of a massive intracranial tuberculous abscess.

We present the case of a young child who developed a massive tuberculous abscess of the posterior fossa while being treated for pulmonary tuberculosis. Clinical improvement after surgical excision of the abscess was followed by recurrence of symptoms of acutely raised intracranial pressure on standard antituberculosis and corticosteroid therapy. Magnetic resonance (MR) imaging of the brain showed that a multiloculated abscess had developed anterior to the excision site of the original abscess. The recurring abscess was partly excised and drained but could not be removed completely because of its proximity to the brain stem. Thalidomide, a potent inhibitor of tumour necrosis factor alpha (TNF-alpha), was added to the treatment regimen and resulted in marked clinical improvement with resolution of the abscess within 4 months. The remaining CT lesion had the appearance of a small granuloma. Both the clinical and the radiological response was maintained after 1 year of antituberculosis treatment.

Tuberculous granulomas in childhood tuberculous meningitis: radiological features and course.

The clinical course and serial cranial computerized tomographic (CT) findings of 202 children with tuberculous meningitis (TBM) admitted to Tygerberg Hospital between 1985 and 1994 were reviewed with regard to the incidence, CT appearance and clinical course of associated intracranial tuberculous granulomas. Thirty-four patients (16.85 per cent) had associated intracranial granulomas. Thirty-eight individual lesions were analysed and classified as meningeal, parenchymal or ependymal according to their central nervous system (CNS) location. Twenty-five patients had round to irregular, brain iso-, hypo- or hyperdense meningeal granulomas with variable degrees of enhancement and peri-lesional hypodensities. Four patients had diffusely enhancing, brain isodense, enplaque-like ependymal granulomas associated with the ventricular ependymal lining. Four patients with miliary tuberculosis and TBM showed multiple small diffusely enhancing, brain iso- or hyperdense parenchymal lesions and associated hypodensities on initial CT. Although granulomas in the meningeal and ependymal group had the propensity to paradoxically enlarge or appear on standard four-drug antituberculosis therapy, the majority resolved uneventfully. Rapid resolution of small parenchymal granulomas associated with miliary tuberculosis occurred in all cases. Most granulomas in this series were co-incidental, asymptomatic CT findings. In rare cases, the development or enlargement of a strategically located granuloma may result in complications.

Adjunctive thalidomide therapy of childhood tuberculous meningitis: possible anti-inflammatory role.

The objective of this study was to determine the safety and tolerability of the immunomodulatory agent thalidomide as adjunct therapy in children with tuberculous meningitis. Children with stage 2 tuberculous meningitis received oral thalidomide for 28 days in a dose-escalating study, in addition to standard four-drug antituberculosis therapy, corticosteroids, and specific treatment of complications such as raised intracranial pressure. Clinical and laboratory evaluations were carried out. Fifteen patients (median age, 34 months) were enrolled. Thalidomide was administered via nasogastric tube in a dosage of 6 mg/kg/day, 12 mg/kg/day, or 24 mg/kg/day. The only adverse events possibly related to the study drug were transient skin rashes in two patients. Levels of tumor necrosis factor-alpha in the cerebrospinal fluid decreased markedly during thalidomide therapy. Clinical outcome and neurologic imaging showed greater improvement than that experienced with historical controls. Thalidomide appeared safe and well tolerated in children with stage 2 tuberculous meningitis and could have important anti-inflammatory effects. These promising results have led us to embark on a randomized, double-blind, placebo-controlled trial of the efficacy of thalidomide in tuberculous meningitis.

Serial lumbar CSF presure measurements and cranial computed tomographic findings in childhood tuberculous meningitits.

Intracranial pressure (ICP) was monitored in 218 consecutive children with hydrocephalus secondary to tuberculous meningitis (TBM). All children underwent cranial computerized tomographic (CT) scanning and continuous lumbar cerebrospinal (CSF) pressure monitoring on admission. Noncommunicating hydrocephalus (37 children), as determined by air encephalography, was treated by ventriculoperitoneal (VP) shunting and communicating hydrocephalus (181 children), by means of daily acetazolamide and frusemide. Response of ICP to treatment in the group with communicating hydrocephalus was assessed by means of repeated CSF pressure monitoring and CT scanning. One hundred and eighty-five of the 218 patients survived the 1st month of treatment. The aim of this study was the retrospective determination of (1) the relationship between ICP measurements and CT findings on admission and (2) the characteristics of the ICP recording which correlated best with the CT criteria of compensated hydrocephalus after the 1st month of treatment. No relationship was found between the level of baseline CSF pressure and the degree of hydrocephalus, as demonstrated by CT scanning, on admission. Seventy-five per cent of the patients with communicating hydrocephalus that survived the 1st month of treatment complied with the CT criteria for compensated hydrocephalus. All these patients had a baseline CSF pressure below 15 mmHg and absence of high-amplitude B waves on the pressure recording done at the end of the 1st month. In this study repeated lumbar CSF pressure monitoring proved to be an effective instrument to assess the response of communicating tuberculous hydrocephalus to medical treatment and also accurately predicted the timing of compensation of the hydrocephalus.