Tuberculous meningitis and HIV
http://www.100md.com
《美国医学杂志》
1 Division of Pediatric Neurology, Department of Pediatrics, Lokmanya Tilak Municipal Medical College and General Hospital, Sion, Mumbai, India
2 Department of Radiology, Lokmanya Tilak Municipal Medical College and General Hospital, Sion, Mumbai, India
3 Department of Microbiology, Lokmanya Tilak Municipal Medical College and General Hospital, Sion, Mumbai, India
Abstract
Objective: To identify factors associated with HIV-infected status in children admitted with tuberculous meningitis (TBM), and to find out whether HIV co-infection affects in-hospital outcome. Methods: This prospective hospital-based study was conducted from May 2000 to August 2003. All consecutive children, aged 1 month to 12 years of age, admitted with a diagnosis of TBM were enrolled. Relationship between 35 features viz., two demographic factors, nine clinical features, 13 neurological features, five laboratory (including cerebrospinal fluid) parameters, six radiological (including computed tomography scan brain) features, and the two outcomes (disabled survivor or death); with HIV-infected status was assessed. Results: Of a total 123 TBM cases enrolled, eight (6.5%) were HIV-infected. There was no significant difference between the two groups, except that more children in the HIV-infected group had Hb< 8 gm/dl: both on bivariate analysis, (OR, 12.0; 95% CI, 2.6-55.9; P = 0.001) and on multivariate analysis (OR, 12.30; 95% CI, 1.9-79.6; P = 0.008). Outcome was similar in both the groups. Conclusion: Only presence of Hb< 8 gm/dl was associated with HIV-infected status. HIV co-infection did not affect the outcome.
Keywords: HIV Infections; Multivariate Analysis; Tomography, X-Ray Computed; Treatment outcome; Tuberculosis; Meningeal
Since the early 1990's, human immunodeficiency virus (HIV) infection is assuming alarming proportions in Mumbai.[1] This HIV epidemic has been accompanied by a corresponding increase in the incidence of tuberculosis (TB) and HIV-related TB cases.[2],[3] Between 1-2% of children with untreated tuberculous infection develop tuberculous meningitis (TBM), a serious illness which has a high rate of neuromorbidity and mortality.[4],[5] Co-infection of TBM and HIV is likely to further complicate this problem.[6]
The objective of the present study was to: (i) find out the incidence of HIV co-infection in children admitted with TBM to our hospital, (ii) identify factors which are associated with HIV-infected status and (iii) find out whether HIV co-infection affects the in-hospital outcome.
Materials and methods
Patient Enrolment
All consecutive children, aged 1 month to 12 years of age, who were admitted to our hospital with a diagnosis of TBM, were enrolled prospectively. The study was conducted from May 2000 to August 2003. The diagnosis of TBM was based on a standard clinical case definition table1 devised by Doerr et al.[7]
At the time of admission, a complete physical examination was performed. Cerebrospinal fluid (CSF) examination and cranial computed tomography (CT) scan were done soon after admission in every child. A standardized data-entry form was used to document demographic data, clinical symptoms and signs, laboratory findings, Mantoux test result, chest radiograph and ultrasound abdomen findings, CSF and Cranial CT scan findings of each patient were conducted at presentation. Histopathological examinations of lymph nodes were carried out wherever appropriate.
Each patient was screened for HIV infection using the enzyme-linked immunosorbent assay (ELISA) test. Pre and post- test counseling was offered to the parents. The ELISA kits were used in strict compliance with the manufacturer's instructions. The kits detected antibodies to both HIV-1 and HIV-2 viruses. They met the minimum standards (sensitivity > 99 %, specificity > 95 %) as recommended by World Health Organization (WHO).[8] The diagnosis of HIV infection was confirmed as per the WHO strategy II: when two ELISA tests based on a different antigen preparation and/or different principle were positive.[9] In children below 18 months of age, HIV infection was confirmed only if they also fulfilled the WHO criteria for symptomatic HIV infection.[10]
Nutritional status was assessed by the Wellcome classification.[11] The severity of the disease at admission was classified into three stages, based on the Medical Research Council (MRC) guidelines[12]: stage I (early) = conscious, non-specific symptoms and no neurological signs; stage II (intermediate) = signs of meningeal irritation with slight or no clouding of sensorium, with or without minor neurological deficit (cranial nerve palsy or limb paresis); and, stage III (advanced) = severe clouding of sensorium, convulsions, focal neurological deficit, and/or involuntary movements. The children's levels of consciousness were assessed by the 15 point Glasgow coma scale (GCS).[13] Cranial CT scan has proved to be a reliable method for the diagnosis of TBM.[4],[14] The degree and significance of hydrocephalus were calculated by ventricular size index (VSI), which is the relation between bi-frontal diameter over the frontal horn diameter.[15] A VSI of 30-38% indicates mild hydrocephalus, 39-45% moderate hydrocephalus, and >45% severe hydrocephalus.[15]
Management of Cases
Standard anti-tuberculous therapy (ATT) [16]: isoniazid (5 mg/kg/day), rifampicin (10 mg/kg/day), pyrazinamide (25 mg/kg/day), and ethambutol (20 mg/kg/day); and other supportive measures (corticosteroids, anticonvulsants and mannitol) were started within 24- 48 hours of admission. Corticosteroids were administered during the first month of therapy and were given as intravenous dexamethasone (0.6 to 1.2 mg/kg/day in three divided doses) for the first seven days, followed by oral prednisolone (2 mg/kg/day in three divided doses) and then gradually tapered over a week. Acetazolamide (20 mg/kg/day orally in three divided doses) was started in children with hydrocephalus to prevent its progression.
During hospitalization, if the patient had deteriorating neurological symptoms and signs, Cranial CT scan was repeated to detect development of hydrocephalus or progression in its severity. A Neurosurgeon would do a ventriculoperitoneal (VP) shunt surgery when Cranial CT scan showed presence of moderate or severe hydrocephalus and it was associated with deteriorating neurological symptoms and signs.[5],[17] If the CSF protein concentration was more than 1 g/l, the hydrocephalus was initially treated by external ventricular drainage using a chamber as high CSF protein content can cause blockage of a VP shunt.[17] After the protein concentration fell below 1 g/l, shunt surgery was done. Family screening was done to detect source of infection. Liver function tests (LFTs) were done weekly, and whenever appropriate, to detect ATT-induced hepatotoxicity.
In children who were HIV seropositive (HIV-P), CSF examination was repeated to detect associated cryptococcal meningitis by direct microscopy on centrifuged deposits with wet and India ink preparation and culture on Sabouraud dextrose agar.
Variables and Outcome Assessment
Almost all the features that were analyzed (except shunt surgery being done, and development of ATT-induced hepatotoxicity) were documented at presentation. The outcome in each patient was assessed in terms of survival or death. Survival meant that the patient was discharged from the hospital: (i) having made a "complete recovery", or (ii) as a "disabled survivor" with neurological sequelae such as altered sensorium, cranial nerve palsy, extrapyramidal movements, focal neurological deficit, mental retardation, optic atrophy, and/or tone abnormalities. Survivors were discharged from hospital after their clinical condition had stabilized and they were able to take feeds either orally or via nasogastric tube.
Data Analysis
The data obtained was analyzed using the Statistical Package for the Social Sciences, version 11 for Windows (SPSS Ltd., Chicago, Illinois, USA). A bivariate analysis was initially performed by the Chi square (c2sub) test (with Yates' correction) to assess the relationship between the 35 variables, and the two outcomes (disabled survivor or death); with the HIV-P status. Wherever appropriate, the odds ratio (OR) was calculated and 95% confidence intervals (CI) was estimated around the OR. A P value <0.05 was considered significant. Two-tailed significance was obtained in all cases. Further, multivariate analysis was performed. Starting with a tentative logistic regression model with all the variables obtained in bivariate analysis, a step down variable selection procedure was used to derive a suitable model predicting the HIV-P status.
Results
Clinical, Laboratory and Radiological Features
Of a total 123 children with TBM who were enrolled, eight (6.5%) were detected to be HIV-P. There was no significant difference (P = 0.060) in mean age of children in the HIV-seronegative (HIV-N) and HIV-P groups: 34.5 months (S.D. 30.4 months, range 2.5 months to 11 years) vs 67.0 months (S.D. 47.3 months, range 5 months to 12 years). The M: F ratio was 1.1: 1 in the HIV-N group and 1.7: 1 in the HIV-P group. There was no significant difference (P = 0.100) in the mean symptom duration in the HIV-N and HIV-P groups: 22.1 days (S.D. 30.1 days, range 1 to 180 days) vs 42.0 days (S.D. 59.4 days, range 1 to 180 days).
On bivariate analysis, clinical and neurological features in both the groups were similar at presentation table2 and table3. Most children (87.8% vs 62.5%) were already in MRC TBM stage III at presentation. Laboratory and radiological features did not differ significantly between the two groups table4, except that more children in the HIV-P group had moderate-severe anemia i.e. Hb < 8 gm/dl (P = 0.001). Three children in the HIV-N group developed ATT-induced hepatotoxicity, which necessitated omitting isoniazid, rifampicin and pyrazinamide, and adding streptomycin till LFTs normalized. Cryptococcal meningitis was not detected in any of the HIV-P cases.
When multivariate analysis was applied table5, again the same variable, viz. Hb < 8 gm/dl, proved to be independently associated ( P = 0.008) to the HIV-P status.
Outcome
Twenty-two (19.1%) HIV-N cases and three (37.5%) HIV-P cases made a complete recovery. table6 shows that there was no significant difference in the two groups as regards disabled survivor ( P = 0.557) or death ( P = 0.779) outcomes. Also, duration of hospital stay in HIV-N and HIV-P children who survived was similar ( P =0.120): 20.3 days (S.D. 11.9 days, range 5 to 68 days) vs. 13.7 days (S.D. 5.1 days, range 7 to 22 days). The mean duration of hospital stay in HIV-N children who died was 21.2 days (range 3 to 90 days, S.D. 17.5 days). Only one patient, a 12-year-old boy, in the HIV-P group died after 10 day hospital stay.
Discussion
To our knowledge, the present study is the first one in India, and probably the second in the world, which has compared the characteristics and outcome of children having TBM with and without HIV co-infection. The present study has revealed that eight (6.5%) of 123 children admitted with TBM were HIV-infected. Except for the significantly increased ( P = 0.008) occurrence of moderate-severe anemia (Hb < 8 gm/dl) in the HIV-infected cases, all other features and in-hospital outcome were similar.
There is scant data on TBM in HIV-infected children. A prospective study by Topley et al[18] from Durban, South Africa has reported that 10 (25%) of 40 children admitted with TBM were HIV-infected. In their study on bivariate analysis, clinical features (including Hb levels) were similar in HIV-negative and HIV-infected children with TBM, but Cranial CT scans showed significantly (P = 0.01) more parenchymal enhancement.[18] Topley et al hospitalized the children for at least six months to assess their treatment outcome. In their experience, the short term response to treatment was reasonable in 6 out of the 10 HIV-infected children. Four of their survivors who had completed treatment left hospital being able to stand or walk. But long-term outcome at six to twelve months was poorer in the HIV-infected children, with 30% dying and the remaining children surviving with moderate or severe handicap.[18] In the present study, patients did not complete treatment in hospital. Since follow-up after discharge from hospital is generally poor in the present setting, the authors could not document the outcome after completion of ATT. The present finding that HIV co-infection does not appear to change the clinical manifestations or the in-hospital outcome of TBM has been reported in adult studies. [19],[20],[21]
Topley et al[18] have not employed multivariate analysis on their data. Multivariate analysis scores over bivariate analysis, as it has the advantage of determining the "independent" effect of each variable while controlling the influence of other variables. The present study confirmed on multivariate analysis, that Hb< 8 gm/dl was independently associated ( P = 0.008) to the HIV-infected status. The authors believe that the probable reason for this increased incidence of moderate-severe anemia is the advanced immunosuppression, and this finding will be useful to clinicians working in resource-constrained settings to suspect HIV co-infection in children admitted with TBM. Anemia is a common feature in HIV-infected children and its pathogenesis is often multi-factorial, including iron deficiency, the anemia of chronic disease, malaria and opportunistic infections.[22],[23] Anemia has also been shown to be a statistically significant predictor of progression to the acquired immunodeficiency syndrome and is independently associated with an increased risk of death in individuals with HIV.[22],[23]
There are no guidelines available about the use of corticosteroids in children with TBM and HIV co-infection. However, the administration of corticosteroids has been shown to significantly improve the survival and intellectual outcome of children with TBM.[24] Hence these were used even if HIV co-infection was present. Topley et al have also similarly used corticosteroids in their study.[19] A recent study from Vietnam has documented that adjunctive treatment with corticosteroids improves survival in patients over 14 years of age with TBM and HIV co-infection.[25]
The strengths of the present study are that: (i) it was prospective, (ii) all patients received uniform treatment, and (iii) 35 variables were analyzed by both bivariate and multivariate analysis to identify those associated with HIV co-infection. However, the present study has its limitations. First, there was a low incidence of HIV co-infection (8/123, 6.5%) in the TBM cases analyzed by the authors. Hence, it must be kept in mind that the sensitivity of the results of the present study may be less than optimal. However, the authors still believe that these results are important, as the present study was carried out when the HIV sero-prevalence in the general population was still < 2%. It is believed that the present study is timely and that its results offer pediatricians in our country an easy and inexpensive guideline to suspect HIV co-infection in children with TBM, viz the presence of moderate-severe anemia (Hb < 8 gm/dl). The only other similar pediatric study is by Topley et al , [18, wherein there was a much higher incidence of HIV co-infection (10/40, 25%) in the TBM cases analyzed. This study was carried out in South Africa in 1995-1997, after the HIV sero-prevalence in the general population had already reached panic levels (>10%).[18] Second, the diagnosis of TBM in the present study patients was based on clinical criteria and not on microbiological criteria. Although definitive diagnosis of TBM depends on the detection of the tubercle bacilli in the CSF, either by smear examination or by bacterial culture; smears are usually positive in fewer than 10% of cases of TBM, while culture for M. tuberculosis takes up to eight weeks and is also often negative.[17] We would like to point out three other pediatric studies in which diagnosis of TBM was based on the clinical criteria devised by Doerr et al.[7],[26],[27] Third, due to financial constraints we could not perform viral antigen detection tests. Hence we followed the Joint UNAIDS-WHO revised recommendations to diagnose HIV disease.[8] Strategy II was followed as it is applicable where the prevalence of infection is 30% or less, as in Mumbai.[1] Combinations of ELISA can provide results as reliable as the ELISA / Western blot combination and at a much lower cost.[8] Fourth, neurological outcome in children who survived was assessed purely on clinical examination at the time of discharge from the hospital. No formal intelligence quotient or audiological evaluations, were performed
Conclusion
The present study documents that there are no significant differences between HIV-infected and HIV-negative children admitted with TBM, except for a lower Hb level (< 8 gm/dl) among HIV-infected children. Also, HIV co-infection does not change the in-hospital response to therapy in childhood TBM.
Acknowledgements
We thank our Dean, Dr. M.E. Yeolekar for granting us permission to publish this study; and our Biostatistician, Mr. Kailas Gandewar, and Dr. D.P. Singh, Reader, Department of Research Methodology, Tata Institute of Social Sciences, Deonar, Mumbai for their help in the statistical analysis of the data.
References
1. Mumbai Districts AIDS Control Society. Reference Manual on HIV infection and AIDS for Training of Health Care Providers. Mumbai: Mumbai Districts AIDS Control Society; 1998.
2. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO Global Surveillance and Monitoring Project. JAMA 1999; 282: 677-686.
3. Anonymous. HIV-related TB cases increasing in Mumbai, India. AIDS Wkly 2000; 20: 19-20.
4. Starke JR. Tuberculosis of the central nervous system in children. Semin Pediatr Neurol 1999; 6: 318-331.
5. Newton RW. Tuberculous meningitis. Arch Dis Child 1994; 70: 364-366.
6. Mahadevan B, Mahadevan S, Serane VT. Prognostic factors in childhood tuberculous meningitis. J Trop Pediatr 2002; 48: 362-365.
7. Doerr CA, Starke JR, Ong LT. Clinical and public health aspects of tuberculous meningitis in children. J Pediatr 1995; 127: 27-33.
8. Joint United Nations Programme on HIV/AIDS (UNAIDS) - World Health Organization. Revised recommendations for the selection and use of HIV antibody tests. Wkly Epidemiol Rec 1997; 72: 81-88.
9. Dodd CL, Greenspan D, Katz MH, Westenhouse JL, Feigal DW, Greenspan JS. Oral candidiasis in HIV infection: pseudomembranous and erythematous candidiasis show similar rates of progression to AIDS. AIDS 1991; 5: 1339-1343.
10. World Health Organization. Guidelines for the Clinical Management of HIV Infection in Children. Geneva: WHO, 1993.
11. Wellcome Trust Working Party. Classification of infantile malnutrition. Lancet 1970; ii: 302-303.
12. Medical Research Council. Streptomycin in Tuberculosis Trials Committee. Streptomycin treatment of tuberculous meningitis. Lancet 1948; i: 582-596.
13. Teasdale GM, Jennett B. Assessment and prognosis of coma after head injury. Acta Neurochir (Wien) 1976; 34: 45-55.
14. Ozates M, Kemaloglu S, Gurkan F, Ozkan U, Hosoglu S, Simsek MM. CT of the brain in tuberculous meningitis. A review of 289 patients. Acta Radiol 2000; 41: 13-17.
15. TerBrugge KG, Rao KC, Lee SH. Hydrocephalus and atrophy. In Lee SH, Rao KC, eds. Cranial Computed Tomography and MRI. New York; McGraw Hill, 1987; 231-262.
16. Indian Academy of Pediatrics. Treatment of childhood tuberculosis: Consensus Statement of IAP Working Group. Indian Pediatr 1997; 34: 1093-1096.
17. Garg RK. Tuberculosis of the central nervous system. Postgrad Med J 1999; 75: 133-140.
18. Topley JM, Bamber S, Coovadia HM, Corr PD. Tuberculous meningitis and co-infection with HIV. Ann Trop Paediatr 1998; 18: 261-266.
19. Berenguer J, Moreno S, Laguna F et al. Tuberculous meningitis in patients infected with the human immunodeficiency virus. N Engl J Med 1992; 326: 668-672.
20. Yechoor VK, Shandera WX, Rodriguez P, Cate TR. Tuberculous meningitis among adults with and without HIV infection. Experience in an urban public hospital [Erratum in: Arch Intern Med 1997; 157: 1022]. Arch Intern Med 1996; 156:1710-1716.
21. Bossi P, Reverdy O, Caumes E et al. Tuberculous meningitis: clinical, biological and x-ray computed tomographic comparison between patients with or without HIV infection (
Abstract). Presse Med 1997; 26: 844-847.
22. Ellaurie M, Burns ER, Rubinstein A. Hematologic manifestations in pediatric HIV infection: severe anemia as a prognostic factor. Am J Pediatr Hematol Oncol 1990; 12: 449-453.
23. Belperio PS, Rhew DC. Prevalence and outcomes of anemia in individuals with human immunodeficiency virus: a systematic review of the literature. Am J Med 2004; 116: 27S-43S.
24. Karak B, Garg RK. Corticosteroids in tuberculous meningitis. Indian Pediatr 1998; 35: 193-194.
25. Thwaites GE, Nguyen DB, Nguyen HD et al. Dexamethasone for the treatment of tuberculous meningitis in adolescents and adults. N Engl J Med 2004; 351: 1741-1751.
26. Yarami A, Gurkan F, Elevli M et al. Central nervous system tuberculosis in children: a review of 214 cases. Pediatrics 1998; 102: e49.
27. Paganini H, Gonzalez F, Santander C, Casimir L, Berberian G, Rosanova MT. Tuberculous meningitis in children: clinical features and outcome in 40 cases. Scand J Infect Dis 2000; 32: 41-45.(Karande Sunil, Gupta Vish)
2 Department of Radiology, Lokmanya Tilak Municipal Medical College and General Hospital, Sion, Mumbai, India
3 Department of Microbiology, Lokmanya Tilak Municipal Medical College and General Hospital, Sion, Mumbai, India
Abstract
Objective: To identify factors associated with HIV-infected status in children admitted with tuberculous meningitis (TBM), and to find out whether HIV co-infection affects in-hospital outcome. Methods: This prospective hospital-based study was conducted from May 2000 to August 2003. All consecutive children, aged 1 month to 12 years of age, admitted with a diagnosis of TBM were enrolled. Relationship between 35 features viz., two demographic factors, nine clinical features, 13 neurological features, five laboratory (including cerebrospinal fluid) parameters, six radiological (including computed tomography scan brain) features, and the two outcomes (disabled survivor or death); with HIV-infected status was assessed. Results: Of a total 123 TBM cases enrolled, eight (6.5%) were HIV-infected. There was no significant difference between the two groups, except that more children in the HIV-infected group had Hb< 8 gm/dl: both on bivariate analysis, (OR, 12.0; 95% CI, 2.6-55.9; P = 0.001) and on multivariate analysis (OR, 12.30; 95% CI, 1.9-79.6; P = 0.008). Outcome was similar in both the groups. Conclusion: Only presence of Hb< 8 gm/dl was associated with HIV-infected status. HIV co-infection did not affect the outcome.
Keywords: HIV Infections; Multivariate Analysis; Tomography, X-Ray Computed; Treatment outcome; Tuberculosis; Meningeal
Since the early 1990's, human immunodeficiency virus (HIV) infection is assuming alarming proportions in Mumbai.[1] This HIV epidemic has been accompanied by a corresponding increase in the incidence of tuberculosis (TB) and HIV-related TB cases.[2],[3] Between 1-2% of children with untreated tuberculous infection develop tuberculous meningitis (TBM), a serious illness which has a high rate of neuromorbidity and mortality.[4],[5] Co-infection of TBM and HIV is likely to further complicate this problem.[6]
The objective of the present study was to: (i) find out the incidence of HIV co-infection in children admitted with TBM to our hospital, (ii) identify factors which are associated with HIV-infected status and (iii) find out whether HIV co-infection affects the in-hospital outcome.
Materials and methods
Patient Enrolment
All consecutive children, aged 1 month to 12 years of age, who were admitted to our hospital with a diagnosis of TBM, were enrolled prospectively. The study was conducted from May 2000 to August 2003. The diagnosis of TBM was based on a standard clinical case definition table1 devised by Doerr et al.[7]
At the time of admission, a complete physical examination was performed. Cerebrospinal fluid (CSF) examination and cranial computed tomography (CT) scan were done soon after admission in every child. A standardized data-entry form was used to document demographic data, clinical symptoms and signs, laboratory findings, Mantoux test result, chest radiograph and ultrasound abdomen findings, CSF and Cranial CT scan findings of each patient were conducted at presentation. Histopathological examinations of lymph nodes were carried out wherever appropriate.
Each patient was screened for HIV infection using the enzyme-linked immunosorbent assay (ELISA) test. Pre and post- test counseling was offered to the parents. The ELISA kits were used in strict compliance with the manufacturer's instructions. The kits detected antibodies to both HIV-1 and HIV-2 viruses. They met the minimum standards (sensitivity > 99 %, specificity > 95 %) as recommended by World Health Organization (WHO).[8] The diagnosis of HIV infection was confirmed as per the WHO strategy II: when two ELISA tests based on a different antigen preparation and/or different principle were positive.[9] In children below 18 months of age, HIV infection was confirmed only if they also fulfilled the WHO criteria for symptomatic HIV infection.[10]
Nutritional status was assessed by the Wellcome classification.[11] The severity of the disease at admission was classified into three stages, based on the Medical Research Council (MRC) guidelines[12]: stage I (early) = conscious, non-specific symptoms and no neurological signs; stage II (intermediate) = signs of meningeal irritation with slight or no clouding of sensorium, with or without minor neurological deficit (cranial nerve palsy or limb paresis); and, stage III (advanced) = severe clouding of sensorium, convulsions, focal neurological deficit, and/or involuntary movements. The children's levels of consciousness were assessed by the 15 point Glasgow coma scale (GCS).[13] Cranial CT scan has proved to be a reliable method for the diagnosis of TBM.[4],[14] The degree and significance of hydrocephalus were calculated by ventricular size index (VSI), which is the relation between bi-frontal diameter over the frontal horn diameter.[15] A VSI of 30-38% indicates mild hydrocephalus, 39-45% moderate hydrocephalus, and >45% severe hydrocephalus.[15]
Management of Cases
Standard anti-tuberculous therapy (ATT) [16]: isoniazid (5 mg/kg/day), rifampicin (10 mg/kg/day), pyrazinamide (25 mg/kg/day), and ethambutol (20 mg/kg/day); and other supportive measures (corticosteroids, anticonvulsants and mannitol) were started within 24- 48 hours of admission. Corticosteroids were administered during the first month of therapy and were given as intravenous dexamethasone (0.6 to 1.2 mg/kg/day in three divided doses) for the first seven days, followed by oral prednisolone (2 mg/kg/day in three divided doses) and then gradually tapered over a week. Acetazolamide (20 mg/kg/day orally in three divided doses) was started in children with hydrocephalus to prevent its progression.
During hospitalization, if the patient had deteriorating neurological symptoms and signs, Cranial CT scan was repeated to detect development of hydrocephalus or progression in its severity. A Neurosurgeon would do a ventriculoperitoneal (VP) shunt surgery when Cranial CT scan showed presence of moderate or severe hydrocephalus and it was associated with deteriorating neurological symptoms and signs.[5],[17] If the CSF protein concentration was more than 1 g/l, the hydrocephalus was initially treated by external ventricular drainage using a chamber as high CSF protein content can cause blockage of a VP shunt.[17] After the protein concentration fell below 1 g/l, shunt surgery was done. Family screening was done to detect source of infection. Liver function tests (LFTs) were done weekly, and whenever appropriate, to detect ATT-induced hepatotoxicity.
In children who were HIV seropositive (HIV-P), CSF examination was repeated to detect associated cryptococcal meningitis by direct microscopy on centrifuged deposits with wet and India ink preparation and culture on Sabouraud dextrose agar.
Variables and Outcome Assessment
Almost all the features that were analyzed (except shunt surgery being done, and development of ATT-induced hepatotoxicity) were documented at presentation. The outcome in each patient was assessed in terms of survival or death. Survival meant that the patient was discharged from the hospital: (i) having made a "complete recovery", or (ii) as a "disabled survivor" with neurological sequelae such as altered sensorium, cranial nerve palsy, extrapyramidal movements, focal neurological deficit, mental retardation, optic atrophy, and/or tone abnormalities. Survivors were discharged from hospital after their clinical condition had stabilized and they were able to take feeds either orally or via nasogastric tube.
Data Analysis
The data obtained was analyzed using the Statistical Package for the Social Sciences, version 11 for Windows (SPSS Ltd., Chicago, Illinois, USA). A bivariate analysis was initially performed by the Chi square (c2sub) test (with Yates' correction) to assess the relationship between the 35 variables, and the two outcomes (disabled survivor or death); with the HIV-P status. Wherever appropriate, the odds ratio (OR) was calculated and 95% confidence intervals (CI) was estimated around the OR. A P value <0.05 was considered significant. Two-tailed significance was obtained in all cases. Further, multivariate analysis was performed. Starting with a tentative logistic regression model with all the variables obtained in bivariate analysis, a step down variable selection procedure was used to derive a suitable model predicting the HIV-P status.
Results
Clinical, Laboratory and Radiological Features
Of a total 123 children with TBM who were enrolled, eight (6.5%) were detected to be HIV-P. There was no significant difference (P = 0.060) in mean age of children in the HIV-seronegative (HIV-N) and HIV-P groups: 34.5 months (S.D. 30.4 months, range 2.5 months to 11 years) vs 67.0 months (S.D. 47.3 months, range 5 months to 12 years). The M: F ratio was 1.1: 1 in the HIV-N group and 1.7: 1 in the HIV-P group. There was no significant difference (P = 0.100) in the mean symptom duration in the HIV-N and HIV-P groups: 22.1 days (S.D. 30.1 days, range 1 to 180 days) vs 42.0 days (S.D. 59.4 days, range 1 to 180 days).
On bivariate analysis, clinical and neurological features in both the groups were similar at presentation table2 and table3. Most children (87.8% vs 62.5%) were already in MRC TBM stage III at presentation. Laboratory and radiological features did not differ significantly between the two groups table4, except that more children in the HIV-P group had moderate-severe anemia i.e. Hb < 8 gm/dl (P = 0.001). Three children in the HIV-N group developed ATT-induced hepatotoxicity, which necessitated omitting isoniazid, rifampicin and pyrazinamide, and adding streptomycin till LFTs normalized. Cryptococcal meningitis was not detected in any of the HIV-P cases.
When multivariate analysis was applied table5, again the same variable, viz. Hb < 8 gm/dl, proved to be independently associated ( P = 0.008) to the HIV-P status.
Outcome
Twenty-two (19.1%) HIV-N cases and three (37.5%) HIV-P cases made a complete recovery. table6 shows that there was no significant difference in the two groups as regards disabled survivor ( P = 0.557) or death ( P = 0.779) outcomes. Also, duration of hospital stay in HIV-N and HIV-P children who survived was similar ( P =0.120): 20.3 days (S.D. 11.9 days, range 5 to 68 days) vs. 13.7 days (S.D. 5.1 days, range 7 to 22 days). The mean duration of hospital stay in HIV-N children who died was 21.2 days (range 3 to 90 days, S.D. 17.5 days). Only one patient, a 12-year-old boy, in the HIV-P group died after 10 day hospital stay.
Discussion
To our knowledge, the present study is the first one in India, and probably the second in the world, which has compared the characteristics and outcome of children having TBM with and without HIV co-infection. The present study has revealed that eight (6.5%) of 123 children admitted with TBM were HIV-infected. Except for the significantly increased ( P = 0.008) occurrence of moderate-severe anemia (Hb < 8 gm/dl) in the HIV-infected cases, all other features and in-hospital outcome were similar.
There is scant data on TBM in HIV-infected children. A prospective study by Topley et al[18] from Durban, South Africa has reported that 10 (25%) of 40 children admitted with TBM were HIV-infected. In their study on bivariate analysis, clinical features (including Hb levels) were similar in HIV-negative and HIV-infected children with TBM, but Cranial CT scans showed significantly (P = 0.01) more parenchymal enhancement.[18] Topley et al hospitalized the children for at least six months to assess their treatment outcome. In their experience, the short term response to treatment was reasonable in 6 out of the 10 HIV-infected children. Four of their survivors who had completed treatment left hospital being able to stand or walk. But long-term outcome at six to twelve months was poorer in the HIV-infected children, with 30% dying and the remaining children surviving with moderate or severe handicap.[18] In the present study, patients did not complete treatment in hospital. Since follow-up after discharge from hospital is generally poor in the present setting, the authors could not document the outcome after completion of ATT. The present finding that HIV co-infection does not appear to change the clinical manifestations or the in-hospital outcome of TBM has been reported in adult studies. [19],[20],[21]
Topley et al[18] have not employed multivariate analysis on their data. Multivariate analysis scores over bivariate analysis, as it has the advantage of determining the "independent" effect of each variable while controlling the influence of other variables. The present study confirmed on multivariate analysis, that Hb< 8 gm/dl was independently associated ( P = 0.008) to the HIV-infected status. The authors believe that the probable reason for this increased incidence of moderate-severe anemia is the advanced immunosuppression, and this finding will be useful to clinicians working in resource-constrained settings to suspect HIV co-infection in children admitted with TBM. Anemia is a common feature in HIV-infected children and its pathogenesis is often multi-factorial, including iron deficiency, the anemia of chronic disease, malaria and opportunistic infections.[22],[23] Anemia has also been shown to be a statistically significant predictor of progression to the acquired immunodeficiency syndrome and is independently associated with an increased risk of death in individuals with HIV.[22],[23]
There are no guidelines available about the use of corticosteroids in children with TBM and HIV co-infection. However, the administration of corticosteroids has been shown to significantly improve the survival and intellectual outcome of children with TBM.[24] Hence these were used even if HIV co-infection was present. Topley et al have also similarly used corticosteroids in their study.[19] A recent study from Vietnam has documented that adjunctive treatment with corticosteroids improves survival in patients over 14 years of age with TBM and HIV co-infection.[25]
The strengths of the present study are that: (i) it was prospective, (ii) all patients received uniform treatment, and (iii) 35 variables were analyzed by both bivariate and multivariate analysis to identify those associated with HIV co-infection. However, the present study has its limitations. First, there was a low incidence of HIV co-infection (8/123, 6.5%) in the TBM cases analyzed by the authors. Hence, it must be kept in mind that the sensitivity of the results of the present study may be less than optimal. However, the authors still believe that these results are important, as the present study was carried out when the HIV sero-prevalence in the general population was still < 2%. It is believed that the present study is timely and that its results offer pediatricians in our country an easy and inexpensive guideline to suspect HIV co-infection in children with TBM, viz the presence of moderate-severe anemia (Hb < 8 gm/dl). The only other similar pediatric study is by Topley et al , [18, wherein there was a much higher incidence of HIV co-infection (10/40, 25%) in the TBM cases analyzed. This study was carried out in South Africa in 1995-1997, after the HIV sero-prevalence in the general population had already reached panic levels (>10%).[18] Second, the diagnosis of TBM in the present study patients was based on clinical criteria and not on microbiological criteria. Although definitive diagnosis of TBM depends on the detection of the tubercle bacilli in the CSF, either by smear examination or by bacterial culture; smears are usually positive in fewer than 10% of cases of TBM, while culture for M. tuberculosis takes up to eight weeks and is also often negative.[17] We would like to point out three other pediatric studies in which diagnosis of TBM was based on the clinical criteria devised by Doerr et al.[7],[26],[27] Third, due to financial constraints we could not perform viral antigen detection tests. Hence we followed the Joint UNAIDS-WHO revised recommendations to diagnose HIV disease.[8] Strategy II was followed as it is applicable where the prevalence of infection is 30% or less, as in Mumbai.[1] Combinations of ELISA can provide results as reliable as the ELISA / Western blot combination and at a much lower cost.[8] Fourth, neurological outcome in children who survived was assessed purely on clinical examination at the time of discharge from the hospital. No formal intelligence quotient or audiological evaluations, were performed
Conclusion
The present study documents that there are no significant differences between HIV-infected and HIV-negative children admitted with TBM, except for a lower Hb level (< 8 gm/dl) among HIV-infected children. Also, HIV co-infection does not change the in-hospital response to therapy in childhood TBM.
Acknowledgements
We thank our Dean, Dr. M.E. Yeolekar for granting us permission to publish this study; and our Biostatistician, Mr. Kailas Gandewar, and Dr. D.P. Singh, Reader, Department of Research Methodology, Tata Institute of Social Sciences, Deonar, Mumbai for their help in the statistical analysis of the data.
References
1. Mumbai Districts AIDS Control Society. Reference Manual on HIV infection and AIDS for Training of Health Care Providers. Mumbai: Mumbai Districts AIDS Control Society; 1998.
2. Dye C, Scheele S, Dolin P, Pathania V, Raviglione MC. Consensus statement. Global burden of tuberculosis: estimated incidence, prevalence, and mortality by country. WHO Global Surveillance and Monitoring Project. JAMA 1999; 282: 677-686.
3. Anonymous. HIV-related TB cases increasing in Mumbai, India. AIDS Wkly 2000; 20: 19-20.
4. Starke JR. Tuberculosis of the central nervous system in children. Semin Pediatr Neurol 1999; 6: 318-331.
5. Newton RW. Tuberculous meningitis. Arch Dis Child 1994; 70: 364-366.
6. Mahadevan B, Mahadevan S, Serane VT. Prognostic factors in childhood tuberculous meningitis. J Trop Pediatr 2002; 48: 362-365.
7. Doerr CA, Starke JR, Ong LT. Clinical and public health aspects of tuberculous meningitis in children. J Pediatr 1995; 127: 27-33.
8. Joint United Nations Programme on HIV/AIDS (UNAIDS) - World Health Organization. Revised recommendations for the selection and use of HIV antibody tests. Wkly Epidemiol Rec 1997; 72: 81-88.
9. Dodd CL, Greenspan D, Katz MH, Westenhouse JL, Feigal DW, Greenspan JS. Oral candidiasis in HIV infection: pseudomembranous and erythematous candidiasis show similar rates of progression to AIDS. AIDS 1991; 5: 1339-1343.
10. World Health Organization. Guidelines for the Clinical Management of HIV Infection in Children. Geneva: WHO, 1993.
11. Wellcome Trust Working Party. Classification of infantile malnutrition. Lancet 1970; ii: 302-303.
12. Medical Research Council. Streptomycin in Tuberculosis Trials Committee. Streptomycin treatment of tuberculous meningitis. Lancet 1948; i: 582-596.
13. Teasdale GM, Jennett B. Assessment and prognosis of coma after head injury. Acta Neurochir (Wien) 1976; 34: 45-55.
14. Ozates M, Kemaloglu S, Gurkan F, Ozkan U, Hosoglu S, Simsek MM. CT of the brain in tuberculous meningitis. A review of 289 patients. Acta Radiol 2000; 41: 13-17.
15. TerBrugge KG, Rao KC, Lee SH. Hydrocephalus and atrophy. In Lee SH, Rao KC, eds. Cranial Computed Tomography and MRI. New York; McGraw Hill, 1987; 231-262.
16. Indian Academy of Pediatrics. Treatment of childhood tuberculosis: Consensus Statement of IAP Working Group. Indian Pediatr 1997; 34: 1093-1096.
17. Garg RK. Tuberculosis of the central nervous system. Postgrad Med J 1999; 75: 133-140.
18. Topley JM, Bamber S, Coovadia HM, Corr PD. Tuberculous meningitis and co-infection with HIV. Ann Trop Paediatr 1998; 18: 261-266.
19. Berenguer J, Moreno S, Laguna F et al. Tuberculous meningitis in patients infected with the human immunodeficiency virus. N Engl J Med 1992; 326: 668-672.
20. Yechoor VK, Shandera WX, Rodriguez P, Cate TR. Tuberculous meningitis among adults with and without HIV infection. Experience in an urban public hospital [Erratum in: Arch Intern Med 1997; 157: 1022]. Arch Intern Med 1996; 156:1710-1716.
21. Bossi P, Reverdy O, Caumes E et al. Tuberculous meningitis: clinical, biological and x-ray computed tomographic comparison between patients with or without HIV infection (
Abstract). Presse Med 1997; 26: 844-847.
22. Ellaurie M, Burns ER, Rubinstein A. Hematologic manifestations in pediatric HIV infection: severe anemia as a prognostic factor. Am J Pediatr Hematol Oncol 1990; 12: 449-453.
23. Belperio PS, Rhew DC. Prevalence and outcomes of anemia in individuals with human immunodeficiency virus: a systematic review of the literature. Am J Med 2004; 116: 27S-43S.
24. Karak B, Garg RK. Corticosteroids in tuberculous meningitis. Indian Pediatr 1998; 35: 193-194.
25. Thwaites GE, Nguyen DB, Nguyen HD et al. Dexamethasone for the treatment of tuberculous meningitis in adolescents and adults. N Engl J Med 2004; 351: 1741-1751.
26. Yarami A, Gurkan F, Elevli M et al. Central nervous system tuberculosis in children: a review of 214 cases. Pediatrics 1998; 102: e49.
27. Paganini H, Gonzalez F, Santander C, Casimir L, Berberian G, Rosanova MT. Tuberculous meningitis in children: clinical features and outcome in 40 cases. Scand J Infect Dis 2000; 32: 41-45.(Karande Sunil, Gupta Vish)