Meningitis is the clinical presentation of Mycobacterium tuberculosis infection with the highest mortality rate. Although HIV infection itself is considered a risk factor for death in patients with tuberculous meningitis (TM), prognostic factors in this population remain unclear [1,2].
Infection by multidrug-resistant strains poses a major challenge for the clinician as it is an additional predictor of mortality. Although the association between infection by these strains and HIV is well known, there is a paucity of data on its impact in the clinical outcome of TM [3,4].
The Infectious Diseases F.J. Muñiz Hospital is a 300-bed institution located in Buenos Aires city, Argentina, which successfully controlled a multidrug-resistant tuberculosis outbreak in the past decade [5,6]. With the aim of characterizing prognostic factors and showing the impact of multidrug-resistant strains on the mortality of HIV-infected patients with culture-confirmed TM, we undertook a retrospective study considering clinical and microbiological aspects.
We retrospectively analysed data from the clinical charts of HIV-infected patients with a confirmed diagnosis of TM by the isolation of M. tuberculosis from cerebrospinal fluid (CSF) attending our institution from 1996 to 2004. During this period, CSF samples were cultured in Lowenstein–Jensen medium or BACTEC. The susceptibility profile was analysed by either the proportion method or the BACTEC radiometric method according to standardized procedures [7–10].
The demographic characteristics, clinical presentation and outcome, neuroimaging findings, CSF alterations, and drug resistance patterns were recorded. Statistical analyses were performed using Statistix software 7.0. HIV infection was documented by HIV-1 antibody enzyme-linked immunosorbent assay and Western blot analysis. A multidrug-resistant strain was defined as M. tuberculosis that was resistant to, at least, isoniazid and rifampicin. The neurological status of patients was described according to the British Medical Research Council (BMRC) system for grading meningitis modified as follows: grade I, patient was conscious and without neurological signs; grade II, patients was confused or had neurological signs (meningeal or focal signs); and grade III, patient was stuporose or comatose. Univariate logistic regression analyses were performed to identify the independent variables that were associated with death during hospitalization. The list of potential variables included extraneural tuberculosis, previous or ongoing treatment for tuberculosis, the BMRC grade, the co-existence of another AIDS-defining illness, infection by a multidrug-resistant strain, the CD4 T-cell count, absence of CSF pleocytosis, CSF glucose level less than 30 mg/dl, a delay in the start of treatment greater than 24 h after hospitalization, and neuroimaging alterations. Only those variables with a significance level of P < 0.05 were considered to be candidate variables for the multivariate logistic regression analysis.
Of a total of 141 CSF samples with positive culture for M. tuberculosis, 101 corresponded to HIV-infected patients. The median age was 33 years, and 68% were men. Thirty-one (30%) had had previous incomplete treatment for pulmonary or extrapulmonary tuberculosis, and 13 (12.8%) were taking antituberculous drugs at the moment of TM diagnosis. The median of CD4 T-cell count was 53 cells/μl.
The BMRC clinical assessment rating for meningitis was grade I for 19 patients (18.8%), grade II for 74 patients (73.3%), and grade III for eight patients (7.9%). The onset of treatment was within the 24 h of admission in 79 patients (78.2%). In 56 patients (54.9%) M. tuberculosis was also isolated from extraneural samples (sputum, bronchoalveolar lavage, pleural effusion, suppurative lesions). Six patients had positive documented blood cultures. Twelve patients (11.8%) had a simultaneous diagnosis of another AIDS-defining illness, and 77.8% had neuroimaging alterations. The global mortality during hospitalization was 63.3% (64 patients).
Forty-nine isolates (48.5%) were not resistant to any drug and 42 (41.6%) were multidrug-resistant strains. Three patients had isolated resistance to isoniazid, four to streptomycin, and three to isoniazid plus pyrazinamide or ethambutol. Infection with multidrug-resistant strains was more frequent in patients with previous incomplete or ongoing non-compliant antituberculous treatment (odds ratio 3.09; 95% confidence interval 1.34–7.08; P < 0.05). The proportion of infection by these strains was similar in patients with less than 50 CD4 T cells/μl and in those with more than 50 cells/μl (P > 0.05).
According to the primary criteria in univariate analysis (Table 1), the following variables were associated with death during hospitalization: BMRC grade II or III at admission, a CD4 T-cell count less than 50 cells/μl and infection caused by multidrug-resistant strains. The three variables remained associated when included in a multivariate logistic regression model.
The high frequency of multidrug-resistant strains in our population allowed us to have an adequate number of patients to show statistically the impact of the infection with these strains on the mortality rate of the disease. The higher mortality rate in patients with multidrug resistant TM may be attributable to the failure of first-line drugs to control the infection and the delay before the susceptibility pattern is known and adequate therapy instituted. When instituted, the treatment of multidrug-resistant TM is less effective because second-line drugs have, with the exception of ethionamide, low CSF penetration [3,4]. In this clinical setting, investigation for new drugs with good CSF penetration is urgently needed.
HIV-infected patients with TM constitute an heterogeneous group. As shown in the multivariate analysis, a CD4 T-cell count less than 50 cells/μl is associated with a higher mortality rate. This fact is independent of the infection by multidrug-resistant strains and may suggest that HIV-infected patients included in series of TM should be stratified according to the CD4 T-cell count when analysing the mortality rate.
As a result of the retrospective design of our study, the evaluation of drug combinations for the treatment of multidrug-resistant strains and reporting the frequency of long-term disability in the surviving patients after hospitalization were not possible.
The identification of risk factors related to infection by multidrug-resistant strains was beyond the scope of this study. Although epidemiological data regarding a family history of tuberculosis and confinement to penal institutions could not be evaluated because of logistical constraints, our institution assists a marginalized population that can be reflected in the high frequency of non-compliance with treatment, as directly observed therapy is an uncommon policy in our country .
In conclusion, interventions that reduce the transmission of multidrug-resistant strains and improve adherence to first and second-line antituberculous regimens are required to reduce the burden of this disease among HIV-positive patients in Argentina.
1. Thwaites G, Bang N, Dung N, Quy H, Oanh D, Thoa N, et al
. The influence of HIV infection on clinical presentation, response to treatment and outcome in adults with tuberculous meningitis. J Infect Dis 2005; 192:2134–2141.
2. Thwaites G, Nguyen D, Hoang T, Do T, Nguyen T, Nguyen Q, et al
. Dexamethasone for the treatment of tuberculous meningitis in adolescents and adults. N Engl J Med 2004; 351:1741–1751.
3. Patel V, Padayatchi N, Bhigjee A, Allen J, Bhagwan B, Moodley A, Mthiyane T. Multidrug-resistant tuberculous meningitis in KwaZulu-Natal, South Africa. Clin Infect Dis 2004; 38:851–856.
4. Thwaites G, Lan N, Dung N, Quy H, Oanh D, Thoa N, et al
. Effect of antituberculosis drug resistance on response to treatment and outcome in adults with tuberculous meningitis. J Infect Dis 2005; 192:79–88.
5. Ritacco V, Di Lonardo M, Reniero A, Ambroggi M, Barrera L, Dambrosi A, et al
. Nosocomial spread of human immunodeficiency virus-related multidrug-resistant tuberculosis in Buenos Aires. J Infect Dis 1997; 176:748–751.
6. Waisman J, Palmero D, Guemes-Gurtubay J, Videla J, Moretti B, Cantero M, et al
. Evaluation of the control measures adopted against an epidemic of AIDS-related multidrug-resistant tuberculosis in a latin-American hospital [in Spanish]. Enferm Infecc Microbiol Clin 2006; 24:71–76.
7. Cetrángolo A, Marchesini L, Isola N, Di Lonardo M, Kantor I, Ambroggi M, et al
. Microbiological aspects of tuberculous meningitis [in Spanish]. Boletín de la OPS 1974; 3:215–221.
8. Canetti G, Froman S, Grosset J, Hauduroy P, Langerova M, Mahler H, et al
. Mycobacteria: laboratory methods for testing drug sensitivity and resistance. Bull WHO 1963; 29:565–578.
9. Laszlo A, Rahman M, Raviglione M, Bustreo F, and the WHO/IUATLD Network of Supranational Reference Laboratorios. Quality assurance programme for drug susceptibility testing of Mycobacterium tuberculosis
in the WHO/IUATLD Supranational Laboratory Network: first round of proficiency testing. Int J Tuberc Lung Dis 1997; 1:231–238.
10. Laszlo A, Gill P, Handzel V, Hodgkin M, Helbecque D. Conventional and radiometric drug susceptibility testing of Mycobacterium tuberculosis
complex. J Clin Microbiol 1983; 18:1335–1339.
11. Tuberculosis in Argentina. Evaluation of direct- smear positive pulmonary TB treatment. Cohort 2003
[in Spanish]. Argentina National TB Control Program, Technical Document PRO/TB; 2005.