Practical Therapeutics

Drugs 39 (2): 224-233, 1990 0012-6667/90/0002-0224/$05.00/0 © ADiS Press Limited All rights reserved. DRUG03297a

Management of Tuberculosis Meningitis Mack R. Holdiness Lakeside Hospital, Metairie, Louisiana, USA

Contents

Summary ................................................................................................................. .. ................ 224 I. CSF Examination and Methods of Disease Detection ..................................................... 225 2. Antituberculosis Drug Penetration into the CSF .............................................................. 226 3. Treatment Regimens ............................................................................................................ 226 4. Therapeutic Use of BCG Vaccine ...................................................................................... 229 5. Use of Steroids in TB Meningitis ....................................................................................... 230 6. Surgical Intervention ............................................................................................................ 230 7. Conclusions and Suggestions for Patient Follow-Up ......... ,.............................................. 231

Summary

This article examines the detection, assessment, and therapeutic modalities available for tuberculosis meningitis. Without appropriate treatment this disease is fatal within 2 months of development, and mortality is closely associated with the stage of disease upon initiation of chemotherapy. Initial lumbar puncture reveals smear-positive acid-fast bacilli in up to 40% in most series; however, repeat examinations increase the yield, via direct smear, to as high as 87%. Analysis of cerebrospinal fluid is described, along with advanced techniques for early detection of this infection. Eight antituberculosis agents have known penetration into the cerebrospinal fluid. The most important prognostic factor is the neurological stage at which the individual presents at the initiation of therapy. Various chemotherapeutic approaches are available but it appears the use of rifampicin (rifampin) with isoniazid-containing regimens gives the best results. Regardless of the therapy undertaken, a significant number of individuals are left with some degree of neurological deficit. The roles of bacillus of Calmette and Guerin (BCG) vaccine, steroids, and neurosurgery in the treatment of this disease are also discussed.

In developing countries as well as in the United States, Mycobacterium tuberculosis is still an important cause of meningitis. In 1985 approximately 5% of 4000 extrapulmonary cases of tuberculosis (TB) reported in the United States were due to TB

meningitis (Ogawa et al. 1987). In addition, with the rising trend of disseminated TB documented in association with the acquired immunodeficiency syndrome (AIDS), there exists ample potential for central nervous system (CNS) involvement (Cen-

Management of Tuberculosis Meningitis

ters for Disease Control 1987; Chaisson & Slutkin 1989). Meningitis requires early detection and institution of appropriate therapy; according to Fitzsimons (1963) early diagnosis and treatment can potentially lead to 100% recovery without any sequelae. However, untreated disease is fatal within approximately 3 to 8 weeks of presentation and carries a high risk of severe neurological sequelae if therapy is delayed (Molavi & LeFrock 1985; Ogawa et al. 1987). Invariably, a focus of infection has occurred elsewhere (usually pulmonary in origin) and bacteria reach the CNS by haematogenous spread or, in some cases, local extension. Retrospective studies confirm that at least 75% of individuals have such an infection at least 12 months before admission for meningitis (Parsons 1979). CNS involvement arises as a complication from foci in the brain or meninges. These foci rupture, leading to clinical meningitis, the physical symptoms of which are described elsewhere in detail (Molavi & LeFrock 1985; Ogawa et al. 1987; Parsons 1979; Phuapradit & Vejjajiva 1987). Limited clinical studies for the treatment of TB meningitis have been published and there is currently no unanimity concerning chemotherapeutic regimens or optimal duration of treatment. It is the intention of this article to examine the current development of various modalities for treatment including the use of bacillus of Calmette and Guerin (BCG) vaccine, steroids and surgical intervention. Cerebrospinal fluid (CSF) findings will also be discussed, along with recent methods for rapid detection of this relentless and often fatal disease.

1. CSF Examination and Methods of Disease Detection In a recent review of the literature, Molavi and LeFrock (1985) observed that mycobacteria were identified via CSF smear in 10 to 40% of patients for most series and by culture in 45 to 90% of cases. Kennedy and Fallon (1979) found initial CSF specimens positive for mycobacteria in 37% on direct smear and 52% on culture, yet repeat lumbar puncture examinations increased the yield to 87% via

225

smear; nevertheless, negative cultures have been documented in autopsy proven cases. Elevated opening pressures and slowly rising CSF protein levels associated with hypoglycorrhacia and a gradual shift in the white blood cell count from neutrophils to lymphocytes over a period of days to weeks are consistent with the progression of disease (Molavi & LeFrock 1985). Pleocytosis is usually less than 500 cells/mm 3 but rare instances of counts as high as 1200 cells/mm 3 exist (Karandanis & Shulman 1976). Although polymorphonuclear predominance was demonstrated in I study in 58% of individuals observed in the initial phase of the disease (Clark et al. 1986), Ogawa et al. (1987) reported lymphocyte predominance in 68% of their patients, with a mean elevated opening pressure of 190mmwater (range 80 to 540mm water). Hypoglycorrhacia occurs in the majority of cases, with Ogawa et al. (1987) reporting a median of 30 mg/dl in 49% of samples. However, depression of CSF glucose may not be found in diabetic patients, thus obscuring the diagnosis. Therefore, simultaneous determinations of blood glucose should be performed such that one is not mislead in the initial evaluation. A CSF/blood glucose ratio less than I : 2 can be regarded as abnormal and should arouse suspicion. CSF protein is usually elevated in the range of 100 to 500 mgfdl. While Ogawa et al. (1987) observed a median protein range of 152 mg/ dl in 48% of patients with this disease, considerable variation in the quantity of CSF protein has been observed, with I series having a protein content of < 100 mg/dl in 25% of patients (Kennedy &. Fallon 1979). Therefore, caution is advised when interpreting CSF data with non-classical findings, as in TB meningitis, so that its diagnosis is not excluded. In advanced stages of the disease with protein levels of 1000 to 1500 mg/dl, spinal block may occur, which correlates with poor prognosis (Molavi & LeFrock 1985). Thus, in the case of smearnegative CSF, hypoglycorrhacia and elevated protein and opening pressures, it may still be difficult to differentiate between TB meningitis and other forms of meningitis (viral, fungal, acute syphilitic, Listeria monocytogenes, and meningeal infiltration

226

with neoplastic cells) [Molavi & LeFrock 1985; Parsons 1979]. Because of this difficulty in differentiating TB from other types of meningitis, biochemical techniques for rapid detection have been investigated. Measurements of adenosine deaminase, radioactive bromide partition testing and quantification of mycobacterial metabolites such as 3-(2'-ketohexyl)-indoline have either not proven satisfactory or do not have high specificity (Daniel 1987, 1988; Editorial 1985). Recently, competitive inhibition enzyme-linked immunosorbent assay (ELISA) of M. tuberculosis extracts have been studied (Bal et al. 1983). ELISA using BCG antigens have 95% specificity and 81 % sensitivity (Sada et al. 1983). Watt et al. (1988) evaluated use of ELISA for mycobacterial antigens and antibodies in the CSF of 29 patients with proven TB meningitis, along with 83 patients with non-TB CNS infections and 15 normal controls. The specificity was 96%, with 4 false negatives occurring in patients with TB meningitis. Thus, failure to demonstrate mycobacterial antigen or antibody did not exclude the diagnosis ofTB meningitis. However, ELISA techniques, for the most part, are too complex for routine use and are mainly of research value. A latex agglutination technique developed by Krambovitis et al. (1984) for the detection of M. tuberculosis antigen in CSF appears more practical, as it is inexpensive, does not require specialised equipment and should be usable in developing countries. It tested positive in 17 of 18 patients with TB meningitis and was negative in 133 of 134 control samples from patients with a variety of nonTB meningitis and other neurological diseases. Only I false positive was obtained from a patient with Haemophilus injluenzae meningitis. Thus, latex agglutination provided a sensitivity of 94.4% and specificity of 99.3%. As a simple test with the ability to provide timely information, it could potentially have widespread use.

2. Antituberculosis Drug Penetration into the CSF A prior survey of the literature (table I) indicates that 8 current antituberculosis drugs penetrate the CSF (Donald & Seifart 1988; Ellard et al.

Drugs 39 (2) 1990

1987; Geiseler et al. 1985; Holdiness 1985a). In inflamed meninges the CSF concentrations of these antibiotics are at least equal to or higher than that in non-inflamed meninges. The minimum inhibitory concentrations (MICs) are in excess of that required for inhibition of M. tuberculosis growth as established by in vitro studies (Holdiness 1985a). With the modern chemotherapeutic regimens available for treatment, intrathecal administration of these agents is not indicated (Holdiness 1984, 1985a).

3. Treatment Regimens In contrast to rapid advan

Management of tuberculosis meningitis.

This article examines the detection, assessment, and therapeutic modalities available for tuberculosis meningitis. Without appropriate treatment this ...
1MB Sizes 0 Downloads 0 Views