Two groups of rabbits with experimental meningitis induced by direct intracisternal inoculation of Streptococcus pneumoniae cells were studied. One group was rendered profoundly leukopenic by nitrogen mustard, and the other had normal leukocyte counts. The two groups had comparable bacterial growth rates (mean generation time, 60 versus 67 min) and ultimate bacterial populations in the cerebrospinal fluid (CSF) (mean log10 CFU, 9.1 versus 8.7); therefore leukocytes did not effectively slow or limit the growth of pneumococci in the CSF in vivo. Increased CSF protein, decreased CSF glucose, and increased CSF lactate levels were similar in both groups, suggesting that leukocytes are not essential for these changes to occur. Quantitative blood cultures revealed identical levels of pneumococcal bacteremia until 13 to 14 h after the initiation of infection, when the leukopenic rabbits showed a larger number of pneumococci in the blood, ultimately exceeding the number reached in nonleukopenic rabbits by 100-fold. Leukocytes therefore limit the extent of pneumococcal bacteremia after infection of the CSF despite their lack of effect on the course or the CSF manifestations of experimental meningitis.

No abstract

Ceftriaxone was highly active in eliminating Escherichia coli from the cerebrospinal fluid of rabbits infected with experimental meningitis. However, concentrations equal to or greater than 10 times the minimal bactericidal concentration had to be achieved to ensure optimal efficacy (rate of kill, 1.5 log10 CFU/ml per h). In contrast to other beta-lactams studied in this model, ceftriaxone concentrations in cerebrospinal fluid progressively increased, whereas serum steady state was obtained by constant infusion. The percent penetration was 2.1% after 1 h of therapy, in contrast to 8.9% after 7 h (P less than 0.001). In vitro time-kill curves done in cerebrospinal fluid or broth more closely predicted the drug concentrations required for a maximum cidal effect in vivo than that predicted by determinations of minimal inhibitory or bactericidal concentrations.

We evaluated the pharmacokinetics and therapeutic elficacies of piperacillin and tazobactam, a j-lactamase inhibitor, given either alone or in different combinations (80:10, 200:10, and 80:25 mg/kg/h), in experimental meningitis due to a strain of KlebsieUla pneumoniae producing the TEM-3 extended-spectrum P-lactamase.Treatment was administered intravenously as a 7-h constant infusion preceded by a bolus of 20%o of the total dose. The mean (± standard deviation) rates of penetration into the cerebrospinal fluid (CSF) of infected animals were 6.7 ± 3.9% 1for piperacillin given alone and 36.3 ± 21.9%o for tazobactam given alone. Combination treatment significantly magnified the concentration of either drug in CSF. Concentrations of bacteria in CSF increased throughout therapy in animals given either drug alone, even at high dosages. In animals given the combination at'dosages of 80:10 and 200/10 mg/kg/h, only a suboptimal reduction of CSF bacterial titers was obtained in vivo, i.e. -0.49 0.34 and -0.73 ± 0.49 log CFU/ml/h, respectively. An increase in the tazobactam dosage within the combination (80:25 mg/kglh) was required in order to obtain a significantly faster elimination of viable organisms from the CSF (-0.97 ± 0.35 log CFU/ml/h). The study shows that ta?obactam is able to provide eflective protection' against piperacillin hydrolysis by the TEM-3 enzyme within the CSF. Appropriate dosage regimens of various P-lactam-tazobactam combinations may deserve comparative studies in experimental meningitis caused by organisms producing extended-spectrum P-lactamases.Gram-negative bacilli producing ,B-lactamases that hydrolyze extended-spectrum cephalosporins have been identified since 1983 (17, 22). Nosocomial infections due to members of the family Enterobacteriaceae that produce these extendedspectrum 13-lactamases have been observed in studies from a number of institutions, predominantly in Europe (3,14,16,18). Therefore, neonatal or neurosurgery-associated meningitis caused by enteric bacilli producing such ,B-lactamases is likely to occur in the future. Antimicrobial chemotherapy of serious infections caused by these organisms is difficult, and novel regimens must be investigated with animal models. On an in vitro basis, the combination of a P-lactam derivative with a 1-lactamase inhibitor is one of the few available options. Tazobactam is a ,B-lactamase inhibitor recently introduced in the clinical setting. In vitro, a combination of tazobactam and piperacillin has been found to be promising against ,1-lactamase-producing members of the Enterobacteriaceae (1, 11-13). In a model of meningitis caused by a strain of Escherichia coli producing the TEM-1 f-lactamase, the standard 8:1 ratio of piperacillin to tazobactam was found to be moderately effective, but other ratios were not investigated (15). Using the same well-standardized rabbit model, we evaluated (i) the kinetics and efficacies of piperacillin and tazobactam, given alone or in combination, in the cerebrospinal fluid (CSF) of rabbits with meningitis due...

Cerebral tuberculosis (TB) was diagnosed in 6 (4%) of 156 HIV-infected patients with TB seen at our institution over 6 years. We describe here the clinical and radiologic features of these cases and of 15 others reported in the literature. Of the 21 patients, 59% were intravenous drug users. Presenting symptoms were fever (76%), confusion (52%), seizures (38%), and headache (38%). Fourteen patients (66%) had previous or active extracerebral TB at presentation. Cranial CT scan showed ring-(62%) or nodular-(24%) enhancing lesions or mixed forms (14%). Among the 12 patients who underwent a brain biopsy, bacteriologic evidence of TB was found in 9. Four patients (19%) died during hospitalization. Among the 17 others who received antituberculous therapy, only 1 developed neurologic sequelae. Five patients also received steroid therapy to control cerebral edema or paradoxical growth of the cerebral mass lesions. TB should be considered as a cause of cerebral mass lesions in HIV-infected patients, especially if tuberculous infection is suspected at other sites.