Enhanced Attenuation of Meningeal Inflammation and Brain Edema by Concomitant Administration of Anti-CD18 Monoclonal Antibodies and Dexamethasone in Experimental Haemophilus Meningitis Xavier Saez-Llorens, Hamid S. Jafari, Carlos Severien, Francisco Parras, Kurt D. Olsen, Eric J. Hansen, Irwin 1. Singer,* and George H. McCracken, Jr. Departments of Pediatrics and Microbiology, University of Texas Southwestern Medical Center, Dallas, Texas 75235-9063; and *Department ofBiochemical and Molecular Pathology, Merck Sharp & Dohme Research Laboratories, Rahway, New Jersey 07065 Abstract Elevated concentrations of proinflammatory substances such as TNF and IL-1 trigger initiation of the inflammatory Antiinflammatory therapy has been shown to reduce the ad- process and provoke profound pathophysiological changes that verse pathophysiological consequences that occur in bacterial result in alteration of cerebrospinal fluid (CSF)l hydrodynam- meningitis and to improve outcome from disease. In the present ics and in activation of the metabolic and coagulation path- study, modulation of two principal steps of the meningeal in- ways (7-9). Dexamethasone has been shown to inhibit produc- flammatory cascade was accomplished by concomitant adminis- tion of these cytokines at transcriptional and translational lev- tration of dexamethasone to diminish overproduction of cyto- els, especially when given before liberation of active bacterial kines in response to a bacterial stimulus and of a monoclonal components into the subarachnoid space (10, 1 1). antibody directed against adhesion-promoting receptors on leu- Influx of leukocytes into CSF can further amplify these kocytes to inhibit recruitment of white blood cells into the sub- pathophysiologic events because their accumulation contrib- arachnoid space. Dexamethasone and antibody therapy pro- utes to a sluggish CSF circulation and decreased resorption, duced a marked attenuation of all indices of meningeal inflam- and because of liberation of cytotoxic products on degranula- mation and reduction of brain water accumulation after H. tion (12). Before entering the subarachnoid space leukocytes influenzae-induced meningitis in rabbits compared with results must adhere to brain endothelial cells presumably through a of each agent given alone and of untreated animals. In addition, family of adhesion-promoting receptors and ligands located on the enhanced host's meningeal inflammatory reaction that fol- leukocyte cell membranes and endothelium that have been pre- lows antibiotic-induced bacterial lysis was profoundly amelior- viously activated by exposure to cytokines (13, 14). Treatment ated when dual therapy was administered without affecting of experimental lapine meningitis induced by intracisternal in- clearance rates of bacteria from cerebrospinal fluid and vascu- oculation of living bacteria (e.g., S. pneumoniae, H. influenzae, lar compartments. The combination of both therapeutic ap- N. meningitidis), their cell wall components (e.g., teichoic proaches may offer a promising mode of treatment to improve acid, lipopolysaccharide), or by human cytokines with mono- further the outcome from bacterial meningitis. (J. Clin. Invest. clonal antibodies (MAbs) directed against the CDl 1/CD 1 8 leu- 1991. 88:2003-2011.) Key words: anti-CD18 antibodies - dexa- kocyte adhesion receptors was shown to decrease transendoth- methasone * experimental meningitis elial passage of polymorphonuclear cells into CSF and to re- duce significantly indices of meningeal inflammation (6, 15). Introduction The use of anti-CD18 MAbs has also been shown to inhibit Antiinflammatory therapy of animals and humans with bacte- polymorphonuclear leukocyte accumulation and plasma leak- rial meningitis has been shown to decrease meningeal inflam- age in rabbits with experimentally-induced inflammatory skin mation, blood-brain barrier permeability, intracranial hyper- lesions (16) and to reduce multiple organ injury and death tension, brain edema, tissue damage, long-term neurological caused by increased leukocyte adhesiveness in an experimental sequelae, and mortality rates associated with this disease (1, 2). model of generalized ischemia-reperfusion injury (17). Likewise, the transient enhancement of the host's inflamma- The aim of this study was to assess whether the combina- tory response that commonly follows initiation of antibiotic tion of dexamethasone (DXM) and anti-CD 1 1/CD 18 antibod- treatment has been significantly attenuated when agents have ies would be more effective than either agent alone in prevent- been used that selectively interfere with production of inflam- ing or attenuating the pathophysiologic events that occur in matory mediators (e.g., dexamethasone, cyclooxygenase inhibi- experimental meningitis induced by H. influenzae type b or its tors) or with influx of leukocytes into the subarachnoid space purified lipooligosaccharide (LOS). (e.g., pentoxifylline, anti-CD 1 8 receptor antibodies) (3-6). Methods Address correspondence to George H. McCracken, Jr., M.D., Dept. of Pediatrics, UT Southwestern Medical Center, 5323 Harry Hines Blvd., Preparation ofHib endotoxin Dallas, TX 75235-9063. Hib endotoxin (LOS) was purified from cells of Hib strain DL42 by Receivedfor publication 16 January 1991 and in revisedform 29 using the hot phenol-water method of Westphal and Jann (18), as July 1991. modified by Johnson and Perry (19). The purity of this LOS prepara- tion was confirmed by SDS-PAGE, followed by silver staining (20) and J. Clin. Invest. © The American Society for Clinical Investigation, Inc. 0021-9738/91/12/2003/09 $2.00 1. Abbreviations used in this paper: CSF, cerebrospinal fluid; DXM, Volume 88, December 1991, 2003-2011 dexamethasone; LOS, lipooligosaccharide. Treatment ofExperimental Meningitis with Anti-CD18 Antibodies and Dexamethasone 2003 Western blot analysis using hyperimmune rat serum to Hib DL42 (21). ELISA assays were conducted to measure the plasma concentration 20 ng of the endotoxin, which is contained in 2 X 106 bacterial cells, of IB4 or OKM-l MAbs. Immulon-2 (Dynatech Laboratories, Inc., was chosen as the inoculum to induce meningitis because this amount Alexandria, VA) 96-well plates were coated with affinity-purified rab- has consistently induced meningeal inflammation in this rabbit model bit anti-mouse IgG2a (Zymed Labs Inc., San Francisco, CA) for detec- (22) and because CSF samples from patients with Hib meningitis con- tion of IB4, or affinity-purified rabbit anti-mouse IgG2b (Zymed Labs) tain from 104 to I0' CFU/ml. for detection of OKM- 1. These trapping antibodies were dissolved at 5 ,ug/ml in 0.1 M NaHCO3 for 1 h at 250C. After this and all subsequent Bacterial strain steps, washing was performed with PBS using a ProPette automated Hib strain DL42 is a non-fl-lactamase-producing strain isolated from plate washer (Cetus Corp., Emeryville, CA). Samples of rabbit plasma the CSF of a child with meningitis. This strain is fully virulent in the containing unknown amounts of IB4 or OKM- 1 were diluted in PBS infant rat model for Hib invasive disease and belongs to LOS antigenic containing 0.05% Tween-20 and 0.1% BSA and added to the wells; group 2, which is the predominant LOS antigenic type among invasive 100-,gl aliquots were incubated for 1 h at 370C in triplicate. Standard Hib isolates (23). The minimum inhibitory concentration of ceftriax- calibration curves were constructed using purified IB4 or OKM-I at one for this Hib strain was 0.01 fIg/ml. concentrations ranging from 10 to 400 ng/ml. After washing, bound A final inoculum of - 1 x 106 colonies/ml taken from a culture IB4 or OKM-I were detected with a 1:500 dilution (in PBS containing which contained bacteria in a log phase of growth was prepared. Ani- 1% BSA) of alkaline phosphatase-conjugated affinity-purified rabbit mals were then intracisternally inoculated with 0.2 ml of this bacterial anti-mouse IgG2a (Zymed Labs) for IB4 and anti-mouse IgG2b for suspension. detection of OKM-1. After incubation for 1 h at 370C and further washing, the quantities of bound conjugate were detected by incubat- Experimental meningitis model ing all wells with a 1 mg/ml solution ofp-nitrophenyl phosphate in 0.1 This study was approved by the Institutional Review Board for Animal M 2,2 amino-methyl-propanediol buffer (pH 10.3, Zymed Labs) for 30 Research of The University of Texas Southwestern Medical Center at min at 250C. The absorbance of each well was determined at 405 nm Dallas. with a UV Max ELISA plate reader (Molecular Devices, Inc., Menlo The rabbit model of experimental meningitis, originally described Park, CA), and the plasma concentrations of IB4 or OKM-I calculated by Dacey and Sande (24), was used in a modified form. New Zealand using the accompanying software (Molecular Devices, Inc.). White male rabbits (2-3 kg) were anesthetized with intramuscular in- jections of ketamine (40 mg/kg) and acepromazine (3 mg/kg) and Study design placed in a stereotactic frame. A spinal needle was introduced into the Three different sets of experiments were conducted: (a) Hib LOS-in- cisterna magna and 0.2 ml ofCSF withdrawn. Animals were then inoc- duced meningitis. (b) Hib-live organism-induced meningitis without ulated with same volume of either Hib LOS or live Hib organisms. antibiotic treatment. (c) Hib-live organism-induced meningitis with Subsequent CSF samples (0.2-0.3 ml) were obtained at different times antibiotic treatment. according to study design. Immediately after collection, the CSF ali- Experiment 1. 36 rabbits received