Antimicrobial Therapy of Experimental Legionella micdadei

Transcription

1 ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Dec. 1985, p /85/ $02.00/0 Copyright C) 1985, American Society for Microbiology Vol. 28, No. 6 Antimicrobial Therapy of Experimental Legionella micdadei Pneumonia in Guinea Pigs A. WILLIAM PASCULLE,l2* JOHN N. DOWLING,2 3 FRANK N. FROLA,1 DAVID A. McDEVITT,3 AND MICHAEL A. LEVI2 Departments of Pathology1 and Medicine,3 School of Medicine, and Department of Infectious Disease and Microbiology,2 Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania Received 7 June 1985/Accepted 30 September 1985 Several antimicrobial agents were evaluated for activity against experimental Legionella micdadei pneumonia in guinea pigs. Erythromycin, rifampin, doxycycline, and sulfamethoxazole- produced significant reductions in mortality. Penicillin, cefazolin, cefoxitin, chloramphenicol, and gentamicin were not efficacious even though, at the doses administered, the peak concentrations of these agents in serum substantially exceeded their MICs for the test strain. It is suggested that the poor performance of the latter group of agents resulted from poor penetration into cells in which L. micdadei was multiplying. Legionella micdadei causes acute purulent pneumonia, usually among patients who are immunosuppressed or otherwise severely compromised (15, 20, 22, 23). Our clinical experience with the initial cases of Pittsburgh pneumonia suggested that cephalosporins, alone or in combination with aminoglycosides, were not therapeutically effective (2, 16). Further studies with embryonated eggs demonstrated that erythromycin, rifampin, and the combination of sulfamethoxazole with protected embryos from death due to L. miedadei (2). Based on this information and the reported efficacy in L. pneumophila pneumonia, erythromycin, with or without rifampin, was successfully used for therapy of L. tnicdadei pneumonia (2, 15, 31). In vitro studies of the susceptibility of L. micdadei with an agar dilution technique revealed that the organism is susceptible to a large number of antinmicrobial agents, many of which were not effective in embryonated eggs (18), suggesting that in vitro susceptibility might not reliably predict the clinical efficacy of antimicrobial agents against L. micdadei. It has already been demonstrated that such is the case for L. pneumophila, which is also sensitive in vitro to a number of agents which are not efficacious in embryonated eggs or guinea pigs (6, 10, 28). Since effective therapy of both Legionnaires disease and Pittsburgh pneumonia has been arrived at empirically, it is unlikely that controlled clinical trials of antimicrobial therapy will be conducted. Accordingly, we studied the efficacy of several antimicrobial compounds for the therapy of experimental L. micdadei pneumonia in guinea pigs. MATERIALS AND METHODS Media. Buffered charcoal yeast extract agar (19) containing 0.1% oa-ketoglutarate (Calbiochem-Behring, La Jolla, Calif.) (5) and 1% bovine serum albumin (ABCYE) was used for the production of primary working cultures of the organism and for the culture of tissues from infected animals. Bacteria. L. micdadei EK, which had originally been isolated in embryonated eggs from lung tissue of a patient with Pittsburgh pneumonia (20) was used for all studies. After its initial isolation, stock cultures of this organism were maintained by passage in embryonated eggs. Stock suspen- * Corresponding author. 730 sions of pooled yolk material from several infected eggs were stored at -70 C. Working bacterial cultures for infection of the animals were prepared by inoculating 0.1 ml of the egg yolk stock culture onto each of several plates of ABCYE. These primary cultures were incubated for 48 h at 35 C in air. Bacteria from several plates were removed and suspended in 0.9% NaCl so that the density of the suspension matched that of a McFarland no. 4 standard, equivalent to about 108 CFU/ml. Animals. Male, Bordetella-free, Hartley strain guinea pigs weighing 250 to 300 g (Hilltop Animal Laboratories, Scottdale, Pa.) were housed five per cage and given free acess to commercial guinea pig chow and water. The baseline weight and rectal temperature of each animal were determined on the three consecutive days preceding an experiment. Antimicrobial agents. All antimicrobial agents, with the exception of rifampin and chloramphenicol, were standard commercial preparations intended for intravenous infusion. Rifampin (Calbiochem-Behring) was dissolved in dimethyl sulfoxide and suspended in water prior to dilution. Chloramphenicol base (Calbiochem-Behring) was used because guinea pigs are not able to hydrolyze chloramphenicol succinate (H. C. Neu, personal communication). This compound was dissolved in methanol and diluted in sterile water for injection. Infection and antimicrobial therapy. Animals were infected by intratracheal inoculation with L. micdadei. Anesthesia was induced by intramuscular administration of 25 mg of ketamine hydrochloride (Parke, Davis & Co., Detroit, Mich.) and 0.25 mg of acepromazine (Ayerst Laboratories, New York, N.Y.). An area on the anterior aspect of the neck was then shaved and anesthetized by subcutaneous injection of 0.2 ml of 0.5% lidocaine solution. A small midline incision was made, and the trachea was exposed by blunt dissection. Subsequently, 0.3 ml of bacterial suspension was slowly introduced into the trachea with a tuberculin syringe fitted with a needle (27 gauge by 0.5 inch [12.5 mm]). Each animal was held erect for 30 to 45 s to ensure that the inoculum was inspired into the lungs before the surgical wound was closed with pediatric surgical clips. Groups of 10 guinea pigs were used for each infected experimental and control group. In early experiments, there

2 VOL. 28, 1985 ANTIMICROBIAL THERAPY OF LEGIONELLA MICDADEI PNEUMONIA 731 were no deaths among animals which were inoculated with 0.2 ml of saline intratracheally, and mock-infected controls were not routinely used thereafter. One group of infected animals received no antimicrobial therapy, while groups of 10 infected animals were given each antimicrobial agent. A group of five uninfected guinea pigs simultaneously received each antimicrobial agent. Insofar as was possible, an antimicrobial agent which was likely to prove efficacious was compared with one or more agents less likely to be therapeutic in each experiment. Antimicrobial therapy was begun 24 h after infection, by which time all of the infected guinea pigs had become febrile. All antimicrobial agents were administered subcutaneously at 8:30 a.m. and 6:00 p.m. for the next 7 consecutive days to both treated animals and antibiotic control groups. The animals were observed for an additional 7 days after cessation of therapy. The weights and rectal temperatures of all animals were monitored daily throughout each experiment. In preliminary experiments, several uninfected animals developed fatal cecitis and colitis after administration of cephalothin or erythromycin. The cecal contents of the animals were negative for Clostridium difficile or C. difficile toxin but contained large numbers of Escherichia coli. In addition, blood cultures of some of these animals grew E. coli. It was determined that this antibiotic-induced bacterial overgrowth of the bowel (9) could be prevented by oral administration of Lactobacillus acidophilus and L. bulgaricus (Lactinex granules; Hynson, Wescott & Dunning, Baltimore, Md.). For this reason, all animals also received 0.5 ml of Lactobacillus suspension twice daily by mouth. All animals which died were autopsied to ensure that death had been due to L. micdadei. The lungs were cultured on ABCYE for L. micdadei and on blood agar (BBL Microbiology Systems, Cockeysville, Md.) for other bacteria. Impression smears made from lung tissue were examined both by Gram and modified acid-fast stains and by direct fluorescent antibody staining with L. micdadei conjugates kindly supplied by the Biological Products Section of the Centers for Disease Control, Atlanta. Ga. The efficacy of each antimicrobial agent was judged primarily by its ability to prevent death during the 14 days following infection with L. micdadei. Since there were no deaths in either sham-infected or uninfected antibiotic control groups, the mortality in each group of treated animals was compared with that of untreated infected animals with the Fisher exact binomial test (one-tailed). For comparison of the times of survival of treated and untreated groups, the survival intervals, recorded in half-day intervals, were ranked, and the Mann-Whitney rank sum test was used. Antibiotic concentrations. On the fifth day of antimicrobial therapy, animals in the antibiotic control group were bled by cardiac puncture for determination of serum antibiotic concentrations. At each of the five hourly intervals following the dose, a different one of the five animals was bled. The concentrations of most agents were determined by disk diffusion assay (24) with either Bacillus sutptilis or Micrococcus luteus as an indicator organism. Gentamicin concentrations in serum were determined by radioimmunoassay, and chloramphenicol concentrations in serum were measured by high-pressure liquid chromatography. The latter two assays were performed by Warren Diven of the Department of Pathology, University of Pittsburgh School of Medicine. Concentrations of sulfamethoxazole and in serum were measured by H. Spiegel (Hoffman-La Roche Inc., Nutley, N.J.) by high-pressure liquid chromatography (29, 30). A composite clearance curve for each antimicrobial agent was constructed by regression analysis of the pooled data from the separate animals, and the serum half-life of each agent was calculated by the least-squares method. RESULTS Intratracheal inoculation of guinea pigs with 3 x 107 CFU of L. micdadei produced acute purulent bronchopneumonia which was similar to the infection seen in humans and to that which we have previously described in guinea pigs (17). Mortality among untreated, infected animals in the various experiments ranged from 60 to 90% (Table 1). All animals that died in both the treated and untreated groups were culture positive for L. micdadei, and organisms could be seen after direct fluorescent antibody staining of lung impression smears. Animals treated with erythromycin, rifampin, sulfamethoxazole-, or high-dose doxycycline were significantly protected from death after infection with L. micdadei (Table 1). Therapy with penicillin G, cefazolin, cefoxitin, chloramphenicol, or gentamicin had no significant effect on mortality. There were no observable side effects among control animals receiving only antimicrobial therapy. It appeared that therapy with the 1-lactam antibiotics or chloramphenicol may have delayed the death of guinea pigs, even though the final mortality of animals to which these agents were administered was no different than that of untreated, infected controls (Table 1). The increased interval until death compared with untreated controls in the same experiment was statistically significant in the case of penicillin and cefoxitin (P < 0.05) but did not reach significance for cefazolin or chloramphenicol (P > 0.05). In contrast, the TABLE 1. Lethality of L. micdadei pneumonia in guinea pigs treated with various antimicrobial agents Expt. Antimicrobial Daily Deaths/total Mean days dose no. in group Pa to death no. agent (mg/kg) (%) (range) 1 None 9/10 (90) 3.2 (2-27) Erythromycin 120 3/10 (30) (2-3.5) Cefazolin 240 9/10 (90) NSb 4.5 (2-8) Cefoxitin /10 (100) NS 4.9 (2-7) 2 None 9/10 (90) 2.4 (2-6) Rifampin 24 4/10 (40) (2-4) Doxycycline 4 6/10 (60) NS 3.6 (2-6.5) 3 None 7/10 (70) 3.6 (2-5.5) Doxycycline 36 1/10 (10) Chloramphenicol 60 8/10 (80) NS 5.0 (2.5-9) 4 None 7/10 (70) 4.1 (1.5-7) Gentamicin 9.6 8/10 (80) NS 2.7 (2-4) 5 None 6/9 (67) 3.0 (2-7) Penicillin /10 (70) NS 7.1 (2.5-13) Sulfamethoxazole- 120/24 0/10 (0) None 9/10 (90) 4.0 (2-8) Sulfamethoxazole- 120/24 4/10 (40) (2-3) '8 Fisher exact test (one-tailed). " NS, Not significant (P > 0.05).

3 732 PASCULLE ET AL. ANTIMICROB. AGENTS CHEMOTHER. TABLE 2. Dosages and resultant concentrations in serum of antimicrobial agents administered to guinea pigs Antimicrobial agent Dose Concentraion (tig/ml) in serum at time (h) after dose: Half-life (mg/kg) (h) Erythromycin ± ± ± ± Rifampin ± ± ± ± Sulfamethoxazole- 60/12 184/1.1 98/0" 81/ /NDb Doxycycline ± ± ± Penicillin G Cefazolin ± ± ± Cefoxitin ± ± ± Chloramphenicol ND Gentamicin a None detectable. b Half-life could not be determined because of low concentrations in serum. tifme until death of animals treated with gentamicin was shorter than that of concurrently infected controls, although this difference was not significant (P > 0.05). The levels of antibiotics in the sera of uninfected animals were remarkably similar to the values reported by Edelstein et al. (6), who measured antibiotic concentrations in the sera of guinea pigs infected with L. pneumophila. The pharmacokinetic data (Table 2) demonstrated that treated animals developed concentrations of most of the antimicrobial agents in their sera which were similar to those expected in humans receiving similar dosages. Doxycycline, however, was an exception. In the first experiment with this agent, a,nmals received a dose of 4 mg/kg per day, which appeared to be only partially efficacious. This dosage did not produce measurable doxycycline concentrations in serum. When animals were treated with a relatively large dose of doxycycline (36 mg/kg per day), the resulting concentrations in serum were comparable to those reported in humans given a 10-fold lower dosage (27). Sulfamethoxazole was the only other agent for which the rate of clearance differed from that qsually seen in humans. The half-life of this agent (2.3 h) was about one-fourth that which has been reported in humans (4) Ṗeak concentrations of each drug in serum were compared with the MICs measured in vitro for the strain of L. micdadei used in these studies (18). The MICs of erythromycin, rifampin, penicillin G, cefazolin, cefoxitin, and gentamicin are less than 1,ug/ml for L. micdadei. The peak concentration of these agents in serum ranged from 15 times (erythromycin) to over 600 times (cefoxitin) the respective MICs. The peak concentrations in serum were approximately equal to the MICs for doxycycline (2.8 pg/ml) and chloramphenicol (1.4,ug/ml). DISCUSSION L. micdadei is susceptible in vitro to most antimicrobial agents (18). The results of the present study, along with those of limited clinical appraisals (2, 16) and studies in eggs (16), indicate that the results of in vitro susceptibility testing do not predict the clinical efficacy of antimicrobial agents in the treatment of Pittsburgh pneumonia. Significant protection from death due to L. micdadei pneumonia was afforded by erythromycin, rifampin, doxycycline, and the combination of sulfamethoxazole. Erythromycin, rifampin, and sulfamethoxazole- have also been shown to protect egg embryos against a lethal L. micdadei inoculum; tetracyclines were not tested in ovo (16). That there is agreement between the results in eggs and guinea pigs is likely due to the fact that the bacteria are largely intracellular in both experimental models. The major disadvantage of the egg model is the difficulty in relating the concentrations of antimicrobial agents achieved in ovo to the dosage regimens and corresponding levels obtained in serum in humans. Our results closely parallel those of Edelstein and his co-workers (6), who used the same intratracheal inoculation model in therapeutic trials for experimental L. pneumophila pneumonia. Despite the efficacy of erythromycin in both in vivo models and human disease, there are disturbing'problems with its clinical use. Some patients with Legionella pneumonia fail to 'respond to erythromycin, and other patients relapse after seemingly adequate courses of therapy. Furthermore, the continued isolation of Legionella organisms after the institution of erythromycin therapy has been described (7, 8). The persistence of viable bacteria in the lung during and after therapy with erythromycin has also been observed in the guinea pig model of L. pneumophila pneumonia (6, 12). It would be prohibitively expensive to determine whether one of the other three effective antimicrobial agents is superior to erythromycin in the guinea pig model. For example, if the true mortality with erythromycin therapy is 30% (Table 1), to have a 90% chance (1-3) of detecting that the mortality of an alternative treatment is 10% at the 5% significance level (a) would require about 65 animals in each therapy group (26). However, parameters other than moratlity can be assessed with in vivo model infections. Edelstein et al. (6) found that rifampin effected more rapid clearance of L. pneumophila from the lungs and more rapid normalization of histologic changes than did erythromycin. Likewise, in a model of L. penumophila pneumonia produced by an aerosolized inoculum, rifampin appeared to be more efficacious than erythromycin in promoting survival and eliminating bacteria from the lungs (12). Despite these apparent advantages of rifampin over erythromycin, rifampin cannot be used alone for therapy of legionellosis because, as we have demonstrated, variants with a high level of resistance to rifampin are selected upon exposure to this agent in vitro (3) İt is not possible to determine the susceptibility of legionellae to either sulfonamides alone or their combination with in vitro, because all media which support the growth of legionellae contain para-aminobenzoic acid (18). We found that the MIC of for L. micdadei was 3.2 pug/ml and that for L. pneumophila was 2.7,g/ml (18); however, these values may be falsely high because it is likely that the BCYE used contained inhibitors of. Additional animal model studies will be required to

4 VOL. 28, 1985 ANTIMICROBIAL THERAPY OF LEGIONELLA MICDADEI PNEUMONIA 733 determine whether one or both of the agents is active alone and to determine whether synergy is afforded by the combination. The legionellae are facultative intracellular pathogens which appear to multiply exclusively within the phagocytic cells of susceptible hosts. All four antimicrobial agents which are efficacious against both L. micdadei and L. pneumophila pneumonias penetrate host phagocytic cells well (13, 21, 25). Thus, efficacy appears to be correlated with attaining an intracellular level of the antimicrobial agent which exceeds the MIC. It is clear that intracellular antimicrobial bactericidal activity is not a requirement for effective therapy. Rifampin is bactericidal for L. micdadei (3) and L. pneumophila (11) in vitro at concentrations which would be exceeded 40-fold intracellularly. Yet Horwitz and Silverstein (14) have demonstrated that neither rifampin nor erythromycin is bactericidal for L. pneumophila within human mononuclear phagocytes. Perhaps factors such as physical separation of the bacteria and the antimicrobial agent within the host cell or diminished activity of intracellular antimicrobial agents may explain these findings. Whether doxycycline or sulfamethoxazole- has bactericidal activity for intracellular legionellae remains to be examined. The inactivity of all P-lactam antibiotics was presumably due to inability to penetrate phagocytes, since the concentrations of these agents in serum were many fold greater than their MICs. The suggestion of a somewhat prolonged survival time of guinea pigs infected with L. micdadei and treated with,b-lactam antibiotics despite no decrease in final mortality suggests that some ineffective antibiotics can favorably influence the course of disease through the killing of extracellular organisms. Alternatively, the small amounts of these drugs which do penetrate host cells may produce a minor amount of intracellular antibacterial activity that is of some benefit to the host. Chloramphenicol and gentamicin are also not effective in treating experimental L. micdadei pneumonia. In the case of chloramphenicol, the concentrations of this agent in serum were approximately equal to the MIC. We do not know whether the failure of chloramphenicol to significantly reduce mortality was due to the failure of the agent to localize in cells or to diminished intracellular activity of the drug. The finding that treated animals died more slowly suggests that some chloramphenicol may have penetrated the phagocytes of the animals. Although gentamicin did not reduce mortality from L. micdadei pneumonia, a high dose of gentamicin (16 mg/day) did reduce mortality and resulted in faster clearance of L. pneumophila from the lungs of guinea pigs infected with one 50% lethal dose by aerosolization (12). Gentamicin was not effective at a lower dose or at the higher dose when the animals were infected with 10 times the 50% lethal dose. In these studies, therapy was administered intramuscularly beginning 24 h after infection, but an additional initial loading dose was given intraperitoneally to produce high levels immediately. High extracellular concentrations of gentamicin may have been present at a time when a significant proportion of bacteria were still extracellular, and bacteria which had not yet been phagocytosed could have been killed effectively. This again suggests that killing of extracellular Legionella organisms may be beneficial to the host in some models in which the inoculum is relatively small and the concentration of largely extracellular antibiotics is great. Inexplicable at the present time is the finding of Bacheson et al. (1) that gentamicin was more effective than erythromycin in killing L. pneumophila in L cells in vitro. We have not found that gentamicin has any effect on the viability of L. micdadei in L cells (unpublished data). The reason for the disparate results is unknown, unless Bacheson et al. were measuring an effect of gentamicin on the killing of extracellular bacteria in this cell culture system. ACKNOWLEDGMENTS We thank Washington Winn, Jr., and Paul Edelstein for their helpful suggestions and Betty Edwards for secretarial assistance. This work was supported in part by Public Health Service grant Al from the National Institute of Allergy and Infectious Diseases. LITERATURE CITED 1. Bacheson, M. A., H. M. Friedman, and C. E. Benson Antimicrobial susceptibility of intracellular Legionella pneumophila. Antimicrob. Agents Chemother. 20: Dowling, J. N Clinical aspects of Pittsburgh pneumonia, p In D. Schlessinger (ed.), Microbiology American Society for Microbiology, Washington, D.C. 3. Dowling, J. N., R. S. Weyant, and A. W. Pasculle Bactericidal activity of antibiotics against Legionella micdadei (Pittsburgh pneumonia agent). Antimicrob. Agents Chemother. 22: Dudley, M. N., R. E. Levitz, R. Quintiliani, J. M. Hikingbotham, and C. H. Nightingale Pharmacokinetics of and sulfamethoxazole in serum and cerebrospinal fluid of adult patients with normal meninges. Antimicrob. Agents Chemother. 26: Edelstein, P. H Improved semiselective medium for isolation of Legionella pneumophila from contaminated clinical and environmental specimens. J. Clin. Microbiol. 14: Edelstein, P. H., K. Calarco, and V. K. Yasui Antimicrobial therapy of experimentally induced Legionnaires' disease in guinea pigs. Am. Rev. Respir. Dis. 130: Edelstein, P. H., R. D. Meyer, and S. M. Finegold Laboratory diagnosis of legionnaires' disease. Am. Rev. Respir. Dis. 121: Edelstein, P. H., R. D. Meyer, and S. M. Finegold Long-term followup of two patients with pulmonary cavitation caused by Legionella pneumophila. Am. Rev. Respir. Dis. 124: Farrar, W. E., Jr., T. H. Kent, and V. B. Elliott Lethal gram-negative superinfection in guinea pigs given bacitracin. J. Bacteriol. 2: Fraser, D. W., I. K. Wachsmuth, C. Bopp, J. C. Feeley, and T. F. Tsai Antibiotic treatment of guineapigs infected with agent of Legionnaires' disease. Lancet i: Gibson, D. H., and R. B. Fitzgeorge Persistence in serum and lungs of guinea pigs of erythromycin, gentamicin, chloramphenicol and rifampicin and their in-vitro activities against Legionella pneumophila. J. Antimicrob. Chemother. 12: Gibson, D. H., R. B. Fitzgeorge, and A. Baskerville Antibiotic therapy of experimental airborne Legionnaires' disease. J. Infection 7: Hand, W. L., W. R. Corwin, T. H. Steinberg, and G. D. Grossman Uptake of antibiotics by human alveolar macrophages. Am. Rev. Respir. Dis. 129: Horwitz, M. A., and S. C. Silverstein Intracellular multiplication of Legionnaires' disease bacteria (Legionella pneumophila) in human monocytes in reversibly inhibited by erythromycin and rifampin. J. Clin. Invest. 71: Muder, R. R., V. L. Yu, and J. J. Zuravleff Pneumonia due to the Pittsburgh pneumonia agent: new clinical perspective with a review of the literature. Medicine (Baltimore) 62: Myerowitz, R. L., A. W. Pasculle, J. N. Dowling, G. J. Pazin, M. Puerzer, R. B. Yee, C. R. Rinaldo, Jr., and T. R. Hakala Opportunistic lung infection due to "Pittsburgh pneumonia agent." N. Engl. J. Med. 301:

5 734 PASCULLE ET AL. 17. Pasculle, A. W Experimental studies of Pittsburgh pneumonia agent, p In D. Schlessinger (ed.), Microbiology American Society for Microbiology, Washington, D.C. 18. Pasculle, A. W., J. N. Dowling, R. S. Weyant, J. M. Sniffen, L. G. Cordes, G. M. Gorman, and J. C. Feeley Susceptibility of Pittsburgh pneumonia agent (Legionella micdadei) and other newly recognized members of the genus Legionella to nineteen antimicrobial agents. Antimicrob. Agents Chemother. 20: Pasculle, A. W., J. C. Feeley, R. J. Gibson, L. G. Cordes, R. L. Myerowitz, C. M. Patton, G. W. Gorman, C. L. Carmack, J. W. Ezzell, and J. N. Dowling Pittsburgh pneumonia agent: direct isolation from human lung tissue. J. Infect. Dis. 141: Pasculle, A. W., R. L. Myerowitz, and C. R. Rinaldo, Jr New bacterial agent of pneumonia isolated from renaltransplant recipients. Lancet ii: Prokesch, R. C., and W. L. Hand Antibiotic entry into human polymorphonuclear leukocytes. Antimicrob. Agents Chemother. 21: Rogers, B. H., G. R. Donowitz, G. K. Walker, S. A. Harding, and M. A. Sande Opportunistic pneumonia. A clinicopathological study of five cass caused by an unidentified acid-fast bacterium. N. Engl. J. Med. 301: Rudin, J. E., and E. J. Wing A comparative study of Legionella micdadei and other nosocomial acquired pneumonia. Chest 86: ANTIMICROB. AGENTS CHEMOTHER. 24. Sabath, L. D., J. I. Casey, P. A. Ruch, L. L. Stumpf, and M. Finland Rapid microassay of gentamicin, kanamycin, neomycin, streptomycin, and vancomycin in serum or plasma. J. Lab. Clin. Med. 78: Seger, R. A., S. Baumgartner, L. X. Tiefenauer, and F. K. Gmunder Chronic granulomatous disease: effect of sulfamethoxazole/ on neutrophil microbicidal function. Helv. Paediatr. Acta 36: Steel, R. G., and J. H. Torrie Principles and procedures of statistics, p McGraw-Hill Book Co., New York. 27. Steigbigel, N. H., C. W. Reed, and M. Finland Absorption and excretion of five tetracycline analogues in normal young men. Am. J. Med. Sci. 255: Toro, J., K. Sawatari, T. Kanda, A. Saito, and K. Hara New 1-lactamase-resistant cephem treatment of guinea pigs infected with Legionella pneumophila. Microbiol. Immunol. 27: Vree, T. B., Y. A. Hekster, A. M. Baars, J. E. Damsma, and E. Van Der Klein Determination of and sulfamethoxazole (co-trimoxazole) in body fluids of man by means of high-performance liquid chromatography. J. Chromatogr. 146: Winfeld, R. E., and T. C. Macasieb Determination of in biological fluids by high-performance liquid chromatography. J. Chromatogr. 9: Wing, E. J., F. J. Schafer, and A. W. Pasculle Successful treatment of Legionella micdadei (Pittsburgh pneumonia agent) pneumonia with erythromycin. Am. J. Med. 71: Downloaded from on November 25, 2018 by guest