Treatment of Human Brucellosis with Netilmicin and Doxycycline

Transcription

1 441 Treatment of Human Brucellosis with Netilmicin and Doxycycline Javier Solera, Alfredo Espinosa, Paloma Geijo, Elisa Martinez-Alfaro, Lourdes Saez, Maria Antonia Sepulveda, and Maria Dolores Ruiz-Rib6, for the GECMEI* From the Department ofmedicine, Unit ofinfectious Diseases, Albacete Hospital, Albacete; Department ofmedicine, Cuenca Hospital, Cuenca; and Department ofmedicine, Toledo Hospital, Toledo, Spain We conducted a prospective, noncomparative, multicenter study to assess the safety and efficacy of doxycycline and netilmicin in the treatment of human brucellosis. The study included 64 patients who had acute brucellosis without endocarditis or neurobrucellosis. The treatment schedule consisted of the administration of 100 mg of doxycycline (or 5 mg/[kg d] if body weight :::;40 kg) twice a day orally for 45 days, plus 300 mg of netilmicin (6 mg/[kg d] if body weight :::;50 kg) intramuscularly once daily for 7 days. Therapeutic failure was noted in 5 patients (7.7%; 95% confidence interval [CI],.5%-17.1%), of whom had spondylitis, 1 had sacroiliitis, and 1 had a splenic abscess that required splenectomy. Relapse was noted in eight patients (1.5%; 95% CI, 5.6%-3.%). When relapse was considered in combination with initial lack of efficacy, 13 patients (1.9%; 95% CI, 1.3%-33.9%) failed to respond to therapy. Fifteen patients (3%; 95% CI, 13.5%-35.%) had adverse effects, and one patient (1.5%) had a treatment-limiting adverse effect. Combination therapy with netilmicinldoxycycline may be effective in treating acute brucellosis. However, prospective controlled trials must confirm these results. For many years, the standard treatment for acute brucellosis has been combination therapy with tetracycline/streptomycin [1]. Although other antibiotics have been used, no substantial improvement in relapse rates has been reported in association with any new treatment regimen in the past 45 years []. In 1986 the WorId Health Organization recommended the use of a 6-week course of doxycycline plus rifampin in the treatment of human brucellosis [3]. However, a recent meta-analysis of eight controlled trials that included 1,059 patients indicated that therapy with a rifampin!doxycycline combination was associated with more relapses than was that with a streptomycin! tetracycline combination (17% and 5%, respectively) [4]. The search for alternative treatments for human brucellosis has been prompted mainly by the frequency of relapses, the potential toxic effects associated with the use of streptomycin and tetracycline, and the inconvenience of parenteral administration of streptomycin. Clinical experience with other aminoglycoside antibiotics in the treatment of brucellosis has been limited [, 5-7]. Lubani et al. [5] treated 1,100 children with brucellosis. No relapse was seen in 89 children receiving com- Received 11 July 1995; revised 0 September Presented in part at the 7th European Congress of Clinical Microbiology and Infectious Diseases (abstract no. 149, p. 88), held 6-30 March 1995 in Vienna. Informed consent was obtained from patients or their parents, and the guidelines of the Ethical Review Committees of the hospitals of Albacete, Cuenca, and Toledo, Spain, were followed in the conduct of this study. * The members of the GECMEI (Grupo de Estudio de Castilla-la Mancha de Enfermedades Infecciosas) who participated in this study are listed in the Appendix. Reprints or correspondence: Dr. Javier Solera, Unidad de Enfermedades Infecciosas, Hospital General, ClHermanos Falco SIN, 0006 Albacete, Spain. Clinical Infectious Diseases 1996;: by The University of Chicago. All rights reserved /96/ $0.00 binations of gentamicin with doxycycline or oxytetracycline. Intramuscular gentamicin (5 mg/[kg' d] ) was given for 5 days bj.d. with doxycycline or oxytetracycline (the latter two were given for 3, 5, or 8 weeks) [5]. Lubani et al. considered that both streptomycin and gentamicin were effective as antibrucella drugs; however, gentamicin was better because it was given for 5 days only, and the side effects of long treatment with streptomycin could be avoided [5]. It has been suggested that netilmicin is the least toxic aminoglycoside available [8]. Madkour has reported that the MIC of netilmicin against clinical isolates of Brucella species is 0.6 jlg/ml [6]. These properties make it attractive for treating human brucellosis. Moreover, two limited clinical studies have shown the efficacy of netilmicin combined with tetracycline or doxycycline [6, 7]. We conducted a prospective, noncomparative, multicenter study to assess the safety and efficacy ofdoxycycline and netilmicin in the treatment of human brucellosis. To our knowledge, this report describes the largest series of patients for whom a doxycycline and netilmicin combination has been used as therapy for acute brucellosis; it is also the first prospective assessment ofa combination regimen ofdoxycycline with an aminoglycoside other than streptomycin used for treatment of adults with brucellosis. Methods Patients Ambulatory and hospitalized patients ~7 years of age who had brucellosis, as defined below, were eligible for the study. The diagnostic criteria were (1) the isolation of a Brucella species from blood or other fluids or tissues or () the finding of a > 1: 160 standard tube agglutination titer of antibodies to Brucella, in association with compatible clinical findings (fe-

2 44 Solera et al. em 1996; (March) ver, sweats, arthralgias, hepatomegaly, splenomegaly, and/or signs offocal disease). Enrollmentwas limitedto patients without endocarditis or neurobrucellosis. Patient-exclusion criteria were pregnancy or nursing; known or suspected hypersensitivity to or other contraindication for tetracyclines oraminoglycosides; severe concomitant disease; and effective antimicrobial therapy within 7 days before enrollment in the study. Patients could be enrolled in the study only once. Study Design This was a prospective, noncomparative, multicenter study to evaluate the efficacy and safety of doxycycline plus netilmicin in the treatment of acute brucellosis. The patients were recruited from three general hospitals in Spain. The study was approved by the institutional review board at each center, and informed consentwas obtainedfrom all patients. The treatment schedule consisted of the administration of doxycycline (Retens, Wassermann, Barcelona) at a dosage of 100 mg (5 mg! [kg'd] ifbody weight :::;;40 kg) twice a day orally for 45 days, plus netilmicin (Netrocin, Schering Plough, Madrid) at a dosage of 300 mg (6 mg/[kg' d] if body weight :::;; 50 kg) intramuscularly once daily for 7 days. For patients with spondylitis, oral treatment with doxycycline was prolonged for days. Patients could not receive any other antibiotics. Patients were categorized according to the end points defined below. Definition of End Points The main end points of interest were therapeutic failure due to lack of efficacy and relapse of brucellosis. Therapeutic failure due to lack of efficacy was defined by symptoms or signs of the disease that persisted at the end of treatment. A relapse of brucellosis was defined by the reappearance of symptoms or signs of the disease or by new positive blood cultures after therapy. Safety was assessed onthe basis ofall reported adverse events, laboratory tests, and other investigations. Clinical adverse events were recorded and evaluated for severity, outcome, and relation to the study drugs. Clinical and Laboratory Assessment Patients were monitored for therapeutic efficacy and signs of drug toxicity by the obtaining of clinical data; determination of complete blood cell counts (with differential and platelet counts) and erythrocyte sedimentation rates; urinalysis; measurements of the creatinine, aspartate aminotransferase, alkaline phosphatase, bilirubin, and electrolyte levels; Brucella serology; and blood culture. The patients were assessed initially, on day 7, and at the end oftherapy. At these visits, the subjects were asked whether they had missed any dosings. After the end of therapy, patients were reassessed (as outpatients) at months 1,, 3, 6, 9, and 1, as well as whenever clinical symptoms reappeared. Diagnosis of spondylitis, sacroiliitis, or hip arthritis was made by appropriate findings on physical examination and with use of radiological, bone scintigraphic, or magnetic resonance imaging studies. Cochlear and vestibular toxicities were assessed clinically. Serum concentrations of netilmicin were monitored by serum assay between days 4 and 7 of treatment. Samples for determining peak concentrations were taken 1 hour after intramuscular injection, and those for determining trough concentrations were obtained immediately before the drug was administered. Serum netilmicin assays were performed with the Abbott Therapeutic Drug Monitoring System (TDX; Abbott Cientifica, Madrid). Microbiological Studies Standard tube agglutination testing, the Rose-Bengal test, and the Coombs test for antibodies to Brucella species were done according to standard methods [9] with commercial reagents (Knickerbocker, Barcelona). Blood cultures were performed as previously reported [9] and incubated for 30 days; BACTEC NR-730 (Becton Dickinson-Spain, Madrid) was used. All isolates were identified as recommended by Hausler et al. [10]. Twenty-two of the isolated strains were sent to a Reference Center (Laboratorio Regional de Brucelosis, Valladolid, Spain) for confirmation and biotyping. All Brucella isolates were identified as Brucella melitensis. Statistical Analysis Confidence intervals for response and relapse were calculated by use ofthe normal approximation to the binomial with Epi Info Version 6 [11]. Results From October 1993 through November 1994, 65 consecutive patients with brucellosis were enrolled in our study. We excluded 1 patient (1.5%), for whom follow-up data were lacking, 7 days after the beginning of treatment. The characteristics of the remaining 64 patients are summarized in table 1. The median follow-up term was 9 months (range, 1-18 months). Fiftyseven patients (89%) were followed up for at least 6 months. Seven patients were lost to follow-up; therefore, data for these patients were censored at the last visit. Two patients discontinued the treatment after 30 and 37 days, respectively, and another required alternative therapy after 4 weeks because of drug intolerance, but they were monitored nonetheless. A patient was considered to be noncompliant because she took < 80% of the doxycycline doses prescribed. Response to Therapy At the end of the first week of treatment, the proportion of patients whose blood yielded a Brucella species was 18% (six of 33). The mean time to defervescence was 3.9 ±.4 days (range, 1-10 days). Lack of therapeutic efficacy was noted in five of the 64 patients (7.7%; 95% CI,.5%-17.1%). The

3 em 1996; (March) Treatment of Human Brucellosis 443 Table 1. Characteristics of64 patients with brucellosis at enrollment in the study. Patients' characteristics Age (y) Mean ± SD Median (range) No. of males (%) No. (%) with risk factor for brucellosis t Occupational exposure Ingestion of unpasteurized dairy products No. (%) with brucellosis previously Duration of symptoms before therapy: median no. of days (range) Weight (kg) Mean ± SD Median (range) No. (%) with positive blood cultures t Median agglutination titer (range) No. (%) with focal disease Sacroiliitis II Spondylitis Peripheral arthritis Bursitis Orchitis Splenic abscess Follow-up Median no. of months (range) No. (%) of patients followed for: <6mo <1 mo * Values are means ± SDs, medians (ranges), or numbers and percentages, as indicated in column one. t Some patients had more than one risk factor. t Brucella was isolated from the bursal fluid of a patient with prepatellar bursitis. Reciprocal of the standard tube agglutination (STA) titer. II One patient had spondylitis and sacroiliitis simultaneously. clinical characteristics ofthese patients are summarized in table. All but one had focal disease at enrollment. Among these, had spondylitis with paravertebral abscess (1 of them had concomitant sacroiliitis), I had sacroiliitis, and 1 had a splenic abscess. The two patients who had spondylitis and the patient who had sacroi1iitis continued to have prominent back pain at the end of treatment. The patient with sacroiliitis had discontinued treatment prematurely after 37 days. Another patient had positive blood cultures and arthralgias after having completed45 days oftherapy. All five patients whose therapy failed (because of lack of efficacy) received maintenance treatment with doxycycline or cotrimoxazole over longer periods (3-8 months). A patient with a splenic abscess required surgery after 90 days of therapy. Overall, the long-term clinical response was favorable. None of these five patients have relapsed since the end of treatment. Relapse Data* 34.7 ± (7-73) 46 (7) 44 (69) 9 (45) 7 (11) 30 (4-365) 67. ± (8-16) 45 (70) 640 (0-0, 480) (34) 1 9 (1-18) 7 (11) 0 (31) Among the 64 patients included in this study, 8 have had a relapse (1.5%; 95% CI, 5.6%-3.%). Of these 8 patients, 5 4 I had a clinical relapse with characteristic clinical findings, and 4 of them had associated brucella bacteremia. Two other patients had only mild clinical symptoms despite new positive bloodcultures. Inone patientbacterial relapse occurredwithout clinical signs or symptoms. A patient with orchitis had focal disease that relapsed in the same location. Another patient had spondylitis in relapse. The clinical characteristics of these patients are summarized in table. All relapses were treated with the previously used antibiotic schedule. Clinical response was excellent in all but one patient. One patient (1.5%) had positive blood cultures after 30 days oftherapy, and treatment was maintained for a total of90 days. When relapse was considered together with initial lack of efficacy, 13 patients (1.9%; 95% CI, 1.3%-33.9%) failed to respond to therapy. Adverse Events Fifteen patients (3%; 95% CI, 13.5%-35.%) had adverse effects: 7, phototoxicity; 4, epigastric discomfort; 3, nausea;, vomiting; I, vertigo; 1, dizziness; 1, gray-brown discoloration of the teeth; and 1, abscess at the injection site. Most of these reactions were classified as mild, and only one patient (1.5%) had an episode ofa treatment-limiting adverse effect. Discoloration of teeth led to discontinuation of doxycycline therapy after 8 days for this 7-year-old patient. The most frequently observed adverse effects were photosensitivity and gastrointestinal complaints, both considered to be related to doxycycline. The mean serum creatinine concentrations at baseline and after 7 days of netilmicin treatment were 0.95 ± 0.16 mg/dl and 0.93 ± 0.15 mg/dl, respectively (P >.). The magnitude of the change is too small to have clinical relevance, but it does demonstrate that there was no trend toward impairment ofrenal function. Plasma concentrations of netilmicin were measured in 1 patients. The mean peak and trough serum netilmicin levels were 11.6 ± 3.8 mg/l (range, mg/l) and 0.6 ± 0.1 mg/l (range, mg/l), respectively. Discussion The key variables in the assessment of treatment regimens ofbrucellosis are the rate of therapeutic failure and the rate of relapse []. The therapeutic failure rate of 7.7% in this study was similar to the rates in our two previous trials [9, 1]. Other studies ofpatients with brucellosis who received treatment with streptomycin for 14-1 days and doxycycline for 4-6 weeks yielded therapeutic failure rates of zero to 3% [13-17]. We considered that the high proportion of initial lack of efficacy in this trial as related to that in other studies may have been due to the high number of patients with focal disease (34%). Our 18% rate of therapeutic failure among patients with focal disease was also similar to rates found in other studies [18]. On the other hand, the 1.5% relapse rate in this study was slightly higher than in previous studies in which streptomycin and doxycycline were given for 14-1 days and days, respectively [9, 1-18]. The higher rates in this study could

4 444 Solera et al. cm 1996; (March) Table. Characteristics of 13 patients with acute brucellosis whose therapy failed or who relapsed. Characteristics before therapy Duration (d) Blood Focal disease Time Patient Age Occupational of culture STA (diagnostic (mo) to no. (y)/sex risk symptoms result titer* procedure) Outcome relapse Comment 71/F Spondylitis, L4-5 Therapy failed Persistent back paint (MRI) 45/F Spondylitis, L3-4; Therapy failed Persistent back paint sacroiliitis (MRI) 3 49/F Splenic abscess Therapy failed Fever, abdominal pain; (CT, MRI) splenectomy 4 391M ,80 Sacroiliitis (MRI) Therapy failed Dox withdrawn at 37 days; persistent buttock paint 5 61M Orchitis Therapy failed Knee arthralgia, positive (ultrasonography) blood cultures after 45 days of treatment 6 7IM 4 + 5,10 None Relapse (B) 7 Dox withdrawn because of side effects after 30 d; TMP-SMZ given for 15 d 7 31M None Relapse (B/C) I Fever, sweats, arthralgias 8 161M ,10 None Relapse (B) 11 By mistake, Dox withdrawn after 30 d 9 31M None Relapse (C/B) 6 Fever, sweats, arthralgias 10 47IM Orchitis Relapse (C) 0.3 Fever, sweats, orchitis (ultrasonography) 11 13/F None Relapse (B) 6 Treatment compliance 80% 1 61M Sacroiliitis (MRI) Relapse (C/B) 11 Fever, sweats, arthromyalgia 13 70IM ,80 None Relapse (C/B) 5 Fever, lumbar pain, spondylitis at L5-S1 (MRI)t NOTE. B = bacterial; C = clinical (see text with regard to clinical relapse); Dox = doxycycline; TMP-SMZ = trimethoprim-sulfamethoxazole; - = negative; + = positive. * Reciprocal of the standard tube agglutination (STA) titer. t Pain was severe; limited movement and spasms in the surrounding muscles. Functional capacity was inadequate to perform the duties of usual occupation (patients and 4) or of self-care (patients 1 and 13). be accounted for by several factors. First, our study differed from most other studies of treatment for brucellosis in that bloodcultures were performedsystematicallyat eachfollow-up visit. Second, reinfection with new strains ofbrucella species is plausible in an area where brucellosis is endemic, such as Spain. Third, positivity of blood cultures before therapy has been identified as an independent predictor ofrelapse [19]. In our study there was a high proportion ofpositive blood cultures at baseline (70%) as compared with the proportion in other studies [5, 14-16]. This fact could have influenced the relapse rate. Fourth, compliance withtreatmentmayhave beensuboptimal, even though drug compliance was emphasized in interviews conducted during the treatment period. In this respect, a patient whose therapy failed and two patients who relapsed were deemed noncompliant (table ). Finally, the treatment regimen- with administration ofdoxycycline for 45 days plus netilmicin on the first 7 days-could be slightly less effective than regimens used in previous studies (involving administration of doxycycline for 45 days plus streptomycin on the first 14 days). As in other studies ofcombination therapy with doxycycline/ streptomycin [9, 1-18], in our trial severe toxic effects were uncommon and most adverse effects were considered to be related to doxycycline. The greatest concern associated with netilmicin treatment is nephrotoxicity [8]. In this study there was no renal toxicity because of the strict exclusion criteria, the relatively young age ofthe patients, and the short duration of treatment. According to previous reports [0, 1], only the height ofthe trough levels seem to be a risk factor for aminoglycoside toxicity. In our study those levels were < mgll, as has been recommended by Parker et al. []. Administration of a tetracycline/streptomycin combination is currentlyconsideredthe treatment ofchoice for acute brucellosis [9, 1-17]. However, streptomycin must be administered by the parenteral route for 14-1 days. Ototoxicity, vestibular and auditory dysfunction, nephrotoxicity, and hypersensitivity reactions can follow the administration of streptomycin [0]. An important advantage oftherapy with this netilmicin/doxycycline regimen is that intramuscular injections are given for 7 days. Although netilmicin [6] and gentamicin [5] have been

5 cm 1996; (March) Treatment of Human Brucellosis 445 administered parenterally in a twice-daily regimen, we gave netilmicin as a single daily intramuscular injection. This mode oftherapy was chosen because against other infectious diseases it has been at least as effective and safe as multiple daily doses and because it may lead to more efficient use of this drug [8, 1, ]. With administration of 6 mg/(kg' d) in intramuscular once-daily doses, netilmicin reached peak concentrations that were ;:::9 times the MIC of B. melitensis [6]. It is possible that a longer netilmicin regimen could decrease relapse and failure rates. However, neither experimental nor clinical data are available about the optimal treatment duration. Although prolonged courses of netilmicin injection have been well tolerated [6], it is particularly important that patients treated for longer than usual periods be carefully monitored for changes in renal, auditory, and vestibular functions, a necessity that could increase the cost of care [0]. In any event, it is desirable to limit the duration oftreatment with aminoglycosides to a short term whenever feasible. The role of a netilmicin/doxycycline combination in the treatment of acute brucellosis requires prospective controlled trials. Although our findings support the notion that treatment with netilmicin plus doxycycline is safe and effective, this combination does not seem to be more effective than that of streptomycin and doxycycline [4]. The advantage of giving intramuscular injections of netilmicin in only the first 7 days of therapy, as compared with 14-1 days of therapy with streptomycin, must be weighed against the greater cost ofnetilmlcm. Acknowledgments The authors are indebted to Belen de la Hoz for editorial assistance; to Emilio Serna, R.N., Maria Luisa Castillejos, R.N., and Dolores Cuenca, R.N., ofthe Unit ofinfectious Diseases, Albacete General Hospital; to the many staff members of the clinical microbiology laboratories and of the division of internal medicine services affiliated with the GECMEI who collected the study data; and to the patients who volunteered to assist in this effort. Appendix The following persons and institutions participated in this trial, enrolled research subjects, or both. Albacete Hospital: Francisco Medrano; Antonio Alamillo; Miguel Angel Barba; Angel Fernandez; Fernando Martinez-Salazar; Jose Antonio Saez-Barcelona; Miguel Torralba; Diego Cebrian; Anunciacion Perez-Tello; and Maria Isabel Serrano. Cuenca Hospital: Jaime Calderon; Maria Enriqueta Peiro; Juan Ruiz-Gonzalez; and Alberto Roldan. Toledo Hospital: Jose Largo and Fernando Cuadra. References 1. Young EJ. Brucella species. In: Mandell GL, Bennett JE, Dolin R, eds. Principles and practice of infectious diseases. 4th ed. New York: Churchill Livingstone, 1995: Hall WHo Modern chemotherapy for brucellosis in humans. Rev Infect Dis 1990; 1: Joint Food and Agriculture Organization, World Health Organization. FAO-WHO Expert Committee on Brucellosis. 6th report. WHO technical report series, no Geneva: World Health Organization, 1986: Solera J, Martinez-Alfaro E, Saez L, Gil A, del Rey Calero J. Treatment of human brucellosis with doxycycline plus rifampin or doxycycline plus streptomycin. A meta-analysis of randomized controlled trials [abstract no. M74]. In: Program and abstracts of the 34th Interscience Conference on Antimicrobial Agents and Chemotherapy (Orlando, FL). Washington, DC: American Society for Microbiology, 1994: Lubani MM, Dudin Kl, Sharda DC, et al. A multicenter therapeutic study of 1100 children with brucellosis. Pediatr Infect Dis J 1989;8: Madkour MM. Treatment. In: Madkour MM, ed. Brucellosis. London: Butterworths, 1989: Gotuzzo E, Carrillo C, Cellillo C. Brucella. In: Gorbach SL, Bartlett JG, Blacklow NR, eds. Infectious diseases. Philadelphia: WB Saunders, 199; Kumana CR, Yuen KY. Parenteral aminoglycoside therapy. Selection, administration and monitoring. Drugs 1994;47: Solera J, Medrano F, Rodriguez M, Geijo P, Paulino J. Ensayo terapeutico comparativo y multicentrico de rifampicina y doxiciclina frente a estreptomicina y doxiciclina en la brucelosis humana. Med Clin (Barc) 1991; 96: Hausler WJ Jr, Moyer NP, Holcomb LA. Brucella. In: Lennette EH, Balows A, Hausler WJ Jr, Shadomy HJ, eds. Manual ofclinical microbiology. 4th ed. Washington, DC: American Society for Microbiology, 1984: Dean AG, Dean JA, Coulombier D, et al. Epi Info, version 6: a word processing, database, and statistics program for epidemiology on microcomputers. Atlanta: Centers for Disease Control and Prevention, Solera J, Rodriguez-Zapata M, Geijo P, et al. Doxycycline-rifampin versus doxycycline-streptomycin in treatment ofhuman brucellosis due to Brucella melitensis. Antimicrob Agents Chemother 1995; 39: Ariza J, Gudiol F, Pallares R, Rufi G, Fernandez-Viladrich P. Comparative trial of rifampin-doxycycline versus tetracycline-streptomycin in the therapy of human brucellosis. Antimicrob Agents Chemother 1985;8: Acocella G, Bertrand A, Beytout J, et al. Comparison of three different regimens in the treatment of acute brucellosis: a multicenter multinational study. J Antimicrob Chemother 1989;3: Colmenero Castillo JD, Hernandez Marquez S, Reguera Iglesias JM, Cabrera F, Ruiz Diaz F, Alonso A. Comparative trial of doxycycline plus streptomycin versus doxycycline plus rifampin for the therapy ofhuman brucellosis. Chemotherapy 1989; 35: Cisneros JM, Viciana P, Colmenero J, Pachon J, Martinez C, Alarcon A. Multicenter prospective study of treatment of Brucella melitensis brucellosis with doxycycline for 6 weeks and streptomycin for weeks. Antimicrob Agents Chemother 1990;34: Montejo JM, Alberola I, Glez-Zarate P, et al. Open, randomized therapeutic trial ofsix antimicrobial regimens in the treatment ofhuman brucellosis. Clin Infect Dis 1993; 16: Ariza J, Gudiol F, Pallares R, et al. Treatment of human brucellosis with doxycycline plus rifampin or doxycycline plus streptomycin: a randomized, double-blind study. Ann Intern Med 199; 117: Ariza J, Corredoira J, Pallares R, et al. Characteristics of and risk factors for relapse of brucellosis in humans. Clin Infect Dis 1995;0: Gilbert DN. Aminoglycosides. In: Mandell GL, Bennett JE, Dolin R, eds. In: Mandell, Douglas and Bennett's principles and practice ofinfectious diseases. 4th ed. New York: Churchill Livingstone, 1995: de Vries PJ, Verkooyen RP, Leguit P, Verbrugh HA. Prospective randomized study of once-daily versus thrice-daily netilmicin regimens in patients with intraabdominal infections. Eur J Clin Microbiol Infect Dis 1990;9: Parker SE, Davey PG. Practicalities of once-daily aminoglycoside dosing. J Antimicrob Chemother 1993; 31 :4-8.