Failure of short antimicrobial treatments for human brucellosis

We read with interest the recent article by Solera et al., inwhich the safety and efcacy of two treatment protocols forpatients with brucellosis, mostly adults, were compared (5).The results of the study showed that the relapse rate of patientsreceiving a 30-day treatment with oral doxycycline, combinedwith a single daily dose of intramuscular gentamicin, was ashigh as 22.9%.Because of the inconveniences of prolonged antimicrobialtherapy, the potential toxicity of prolonged tetracycline expo-sure, and the need for parenteral administration of aminogly-cosides, shorter antimicrobial treatments for pediatric brucel-losis are required. In 1989, Lubani et al. reported that norelapses among 71 children receiving gentamicin twice daily for5 days combined with either doxycycline, oxytetracycline, ri-fampin or co-trimoxazole for 3 weeks were observed, suggest-ing that short antimicrobial regimens can also be efcacious(2).We have recently evaluated the efcacy of treating childrenwith brucellosis with 5 mg of gentamicin per kg of body weightper day (maximum 300 mg) given intramuscularly as a singledaily dose for 5 days, combined with (i) 3-week therapy ofdoxycycline (5 mg/kg/day in two divided doses, maximum 200mg) in children older than 8 years of age or (ii) co-trimoxazole(10 and 50 mg/kg/day in two divided doses) in children youngerthan 8 years. Diagnostic criteria included a positive blood cul-ture with the BACTEC 9240 Peds Plus medium (Becton D only 1 of these 2 children was symptomatic. The fail-ure rate of both regimens combined was then 33.3% (95%condence interval [CI] 9.5 to 57.2%).The failure rate found in our study is higher than that foundby Solera et al. after 30 days of doxycycline and is substantiallydifferent from the experience reported by Lubani et al. (2). Itshould be pointed out that the use of sensitive blood culturemethods resulted in the detection of relapse in three of ourpatients and treatment failure in another that would have goneunnoticed by clinical or serologic surveillance. It is then pos-sible that a fraction of patients considered to be treatmentsuccesses in other studies in which less sensitive bacteriologicalmethods were used in fact failed to eradicate the organism (3,4).Nowadays, treatment strategies based on short antibioticcourses are being developed for infectious diseases to reducecosts and inconveniences of prolonged hospitalizations (1).The ultimate goal, however, is to achieve cure rates compara-ble to those observed with traditional therapeutic modalities(1). The preliminary results of our study, as well as those ofSolera et al., preclude the recommendation of gentamicin as asingle daily dose combined with short courses of either doxy-cycline or co-trimoxazole for adult and pediatric patients withbrucellosis. Furthermore, when treatment protocols for bru-cellosis are evaluated, sensitive blood culture methods shouldbe used, since persistent bacteremia can occur in asymptomaticindividuals.