AMERICAN JOURNAL OF EPIDEMIOLOGY Copyright 97 by The Johns Hopkins University Vol, Printed in USA. CANINE BRUCELLOSIS IN FLORIDA: SEROLOGIC SURVEY OF POUND DOGS, ANIMAL SHELTER WORKERS AND VETERINARIANS GERALD L. HOFF' AND JAMES B. NICHOLS' (Received for publication February, 97) Hoff. G. L. (Veterinary Public Health Section, Florida Division of Health, P. O. Box, Jacksonville, Fla. ) and J. B. Nichols. Canine brucellosis in Florida: serologic survey of pound dogs, animal shelter workers and veterinarians. Am J Epidemiol : 5-9. A serologic survey for agglutinins to Brucella cam's was conducted with 7 dogs in pounds, 67 animal shelter workers and veterinarians. Dogs and animal shelter workers were surveyed from establishments in 6 counties. Ten dogs (.65%) and one worker (.59%) possessed agglutinin titers of : or greater to B. canis. None of the veterinarians had titers to B. canis. agglutinins; Brucella canis; brucellosis; zoonoses The economic importance of Brucella canis as an etiologic agent of abortions, embryonic death and poor reproductive performance in breeding kennel dogs has been appreciated only in the last two decades (, ). The disease of canine brucellosis has been recognized in many breeds of dogs from at least 8 states (, ). Transmission is believed to be by direct contact with expelled tissues or vaginal discharge from a recently aborted female (5). Venereal transmission may also occur. The bacteria have been isolated from dog urine but the significance of this observation to disease transmission remains undetermined (6). Congenital infection with Veterinary Public Health Section, Bureau of Preventable Diseases, Florida Division of Health, P.. Box, Jacksonville, Florida. (Address for reprints.) "Formerly, Veterinary Public Health Section, Bureau of Preventable Diseases, Florida Division of Health. The cooperation of the various animal pound and shelter personnel and participating veterinarians is gratefully acknowledged. Gratitude is also extended to Dr. Wayne Hemmert, Assistant State Epidemiologist, and the personnel of the health departments of the counties involved for their assistance in obtaining human sera. Special thanks are given to Dr. G. M. Brown, USDA, for providing the antigen and reference antisera and for performing confirmatory testing. 5 B. canis has been reported and the organism has been isolated from milk of infected females (5). The infection rate in male and female dogs is essentially equal (, 7). Most infected dogs are free of clinical signs, though many suffer reproductive failures and loss of vigor. Infected dogs are afebrile and commonly have prolonged bacteriemias of to months; the bacteria is associated with the leukocytes (). B. canis can persist in various tissues for long periods of time despite antibiotic treatment (, 8). Lymph nodes are the most consistent source for isolation of B. canis from abacteriemic dogs (8). In dogs, the disease is characterized by generalized lymphadenitis and splenitis, early embryonic deaths or abortion (usually undetectable, overt abortions at approximately 5 days of gestation) and prolonged vaginal discharge following abortion (). In male dogs epididymitis, dermatitis of the scrotum and testicular atrophy, often unilateral, are common and sterility may develop (6). While the disease is easily recognizable and controllable in kennel situations (9), little is known concerning B. canis infec- Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on March, 6
6 HOFF AND NICHOLS tion in stray or pet dogs. A study conducted at the University of Minnesota Medical Center found no evidence of infection in 5 mixed breed dogs (). Agglutinin titers of : or greater have been reported, however, in 6 of (9.9 per cent) hunting dogs and in 7 of 5 (.6 per cent) mixed breed dogs in Alabama and Georgia (). More recently, it has been recognized that canine brucellosis is a zoonotic disease. Eleven cases of B. canis infection in man have been reported although six of these cases were in laboratory personnel working with the organism (, -). Of the other five cases, one was in an animal caretaker and the remaining were presumably acquired from pet dogs including one case from Jacksonville, Florida, in 97. In three cases, the pet dogs were bacteriemic or seropositive to B. canis although the mode of transmission from dog to owner in these cases was not determined. In addition, a survey of 8 military recruits revealed that five (. per cent) men had agglutinins to B. canis, however, medical histories were not available to determine if any of the five had experienced clinical B. canis infection (5). Canine brucellosis in man may be asymptomatic or cause overt disease. Symptoms have included lowgrade fever, malaise, fatigue, headache, night sweats, chills, lymphadenopathy, splenomegaly, generalized macular rash, chest pains, pharyngitis, thrombophlebitis, pain in legs and arms, and abdominal abscess (6). In order to ascertain the relative importance of canine brucellosis as a zoonotic disease, the Veterinary Public Health Section of the Florida Division of Health instituted a serologic survey for B. canis agglutinins in pound dogs, animal shelter employees and veterinarians. The results of that survey are summarized in the present report. MATERIALS AND METHODS Through the cooperation of various county or city animal pounds and shelters, dogs were made available for this study; for a variety of reasons, only dogs scheduled for euthanasia were used. Three to 5 ml of blood were drawn from the cephalic vein. Blood samples from personnel working at the animal shelters and from government and private veterinarians were also obtained. The blood samples were refrigerated overnight and the serum was removed after centrifugation for 5 min at 5 rpm. Using a serologic procedure developed by the United States Department of Agriculture (USDA), Biologic Reagents Section, Veterinary Services Diagnostic Laboratory, Ames, Iowa, non-hemolytic sera were initially screened at a :5 dilution for reactivity to B. canis. This dilution was achieved by placing. ml of serum in a - x -mm disposable tissue culture tube and then adding. ml of a. M.5 per cent sodium chloride -mercaptoethanol solution. The -mercaptoethanol was used to reduce heterospecific agglutinin responses of the 9S type. To each tube was added. ml of B. canis (serial lot #6) antigen, supplied by the USDA. Positive and negative reference antisera were also supplied and employed as controls. The tubes were then incubated (hot air) at 7C for 8 ± hr. Because of the mucoid nature of the bacterium, agglutination is not firm and therefore the results of the test were read by observing the clearing of the supernatant as compared to the controls. Based on the amount of clearing, the reactions were recorded as complete, incomplete or negative. All sera which gave a complete or incomplete reaction were then titrated in a twofold dilution series starting with :5. Only those serum samples which had complete reactions at : or greater were considered positive. Selected samples having no partial or complete reactions to B. canis were forwarded to the USDA for confirmation. RESULTS Dogs and animal workers were bled at shelters in 6 of the more populated coun- Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on March, 6
CANINE BRUCELLOSIS IN FLORIDA ties of Florida (figure ). The vast majority of the dogs were of mixed breeds; 7 animals were tested of which 9 were male and 55 were female. Ten dogs, four males and six females, were considered positive for agglutinins to B. canis for an overall agglutinin prevalence of.65 per cent (tables and ). Four of the positive dogs were from County, two from Dade County, and one from each of Broward, Escambia, Orange and Palm Beach Counties. The agglutinin prevalence rates in the FIGURE. Florida counties included in serologic survey for Brucella canis in pound dogs. Dots indicate counties where all dogs were serologically negative. Lines indicate counties with at least one dog having an agglutination titer of : or greater. The number of positive dogs in these counties is indicated. dogs tested from these six counties ranged from.7 to 9.5 per cent. A total of 67 dog shelter personnel were surveyed for agglutinins to B. canis. There were 6 men and 5 women from whom blood samples were obtained. Ten percent of these people ( men and women) had duties which did not involve the handling or maintenance of the dogs. Forty-four per cent of the people had less than one year of experience working with dogs either on their present job or in previous employment. However, 8 of the 67 (76.65 per cent) of the people sampled currently owned a dog (average. dogs per person). Only one individual was found to have agglutinins against B. canis (table ). This TABLE Sex and age distribution of pound dogs tested for agglutinins to Brucella canis Age (years)* 5+ Undetermined! Total tested 5 5 9 s positive tested 5 5 6 5 55 s positive * Estimated from tooth wear. t Undetermined age means adult dogs but no specific age. TABLE Demographic information on ten Florida dogs with agglutination titers > :) to Brucella canis Dog CB-7 CB-6 CB-67 CB-69 CB-96 CB- CB- CB- CB-6 CB- Sex Age (year)* * Estimated from tooth wear. t Mixed breed, with dominant identifiable breed of the mixture. Breed Miied-colliet German shepherd Dalmatian Beagle Shetland sheepdog Mixed-hound County Orange Broward Dade Dade Palm Beach Escambia 6 Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on March, 6
8 HOFF AND NICHOLS TABLE Number of animal shelter personnel tested for agglutinins to Brucella canis, by primary job duty Administration Clerical Kennel worker Animal warden Total tested 77 7 67 Positive was a 5 year-old male animal warden from Brevard County who previously had worked in the kennel operation of that shelter. He had been employed three years and had no prior job experience with dogs. At the time of the survey, he owned two terriers, a male and a female. These pet dogs were not tested but the female had never had a litter or an abortion. The man had no medical history of an illness suggestive of brucellosis and he had no antibodies to other Brucella sp. Four government veterinarians and 9 veterinarians in private practice were surveyed. Of those in private practice, worked with only small animals while nine had mixed large and small animal practices. Agglutinins to B. canis were not detected in any of the veterinarians, although six had titers to other Brucella sp. (7). DISCUSSION The association of human brucellosis and Brucella sp. infections in dogs has been recognized by many investigators (8, 9) but the etiologic role of B. canis has been appreciated only recently. Transmission of this organism from infected dogs to people appears to occur infrequently (5, 6). The results of the present survey showing a.59 per cent agglutinin rate in animal shelter employees are in agreement with these earlier observations. The fact that none of the veterinarians surveyed had agglutinins to B. canis may be a reflection of sample size and/or the type of exposure they have to infected animals as opposed to animal shelter personnel. As the significance of agglutination titers in dogs of less than : is uncertain (, 5), the results of the present study are probably a conservative estimate of B. canis activity in pound dogs in Florida. The.65 per cent agglutinin prevalence rate is in agreement with a. per cent rate observed in a study in Alabama and Georgia when only animals with titers of : or greater are considered (). Since the animal shelters involved in this study have a combined yearly dog turnover rate of approximately, animals, the animal shelter workers are potentially exposed to 75 infected dogs yearly. How many of these animals would be capable of transmitting B. canis is unknown. One would expect a higher agglutinin prevalence rate in the shelter personnel if the disease is readily transmissible from dogs to people in pound situations. However, because all studied cases of B. canis infection in man have had high agglutination titers which subsequently became negative within a few months after treatment or recovery, prevalence surveys, such as in military recruits (5) and the present investigation, may only indicate the presence of infection at the time of the survey and may not be a good retrospective epidemiologic investigative method. Further study is required to interpret the clinical significance of agglutinin levels in man and evaluate the use of serologic surveys. REFERENCES. Morisset R, Spink WW: Epidemic canine brucellosis due to a new species, Brucella canis. Lancet :-, 969. Hall WH: Epidemic brucellosis in beagles. J Infect Dis :65-67, 97. Carmichael LE, Kenney RM: Canine abortion caused by Brucella canis. J Am Vet Med Assoc 5:65-66, 968. Lewis GE: A serological survey of 65 dogs to detect titers for Brucella canis (Brucella suis, Type 5). J Am Animal Hosp Assoc 8:-7, 97 Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on March, 6
CANINE BRUCELLOSIS IN FLORIDA 9 5. Moore JA: Brucella canis infection in dogs. J Am Vet Med Assoc 55:-7, 969 6. Moore JA, Kakuk TJ: dogs naturally infected with Brucella canis. J Am Vet Med Assoc 55:5-58, 969 7. Moore JA, Gupta BN, Conner GH: Eradication of Brucella canis infection from a dog colony. J Am Vet Med Assoc 5:5-57, 968 8. Lewis GE, Crumrine MH, Jennings PB, et al: Therapeutic value of tetracycline and ampicillin in dogs infected with Brucella canis. J Am Vet Med Assoc 6:9-, 97 9. Pickerill PA, Carmichael LE: Canine brucellosis: control programs in commerical kennels and effect on reproduction. J Am Vet Med Assoc 6:67-65, 97. Center for Disease Control: Annual Brucellosis Summary, 968. Center for Disease Control: Annual Brucellosis Summary, 969. Ewin WW Jr: Personal communication. Tennessee State Health Department, Nashville, Tennessee, 97. Ginter M: Personal communication. Consolidated City of Jacksonville Department of Public Health, Jacksonville, Florida, 97. Swenson RM, Carmichael LE, Cundy KR: Human infection with Brucella canis. Ann Lntem Med 76:5-8, 97 5. Lewis GE, Anderson JK: The incidence of Brucella canis antibodies in sera of military recruits. Am J Public Hlth 6:-5, 97 6. Center for Disease Control: Annual Brucellosis Summary, 97 7. Hoff GL, Nichols JB: Results of serologic survey for canine brucellosis in Florida veterinarians. Fla Vet J :, 97 8. Nicoletti PL, Quinn BR, Minor PW: Canine to human transmission of brucellosis. NY State J Med 67:886-887, 967 9. Robertson MG: Brucella infection transmitted by dog bite. JAMA 5:75-75, 97 Downloaded from http://aje.oxfordjournals.org/ at Pennsylvania State University on March, 6