Serosurvey of Dogs for Human, Livestock, and Wildlife Pathogens, Uganda

To the Editor: Domestic dogs live in close association with humans and livestock, participating in the transmission of diseases of zoonotic, veterinary, and conservation interest (1,2). Most households in Uganda have traditionally kept dogs for hunting and for help with herding, security, and guarding livestock. Most dogs receive no prophylactic measures (e.g., vaccinations) and roam freely; this situation exposes them to pathogens from eating garbage, rodents, and stillborn animals and other carcasses and through inhalation during scent communication. Thus, dogs are a reservoir for certain pathogens and a useful sentinel for others (3). 
 
In 2011, serum samples were obtained from 116 mixed-breed dogs during a rabies vaccination campaign in and near 3 national parks in southwestern Uganda; the dogs were >4 months of age and were voluntarily brought in by their owners (Figure, Appendix, Table). Two of the parks, Bwindi Impenetrable (BI) and Mgahinga Gorilla (MG), have some of the most biologically diverse tropical forests in eastern Africa and are home to mountain gorillas. The third park, Queen Elizabeth (QE), is home to populations of protected carnivores and ungulates. The parks lie within a densely populated rural landscape; in some areas, the population is as high as 500 persons/km2. 
 
 
 
Figure 
 
Map of Uganda showing 3 areas where a serosurvery for human and animal pathogens was conducted among dogs. 1, Queen Elizabeth National Park; 2, Bwindi Impenetrable National Park; 3, Mgahinga Gorilla National Park. 
 
 
 
 
 
Table 
 
Methodology and seroprevalence for selected pathogens in rural dogs in 3 national parks, Uganda, 2011* 
 
 
 
Of the 116 sampled dogs, 4 had been vaccinated against rabies by the authors in 2010 in QE (not included in rabies results), and 11 (all males) had been castrated by local animal healers before serum samples were obtained. The samples were used to test for seroprevalence rates to rabies virus (RABV), canine distemper virus (CDV), canine parvovirus (CPV), Leptospira interrogans, Leishmania sp., Toxoplasma gondii, and Neospora caninum (Table). Seroprevalence rates ranged from 20% to 100% (Table). CPV seroprevalence was higher in BI and QE than in MG (χ2 >12.6, p<0.001); T. gondii seroprevalence was higher in BI than in MG (Fisher p = 0.002); and RABV seroprevalence was higher in castrated than noncastrated dogs (50% vs. 10%; Fisher p = 0.005). 
 
For humans, the domestic dog is the main source of exposure to RABV. The possibility that the presence of the rabies titers in the dog serum samples was due to a previous vaccination can be ruled out because the only previous recent campaign in the area was conducted by the authors. Antibodies against RABV in apparently healthy dogs have been reported in Africa (6), and rabies seems to be not invariably fatal in dogs. Dogs that have recovered from a rabies infection are prone to shed RABV in their saliva for long periods (7). Antibodies against RABV were more frequently found in castrated dogs. This finding may be due to an increase in virus-related deaths among noncastrated dogs; such dogs tend to be more aggressive and to roam, so they may come more frequently into contact with pathogenic RABV strains. 
 
Results indicate that both CDV and CPV are actively circulating in the studied dog populations. High CDV seroprevalence rates have been reported among other rural dog populations in Africa (8). Sick, debilitated pups are at high risk for predation by wild carnivores, so spillover may take place. A dog population exhibiting similar characteristics to the population we studied was believed to be the origin of the 1994 CDV epidemic among Serengeti wildlife (8). Furthermore, carnivores use feces for scent communication, so the probability of infection by CPV in wild carnivores in the study area may also be high. 
 
In developing countries, leptospirosis is emerging as a major public health problem and also causes enormous economic losses because of disease in livestock (9). The most commonly detected serovars in this study were those that have rats and dogs as reservoirs (Table​(Table).). Visceral leishmaniasis in humans is also a major health problem in several areas of eastern Africa, where the number of cases has dramatically increased during the past 20 years. Transmission of Leishmania donovani in eastern Africa may take place through anthroponotic or zoonotic cycles, although, to our knowledge, no reservoir host had been identified (10). 
 
The mean T. gondii seroprevalence detected during this survey appears to be the highest reported for dogs worldwide. This protozoon has implications for human and animal health, and dogs, who probably become infected with T. gondii when eating raw meat, are a good sentinel for environmental contamination by this parasite. On the other hand, dogs serve as the definitive host for N. caninum, which is a major cause of abortions in cattle and causes economic losses wherever it is enzootic. 
 
Some of these diseases may also have implications for the conservation of endangered mountain gorillas. Diseases such as leptospirosis, toxoplasmosis, and especially, rabies could be fatal for gorillas, and there are unpublished reports of fights between hunting dogs and gorillas. 
 
Our work should serve as a first step toward the establishment of preventive strategies for improvements in the health of humans and domestic animals living in rural Uganda and for the health of the country’s unique wildlife. Tracing the role of dogs in the cycle of the studied pathogens is crucial for the design of control programs.

To the Editor: Domestic dogs live in close association with humans and livestock, participating in the transmission of diseases of zoonotic, veterinary, and conservation interest (1,2). Most households in Uganda have traditionally kept dogs for hunting and for help with herding, security, and guarding livestock. Most dogs receive no prophylactic measures (e.g., vaccinations) and roam freely; this situation exposes them to pathogens from eating garbage, rodents, and stillborn animals and other carcasses and through inhalation during scent communication. Thus, dogs are a reservoir for certain pathogens and a useful sentinel for others (3).
In 2011, serum samples were obtained from 116 mixed-breed dogs during a rabies vaccination campaign in and near 3 national parks in southwestern Uganda; the dogs were >4 months of age and were voluntarily brought in by their owners (Figure, Appendix, wwwnc.cdc.gov/EID/ article/19/4/12-1143-F1.htm; Table). Two of the parks, Bwindi Impenetrable (BI) and Mgahinga Gorilla (MG), have some of the most biologically diverse tropical forests in eastern Africa and are home to mountain gorillas. The third park, Queen Elizabeth (QE), is home to populations of protected carnivores and ungulates. The parks lie within a densely populated rural landscape; in some areas, the population is as high as 500 persons/km 2 .
Of the 116 sampled dogs, 4 had been vaccinated against rabies by the authors in 2010 in QE (not included in rabies results), and 11 (all males) had been castrated by local animal healers before serum samples were obtained. The samples were used to test for seroprevalence rates to rabies virus (RABV), canine distemper virus (CDV), canine parvovirus (CPV), Leptospira interrogans, Leishmania sp., Toxoplasma gondii, and Neospora caninum (Table). Seroprevalence rates ranged from 20% to 100% (Table). CPV seroprevalence was higher in BI and QE than in MG (χ 2 >12.6, p<0.001); T. gondii seroprevalence was higher in BI than in MG (Fisher p = 0.002); and RABV seroprevalence was higher in castrated than noncastrated dogs (50% vs. 10%; Fisher p = 0.005).
For humans, the domestic dog is the main source of exposure to RABV. The possibility that the presence of the rabies titers in the dog serum samples was due to a previous vaccination can be ruled out because the only previous recent campaign in the area was conducted by the authors. Antibodies against RABV in apparently healthy dogs have been reported in Africa (6), and rabies seems to be not invariably fatal in dogs. Dogs that have recovered from a rabies infection are prone to shed RABV in their saliva for long periods (7). Antibodies against RABV were more frequently found in castrated dogs. This finding may be due to an increase in virus-related deaths among noncastrated dogs; such dogs tend to be more aggressive and to roam, so they may come more frequently into contact with pathogenic RABV strains.
Results indicate that both CDV and CPV are actively circulating in the studied dog populations. High CDV seroprevalence rates have been reported among other rural dog populations in Africa (8). Sick, debilitated pups are at high risk for predation by wild carnivores, so spillover may take place. A dog population exhibiting similar characteristics to the population we studied was believed to be the origin of the 1994 CDV epidemic among Serengeti wildlife (8). Furthermore, carnivores use feces for scent communication, so the probability of infection by CPV in wild carnivores in the study area may also be high.
In developing countries, leptospirosis is emerging as a major public health problem and also causes enormous economic losses because of disease in livestock (9). The most commonly detected serovars in this study were those that have rats and dogs as reservoirs (Table). Visceral leishmaniasis in humans is also a major health problem in several areas of eastern Africa, where the number of cases has dramatically increased during the past 20 years. Transmission of Leishmania donovani in eastern Africa may take place through anthroponotic or zoonotic cycles, although, to our knowledge, no reservoir host had been identified (10).
The mean T. gondii seroprevalence detected during this survey appears to be the highest reported for dogs worldwide. This protozoon has implications for human and animal health, and dogs, who probably become infected with T. gondii when eating raw meat, are a good sentinel for environmental contamination by this parasite. On the other hand, dogs serve as the definitive host for N. caninum, which is a major cause of abortions in cattle and causes economic losses wherever it is enzootic.
Some of these diseases may also have implications for the conservation of endangered mountain gorillas. Diseases such as leptospirosis, toxoplasmosis, and especially, rabies could be fatal for gorillas, and there are unpublished reports of fights between hunting dogs and gorillas.
Our work should serve as a first step toward the establishment of preventive strategies for improvements in the health of humans and domestic animals living in rural Uganda and for the health of the country's unique wildlife. Tracing the role of dogs in the cycle of the studied pathogens is crucial for the design of control programs.

Acknowledgments
We thank Steven Rubanga, Abdulhameed Kateregga, and all the staff of Conservation through Public Health; the Mgahinga Community Development Organization; Martin Esau; District Veterinary Officers from Kanungu, Kasese, and Kisoro local governments; and Vicky Ticó for help with field work. We are indebted to Ingenasa for contributing to this study by freely providing the ELISA kits.
This study received partial support from project CGL2010-17931. We are grateful to the John D. and Catherine T.  To the Editor: Iatrogenic Creutzfeldt-Jakob disease (iCJD) is an acquired form of prion disease that has been declining in incidence since the mid-1990s (1). Worldwide, at least 226 cases of iCJD, including 29 US cases, have been associated with administration of contaminated human growth hormone (hGH) from cadavers. Reported incubation periods ranged from 5 to 42 years (mean 17 years) (2). Commercially produced cadaveric hGH has been associated with only 1 previously reported case of iCJD: CJD developed in a 39-yearold Austrian man ≈22 years after he received commercial cadaveric hGH (Crescormon, Kabivitrum, Stockholm, Sweden) during 1984-1985 (3). We report a second case of probable iCJD acquired through treatment with commercial cadaveric hGH.
The patient was born at 32 weeks' gestation with subsequent developmental delay, agenesis of the corpus callosum, and panhypopituitarism. He demonstrated clinical and laboratory signs of growth hormone deficiency but was denied treatment with hGH through the US government-supported National Hormone and Pituitary Program (NHPP) because he did not meet the height requirement. Treatment with commercial cadaveric hGH began when he was 5.8 years of age and continued for 23 months (1983)(1984)(1985). He received 1.5 units intramuscularly 3× per week and was primarily treated with Asellacrin (Ares-Serono, Geneva, Switzerland). In early 1984, for an unspecified duration, he received Crescormon (Kabivitrum) because of an Asellacrin shortage. Treatment was halted in 1985 because of iCJD concerns and resumed 2 years later with recombinant hGH.
At age 33, 26.5 years (range 25.5-28 years) after the midpoint of commercial cadaveric hGH treatment, dizziness and gait imbalance developed, causing a fall. The patient's mental status also began declining, and he never returned to his baseline status. Six months after illness onset, he experienced hallucinations, weakness of lower extremities, and limb ataxia. Seven months after the fall, he entered a state of akinetic mutism; he died 9 months after symptom onset. A lumbar puncture, performed 8 months after illness onset, demonstrated 14-3-3 proteins and an elevated cerebrospinal fluid (CSF) tau level of 14,111 pg/mL (decision point 1,150 pg/mL) (4), although the specimen was contaminated with blood (39,375 erythrocytes/μL). Electroencephalogram demonstrated severe diffuse encephalopathy. Two brain magnetic resonance imaging studies performed 8 months after illness onset indicated probable CJD, given lack of prior metabolic and anoxic insults (Figure). The patient was discharged from a referral hospital with this diagnosis; no postmortem analysis was conducted.
On the basis of World Health Organization criteria, we conclude that this patient had probable iCJD as a result of hGH treatment (5). The patient's condition was treated with 2 different formulations of commercial cadaveric hGH, including one of the same brands in the same year as that of the first reported patient with iCJD associated with commercial cadaveric hGH (3). The patient's incubation period (25.5-28 years) is well within expectations (1).
Despite an ongoing active surveillance program that identified ≈3,500 of ≈4,500 post-1977 cadaveric hGH recipients in the US NHPP, all 29 CJD infections in NHPP recipients occurred among the estimated ≈2,700 pre-1977 recipients (1,2). This significant reduction in iCJD was attributed to the 1977 introduction of a highly selective, column chromatography step in the hormone purification protocol that can markedly reduce prion infectivity (1,2). As shown by the many iCJD cases linked to hGH in France, the efficacy of column chromatography purification steps may vary (1). Commercially derived cadaveric hGH was produced in different laboratories from those that produced NHPP-distributed hGH, and sufficient details regarding sourcing and production methods of