Click for next page ( 70


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 69
4 Companion Animals as Sentinels Companion animals have been used as surrogates for humans in exposure assessments. Where humans are exposed to contaminants in complex environ- ments (e.g., in the home or in the work place), it can be difficult to estimate their exposure with conventional procedures of measuring ambient concentra- tions of the contaminants and calculating their intakes from the contaminated media. One approach to solving the problem is to use animals exposed in the same environments as surrogate monitors; tissues of the animals are taken for analysis and used to provide an integrated measure of the animals' exposure. If the animals' contact with the contaminated media is sufficiently similar to that of the humans, the animals' exposure might provide a reasonable indirect measure of the humans' exposure. Most examples of such animal sentinel systems involve the use of domestic or companion animals. For example, pet dogs have been used as surrogate monitors of human exposure to asbestos (Glickman et al., 1983) and lead (Thomas et al., 1976; Kucera, 1988~; studies are under way with dogs as surrogate monitors of human exposure to tetra- chlorodibenzopdioxin (Schilling and Stehr-Green, 1987), and radon (Schuckel, 1990~. Blood and other tissues of companion animals often are sampled, e.g., at surgery or slaughter. Pet animals generally are not sacrificed for study, but most pet animals have relatively short lives, and their tissues can be sampled when they die (see, e.g., Glickman et al., 1983~. Although pet animals occupy the same environments as their owners and are expected to be exposed in broadly similar ways, exposures of pets are not identical. Among other differ- ences, animals usually have greater contact with soil, house dust, and floor surfaces than do humans, and they are more likely to ingest contaminants when cleaning or grooming themselves. Differences in animal metabolism and pharmacokinetics also mean different relationships between exposure and tissue concentrations, but these could be adjusted for using modeling tech- niques (Andersen, 1987~. 69

OCR for page 69
70 ANIMALS AS SENTINELS DESCRIPTIVE EPIDEMIOLOGIC STUDIES Veter~y Medicme Data P~7' The Veterinary Medicine Data Program (VMDP), sponsored by the Na- tional Cancer Institute, was initiated in 1964 (Priester and McKay, 1980~. Selected data on animals (some of which are food animals) treated at partici- pating veterinary schools are entered on magnetic tape and sent to a central processing unit at Purdue University, where they are summarized, and quar- terly reports are issued. A standardized coding scheme is used to record diagnoses and operations by each school. The population at risk in the VMDP system is unknown, so true prevalence or incidence rates cannot be determined. Proportionate morbidity ratios (PMRs) are used, however; the numerator is the number of cases of a specified disease observed at participat- ing veterinary schools, and the denominator is the total number of animal visits at the schools. The ratios can be expressed as risk in animal-years and are age-adjusted, allowing for a comparison of risk between geographic re- gions and periods. Cases of cancer in animals recorded by the VMDP can be used as a start- ing point for more extensive, controlled analytic studies or to pinpoint unex- pected clusters or an increased frequency of an environmentally related dis- ease. Data from the VMDP were used to calculate PMRs for animal cancers by site or type for 8,760 pet dogs (Hayes et al., 1981~. A significant positive correlation was noted between the PMRs for canine bladder cancer and over- aD industrial activity in the host county of the veterinary school. Mortality from bladder cancer among white men and women in the same counties showed similar correlations with industrial activity. Similar patterns in humans and animals suggest that ambient exposures are more important than occupa- tional exposures in the risk of bladder cancer and that the dog might be a sensitive sentinel for the presence of bladder carcinogens. However, addition- al analytic environmental epidemiologic studies are needed to identify specific residential and environmental exposures associated with an increased risk of bladder cancer. Poisorl-Control Centers Animal poison-control centers (APCCs) serve as epidemiologic networks to monitor the prevalence of animal exposure to diverse chemicals and other toxicants (e.g., pesticides, feed additives, human and animal drugs, and house- hold products). Most cases reported to APCCs involve companion animals;

OCR for page 69
COMPANION ANIMALS AS SENTINELS 71 however, the largest numbers of animals involved are livestock and poultry, because more animals per reported case are involved. In 1988, the classes of toxic agents most commonly reported were rodenti- cides (15.7%) and insecticides (14.5%) (Trammel and Buck, 1990~. Plants, human medicines, and household products also were reported frequently. The home was the reported location of the exposure in 68% of the incidents; the yard and garage accounted for an additional 10%. Systematic evaluation of reported animal toxicoses might permit identification of unsuspected hazards in the human environment that might otherwise go unnoticed. ANALY7IC EPIDEMIOLOGIC STUDIES Canine Mesothelioma and Asbestos Exposure Epidemiologic evidence indicates that asbestos is a causal factor in human mesothelioma and that the latent period for cancer development after occupa- tional exposure usually exceeds 20 years (Selikoff et al., 1980~. However, up to 40% of persons with mesothelioma have no recorded history of exposure to asbestos (McDonald and McDonald, 1977~. Pet dogs with spontaneous mesothelioma were used to identify environmen- tal exposures that might increase their owners' risk of asbestos-related disease (Glickman et al., 1983~. The animals were selected because they share human domiciles, but do not indulge in activities that confound interpretation of the results of human epidemiologic studies (e.g., smoking and working). Eighteen histologically confirmed canine mesotheliomas were diagnosed at the Veteri- nary Hospital of the University of Pennsylvania, in Philadelphia, from April 1977 to December 1981. Sixteen owners of dogs with mesothelioma and 32 owners of age-, breed-, and sex-matched controls were interviewed to deter- mine their occupations and their dogs' medical history, life style, diet, and exposure to asbestos. Mesothelioma in the dogs was significantly associated with household members' asbestos-related occupation or hobby and the use of flea repellents (powders containing asbestos-contaminated talc). In addition, there was a trend toward increased risk of mesothelioma with urban resi- dence. Lung tissue from three dogs with mesothelioma and one dog with squamous cell carcinoma of the lung had higher concentrations of chrysotile asbestos fibers than lung tissue from control dogs. The latent period for mesothelioma in the dog after asbestos exposure is probably less than 8 years (which is considerably less than the latency in humans). Therefore, if canine mesothelioma were reportable (to a local health department or to a federal agency), coordinated efforts could be made to identify and control asbestos

OCR for page 69
72 ANIMALS AS SENTINELS sources in the reporting households, and household members could be screened for early radiographic signs of asbestos-related disease. Canine Bladder Cancer and Insecticide Exposure A case-control study of bladder cancer in pet dogs was conducted to assess exposure to insecticides and passive smoking as potential bladder carcinogens and to determine whether obesity increases the risk of bladder cancer in dogs exposed to insecticides (Glickman et al., 1989~. Histologically confirmed transitional-cell carcinoma (TCC) of the bladder was diagnosed in 89 pet dogs from January 1982 to June 1985 at the Veterinary Hospital and Surgical Pa- thology Service of the University of Pennsylvania in Philadelphia. The dogs were stratified by year of diagnoses, sex, breed size (based on ideal body weight), and age. Control dogs were selected with a random search of records of all other dogs with biopsy reports until a comparable number of dogs with similar characteristics were identified. The owners of the dogs were interviewed to obtain medical, residential, and diet histories of the dogs, and information was requested on exposures to specific household and environmental chemicals and on smoking patterns of household members. Preliminary analyses revealed that the risk of TCC was not related to lifetime passive exposure to tobacco smoke (Table 4-1~. TABLE 41 Odds Ratio for Lifetime Pack-Years of Smoking by Household Members and Risk of Transitional-Cell Carcinoma in Exposed Pet Dogs Pack-Years Number Number of at Home of Cases Controls Odds Ratio* 0 26 32 1.0 1-3,000 18 18 1.2 >3,000 14 21 0.8 *x2 for trend = 0.11; p = 0.7. Source: Glickman, 1989. There was a significant dose-response relationship between TCC risk and lifetime exposure to tick and flea dips (Table 4-2~. Because some insecticides are stored in fat deposits in the body, and fat can promote tumors, the dogs'

OCR for page 69
COMPANrION ANIMALS AS SENTINELS 73 TABLE 42 Odds Ratio for Topical Tick- and Flea-Dip Exposures and Risk of Transitional-CellCarcinomain Pet Dogs Applications/ Number Number of Year of Cases Controls Odds Ratio* 0 23 42 1.0 1-2 14 23 1.1 >2 18 11 3.0 *x2 for trend = 6.1; p = 0.1. Source: Glickman, 1989. TABLE 43 Odds Ratios for Insecticide Exposure, Body Conformation, and Risk of Transistional-Cell Carcinoma in Pet Dogs Tick- and Flea-Dip Exposure Body Conformation No Yes Thin or average 1.0 1.4 (11/24)a 16/25 Overweight or obese 1.5 9.8 (12/18) (18/4)b aNumbers of cases/number of controls bp = 0.0003 Source: Glickman, 1989 body conformations a year before TCC diagnosis (as reported by owners) were evaluated in relation to the risk associated with insecticide exposure (Table 4-3~. On the basis of an odds ratio of 2.2 for any tick- and flea-dip exposure and a prevalence of use of dips of 41% among the control dogs in the study, the proportion of TCC cases in the pet-dog population that can be attributed to dip use is 33%. That does not, however, include the risk of TCC associated with other sources of insecticides, including flea shampoos, lawn products, and other household products. Tick and flea dips for dogs contain a variety of active , i] ngredients and sol-

OCR for page 69
74 ANIMALS AS SENTINELS vents that might be carcinogenic to the animals. Furthermore, each time a pet animal is treated with a tick and flea dip, substantial human exposure is likely to occur, primarily by absorption through the skin while handling the pet. A 1987 survey of California pet handlers revealed numerous symptoms associat- ed with occupational exposure to flea-control products (Ames et al., 1989~. The information from this survey has direct human health implications, in that the ingredients in tick and flea dips are commonly used as indoor and outdoor household insecticides. The role of insecticides in the etiology of human bladder cancer should be explored, especially for persons involved in the manufacture or application of insecticides. Lead Poisoning Lead poisoning is clinically and epidemiologically similar in dogs and hu- man infants, and Thomas et al. (1976) suggested using pet dogs as indicators of high blood-lead concentrations in children. Blood-lead concentrations were measured in 119 children and 94 pet dogs in 83 low-income suburban Illinois families. A significant diagnostic blood-lead concentration in a family dog increased the probability of finding a child in the family with similarly in- creased blood-lead concentration sixfold. In addition, a family dog with a history of pica also increased the likelihood of finding a child in the family with pica. Thomas et al. concluded that family dogs could be useful sentinels of lead poisoning in children, and veterinarians seeing dogs in clinical situa- tions might have a public-health responsibility to report lead poisoning. Marino et al. (1990) recently reported a case of lead poisoning from paint that was first noted by a family veterinarian in the family 10-year-old dog. Blood was sampled from another pet dog in the household; it also had elevat- ed blood lead. Medical examinations showed the family, including two chil- dren, and a babysitter and her children suffered lead poisoning. The Veterinary Medicine Data Program has been used to describe patterns of lead poisoning in cattle, horses, cats, and dogs (Priester and Hayes, 1974~. Breast Cancer u' Dogs Humans and their pet dogs often consume many of the same foods. Son- nenschein (in press) found that table food provides an average of 33.7% of the total calories that a pet dog consumes; the percentage is considerably higher (as much as 100%) for small dogs. Pet dogs can be used to evaluate the

OCR for page 69
COMPANION ANIMALS AS SENTINELS 75 effects of diet on breast cancer and on recurrence rates or survival times after diagnosis of cancer. Diet and Risk of Breast Cancer Obesity and high-fat diets have been associated with an increase in the incidence of breast cancer in laboratory animals (Rogers and Longnecker, 1988~. Results of human studies have been inconsistent, possibly because childhood nutrition might be more relevant than adult nutrition in the devel- opment of breast cancer (Rohan and Bain, 1987~. The potential connection between dietary fat and breast-cancer development is important and deserves vigorous investigation (Schatzkin et al., 1989~. Sonnenschein et al. (1987) conducted a case-control study with 150 female pet dogs that had breast cancer and two control groups 147 cancer controls (i.e., dogs with other cancers) and 131 noncancer controls. Dogs were matched for neuter status, sex, age, and breed size (based on ideal body weight). Owners were interviewed to obtain dietary, management (e.g., hous- ing and care), medical, and reproductive histories. The mean age of the test dogs was 10.5 ~ 2.5 years; 56 (37%) were spayed (by ovariohysterectomy) before diagnosis. Multiple logistic-regression analysis of fat, protein, and carbohydrate intake showed inconsistent associations for the two control groups and for intact and spayed dogs. In contrast, being underweight at 1 year was strongly protective, particularly for spayed dogs, whether cancer control dogs were used (odds ratio [OR], 0.04; 95% confidence interval [CI], 0.004-0.4) or noncancer controls were used (OR, 0.04; 95% CI, 0.004-0.5~. Being underweight as an adult had a weak protective effect for spayed dogs in the cancer controls (OR, 1.25; 95% CI, 0.04 3.5~. These findings suggest that nutritional status (and therefore body conformation) modulates hormonal concentrations at a critical period in reproductive development and thus modifies the risk of breast cancer. Diet and Survival with Breast Cancer Recent studies examining the relationship between dietary habits and prog- nostic factors for breast cancer in women suggest that the dietary patterns of the western world c.g., high intake of fat and low intake of carbohydrates and fiberaffect some prognostic factors in breast cancer, such as tumor size and estrogen-receptor content of the tumor (Holm et al., 1989~. Obesity and increased fat intake also have been associated with decreased survival of

OCR for page 69
76 ANIMALS AS SENTINELS women with breast cancer, and investigators have begun to study the efficacy of reduced fat intake as a component of breast-cancer treatment (Chlebowski et al., 1987; Boyar et al., 1988~. To define those relationships further in an animal sentinel, Shofer et al. (1989) identified a cohort of 145 female pet dogs with histologically confirmed mammary carcinoma. Information similar to that In the Sonnenschein et al. (1987) study was collected. A histologic-malignancy score was derived for each animal according to seven pathologic criteria. The mean age of the dogs was 10.4 ~ 2.5 years; 41% had been spayed before diagnosis. Estimates of survival indicated that no dietary or nutritional factor alone was statistically significant ~n ~ 0.005~. However, median survival for dogs with more than 27% of their total calories derived from protein was 2.4 years, compared with 1.3 years and 1 year for dogs with 23-27% and less than 23%, respectively ~ = 0.06~. When those data were fitted to a proportional-haz- ards model, recurrence, histologic score and tumor type, percentage of calo- ries from protein, and history of pseudopregnancy were significantly associat- ed with survival. Predicted 2-year survival rates for dogs with 10, 25, and 40% of total calories derived from protein in their diets were 26, 54, and 75%, respectively. Animal sentinel data can help to define the relationship between dietary fat and human breast cancer, and they might provide a useful model to clarify the nature of the association between increased fat consumption and risk of devel- oping breast cancer and help to improve disease management after traditional treatments. Birds and PotytetraJluoroe~ylene Exposure Blandford et al. (1975) reported that five cockatiels died within 0.5 hour after a frying pan that was coated With polytetrafluoroethylene (PTEE) burned in a room in which they were caged. PTFE is commonly used to coat cook- ware (e.g., Teflon and Silverstone~ products). The owner of the cockatiels developed shortness of breath, shivering, dizziness, nausea, and tightness in the chest all symptoms of "polymer fume fever." Blandford et al. noted the particular susceptibility of parakeets to exposure to PTFE pyrolysis products and cautioned pet-bird owners against keeping caged birds in cooking areas; they further warned humans to avoid exposure to such products. Wells and Slocombe (1982) also noted that birds were more susceptible to poisoning from PTFE pyrolysis products than were small mammals in a study of acute toxicosis of parakeets caused by heated, PTFE-coated cookware.

OCR for page 69
COMPANION ANIMALS AS SENTINELS 77 Lo - Level Radiation arid Cam:er Reif et al. (1983) investigated the relationship between exposure to low- level radiation from uranium-mill tailings and canine cancer in Mesa County, Colo. The county had been the site of extensive contamination of residential properties with radioactive tailings used for fill and other purposes. Human leukemia incidence in the county was twice the state incidence. A cancer registry was established that collected incidence data from the nine practicing veterinarians serving the county. All homes in the county were surveyed for gamma radiation. Analysis of 212 cancer cases in dogs and an equal number of noncancer controls showed no increase in cancer risk associ- ated with residence in a home contaminated with tailings. Similarly, the risk of specific cancers, such as malignant lymphoma, was not increased in dogs. Further studies of human cancer over a longer period also showed no increase in leukemia rates in the county. Cager u' r~etnam-Senace Dogs During the Vietnam War, dogs and their handlers were exposed to many infectious agents, insecticides, phenoxy herbicides, and therapeutic drugs. Military dogs have been used as sentinels for the presence of zoonotic infec- tious agents in their handlers in southeast Asia. Hayes et al. (1990) hypothe- sized that military dogs might be useful in assessing increased cancer risk in human Vietnam veterans. They examined necropsy records of 1,167 nonneu- tered male dogs that served in Vietnam and 1,409 that served in the United States. Dogs working in Vietnam were 1.9 times as likely to have testicular seminomas and testicular dysfunction. Hayes et al. concluded that further research is warranted and that the testis should be made a priority site in the study of cancers related to Vietnam experience. Animal Neoplasm Registry As described in Chapter 3, National Animal Poison Information Network and VMDP collect data on animals, some of which are companion animals. Another program, the Animal Neoplasm Registry (ANR), was established in July 1963 and ended in July 1983; it operated in and was sponsored by Ala- meda and Contra Costa Counties in California. The ANR was the first popu- lation-based animal-tumor registry in the world (Schneider, 1975~. A popula- tion-based human-tumor registry already operated in the same area of Califor-

OCR for page 69
78 ANIMALS AS SENTINELS nia; it was used for comparison with ANR to reveal information on the etiolo- gy and pathogenesis of virally caused cancer that might be shared or transmit- ted between animals and humans. The numbers of dogs, cats, and persons estimated to be included in the populations at risk in 1970 are shown in Table ~4. A comparison of cancer incidence rates for primary anatomic sites in each species is summarized in Table ~5. TABLE '4 Numbers and Ratios of Dogs, Cats, and Persons in Populations at Risk Alameda arid Contra Costa Counties, 1970 Contra Alameda Costa County County Combined Numbers Dogs 131,329 93,4136 224,815 Cats 89,138 62,038 151,176 Persons 1,073,184 558,389 1,631,573 Ratios Persons/dog 8.2 Persons/cat 12.0 6.0 7.3 9.0 10.8 Source: Schneider, 1975 On the basis of cases of cancer detected by the ANR, several studies were conducted to investigate the association between specific human and animal cancers. In one retrospective study, no association was found between the occurrence of cancer in humans and the occurrence of the same cancers in animals in the same households (Schneider et al., 1968~. In another study, 221 households identified as owning cats that had developed malignant lymphoma, a viral disease, were compared with matched control households containing cats without the disease (Schneider, 1972~. No difference was found in human cancer rates between the case and control households. Followup analytic epidemiologic studies of cancer cases in animals identi- fied through the ANR have found neither common determinants of cancer in humans and animals nor a cancer-causing agent that was transmissible from

OCR for page 69
COMPANION ANIA~1LS AS SENTINELS 79 TABLE ~5 Age-Adjusted Incidence Ratesa of Cancer, by Primary Site in Humans, Dogs, and Cats in San FranciscmOakland Area Primary Site or Cancer Rate per 100,000 of Each Species Per Year Humanb DogC CatC Mouth and pharynx 12.6 20.2 13.0 Digestive system 73.6 24.8 20.5 Respiratory system 47.1 8.0 8.1 Bones and joints 0.9 5.3 3.5 Soft tissues 2.1 27.2 13.1 Melanomas of skin 5.2 10.8 2.1 Breast female 81.1 90.8 25.6 male 0.8 3.3 0.2 Genital system female 59.2 3.4 0.58 male 55.6 39.4 0.0 Urinary system 18.4 4.2 2.4 Eye and orbit 0.9 0.8 1.6 Nervous system 5.8 2.0 0.9 Endocrine system 5.5 6.0 0.2 Leukemias and lymphomas 21.3 25.7 179.6 Unknown 8.4 7.4 4.5 All sites 300.3 213.0 264.3 aRate for each species age-adjusted to 1950 U.S. human census popula- tion standard. bl969-1971, from Cutler and Young, 1975. CJuly 1967-June 1974. reincludes melanomas, but not other skin cancers. Source: Schneider, 1976. animals to humans. The ANR did provide information on risk factors for some animal neoplasms, such as breast cancer, and suggested preventive methods.

OCR for page 69
SO ANIMALS AS SENTINELS SUMMARY Veterinary epidemiologic studies have several advantages over human epidemiologic studies, including lower cost, shorter latency of disease develop- ment, and greater ease of obtaining tissue and necropsy data. In comparison with laboratory-animal studies, animal sentinel studies more closely parallel human exposure conditions. Despite those advantages, limitations of animal sentinel studies are evident. Use of veterinary epidemiologic data permits collection of data on a large number of cases; but the data must represent minimal estimates, because the number of cases not diagnosed cannot be known. In contrast with human records, no birth or death records are collected, and many diseased animals are not taken for medical treatment. Furthermore, risk factors of sex and breed are calculated not on the total animal population (which usually is unknown), but on the number of animals seen at participating veterinary hospitals and clinics. Although observations of those cases are made by veter- inarians, rather than physicians, veterinarians can identify in animals disease conditions that exhibit the same etiology and development in humans. The physical size of an animal model used might be a disadvantage; data on ultrastructure, anatomy, physiology, and pathology of species or breeds might limit the usefulness of the data collected in some programs (Mulvihill, 1972~. The genetic makeup and environmental factors that influence the course of disease development often are unknown in the animals used in veterinary studies (as opposed to laboratory animals); therefore, excesses of specific defects in some animals might be attributed to genetic or environmen- tal factors or both.