Veterinary cancer registries in companion animal cancer: a review

Authors


L. B. Brønden
Department of Small Animal
Clinical Sciences
The Royal Veterinary and
Agricultural University
Dyrlaegevej 16
DK-1870 Frederiksberg C
Denmark
e-mail: lobb@kvl.dk

Abstract

Current and prior veterinary cancer registries are few in number and scattered. Different inclusion criteria, dissimilar collection methods and variable reference population estimation methods pose obstacles in the comparisons between veterinary and human cancer registries. Veterinary cancer registries have yielded information on the risk and incidence of different cancer types in certain breeds and geographical regions, as well as provided information on genetic and environmental risk factors in some cancers. The objective of this article is to review the prior and current veterinary cancer registries, the information they have contributed and to discuss different issues relating to their structure including inclusion criteria, study populations, reference populations utilized in evaluations, recorded variables and the outcome from these.

Introduction

Cancer is one of the most important causes of death in dogs.1 Mortality studies show that death is caused by cancer in approximately 15–30% of dogs and 26% in cats2–5 and that cancer is the most common fatal disease.4 A marked variation between the countries where the studies were performed and a significant difference between different breeds were evident.2,3,5–7

According to Baillière’s veterinary medical dictionary,8 a registry is defined as a central agency for collection of pathologic material and related data in a specified field of pathology, so organized that the data can be properly processed and made available for study. However, the following registries have also been based on related data, such as pathology reports or clinical findings not accompanied by pathological material.

While cancer registries in human medicine have existed and evolved since the 1940s, now numbering 449 individual registries, veterinary cancer registries have existed in smaller numbers and have often been short lived and sporadic in nature. There is little up-to-date information available on the incidence of different types of cancer in the companion animal population anywhere in the world.9,10

Companion animal cancer registries provide information for evaluation of incidence and relative risk estimates along with data for epidemiological studies of spontaneous companion animal cancers. This information includes risk factor identification, treatment evaluation and location of patients for clinical trials as well as for cases in case–control studies. In addition, canine and feline cancers may be utilized comparatively as spontaneous animal models and sentinels and as such facilitate the identification of risk factors in carcinogenesis.11,12 Variations in cancer distribution between different geographic areas and correlations made between neoplasia in companion animals and humans may support and develop cancer research.2,13

Cancer registries have yielded descriptive studies on the distribution of neoplastic diseases between groups within a population but can provide the information needed for analytical studies. In many analytical studies, the focus is the identification of risk factors by comparing groups of affected animals with healthy animals. Often studies are conducted in a retrospective manner, but a true registry is prospective in order to ensure uniform submissions.

The purpose of this article is to review the current and prior veterinary cancer registries, assess the information they have contributed and to discuss their different structures, including inclusion criteria, study populations, reference populations utilized in evaluations, recorded variables and outcome from these, with an emphasis on registries holding information on companion animal neoplastic diseases.

Former and existing registries

Numerous surveys and studies from extensive data collections have been performed, covering broad fields in oncology research, including genetic studies.14,15 These studies can yield specific information but lack a population-oriented approach, and therefore, the results often cannot be extrapolated to cover the entire population.

The first companion animal cancer registries were started in the early 1960s. Some are closed, while others still accept submissions. In addition, there exist animal cancer registries of lower animal species16–18 and on neoplasms seen in experimental animal studies.19,20

A list of the veterinary cancer registries for companion animals from which data have been published is presented with date of initiation, species included and level of investigation in Table 1.

Table 1.  Current and prior veterinary cancer registries for companion animals
 StartDurationNumber of entriesSpeciesDiscretion levelNomenclature/classification
  • NCI, National Cancer Institute; WHO, World Health Organization.

  • a

    NCI: Standard Nomenclature of Veterinary Diseases and Operations. WHO: Manual of the International Statistical Classification of Diseases, Injuries, and Causes of Death (1955 revision, Geneva 1957).

  • b

    Except for canine mammary tumours – other classification scheme.

Kansas University Neoplasm Registry records21–24196111 years3837 dogs, 287 horsesCanine, equineHistopathology 
California Animal Neoplasm Registry26,2819633 yearsMore than 30 000Canine, feline, avian, bovine, equine, rats, other small rodentsHistopathologyNCI/WHOa
Tulsa Registry of Canine and Feline Neoplasms2519725 years Canine, felineHistopathologyNCIb
Purdue Comparative Oncology Program411979 More than 20 000 NCI/WHO
Cancer Registry and Surveillance System for Companion Animals Cornell 421980Still collecting (25 years)170 000Canine, felineHistopathology 
Norwegian cancer project1,50–53,56,57,65,791990Still collecting (15 years)14 401CanineHistopathology 
VetCancer Registry491994Still collecting (11 years)Almost 10 000Canine, feline 
Danish Veterinary Cancer Registry582005Still collecting (1 year)1000Canine, felineAll diagnostic modalities 

Inactive registries

At Kansas University, surgical biopsy samples and clinical data from tumours in dogs and horses presented at the university clinic were prospectively gathered from 1961 to 1972. Data published from this registry consist of overview epidemiologic articles on squamous cell carcinomas, sebaceous gland carcinomas and basal cell tumours.21–24 The biopsies were collected and later evaluated histopathologically at the university.21–24

The Tulsa Registry of Canine and Feline Neoplasms in Oklahoma was active from 1972 to 1977 and offered free cytology and histopathology.25 Publications from this registry include a descriptive review listing the occurrence of the different tumour types in cats and dogs according to gender. A special emphasis was placed on the most frequently submitted neoplasm in dogs, in this case, mammary tumours, which occurred most frequently in Pointers and Poodles. In cats, haematopoetic tissue was the most frequently affected.25 A publication on the number of submissions and use of diagnostic methods utilized by submitting veterinarians showed that admissions to the registry increased over a 3-year period, and this was declared likely to be because of greater awareness by clients and veterinarians.

The California Animal Neoplasm Registry

One of the largest, best known and most often cited veterinary cancer registries is the California Animal Neoplasm Registry (CANR). For 3 years after 1963, data were collected from a well-defined study area, recording more than 30 000 malignant neoplastic cases on which histopathology was performed free of charge.26 This data set has frequently been used as reference data set in the evaluation of cancer incidence in other studies and surveys as well as in teaching materials.27

Publications derived from the CANR include analytical studies, case reports and morbidity studies. These include estimates of cancer incidence from the reported cases in dogs and cats according to age, sex and breed, showing that skin is the most frequently affected tissue in both cats and dogs and that purebred dogs seem to be more prone to develop neoplastic diseases in all sites.28 Selected cancers were evaluated to investigate the usability of canine and feline cancers as models for human cancers in the cases of mammary cancer, leukaemia and lymphoma.28 From this registry, risk factors and geographic trends have been considered and investigated. A comparison between cancer incidence in humans and companion animals in the same households was made, suspecting a viral aetiology, but no overall association of human and canine cancer was found.29 Likewise, no association was found between feline malignant lymphoma and contemporaneous occurrence of human cancer.30 With data from the CANR, Dorn et al. showed that white cats have a 13.4 times greater risk for developing cutaneous squamous cell carcinomas compared with nonwhite cats.31 Another study showed that Dachshund breeds had a greater risk of developing mammary cancer than other breeds.32

Purdue Comparative Oncology Program

The Purdue Comparative Oncology Program (PCOP) was initiated in 1979 with the intention to improve cancer treatment and methods for companion animal cancer patient care. It consists of a registry of biological and demographic data for animal tumour specimens submitted for diagnosis by veterinarians; more than 100 veterinarians voluntarily submit samples for free histopathological evaluation.33,34

A few publications crediting this registry have been published relating to canine bladder cancer, among others. The role of topical flea and tick pesticides in the development of transitional cell carcinoma (TCC) was investigated; however, no association was confirmed.35 A significant increase in the development of TCC in dogs exposed to gardens treated with herbicides alone or with both insecticides and herbicides was found using data from the registry.36 A number of neoplasias have been identified and investigated for their production of cyclooxygenase (COX)-1 and COX-2 with the intention to predict tumour response to use of COX inhibitors in anticancer treatment.37 However, further studies concerning the predictive value of neoplastic concentrations of prostaglandin 2 and COX-2 on antineoplastic activity of COX inhibitor treatment showed that these were not useful.38 Articles providing an overview of cancer incidences of dogs and cats from this cancer registry have not been published despite the fact that evaluations have been made of the pet population in the catchment area of the registry.39,40 There is currently no mention of this registry on the PCOP homepage.41

Cancer Registry and Surveillance System for Companion Animals, Cornell

This registry has collected and analysed biopsy reports from two diagnostic laboratories, receiving samples from four counties in New York State.42 Data have been gathered since 1980, and more than 170 000 pathology reports on neoplastic diseases have been collected so far.42 Information is shared via a homepage which, among other things, displays the geographical distribution of cancer types on maps of the defined region.42 Evaluation of data of 25 years worth has revealed temporal changes in cancer occurrence in the form of decreases in the occurrence of canine and feline mammary cancer and an increase in the number of canine lymphomas.42 Published geographical observations locating cancers such as lymphoma, mammary carcinoma and mast cell tumours in both cats and dogs to specific areas in New York state can be seen on the homepage, which was last updated in April 2005.43

Active registries

The Veterinary Medical Database

Another very large data compilation is The Veterinary Medical Database (VMDB). It was initiated in 1964 and is still accepting submissions on all diseases in many species. It covers a large area, as registrations have been gathered from 26 universities with teaching hospitals across the USA and Canada.44 When originally initiated by The National Cancer Institute, the registry was situated at Michigan State University and worked as a collection of information concerning cancer in animals. It was hosted at Purdue University until 2006 and expanded to include all clinical conditions. The registry has recently been moved to Illinois and holds a total of 7 million cases from a number of species.45,46 It is diagnosis-based and receives only data and no specimen samples. At least 72 articles have been published utilizing data directly from the VMDB and a multitude of articles have subsequently cited these.45 In canine cancer research, articles have been published on investigations into the prognosis after surgery for mast cell tumours, which revealed no difference in outcome with or without tumour-free margin in cutaneous mast cell tumours.47 Another study summed up clinical and pathological aspects of spontaneous canine prostate carcinoma based on cases derived from the VMDB caseload.48 However, very few hospitals have maintained their registrations, and only eight have submitted data in this century, the latest in 2005.

The VetCancer Registry

The international web-based VetCancer Registry was initiated in 1994 with the development of the International Veterinary Brain Tumour Registry (IBTR).49 With a successful initial registration of more than 200 brain tumours, the registry was expanded to include all neoplasias, and the IBTR was incorporated in this general registry.49 The registry only accepts cases with diagnoses confirmed by histopathology and works as an anonymous link between investigators and contributors.49 The VetCancer Registry is an international registry based on voluntary data submission, and both submission and extraction of data is free.49 The registry is currently still receiving cases (pers. comm., Steve Steinberg, 6 January 2007). No publications from this registry are known to the authors.

The Norwegian Canine Cancer Registry

In Norway, a cancer registration project was initiated for canine cancer in a defined geographical region in 1990. This incidence registry includes both malignant and benign neoplasias and originally provided free histopathological evaluation to practitioners in the area when submitting cases for the registry. In 1998, the register became a national register, and free histopathology is no longer available for the practitioners.50,51 The registry has reported information on the geographical distribution, reproduction status, nutritional status, prior hormonal treatment and concurrent diseases.50 Comparisons to other surveys and epidemiological studies were last made in 2001.1,50,52 Informational articles have been published in both veterinary, breeder and medical periodicals in the 1990s.51,53 An incidence study of mammary tumours in boxers, bichon frises and Bernese mountain dogs revealed that boxers had the highest relative risk and a high population-based incidence risk.54 The association of the development of canine mammary cancer and the use of medroxyprogesterone acetate was confirmed in another study.55 Further studies have been performed on mammary cancer in boxers showing a general high risk but unable to confirm any familiar relation in the pathogenesis.56 Bone tumours were also investigated, and the Irish wolfhound, Bull mastiff and Afghan breeds were reported as having high relative risk estimates.57 An overview article reported that the most frequently encountered breeds in the registry were boxer, Flat coated retriever, Giant schnauzer, English cocker spaniel and Rottweiler when breed distribution in the entire population was taken into consideration. Neoplasias of the skin were by far the most common type of tumour encountered in this registry, followed by mammary, oral and testicular neoplasias.52

The Danish Veterinary Cancer Registry

In 2005, a new veterinary cancer registry was established at the Royal Veterinary and Agricultural University in Denmark. Veterinarians voluntarily submit data to this web-based data compilation. It is an incidence registry, and inclusion of a case is based on clinical information, cytology, diagnostic imaging and histopathology when available. The structure is based on the human cancer registry in Denmark, one of the oldest cancer registries still in function. So far, more than 1000 submissions have been collected, and newsletters containing epidemiological overviews are published on a regular basis.58 Preliminary data show increased relative risk in breeds such as Bull terrier, American Bull dogs and Boxers and the most frequently occurring neoplasm to be mammary tumours. The data have been correlated for breed distribution in the country, as detailed information has been obtained from the Danish dog registry on the 600 000 resident dogs.

Human registries

Collection and evaluation of data on cancer in human beings have taken place for decades. One of the earliest successful ones was the National Danish Cancer Registry. It was initiated on a voluntary basis in 1942, and since 1987, registration of all neoplasias to a national database has been made mandatory by law. Submissions are forwarded directly to the Danish Cancer Registry from general practitioners, specialists, hospital departments, histopathology laboratories and forensic departments, whenever cancer is diagnosed in a patient. Diagnosis and entry into the database can be based not only on cytology and/or histopathology but also clinical diagnosis or diagnostic imaging information. Links to the National Hospital Patient Registry, the Danish Registry of Causes of Death and the Civil Registration System ensure that all registered cases are linked to the correct names and demographic data. This substantial database yields yearly publications assembling all current data. An Internet homepage with elaborate information on relevant surveys and research results is maintained by the Danish National Board of Health.59

Discussion

Data input

The multitude of data collection methods, i.e. use of nomenclature and classification schemes, and inclusion criteria challenge the comparability of the information gathered in the existing veterinary cancer registries.

Standard nomenclature and classification is applied in the CANR,26 in the Tulsa Registry,25 in the PCOP and in the Danish Veterinary Cancer Registry to facilitate comparisons between human and animal neoplastic disease.11 Older registries may hold legacy data that no longer correspond to the division and subdivisions of the neoplasias that are utilized today owing to changes, new or redefined classification systems.45

Within the veterinary community, no universally accepted classification system exists, and an agreement to follow a specific classification system scheme (e.g. World Health Organization’s International Classification of Disease for Oncology system) worldwide is desirable.11,45

The registries also differ in their inclusion criteria as some include only neoplasias with a histopathologically verified diagnosis. In these registries, samples from non-neoplastic lesions represent 20–40% of the samples, the highest percentages in those registries where free histopathology was offered – the Norwegian, the CANR and the Tulsa registries.25,26,49,50 However, some diagnoses are more easily made by cytological evaluation or other diagnostic modalities rather than by histopathology, and an underestimation of tumour types readily diagnosed by means other than histopathology may take place. Some registries accept results from other diagnostic modalities when including cases.58

All these differences must be taken into consideration when comparing the outcome of the different registries, and results cannot be readily compared.

Data evaluation

Registries are by nature prospective observational studies, and the evaluation of this type of data collections is strong in the identification of intermediate or high levels of risk but weak at identifying very low levels of risk.60 If latency periods are long, it can be difficult to locate specific agents or to trace specific exposures owing to the mixtures of environmental exposures that the individuals are submitted to. In observational studies, risk factors cannot be controlled and confounding influences are not avoided by randomization procedures as in a laboratory setting.60

Completeness of reporting is a major concern in morbidity studies. Underreporting of cases will result in an underestimate of the true incidence, while incidence rates based on data that include unconfirmed diagnoses may overestimate true incidence.26 Overreporting of specific neoplasias may occur because of screening or special interests, new diagnostic modalities or free histopathology in a given registry or smaller study within a registry.10 Numbers, types and stage of development of neoplasia cases presented to referral institutions may not be representative of those in the general population. In the Norwegian Canine Cancer Registry, the CANR and the Tulsa Registry, free histopathology was offered to the submitting veterinarians. This may have not only altered the number and types of submissions but also facilitated the gathering of a large amount of data.

Collection methods and management of data input may severely influence the output that may be withdrawn from a registry. A simple uniform collection of data is crucial. Preferably, all registries would operate with the same inclusion criteria and classification schemes.

Reference populations

The usefulness of registry data depends on the quality of the characterization of the population at risk and realistic estimates of the proportion of cases that the registry will identify.39,40 The calculation of incidence rates is based on two numbers; the total number of animals in the reference population at risk and the actual total number of affected animals in this population.13 Often, both the numerator and denominator are estimated numbers. The estimated incidence rates therefore may be unreliable, and this may enhance or conceal smaller differences and subtle changes in the occurrence of neoplasia.25,61 Some authors choose to calculate the relative risk, where one type of neoplastic disease is seen as a percentage of all neoplasias within one breed.50

If unknown, the population at risk can be difficult to estimate, and several different approaches have been undertaken. In the CANR, a reference population was estimated based on a telephone household survey calculating the number of dogs and cats according to the number of people in the catchment area.26,39,46,62 Some registries have included only cases from defined populations in a specified and limited geographical area.25,42,58,63

Enumeration of licensed or vaccinated animals in a region may lead to an underestimation because not all owners take advantage of these opportunities.25 In the case of the Danish Veterinary Cancer Registry, researchers have the advantage of the Danish Dog Registry, which holds detailed information on the vast majority of dogs in Denmark. Registration of dogs is a legal requirement, and cat owners are also very willing to register their animals in the existing cat registries in the country.

Extrapolating population estimates from one period or from one geographic region to another may be inappropriate because of temporal and regional variations in companion animal ownership and in distribution and frequency of cancer.39,40 Studies have shown variations in type and occurrence of neoplasia between geographical areas and between different breeds.25,64,65 This emphasizes the importance of the ability to ascertain variations in breed distribution, as these may change markedly and unpredictably.66

Knowledge of the reference population is essential, but an insight into which proportion of these animals are taken to the veterinarian and diagnosed with neoplastic disease is equally important. Teclaw et al. found that approximately 20% of dogs and 36% of cats had not been seen by a veterinarian in 12 months before a large round of interviews.40 They also found that dogs were more likely to have been seen by a veterinarian if they were less than 1 year old or older than 7 years.40 This suggests that the number of cats is often underestimated and the number of very young and older dogs overestimated if a population estimate is based on the number of veterinary visits. Dorn et al.63 had a probability sample survey performed on households in the catchment area by the Human Population Laboratory in 1965. They included only animals with a history of being seen or treated by a veterinarian in the estimated population at risk. This resulted in an inclusion of only 87% of the dogs and 75% of the cats,63 illustrating that a hidden population must be taken into consideration when evaluating data derived from veterinary cancer registers.

The challenge of estimating a reference population and the different approaches utilized complicates the comparability of the registries and the information reported by these.

Comparative aspects and risk factor identification

Comparative oncology is concerned with collecting clinical and scientific information from a variety of related disciplines to expand the understanding of aetiology, diagnosis and the clinical course of neoplasia in animals and human beings.67

The spontaneous neoplasms of companion animals may provide useful models for studying the health effects of environmental hazards on human beings and thereby constitute sentinel species on human cancer biology and translational therapeutics.13,68–71 The relatively high incidence rate of some cancers, similar biological behaviour, large body size, comparable responses to cytotoxic agents, shorter overall lifespan and shorter latency periods in companion animals are factors that contribute to the advantages of the companion animal as a model.60 A number of neoplastic diseases have already been recognized to share numerous traits with human counterparts in terms of biological constitution, distribution and therapy responses. These include lymphoma/leukaemia, soft tissue sarcoma, melanoma, mammary tumours, canine bladder cancer (TCC), prostatic carcinoma and osteosarcoma.71,72 Osteosarcoma in dogs is histologically indistinguishable from human tumours in children and adolescents, and melanoma in dogs closely resembles human melanoma both in biology and response to treatment.71,72 Non-Hodgkin’s lymphoma (NHL) in the dog has a similar biological constitution and prognostic behaviour as NHL in humans, which strengthens the model potential of this species.71 Several features of feline malignant lymphomas indicate that they would provide a good model for study of human leukaemia and lymphoma.28 Similarities in distribution of bladder cancer in companion animal dogs and human beings have been described.73

In a study where malignancies of oral, nasal, lung and bladder tissues were selected owing to their potentially environmental aetiology, estimates of canine cancer incidence were compared with human cancer incidence within a specified geographical area. A correlation was found whereby changes in the canine proportionate incidence ratios preceded human incidence rates by 2 years, suggesting that fluctuations in proportionate incidence ratios in neoplasia in dogs may be useful in predicting changes in human neoplastic patterns.13 Research has been published on concurrent neoplasia in human beings and companion animals in the same household, and zoonotic transmission has been suspected. However, so far, no causative connections have been documented.30,74,75

A ‘standard of care’ has not been established for most neoplastic diseases in animals. This allows for some latitude in prospective clinical trials, and with the increase in new therapeutic agents, the companion animal model may be useful populations to test new agents with benefits for both humans and animals.71 Veterinary cancer registries providing information on cancer in the existing companion animal population can provide information on the effect of long-term exposure to appropriate offending agents as opposed to a classical trial setting where chemicals and other agents are administered to experimental animals.76 Findings in such inbred animal populations in an artificial laboratory environment may not be applicable to naturally occurring tumours in the human population.71 In a study where a comparison of canine oropharyngeal malignancies was made, marked differences between different geographic locations were found, showing that the prevalence of canine tonsillar squamous cell carcinoma to be 15 times greater in London compared with Melbourne and four times greater than in other areas of south-east England.64

Animals may be used as models for some of the cancers that occur infrequently in humans.12 Annual incidence rates for certain neoplasias, such as canine osteosarcoma, canine soft tissue sarcoma and feline NHL, are significantly higher than that in humans, and their relative abundance highlights their model potential.71 The role of genes in cancer development may also be studied in companion animals, as veterinary epidemiological studies have disclosed clustering of tumour cases within certain breeds and families.77,78

In theory, one of the main purposes of cancer epidemiology is the detection and identification of risk factors.12 Often, more elaborate data are needed than that are generated in registry files. However, cases for case–control studies can potentially be identified. Specific cancers, areas or breeds are easily isolated and further investigations can be undertaken. Even cases suitable for ongoing trials or investigations could possibly be identified for inclusion as they are reported. Using companion animals as sentinels offers the advantage of assessing possible environmental exposures in an outbred population, without the potential confounding effects of alcohol, tobacco and occupational exposures to carcinogenic agents.13

Risk factors ascertained from veterinary cancer registries

A list of risk factors that has been investigated with registry material as its background is presented in Table 2. This list could potentially be much longer, as a large number of studies on both exogenous and endogenous risk factors have been conducted using data from hospital files and surveys instead of utilizing the data compilations already gathered.

Table 2.  Research of risk factors in the development of neoplastic diseases where data have been derived from veterinary cancer registries
Offending agentStudy typeCancerFindingsSpeciesRegistry and reference
  1. OS, osteosarcoma; SCC, squamous cell carcinoma; TCC, transitional cell carcinoma.

Topical flea and tick pesticidesObservational, case–controlTCCNo association was confirmedDogsPCOP35
Herbicide exposureObservational, case–controlTCCRisk of TCC significantly increased if dogs exposed to gardens treated with both herbicides and insecticides or herbicides aloneDogsPCOP36
Human cancerObservational, retrospective, case–controlLymphomaNo association was confirmedCatsCANR30
Human cancerObservational, retrospective, case–controlAll malignanciesNo overall association was confirmedDogsCANR29
Fur colour (white)Observational, case–controlSCCWhite cats have a 13.4 times greater risk for developing cutaneous SCC compared with nonwhite catsCatsCANR31
NeuteringObservational, case–controlMammary tumoursNeutered bitches had 12% of the mammary cancer risk as compared with intact animals. Early spaying was shown to have a sparing effectDogsCANR80
Concurrent immune-mediated diseaseObservational, case–controlLymphomaDogs with immune-mediated thrombocytopenia had a greater occurrence of lymphoma than dogs withoutDogsVMDB81
Height and weightObservational, case–control, retrospectiveOSA stronger and more consistent association of OS was found with increasing height than with increasing weightDogsVMDB82

Removal of risk factors or the development of more resistant animals by breeding procedures may lead to the prevention of some cancers in the companion animal species.12

Conclusion

A comprehensive veterinary cancer registry has the potential to increase our knowledge of the distribution of neoplasia in the companion animal population while at the same time being an instrument with which we are able to monitor fluctuations in cancer occurrence. In addition, animals suffering from specific neoplastic diseases may be selected and utilized as spontaneous animal models for human cancers. Veterinary practitioners may benefit from a registry by obtaining data for specific breeds, age groups or geographic areas, as these may vary considerably from surveys and registries quoted in books and journals.

An important contribution that veterinary cancer registries can provide is data on identification of geographical differences and high or low animal risk groups, providing insight into the aetiology of neoplasia. A computerized database, easily accessed by the users with an internationally standardized format, a unique extensive set of guidelines and a single up-to-date and change-ready coding system corresponding with human cancer registry coding, is desirable.11,45

A major challenge in the creation of a registry is to communicate its existence to the veterinarians and to create incentives for the participants to join and maintain this valuable research tool.11 Ideally, all veterinarians would report equally and all neoplastic lesions diagnosed by histopathology. If reporting to a central national veterinary cancer registry was made a legal requirement and if extracting data directly from the veterinarian’s electronic journal systems was made possible as well as cross-reference to information from pathology laboratories and companion animal registries, a high level of diagnostic value and a high number of registrations would be ensured and thereby a valid numerator will be added to the equation in the calculation of true incidence of neoplastic diseases.

Ancillary