Docsity
Docsity

Prepare for your exams
Prepare for your exams

Study with the several resources on Docsity


Earn points to download
Earn points to download

Earn points by helping other students or get them with a premium plan


Guidelines and tips
Guidelines and tips

Cancer in Wildlife: Understanding the Connection between Wildlife and Human Health, Slides of Pathology

The prevalence of cancer in wildlife, focusing on the role of pathogens, environmental toxicology, and reproductive factors. The document also discusses the importance of studying wildlife cancer as a sentinel for both animal and human health. New examples of spontaneous cancers in wildlife are presented, along with potential mechanisms of oncogenesis. The document emphasizes the need for greater attention to wildlife cancer and its implications for wildlife and human health.

What you will learn

  • What are the implications of wildlife cancer for human health?
  • What are the commonalities between cancer in wildlife and humans?
  • How can studying wildlife cancer provide insights into potentially unique non-age-related mechanisms of carcinogenesis?
  • How do pathogens contribute to cancer in wildlife?
  • What role does environmental toxicology play in wildlife cancer?

Typology: Slides

2021/2022

Uploaded on 09/12/2022

journalyyy
journalyyy 🇬🇧

4.7

(12)

215 documents

1 / 16

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
The prevalence and variety of neoplastic diseases in wild-
life are likely under-reported and are certainly understud-
ied. Phenotypically similar forms of cancer exist across
disparate species, underscoring the universal nature of
cancer. A better understanding of both shared and unique
aspects of cancer across wildlife species will undoubtedly
shed light on potential mechanisms of oncogenesis. This
comparison may even untangle some of the complexities
regarding human cancer. Certainly, the large number of
animal populations inhabiting distinct environments pro-
vide many natural ‘experiments’ such as that examined
by Mǿller etal.1, where cancer in a broad spectrum of
bird species was shown to be tightly linked to the natu-
ral history and immunity of each species. Wild animals
are profoundly influenced not only by selective pressures
within their own species but also by their interface with
humans. Information gained from monitoring diseases in
free-ranging wildlife is of paramount importance, as we
often occupy the same habitat. A growing human popula-
tion places increasing demand on many shared resources.
For example, globally, water resources are being limited by
both access (human development) and a shrinking sup-
ply, so that the demand for adequate wastewater treatment
will become increasingly important over the coming dec-
ades (Fig.1). Furthermore, an increasing human popula-
tion will usher in global changes that wildlife already face,
such as crowding, mixing of subpopulations and increased
environmental complexity. How these changes modulate
the pathophysiology of cancer in wildlife species will also
be extremely important to identify, as the implications for
wildlife and human health are likely to be similar.
This Review describes some of the cancers affecting
wildlife populations and what we know about the role of
infectious agents, environmental toxins and reproductive
or metabolic disturbances in contributing to the onco-
genic processes (Table1). Although neoplasia occurs in
nearly all multicellular organisms2, this Review primarily
addresses instances of such disease in wild vertebrates.
We also describe new examples of spontaneous cancers
that have been identified in both captive and wildlife spe-
cies since the potential impact of cancer on wildlife was
first reviewed by McAloose and Newton3. Such diseases
are an increasing concern for protected and endangered
species and are a unique frame of reference for pressures
that drive neoplastic transformation.
Pathogen-driven cancer
Infections in humans are considered causal for approxi-
mately 15% of the global cancer burden2. Owing to a lack
of vigilant screening, which is often applied to human
cancer, the number of wildlife cancers caused by infec-
tion is not known, nor has it been possible to determine
whether the number of cancers associated with patho-
gens is similar in humans and wildlife. However, there
Cancer in wildlife: patterns of
emergence
PatriciaA.Pesavento1*, DalenAgnew2, MichaelK.Keel1 and KevinD.Woolard1
Abstract | Cancer is ubiquitous in wildlife, affecting animals from bivalves to pachyderms and
cetaceans. Reports of increasing frequency demonstrate that neoplasia is associated with
substantial mortality in wildlife species. Anthropogenic activities and global weather changes are
shaping new geographical limitations for many species, and alterations in living niches are
associated with visible examples of genetic bottlenecks, toxin exposures, oncogenic pathogens,
stress and immunosuppression, which can all contribute to cancers in wild species. Nations that
devote resources to monitoring the health of wildlife often do so for human-centric reasons,
including for the prediction of the potential for zoonotic disease, shared contaminants, chemicals
and medications, and for observing the effect of exposure from crowding and loss of habitat.
Given the increasing human footprint on land and in the sea, wildlife conservation should also
become a more important motivating factor. Greater attention to the patterns of the emergence
of wildlife cancer is imperative because growing numbers of species are existing at the interface
between humans and the environment, making wildlife sentinels for both animal and human
health. Therefore, monitoring wildlife cancers could offer interesting and novel insights into
potentially unique non-age-related mechanisms of carcinogenesis across species.
1Department of Pathology,
Microbiology, and
Immunology, School of
Veterinary Medicine,
University of California,
Davis, CA, USA.
2Veterinary Diagnostic
Laboratory, Department of
Pathobiology and Diagnostic
Investigation, College of
Veterinary Medicine,
Michigan State University,
East Lansing, MI, USA.
*e-mail: papesavento@
ucdavis.edu
https://doi.org/10.1038/
s41568-018-0045-0
www.nature.co m/nrc
Reviews
646
|
OCTOBER 2018
|
vOlumE 18
pf3
pf4
pf5
pf8
pf9
pfa
pfd
pfe
pff

Partial preview of the text

Download Cancer in Wildlife: Understanding the Connection between Wildlife and Human Health and more Slides Pathology in PDF only on Docsity!

The prevalence and variety of neoplastic diseases in wild-

life are likely under-reported and are certainly understud-

ied. Phenotypically similar forms of cancer exist across

disparate species, underscoring the universal nature of

cancer. A better understanding of both shared and unique

aspects of cancer across wildlife species will undoubtedly

shed light on potential mechanisms of oncogenesis. This

comparison may even untangle some of the complexities

regarding human cancer. Certainly, the large number of

animal populations inhabiting distinct environments pro-

vide many natural ‘experiments’ such as that examined

by Mǿller et al.^1 , where cancer in a broad spectrum of

bird species was shown to be tightly linked to the natu-

ral history and immunity of each species. Wild animals

are profoundly influenced not only by selective pressures

within their own species but also by their interface with

humans. Information gained from monitoring diseases in

free-ranging wildlife is of paramount importance, as we

often occupy the same habitat. A growing human popula-

tion places increasing demand on many shared resources.

For example, globally, water resources are being limited by

both access (human development) and a shrinking sup-

ply, so that the demand for adequate wastewater treatment

will become increasingly important over the coming dec-

ades (Fig. 1). Furthermore, an increasing human popula-

tion will usher in global changes that wildlife already face,

such as crowding, mixing of subpopulations and increased

environmental complexity. How these changes modulate

the pathophysiology of cancer in wildlife species will also

be extremely important to identify, as the implications for

wildlife and human health are likely to be similar.

This Review describes some of the cancers affecting

wildlife populations and what we know about the role of

infectious agents, environmental toxins and reproductive

or metabolic disturbances in contributing to the onco-

genic processes (Table 1). Although neoplasia occurs in

nearly all multicellular organisms^2 , this Review primarily

addresses instances of such disease in wild vertebrates.

We also describe new examples of spontaneous cancers

that have been identified in both captive and wildlife spe-

cies since the potential impact of cancer on wildlife was

first reviewed by McAloose and Newton^3. Such diseases

are an increasing concern for protected and endangered

species and are a unique frame of reference for pressures

that drive neoplastic transformation.

Pathogen-driven cancer

Infections in humans are considered causal for approxi-

mately 15% of the global cancer burden^2. Owing to a lack

of vigilant screening, which is often applied to human

cancer, the number of wildlife cancers caused by infec-

tion is not known, nor has it been possible to determine

whether the number of cancers associated with patho-

gens is similar in humans and wildlife. However, there

Cancer in wildlife: patterns of

emergence

Patricia A. Pesavento 1 *, Dalen Agnew 2 , Michael K. Keel 1 and Kevin D. Woolard 1

Abstract | Cancer is ubiquitous in wildlife, affecting animals from bivalves to pachyderms and

cetaceans. Reports of increasing frequency demonstrate that neoplasia is associated with

substantial mortality in wildlife species. Anthropogenic activities and global weather changes are

shaping new geographical limitations for many species, and alterations in living niches are

associated with visible examples of genetic bottlenecks, toxin exposures, oncogenic pathogens,

stress and immunosuppression, which can all contribute to cancers in wild species. Nations that

devote resources to monitoring the health of wildlife often do so for human-centric reasons,

including for the prediction of the potential for zoonotic disease, shared contaminants, chemicals

and medications, and for observing the effect of exposure from crowding and loss of habitat.

Given the increasing human footprint on land and in the sea, wildlife conservation should also

become a more important motivating factor. Greater attention to the patterns of the emergence

of wildlife cancer is imperative because growing numbers of species are existing at the interface

between humans and the environment, making wildlife sentinels for both animal and human

health. Therefore, monitoring wildlife cancers could offer interesting and novel insights into

potentially unique non-age-related mechanisms of carcinogenesis across species.

1 Department of Pathology,

Microbiology, and

Immunology, School of

Veterinary Medicine,

University of California,

Davis, CA, USA.

2 Veterinary Diagnostic

Laboratory, Department of

Pathobiology and Diagnostic

Investigation, College of

Veterinary Medicine,

Michigan State University,

East Lansing, MI, USA.

*e-mail: papesavento@

ucdavis.edu

https://doi.org/10.1038/ s41568-018-0045-

www.nature.com/nrc

646 |^ OCTOBER 2018 |^ vOlumE 18

is sound reasoning to hypothesize that wildlife might

harbour an even greater burden of pathogen-associated

cancer as a result of multiple factors, such as a lack of

prophylaxis, higher burden of infectious disease, chronic

stress and crowding. Despite research lagging in the

identification of oncogenic microorganisms (viruses,

bacteria or parasites) in domestic and non-domestic

animal species, recent molecular strategies have been

successful in revealing not only microorganism associa-

tion with certain wildlife cancers but also interesting and

novel patterns of pathogenesis.

Paradoxically, cancers associated with microorgan-

isms were first described in animals, and subsequent

research led the way for recognition of this common

mechanism of oncogenesis in humans 4. Studies of

cancers with an infectious origin have provided insights

into cell proliferation and pathogen lifestyles that chal-

lenged pre-existing dogmas of pathogen-driven diseases

(box 1); however, proving transformation by a pathogen

in the laboratory versus a naturally occurring infection in

the wild is quite distinct 5,6^. Therefore, uncovering

pathogen-driven oncogenesis is challenging, even with

extensive epidemiology and considerable research invest-

ment. For example, documenting the role of human pap-

illomavirus (PV) in cervical cancer beyond a reasonable

doubt or as a necessary cause took almost 50 years of

epidemiology, seroepidemiology, molecular investigation

and pathogenesis studies^7. Despite the sizeable challenges

in demonstrating causality in wildlife disease, these types

of studies provide valuable insight for global health.

1950 2017

Growth in suburbia

In human populated areas, there is exposure to agricultural, landscaping, food, drug and sewage contaminants.

Transport links limit roaming range and could limit mating choices, which has the potential to create genetic bottlenecks, even in non-endangered species.

Consistent water sources are important and can be either non-treated or treated. Treated water can be inappropriate for drinking.

Are animals in suburbia more or less stressed than their wild counterparts? Research on the effects of stress in wildlife is limited.

Exposure to pathogens, including those able to infect multiple host species, is more common in densely populated areas.

Coexistence of multiple species generates an intensive human–wildlife interface in which several species share shelter, water and food.

Poison

Fig. 1 |^ Oncogenic pressures at the human–animal interface using suburbia as an example. Suburbs or residential areas

are an expanding and unique living niche for wildlife. The figure shows the USA, where there has been a large and rapid

growth in suburbia from the 1950s until today. The pale orange areas depict the expansion from major and smaller cities

(grey). Raccoons, foxes, opossums and other mesopredators are much denser in population in suburban areas than in any

rural or wild landscape. Animals that live in suburbia share food and water sources, such as domestic waste, cat food,

fountains, sprinklers and feeders, with domestic animals and humans. Access to food and water is dictated by man-made

structures, such as roads, buildings or railways, that restrict natural roaming ranges. For example, raccoons in the wilderness

range up to 12,000 acres, whereas in suburbia in the USA, their range is only 3–4 blocks (the smallest area that is surrounded

by streets). Given these limitations in roaming, there is considerable potential for creating genetic islands or bottlenecks,

even in this non-endangered species. Higher population densities (of animals and humans) can support pathogen exposure,

stress, pathogen load and cross-species transmission. Pathogens that can jump species, for example, canine parvovirus and

canine distemper virus, could hypothetically be very successful in this intensive housing situation. Infection with both of these

pathogens can result in immunosuppression in the host and could, therefore, alter lifetime immunity to oncogenic pathogens.

Canine distemper virus

an enveloped single-stranded

negative RNa virus of the

family Paramyxoviridae related

to the viruses that cause

measles in humans. it is also

referred to as carnivore

distemper virus, as it causes

systemic disease in a wide

variety of animal families,

including domestic and wild

dogs, coyotes, foxes, pandas,

wolves, ferrets, skunks,

raccoons, large cats and

pinnipeds.

NATuRE REvIEwS |^ CANCeR

vOlumE 18 |^ OCTOBER 2018 |^647

destruction, but chronic infestation can be associated

with cancer. Inflammation associated with the ear mite

(Otodectes cynotis) has been implicated in the develop-

ment of ceruminous gland hyperplasia and carcinoma in

feral, adult Santa Catalina Island foxes (Urocyon littoralis

catalinae)^14. This is an important discovery in the unique

circumstance in which the fox population is isolated on

a single island and can therefore be tracked and treated.

Because these foxes are Near Threatened (International

Union for Conservation of Nature (IUCN) Red List),

a field trial was conducted to assess the impact of treat-

ment of ceruminous gland hyperplasia with the pes-

ticide acaricide, which kills ticks and mites 15. Initially,

these foxes had a 100% mite infestation rate, half of

which developed ear tumours. Treatment significantly

reduced hyperplasia compared with untreated controls^15.

The researchers proposed that the long-term presence

of mites is associated with epithelial hyperplasia, and

in support of causality, removal of the parasite burden

resulted in reversal of tumour development. Ultimately,

these findings have altered management decisions,

chronic gastritis in humans are all considered mecha-

nisms for bacterial Helicobacter-associated gastric car-

cinoma and B cell lymphoma^10. Yet, H. pylori-encoded

cytotoxicity-associated immunodominant antigen

(CagA) can directly manipulate host proliferation and

apoptotic pathways 11,12^. Chronic gastritis associated

with Helicobacter infection that results in adenocarci-

noma has been reproduced under laboratory condi-

tions in infected ferrets (Mustela putorius furo), gerbils

(Meriones unguiculatus) and hamsters (Mesocricetus

auratus), and there is some evidence to suggest that

mucosa-associated lymphoid tissue (MALT) lymphoma

is associated with H. heilmannii infection in big cats

such as tigers (Panthera tigris) and cheetahs (Acinonyx

jubatus)^13. Taken together, the available data suggest that

susceptibility to inflammation and subsequent neoplasia

vary among different hosts as well as among different

Helicobacter species.

Wild animals commonly have a burden of macro-

parasites (such as helminths and arthropods). The con-

sequences of infestation can be direct, such as tissue

Box 1 | Oncovirus–host co-evolution

viral taxonomists estimate that for most persistent oncogenic viruses, co-evolution with their host animal lineage has occurred slowly over millions of years. For example, papillomaviruses (Pvs) are an ancient virus group containing 49 genera and over 300 virus species in animals ranging from fish to marine birds and mammals (http://pave.niaid.nih.gov/). In the vast majority of Pv infections, the infected host does not exhibit noticeable symptoms, but disruption of this apparently metastable coexistence can occur, for example, under conditions of immunosuppression. unfortunately, a lack of widespread infection in healthy animals has limited our understanding of how persistent infections become oncogenic or have other potential sequelae. There is also no experimental model that can provide insights into the often decades-long metastable state of a potentially oncogenic viral infection. Koch’s postulates for establishing a relationship between a pathogen and a given disease are based upon the criteria that the pathogen should cause disease in all infected individuals and that the disease is not caused by other agents^168. As Pvs, polyomaviruses and herpesviruses typically infect the entire human population and cause cancer in, at most, only a small fraction of infected individuals, they cannot be explained by Koch’s postulates. Cancer virology has instead utilized Hill’s criteria for causation, whereby the more certain an association between a factor and an effect is, the greater the probability it is a causal relationship, an example of this being the finding that viral sequences are associated with particular forms of cancer^169. Careful analysis of outbreaks and species-specific cancers can identify tumour-associated pathogens, and although we cannot infer causation lightly, they do provide insight into host susceptibility, the environment and the evolution of pathogens. The image in part a shows the ultrastructure of a chondroblast surrounded by scant chondroid matrix from a seabird infected with an avian Pv. In part b , a higher power magnification of part a , paracrystalline arrays of Pv (virus diameter 46–48 nm) can be seen in the nucleus. Scale bar = 2 μm.

a

b

NATuRE REvIEwS |^ CANCeR

vOlumE 18 |^ OCTOBER 2018 |^649

with any fox caught during routine trapping now being

treated with acaricide.

In another example, ring-necked pheasants

(Phasianus colchicus) with intestinal nematode infesta-

tion (Heterakis spp.) frequently develop severe nodular

typhlocolitis, a chronic inflammatory condition that can

develop into local and metastatic sarcomas^16. These have

been primarily identified in captive pheasants, suggest-

ing contributions from stress, higher parasite loads and

altered diets^16.

Direct cell transformation

In tissue culture, cellular transformation can be achieved

by disruption of a limited number of cellular regulatory

pathways^17. Some viruses encode proteins that can target

regulatory pathways and disrupt cell cycle, metabolic or

apoptotic control, which can all contribute to transforma-

tion. There are commonalities among oncogenic patho-

gens that directly transform cells and that have served as

a guide for oncogenic viral discovery. These common-

alities are that pathogens associated with cancer have

been persistent (usually lifelong) infections, regardless

of their host species; transformation is rare as a sequela of

infection compared with the prevalence of infected hosts

and is most likely to occur in immunosuppressed ani-

mals; and either tissue stem cells are the initial target cells

of oncogenic viruses or ‘stemness’ results from infection.

However, these are generalizations and not limited,

among pathogens, to viruses, although viruses form the

vast majority of oncogenic pathogens. The causal onco-

genic viral families that have emerged, in all animals

and humans, include Papillomaviridae, Polyomaviridae,

Retroviridae and Herpesviridae. Reduced genetic diver-

sity that arises when a new population is formed from a

small number of individuals from a much larger popula-

tion (known as the founder effect) appears to have impli-

cations for susceptibility to pathogens that can result in

neoplasia. For example, more highly inbred California

sea lions (Zalophus californianus) are susceptible to bac-

terial infection, helminthiasis, intoxication by harmful

algal blooms and neoplasia; among these susceptibilities,

the highest correlation was observed between the degree

of genetic relatedness and the presence of urogenital car-

cinoma^18. However, it should be noted that the overriding

mechanism for urogenital carcinogenesis in California

sea lions is not yet known, and development of these

tumours could involve virus infection, environmental

toxins or a combination of these and other factors.

Polyomaviruses and papillomaviruses. Polyomavirus

(PyV) infections appear nearly ubiquitous in wildlife

and have been identified in mammals, invertebrates,

fish, amphibians, reptiles and birds^19. Well-studied PyVs

include the simian virus 40 (SV40) and murine PyV in

laboratory animals and Merkel cell PyV (MCPyV), JC

PyV and BK PyV in humans 20,21^. All PyVs encode the

potentially oncogenic tumour antigen (T antigen or T Ag)

proteins, which have a pivotal function in both virus

replication and control of the host cell cycle19,21. On the

basis of experimental studies, many viruses in this family

cause tumours in immunosuppressed animals, but the

vast majority of naturally occurring infections are clin-

ically silent, with exceedingly rare transformation 20,^.

Examples of tumours caused by PyV infection include

Merkel cell skin carcinomas in humans22,23^ (box 2), brain

tumours in wild raccoons (Procyon lotor) 24,25^ and a

variety of tumours in laboratory animal species26–28.

Raccoon polyomavirus (RacPyV) associated with the

occurrence of cancer was first reported in 2013 (ReF.^24 ).

There is now strong evidence that RacPyV causes a series

of recently identified, naturally occurring neuroglial

tumours in raccoons. Tumours in raccoons are not com-

mon, but to date, 23 cases of tumours have been recorded

on the West Coast of the USA and Canada^29. It is unlikely

that these tumours occurred before 2013, because nec-

ropsy, including evaluation of the brain, is commonly

performed on raccoons in many diagnostic laboratories

in the USA because they are known to carry rabies on

the East Coast. In 3 years (2013–2016), the number of

raccoons with tumours represented 15% of the total

number of raccoons that had necropsies performed

regionally (northern California, USA). The number is

heavily biased because animals submitted for necropsy

evaluation are more likely to have clinical signs referable

to the neurological system and to be unafraid of humans

but is notably high nonetheless in the context of the pre-

vious and fairly consistent screening by necropsy. The

tumours are invariably found crossing the nasal–brain

barrier and affecting the olfactory tract and frontal brain

lobes^24. RacPyV has been detectable in all the identified

tumours to date and is present in high copy number as

measured by both quantitative PCR and in situ hybridi-

zation25,30. Metastases occurred in only a single case, and

virus was detectable in all neoplastic cells, within all met-

astatic foci. Primary cell culture studies suggest that the

target cell of transformation is the multipotent neuroglial

stem cells of the subventricular zone of the brain^30 (Fig. 2).

Transcriptome analyses have revealed that T Ags are

highly expressed in these tumours, but structural genes

Box 2 | Merkel cell polyomavirus

merkel cell polyomavirus (mCPyv) is a common, almost universal infection in humans^170. In rare cases, mCPyv is found to be the cause of a very aggressive form of skin cancer called merkel cell carcinoma, where viral DNA is integrated into the tumour genome^27. Approximately half of the mCPyv genome encodes structural viral proteins (vPs), and the other half encodes viral tumour antigens (T antigens or T Ags). Among several established criteria for causality is the key finding that vP expression is uncoupled (absent) compared with the dysregulated or upregulated expression of T Ags. Expression of a T Ag alone is sufficient, in experimental models, to drive tumour formation^20. Several viral T Ag–host protein interactions occur, among which are the binding and inhibition of the tumour suppressors RB, p53 and protein phosphatase 2A (PP2A) (see the figure). Analyses of the genomes of polyomaviral-associated versus non-viral-associated merkel cell tumours have demonstrated that virus-negative tumours have a high burden of somatic gene mutations, whereas mCPyv-positive tumours have few. Presumably, the T Ags of mCPyv are capable of efficiently hijacking cellular processes to drive tumorigenesis in a manner that is comparable to inducing somatic gene alterations^23.

p53 binding site

Origin binding domain

RB DnaJ binding site

PP2A binding site

CR1 Helicase

www.nature.com/nrc

650 |^ OCTOBER 2018 |^ vOlumE 18

urogenital carcinoma in a captive South American fur

seal^53. However, the criteria for causality remain limited

to detection of the viral genome in tumour tissue. Like

all persistent viruses, establishment of causality faces the

obstacle that the virus is also detected in animals with-

out cancer 54. Environmental toxin exposure, a genetic

basis and variation in hormone receptor expression are

other factors that have also been identified as potential

contributors to urogenital carcinoma occurrence in

Californian sea lions55,56.

Fibropapillomatosis found specifically in marine tur-

tles is associated with variants of Chelonid alphaherpes-

virus 5 (ChHV5)^57. Dermal fibropapillomas and fibromas

have been documented to various degrees in all marine

turtle species, but the disease is most prevalent in green

turtles (Chelonia mydas) 3,58^. Although most tumours

appear benign, they can impair vision, obstruct the mouth

or cloaca and impede swimming. Moreover, the forma-

tion of tumours in internal organs occurs in severe cases.

Fibropapillomatosis is transmissible by experimental

infection 59 , suggesting that the associated herpesvirus

plays a primary aetiological role; however, environmen-

tal cofactors have also been proposed to contribute60–62.

Regardless of the role of the virus, ecology studies on

ocean life have had to consider a new dimension in

potential transmission routes because water immersion

is such a uniquely shared space. Waterborne infection is

possible, as herpesviruses can be stable in water^57 , and

ChHV5 DNA has been detected in marine leeches (genus

Ozobranchus) in breeding grounds of the green turtle^63.

Infection of wild mountain gorillas with a specific

strain of lymphocryptovirus 1 (GbbLCV-1) was recently

discovered^64. GbbLCV-1 is closely related to the human,

oncogenic Epstein–Barr virus (EBV), thus explaining

why previous reports detected ubiquitous EBV expo-

sure, by serological assay, among gorillas 65. Similar to

EBV infection in humans, GbbLCV-1 primary infection

occurs in infants, and latent infection was detected in

peripheral white blood cells^64. B cell lymphomas, a subset

of which are caused by EBV infection in humans, were

diagnosed in a single mountain gorilla within the group

examined in this particular study, and GbbLCV-1 was

detected by PCR in blood collected from this individual

at the time of necropsy^64.

Viral integration

Both DNA and RNA viruses can disrupt the host genome

by integration, but examples of integration causing can-

cer among wildlife are currently limited to retroviruses.

Studies of transmissible exogenous retroviruses have led

to the recognition that all animals carry a large number

of integrated endogenous retroviruses (ERVs)^66. ERVs

make up 5% and 4.7% of the mouse and human genome,

respectively, and many of these so-called fossil viruses

retain transcriptional and (at least in mice) replication

capacity^67. Given the ability of ERVs to move within the

genome, distinguishing them from exogenous viruses

and identifying their potential contributions to cancer

are vital. There are important examples of studies in

wildlife species that have achieved a requisite combina-

tion of epidemiological, transmission and in vitro data

supporting causality.

Jaagsiekte sheep retrovirus (JSRV) and enzootic nasal

tumour virus (ENTV) are the causative agents of pul-

monary and nasal carcinomas, respectively, in domestic

sheep and goats 68. Feral bighorn sheep spontaneously

develop similar paranasal sinus tumours, and histo-

logical features and geographic clustering support an

infectious aetiology69,70. The most convincing evidence

for an infectious origin is that the tumours in bighorn

sheep are transmissible through inoculation with a

cell-free filtrate: within 18 months of inoculation, one

of four bighorn sheep and, intriguingly, four of four

domestic sheep developed tumours within ethmoid

sinuses or nasal conchae^69. However, PCR and immuno-

histochemistry failed to demonstrate the presence of

ENTV or JSRV, so metagenomics, deep-sequencing and

cell culture studies designed to uncover a viral cause are

ongoing (S. L. Quackenbush, personal communication).

Gammaretroviruses (KoRVs) found in koalas

(Phascolarctos cinereus) are in a unique state of evolu-

tionary adaptation. At least three variants have been

identified, KoRV-A, KoRV-B and KoRV-J, each with

varying associations with cancer 71. KoRV-A is endoge-

nous and is rarely associated with cancer, while KoRV-B

(a presently exogenous virus) is associated with lym-

phoma 72. The lack of genetic diversity in the KoRV-A

endogenous virus appears to be associated with its

inability to induce tumours 73. Whether KoRV-J is also

associated with increased cancer incidence is still being

investigated. Therefore, KoRVs might provide the

opportunity to witness a retrovirus becoming embedded

in the genome of its host, a phenomenon that has been

only theorized in other species^74.

Herpesviridae, Papillomaviridae and Retroviridae are

all associated with the development of tumours in fish,

but retroviruses are particularly well represented, with

suggested involvement in 13 different piscine tumours^75.

Compelling evidence (in the form of sequencing, iso-

lation and transmission studies) for a causal role exists

only for a few of these. Walleye dermal sarcoma virus

(WDSV), an exogenous retrovirus that causes dermal

sarcoma in walleye (Sander vitreus), encodes a cyclin-

C-like homologue postulated to interfere with cell cycle

regulation76,77. Experimental transmission with cell-free

tumour filtrates has been successful by numerous trans-

mission routes (including intramuscular, oral and top-

ical)^78 and has expanded the potential host range of the

virus to other fish species^79. The study of fish oncogenic

retroviruses has an interesting and potentially exploit-

able angle on host tumour immunity. Many of these

retroviral-associated fish tumours develop season-

ally, and their growth is linked to temperature-driven

changes in the immune response of these fish (which

are ectotherms), providing potentially valuable models

of tumour regression^80.

Toxin-related cancer

Wildlife populations encounter numerous anthropo-

genic pollutants, including industrial and agricultural

waste, radiation and, increasingly, pharmaceutical con-

taminants 81,82^. These pollutants coexist with biota in a

complex ecosystem. As we move towards a more inten-

sive agrarian and civic lifestyle worldwide, concentrated

Cloaca

The caudal opening in reptiles,

amphibians and birds used for

digestive, reproductive and

urinary tract excretions.

Retroviruses

RNa viruses that utilize reverse

transcriptase to generate a

complementary DNa strand

from the RNa template, which

is then integrated into the

genome of the infected cell.

Nasal conchae

also called nasal turbinates.

Convoluted, curled thin bones

covered by respiratory

epithelium that protrude into

the breathing passage of

animals.

Ectotherms

animals dependent on

exogenous heat to maintain

body temperature.

Agrarian

Relating to farmland,

agriculture or the cultivation of

land for crops.

www.nature.com/nrc

652 |^ OCTOBER 2018 |^ vOlumE 18

exposures are more likely for all. Ageing infrastructure,

a global lack of governmental priority and regional water

shortages may see the spread of environmental contam-

inants into water bodies or groundwater, affecting not

only humans but also wildlife. Thus, studying the impact

of environmental toxicology on cancer in wildlife has the

potential not only to identify causal relationships that

may affect humans but also to compare affected ver-

sus resistant species to identify evolutionary strategies

inhibiting oncogenesis. Wildlife now frequently exist

within changing environments, where large regions

of mixed use (industrial, agricultural and residential)

are increasingly becoming divided into more focused

microenvironments, which reduce the ecosystem com-

plexity and biodiversity. How this change will affect

cancer incidence and how it relates to concentrated

environmental pollution highlight the importance of

disease surveillance in wild fauna.

Pollution may affect the cancer prevalence across

wildlife species by diverse means. Historically, empha-

sis has been placed on the direct roles of xenobiotics

or radiation-inducing somatic mutations that disrupt

oncogenes or tumour suppressor genes 83. While these

are undoubtedly important, with mounting cases of

documented contribution to wildlife cancer, these are

perhaps over-represented because they are the easiest to

identify as causal84–86.

Organochlorines and cancer

Beginning in the 1980s, hundreds of stranded California

sea lions have been examined by the marine stranding

network^55. These animals display a strikingly high prev-

alence of metastatic urogenital carcinomas, with up to

26% of adults having these tumours 55. This aggressive

and metastatic tumour appears to arise from the cervix

and vagina in females and from the penis, prepuce and

urethra in males. It exhibits similarities to human cervical

intraepithelial neoplasia (CIN), with lesions that resem-

ble human CIN grades I to III 55,87,88^. Organochlorines

(OCs) and polycyclic aromatic hydrocarbons (PAHs)

are putative endocrine disrupting compounds (EDCs) that

have both been associated with the development of these

tumours 89. While the mechanism of cancer formation

in these animals remains complex (including potential

contribution from herpesvirus infection), OCs remain

relevant as potential factors, as they are environmentally

persistent molecules that bioaccumulate and biomagnify

within predators^90.

Additional evidence for a contributory role of OCs

in wildlife cancer can be found within beluga whales

(Delphinapterus leucas) of the St. Lawrence Estuary,

Quebec, Canada 3,90^. Over half of all cancers identified

worldwide in toothed whales occur within this iso-

lated population of approximately 900 beluga whales 90.

Cancers are most often of gastrointestinal origin, which

is important given the whales propensity to feed within

the benthic zone, which is associated with contaminated

sediment^90. While OCs have largely been banned in the

USA since the 1980s, their environmental persistence,

particularly in aquatic sediment, remains a threat to

wildlife^91. Furthermore, in many countries, OCs are con-

tinually used as pesticides owing to their low cost and

efficacy. Sediment containing high levels of OCs from

regional runoff is often identified within the gastric con-

tents of beluga whales suffering from these malignan-

cies^90. These findings are of particular interest as many

scientists have linked OC exposure to gastric cancer in

humans 92,93^. Accumulated OCs and PAHs in sediment

have also been associated with increased liver and skin

tumours in a variety of fish, including English sole,

European flounder and brown bullhead 90,94^. Identified

and proposed mechanisms of carcinogenesis from OCs

and PAHs are shown in Fig. 3.

Endocrine disrupting compounds

EDCs may mimic vertebrate sex hormones, functioning

as either an agonist or antagonist within a particular hor-

monal pathway, or may alter hormonal action through

perturbations in hormone production, receptor bind-

ing or metabolism 95. The World Health Organization

(WHO) has recognized EDCs as a widespread envi-

ronmental contaminant owing to their multiple trans-

mission routes — ingestion, inhalation, trans-dermal

contact and trans-placental spread 96. It has designated

EDCs as a global emerging issue, with concern over the

effects of these compounds on fertility, birth defects and

cancer formation. The strongest link between EDCs

and human cancer has been demonstrated with breast

cancer, but potential involvement of biologically active

EDCs in both male and female reproductive tumours is

also a concern^97.

Wildlife encounter EDCs in much the same way that

humans do — through ingestion, inhalation or skin

contact. They are exposed to widely varying amounts

of EDCs, and documentation of regional variation is

important because, at nanomolar concentrations, EDCs

primarily act through binding of nuclear receptors, but

at millimolar concentrations, they may affect cell behav-

iour through epigenetic disruption^98. EDCs implicated

in wildlife cancer include dioxins, which are associated

with an increasing prevalence of non-Hodgkin lym-

phoma in humans 99. Dioxins have been detected from

blubber and liver tissue in a single case of naturally

occurring hepatosplenic lymphoma in a bottlenose dol-

phin and can drive dolphin lymphocyte proliferation

in vitro 100. In addition, examination of white suckers,

a freshwater fish species in Lake Michigan tributaries,

also indicates increased hepatic tumour prevalence

associated with EDC levels^101.

Of increasing interest is the role of epigenetic dis-

ruption caused by EDCs. Indeed, this is now viewed as

one possible explanation for their persistent biological

effects. As endocrine hormonal mimics, EDCs princi-

pally act through binding of nuclear receptors, such as

the oestrogen, androgen or thyroid receptors, to drive

gene transcription 102. Nuclear receptor activation may

increase or decrease expression of key epigenetic path-

ways, including expression of DNA methyltransferases

(DNMTs) and genes involved in the conversion of

5-methylcytosine (5-mC; such as tet methylcytosine

dioxygenase 1 (TET1) and TET2) and in histone mod-

ifications (such as histone acetyltransferase KAT2A

(also known as GCN5) and histone deacetylase 1

(HDAC1))103,104. Both nuclear receptors and EDCs may

Xenobiotics

any substance (synthetic or

natural) that is not naturally

present in the body of an

organism.

Endocrine disrupting

compounds

(eDCs). a broad category of

mostly man-made substances

that are present in pesticides,

plastics, personal care

products, metals and

pharmaceuticals, among many

other items, which result in

altered hormonal activity via

agonistic and antagonistic

receptor binding.

Bioaccumulate

When a substance becomes

concentrated within the body

of a living thing. if the source of

the substance is from water,

this is specifically referred to as

bioconcentration.

Biomagnify

The increasing concentration of

a substance within the tissues

of an organism acquired

through predatory acquisition

(a food chain).

Benthic zone

The lowest ecological regions

of a body of water, such as the

sediment surface.

NATuRE REvIEwS |^ CANCeR

vOlumE 18 |^ OCTOBER 2018 |^653

reproductive system may have increased impact for

small populations. Smaller populations of free-ranging

animals replicate, to some degree, the limitations of zoo

populations. Captive animals receive medical treatment,

are on standardized diets and are held in high popula-

tion densities, and their reproduction is often controlled

medically or by separation; however, increasingly,

wild populations are being more individually managed

than in the past, making the differences between wild

populations and captive animals less disparate.

Limited genetic diversity in small gene pools can lead

to decreased fitness in populations but can also allow

the rapid increase in prevalence of deleterious genes^111.

Small populations can also limit breeding opportunities,

leading to abnormal social structures, fewer pregnancies,

as well as pregnancies that occur later in the animal’s

life, and a corresponding decrease in overall reproduc-

tive health 112,113^. The concept of evolutionary mismatch

provides a model for these types of population effects

and can be likened to the neoplastic diseases associ-

ated with nulliparity in women114,115. Many fragmented

populations are also exposed to environments contami-

nated with toxins affecting reproduction, making it dif-

ficult to tease out the relative impact each factor might

have. Furthermore, rates of potentially cancer-causing

pathogen transmission might be anticipated to be dif-

ferent between smaller, denser populations and larger,

free-ranging populations. There might also be a greater

exposure to zoonotic agents as humans impinge on

these populations.

An example of the effects of limited genetic diversity

is seen in the Florida panther (Puma concolor coryi), an

Endangered (U.S. Fish and Wildlife Service Endangered

Species) and inbred species, with one study showing

that 11 of 17 (65%) individuals had cryptorchidism^116.

Cryptorchidism is a highly heritable trait that is also

associated with exposure to endocrine disruptors,

such as 1,1-dichloro-2,2-bis(p-chlorophenyl)ethylene

(p,p’-DDE), a breakdown product of dichlorodiphenyl-

trichloroethane (DDT), mercury and PCBs 117. In

addition to the negative effects on spermatogenesis, cryp-

torchidism is linked to the development of seminomas

and Sertoli cell tumours, although no testicular cancer has

been specifically identified in the limited surveys so far

carried out on Florida panthers116,118. A similar high rate

of cryptorchidism has been reported in a population of

Sitka black-tailed deer (Odocoileus hemionus sitkensis)

restricted geographically to Kodiak Island (Alaska,

USA), although reproductive system cancers were also

not identified in this limited observational study^119.

A Near Threatened (IUCN Red List) species, the

jaguar (Panthera onca), whose population is carefully

managed in captivity, has shown an alarmingly high

incidence of ovarian carcinoma in USA captive pop-

ulations, leading to death or euthanasia 120–124^. These

findings are reminiscent of women with mutations in

BRCA1 or BRCA2 genes 125 , and although a large num-

ber of candidate genes have now been found in jaguars,

further analysis will be required to identify specific

mutations in the genes correlating with tumorigenesis^123.

Similar cancers have been only anecdotally reported in

free-ranging jaguars, highlighting the challenges of such

investigations in the field and the value of zoo collections

in better understanding the interactions of genetics,

environment and disease. Furthermore, while genetic

analyses are possible from small samples of skin, blood,

hair or even faeces, substantial political impediments to

sharing biological materials across international bor-

ders make it difficult or impossible to effectively col-

laborate with conservation scientists in countries where

these species are free-ranging. Laws intended to protect

endangered species can often make the science required

to effect such protection unfeasible^126.

A more subtle effect of less frequent pregnancy is a

reduction in future fertility, an increasingly recognized

phenomenon of evolutionary mismatch seen in captive

populations of wild mammals and often described as ‘use

it or lose it’115,127. Hormonal stimulation, particularly by

oestrogen, can produce marked proliferative responses

from the uterine, cervical, vaginal and mammary epi-

thelium and smooth muscle^128. Such stimulation would

be infrequent in the wild, as females would become

pregnant on their first estrous cycle, spend a consider-

able time span being pregnant or lactating and would

then become pregnant again shortly after they began to

cycle again. One study of the reproductive pathology of

white rhinoceros estimates that a free-ranging female

would have 30–90 estrous cycles in a typical lifetime,

with the majority of its adult lifespan spent in lactation

or pregnancy 129. However, females held in captivity

in a non-reproductive state (that is, separated from a

fertile male) would experience greater than 300 cycles,

exposing the reproductive tract to an oestrogen-rich,

proliferative hormonal milieu up to ten times more

frequently. This leads to progressively decreased fertil-

ity and an increase in proliferative and neoplastic dis-

eases, collectively described as asymmetric reproductive

ageing^129 (Fig. 4).

An example highlighting how pregnancy can provide

a natural protective mechanism against asymmetric

reproductive ageing processes is the observation that

leiomyomas and cystic endometrial hyperplasia have been

reported in nearly all nulliparous captive white rhinos

yet are very rare in multiparous reproductively active

animals 129. This scenario is poignantly demonstrated

by the critically endangered northern white rhino

(Ceratotherium simum cottoni), in which such lesions

have been described in nearly all remaining females,

making this species functionally extinct 129. A similar

prevalence of these lesions has also been reported in the

Indian rhinoceros (Rhinoceros unicornis) 130. Captive,

aged chimpanzees, largely kept in non-reproductive

conditions, have a leiomyoma incidence of 40–62%

(4/10 in one study and 20/32 in another) 131,132^ , yet the

number in wild populations has not been accurately

estimated. In addition to environmental toxins, the inci-

dence of reproductive system cancers in manatees may

also be associated with increasing urban encroachment

and fragmentation of their range, leading to isolation

of individuals and difficulty in finding mates 133. The

continued monitoring of a wild population of moun-

tain gorillas in central Africa provides one of the few

opportunities to compare free-ranging and captive

populations: interestingly, no leiomyomas have been

Evolutionary mismatch

a concept in evolutionary

biology referring to the

presence of once beneficial

traits in a population that,

owing to rapid environmental

change, are no longer

beneficial but harmful.

Cryptorchidism

The absence of one or both

testes from the scrotum,

usually resulting from a failure

to descend during

development.

Seminomas

a type of germ cell tumour of

the testicle.

Sertoli cell tumours

a sex cord-gonadal stromal

tumour composed of Sertoli

cells, which line the

seminiferous tubules and help

in the development of sperm.

These are typically benign and

often hormonally active.

Estrous cycle

The recurring cyclic variation in

reproductive hormones (for

example, oestrogen and

progesterone) in the

mammalian female that

controls behaviour,

reproductive organ

morphology, ovulation and

conception.

Leiomyomas

When present in the

reproductive tract (vagina,

cervix, uterus, oviduct or

ovary), these are hormonally

responsive benign smooth

muscle tumours. in humans,

these are also known as

fibroids.

Cystic endometrial

hyperplasia

a condition of excessive

proliferation of the glandular

epithelium of the uterus,

typically associated with

excessive progesterone and/or

oestrogen stimulation.

Nulliparous

an animal that has never given

birth.

Multiparous

an animal that has given birth

multiple times.

NATuRE REvIEwS |^ CANCeR

vOlumE 18 |^ OCTOBER 2018 |^655

reported in the free-ranging population of approxi-

mately 900 individuals, while one of nine reproductive

tracts available for evaluation from captive gorillas had

leiomyomas^131. Similar effects have been seen in women,

where early pregnancy and numbers of pregnancies sig-

nificantly decrease the risk of some subtypes of breast

cancers later in life, and multiparous women have a

greatly reduced risk of leiomyomas than nulliparous

women134–136.

In sexually intact domestic dogs and cats, when they

cycle but do not conceive, they are exposed not only to

high levels of oestrogen but also to high endogenous lev-

els of progesterone during diestrus. In these species, and

in wildlife species with similar cycles such as African

painted dogs or lions, the endogenous high hormonal

levels are also suspected to predispose these animals to

increased mammary cancer 137,138^. The use of progestin

contraceptives in zoos only increases this effect, resulting

in many mammary cancers in tigers as well as other car-

nivores, unless animals are periodically allowed to breed

and carry a fetus to term^139. While contraceptives are not

commonly used in free-ranging populations, similar

effects might be expected in some species if breeding

opportunities are less frequent.

Another manifestation of human impact on animal

populations may be the loss of predators in many eco-

systems; large predators are often targeted by humans

as a result of hunting for sport, out of fear for personal

safety or to protect livestock 140. It is likely that in these

cases, prey animals will have longer lifespans, with

corresponding increases in cancer, similar to humans

when the reduction of infectious diseases and modern

health care increased life expectancies. An example of

this is a semi-wild population of fallow deer managed

in an urban park island effectively free from predation

for over 100 years where population control was finally

achieved following sterilization of all the males (D. A.,

unpublished observations). Ten to 12 years after this

population control, many females were found to be

nulliparous or had only produced fawns much earlier

in their lifetime, and most had developed endometrial

hyperplasia, endometrial carcinoma, cervical carcino-

mas and leiomyomas. However, as might be expected,

multiple factors are likely to play a role, including age,

prolonged hormonal stimulation without pregnancy

(asymmetric reproductive ageing), potential toxins in

an urban environment and genetics (inbreeding).

Transmissible cancer

One of the most unique cancers in the animal and

human world is transmissible venereal tumour (TVT)

of the dog, which is also seen in other canids 3,141^. In

this disease, a neoplastic population of cells is trans-

mitted during coitus, upon which the cells implant on

the mucosa of the new host and persist until a vigorous

host immune response causes tumour regression 142,^.

Recent work shows that the tumour regression (either

spontaneous or following vincristine chemotherapy) is

sequential and dependent on host innate immunity 144.

After tumour development, the host innate immune sys-

tem is activated, leading to remodelling of host epithe-

lium, immune cell infiltration of the tumour, cell cycle

arrest within the tumour cells and, finally, repair of host

tissue damage. Specific cytokines (especially C-C motif

chemokine ligand 5 (CCL5; also known as RANTES))

were identified as key players, in addition to changes in

gene expression associated with methylation of tumour

cell DNA 144. Since 1876 when it was first described 145 ,

TVT was thought to be the only naturally transmissible

tumour in nature; however, in 1996, a new transmissi-

ble tumour was identified, affecting the already endan-

gered Tasmanian devil (Sarcophilus harrisii), known as

devil facial tumour disease (DFTD)^146. This tumour is

transmitted between devils during social interactions,

including mating. Such interactions are often violent

and may result in oral and facial wounds where the

tumour cells are implanted. Subsequent proliferation of

the tumours leads to disfigurement and facial damage

sufficient to cause starvation and death^147. Intriguingly,

tumours affect the most reproductively fit animals,

Diestrus

a non-receptive phase of the

estrous cycle dominated by

progesterone production.

Pregnant Non-pregnant

Natural state Hyperplasia Progression to cancer

Cancer

Parturition and endometrial remodelling

Conception and pregnancy

Repeated cycles of hormonal stimulation

Repeated cycles without pregnancy or remodelling

Inflammation and hyperplasia

Fig. 4 |^ Reproductive system cancers in wildlife. In the normal, natural state, adult wild animals are typically pregnant or

lactating. During brief non-pregnant states, animals have short hormonally driven estrous cycles, which typically result in

another pregnancy. Pregnancy, parturition and post-partum involution result in extensive remodelling of the uterine

endometrium and removal of hyperplastic tissue. However, if pregnancy does not occur regularly owing to factors such as

skewing of sex ratios, captivity or other abnormal pressures on a population, repeated estrous cycles may occur with

elevated and prolonged hormonal exposure, leading to hyperplasia without remodelling as might occur during

parturition and involution. This phenomenon of asynchronous reproductive ageing can lead to inflammation, infertility

and hormonally driven cancers, such as leiomyomas.

www.nature.com/nrc

656 |^ OCTOBER 2018 |^ vOlumE 18

In other cases, the origins are less certain. Regardless,

disease threats, including cancer, continue to place our

wild populations at risk.

Future directions and conclusions

Owing to the difficulty in acquiring case material dur-

ing an outbreak and the relative rarity of recognized

disease outbreaks in wild animals (both captive and

free-ranging), support for archiving of biological speci-

men collections is critical. This is important for animals

collected during mortality events, as well as collections

of samples from unaffected animals, as the latter can

provide very important baseline data. Potential dis-

ease outbreaks are often identified only after anecdo-

tal reports accumulate to a critical level and valuable

biological material has not been collected, discarded or

lost. Continuous monitoring of cancers in wild animals

is needed, as well as collation and archiving of tissues

(biobanks) to test the potential cause or causes of those

cancers^160. In addition, documentation of the wide diver-

sity of endocrine and reproductive patterns, develop-

ment of tools to estimate stress and studies of the genetic

or environmental associated immunosuppression in the

animal kingdom will improve our ability to correlate

disturbances in the natural history of individual species

with increased incidences of cancer.

The study of how evolutionary bottlenecks affect

cancer predisposition and outcome in wildlife also has

ramifications for human cancers. Founder effects have

been identified that predispose ethnic populations to

breast, ovarian, prostate and colorectal cancers161–164. As

our ability to interrogate complex data sets in the age of

‘big data’ grows, it is possible to evaluate how cancer is

affected by ancestral and spontaneous genetic polymor-

phisms as well as the environment. Favé et al. 165 recently

published an intriguing method for the computational

analysis of genetic and environmental interactions in

determining disease. Performing similar analyses across

wildlife cancer would undoubtedly reveal more infor-

mation regarding the role of putative tumour suppres-

sors and oncogenes and how our changing environment

dictates outcomes like cancer.

Tumour prevalence varies greatly across the wide

variety of animals, and studying wildlife cancer can

provide us with an appreciation of the varied mech-

anisms of oncogenesis. The so-called Peto’s paradox

asks why cancer incidence is not directly related to the

number of cells in an organism and lifespan 166. If we

believe that cancer exists as a result of the combined

influences of endogenous and exogenous genomic

stressors, chronic inflammation and infectious dis-

ease, the immune system, and inherent variation in

cell biology, assessing the contribution of multiple fac-

tors, such as pathogen infection and environmental

pollutants on cancer formation, is a prime example

of how wildlife oncology stands to address the ques-

tion of Peto’s paradox. An example highlighting this

is why California sea lions are susceptible to urogeni-

tal carcinoma, but other marine mammals that share

exposure to the same (or similar) environment are not.

Therefore, examining several animal species within a

particular ecosystem is likely to help reveal both mech-

anisms of cancer resistance as well as species-specific

biological processes that are requisite for cancer

formation 167.

The role wildlife can play as both a sentinel for novel

threats to humankind and as models for as yet unrec-

ognized or poorly understood mechanisms of cancer

development cannot be underestimated. Wild popula-

tions provide clues to the very basic questions regarding

the origin, evolution and likely consequences of cancer

on a population level. Studies of cancers in wild popu-

lations due to the interplay between host defences and

infectious agents, environmental toxins (natural and

man-made) and the disturbed proliferative capacity of

the reproductive tract can provide useful information for

human health and make the challenges associated with

this research worth the effort. Furthermore, the benefit

such a study provides for the preservation of the wildlife

in our shared world is all the more reason to turn our

investigative resources away from the bench and towards

the wilderness.

Published online 16 August 2018

  1. Møller, A. P., Erritzøe, J. & Soler, J. J. Life history, immunity, Peto’s paradox and tumours in birds. J. Evol. Biol. 30 , 960–967 (2017).

  2. Ewald, P. W. & Swain Ewald, H. A. Infection and cancer in multicellular organisms. Phil. Trans. R. Soc. B 370 , 20140224 (2015). This paper reviews the diversity of known pathogen-derived neoplasms among all types of multicellular organisms and also addresses the challenges in recognizing cancer in nature.

  3. McAloose, D. & Newton, A. L. Wildlife cancer: a conservation perspective. Nat. Rev. Cancer 9 , 517–526 (2009). This Review highlights how wildlife cancer is an important threat to some species and establishes tumours in wildlife as sentinels of host changes and the ecosystem.

  4. Rous, P. A transmissible avian neoplasm (Sarcoma of the common fowl). J. Exp. Med. 12 , 696– (1910).

  5. Fredricks, D. N. & Relman, D. A. Sequence-based identification of microbial pathogens: a reconsideration of Koch’s postulates. Clin. Microbiol. Rev. 9 , 18–33 (1996).

  6. Pagano, J. S. et al. Infectious agents and cancer: criteria for a causal relation. Semin. Cancer Biol. 14 , 453–471 (2004).

  7. zur Hausen, H. Papillomaviruses in the causation of human cancers — a brief historical account. Virology 384 , 260–265 (2009).

  8. Lempp, C. et al. Pathological findings in the red fox (Vulpes vulpes), stone marten (Martes foina) and raccoon dog (Nyctereutes procyonoides), with special emphasis on infectious and zoonotic agents in Northern Germany. PLoS ONE 12 , e0175469 (2017).

  9. Balkwill, F. & Mantovani, A. Inflammation and cancer: back to Virchow? Lancet 357 , 539–545 (2001).

  10. Gobert, A. P. & Wilson, K. T. Human and Helicobacter pylori interactions determine the outcome of gastric diseases. Curr. Top. Microbiol. Immunol. 400 , 27– (2017).

  11. Gagnaire, A., Nadel, B., Raoult, D., Neefjes, J. & Gorvel, J. P. Collateral damage: insights into bacterial mechanisms that predispose host cells to cancer. Nat. Rev. Microbiol. 15 , 109–128 (2017).

  12. Sibony, M. & Jones, N. L. Recent advances in Helicobacter pylori pathogenesis. Curr. Opin. Gastroenterol. 28 , 30–35 (2012).

  13. Bridgeford, E. C. et al. Gastric Helicobacter species as a cause of feline gastric lymphoma: a viable hypothesis. Vet. Immunol. Immunopathol. 123 , 106–113 (2008).

  14. Vickers, T. W. et al. Pathology and epidemiology of ceruminous gland tumors among endangered Santa

Catalina Island foxes ( Urocyon littoralis catalinae ) in the Channel Islands, USA. PLoS ONE 10 , 1– (2015).

  1. Moriarty, M. E. et al. Ear mite removal in the Santa Catalina Island fox ( Urocyon littoralis catalinae ): controlling risk factors for cancer development. PLoS ONE 10 , 1–15 (2015).
  2. Himmel, L. & Cianciolo, R. Nodular typhlocolitis, heterakiasis, and mesenchymal neoplasia in a ring-necked pheasant ( Phasianus colchicus ) with immunohistochemical characterization of visceral metastases. J. Vet. Diagn. Invest. 29 , 561– (2017).
  3. Hahn, W. C. et al. Creation of human tumour cells with defined genetic elements. Nature 400 , 464– (1999).
  4. Acevedo-Whitehouse, K., Gulland, F. M., Greig, D. J. & Amos, W. Inbreeding: disease susceptibility in California sea lions. Nature 422 , 35 (2003).
  5. Buck, C. B. et al. The ancient evolutionary history of polyomaviruses. PLoS Pathol. 12 , e1005574 (2016).
  6. DeCaprio, J. A. & Garcea, R. L. A cornucopia of human polyomaviruses. Nat. Rev. Microbiol. 11 , 264– (2013).
  7. Knowles, W. A. in Polyomaviruses and Human Diseases. (ed. Ahsan, N.) 1–25 (Springer-Verlag, New York, 2005).

www.nature.com/nrc

658 |^ OCTOBER 2018 |^ vOlumE 18

  1. Feng, H., Shuda, M., Chang, Y. & Moore, P. S. Clonal integration of a polyomavirus in human Merkel cell carcinoma. Science 319 , 1096–1100 (2008). This paper uses a focused subtractive viral discovery scheme to find an integrated causative virus in the case of Merkel cell carcinoma of humans. The basis for the search was an association of the cancer with immunosuppression, which underscores the importance of host immunity in the case of viral-associated oncogenesis.

  2. Starrett, G. J. et al. Merkel cell polyomavirus exhibits dominant control of the tumor genome and transcriptome in virus-associated Merkel cell carcinoma. mBio 8 , e02079 (2017). This paper shows how oncogenic proteins of viruses, in this case the T Ag of PyVs, are powerful substitutes for the numerous genetic mutations that usually precede transformation in non-virus- associated cancers.

  3. Dela Cruz, F. N. J. et al. Novel polyomavirus associated with brain tumors in free-ranging raccoons, western United States. Emerg. Infect. Dis. 19 , 77– (2013).

  4. Brostoff, T., Dela Cruz, F. N. J., Church, M. E., Woolard, K. D. & Pesavento, P. A. The raccoon polyomavirus genome and tumor antigen transcription are stable and abundant in neuroglial tumors. J. Virol. 88 , 12816–12824 (2014). The data presented in this paper fulfil some of the criteria for viral causality in a wildlife cancer. The stability of the virus as an episome is considered a different oncogenic mechanism than that of the other known PyV associated with cancer (MCPyV).

  5. Stewart, S. E., Eddy, B. E., Gochenour, A. M., Borgese, N. G. & Grubbs, G. E. The induction of neoplasms with a substance released from mouse tumors by tissue culture. Virology 3 , 380–400 (1957).

  6. Gross, L. A filterable agent, recovered from Ak leukemic extracts, causing salivary gland carcinomas in C3H mice. Proc. Soc. Exp. Biol. Med. 83 , 414– (1953).

  7. Cheng, J., DeCaprio, J. A., Fluck, M. M. & Schaffhausen, B. S. Cellular transformation by simian virus 40 and murine polyoma virus T antigens. Semin. Cancer Biol. 19 , 218–228 (2009).

  8. Church, M. E. et al. Exposure to raccoon polyomavirus (RacPyV) in free-ranging North American raccoons ( Procyon lotor ). Virology 489 , 292–299 (2016).

  9. Church, M. E. et al. BRD4 is associated with raccoon polyomavirus genome and mediates viral gene transcription and maintenance of a stem cell state in neuroglial tumour cells. J. Gen. Virol. 97 , 2939– (2016).

  10. Colegrove, K. M. et al. Polyomavirus infection in a free-ranging California sea lion ( Zalophus californianus ) with intestinal T cell lymphoma. J. Vet. Diagn. Invest. 22 , 628–632 (2010).

  11. Wellehan, J. F. J. et al. Characterization of California sea lion polyomavirus 1: expansion of the known host range of the Polyomaviridae to Carnivora. Infect. Genet. Evol. 11 , 987–996 (2011).

  12. Leendertz, F. H. et al. African great apes are naturally infected with polyomaviruses closely related to Merkel cell polyomavirus. J. Virol. 85 , 916–924 (2011).

  13. Hill, S. C. et al. Discovery of a polyomavirus in European badgers ( Meles meles ) and the evolution of host range in the family Polyomaviridae. J. Gen. Virol. 96 , 1411–1422 (2015).

  14. Varsani, A. et al. Identification of a polyomavirus in Weddell seal ( Leptonychotes weddellii ) from the Ross Sea (Antarctica). Arch. Virol. 162 , 1403– (2017).

  15. Siqueira, J. D. et al. Endemic infection of stranded southern sea otters ( Enhydra lutris nereis ) with novel parvovirus, polyomavirus, and adenovirus. J. Wildl. Dis. 53 , 532–542 (2017).

  16. Nainys, J., Timinskas, A., Schneider, J., Ulrich, R. G. & Gedvilaite, A. Identification of two novel members of the tentative genus Wukipolyomavirus in wild rodents. PLoS ONE 10 , e0140916 (2015).

  17. Kobayashi, S. et al. Detection of novel polyomaviruses in fruit bats in Indonesia. Arch. Virol. 160 , 1075–1082 (2015).

  18. Cantalupo, P. G., Buck, C. B. & Pipas, J. M. Complete genome sequence of a polyomavirus recovered from a pomona leaf-nosed bat ( Hipposideros pomona ) metagenome data set. Genome Announc. 5 , e01053-16 (2017).

  19. Varsani, A. et al. A novel papillomavirus in Adélie penguin ( Pygoscelis adeliae ) faeces sampled at the Cape Crozier colony. Antarctica. J. Gen. Virol. 95 , 1352–1365 (2014).

  20. Peretti, A., FitzGerald, P. C., Bliskovsky, V., Pastrana, D. V. & Buck, C. B. Genome sequence of a fish-associated polyomavirus, Black Sea Bass ( Centropristis striata ) polyomavirus 1. Genome Announc. 3 , e01476-14 (2015).

  21. Woolford, L. et al. A novel virus detected in papillomas and carcinomas of the endangered western barred bandicoot ( Perameles bougainville ) exhibits genomic features of both the Papillomaviridae and Polyomaviridae. J. Virol. 81 , 13280–13289 (2007).

  22. Woolford, L. et al. Cutaneous papillomatosis and carcinomatosis in the Western barred bandicoot ( Perameles bougainville ). Vet. Pathol. 45 , 95– (2008).

  23. van Dyk, E. et al. Detection of bovine papillomavirus DNA in sarcoid-affected and healthy free-roaming zebra ( Equus zebra ) populations in South Africa. J. Virol. Methods 158 , 141–151 (2009).

  24. Lindsey, C. L. et al. Bovine papillomavirus DNA in milk, blood, urine, semen, and spermatozoa of bovine papillomavirus-infected animals. Gen. Mol. Res. 8 , 310–318 (2009).

  25. Rector, A. & Van Ranst, M. Animal papillomaviruses. Virology 445 , 213–223 (2013).

  26. Gaynor, A. M., Fish, S., Duerr, R. S., Cruz, F. N. J. & Pesavento, P. A. Identification of a novel papillomavirus in a Northern Fulmar ( Fulmarus glacialis ) with viral production in cartilage. Vet. Pathol. 52 , 553–561 (2015).

  27. Moore, P. S. & Chang, Y. The conundrum of causality in tumor virology: the cases of KSHV and MCV. Semin. Cancer Biol. 26 , 4–12 (2014).

  28. Gulland, F. M., Trupkiewicz, J. G., Spraker, T. R. & Lowenstine, L. J. Metastatic carcinoma of probable transitional cell origin in 66 free-living California sea lions ( Zalophus californianus ), 1979 to 1994. J. Wildl. Dis. 32 , 250–258 (1996).

  29. Lipscomb, T. P. et al. Common metastatic carcinoma of California sea lions ( Zalophus californianus ): evidence of genital origin and association with novel gammaherpesvirus. Vet. Pathol. 37 , 609–617 (2000).

  30. King, D. P. et al. Otarine herpesvirus-1: a novel gammaherpesvirus associated with urogenital carcinoma in California sea lions ( Zalophus californianus ). Vet. Microbiol. 86 , 131–137 (2002).

  31. Buckles, E. L. et al. Otarine Herpesvirus-1, not papillomavirus, is associated with endemic tumours in California sea lions (Zalophus californianus). J. Comp. Pathol. 135 , 183–189 (2006).

  32. Dagleish, M. P. et al. The first report of otarine herpesvirus-1-associated urogenital carcinoma in a South American fur seal ( Arctocephalus australis ). J. Comp. Pathol. 149 , 119–125 (2013).

  33. Buckles, E. L. et al. Age-prevalence of Otarine Herpesvirus-1, a tumor-associated virus, and possibility of its sexual transmission in California sea lions. Vet. Microbiol. 120 , 1–8 (2007).

  34. Browning, H. M., Gulland, F. M., Hammond, J. A., Colegrove, K. M. & Hall, A. J. Common cancer in a wild animal: the California sea lion ( Zalophus californianus ) as an emerging model for carcinogenesis. Philos. Trans. R Soc. Lond. B Biol. Sci. 370 , 1673 (2015). This study highlights how urogenital carcinomas in the California sea lion provide a remarkable model of the complex interaction of environment, pathogens and genetics that are likely involved in most wildlife cancers and that are among the most studied.

  35. Browning, H. M. et al. Evidence for a genetic basis of urogenital carcinoma in the wild California sea lion. Proc. Biol. Sci. 281 , 20140240 (2014).

  36. Curry, S. S. et al. Persistent infectivity of a disease-associated herpesvirus in green turtles after exposure to seawater. J. Wildl. Dis. 36 , 792– (2000).

  37. Ene, A. et al. Distribution of chelonid fibropapillomatosis-associated herpesvirus variants in Florida: molecular genetic evidence for infection of turtles following recruitment to neritic developmental habitats. J. Wildl. Dis. 41 , 489–497 (2005).

  38. Herbst, L. H. et al. Experimental transmission of Green turtle fibropapillomatosis using cell-free tumor extracts. Dis. Aquat. Organ. 22 , 1–12 (1995).

  39. Foley, A. M., Schroeder, B. A., Redlow, A. E., Fick-Child, K. J. & Teas, W. G. Fibropapillomatosis in stranded green turtles ( Chelonia mydas ) from the eastern United States (1980–1998): trends and associations with environmental factors. J. Wildl. Dis. 41 , 29–41 (2005).

  40. Van Houtan, K. S., Hargrove, S. K. & Balazs, G. H. Land use, macroalgae, and a tumor-forming disease in marine turtles. PLoS ONE 5 , 1–9 (2010).

  41. dos Santos, R. G. et al. Relationship between fibropapillomatosis and environmental quality: a case study with Chelonia mydas off Brazil. Dis. Aquat. Organ. 89 , 87–95 (2010).

  42. Greenblatt, R. J. et al. The Ozobranchus leech is a candidate mechanical vector for the fibropapilloma- associated turtle herpesvirus found latently infecting skin tumors on Hawaiian green turtles ( Chelonia mydas ). J. Virol. 321 , 101–110 (2004).

  43. Smiley Evans, T. et al. Mountain gorilla lymphocryptovirus has Epstein-Barr virus-like epidemiology and pathology in infants. Sci. Rep. 7 , 5352 (2017).

  44. Gilardi, K., Whittier, C. & Cranfield, M. in 60th Annual International Conference of the Wildlife Disease Association (Quebec City, Quebec, Canada, 2011).

  45. Hayward, A., Cornwallis, C. K. & Jern, P. Pan-vertebrate comparative genomics unmasks retrovirus macroevolution. Proc. Natl Acad. Sci. USA 112 , 464–469 (2015).

  46. Kassiotis, G. Endogenous retroviruses and the development of cancer. J. Immunol. 92 , 1343– (2014).

  47. Youssef, G., Wallace, W. A., Dagleish, M. P., Cousens, C. & Griffiths, D. J. Ovine pulmonary adenocarcinoma: a large animal model for human lung cancer. ILAR J. 56 , 99–115 (2015).

  48. Fox, K. A. et al. Experimental transmission of bighorn sheep sinus tumors to bighorn sheep ( Ovis canadensis canadensis ) and domestic sheep. Vet. Pathol. 53 , 1164–1171 (2016).

  49. Fox, K. A. et al. Paranasal sinus masses of Rocky Mountain bighorn sheep ( Ovis canadensis canadensis ). Vet. Pathol. 48 , 706–712 (2011).

  50. Kinney, M. E. & Pye, G. W. Koala retroviruses: a review. J. Zoo Wildl. Med. 47 , 387–396 (2016).

  51. Xu, W. et al. An exogenous retrovirus isolated from koalas with malignant neoplasias in a US zoo. Proc. Natl Acad. Sci. USA 110 , 11547– (2013).

  52. Shojima, T. et al. Identification of a novel subgroup of koala retrovirus from koalas in Japanese zoos. J. Virol. 87 , 9943–9948 (2013).

  53. Xu, W. & Eiden, M. Koala retroviruses: evolution and disease dynamics. Ann. Rev. Virol. 2 , 119– (2015). This paper reveals how koala retroviruses are likely in a unique transient state, as they are presently evolving into endogenous retroviruses, providing an opportunity to study this phenomenon.

  54. Coffee, L. L., Casey, J. W. & Bowser, P. R. Pathology of tumors in fish associated with retroviruses: a review. Vet. Pathol. 50 , 390–403 (2013).

  55. Rovnak, J. & Quackenbush, S. L. Walleye dermal sarcoma virus: molecular biology and oncogenesis. Viruses 2 , 1984–1999 (2010).

  56. LaPierre, L. A., Casey, J. W. & Holzschu, D. L. Walleye retroviruses associated with skin tumors and hyperplasias encode cyclin D homologs. J. Virol. 72 , 8765–8767 (1998).

  57. Martineau, D., Bowser, P. R., Wooster, G. A. & Armstrong, L. D. Experimental transmission of a dermal sarcoma in fingerling walleyes ( Stizostedion vitreum vitreum ). Vet. Pathol. 27 , 230–234 (1990).

  58. Bowser, P. R. et al. Swimbladder leiomyosarcoma in Atlantic salmon ( Salmo salar ) in North America. J. Wildl. Dis. 48 , 795–798 (2012).

  59. Abram, Q. H., Dixon, B. & Katzenback, B. A. Impacts of low temperature on the teleost immune system. Biology 22 , 4 (2017).

  60. Petrie, B., Barden, R. & Kasprzyk-Hordern, B. A review on emerging contaminants in wastewaters and the environment: current knowledge, understudied areas and recommendations for future monitoring. Water Res. 72 , 3–27 (2015).

  61. Wilkinson, J., Hooda, P. S., Barker, J., Barton, S. & Swinden, J. Occurrence, fate and transformation of emerging contaminants in water: an overarching review of the field. Environ. Pollut. 231 , 954– (2017).

  62. Matson, C. W. et al. Evolutionary toxicology: population-level effects of chronic contaminant exposure on the marsh frogs ( Rana ridibunda ) of Azerbaijan. Environ. Health Perspect. 114 , 547– (2006).

  63. Hinton, D. E. et al. Resolving mechanisms of toxicity while pursuing ecotoxicological relevance? Mar. Pollut. Bull. 51 , 635–648 (2005).

  64. Martineau, D. et al. Cancer in wildlife, a case study: beluga from the St. Lawrence Estuary, Québec, Canada. Environ. Health Perspect. 110 , 285– (2002).

NATuRE REvIEwS |^ CANCeR

vOlumE 18 |^ OCTOBER 2018 |^659

  1. Metzger, M. J. et al. Widespread transmission of independent cancer lineages within multiple bivalve species. Nature 534 , 705–709 (2016).
  2. Muehlenbachs, A. et al. Malignant transformation of Hymenolepis nana in a human host. N. Engl. J. Med. 373 , 1845–1852 (2015).
  3. Conn, D. B. Malignant transformation of Hymenolepis nana in a human host. N. Engl. J. Med. 374 , 1293 (2016).
  4. Abegglen, L. M. et al. Potential mechanisms for cancer resistance in elephants and comparative cellular response to DNA damage in humans. J. Am. Vet. Med. Assoc. 314 , 1850–1860 (2015).
  5. Pessier, A. P., Stern, J. K. & Witte, C. L. TP53 gene and cancer resistance in elephants. JAMA 315 , 1789 (2016).
  6. Roher, D. P. & Nielsen, S. W. Carcinoma in the urinary bladder of a white-tailed deer ( Odocoileus virginianus ). J. Wildl. Dis. 18 , 361–363 (1982).
  7. Wobeser, G. & Wobeser, A. G. Carcass disappearance and estimation of mortality in a simulated die-off of small birds. J. Wildl. Dis. 28 , 548–554 (1992). This article describes the substantial rate of carcass attrition and highlights the difficulty in detecting moribund or dead animals — one of the greatest challenges in wildlife disease surveillance — which is likely exacerbated in instances with a relatively low prevalence of disease at a given time.
  8. Beringer, J., Hansen, L. P. & Stallknech, D. E. An epizootic of hemorrhagic disease in white-tailed deer in Missouri. J. Wildl. Dis. 36 , 588–591 (2000).
  9. Moresco, A. & Agnew, D. W. Reproductive health surveillance in zoo and wildlife medicine. J. Zoo Wildl. Med. 44 , S26–33 (2013).
  10. Karlitz, J. J. et al. Colorectal cancer incidence rates in the Louisiana Acadian Parishes demonstrated to be among the highest in the United States. Clin. Transl Gastroenterol. 5 , e60 (2014).
  11. Mantere, T. et al. Finnish Fanconi anemia mutations and hereditary predisposition to breast

and prostate cancer. Clin. Genet. 88 , 68– (2015).

  1. Im, K. M. et al. Haplotype structure in Ashkenazi Jewish BRCA1 and BRCA2 mutation carriers. Hum. Genet. 130 , 685–699 (2011).

  2. Belanger, M. H. et al. A targeted analysis identifies a high frequency of BRCA1 and BRCA2 mutation carriers in women with ovarian cancer from a founder population. J. Ovarian Res. 8 , 1 (2015).

  3. Favé, M. J. et al. Gene-by-environment interactions in urban populations modulate risk phenotypes. Nat. Commun. 9 , 827 (2018).

  4. Caulin, A. F. & Maley, C. C. Peto’s Paradox: evolution’s prescription for cancer prevention. Trends Ecol. Evol. 26 , 175–182 (2011).

  5. Seluanov, A., Gladyshev, V. N., Vijg, J. & Gorbunova, V. Mechanisms of cancer resistance in long-lived mammals. Nat. Rev. Cancer (2018).

  6. Koch, R. Die Aetiologie der Tuberkulose. Mitt. k. Gesundh-Amte 2 , 1–88 (1884).

  7. Hill, A. B. The environment and disease: association or causation? Proc. R. Soc. Med. 58 , 295–300 (1965).

  8. Kean, J. M., Rao, S., Wang, M., Garcea, R. L. & Atwood, W. J. Seroepidemiology of human polyomaviruses. PLoS Pathol. 5 , e1000363 (2009).

  9. Namiesnik, J. et al. Concentration of bioactive compounds in mussels Mytilus galloprovincialis as an indicator of pollution. Chemosphere 73 , 938– (2008).

  10. Kakehashi, A., Wei, M., Fukushima, S. & Wanibuchi, H. Oxidative stress in the carcinogenicity of chemical carcinogens. Cancers 5 , 1332–1354 (2013).

  11. Spink, D. C. et al. Induction of CYP1A1 and CYP1B by benzo(k)fluoranthene and benzo(a)pyrene in T-47D human breast cancer cells: roles of PAH interactions and PAH metabolites. Toxicol. Appl. Pharmacol. 226 , 213–224 (2008).

  12. Tarantini, A. et al. Relative contribution of DNA strand breaks and DNA adducts to the genotoxicity of benzo[a]pyrene as a pure compound and in complex mixtures. Mutat. Res. 671 , 67–75 (2009).

  13. Chaudhuri, L. et al. Polychlorinated biphenyl induced ROS signaling delays the entry of quiescent human breast epithelial cells into the proliferative cycle. Free Radic. Biol. Med. 49 , 40–49 (2010).

  14. Ng, H. W., Perkins, R., Tong, W. & Hong, H. Versatility or promiscuity: the estrogen receptors, control of ligand selectivity and an update on subtype selective ligands. Int. J. Env. Res. Publ. Health 11 , 8709– (2014).

  15. Karoutsou, E., Karoutsos, P. & Karoutsos, D. Endocrine disruptors and carcinogenesis. Arch. Canc. Res. 5 , 1–8 (2017).

Acknowledgements

The authors are grateful to B. Stacy (University of Florida), K. Colegrove (University of Illinois) and S. L. Quackenbush (University of Colorado) for responding to their requests for additional information and to J. Crum (West Virginia Division of Natural Resources) for his contribution of cancer cases in white-tailed deer. The authors are also deeply grateful to their colleagues at the University of California Davis and Michigan State University, East Lansing, for comments on the manuscript and their support.

Author contributions

In addition to contributions in research, P.A.P., D.A. and K.D.W. all contributed to the writing, reviewing and editing of the manuscript. M.K.K. was instrumental in drafting the man- uscript and in providing observations from morbidity and mortality investigations.

Competing interests

The authors declare no competing interests.

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Reviewer information

Nature Reviews Cancer thanks A. Boddy, J. Landolfi and D. McAloose for their contribution to the peer review of this work.

NATuRE REvIEwS |^ CANCeR

vOlumE 18 |^ OCTOBER 2018 |^661