Beyond Mice and Men: Alternative Vertebrate and Invertebrate Models in Cancer Biology

Abstract

While cancer looms large in the public and scientific minds, this can result in a quite narrow view of its range and manifestations. The centering of human clinical investigations and the popularity of mouse models leave many unaware of fascinating findings about cancer in other species. Comparative oncology shows that cancer can be found not just in mammals but widely across the tree of animals, a realization that broadens views of the disease in fundamentally important ways. Examples include protective mechanisms of cancer-resistant species, transmissible cancers, and questions about whether cancer risk is intrinsic to metazoan life. I highlight strong evidence for cancer in invertebrates, focusing on Drosophila, which not only show many hallmarks of cancer but also induce host responses mimicking those of human patients, including antitumor immunity. A wide-ranging study of cancer-bearing species enhances the potential for transformative advances in battling this ancient disease.

INTRODUCTION

Cancer can dominate views of health and mortality, especially in the Global North. Four in 10 Americans will be diagnosed with cancer, and nearly everyone has a loved one affected by it. Cancer is a public enemy whose threat is appreciated by all, regardless of their understanding of the underlying disease mechanisms. However, this permeation of cancer awareness means that much about it is tacitly accepted.

Everyone knows that humans, if they do not die from other causes, will get cancer. Many are aware that their pet dogs and cats can also develop tumors, and some have read that scientists use mice and rats to study cancer in the laboratory. But few have wondered whether other animals beyond mammals can develop this disease—and what we might learn from these creatures.

The field that studies how different species do or do not experience cancer is called comparative oncology (succinctly reviewed in Vincze et al. 2025). Comparative oncology has a long history but has been overshadowed in recent decades by laboratory cancer models including cultured cells and mice. The foundational impact of these focused models is acknowledged, but there is now interest in what may be learned by exploring cancer more widely across the animal kingdom. To give one striking example, although human cancer is not a communicable disease, Tasmanian devils (Sarcophilus harrisii) can pass a carcinoma from one individual to another through biting. This transmissible cancer is decimating the species in the wild; transmissible cancers also exist in feral dogs and devastate populations of marine mussels and clams (Hart et al. 2023, Murchison 2008, Ostrander et al. 2016) (Figure 1b). Such phenomena illustrate how comparative oncology can broaden views of metastasis, tumor evolution, and immune defense. They dramatize the impactful perspectives that come from considering cancer beyond the familiar.

Figure 1.

Cancer in vertebrates. (a) Simplified phylogeny of vertebrates, with average neoplasia and malignancy rates of familiar groups, from Vincze et al. (2022). Birds and reptiles are measured together in the clade of sauropsids. “Present” indicates that both neoplasia and malignancy are found, but rates have not been evaluated. (b) Rates of malignancy in select tetrapod species from Compton et al. (2025). Tree information is from https://timetree.org.

To hazard an analogy, comparative oncology can bring benefits akin to those of learning a second language. If one only speaks English, one might never know that some languages (e.g., Romance languages) assign gender to nouns. One might further never know that other languages (e.g., Mandarin Chinese) convey information through spoken tones as well as words. So a comparative approach can bring to light elements that are, but should not be, taken for granted; it can also sharpen appreciation of truly core elements. Moreover, though languages do not develop via a biological evolutionary model, they nevertheless show descent with divergence over time, and they can preserve elements that reveal shared heritage. Comparative oncology can similarly yield insights into how cancer has shaped animal evolutionary history.

In the millennia-long human quest to understand cancer and mitigate its effects, new techniques are creating remarkable opportunities to leverage information about the breadth of cancer across the animal kingdom, and can expand experimental work to species not conventionally considered within the realm of oncology. To extend the analogy above, a scientist studying only cancer in mammals might be akin to a Spanish speaker studying only the closely related Romance languages. The Spanish speaker would miss insights about fundamental language structures from Sino-Tibetan and African languages, akin to what the scientist could learn from a study of avian and ectothermic vertebrates as well as invertebrate species. As the latter are more widely diverged from the familiar but nevertheless share essential elements, they provide a rich opportunity for new and transformational discoveries. This review therefore specifically argues for the value of a broad scope of comparison within comparative oncology.

HOW SHOULD WE DEFINE CANCER?

When considering cancer across the tree of life, it is necessary to define terms and to ask whether the terminology may intrinsically limit the species to which studies may be applied. Clear definitions of cancer and tumor are more elusive than generally appreciated. For instance, the standard textbook The Biology of Cancer (Weinberg 2023) offers largely circular definitions of cancer and neoplasia that invoke tumors, and the word tumor is not specifically defined. The murkiness can be understood by appreciating the history of cancer terminology, originating with Hippocrates’ karkinos and oncos 2,400 years ago and subsequently shaped by clinicians in ways that do not always map on to current understanding of molecular mechanisms. One example is the term carcinoma in situ, commonly used to describe a dysplastic set of noninvasive cells in breast tissue that are not in fact carcinomas (which are defined as malignant; see below). A separate point of contention is whether a tumor must be metastatic to be truly cancerous, wherein metastasis involves not just crossing a basement membrane at its site of origin (invasion) but undergoing intravasion into blood vessels followed by extravasation and growth at a secondary site. Such a definition would intrinsically rule out any possible recognition of cancer in invertebrates, which do not have vascular circulatory systems; a similar objection applies to the angiogenesis criterion offered by the famed “Hallmarks of Cancer” review (Hanahan & Weinberg 2000). Overall, the origins of cancer-related terms in human clinical medicine can present a challenge to broad comparative oncology, particularly when considering species with anatomies and life histories that differ from our own.

For this review, we use the following definitions, which approach consensus without relying on criteria that would restrict applicability to vertebrates. A tumor is the accumulation of excess cells—due to increased proliferation, growth, and/or reduced cell death—in a time or place that is inappropriate to the animal, with morphological changes that indicate changes beyond hyperplastic growth alone; this can also be termed neoplasia. When such cells leave their normal site and cross a basement membrane to invade other tissues, the tumor can be termed malignant; when they do not, the tumor is called benign. Though often used as a synonym for the malignant tumor itself, the term cancer may be also usefully considered as the whole-body disease of disrupted physiology that is caused by malignant tumors.

A BRIEF HISTORY OF COMPARATIVE ONCOLOGY

Comparative oncology as a discipline dates back centuries, and it is inextricably intertwined with foundational advances in cancer biology. A list of 13 unanswered questions about cancer posed in 1802 by a British scientific society includes “Query 10: Are brute creatures subject to any disease resembling cancer in the human body?” (Dawe 1969, The Medical Committee of the Society for Investigating the Nature and Cure of Cancer 1967). An affirmative answer could be found in descriptions of dog tumors by Jean-Baptiste Huzard in 1792 (Baba & Catoi 2007). Detailed observations were subsequently made by Johannes Müller, whose 1838 work pioneering the microscopic analysis of tumors—to which we owe the distinction between benign and malignant—included samples from domestic animals (Hajdu 2004). Muller’s work influenced Rudolf Virchow, the famous originator of cell theory, whose foundational work on cellular pathology examined tumors in cattle as well as dogs. Virchow’s work prompted his striking 1858 declaration, “Between animal and human medicine there are no dividing lines—nor should there be” (Klauder 1958).

Consistent with Virchow’s program, the late nineteenth and early twentieth centuries saw a rapid recognition of tumor-like growths in nonmammalian species. From oysters in 1887 to trout in 1903 (Dawe 1969), what these reports lacked in pathological detail was countered by the confidence that there was no reason to limit cancer to domestic or even wild mammals. These observations aroused skepticism, and it is instructive that skepticism also extended to reports of tumors in mice (Dunn 1962). By not succumbing to skeptics, scientists paved the way for the model organism contention that a variety of animals could be useful avenues for human cancer research.

The open-minded approach bore immediate fruit whose roots are often forgotten today. The discovery of oncogenes, culminating in the cloning of Src and the 1989 Nobel Prize given to Bishop and Varmus, began when a Long Island farmer brought her prize hen to Peyton Rous at the Rockefeller Institute in 1909 (Rous 1911). Rous’s recognition that this chicken bore a sarcoma, and his isolation of the transforming principle from it (the eponymous Rous sarcoma virus), led to his own 1966 Nobel Prize. Perhaps the earliest experiment in cancer genetics came in the 1920s when investigators found that intercrossing two closely related fish of the genus Xiphophorus led to hybrids that develop spontaneous and lethal melanomas (Gordon 1927, Johnson & Phadnis 2021); these melanomas were later found to be driven by a mutated epidermal growth factor (EGF) receptor (Wittbrodt et al. 1989). Thus, nonmammalian tumors lay at the birth of cancer genetics and molecular mechanisms in experimental oncology.

Recognizing the value of such investigations, in 1966 the National Cancer Institute established the Registry of Tumors in Lower Animals (RTLA) at the Smithsonian Institute. Though repositories of mammalian tumors had existed since the nineteenth century, the RTLA was specifically charged with the collection of specimens to “facilitate the comparative study of tumorigenesis and related disorders in invertebrate and poikilothermic vertebrate animals” (Harshbarger 1969). Donated tumors were preserved and underwent expert analysis to distinguish likely neoplastic lesions from hyperplasia, granulomas, and other nononcogenic processes (Harshbarger 1984). When the RTLA was closed in 2007, it contained more than 7,000 specimens; the collection is still extant though not actively managed.

APPROACHES WITHIN COMPARATIVE ONCOLOGY

Comparative oncology today has moved past specimen collection to embrace a variety of approaches. The first looks for tumors in wild animals (Giraudeau et al. 2024, Madsen et al. 2017, Pesavento et al. 2018). While of compelling relevance, this approach has an obvious challenge beyond logistics. Selection bias eliminates most cancer-harboring individuals from any surveyed population, both because malignant tumors are lethal in themselves and because any sublethal loss of fitness may lead to culling from the population by predation or disease. In addition to the likelihood that wild populations may not reach age ranges in which cancer may be prevalent, less obvious challenges [e.g., the rapid rate of carcass disappearance (Vittecoq et al. 2013) and bias toward visible benign surface tumors] also exist. Overall, there are quite limited data on animal cancer in the wild; ideally this gap would be closed. Wild animals can also serve as sentinels for carcinogenic pollution, particularly in aquatic environments, where cancer rates in fish as well as marine mammals increase in habitats exposed to industrial or agricultural use (Baines et al. 2021). For land animals, in addition to charismatic megafauna that are actively managed, global wildlife health surveillance programs in place to detect zoonotic disease emergence have the potential to contribute information about cancer rates.

A second approach surveys cancer in animals held in zoos and similar facilities. More than 1,200 zoos share a medical database, managed by the nonprofit organization Species360, that records routine monitoring of animals under care as well as necropsies where cancer can be identified (Compton et al. 2025, Vincze et al. 2022). This database covers 21,000 species, albeit with a bias toward mammals and birds, and provides fascinating information about relative cancer risk. In addition to limitations, the data have many strengths, one of which is that the age of the individual is usually known. A second is that they allow a comparison of neoplasia versus malignancy rates, which reveals that there are not species that get solely benign tumors. It must be remembered that animals held in zoos are immune to predation, leading to average lifetimes that far exceed those of their wild counterparts. They also subsist on a different diet than that of their natural habitat and trade native health stressors (e.g., parasites) for artificial (e.g., altered light and temperature, human interaction). The same, of course, is true in the laboratory: One estimate is that most wild mice die within less than a year, a fraction of their 2–3-year lifespan in captivity and before most naturally occurring tumors have been documented in the species (Anisimov et al. 2005). Interestingly, a recent analysis found that most diagnoses of neoplasia in zoo necropsies occurred prior to the average lifespan of the animal, that is, not preferentially in old individuals (Compton et al. 2025).

A third approach is veterinary oncology, which is powered by the close medical observation of pets (mostly small mammals such as cats and dogs) as well as livestock. Perhaps 4 million dogs are diagnosed with cancer in the United States each year, a risk that elevates sharply with age; malignancy is the major source of mortality in aged dogs in several Western countries (LeBlanc & Mazcko 2020, Ostrander et al. 2019). In addition to documenting naturally occurring tumors in animals that share an environment with humans, research with companion animals involves clinical trials, including testing therapies that go beyond the options limited for human patients by governmental regulations (LeBlanc & Mazcko 2020, Lenz & Atherton 2025). Canine trials on tumors such as osteosarcoma, melanoma, and glioma were included in a comparative oncology component of the multibillion dollar US Cancer Moonshot Initiative.

A fourth approach is to bring wild animal tissue into the lab by isolating cells and immortalizing them through well-established culture methods. Such cells are thereafter an inexhaustible source of material for cancer-related assays. Fresh samples of vertebrate skin can yield fibroblasts, which can be either directly studied or reprogrammed to other tissue fates. This approach has been powerful in revealing species-specific aspects of tumor-relevant biology (e.g., Firsanov et al. 2024), with the caveat that it is limited to cell-intrinsic mechanisms and that the fibroblast is not the cell of origin for most tumors. Moreover, not all animal tissues are prone to convenient cell immortalization.

Finally, the fifth approach is to use diverse animal species in laboratory cancer research. Mice have been the overwhelmingly favorite laboratory model, displacing the rat during the revolution in transgenesis and knockout technologies. Alongside the many advantages there are also relevant limitations (e.g., Anisimov et al. 2005, Bulls et al. 2024, Cheon & Orsulic 2011, Day et al. 2015, Rangarajan et al. 2004, Tao & Reese 2017). Other well-known model organisms, such as the zebrafish and the fruit fly (Drosophila) (Astell & Sieger 2020, Bilder et al. 2021, Cagan et al. 2019, Chatterjee & Deng 2019, Weiss et al. 2022), are now explicitly studied as in vivo cancer models; the cost, generation time, and ethical regulations for these species are lower than those for rodents. Excitingly, as new technologies for genome engineering and sequencing have emerged, this range can expand to nonmodel species with phylogenetically interesting positions.

IS CANCER RISK INTRINSIC TO THE METAZOAN LIFESTYLE?

A central question of comparative oncology is, Can all animals get cancer? To form a fertile adult, any metazoan must create organs by stimulating cell division but then must cease proliferation when the appropriate size is reached, which often also involves programmed cell death. These cells must differentiate into appropriate fates, including exiting from proliferative progenitor status, in a manner coordinated with surrounding neighbors. The cells must further organize physically in an appropriate manner, including being separated from other tissues by a basement membrane. Cancer results from the misregulation of these normal programs of proliferation, differentiation, and morphogenesis that ensure both embryonic development and adult homeostasis. Since core molecular pathways—signaling and otherwise—that control proliferation, differentiation, and morphogenesis are conserved and evolved prior to the separation of vertebrate and invertebrate lineages, a reasonable perspective is that central phenotypes of cancer cells that corrupt these processes—e.g., uncontrolled proliferation, aberrant differentiation, and invasive behavior—are a latent risk to any metazoan species. Thus, for the purpose of this review, one could reframe the central question as follows: Is there a reason to think that some animals cannot get cancer?

For the myriad species where there are no reports of cancer-stricken individuals, one can easily invoke a sampling problem. A 1952 paper (Brown et al. 1952) provides an object lesson: It begins, “It is commonly stated that cancer does not occur in the Eskimos, and to our knowledge no case has so far been reported” before going on to document the first case. No one today would consider that any human ethnic group is immune to cancer. Might the same be true for branches of the animal tree of life? Certainly for many species, no one has bothered to look, a problem fed by the perception of cancer as a disease of mammals and amplified by animal care regulations that require closer health monitoring of mammals than other vertebrates, with very few required for invertebrates. Moreover, observers are trained to recognize mammalian tumors. Such biases are just one obstacle to a comprehensive evaluation of relative cancer susceptibility.

Nevertheless, we can say that there are animals that do not, and probably cannot, get cancer. The nematode Caenorhabditis elegans has been studied intensively in laboratories for over 50 years, including with saturating genetic screens that have yielded fundamental discoveries in cancer-relevant mechanisms such as apoptosis, Ras signaling, and cell invasion (Kimble & Nüsslein-Volhard 2022). Nevertheless, no mutant phenotypes have been seen that fit the definition of cancer used here. The organism’s well-known life history provides some explanation. C. elegans consists of only 959 cells, with the longest somatic lineage having just 10 divisions; it has no somatic stem cells, limiting the possibility of genetic changes within any lineage over the animal’s 2–3-week lifespan. Interestingly, somatic hyperplasia has in fact been achieved through careful hypomorphic reduction of SWI/SNF-mediated differentiation in combination with inactivation of G1/S inhibitors, which in one lineage retains the mutant cell in a proliferative precursor state. These studies reveal a complex network that ensures the robust cell cycle exit associated with the invariant cell lineage of the animal (Ruijtenberg & van den Heuvel 2015).

This instructive example illustrates that small, short-lived species that do not require somatic stem cells are unlikely to suffer from cancer in the wild, although it also reveals that related phenotypes might indeed be experimentally induced. While such lifestyles are much more prevalent among invertebrates, they are certainly not exclusive to the vertebrate clade. For instance, lobsters can live for more than 100 years, growing throughout this period to reach more than three feet long. The longest-lived noncolonial animal is thought to be a 507-year-old clam, of a clade that likely produces immune cells throughout their lifetime. And indeed, tumors have been reported in related bivalve species (see below).

Rather than sampling error, short lifetime, or inherent inability to be transformed, a lack of reported cancer in a species can have a fourth interpretation: that the species has evolved potent anticancer defenses. Some species, such as naked mole rats (Heterocephalus glaber), are nearly though not entirely cancer-free, while the large and long-lived African elephant has cancer rates that are much lower than those of humans. Rather than assuming that these species have some biology that is intrinsically incompatible with cancer formation, their position on the phylogenetic tree near cancer-experiencing species has prompted a fruitful investigation of molecular mechanisms by which they resist tumors. Thus, how are we to view the invertebrates, where arguably clear cases of cancer—as well as cancer-fighting defenses—are reported? If we adopt the view that invertebrate species are generally incapable of oncogenesis, we risk missing interesting lessons about cancer’s breadth as well as potentially informative anticancer strategies that have evolved in this lineage.

CANCER ACROSS THE TREE OF ANIMAL LIFE

Extensive current data on cancer across animal phyla allow two important points to emerge. First, cancer is ubiquitous across the vertebrates, but with varying prevalence that highlights unexpectedly cancer-resistant species (Figure 2). Second, tumors have been described in a number of invertebrate species, and two examples—Drosophila and bivalves—have been particularly well-studied with modern methods (Figure 1). Recent publications provide an overview of vertebrate data (Aktipis et al. 2015, Boddy et al. 2020, Compton et al. 2025, Vincze et al. 2022), which I summarize briefly below.

Figure 2.

Cancer across the tree of animal life. (a) Simplified phylogeny of animals, indicating taxa in which cancer has been identified (with a bracketed 1 indicating a single report in the literature), No? indicating that there is no well-documented report to date in a taxon subjected to some examination, or ND (not determined) where no data on the taxon are available. Within the chordate phylum, vertebrates, tunicates, and lancelets are distinguished. (b) Examples of species with transmissible cancers. Tree information is from https://timetree.org

Mammals

Among the vertebrates, mammals are the most carefully analyzed, primarily using the Species360 database (Vincze et al. 2022). This database includes postmortem evaluations of nearly 12,000 individuals in which cancer is noted if suspected of contributing to demise (i.e., so-called incidentalomas are not recorded). Neoplasia is found widely across the mammals, with a median prevalence of 12% at death. However, there is substantial diversity of rates among the major orders, as well as between individual species. This ranges from ferrets (63% dead with a neoplasm, largely lymphoma) and opossums (56% dead with a neoplasm, largely lung adenocarcinoma) to the common porpoise and Rodrigues fruit bat (Pteropus rodricensis), with neoplasia prevalences of ~1%. Among this diversity, it appears that carnivorous mammals, especially those that consume other mammals, face the highest cancer-related mortality.

The most famous issue in comparative oncology is Peto’s paradox, which points out the contrast with a prima facie expectation that large and long-lived animals (such as elephants, which likely have 100 times more cells than humans) should get more cancer, due to the greater number of chances for an individual cell to acquire a set of oncogenic mutations. Analyses of the mammalian dataset, as well as its expansion to other vertebrates, have led to ongoing debates over whether any relationship between cancer rate and body size or average lifespan exists (Dujon et al. 2024, Lynch 2025, Metzger 2025).

What is clear from the survey is that certain species within each vertebrate clade have remarkably low cancer rates. This observation has sparked the exciting discovery that species have evolved different tumor-suppressive mechanisms, and lessons learned may perhaps inspire new therapeutic approaches (Seluanov et al. 2018). Several examples will whet readers’ appetites, but it is evident that these species’ potent resistance involves mechanisms beyond those listed below.

The rarity of cancer in the enormous and long-lived bowhead whale (Balaena mysticetus) prompted analysis of immortalized fibroblasts from this species, which revealed high efficiency and accuracy in repairing DNA double-strand breaks compared to other mammals (Firsanov et al. 2024). Interestingly, transformation of whale fibroblasts to malignancy actually required fewer genetic changes than for human cells, and apoptotic activity mediated by the central DNA damage response molecule and tumor suppressor p53 was also low. The hypothesis that whales may suppress mutagenic damage by elevating levels of two conserved DNA repair proteins—CIRBP and RPA2, which can improve DNA repair when provided to human cells—points to a potential therapeutic direction.

The small but long-lived naked mole rat is famed for being almost entirely cancer free, with only a few confirmed cases despite widespread breeding. One mechanism implicated in protection is the secretion of a specific, higher-molecular-weight form of hyaluronic acid (HA) into the extracellular matrix (Tian et al. 2013). This HA causes cultured naked mole rat cells to be highly sensitive to contact inhibition of proliferation, a phenotype associated with heightened growth control in vivo. Strikingly, mammalian cancer cells engineered to produce the naked mole rat form of HA showed inhibited proliferation and induction of apoptosis, consistent with a model in which high-molecular-weight HA can be tumor suppressive in transformed cells.

Blind mole rats, which are only distantly related to naked mole rats, have strong cancer resistance that involves an entirely different phenomenon (Zhao et al. 2021). These animals possess naturally low levels of DNA methyltransferase 1 (DNMT1), which leads to activation of retrotransposons during cellular hyperplasia in culture. The resultant cytoplasmic nucleic acids then trigger interferon-induced cell death, which could eliminate an overgrowing population and reduce risk of transformation in vivo.

This small sample provides just a taste of unexpected mechanisms by which exotic mammals may resist cancer. Bats and elephants provide yet other examples (Abegglen et al. 2015, Athar et al. 2025, Karpinski et al. 2025, Koh et al. 2019, Sulak et al. 2016), which include both tumor cell–autonomous and host-driven processes. The diversity of these mechanisms emphasizes how many species have independently evolved cancer defense strategies. One can only begin to imagine how many more distinct mechanisms exist to regulate cancer prevalence in other animals across the tree of life. Comparative oncology studies can thus yield a smorgasbord of therapeutic avenues to pursue toward clinical ends.

Nonmammalian Vertebrates

The literature on nonmammalian vertebrates is less prominent, but it reemphasizes the presence of tumors throughout these groups (Compton et al. 2025). In fact, the fossil record preserves evidence of osteosarcomas in ancestral turtles from the Triassic, dating back 240 million years (Haridy et al. 2019), as well as in later dinosaurs (Ekhtiari et al. 2020). Although all extant nonmammalian vertebrates have a lower average prevalence of neoplasia compared to mammals, many bird tumors are malignant. Blood cancers are the most common, but others (including sarcomas as in Rous’s hen) are found both in pets and in the wild. Reptiles show a slightly greater neoplastic rate than birds, with prominent blood cancers as well as carcinomas of the skin, liver, and bile duct (Christman et al. 2017). Snakes showed a particularly high frequency of neoplasia as well as malignancy, including hematopoietic cancer as well as renal adenocarcinomas. By contrast, turtles are represented in the zoo database by a single nonmalignant neoplasia; neoplasia in crocodilians is also remarkably low (Bulls et al. 2024). However, there is an independent report of a malignant osteosarcoma in a land turtle (Bert et al. 2024), while benign fibropapillomatosis of viral etiology is widely known to impact sea turtles (Jones et al. 2016). Amphibians exhibit the lowest neoplasia and cancer prevalence of the vertebrates; they are relatively understudied, but reports exist of renal adenocarcinoma in the leopard frog and melanomas in axolotls (Ambystoma mexicanum) (Harshbarger et al. 1999, Ruben et al. 2007); invasive lesions were found in tiger salamanders (Ambystoma tigrinum) from a sewage-contaminated lagoon (Rose & Harshbarger 1977).

Fish are not central to the Species360 database, but reports of tumors in bony fish (teleosts) in the wild are widespread (e.g., Romano, Luis Alberto & Pedros, Vergneau-Grosset et al. 2017, Virginia Fonseca 2020). An aquarium database with 883 fish necropsies showed a spontaneous neoplasm rate of 6%. Carcinomas exist, but melanomas beyond those found in Xiphophorus deserve special mention, as the laboratory model organism zebrafish has become a powerful model for studying melanoma as well as other tumor types (van Rooijen et al. 2017, Weiss et al. 2022). While there was an early misconception that sharks and rays (chondrichthyans) rarely developed cancer, this belief has been shown to be inaccurate. Tumors are reported in six of the nine chondrichthyan orders, with about one-third categorized as malignant (Huveneers et al. 2016).

Finally, the order of agnathans (nonjawed vertebrates) is of particular interest. There are only two extant members of this order: the lamprey and the hagfish. Interestingly, the RTLA documents tumors in both, including hepatocellular carcinoma, and describes a lamprey melanoma with multiple subcutaneous, renal, and branchial metastases (Harshbarger 1970). Thus, clear tumors are found throughout vertebrates, even in the lineage related most distantly to mammals and most closely to invertebrates.

Invertebrates

As we consider cancers in invertebrates, let us remind ourselves of the obstacles. First, invertebrates are not monitored by zoo health programs and are rarely treated by veterinarians. Second, only a few edible marine species are harvested in a manner that might allow chance notice of a neoplastic lesion. Third, invertebrate anatomy not only presents some challenges to conventional cancer definitions but also is unfamiliar to nonspecialists, while invertebrate specialists are unfamiliar with cancer pathology. This latter point has caused old case reports with limited documentation to be viewed with skepticism.

One goal of the RTLA was to more rigorously examine reports of tumors in invertebrates. Their adjudication categorized hundreds of examples of invertebrate neoplasia, including in insects, crustaceans, and mollusks (Newton & Lewbart 2017). Most invertebrate tumors were benign, including papillomar growths on lobster, shrimp, and a planarian; benign neoplasms have also been reported in slugs and other gastropods. Malignant solid tumors were categorized in several bivalve species, and one report described a gut carcinoma in king crab with distinct invasiveness. Among the flatworms, one RTLA submission in a trematode was characterized as a ganglioneuroblastoma, while MMP-expressing tumors that invade and destroy surrounding tissue occur spontaneously and can be experimentally induced in planarians (Hall et al. 1986, Voura et al. 2017). There is also the remarkable case of an immunocompromised patient whose lymph node carried malignant cells originating not from his own body but from that of a parasitic tapeworm that he carried (Muehlenbachs et al. 2015), although the prevalence of tumors in tapeworms remains unexplored. Anomalous growths in corals have also been proposed to be neoplastic although this is controversial, as are reports in tunicates and echinoderms. No reports of tumors in sponges exist, and there is only one of a potential tumor in the cephalochordate Amphioxus (Aktipis et al. 2015). As mentioned previously, only the intensively studied C. elegans has evidence of absence; for some taxa above there is currently absence of evidence.

More recently, studies of tumor-like growths in a nonbilaterian cnidarian have emerged. Both wild-caught and laboratory-raised strains of Hydra can exhibit abnormal growths, which result from ectopically proliferating cells accumulating in the ectoderm (Boutry et al. 2023, Domazet-Lošo et al. 2014). These cells appear related to normal interstitial stem cells, but they exhibit a distinct gene expression profile. In one strain, tumor growth is associated with a distinctive microbiome, but that is not the case in another. These tumors can be grafted into a healthy host, where they proliferate and invade the normal Hydra tissue. Interestingly, they can also be transmitted to progeny during budding in this asexually reproducing species. Again, the organism’s life history, including its reputation as an “immortal animal” (Boehm et al. 2013), encourages one to consider how to apply conventional definitions of a tumor.

Drosophila tumors and the hallmarks of cancer.

The best-studied tumors of invertebrates are found in Drosophila. These flies are also arguably the best-studied animals on Earth, with 125 years of laboratory research using classical and cutting-edge methods that yielded seven Nobel Prizes (Bilder & Irvine 2017). Spontaneous tumors are found in the guts of ~5–10% of aged wild-type laboratory flies, starting about three-quarters of the way through the ~9-week lifespan (Siudeja et al. 2015). These tumors initiate in intestinal stem cells and are caused by various forms of genetic instability (Riddiford et al. 2021). Tumors with similar genotypes are driven by autocrine EGF signaling and stimulate IL-6-like mediated inflammation from neighboring cells; they disrupt architecture leading to tissue dysplasia (Patel et al. 2015). Spontaneous lethal tumors can also appear in laboratory larvae, through loss of heterozygosity uncovering a mutant tumor-suppressor gene (TSG) allele maintained at a low rate in natural populations; there are no measurements of tumor rates in wild larvae (Golubovsky et al. 2006, Roegiers et al. 2009). Though those observed to date are not described as invasive, they fulfill many cellular and molecular criteria associated with benign tumors.

Fly tumors can be conveniently created in the laboratory with genetically engineered fly models (GEFMs) that overexpress oncogenes or knock out TSGs in a tissue-specific manner. These approaches and information about the tumors are detailed in reviews (Bilder 2004, Cagan et al. 2019, Chatterjee & Deng 2019, Gonzalez 2013); I note some pertinent information here: (a) Misregulation of a single fly gene can be sufficient to transform cells with many hallmarks of cancer (see below), and these phenotypes, particularly malignant phenotypes, can be exacerbated when a cooperating oncogene such as activated Ras is added; (b) transformation can take place either in developing organ primordia of the larva (both epithelial and neural) or in adult stem cell populations that proliferate throughout the animal’s lifetime; and (c) some genes that transform fly cells are also frequent drivers of human tumors, while others are not commonly mutated (though still conserved regulators of human biology).

It can be instructive to evaluate malignant GEFMs against the well-known hallmarks of cancer, as originally defined by Hanahan & Weinberg (2000). The examples below include those drawn from epithelial cells overexpressing activated Ras along with loss of the tumor suppressor Scribble (Dillard et al. 2021, Pagliarini & Xu 2003), although different GEFMs including those that transform neural stem cell populations can act similarly.

The first three cancer hallmarks describe mechanisms that drive inexorable tumor growth (sustained proliferative signaling, evading growth suppressors, resisting cell death). Scrib/Ras tumors activate the JNK and Hippo signaling pathways, which fuel tumor growth in part through an autocrine IL-6-like/STAT feedback loop, as well as upregulating apoptosis inhibitors (Dillard et al. 2021). Phenotypically, malignant fly cells essentially become immortalized: Though they kill the host, they can be transplanted into a new host, where they continue their lethal growth (Rossi & Gonzalez 2015). Similar cells have been passaged over many years with unceasing proliferation. The tumor cells have thus achieved a fourth hallmark: enabling replicative immortality, although Drosophila use a transposon-based mechanism to prevent telomere shortening (Mason et al. 2008) rather than the telomerase-based mechanism used by humans and other mammals. At any rate, like human malignant tumors, these fly tumors are invariably lethal.

With the final two hallmarks, we encounter the issue of anatomical differences. The fifth hallmark is inducing angiogenesis. This cannot strictly apply to invertebrates because they do not use vasculature to deliver blood to cells but instead bathe their internal organs in an open circulatory system (the hemolymph). Nevertheless, insects face the same challenges in oxygen supply that humans do, and they solve this challenge by supplying oxygen to cells through a tubular network called the trachea (Hayashi & Kondo 2018). The tracheal network is even constructed by the homologous VEGF and PDGF signaling pathway that shapes vertebrate vasculature. And indeed, some GEFMs upregulate PDGF/VEGF ligands to attract extra oxygen supply to the growing tumor (Calleja et al. 2016, Grifoni et al. 2015, Tamamouna et al. 2021).

The sixth hallmark is activating invasion and metastasis, which is a central discriminator between benign and malignant tumors. It is clear that many fly GEFMs meet the invasion criterion (Sharpe et al. 2023). Fly epithelia are bounded by a basement membrane that shares high functional, ultrastructural, and molecular conservation with that of humans. Many fly tumors upregulate matrix metalloproteases that degrade the basement membrane, and they activate motility programs that allow movement out of the primary site, even migrating onto neighboring organs. Moreover, conserved regulators of the epithelial-mesenchymal transition can enhance long-distance dispersion of fly tumors (Campbell et al. 2019). Yet here the question of anatomy arises. Metastasis by definition encompasses steps beyond invasion, including intravasation into and extravasation from a vasculature that the insect simply does not possess. Thus, while flies cannot be said to execute a conventional metastatic process, they certainly exhibit invasive migration out of the tissue of origin.

As the six hallmarks expand with regularly updated reviews on, for example, “next generations” (Hanahan & Weinberg 2011) and “new dimensions” (Hanahan 2022), we wonder whether fly tumors will keep pace in the manner shown above: not a 100% match, but many thematic parallels when considering analogous rather than strictly homologous biology. Certainly features such as avoiding immune destruction (see below), dedifferentiation, and tumor-promoting inflammation are already evident; recent papers demonstrate nonmutational epigenetic reprogramming, altered metabolism, and senescent cell presence as well (Datta & Bangi 2024, Floc’hlay et al. 2023, Parreno et al. 2024, Wong & Verheyen 2021).

Drosophila display tumor–host interactions analogous to human cancer patients.

The latest hallmarks review is titled “Embracing Cancer Complexity: Hallmarks of Systemic Disease,” highlighting tumor–host interactions and systemic impacts of cancer including lethality unrelated to the “mass effect” (Swanton et al. 2024). We can neatly place remarkable recent findings about fly cancer in this perspective as well. Astonishingly, the last decade has identified case upon case where fly tumors induce in their own hosts systemic manifestations that mimic those seen in human cancer patients. These phenomena include critical drivers of human mortality such as cancer cachexia and coagulopathy (detailed in, e.g., Bilder et al. 2021, García-López et al. 2021, Hsi et al. 2023, Liu et al. 2022). Recent work has also identified (a) an early tumor-eliminating process mediated by circulating fly TNF (de Vreede et al. 2022); (b) renal failure caused by tumor-induced signals, at least one of which is conserved in mammals (Kwok et al. 2024; Xu et al. 2023, 2024); (c) evidence for unappreciated cachexia mechanisms conserved with mammals discovered through new fly studies (Liu et al. 2025); and finally, (d) that a fly tumor accelerates mortality by permeabilizing the blood–brain barrier through IL-6-like inflammation, with a similar phenotype demonstrated in one mouse cancer model (Kim et al. 2021).

Excitingly, flies also show antitumor immunity. Although differing biology again comes into play, flies display the therapeutically critical response of the vertebrate immune system in fighting cancer. Insects have no adaptive immune system and instead defend themselves from nonself pathogens via a robust innate immune system with cellular and humoral arms (Banerjee et al. 2019). The cellular arm includes macrophages resembling those of mammals; the humoral arm is activated by the conserved Toll and the semiconserved IMD (which activates NF-κB via receptors with no direct mammalian ortholog) signaling pathways. The output of both pathways is NF-κB transcription factors that upregulate antimicrobial peptides (AMPs) secreted into circulation. A seminal study found that fly macrophages recognize malignant epithelial tumors and restrict their progression (Pastor-Pareja et al. 2008). In one tumor model, macrophages activate Toll signaling in humoral effector organs to secrete AMPs, which cooperate with tumor-associated macrophages to specifically kill tumor cells (Parisi et al. 2014, Parvy et al. 2019). Related immune reactions are seen against blood cell tumors and grafted neural tumors (Hirata et al. 2025, Voutyraki et al. 2023). So even without T cells and B cells, and even in a relatively short-lived species with a modest propensity for cancer in the wild, an invertebrate has evolved an immune regulatory axis to recognize tumors and limit their growth.

To recap, in this well-studied organism, an invertebrate that spontaneously forms tumors can be genetically manipulated to transform cells that not only fulfill many of the hallmarks of cancer but also induce host responses that resemble familiar and lethal manifestations of human cancer such as cachexia and coagulopathy. Moreover, the fly has a mechanism to surveil and eliminate small clones of malignant cells, and it also raises an immune response that fights larger tumors in ways at least superficially related to human tumor immunology. These provocative findings suggest a profound evolutionary connection between cancer-like phenomena and the normal physiology of the fly. They also raise the possibility of deep homology in the way that animals across the phylogenetic tree respond to malignant growth.

Neoplasias in bivalves.

In the wild, the most prevalent invertebrate neoplasias are found in bivalve mollusks. These cancers affect a wide variety of species including clams, oysters, mussels, and cockles. One such neoplasia, termed transmissible disseminated neoplasia (TDN), has led to massive die-offs in commercially important regions. Before dealing with TDN, I discuss conventional neoplasias of this taxon. Large numbers of bivalves are harvested for human consumption, and many are inspected before being served. Moreover, some species are quite long-lived—as mentioned previously, a clam more than 500 years old is the oldest individual animal on record—and expected to contain stem cells that can produce at least blood cells throughout the lifetime. All of this may contribute to the substantial list of neoplasias, which fall into three general categories (Carballal et al. 2015, Newton & Lewbart 2017).

The first are epithelial tumors, with descriptions of both benign and malignant neoplasms. These form on various organs, but they include invasive lesions with evident mitotic figures and associated inflammation. The second affect cells of the germinal epithelium, a tissue producing both soma and germline that must proliferate to ensure reproduction. In these gonadal neoplasms, normal tissue is replaced with cells showing altered cellular, nuclear, and nucleolar morphology. Small foci initially proliferate to fill the reproductive follicle but then later invade into other gonadal organs.

The third are disseminated neoplasias (DNs). In this lethal disease, anaplastic cells proliferate within the affected animal, gradually disrupting normal tissues and replacing them. Across 15 species worldwide, the condition presents with common features: large, rounded cells showing a high nucleus-to-cytoplasm ratio, diminished organellar contents, and frequent mitotic figures that are seldom present in healthy cells. DN cells consistently show varied chromosomal imbalances as well as distinct mutation signatures (including independent events duplicating the mitochondrial control region). The DN cell of origin is likely the hemocyte, a circulating blood cell of invertebrates. However, DN cells are distinctly transformed and exhibit reduced cell death, reduced phagocytic ability, and inhibition of p53 activity. For all of these reasons, investigators have likened DN to a leukemia.

A bivalve transmissible neoplasia.

Exciting recent findings demonstrated that some, though not all, cases of DN result from horizontal transmission of clonal cancer cells from one member of a clam population to another (Metzger et al. 2015). Affected animals across a large range shared neoplastic cells with nearly identical genotypes that were distinct from the hosts’ own genotype. To date, eight independent lines of TDN behaving similarly have been characterized worldwide. These TDNs affect a variety of bivalve species, but they are species restricted, except in a striking case where one species carries a neoplasia derived from a second sympatric species that apparently can resist its own neoplastic cells (Metzger et al. 2016). Neoplastic cells can likely survive in the water column and are taken up by the host but then aggressively replace the host tissues (Hart et al. 2025). Three distinct TDNs (one of clams and two of cockles) have now been examined genomically (Bruzos et al. 2023, Hart et al. 2023). The lineages have certain similarities (hemocyte origin, pronounced genetic instability) but differ in other aspects (two separate putative mechanisms of p53 functional interference, only one lineage is associated with mobile element reactivation).

Thus, TDN is an invertebrate transmissible cancer, analogous to the clonal vertebrate transmissible tumors of Tasmanian devils and canines (Figure 1b). It can arise repeatedly and independently in different bivalve species, presumably originating from a specific nontransmissible DN of an individual. Certainly much is to be learned about fundamental biology [e.g., the bivalve functional equivalent of major histocompatibility class (MHC) systems that prevents most TDNs from crossing species], but TDN also raises fascinating larger questions about how an animal treats cancer on the self–nonself continuum of immune surveillance. It also raises questions about cancer evolution: TDN lineages are highly genetically unstable, even one that is dated to originate several hundred years ago. This provides an interesting comparison with the genomes of the transmissible tumors in Tasmanian devils (Stammnitz et al. 2023), perhaps giving insight into selection against vertebrate tumors by adaptive immune mechanisms that are not present in invertebrates. The availability of many TDN lines will allow comparison of the remarkable ability of cancer cells to transcend the boundaries of an individual animal. Finally, given its prevalence and high lethality, it is reasonable to expect that TDN is exerting a high selective pressure in a short amount of time, and this could reveal elements of tumor defenses in the host as well as perhaps evasive responses from the genetically mutable tumor population.

CONCLUSIONS

Comparative oncology shows us that animal diversity has provided a plethora of natural experiments that have much to tell us about the monumental problem of cancer. Expanding surveys of captive animals will sort out species that are extremely susceptible or resistant to cancer through intrinsic biology rather than extrinsic factors, particularly as nonmammalian vertebrates are scrutinized. Close examination might also identify species in which metastases are enriched or underrepresented.

It is clear that research on cancer-resistant species will lead to new discoveries that are fundamentally different from those in popular rodent models. On the one hand, one could consider therapeutic export of exogenous tumor suppressive mechanisms into the human context [e.g., the provision of naked mole rate HA synthase, which extends the health span in mice (Zhang et al. 2023)]. On the other hand, one could learn deeper information about known elements of human tumor biology (e.g., the multiple mechanisms that downregulate MHC in the mammalian transmissible cancers). For exotic vertebrates, analysis of cultured cells for autonomous properties will be enhanced by in vivo studies, given the appreciation of cancer as a systemic disease and systemic anticancer defenses including immunity. A striking question concerns how evolution of diversity in immune systems plays out in terms of tumor immunology. A considered choice of taxa could yield exciting research programs. Finally, there are well-established nonmammalian models like the zebrafish that will continue to yield insights complementary to those that rodent models supply.

What can we learn from invertebrate cancers? The track record of Drosophila research (e.g., recent Nobel Prizes for widely conserved mechanisms such as developmental and oncogenic signaling pathways, innate immune signaling, and circadian rhythms) indicates that the answer is more than we might think. One has only to consider cell competition, a phenomenon studied for 50 years in the fly that is now emerging as a relevant player in human cancer (van Neerven & Vermeulen 2023). It is worth remembering that fly tumors do not purport to be something unattainable—an exact model for human tumors—but instead a rapid and rigorous in vivo experimental system that can point to conserved pathways and mechanisms of cancer biology. Certainly the striking similarities now documented between tumor–host interactions in flies and human patients fall into this category. To reiterate the familiar phrase, all models are wrong; some are useful. In this light, Drosophila has proven its utility time and again.

Future investigations could go far beyond the fruit fly. Long-lived insects include ant and termite queens that thrive for decades while laying over 100 million eggs. Considering the stem cells needed to support such homeostasis and proliferation, might they be protected from transformation by undiscovered anticancer mechanisms? Are there extra cancer risks associated with invertebrate (and other) species undergoing indeterminate growth or species primed for regeneration and degrowth like Planaria (Hariharan et al. 2015)?

Overall, interrogating the presence and potential of cancer in invertebrates and nonjawed vertebrates allows us to ask fascinating questions about both overlapping and distinct mechanisms that react to general interruptions of epithelial homeostasis—such as sterile and septic injury—compared to a rogue “wound that does not heal” (Dvorak 1986). Further, it will provide insight into how the potential and reality of this disease has shaped the evolution of the animal kingdom, as well as how animals have evolved anticancer defenses. It is this broad view of comparative oncology—not only in terms of species, but in terms of fascinating questions—that will truly realize the promise of the field.