Tollis et al.

BMC Biology (2017) 15:60

DOI 10.1186/s12915-017-0401-7

QUESTION AND ANSWER Open Access

Peto’s Paradox: how has evolution solved

the problem of cancer prevention?
Marc Tollis, Amy M. Boddy and Carlo C. Maley*

Abstract normal checks on carcinogenesis [3]. If every cell

division carries a certain chance that a cancer-causing
The risk of developing cancer should theoretically somatic mutation could occur, then the risk of develop-
increase with both the number of cells and the ing cancer should be a function of the number of cell
lifespan of an organism. However, gigantic animals do divisions in an organism’s lifetime [4]. Therefore, large
not get more cancer than humans, suggesting that bodied and long-lived organisms should face a higher
super-human cancer suppression has evolved lifetime risk of cancer simply due to the fact that their
numerous times across the tree of life. This is the bodies contain more cells and will undergo more cell
essence and promise of Peto’s Paradox. We discuss divisions over the course of their lifespan (Fig. 1). How-
what is known about Peto’s Paradox and provide ever, a 2015 study that compared cancer incidence from
hints of what is yet to be discovered. zoo necropsy data for 36 mammals found that a higher
risk of cancer does not correlate with increased body
mass or lifespan [5]. In fact, the evidence suggested that
What is Peto’s Paradox?
larger long-lived mammals actually get less cancer. This
Peto’s Paradox is named after epidemiologist Richard
has profound implications for our understanding of how
Peto, who noted the relationship between time and
nature has solved the cancer problem over the course of
cancer when he was studying how tumors form in mice.
evolution.
Peto observed that the probability of cancer progression
was related to the duration of exposure to the carcino-
gen benzpyrene [1]. He later added body mass to the How does one go about solving the paradox?
equation, when he wondered why humans both contain From one perspective, the solution to Peto’s Paradox is
1000 times more cells and live 30 times longer than quite simple: evolution [6]. When individuals in popula-
mice, yet the two species do not suffer incredibly differ- tions are exposed to the selective pressure of cancer risk,
ent probabilities of developing cancer [2]. Further, the population must evolve cancer suppression as an
cancer was not a major cause of mortality for large and adaptation or else suffer fitness costs and possibly
long-lived wild animals, despite the increased theoretical extinction. But that only tells us that evolution has
risks. How can this be? found a solution to the paradox, not how those animals
are suppressing cancer. Discovering the mechanisms
underlying these solutions to Peto’s Paradox requires the
Why is it a paradox? tools of numerous subfields of biology including genom-
In a multicellular organism, cells must go through a cell ics, comparative methods, and experiments with cells.
cycle that includes growth and division. Every time a For instance, genomic analyses revealed that the African
human cell divides, it must copy its six billion base pairs savannah elephant (Loxodonta africana) genome con-
of DNA, and it inevitably makes some mistakes. These tains 20 copies, or 40 alleles, of the most famous tumor
mistakes are called somatic mutations. Some somatic suppressor gene TP53 [5, 7]. The human genome
mutations may occur in genetic pathways that control contains only one TP53 copy, and two functional TP53
cell proliferation, DNA repair, apoptosis, telomere ero- alleles are required for proper checks on cancer progres-
sion, and growth of new blood vessels, disrupting the sion. When cells become stressed and incur DNA
damage, they can either try to repair the DNA or they
* Correspondence: [email protected]
Virginia G. Piper Center for Personalized Diagnostics, Biodesign Institute, can undergo apopotosis, or self-destruction. The protein
Arizona State University, 727 E. Tyler St., Tempe, AZ 85287-5001, USA produced by the TP53 gene is necessary to turn on this
© Maley et al. 2017 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Tollis et al. BMC Biology (2017) 15:60 Page 2 of 5

Fig. 1. An illustration of Peto’s Paradox. Cancer is a disease of uncontrolled cell growth and division, and the risk of developing cancer increases with the
number of cell divisions during the lifetime of an organism. Thus, the expected cancer rate for large and/or long-lived species is higher than for smaller
short-lived ones. The solid red line indicates a linear relationship between cancer rate and (body mass)*(lifespan) and the dashed red line represents an
approximation of the expected cancer rate assuming a model describing the probability of an individual developing colorectal cancer after a given
number of cell divisions [4]. The solid blue line represents the observation that there is no relationship between cancer risk and (body mass)*(lifespan) [5].
For instance, cancer risk, which is 11–25% in the human population, is not vastly different between mice and humans. In contrast, cancer risk was
estimated to be 5% in elephants [5]. Metastatic cancer was found in a duck-billed dinosaur [26], suggesting cancer was common enough in that lineage
to be preserved in the fossil record, but not in other species of large dinosaurs. While adult body mass is approximately the same for the dinosaur and
the elephant, duck-billed dinosaurs are thought to have had a shorter lifespan [28, 31]. This suggests that the trade-offs between reproduction and
growth and cancer defense mechanisms [22] left these dinosaurs more susceptible to cancer than elephants

apoptotic pathway. Humans with one defective TP53 which has a lifespan of over 200 years [9]. In popula-
allele have Li Fraumeni syndrome and a ~90% lifetime tions, large body size is often associated with higher
risk of many cancers, because they cannot properly shut fitness, conferring greater access to resources or mates
down cells with DNA damage. Meanwhile, experiments and better predator avoidance. It is not surprising, there-
revealed that elephant cells exposed to ionizing radiation fore, that large body size has evolved again and again
behave in a manner consistent with what you would throughout evolutionary time—a trend first realized in
expect with all those TP53 copies—they are much more the fossil record and known as Cope’s Rule [10]. Cope’s
likely to switch on the apoptotic pathway and therefore Rule applies widely across life from diverse marine taxa
destroy cells rather than accumulate carcinogenic [11] to the extinct giant dinosaurs [12]. Large body size
mutations [5, 7]. has evolved independently in 10 of the 11 placental
mammalian orders [13]: think polar bears and hippopot-
How many different solutions to Peto’s Paradox ami, walruses and giraffes, elephants and whales. Since
are there? many lineages faced the trade-off between large body
There are likely many solutions to Peto’s Paradox in na- size and cancer risk during their evolution, there have
ture, because large body size has evolved independently likely been many different pathways in which cancer
so many times across the history of life. We know that suppression has evolved.
whales did not evolve the extra copies of TP53 like
elephants [8, 9]. In fact, there is no evidence that whales What solutions to the paradox do we know?
evolved extra copies of any tumor suppressor gene—e- In the example of the elephant given above, greater
ven the gigantic bowhead whale (Balaena mysticetus), tumor suppressor gene redundancy provides better
Tollis et al. BMC Biology (2017) 15:60 Page 3 of 5

checks on the cell cycle, in effect ‘blowing up’ cells with doubt require substantial effort to translate recent dis-
DNA damage and preventing cancer [5, 7]. Larger ani- coveries into effective therapies for humans.
mals have a lower metabolic rate than smaller ones, and
the mutagenic agents supplied by fast metabolisms—- But wait, why haven’t all animals evolved extra
most notably reactive oxygen species that can damage tumor suppression mechanisms?
DNA—are simply less common in large animals. This Cancer is a potential problem for all multicellular life
may also resolve Peto’s Paradox in some species. Some and there is no expectation that a species should be
exceptionally long-lived rodents such as naked mole rats completely cancer free; in fact, elephants still get can-
and blind mole rats are famous for having very low can- cer—about 5% of deaths in zoos according to one study
cer rates. In the case of naked mole rats (Heterocephalus [5]. Cancer has also been found in whales [21]. There
glaber), this seems to be due to a form of hyaluronic are a few potential reasons that cancer is still a problem
acid and a super-sensitive CDKN2A tumor suppressor for multicellular animals. First, cancer defense mecha-
pathway that suppresses proliferation of naked mole rat nisms, such as DNA repair, cell cycle control, and im-
cells [14, 15]. Blind mole rats (Spalax judaei and Spalax mune function, can be costly. There are likely energetic
golani) have a different mechanism of cancer suppres- trade-offs between cancer suppression and other import-
sion. Over-proliferation of blind mole rat cells triggers ant life history components, such as reproduction and
massive necrotic cell death that destroys both the prolif- growth [22]. Cancer is a disease of ageing populations
erating cells and their neighbors [16]. both because there is weaker selection to avoid problems
There are a variety of hypotheses for other potential after reproduction [23], and because it takes time to ac-
solutions to Peto’s Paradox. One attractive hypothesis is cumulate all the mutations necessary to cause a cancer.
the prediction of ‘hypertumors’ [17]: bigger tumors in For animals that are short-lived (such as mice), it doesn’t
bigger animals take longer to grow and are susceptible make much sense to invest much in cancer defense
to ‘cheater’ cells, which can take advantage of the tu- mechanisms. These animals are more likely to die of
mor’s angiogenic properties and lower the fitness of the other extrinsic causes (such as predators) than of cancer.
whole tumor. The effect of hypertumors could be to Second, benefits early in life that increase an organism’s
lower the overall lethality of cancer in the bodies of large fitness may lead to disease susceptibility later in life, an
animals. Other potential solutions could be: large ani- evolutionary term called antagonistic pleiotropy [24].
mals have increased immunocompetence with better For example, there may be a genetic variant that allows
surveillance and attacking of neoplastic cells, or they an organism to get big fast—increasing its mating poten-
may have shorter telomeres which would limit the num- tial and decreasing the likelihood it will be killed by a
ber of cell divisions and thus the risk of cancer. How- predator—but this same genetic variant may also lead to
ever, these solutions have not yet been observed in cancer susceptibility as the animal ages.
large-bodied species and more research in these areas is
needed. Did dinosaurs get cancer?
Birds evolved from a lineage of theropod dinosaurs dur-
ing the Jurassic Period and they get cancer today [20], so
How can you translate a solution in some other we can say that, yes, absolutely, dinosaurs did and still
species to prevent cancer in humans? do get cancer. Cancer incidence in extant birds is lower
Ideally, comparative studies could highlight potential tar- than in mammals [25], and the reasons for this are not
gets where the genetic mechanisms underlying cancer well known and are worthy of future study. There is a
suppression in one species could be transferred to an- fossil record of cancer for extinct non-avian dinosaurs as
other, with clinical implications. For instance, it was well; one study examined >10,000 vertebrae from >700
found that genetically altering mice to overexpress a individuals across the dinosaur phylogeny [26]. The
form of the TP53 protein conferred a cancer-suppressive study concluded that the only examined non-avian dino-
phenotype; however, these mice also displayed a prema- saurs with bone neoplasms were large, elephant-sized
ture ageing phenotype [18]. Surprisingly, another study herbivores classified as hadrosaurs, also known as duck-
created ‘super p53’ mice which contained extra copies of billed dinosaurs, including one instance of metastatic
the TP53 gene—similar to the elephant genome—under cancer in a caudal vertebrae (tail bone) from an individ-
their normal promoters, and these mice revealed an en- ual Edmontosaurus. A later study described a set of be-
hanced DNA damage response and cancer suppression nign neoplasms in a titanosaur [27], which was a
without the ageing effect [19]. Work is now underway to member of the sauropod family that included the largest
develop medicines based on the TP53 pathway. While terrestrial animals in Earth’s history. Whether the exclu-
the search for solutions to Peto’s Paradox across a diver- sivity of non-avian dinosaurian malignant cancers to
sity of species is still in progress [5, 7, 9, 20], it will no hadrosaurs is due to a genetic component or some
Tollis et al. BMC Biology (2017) 15:60 Page 4 of 5

vagary or artifact of the fossil record is not known. How- find new therapeutic targets and approaches to cancer
ever, similar sample sizes across many dinosaur clades prevention to save human lives. Cancer has been part of
were examined, so it appears to be statistically signifi- the story of the evolution of multicellularity [29], and it
cant, and cancer may have occurred often enough in is obvious that many lineages have evolved ways to cope
hadrosaurs so that some affected individuals became fos- with this disease. Now we are living in the midst of
silized and eventually observed by paleontologists. Earth’s sixth mass extinction, with extinction rates pos-
sibly 1000 times the historical rate [30]. Investigations
Why did hadrosaurs get so much cancer but not into Peto’s Paradox can help cancer prevention research
other large dinosaurs? as well as foster an appreciation for biodiversity and the
The answer may lie in life history differences. Elephants need to conserve species across the planet.
have slow life histories, taking decades to reach adult
Acknowledgements
size and investing heavily in the rearing of their off- This work was supported in part by NIH grants P01 CA91955, R01 CA149566,
spring. It makes sense that natural selection led to the R01 CA170595, R01 CA185138, and R01 CA140657 as well as CDMRP Breast
evolution of elephants with a means to suppress cancer, Cancer Research Program Award BC132057. The findings, opinions, and
recommendations expressed here are those of the authors and not
because only those that could suppress cancer lived long necessarily those of the universities where the research was performed or
enough and grew large enough to out-compete their ri- the National Institutes of Health.
vals. Can we apply this logic to extinct dinosaurs? The
fact that one out of 16 Edmontosaurus specimens had Authors’ contributions
All authors contributed to the writing of this manuscript and agreed to its
metastatic cancer [26] suggests that cancer occurred at a final content.
rate sufficient to ensure preservation in the fossil record.
Fortunately, some species of hadrosaur—including Competing interests
Edmontosaurus—left behind rich fossil deposits, giving The authors declare that they have no competing interests.

paleontologists a window into hadrosaurian demograph-

ics [28, 29]. Hadrosaurs lived very different lives to our Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
extant mammalian giants. They laid many eggs at a time published maps and institutional affiliations.
in huge nests, suggesting greater reproductive output,
and grew very rapidly, with some species reaching skel-
etal maturity in as little as 8 years. In addition, hadro-
References
saurs may have had much shorter lifespans than modern 1. Peto R, Roe FJ, Lee PN, Levy L, Clack J. Cancer and ageing in mice and men.
gigantic mammals, with senescence beginning soon after Br J Cancer. 1975;32:411–26.
skeletal maturity and some fossil deposits lacking any 2. Peto R. Epidemiology, multistage models, and short-term mutagenicity
tests. In: Hiatt HH, Watson JD, Winstein JA, editors. The origins of human
adults over ~16 years of age [28]. High reproductive out- cancer. New York: Cold Spring Harbor Laboratory; 1977. p. 1403–28.
put, rapid growth rates, and large body sizes may have 3. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell.
been traits necessary for species survival during the Late 2011;144:646–74.
4. Caulin AF, Graham TA, Wang L-S, Maley CC. Solutions to Peto's paradox
Cretaceous Period ~70 million years ago, when preda- revealed by mathematical modelling and cross-species cancer gene
tors like tyrannosaurs roamed the ancient floodplains. analysis. Phil Trans R Soc B. 2015;370:20140222.
Perhaps the trade-offs associated with these traits, such 5. Abegglen LM, Caulin AF, Chan A, Lee K, Robinson R, Campbell MS, et al.
Potential mechanisms for cancer resistance in elephants and comparative
as lack of sufficient DNA repair mechanisms [22], left cellular response to DNA damage in humans. JAMA. 2015;314:1850–60.
hadrosaurs susceptible to cancer once reaching a certain 6. Roche B, Hochberg ME, Caulin AF, Maley CC, Gatenby RA, Misse D, et al.
size. Other large dinosaurs may have had slower life his- Natural resistance to cancers: a Darwinian hypothesis to explain Peto’s
paradox. BMC Cancer. 2012;12:387.
tory strategies and so were under selection to evolve 7. Sulak M, Fong L, Mika K, Chigurupati S, Yon L, Mongan NP, et al. TP53 copy
more effective cancer suppression mechanisms. number expansion is associated with the evolution of increased body size
and an enhanced DNA damage response in elephants. Elife. 2016;5:1850.
8. Yim H-S, Cho YS, Guang X, Kang SG, Jeong J-Y, Cha S-S, et al. Minke whale
Why should I care about Peto’s Paradox? genome and aquatic adaptation in cetaceans. Nat Genet. 2014;46:88–92.
If cancer suppression has repeatedly evolved as a trait, 9. Keane M, Semeiks J, Webb AE, Li YI, Quesada V, Craig T, et al. Insights into
then basic research into the life history, genomics, and the evolution of longevity from the bowhead whale genome. Cell Rep.
2015;10:112–22.
cell biology of many different organisms can ultimately 10. Kingsolver JG, Pfennig DW. Individual level selection as a cause of Cope's
lead to better therapeutic strategies. If Peto’s Paradox Rule of phyletic size increase. Evolution. 2004;58:1608–12.
holds across the tree of life, and large long lived organ- 11. Heim NA, Knope ML, Schaal EK, Wang SC, Payne JL. Cope’s rule in the
evolution of marine animals. Science. 2015;347:867–70.
isms do in fact get less cancer than their body size war- 12. Benson RBJ, Campione NE, Carrano MT, Mannion PD, Sullivan C, Upchurch
rants, then it implies numerous mechanisms exist to P, et al. Rates of dinosaur body mass evolution indicate 170 million years of
prevent neoplastic progression and battle cancer. Every sustained ecological innovation on the avian stem lineage. Plos Biol. 2014;
12, e1001853.
time we discover a potential mechanism for cancer sup- 13. Baker J, Meade A, Pagel M, Venditti C. Adaptive evolution toward larger size
pression in a species, there is the chance that we can in mammals. Proc Natl Acad Sci U S A. 2015;112:5093–8.
Tollis et al. BMC Biology (2017) 15:60 Page 5 of 5

14. Seluanov A, Hine C, Azpurua J, Feigenson M, Bozzella M, Mao Z, et al.

Hypersensitivity to contact inhibition provides a clue to cancer resistance of
naked mole-rat. Proc Natl Acad Sci U S A. 2009;106:19352–7.
15. Tian X, Azpurua J, Hine C, Vaidya A, Myakishev-Rempel M, Ablaeva J, et al.
High-molecular-mass hyaluronan mediates the cancer resistance of the
naked mole rat. Nature. 2013;499:346–9.
16. Gorbunova V, Hine C, Tian X, Ablaeva J, Gudkov AV, Nevo E, et al. Cancer
resistance in the blind mole rat is mediated by concerted necrotic cell
death mechanism. Proc Natl Acad Sci U S A. 2012;109:19392–6.
17. Nagy JD, Victor EM, Cropper JH. Why don't all whales have cancer? A novel
hypothesis resolving Peto's paradox. Am Zool. 2007;47:317–28.
18. Tyner SD, Venkatachalam S, Choi J, Jones S, Ghebranious N, Igelmann H,
et al. p53 mutant mice that display early ageing-associated phenotypes.
Nature. 2002;415:45–53.
19. García-Cao I, Cao MG, Caballero JM, Criado LM, Klatt P, Flores JM, et al.
“Super p53” mice exhibit enhanced DNA damage response, are tumor
resistant and age normally. EMBO J. 2002;21:6225–35.
20. Møller AP, Erritzøe J, Soler JJ. Life history, immunity, Peto's paradox and
tumours in birds. J Evol Biol. 2017;30:960–7.
21. Martineau D, Lemberger K, Dallaire A, Labelle P, Lipscomb TP, Michel P,
et al. Cancer in wildlife, a case study: beluga from the St. Lawrence estuary,
Québec, Canada. Environ Health Perspect. 2002;110:285–92.
22. Boddy AM, Kokko H, Breden F, Wilkinson GS, Aktipis CA. Cancer
susceptibility and reproductive trade-offs: a model of the evolution of
cancer defences. Phil Trans R Soc B. 2015;370:20140220.
23. Brown JS, Aktipis CA. Inclusive fitness effects can select for cancer
suppression into old age. Phil Trans R Soc B. 2015;370:20150160.
24. Williams GC. Pleiotropy, natural selection, and the evolution of senescence.
Evolution. 1957;11:398–411.
25. Effron M, Griner L, Benirschke K. Nature and rate of neoplasia found in
captive wild mammals, birds, and reptiles at necropsy. J Natl Cancer Inst.
1977;59:185–98.
26. Rothschild BM, Tanke DH, Helbling M, Martin LD. Epidemiologic study of
tumors in dinosaurs. Naturwissenschaften. 2003;90:495–500.
27. Barbosa FH de S, Pereira PVLG de C, Bergqvist LP, Rothschild BM. Multiple
neoplasms in a single sauropod dinosaur from the Upper Cretaceous of
Brazil. Cretaceous Res. 2016;62:13–7.
28. Woodward HN, Freedman Fowler EA, Farlow JO, Horner JR. Maiasaura, a
model organism for extinct vertebrate population biology: a large sample
statistical assessment of growth dynamics and survivorship. Paleobiology.
2015;41:503–27.
29. Aktipis CA, Boddy AM, Jansen G, Hibner U, Hochberg ME, Maley CC, et al.
Cancer across the tree of life: cooperation and cheating in multicellularity.
Phil Trans R Soc B. 2015;370:20140219–9.
30. De Vos JM, Joppa LN, Gittleman JL, Stephens PR, Pimm SL. Estimating the
normal background rate of species extinction. Conserv Biol. 2015;29:452–62.
31. Vanderven E, Burns ME, Currie PJ, Farke A. Histologic growth dynamic study
of Edmontosaurus regalis (Dinosauria: Hadrosauridae) from a bonebed
assemblage of the Upper Cretaceous Horseshoe Canyon Formation,
Edmonton, Alberta, Canada. Can J of Earth Sci. 2014;51:1023–33.