Theoretical Biology and Medical

Modelling BioMed Central

Research Open Access

The biological sense of cancer: a hypothesis
Raúl A Ruggiero* and Oscar D Bustuoabad

Address: División Medicina Experimental, Instituto de Investigaciones Hematológicas, Academia Nacional de Medicina de Buenos Aires, Pacheco
de Melo 3081, 1425 Buenos Aires, Argentina
Email: Raúl A Ruggiero* - [email protected]; Oscar D Bustuoabad - [email protected]
* Corresponding author

Published: 15 December 2006 Received: 25 September 2006

Accepted: 15 December 2006
Theoretical Biology and Medical Modelling 2006, 3:43 doi:10.1186/1742-4682-3-43
This article is available from: http://www.tbiomed.com/content/3/1/43
© 2006 Ruggiero and Bustuoabad; licensee BioMed Central Ltd.
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0),
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Abstract
Background: Most theories about cancer proposed during the last century share a common
denominator: cancer is believed to be a biological nonsense for the organism in which it originates,
since cancer cells are believed to be ones evading the rules that control normal cell proliferation
and differentiation. In this essay, we have challenged this interpretation on the basis that,
throughout the animal kingdom, cancer seems to arise only in injured organs and tissues that display
lost or diminished regenerative ability.
Hypothesis: According to our hypothesis, a tumor cell would be the only one able to respond to
the demand to proliferate in the organ of origin. It would be surrounded by "normal" aged cells that
cannot respond to that signal. According to this interpretation, cancer would have a profound
biological sense: it would be the ultimate way to attempt to restore organ functions and structures
that have been lost or altered by aging or noxious environmental agents. In this way, the features
commonly associated with tumor cells could be reinterpreted as progressively acquired
adaptations for responding to a permanent regenerative signal in the context of tissue injury.
Analogously, several embryo developmental stages could be dependent on cellular damage and
death, which together disrupt the field topography. However, unlike normal structures, cancer
would have no physiological value, because the usually poor or non-functional nature of its cells
would make their reparative task unattainable.
Conclusion: The hypothesis advanced in this essay might have significant practical implications. All
conventional therapies against cancer attempt to kill all cancer cells. However, according to our
hypothesis, the problem might not be solved even if all the tumor cells were eradicated. In effect,
if the organ failure remained, new tumor cells would emerge and the tumor would reinitiate its
progressive growth in response to the permanent regenerative signal of the non-restored organ.
Therefore, efficient anti-cancer therapy should combine an attack against the tumor cells
themselves with the correction of the organ failure, which, according to this hypothesis, is
fundamental to the origin of the cancer.

Background appeared on earth [1,2] and human beings since prehis-

Cancers as well as benign neoplasias are very old diseases, toric times [1,3]. Written records concerning cancer can be
which have afflicted animals since long before man traced to ancient Egypt [4]. However, there is consensus

Page 1 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

that only during the past 100 years has a truly scientific echinoderms, nematodes, sipunculides [17-20], etc. or
approach to malignant diseases emerged as a result of the parts of the body, as in the upper body regions of Planaria,
mounting and concerted efforts of clinical physicians, phylum Platyhelminthes [21]; hind limbs of urodele
experimentalists and theoretical scientists. Since the late amphibians [13,22]; etc. Conversely, cancer is relatively
1970, different alterations in cellular genes as well as in frequent in animals that display weak regenerative ability
several intracellular transducing signaling pathways have throughout their bodies, such as vertebrates others than
been identified in cancer cells, and on this basis a unified urodele amphibians, arachnids, insects [13,19,23-26],
genetic theory of carcinogenesis has been advanced [5-8]. etc., "weak" meaning the ability to repair or regenerate rel-
atively simple structures only, as in compensatory hyper-
This theory states that cancer starts and ends with the plasia of the liver, skin regeneration, etc. A similar
malignant cell, in which genetic changes lead to constitu- relatively high frequency of tumors has been observed in
tive activation of some genes (oncogenes) and/or inacti- the body regions of urodele amphibians that cannot
vation of others (anti-oncogenes or tumor suppressor regenerate [27,28].
genes). allowing that cell to evade – in all or in some
microenvironments – the mechanisms controlling cell 2) In animals in which cancer is relatively frequent, cancer
proliferation. These genetic changes would define the incidence rises exponentially with age [29]. In addition,
molecular and cellular attributes of the cancer cell, which, when cancer develops in young animals, it is usually asso-
in turn, should be the target of specific therapies against ciated with injured organs and tissues such as cirrhotic
cancer. This theory has the enormous merit of unifying, liver, gastric tissues exhibiting chronic atrophic gastritis,
through an immediate common cause, the numerous dif- radiation-damaged skin, colon displaying ulcerative coli-
ferent mediate causes of cancer such as chemicals, radia- tis, breasts of nulliparous women, non-secreting prostate
tion, viruses, etc. However, it has some theoretical alveoli, etc., which may have exhausted or diminished
difficulties, which have been addressed [9-11] by authors their regenerative abilities [13,30,31].
who have also emphasized that cancer remains a major
cause of morbidity and mortality, despite the explosive 3) In animals displaying a strong regenerative ability,
development of our knowledge about the molecular reparative or/and regenerative mechanisms remain fairly
mechanisms associated with the control of cell cycle and efficient throughout life [32]. On the other hand, in ani-
survival [12]. Of course, these theoretical difficulties and mals displaying a weak regenerative ability, reparative or/
the persistent failure in treating cancer do not necessarily and regenerative mechanisms are efficient mainly during
imply that the unified genetic theory of carcinogenesis is youth; as these animals age, cellular loss increases and
incorrect. However, they encourage us to explore other those mechanisms wane progressively [33].
possible theoretical approaches.
Corollaries
In this paper, on the basis of ideas advanced by Prehn, 1) Throughout the animal kingdom, cancer is rarely – if
Zajicek, Bissell, Duesberg, Sonnenschein and Soto [9,13- ever – induced in organs (or tissues) displaying an effi-
16] among others, we propose a hypothesis of cancer that cient reparative or regenerative mechanism, "efficient"
does not consider it an autonomous entity disobeying the meaning the ability of organs and tissues to regenerate
mechanisms controlling cell proliferation, but one themselves numerically and functionally. In effect, when
dependent on a reparative signal originating in the partic- these mechanisms remain fairly efficient throughout life –
ular environment of an injured tissue with diminished or even under the action of putative noxious agents – as they
exhausted reparative ability. Hopefully, this hypothesis do in animals displaying strong regenerative ability, can-
might help to reconcile some apparently contradictory cer never (or almost never) occurs. When they remain effi-
approaches entailed in the unified genetic and some alter- cient only during youth – and even during youth, some
native theories of carcinogenesis, improving our under- noxious agents can deplete them – as they do in animals
standing of the relationship among aging, regeneration displaying weak regenerative ability, cancer occurs mainly
and cancer. in aging individuals and also in injured organs from
young individuals that may have exhausted their regener-
Postulates ative ability because of the action of those noxious agents.
This hypothesis is based on three postulates:
2) Homeostasis in organs or tissues with mitotic potential
1) Throughout the animal kingdom, cancer is rarely – if would be maintained by regulatory fields, "regulatory
ever – produced in body regions displaying strong regener- field" meaning the existence of inhibitory and stimulatory
ative ability, "strong" meaning the ability to regenerate signals for cell proliferation and differentiation within the
complex structures such as a whole limb. These regions space of an organ or tissue. Both types of signal, regardless
can encompass the whole body, as in sponges, cnidarians, of their molecular nature, would not be symmetric. In

Page 2 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

effect, when a reparative or regenerative mechanism is mitotic ability would emerge by chance. This new variant
efficient, all cellular loss is compensated by cellular divi- would begin to divide; and if it were poorly functional or
sion until the organ attains its original size and function, non-functional, the organ would be numerically but not
after which all new mitoses are inhibited. This inhibitory functionally restored. In consequence, it would not score
signal, associated with the "right" number of normal the regeneration as effective and it would continue to send
functional cells located in the "right" place, must be mitotic signals to restore the lost or diminished organ
obeyed not only by the normal cells of the organ but function. As a result, the new variant would grow over and
also by all putative anomalous cells that could have over and the outcome would be a tumor. On the other
emerged within the organ by chance, injury or other hand, if the emergent new variant were functionally
cause. In effect, if these anomalous cells could disobey the active, normal function might be restored and this
inhibitory signal and grow autonomously, cancer could "restored" organ might, in most cases, mimic the negative
develop rather easily in an organ exhibiting an efficient regulatory field associated with the intact organ, after
reparative or regenerative mechanism, contradicting corol- which further mitosis would be halted. In a few cases,
lary 1. In contrast, the mere existence of an organ display- however, the new variant – even if functional – might be
ing an inefficient reparative mechanism means that some unable to mimic that negative regulatory field (for exam-
or all of their cells could occasionally be non-responsive ple, because of aberrant cellular features not directly
to the stimulatory signal associated with (or produced by) related to function) and in such cases a tumor would also
the "less than right" number of functional cells of that be produced. In the case of poorly functional or non-func-
organ. The concept of the "right" number of cells in the tional variants, the tumor would be poorly functional or
"right" place can be elucidated by the following example: non-functional, as most tumors are. On the other hand, in
when a liver is intact, no proliferation of hepatocytes the special cases of functional variants producing tumors,
occurs; when it is partially excised and regenerative ability they would be functioning ones, such as some adenomas
is normal, proliferation occurs until the liver attains its or some papillary and follicular carcinomas of the thy-
original size and function. The number of hepatocytes in roid.
the intact liver would be the "right" number of functional
cells, which would induce or produce an inhibitory sig- Many authors have highlighted the critical importance of
nal(s) for the hepatocytes. Proliferation of hepatocytes injury in the development of cancer [31,34-37], and the
after partial hepatectomy would not be prevented by idea that cancer actually behaves as a wound healing proc-
ectopic implantation of liver cells, meaning that these ess has been suggested by Dvorak [38]. Others have chal-
ectopic cells would not be in the "right" place for sending lenged this interpretation [39,40], but a critical
inhibitory signals to prevent hepatocyte proliferation in examination of their data reveals that they scored only
the remnant liver. massive necrosis and overt degenerative changes as
"injury", dismissing less evident injuries such as lost or
Origin of tumor cells diminished function of the whole organ or part of the
What, according to this hypothesis, is the putative origin organ, apoptosis, cellular senescence, etc. These are as rel-
of cancer? evant as massive or overt injury for this hypothesis,
because both demand a regenerative response.
We have said that cancer would not be induced in organs
(or tissues) exhibiting an efficient regenerative mecha- Cellular heterogeneity, and a genomic instability phase
nism. However, when an organism becomes aged and its during stages of high-grade dysplasia prior to the acquisi-
regenerative ability is progressively lost, any injury caus- tion of a frankly malignant phenotype, are two well-doc-
ing loss of cells or cellular function cannot be compen- umented (though so far unexplained) phenomena
sated by cellular division. In consequence, the original [33,41]. Similarly well-documented are the picture of a
size and function of the organ cannot be restored. tumor arising in a tissue surrounded by "normal" arrested
cells, and the existence of factors involved in organ and
We suggest that this situation induces a "crisis", which, tissue regeneration that enhance or are necessary for
through putative danger signals resulting from retardation tumor growth [15,36,42]. Moreover, under certain condi-
of tissue repair, acceleration of cell loss and functional tions, the immune response might play a role in tissue
compromise, might create an environment capable of regeneration, and in that case it would stimulate rather
promoting some degree of variability in the remaining than inhibit tumor growth [43,44].
live but arrested cells of the injured organ. The outcome of
this situation would be the emergence of some genetically In summary, according to this hypothesis, cancer would
and/or epigenetically modified cell variants. Most of these originate on the basis of three conditions:
would still lack the ability to divide in response to the
organ demand, but sooner or later a variant bearing that

Page 3 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

a) An injury of the affected organ (or tissue), "injury" upon stimulation [47]. Therefore, while the neuronal tis-
meaning not only partial removal of the organ, massive sue of the brain remains intact, no extracellular inhibitory
necrosis or extensive degenerative change but also less evi- signals seem to be necessary to keep its cells arrested. On
dent deleterious effects such as lost or diminished func- the other hand, when that tissue is injured, probably no
tion of the whole or a part of the organ, apoptosis, cellular stimulatory signals will be generated. In consequence,
senescence, etc. according to the hypothesis, no primary condition exists
for tumor initiation.
b) The impossibility of restoring the injury to that organ,
and the consequent existence of a permanent reparative Properties of tumor growth
signal to the remaining live cells. Since tumor growth does not restore the negative regula-
tory field associated with the intact organ, the "crisis"
c) The existence or emergence of atypical cells able to would persist and, as a consequence, new variants would
respond to the mitotic reparative signal of the injured be forced to emerge continuously by chance in the "nor-
organ but unable to mimic the negative regulatory field mal" resting tissue as well as within the growing tumor. In
associated with the intact organ. fact, new cellular variants have been found in the "nor-
mal" tissue surrounding a tumor [48,49]. In the same way,
Our hypothesis about the origin of cancer seems to work new variants continuously emerging in the tumor itself
regardless of which hypothesis we adopt for the control of could account for the cellular heterogeneity typically
the cell proliferation. In effect, if we adopt the stimulatory observed in both experimental and clinical tumors [50].
or positive hypothesis [45], the regenerative signals will
be represented by different kinds of growth factors In addition, since the speed of regeneration of a partially
depending on the tissue or organ involved. In the same removed organ or tissue is greatest at the outset of the
way, the diminished or lost expression of at least one of process, when the lack of function is maximal [51], our
the numerous molecular steps in the growth factor signal- hypothesis would predict that the more undifferentiated
ing pathway in normal aged cells – and, conversely, the and non-functional the tumor cell, the faster its growth,
existence of a responsive pathway in cancer cells – might because for all practical purposes, "regeneration" by non-
explain why the latter can proliferate in an organ where functional tumor cells would always simulate the outset
normal aged cells cannot. On the other hand, if we adopt of the normal regeneration process. The faster growth of
the inhibitory or negative hypothesis [45], the regenera- more undifferentiated tumors compared with more differ-
tive signals will be represented by the absence of some entiated ones is a common but not yet satisfactorily
kinds of inhibitory factors (chalones, TGF-β among oth- explained phenomenon in tumor biology [46,52].
ers). In the same way, the constitutive expression of at
least one step in the inhibitory signaling pathway in nor- The nature of the tumor cell
mal aged cells – and, conversely, the absence of such con- The most intriguing consequence of this hypothesis con-
stitutive expression in tumor cells – might explain why cerns the nature of the tumor cell itself. During the past
tumor cells can proliferate while normal aged cells can- century, many quite different theories and hypotheses
not. about cancer have been proposed (reviewed in
[45,46,51]). Despite their wide differences, most of these
A plausible objection may be raised about the origin of accounts agree that a frank or true tumor cell is autono-
cancer postulated by this hypothesis. If cancers originate mous, meaning that it is not subject to the rules and regu-
in injured organs or tissues that have exhausted or dimin- lations that control normal cell proliferation and survival.
ished regenerative capacities, they should be much more The concept of autonomy was originally enunciated in a
frequent in organs or tissues that display poor or null biological sense (classical definition of Ewing [53]), but
regenerative ability from birth. An obvious example is the main goal of experimental oncology has been "to
neuronal tissue in the human brain; however, this tissue understand it in the molecular sense", that is "to elucidate
actually exhibits fewer tumors than other organs and tis- the molecular definition of the cancer cell regardless of its
sues such as colon, breast, lung and skin [12,46]. The environment" [46].
answer to this objection might be as follows: as stated in
corollary 2, "regulatory fields" seem to be necessary to con- With the help of new molecular technologies, several
trol the proliferation of cells with mitotic potential, which intracellular transducing pathways have been elucidated
are found in almost all body organs and tissues. However, in the last 25 years and progress in dissecting these path-
the theory does not require that "regulatory fields" control ways "has begun to lay out a circuitry that will likely
the proliferation of postmitotic cells such as brain neu- mimic electronic integrated circuits in complexity and
rons, because they would not proliferate on their own, as finesse, where transistors are replaced by proteins (e.g.
shown by their inability to re-enter the cell cycle even kinases and phosphatases) and the electrons by phos-

Page 4 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

phates and lipids, among others" [6]. Some of these path- are many reported genetic and even heritable epigenetic
ways transmit stimulatory growth signals from the changes in different tumors [55,56], but these changes
extracellular medium to the nucleus, such as the mitogen- might not be the origin of cancer. Instead, they could be
activated protein kinase (MAP-kinase) cascade. Others reinterpreted as adaptations of cancer cells that enable
transmit inhibitory signals (most of them funneled them to respond to the demand of the aged organ to pro-
through the retinoblastoma protein, pRB, and its two rel- liferate in response to injury. Claims that several puta-
atives, p107 and p130), death signals (such as that initi- tively oncogenic mutations could be the result rather than
ated by Fas L), survival signals (such as that initiated by the cause of cancer are available in the literature [9-11,57].
IGF-1), etc. [6,54]. In this context, the constitutive expres-
sion of any step(s) in the stimulatory and/or survival sig- According to our hypothesis, any non-functional (and a
naling pathways (most of them related to the expression few aberrant functional) but mitotically active variant
of known "protooncogenes"), or the constitutive block- present in an injured "aged" organ – with exhausted or
ade of any step(s) in the inhibitory and/or death signaling diminished regenerative capacity – could behave as a
pathways (most of them related to the expression of some tumor cell. But the same cell put into a "young" organ
known "antioncogenes"), or a combination of both, with an intact regenerative capacity would behave as a
would confer the capacity for autonomous growth on the normal cell. Moreover, in very special situations, even
cell. absolutely normal functional cells could behave as tumor
cells. For example, when an inert foreign body (such as a
Some authors have claimed that this autonomy is not glass cylinder) is subcutaneously implanted in a mouse,
absolute but relative, meaning that the expression of some tissue homeostasis is disrupted and, in consequence, a
oncogenes or the silencing of some antioncogenes may regenerative signal must be produced. If the tissue is
generate cancer in some but not all environments. This "young", absolutely normal cells will proliferate to repair
contention was originally suggested by the classical exper- it, but the presence of the foreign body would not allow
iments of Brinster and Mintz and Illmense, demonstrating the repair to be effected. Therefore, the regenerative signal
that the malignant potential of teratocarcinoma cells would continue (presumably because although there are
could be constrained if they were injected into the blasto- sufficient normal functional cells to heal the injury, they
cyst; the resulting mice contained tumor-free tissues are in the "wrong" place), and a tumor-like proliferation
derived from the teratocarcinoma cells [15]. Further evi- of exclusively normal cells would result. The "crisis" gen-
dence is available to support this claim. For example, erated by the unresolved disruption of homeostasis would
infection of adult chickens with Rous Sarcoma Virus persist, and eventually new non-functional variants
(RSV) leads to malignant transformation associated with would emerge, better adapted to respond to the regenera-
the expression of the oncogene v-src; however, infection tive stimulus; these would be the origin of the late sarco-
of chick embryos in ovo with RSV does not lead to malig- mas observed in such cases [58,59]. The existence of a
nant transformation, even though v-src is both expressed tumor-like proliferation of normal mesenchymal cells,
and active [15]. In the same way, expression of v-myc and relatively early after foreign body implantation, is a well-
c-myc is typical of some tumors, but myc is also expressed documented observation [58].
in echinoderms, which never develop tumors [17]. In any
case, irrespective of whether a tumor cell is considered Our suggestion that a tumor cell is not autonomous but
absolutely or relatively autonomous, there is a consensus dependent on a reparative or regenerative signal originat-
that it has molecular anomalies that allow it to escape – in ing in an "aged" organ or tissue seems heretical, because it
all or in some environments – from the regulatory mech- contradicts the classical definition of Ewing ("A neoplasm
anisms that inhibit normal cell proliferation in those is an autonomous, or relatively autonomous, growth of
environments. tissue"), which has guided cancer research for the last 60
or more years [53]. However, closer examination of
However, if the hypothesis advanced in this paper were Ewing's proposition reveals that it is a postulate rather
true, a tumor cell would not be one ignoring the mecha- than a true definition. First, pathologists do not use it as
nisms that control normal cell proliferation. In fact, in the an operational tool to diagnose the presence of a tumor;
injured organ where tumor originates, the tumor cell in fact, "the means to diagnose cancer have not changed
would be the only one able to respond to the organ that much since"... the 19th century, "when pathologists
demand to proliferate, surrounded by "normal" aged cells began describing the histological pattern of tumors using
that cannot respond to that signal. In this way, any attempt the light microscope" [45]. Second, if nobody knows
to find the molecular definition of the cancer cell, meaning the exactly what the mechanisms control normal cell prolifer-
molecular anomalies that allow the tumor cell to escape from ation [45], how can anyone be absolutely sure that cancer
the inhibitory signals of normal cell proliferation, might be an cells are disobeying those mechanisms? Some years ago,
attempt to find something that does not exist. Of course, there Dr Joseph Aub suggested that the "ugly word autonomy"

Page 5 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

be dropped, because while one can prove dependency, other interpretations [62,65], which in fact might comple-
one is never certain of autonomy [60]. ment ours) by assuming that the true basic unit at risk of car-
cinogenesis is the tissue or organ as a whole rather than the
The riddle of the blue whale and the mouse individual cell. In effect, according to the hypothesis, can-
The unified genetic as well as some (but not all) alterna- cer originates in organs or tissues that have exhausted or
tive theories of carcinogenesis share the idea that the diminished their regenerative capacities, and this would
malignant cell is the physiological and anatomical unit of occur when all or a critical proportion of their cells have
cancer disease. Implicit in this contention is the assump- partially or wholly lost that capacity. In such a case, if an
tion that the probability of origin of an aberrant, neoplas- organ were x times larger than another one, the probabil-
tic cell lineage is the same per unit of cell population, ity that its regenerative capacity is critically diminished
regardless of species or cell type concerned. would be x times lower, because an x times greater
number of cells would have to be affected to depress that
However, this assumption evokes one of the most intrigu- capacity. This lower probability would balance the pro-
ing riddles in cancer research, which remains unsolved. portionally higher number of their cells that could be
This riddle, stated by Dawe [20] some years ago, asks: transformed. As a result, if the unit at risk is, for example,
"Why don't extremely large animals develop neoplasms one liver rather than 109 (mouse) as opposed to 3 × 1015
with a much higher incidence than very small ones since (blue whale) liver cells, then the whale will be at no
the cell population at risk is greater by several orders of greater risk of developing liver cancer than the mouse, or
magnitude?" As an extreme example, let us consider the any other animal with an equally efficient defense mech-
blue whale and the mouse. "If one takes the weight of the anism against neoplasia. The idea that cancer is an organ
mouse as 30 g and that of the blue whale as 100 tons, the or tissue disease rather than a cellular one has been advo-
whale is equivalent to 3,030,303 mice. Then, if one cated especially by the group of Sonnenschein and Soto
accounts for differences of lifespan (65 years for the blue [45].
whale, 3 years for the mouse), the ratio of weight-year
units per whale to weight-year units per mouse is about Tumor progression. Invasion and metastases
66,670,000" [20]. We should therefore expect the blue Sooner or later, tumor growth will be restrained by the
whale to develop neoplasias about 3 × 106 and 6.6 × 107 rather rigid architecture of the organ or tissue in which the
times more often than the mouse per unit time and per tumor originated (first tissue). However, the persistent
lifespan, respectively. Since about 40% of wild mice kept "crisis" will force the emergence of new variants with the
under laboratory observation develop spontaneous neo- ability to disrupt that architecture, so growth can be re-ini-
plasias during their lives [61], we should expect each blue tiated. When these new variants reach the basal mem-
whale to develop about 2.6 × 107 neoplasms per lifespan. brane, they would eventually be able to disrupt it,
It is clear that these expectations do not match reality: the allowing the tumor cells to invade another tissue (second
incidence of neoplasia in whales, as in most mammals, is tissue). The claim that cancer cells can produce enzymes
roughly similar to that in mice. Therefore, the incidence of that destroy the matrix barriers surrounding the tumor,
neoplasia is not a simple function of protoplasm mass at permitting invasion into surrounding tissues, has signifi-
risk per unit time. In fact, the greater the body size of the cant experimental support [66,67].
animal, the greater seems to be its resistance to oncogene-
sis on a unit weight per unit time basis. Assuming that the second tissue is not injured and that its
regenerative capacity is intact, the invading tumor cells
Some ad hoc hypotheses have been invoked to account for would face an inhibitory signal from the second tissue
this fact on the assumption that the individual cell in an which – according to corollary 2 – they could not disobey.
organ or tissue is the unit at risk of carcinogenesis. For exam- At that point, the tumor cells might remain arrested indef-
ple, the animal fat depots might sequester fat-soluble car- initely. Alternatively, the arrested tumor cells might pro-
cinogens with an efficiency proportional to animal's size duce – directly, by releasing inhibitory factors, or
and thereby proportionately diminish the exposure of indirectly, by attracting inflammatory cells that in turn
other tissues. Another possibility is that the efficiency of release inhibitory factors – a lowering of the regenerative
defenses against neoplasia, such as mechanisms of DNA capacity of the second tissue. If an injury were incurred in
repair, cellular resistance to metabolism and mutagenic the second tissue, simultaneously or subsequently – most
activation of putative carcinogens, immunological sur- probably associated with the pre-acquired ability of the
veillance, etc., could be proportional to animal size. While tumor cells to disrupt the architecture of the first tissue –
these invoked mechanisms remain largely undemon- a stimulatory signal would appear, aimed at repairing the
strated as general rules [62-64], the hypothesis of cancer injured tissue. Since the regenerative capacity of the tissue
that we present in this paper could offer a relatively easy would thereby become exhausted or diminished, the
solution of the riddle (although not necessarily excluding tumor cells would have a selective advantage over normal

Page 6 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

cells to proliferate. Examples of this selective advantage erto "dormant tumor cells" could respond. They would
have been documented [68,69]. thus resume their progressive growth.

However, the tumor cells did not originate in the second Our hypothesis could operate not only for primary
tissue, and since repair or regeneration processes in differ- tumors but also for dormant metastases, the main clinical
ent tissues are generally independent of each other [45], problem. In effect, as stated in the preceding section
stimulatory signals from one tissue would not usually ("Tumor progression. Invasion and metastases"), when
induce the proliferation of cells from another. Why, then, tumor cells invade a second intact tissue, they would face
could the growth of tumor cells from the first tissue actu- an inhibitory signal that they could not disobey. At that
ally be stimulated by the stimulatory signal of the second? point, if these invading tumor cells were not able by them-
We suggest that the less the tumor resembles the primary selves to injure and deplete the regenerative capacity of
tissue (presumably the more undifferentiated it is), the that second tissue, they might remain arrested indefi-
more likely it would be to respond to the stimulatory sig- nitely, behaving as dormant metastases. Dormant metas-
nal of the second tissue and thus to grow in it. The same tases may awaken as a dormant primary does, even years
procedure could also explain why tumor cells can grow in or decades after the tumor cells were seeded in the second
distant organs (metastases), assuming that they can reach tissue, when this tissue becomes aged and loses its regen-
those organs. erative ability.

On the other hand, more differentiated tumor cells from The induction of tumor dormancy in secondary tumor
the first tissue could hardly grow in the second tissue implants in the presence of a primary growing tumor
unless the stimulatory signal from the first had reached (concomitant resistance phenomenon [73-75]) might
the orbit of the second. In that case, the tumor cells would also be interpreted according to this hypothesis, by
grow in the injured second tissue under the guidance of assuming that the local regenerative signal(s) promoting
the stimulatory signal from the first. This particular case tumor growth, generated at the site of secondary tumor
can be illustrated by the behavior of so-called hormo- implantation, could be counteracted by a diffusible inhib-
nally-conditioned tumors growing in secondary tissues or itory factor(s) produced or induced by the large primary
organs [60,70]. tumor [76].

Tumor dormancy Transplantability of tumors

Tumors can occasionally remain dormant for several The hypothesis advanced in this paper postulates that a
years, even decades; but suddenly, often in association tumor cell is never autonomous even in the case of inva-
with surgical stress or another injury, they can awake and sive and metastatic tumors. In effect, the mere existence of
resume progressive growth [46,71]. Some hypotheses heritable changes (genetic and/or epigenetic) that endow
have been advocated to explain this phenomenon [72] a cell with the ability to evade the rules controlling normal
but its nature remains obscure. cell proliferation would mean that these changes could
appear by chance in a normal cell within an organ with
According to the hypothesis presented in this paper, can- intact regenerative capacity. But if it were possible, cancer
cers originate in injured organs or tissues with exhausted could develop rather easily in that organ, contradicting
or diminished regenerative capacities. However, if this corollary 1.
exhausted or diminished capacity could sometimes be
recovered, normal cells could reassume their mitotic In this section, we consider an apparently fatal objection
potential and divide in response to the regenerative signal. to the hypothesis, which is one of the milestones in the
Of course, tumor cells would also divide in response to development of conventional ideas about cancer: the
that signal, but as the organ attained its "right size and transplantability of experimental tumors. In 1877, Novin-
function" – as a result of the growth of normal function- sky successfully transplanted tumors from adult to young
ally active cells – all new mitosis would be stopped, dogs for the first time. These experiments were reproduced
including that in tumor cells, according to corollary 2. That in 1888 by Moreau and later by Loeb and Jensen, using rat
could be the mechanism underlying the induction of a and murine tumors [51]. These pioneering experiments,
dormant tumor. The hypothesis could also offer a plausi- which became universal laboratory practice for more than
ble explanation for the awakening of the dormant tumor. a century, demonstrated that only a small fragment of a
In effect, after years or decades of dormancy, the organ tumor or a relatively small number of tumor cells dis-
could become aged, and therefore its regenerative ability persed in a physiological saline will suffice to transplant
could decrease irreversibly. In that situation, any injury that tumor from a donor to a recipient host. This implies
would induce a reparative signal to which only the hith- that the growth of a tumor does not need to be supported
by any tissue, organ or organismic pathological condition,

Page 7 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

but only by the nature of the tumor cells themselves. In if quite different from the organ of origin, tends to be
other words, tumor cells are autonomous, and this claim more similar to that organ than to others; in consequence,
means that our hypothesis would be false. However, the it would respond to a regenerative signal from the former
whole of this apparently fatal objection pivots on the better than to one from the latter, resulting in faster tumor
ambiguity of the word "autonomy". growth.

We can accept that tumor cells are deemed "autonomous" Carcinogenesis in vitro
if their inoculation into an appropriate recipient host is Carcinogenesis in vitro can also be considered an objec-
enough to induce new tumor growth (the first meaning of tion to our hypothesis. In effect, when "transformed cells"
autonomy). But this does not contradict our hypothesis, are produced in culture – spontaneously or induced by a
because the new tumor growth need not be accomplished given carcinogen – they are assumed to be endowed with
by evading the rules controlling normal cell proliferation the ability to evade normal inhibitory signals when
in the recipient host (the second meaning of autonomy). implanted into the organism. If this were true, it would be
That is, we can accept that tumor cells are autonomous in contradictory to corollary 2, because according to that cor-
the first sense, but not in the second sense. According to ollary no body cell can evade such signals.
our hypothesis, the mechanisms involved in tumor trans-
plantation would not differ markedly from those used by However, this conclusion is not unavoidable. It could
a tumor to invade adjacent or distant organs or tissues alternatively be proposed that so-called carcinogenesis in
within its primary host. In neither case would the tumor vitro produces cells with particular features that enable
cells be autonomous in the second sense of "autonomy", them to disrupt homeostasis in the organ or tissue into
because they would have to injure the recipient organ or which they are eventually implanted. This situation would
tissue and to eliminate or reduce its regenerative capacity initiate regenerative signals, which could be detected and
as a prerequisite for regenerative signals produced by the utilized by the in vitro" transformed" cells, promoting
injured organ or tissue to promote tumor growth. growth in a setting in which normal cells would have been
prevented from growing. That is, the putative objection of
Our contention concerning the mechanisms underlying "carcinogenesis in vitro" could be reducible to the objec-
tumor transplantation have significant experimental sup- tion of "transplantability of tumors", which we addressed
port: in the preceding section.

a) Benign tumors, which are not invasive and commonly Carcinogens

produce little damage to host tissues, seldom – if ever – In this section we will consider another apparently fatal
grow when transplanted into another host [77]. objection to the hypothesis presented in this paper: the
existence of carcinogens. As Miller and Miller proposed
b) In chickens, tumors induced by Rous sarcoma virus [46]: "a carcinogen is an agent whose administration to
(RSV) typically form at the viral injection site but not at previously untreated animals leads to a statistically signif-
distant sites; the wound associated with the injection icantly increased incidence of malignant neoplasms as
seems to be required for local tumor growth, because compared with that in appropriate control animals". The
additional tumors can be induced at distant sites simply most prevalent interpretation of this definition, mainly
by wounding the infected birds [15]. based on the putative mode of action of chemicals, radia-
tion and oncogenic viruses, suggests that most carcino-
c) The liver of a young rat, but not of an aged rat in which gens exert their critical effects by inducing genetic changes
regenerative capacity is diminished or lost, can normalize that endow the affected cells with the ability to grow inde-
the morphology and growth capacity of transplanted pendently of the mechanisms controlling normal cell pro-
hepatocarcinoma cells. The most successful normaliza- liferation. If this were the case, a cancer cell could emerge
tion occurred when cells were transplanted into the spleen in the middle of an otherwise normal organ or tissue,
and filtered as solitary cells into the liver without disrupt- directly contradicting corollary 1.
ing normal liver architecture. On the other hand, when
this architecture was disrupted by transplanting a greater However, closer examination of the available data sug-
number of malignant cells directly into the liver, normal- gests that this prevalent view is not as straightforward as is
ization was less likely to occur [78]. usually thought. In effect, cancer development with chem-
icals, radiation, DNA viruses and retroviruses in humans
d) Upon transplantation, tumors usually grow in anatom- and animals that lack oncogenes is a very prolonged proc-
ically correct (orthotopic) organs better than in hetero- ess, often lasting one third to two thirds of the life span of
topic ones [79]. This observation can be interpreted by the organism. This long period of development is associ-
assuming that an invasive and transplantable tumor, even ated with many adaptive cellular proliferative responses

Page 8 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

that may show a slow evolution to cancer [35]. For exam- In effect, although signals from v-onc genes have a domi-
ple, after treatment of rats with many different types of nant role in transformation, changes in cellular genes are
chemical hepatocarcinogens, rapid inhibition of cell pro- also required for transformation to occur. This contribu-
liferation and cellular death was observed in the liver. This tion is highlighted by the fact that some v-onc genes fail
early effect was followed by the appearance of clones of to transform certain kinds of primary cell cultures but can
resistant hepatocytes, which proliferated vigorously in transform established cell lines derived from them. Simi-
response to a proliferative stimulus in the hostile environ- larly, some cellular lineages can be both infected and
ment created by the carcinogen, in which the vast majority transformed, while others can be infected but not trans-
of hepatocytes, the non-resistant ones, were inhibited or formed, by a particular retrovirus carrying a v-onc gene
dead. The resistant hepatocyte nodules have physiological [81]. Furthermore, transgenic animals are usually suscep-
value; but later, as the carcinogen-mediated injury per- tible to spontaneous tumors involving the tissue (or tis-
sists, they can evolve into fully transformed cells [35,36]. sues) in which the transgenic oncogene is expressed.
Similarly, in Africa and Asia, infection with the hepatitis B However, in most cases, only a fraction of the animals
DNA virus early in life is associated with the appearance develop tumors from only a small subset of cells in the
of hepatocellular carcinomas 25 or 30 years later. Preven- infected tissue, and a long latent period is required, indi-
tion of this disease has been achieved by a vaccine against cating that expression of the transgenic oncogene is not
the virus, thus preventing hepatitis and the resulting dam- sufficient for tumor development. Similar conclusions
age to the liver. This damage, caused by the cytolysis of can be drawn from studies in which a tumor suppressor
virally-infected hepatocytes and the aberrant compensa- gene has been selectively disrupted alone or in association
tory proliferation of the surviving hepatocytes, seems to with the constitutive expression of a transgenic oncogene
be essential for the development of liver tumors since it is [81-83].
the common denominator of both virally- and non-
virally-associated hepatocellular carcinomas [45,80]. A clue to understanding the transforming effect of retrovi-
ruses carrying oncogenes to their target cells might be the
On the other hand, carcinogenesis by retroviruses that existence of a common denominator among the different
carry oncogenes or v-onc genes, such as Abelson murine lineages that are both infected and transformed by differ-
leukemia virus (Ab-MLV), Rous sarcoma virus (RSV), ent retroviruses. In all these lineages, expression of the
Avian erythroblastosis virus (AEV) etc., offers at first particular v-onc gene interferes primarily with the normal
glance a very different picture, because of their ability to differentiation of the cells that will be transformed. Con-
induce tumors rapidly and to transform cells in vitro. Reli- versely, when expression of the v-onc gene fails to arrest
able experiments, including the use of mutants lacking v- the differentiation of the infected cell, no transformation
onc genes and transfection assays using cDNA of v-onc occurs [81]. For example, Abelson murine leukemia virus
genes, have unambiguously demonstrated that these (Ab-MLV), a virus that normally arrests differentiation of
genes are both necessary and sufficient for the transform- pre-B cells, induces pre-B lymphomas from a small subset
ing ability of such viruses. In addition, use of temperature- of the infected pre-B cells. In contrast, Ab-MLV infects
sensitive mutants has shown that the expression of pro- erythroid precursors but does not arrest their differentia-
tein(s) encoded by the v-onc gene(s) is essential for the tion and never induces transformation in this lineage. In
expression of the neoplastic phenotype. Furthermore, sev- fact, expression of the v-abl gene (the v-onc gene of Ab-
eral systems of regulation of gene expression in transgenic MLV) can stimulate erythropoeitin-independent differen-
mice have allowed controlled models of neoplasia initi- tiation of erythroid cells. Presumably, this reflects the abil-
ated by numerous oncogenes to be developed in a variety ity of v-Abl protein to mimic signals normally transmitted
of tissues [15,81]. Retroviruses that carry oncogenes are via the Epo receptor in a situation where the oncoprotein
not a significant cause of naturally-occurring tumors. cannot stimulate continued growth [81].
However, most researchers, stimulated mainly by the dis-
covery in normal cells of protooncogenes homologous to On the basis of the above considerations, we will now
viral oncogenes, have assumed that in all cancers, inde- advance an interpretation of retroviral carcinogenesis
pendently of their etiology and the duration of the prene- according to the postulates of our hypothesis. Consider a
oplastic process, the critical step driving a normal cell into schematic representation of a single hematopoietic nor-
a neoplastic one must be similar to that carried out by mal cell lineage, comprising a stem cell, some undifferen-
these retroviruses on their target cells [81]. If this were tiated mitotically active cells and some differentiated and
absolutely true, the hypothesis advanced in this paper functional postmitotic cells. The regulation of cell matura-
would again have to be rejected, because that critical step tion and turnover in a lineage is not completely under-
would be a single intracellular event independent of the stood at the molecular level. Nevertheless, the
environment in which the affected cell resides. However, differentiated cells of the lineage somehow control the
the final word may not have been said yet. proliferation of the less differentiated ones [51,84]. For

Page 9 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

example, when a differentiated cell dies, a restorative sig- phenolic compounds released from the wound trigger
nal is generated that induces an undifferentiated cell to both the attachment of A. tumefaciens to plant cells and
divide; one of the resulting cells will differentiate into a the expression of the vir regulon, which is necessary for
functional postmitotic cell while the other will remain transferring the oncogenic T-DNA from the bacterium to
undifferentiated, restoring the original function and struc- the cells [86,89]. However, no role in the proliferation of
ture of the lineage. transformed plant cells has been attributed to wounding,
since crown gall tumor growth has usually been assumed
However, when a retrovirus carrying a v-onc gene infects to depend only on the plant growth hormones produced
undifferentiated cells and arrests their differentiation, the by the proper transformed cells.
normal program of tissue regeneration will be damaged.
In effect, although all differentiated functional cells die, This interpretation contradicts the concept of tumor cells
promoting a strong regenerative signal, no undifferenti- advocated in this paper. However, more recent evidence
ated cell can now differentiate into functional cells, mean- seems to offer a different picture. A. tumefaciens was inoc-
ing that this lineage would lose its regenerative capacity. ulated in unwounded tobacco seedlings and new molecu-
Presumably, at early stages of infection, cells that fail to lar technologies were used to demonstrate that vir gene
differentiate could only divide three or four times before induction, T-DNA transfer and plant cell transformation
dying. At that moment, the stem cell would begin to were produced as they are in wounded plants. In contrast
divide to compensate the loss of undifferentiated cells, to wound sites, the transformed plant cells could not pro-
but these new undifferentiated cells would again be duce tumors [88], suggesting that, as long as tissue architec-
infected with the virus, rendering them unable to differen- ture is not disrupted, negative regulatory signals prevent
tiate. As a result, a "crisis" would generate a state of varia- growth of the transformed cells. On the other hand, such
bility, and undifferentiated variants not committed to die negative regulatory signals would tend to be reduced at
after a few mitoses would sooner or later emerge. These wound sites, and proliferation of transformed cells could
variants would divide over and over in response to the be initiated in consequence. Since growing galls retard or
regenerative signal, thus generating a neoplastic growth. inhibit the development of normal host tissues [90],
This suggests that the expression of a v-onc gene could be transformed cells would have a selective advantage to pro-
interpreted otherwise than as a single intracellular event liferate, and in consequence the wound would tend to be
that directly drives a normal cell into an autonomous one, filled only with transformed cells, which (as opposed to
as it usually is. Instead, this expression could be a power- normal wound-healing meristematic cells) display a lim-
ful force primarily arresting normal cell differentiation. ited ability to differentiate [86,88]. From that moment,
Only on that basis would a tumor emerge in a subset of tumor growth could proceed as described in the section
those arrested cells. That an impediment to normal cellu- "Origin of tumor cells", suggesting that the hypothesis
lar differentiation is an essential element in the formation presented in this paper might work even beyond the ani-
of malignant tumors has recently been suggested by Har- mal kingdom.
ris [85].
Anti-cancer treatments
All the above considerations suggest that carcinogenesis Despite many years of basic and clinical research and trials
induced by chemicals, radiation and oncogenic viruses, of promising new therapies, most cancers are resistant to
even retroviruses carrying viral oncogenes, considered as therapy at presentation or become resistant after an initial
the paradigm of the unified genetic theory of cancer, response [12,91,92]. All current conventional therapies
might be reinterpreted according to the postulates of the against cancer attempt to kill all cancer cells with minimal
hypothesis advanced in this paper. toxic side effects. A similar aim is pursued by some of the
new anti-cancer trials. However, according to our hypoth-
Plant tumors esis, even if all tumor cells were eradicated, the problem
It has long been known that the induction of crown gall might not be solved. In effect, if the organ failure
tumors by Agrobacterium tumefaciens in a wide variety of remained, new tumor cells would emerge and the progres-
plants depends on the existence of a wound, because inoc- sive tumor growth would be re-initiated in response to the
ulating the bacterium into intact plants rarely, if ever, permanent regenerative signal of the non-restored organ.
causes tumors [86-88]. However, the precise role of
wounding in each step of the tumorigenic process remains A theoretically attractive approach would be to make
unclear. tumor cells functional, because in that case the organ
function would be restored and no regenerative signal
The conventional interpretation states that the wound is would remain to promote new cellular growth. This ther-
necessary for transformation but not for tumor growth apeutic schedule is exemplified by the successful treat-
itself. In effect, previous experiments have suggested that ment of acute promyelocytic leukemia by retinoic acid-

Page 10 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

based therapies, resulting in the induction of differentia- Conclusion

tion of promyelocytes to mature cells [93]. However, in Despite their obvious differences and with few exceptions
general, attempts to differentiate tumor cells in vitro or in [14], most theories about cancer proposed during the last
vivo by using differentiating molecules have proved very century share a common denominator: cancer is believed
difficult [15,52]. On the other hand, full-blown tumor to be a biological nonsense for the organism in which it
cells become functional or behave as normal cells in the originates, since cancer cells are believed to be ones that
context of strong regenerative fields, stressing that the evade the rules controlling normal cell proliferation and
field in which tumor cells originate, and not only the differentiation. According to that understanding, cancer
tumor cells themselves, should be an important objective cells have usually been considered as cells returning to a
of our studies [15]. The possibility of manipulating the more primitive, unicellular, condition of life. Different
microenvironment and therefore the growth status of the features of tumor cells such as asocial behavior, reduced
cell has been demonstrated, for example, with cells that need for putative growth factors, ability to grow in over-
overexpress the hyaluronan receptor [94]. crowded settings, lack of contact inhibition, etc., have
been interpreted by most researchers as confirmation of
Does this mean that an efficient anti-cancer treatment that primitive way of life.
could be achieved by correcting the organ failure only?
Perhaps it depends on the stage of tumor progression. In In contrast, according to the hypothesis advanced in this
effect, to correct organ failure means that normal cells can paper, cancer would have a profound biological sense: it
resume their previously lost regenerative capacities and would be the ultimate attempt to restore the organ func-
normal organ function and/or structure can be restored. tion and structure that have been lost or altered by aging
Theoretically, this correction would stop tumor growth or environmental noxious agents, that is, an attempt to
because the regenerative signals – which, according to our evade the aging and the death of the organ or the organ-
hypothesis, always guide that growth – would have disap- ism (if the organ is essential for survival). Therefore, the
peared. However, that hopeful result could only be above-mentioned features of tumor cells could be reinter-
achieved if we faced tumor cells at the beginning of tumor preted as progressively acquired adaptations for respond-
growth. Later, when the tumor cells have evolved to ing to a permanent regenerative signal in the context of
become invasive and metastatic, they will have acquired tissue injury, just as several embryonic developmental
particular features endowing them with the capacity to stages such as morphological differentiation and mode-
produce new organ failures. In consequence, to correct the ling events could depend on cellular damage and death
organ failure only would be a transient solution; it would together with disruption of the topographic field [96].
work only until tumor cells damaged the organ again, However, unlike normal structures, cancer would have no
demanding new tumor growth. physiological value, because the usually poorly func-
tional or non-functional nature of its cells would make
A better putative therapy against full-blown tumor cells their reparative task unattainable.
might combine the correction of organ failure with an
attack against the tumor cells themselves. Perhaps it Under special circumstances, however, the attempt of
would not be necessary to eradicate all the tumor cells; tumors to correct organ failure or to evade death could
perhaps it would be enough to lower their number below have been successful. For example, fossil fish of the genus
a critical threshold (tumor dose 50?) by cytostatic rather Pachylebias that lived 8 million years ago adopted pachy-
than cytotoxic therapies, thus avoiding or reducing the ostosis to facilitate immersion in the hyper-saline water of
occurrence of new organ injuries that would vitiate the Mediterranean Sea, "through the development of dif-
attempts to correct the organ failure. Below this threshold, fuse hyperostosis that did not differ from a neoplastic
cancer cells would presumably be unable to exert any del- form of benign tumor originating from bone tissue" [1].
eterious effect on the organ. In consequence, if the organ Similarly, mammals of the Sirenidae group from the Oli-
failure were corrected, no new failure would be expected gocene acquired tumor-like forms "in the axial skeleton to
and any remaining tumor cells would therefore behave as consent browsing on the bottom in shallow waters" [1].
normal ones.
In other cases, tumors appeared as products of inter-spe-
However, to correct organ failure will demand under- cific associations between pairs of organisms. The classic
standing of the so far elusive nature of the regulatory fields examples are the insect-induced plant gall tumors, which
operating in normal organs. Hopefully, the study of serve a reproductive function for some groups of insects.
wound and wound healing-like phenomena in three- They represent a re-differentiation and neo-formation of
dimensional organotypic cultures [95], which maintain or host tissues, characterized by morphological and histolog-
mimic the natural organ structure, may provide valuable ical changes elicited by the developing insect that are
insights into the basic mechanisms of those fields. unique for both the inducing insect and the affected plant

Page 11 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

organ. In the more highly developed galls, these self-lim- References

iting neoplastic growths are almost comparable in the 1. Capasso LL: Antiquity of cancer. Int J Cancer 2005, 113:2-13.
2. Rothschild BM, Witzke BJ, Hershkovitz I: Metastatic cancer in the
determinate growth of their structures to a leaf or a fruit Jurassic. Lancet 1999, 354:398.
[87]. These considerations suggest that some neoplasms 3. Czarnetzki A, Schwaderer E, Pusch CM: Fossil record of meningi-
could have been adopted as a biological strategy to oma. Lancet 2003, 362:408.
4. Shimkin MB: Contrary to Nature Washington: Public Health Service,
increase the adaptation of some organisms to difficult National Institute of Health; 1977.
environmental conditions, allowing the "pathology" to 5. Bishop JM: The molecular biology of RNA tumor viruses: a
survive for millions of years. physician's guide. Basic science for clinicians. N Engl J Med
1980, 303:675-682.
6. Hanahan D, Weinberg RA: The hallmarks of cancer. Cell 2000,
Numerous predictions of the hypothesis advanced in this 100:57-70.
7. Bignold LP: The mutator phenotype theory can explain the
manuscript might be experimentally testable. For exam- complex morphology and behavior of cancer. Cell Mol Life Sci
ple: (I) the close relationship between strong regenerative 2002, 59:950-958.
ability and absence of tumors throughout the animal 8. Frank SA: Age-specific incidence of inherited versus sporadic
cancers: a test for the multistage theory of carcinogenesis.
kingdom; (II) the existence of an injury and a decreased Proc Natl Acad Sci USA 2005, 102:1071-1075.
regenerative capacity in a whole organ or tissue, or in a 9. Duesberg P, Rasnick D: Aneuplody, the somatic mutation that
makes cancer a species of its own. Cell Motil Cytoskel 2000,
part of that organ or tissue, before the emergence therein 47:81-107.
of a primary or secondary metastatic tumor; (III) the exist- 10. Sonnenschein C, Soto AM: Somatic mutation theory of carcino-
ence of danger signals resulting from a retardation of tis- genesis: why it should be dropped and replaced. Mol Carcino-
genesis 2000, 29:205-211.
sue repair, acceleration of cell loss and functional 11. Prehn RT: The role of mutation in the new cancer paradigm.
compromise, inducing hereditable changes (genetic or Cancer Cell Int 2005, 5:9-14.
epigenetic) of some kind in aging or injured cells with 12. Bailar JC III, Gornik HL: Cancer undefeated. N Engl J Med 1997,
336:1569-1574.
diminished or exhausted regenerative capacity; (IV) the 13. Prehn RT: Regeneration versus neoplastic growth. Carcinogen-
existence of cellular heterogeneity and a genomic instabil- esis 1997, 18:1439-1444.
14. Zajicek G: A new cancer hypothesis. Med Hypoth 1996,
ity phase in an organ or tissue before the acquisition of a 47:111-115.
frankly malignant phenotype, and in normal tissues sur- 15. Kenny PA, Bissell MJ: Tumor reversion: correction of malignant
rounding a tumor; (V) the ability of tumor cells, and the behavior by microenvironmental cues. Int J Cancer 2003,
107:688-695.
inability of surrounding normally aging cells, to respond 16. Maffini MV, Soto AM, Calabro JM, Ucci AA, Sonnenschein C: The
to local mitogenic or regenerative signals of the tissue in stroma as a crucial target in rat mammary gland carcinogen-
which the tumor has emerged; (VI) the capacity of trans- esis. J Cell Sci 2004, 117:1495-1502.
17. Walker CW, Boom JDG, Marsh AG: First non-vertebrate mem-
plantable tumor cells to injure and to diminish the regen- ber of the myc gene family is seasonally expressed in an
erative ability of the organ or tissue in which they can invertebrate testis. Oncogene 1992, 7:2007-2012.
18. Wellings SR: Neoplasia and primitive vertebrate phylogeny:
grow; (VII) the injurious action of carcinogens on cells of echinoderms, prevertebrates and fishes. A review. Natl Can-
organs and tissues as a prerequisite for inducing neoplas- cer Inst Monograph 1969, 31:59-128.
tic growth, injury meaning not only partial removal of the 19. Sparks AK: Review of tumors and tumor-like conditions in
Protozoa, Coelenterata, Platyhelminthes, Annelida, Sipun-
organ, massive necrosis or extensive degenerative change, culida and Arthropoda excluding Insects. Natl Cancer Inst Mon-
but also less evident deleterious effects such as lost or ograph 1969, 31:671-682.
20. Dawe CJ: Phylogeny and oncogeny. Natl Cancer Inst Monograph
diminished function of the whole organ or part of the 1969, 31:1-40.
organ, apoptosis, cellular senescence, etc. 21. Hall F, Morita M, Best JB: Neoplastic transformation in the
planarian: Cocarcinogenesis and histopathology. J Exp Zool
1986, 240:211-227.
Competing interests 22. Del Rio-Tsonis K, Tsonis PA: Amphibian tissue regeneration: a
The author(s) declare that they have no competing inter- model for cancer regulation (review). Int J Oncol 1992,
ests. 1:161-164.
23. Mechler BM, Strand D, Kalmes A, Merz R, Schmidt M, Torok I: Dro-
sophila as a model system for molecular analysis of tumori-
Authors' contributions genesis. Environ Health Perspect 1991, 93:63-71.
24. Taylor RL: Formation of tumor-like lesions in the cockroach
The two authors contributed equally to this work and they Leucophaea maderae after decapitation. Natl Cancer Inst Mono-
read and approved the final manuscript. graph 1969, 31:419-432.
25. Lopez A: Tumoral pathology in spiders. J Invertebr Pathol 1979,
34:224-230.
Acknowledgements 26. Caussinus E, Gonzalez C: Induction of tumor growth by altered
We are grateful to Dr Richmond T. Prehn and Dr Carlos M. Galmarini for stem-cell asymmetric división in Drosophila melanogaster.
their helpful and critical discussion of the manuscript and to Miss Victoria Nature Genet 2005, 37:1027-1028.
Ruival for excellent technical assistance. This work was supported by 27. Zilakos NP, Tsonis PA, Del Rio-Tsonis K, Parchment RE: Newt
squamous carcinoma proves phylogenetic conservation of
CONICET (Consejo Nacional de Investigaciones Científicas y Técnicas, tumors as caricatures of tissue renewal. Cancer Res 1992,
Argentina). Both authors are members of the Research Career, CONICET. 52:4858-4865.
This article is dedicated to the memory of an intelligent, honest and 28. Okanoto M: Simultaneous demonstration of lens regenera-
extraordinarily generous man, Mr. Juan J. Portaluppi, the technician's chief tion from dorsal iris and tumor production from ventral iris
in the same newt eye after carcinogen administration. Differ-
of our laboratory for almost 50 years. entiation 1997, 61:285-292.

Page 12 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

29. Kitagawa M, Utsuyama M, Kurata M, Yamamoto K, Yuasa Y, Ishikawa 57. Sato M, Vaughan MB, Girard L, Peyton M, Lee W, Shames DS, Rami-
Y, Arai T, Hirokawa K: Cancer and aging: symposium of the 27th rez RD, Sunaga N, Gazdar AF, Shay JW, Minna JD: Multiple onco-
annual meeting of the Japanese society for biomedical ger- genic changes (K-RAS (v 12), p53 knockdown, mutant
ontology. Tokyo Cancer Immunol Immunother 2005, 54:623-634. EGFRs, p16 bypass, telomerase) are not sufficient to confer
30. Karin M, Lawrence T, Nizet V: Innate immunity gone awry: link- a full malignant phenotype on human bronchial epithelial
ing microbial infections to chronic inflammation and cancer. cells. Cancer Res 2006, 66:2116-2128.
Cell 2006, 124:823-836. 58. Brand KG: Cancer associated with asbestosis, schistosomiasis,
31. Harrison SA, Bacon BR: Relation of hemochromatosis with foreign bodies and scars. In Cancer: A Comprehensive Treatise
hepatocellular carcinoma: epidemiology, natural history, Edited by: Becker FF. New York: Plenum Press; 1982:661-692.
pathophysiology, screening, treatment and prevention. Med 59. Pasqualini CD, Ruggiero RA, Bustuoabad OD, Nepomnaschy I, Piaz-
Clin North Am 2005, 89:391-409. zon I: Experimental onco-immunology revisited. Curr Cancer
32. Zamaraev VN: Does regenerative capacity decrease with age? Therapy Rev 2005, 1:289-298.
Sov J Dev Biol 1974, 4:495-503. 60. Furth J: Conditioned and autonomous neoplasms: a review.
33. Sharpless NE, DePinho RA: Telomeres, stem cells, senescence Cancer Res 1953, 13:477-492.
and cancer. J Clin Invest 2004, 113:160-168. 61. Andervont HB, Dunn TB: Occurrence of tumors in wild house
34. Haddow A: Cellular inhibition and the origin of cancer. Acta mice. J Natl Cancer Inst 1962, 28:1153-1163.
Unio Int Contra Cancrum 1938, 3:342-352. 62. Hahn WC, Weinberg RA: Rules for making human tumor cells.
35. Farber E, Rubin H: Cellular adaptation in the origin and devel- N Engl J Med 2002, 347:1593-1603.
opment of cancer. Cancer Res 1991, 51:2751-2761. 63. Rudo K, Meyers WC, Dauterman W, Langenbach R: Comparison of
36. Farber E: Cell proliferation as a major risk factor for cancer: a human and rat hepatocyte metabolism and mutagenic acti-
concept of doubtful validity. Cancer Res 1995, 55:3759-3762. vation of 2-acetylaminofluorene. Cancer Res 1987,
37. Simpson CJ, Bissell MJ, Werb Z: Mammary gland tumor forma- 47:5861-5867.
tion in transgenic mice overexpressing stromelysin-1. Semin 64. Hsu IC, Harris CC, Lipsky MM, Snyder S, Trump BF: Cell and spe-
Cancer Biol 1995, 6:159-163. cies differences in metabolic activation of chemical carcino-
38. Dvorak HF: Tumors: wounds that do not heal. Similarities gens. Mutat Res 1987, 177:1-7.
between tumor stroma generation and wound healing. N 65. Holliday R: Aging is no longer an unsolved problem in biology.
Engl J Med 1986, 315:1650-1691. Ann N Y Acad Sci 2006, 1067:1-9.
39. Hoel DG, Haseman JK, Hogan MD, Huff J, Mc Connell E: The impact 66. Coussens LM, Fingleton B, Matrisian LM: Matrix metalloprotein-
of toxicity on carcinogeneity studies: implications for risk ase inhibitors and cancer: trials and tribulations. Science 2002,
assessment. Carcinogenesis 1988, 9:2043-2052. 295:2387-2392.
40. Asare GA, Paterson AC, Kero MC, Khan S, Mossanda KS: Iron-free 67. Tan X, Egami H, Abe M, Nozawa F, Hirota M, Ogawa M: Involve-
neoplastic nodules and hepatocellular carcinoma without ment of MMP-7 in invasion of pancreatic cancer cells through
cirrhosis in Wistar rats fed a diet high in iron. J Pathol 2006, activation of the EGFR mediated MEK-ERK signal transduc-
208:82-90. tion pathway. J Clin Pathol 2005, 58:1242-1248.
41. Nery R: Carcinogenic mechanisms: a critical review and a 68. Elias D, De Baere T, Roche A, Ducreux M, Leclere J, Lasser P: During
suggestion that oncogenesis may be adaptive ontogenesis. liver regeneration following right portal embolization the
Chem Biol Interact 1976, 12:145-169. growth rate of liver metastases is more rapid than that of the
42. Schindel DT, Grosfeld JL: Hepatic resection enhances growth of liver parenchyma. Br J Surg 1999, 86:784-788.
residual intrahepatic and subcutaneous hepatoma, which is 69. García Alonso I, Palomares T, Alonso A, Portugal V, Castro B, Cara-
inhibited by octreotide. J Pediatr Surg 1997, 32:995-998. mes J, Mendez J: Effect of hepatic resection on development of
43. Pliskin ME, Prehn RT: Stimulation of liver regeneration and liver metastases. Rev Esp Enferm Dig 2003, 95:765-770.
compensatory kidney hyperplasia by passive transfer of 70. Li R, Wheeler T, Dai H, Frolov A, Thompson T, Ayala G: High level
spleen cells. J Reticuloendothel Soc 1975, 17:290-299. of androgen receptor is associated with aggressive clinico-
44. Prehn RT: An adaptive immune reaction may be necessary for pathologic features and decreased biochemical recurrence-
cancer development. Theor Biol Med Model 2006, 3:6. free survival in prostate: cancer patients treated with radical
45. Sonneschein C, Soto AM: The Society of Cells New York: Springer-Ver- prostatectomy. Am J Surg Pathol 2004, 28:928-934.
lag; 1999. 71. El Saghir NS, Elhajj II, Geara FB, Hourani MH: Trauma-associated
46. Pitot HC: Fundamentals of Oncology New York: Marcel Dekker Inc; growth of suspected micrometastasis. BMC Cancer 2005, 5:94.
1978. 72. Glick AB, Yuspa SH: Tissue homeostasis and the control of the
47. Geneser F: Histología. Buenos Aires: Editorial Médica Panamericana 2000. neoplastic phenotype in epithelial cancers. Semin Cancer Biol
(translated into Spanish from Histologi Copenhagen: Munksgaard; 2005, 15:75-83.
1999) 73. Gorelik E: Concomitant tumor immunity and resistance to a
48. Deng G, Lu Y, Zlotnikov G, Thor AD, Smith HS: Loss of heterozy- second tumor challenge. Adv Cancer Res 1983, 39:71-120.
gosity in normal tissue adjacent to breast carcinoma. Science 74. Ruggiero RA, Bustuoabad OD, Cramer P, Bonfil RD, Pasqualini CD:
1996, 274:2057-2059. Correlation between seric antitumor activity and concomi-
49. Mirsalis JC, Shimon JA, Johnson A, Fairchild D, Kanazawa N, Nguyen tant resistance in mice bearing non-immunogenic tumors.
T, de Boer J, Glickman B, Winegar RA: Evaluation of mutant fre- Cancer Res 1990, 50:7159-7165.
quencies of chemically induced tumors and normal tissues in 75. O'Reilly MS, Holmgren L, Sage EH, Folkman J: Angiostatin: a novel
I/c II transgenic mice. Environ Mol Mutagen 2005, 45:17-35. angiogenesis inhibitor that mediates the suppression of
50. Lengauer C, Kinzler KW, Vogelstein B: Genetic instabilities in metastases by a Lewis lung carcinoma. Cell 1994, 79:315-328.
human cancers. Nature 1998, 396:643-649. 76. Prehn RT: Two competing influences that may explain con-
51. Suss R, Kinzel V, Scribner JD: Cancer. Experiments and Concepts New comitant tumor resistance. Cancer Res 1993, 53:3266-3269.
York, Heidelberg, Berlin: Springer-Verlag; 1973. 77. Shimosato Y, Kameya T, Nagai K, Hirohashi S, Koide T, Hayashi H,
52. Cotran RS, Kumar V, Collins T, (editors): Tissue repair: cellular Nomura T: Transplantation of human tumors in nude mice. J
growth, fibrosis and wound healing. In Robbins Pathologic basis of Natl Cancer Inst 1976, 56:1251-1260.
disease 6th edition. New York: WB Saunders; 1999:89-112. 78. Rubin H: What keeps cells in tissues behaving normally in the
53. Ewing J: Neoplastic Disease: a Treatise on Tumors Philadelphia: Saunders; face of myriad mutations? BioEssays 2006, 28:515-524.
1940. 79. Nathanson SD, Nelson LT, Lee M: A spontaneous subcutaneous
54. Hahn WC, Weinberg RA: Modelling the molecular circuitry of tumor in C57BL/6 mice that metastatize to the liver. Clin Exp
cancer. Nature Rev Cancer 2002, 2:331-341. Metastasis 1993, 11:45-54.
55. Ionov Y, Peinado MA, Malkhosyan S, Shibata D, Perucho M: Ubiqui- 80. Maeda S, Kamata H, Luo JL, Leffert H, Karin M: IKKβ couples hepa-
tous somatic mutations in simple repeated sequences reveal tocyte death to cytokine-driven compensatory proliferation
a new mechanism for colonic carcinogenesis. Nature 1993, that promotes chemical hepatocarcinogenesis. Cell 2005,
363:558-561. 121:977-990.
56. Jaffe LF: Epigenetic theories of cancer initiation. Adv Cancer Res
2004, 90:209-230.

Page 13 of 14
(page number not for citation purposes)
Theoretical Biology and Medical Modelling 2006, 3:43 http://www.tbiomed.com/content/3/1/43

81. Rosenberg N, Jolicoeur P: Retoviral Pathogenesis. In Retroviruses

Edited by: Coffin JM, Hughes SH, Varmus H. New York: Cold Spring
Harbor Laboratory Press; 1997:475-585.
82. Adams JM, Cory S: Transgenic models of tumor development.
Science 1991, 254:1161-1167.
83. Suzuki A, Itami S, Ohishi M, Hamada K, Inoue T, Komazawa N, Senoo
H, Sasaki T, Takeda J, Manabe M, Wah Mak T, Nakano T: Keratino-
cyte-specific Pten deficiency results in epidermal hyperpla-
sia, accelerated hair follicle morphogenesis and tumor
formation. Cancer Res 2003, 63:674-681.
84. Pierce GB, Speers WC: Tumors as caricatures of the process of
tissue renewal: Prospects for therapy by directing differenti-
ation. Cancer Res 1988, 48:1996-2004.
85. Harris H: A long view of fashions in cancer research. BioEssays
2005, 27:833-838.
86. Formica JV: Crown gall neoplasms. In The Pathobiology of Neoplasia
Edited by: Sirica AE. New York: Plenum Press; 1989:497-512.
87. Bayer MH, Kaiser HE, Micozzi MS: Abnormal growth processes
in plants and animals: a comparison. In Vivo 1994, 8:3-16.
88. Brencic A, Angert ER, Winans SC: Unwounded plants elicit Ago-
bacterium vir induction and T-DNA transfer: transformed
plant cells produce opines yet are tumor free. Mol Microbiol
2005, 57:1522-1531.
89. Zhu J, Oger PM, Schrammeijer B, Houykaas PJJ, Farrand SK, Winans
SC: The bases of crown gall tumorigenesis. J Bacteriol 2000,
182:3885-3895.
90. Ullrich CI, Aloni R: Vascularization is a general requirement
for growth of plant and animal tumors. J Exp Botany 2000,
51:1951-1960.
91. Galmarini CM, Galmarini FC: Multidrug resistance in cancer
therapy: role of the microenvironment. Curr Opin Investig Drugs
2003, 4:1416-1421.
92. Cranmer LD, Trevor KT, Hersh EM: Clinical applications of den-
dritic cell vaccination in the treatment of cancer. Cancer
Immunol Immunother 2004, 53:275-306.
93. Tallman MS, Nabhan C, Feusner JH, Rowe JM: Acute promyelo-
cytic leukemia: evolving therapeutic strategies. Blood 2002,
99:759-767.
94. Schipper H, Turley EA, Braun M: A new biological framework for
cancer research. Lancet 1996, 348:1149-1151.
95. Leclerc E, Corlu A, Griscom L, Baudoin R, Legallais C: Guidance of
liver and kidney organotypic cultures inside rectangular sili-
cone microchannels. Biomaterials 2006, 27:4109-4119.
96. Bustuoabad OD, Correa JE: Injury and death in the embryo
development process: hypothesis of biological self-organiza-
tion. Frontier Perspectives 2004, 13:34-38.

Publish with Bio Med Central and every

scientist can read your work free of charge
"BioMed Central will be the most significant development for
disseminating the results of biomedical researc h in our lifetime."
Sir Paul Nurse, Cancer Research UK

Your research papers will be:

available free of charge to the entire biomedical community
peer reviewed and published immediately upon acceptance
cited in PubMed and archived on PubMed Central
yours — you keep the copyright

Submit your manuscript here: BioMedcentral

http://www.biomedcentral.com/info/publishing_adv.asp

Page 14 of 14
(page number not for citation purposes)