Over the last several decades, improved awareness of the prevalence of carcinogens in the environment, along with a growing appreciation of the complexity of the carcinogenesis process, has shifted policy on cancer risk from one of strict avoidance of carcinogens to one of adherence to exposure limits deemed “safe” based on quantitative risk estimation. Meanwhile, given the mutagenic nature of most carcinogens, attention has gravitated to developing a genetic rationale for measuring and comparing risks. This focus has culminated in the now well-established multistage mutational paradigm, which holds that a stepwise sequence of mutations drives cell “initiation” and the subsequent “transformation” of an initiated cell into a cancer cell, and that, once created, a cancer cell will inevitably undergo “progression” to become overt disease. Unanticipated by this paradigm is the effect progression-phase population- and tissue-level bottleneck events may have on this process. Attesting to this is the prevalence of tumor dormancy, a state of arrested growth of an otherwise fully malignant, often microscopic cancer mass, maintained by interactions among cancer cells and between cancer and host cells. The proper inclusion of such progression-modifying influences would clearly behoove risk estimation and improve our understanding of the natural history of cancer by accounting for the less-than-certain risk of eventual cancer disease even when cancer cells are present. Such an improved understanding, in turn, stands to better inform policy-making and influence such clinical practice decisions as whether to treat the increasingly smaller tumors detectable with advancing technologies. Cancer Res; 74(3); 659–64. ©2013 AACR.

The issue of cancer risk from exposure to environmental carcinogens impacts millions of individuals worldwide. Policy considerations include the setting of appropriate standards for exposure in the workplace, home, and clinic, along with assessing tradeoffs among alternative energy sources and classifying carcinogens to which the public may be exposed. Of special interest to the U.S. Department of Energy (DOE) is the cancer risk associated with the post–Cold War cleanup of nuclear waste; more generally, the longstanding question of whether there exist radiation dose or dose-rate thresholds below which there is no increase, or perhaps even a decrease, in cancer risk (1, 2). The National Aeronautics and Space Administration (NASA) seeks to estimate the excess radiogenic cancer risk from long-duration space missions (3, 4). In response to workplace concerns, the National Institute for Occupational Safety and Health upgraded an earlier policy that called for “no detectable exposure levels for proven carcinogenic substances” (5). It now recommends chemical carcinogen exposure limits go beyond “no effects” to include exposure levels at which there may be residual risks. RISC-RAD, an 11-country project funded by the European Commission, is dedicated to understanding carcinogenesis risk “at low doses and dose rates and the effects of radiation quality at exposures below those at which direct information is available.” With the now widespread implementation of airport x-ray backscatter screening procedures and the high-profile Fukushima nuclear power plant disaster, considerable public attention has been drawn to these issues.

In parallel with these developments, efforts have been underway to fully understand the events leading to cancer cell generation. Carcinogenesis, the sequence of events that convert a normal cell into an expanding, invasive neoplastic mass, is commonly modeled as initially involving a series of cell mutations followed by clonal expansions (6, 7). These so-called “natural history” models of carcinogenesis take on various forms, but are inspired by the basic two-stage mutational paradigm. By this paradigm, a chance DNA-altering event (e.g., a point mutation, deletion, translocation, copy number change, or heritable epigenetic alteration) “initiates” a cell, that is, alters its growth characteristics and predisposes it for later carcinogenic conversion. In a phase called “promotion,” the progeny of this initiated cell go on to display modified growth characteristics and altered propensities to undergo further spontaneous or induced genetic change. This cycle of DNA alteration and clonal expansion is revisited a number of times. Ultimately, a cell in the clone undergoes a “transformation” event, becoming a frank cancer cell, which expands within the host tissues to become overt disease (Fig. 1). As carcinogenesis proceeds, what begins as a cell-level series of events graduates to include population-level effects shortly after initiation, then comes to include tissue-level influences as the nascent tumor grows and acquires larger-scale properties, such as invasiveness and metastatic potential, upon interaction with the host.

Figure 1.

Carcinogenesis timeline. Shown is the standard natural history timeline for cancer development. By this timeline, a chance DNA-altering event (e.g., a point mutation, deletion, translocation, copy number change, or heritable epigenetic alteration) “initiates” a cell, that is, alters its growth characteristics and predisposes it for later carcinogenic conversion. In a phase called “promotion,” the progeny of this initiated cell go on to display modified growth characteristics and altered propensities to undergo further spontaneous or induced genetic change. This cycle of DNA alteration and clonal expansion is revisited a number of times. Ultimately, a single cell in the clone undergoes a “transformation” event, becoming a frank cancer cell, which expands within the host tissues to become overt disease. As shown, what is classically understood to begin as a cell-level series of events graduates to include population-level effects shortly after initiation, then comes to include tissue-level influences as the nascent tumor grows and acquires larger-scale properties, e.g., invasiveness and metastatic potential, through interaction with the host. The periods of applicability are shown as solid red, green, and blue lines, respectively. It is becoming increasingly clear, however, that each of these scales of influence may extend across the carcinogenesis spectrum, as indicated by the dashed extensions of these lines.

Figure 1.

Carcinogenesis timeline. Shown is the standard natural history timeline for cancer development. By this timeline, a chance DNA-altering event (e.g., a point mutation, deletion, translocation, copy number change, or heritable epigenetic alteration) “initiates” a cell, that is, alters its growth characteristics and predisposes it for later carcinogenic conversion. In a phase called “promotion,” the progeny of this initiated cell go on to display modified growth characteristics and altered propensities to undergo further spontaneous or induced genetic change. This cycle of DNA alteration and clonal expansion is revisited a number of times. Ultimately, a single cell in the clone undergoes a “transformation” event, becoming a frank cancer cell, which expands within the host tissues to become overt disease. As shown, what is classically understood to begin as a cell-level series of events graduates to include population-level effects shortly after initiation, then comes to include tissue-level influences as the nascent tumor grows and acquires larger-scale properties, e.g., invasiveness and metastatic potential, through interaction with the host. The periods of applicability are shown as solid red, green, and blue lines, respectively. It is becoming increasingly clear, however, that each of these scales of influence may extend across the carcinogenesis spectrum, as indicated by the dashed extensions of these lines.

Close modal

Work in colon carcinogenesis has refined our understanding of the malignant transformation event. It is known, for instance, that sequential mutations of key genes, adenomatous polyposis coli (APC), Ras/Raf, and p53, can transform a colonic crypt cell (8). After a cancer cell is created, the progeny population evolves over a phase termed progression, during which a number of the characteristic features of cancer, for example, tumor angiogenesis, resistance to cell death, and insensitivity to growth-inhibitory signaling, are acquired or enhanced (9). But the rapid accumulation of these and other molecular insights into carcinogenesis gained over the last 50 years would only gradually win favor for mechanistic risk modeling to become the guide for public policy on carcinogen exposure. With the renaissance in attitude that took place in the late 1970s, a “precautionary principle” that arbitrarily sought exposure limits well below the No Observed Adverse Effect Level, finally gave way to a “quantitative risk assessment” objective that made experimentation and elucidation of molecular mechanism the gold standard (10).

Even though the embrace of molecular mechanism for risk-based policymaking is now the mainstay, unexpected challenges to the carcinogenesis paradigm itself have begun to surface. As a case in point, research into cancer risk from radiation exposures in the low-dose realm is tackling the perplexing question of whether the risk at very low doses remains dose proportional (11), or instead deviates higher (12) or lower (13), perhaps to the point where the risk is actually reduced (14). However this plays out, these investigations have served to bring to the risk modeling forefront a growing sentiment that the multistage mutational paradigm, with the rather simplistic monotonic association it makes between cellular DNA damage and risk, neglects contributions from other biological scales of influence.

Complicating matters further, mechanistic insights into carcinogenesis are frequently derived indirectly from the study of cellular and molecular “surrogates” expected to correlate with cancer risk, for example, the extent of DNA damage and repair, oxidative stress, and genomic instability (15–17). Likewise, at the other end of the spectrum, conclusions as to how cells advance through the steps leading to cancer presentation are inferred from epidemiologic data on cancer incidence (18) in connection with cell culture and animal natural history models of tumor development (19, 20). Amidst the various linkages thus made, one presumption that has been consistently applied from the multistage mutational paradigm is that a cancer cell, once created, will eventually manifest as a cancer case (6, 21–23). By this standard, once a cell transforms to become a cancer cell, its progression to manifest disease is inevitable and the risk of eventual cancer incidence certain (left curve, Fig. 2).

Figure 2.

Risk implications comparing the standard carcinogenesis-risk paradigm to that accounting for progression-level effects. The cancer risk of relevance for policy-making and quantitative model building is that of disease presentation. By the time cancer presents, a variety of expansion-limiting progression-phase bottlenecks have been successfully traversed by the cancer cell population, any one of which could have blocked the carcinogenesis process. Beyond the oncogenic mutations that must be acquired for a cell to transform, population-level acquisitions, including angiogenic and invasive capacity, immune escape, and stromal permissiveness, must be in place for transformed cells to advance to become symptomatic disease (waving curve, extending left to right). Interruption of tumor progression through any one of these bottlenecks could be decisive in whether overt disease ever arises. Notwithstanding, standard risk modeling approaches typically assume cancer cells, once created, progress inevitably to disease presentation, and cancer incidence rates are therefore a direct reflection of the numbers and natures of the genetic alterations required to produce a cancer cell. This depiction of the carcinogenesis process has the effect of assigning the entire risk of disease presentation to that of transformation of an initiated cell to a cancer cell (leftmost curve, steeply rising), an event that autopsy evidence refutes, suggesting transformation is but a first step toward manifest disease (portion of waving curve that lies below the leftmost curve). An alternative risk model that includes posttransformation determinants of carcinogenesis from the outset would provide not only more realistic risk estimates, but also an improved understanding of both the pretumor-cell and posttumor-cell contributions to the overall carcinogenesis process.

Figure 2.

Risk implications comparing the standard carcinogenesis-risk paradigm to that accounting for progression-level effects. The cancer risk of relevance for policy-making and quantitative model building is that of disease presentation. By the time cancer presents, a variety of expansion-limiting progression-phase bottlenecks have been successfully traversed by the cancer cell population, any one of which could have blocked the carcinogenesis process. Beyond the oncogenic mutations that must be acquired for a cell to transform, population-level acquisitions, including angiogenic and invasive capacity, immune escape, and stromal permissiveness, must be in place for transformed cells to advance to become symptomatic disease (waving curve, extending left to right). Interruption of tumor progression through any one of these bottlenecks could be decisive in whether overt disease ever arises. Notwithstanding, standard risk modeling approaches typically assume cancer cells, once created, progress inevitably to disease presentation, and cancer incidence rates are therefore a direct reflection of the numbers and natures of the genetic alterations required to produce a cancer cell. This depiction of the carcinogenesis process has the effect of assigning the entire risk of disease presentation to that of transformation of an initiated cell to a cancer cell (leftmost curve, steeply rising), an event that autopsy evidence refutes, suggesting transformation is but a first step toward manifest disease (portion of waving curve that lies below the leftmost curve). An alternative risk model that includes posttransformation determinants of carcinogenesis from the outset would provide not only more realistic risk estimates, but also an improved understanding of both the pretumor-cell and posttumor-cell contributions to the overall carcinogenesis process.

Close modal

Countering this notion, a vast number of direct observations at the level of tumor progression are showing interactions of cancer cells with themselves and with cells of the host, may not only promote, but also slow, stop, or even reverse a carcinogenesis course (24–30). One is confronted with the realization that cancer risk determination from exposures to carcinogenic agents not only requires a multiscale effort to link the various levels at which data are available (from molecular to cellular to tissue to epidemiologic), but that the risk itself is influenced at multiple scales, having substantial population- and tissue-level components (Figs. 1 and 2, right curve). Well established are the supportive and/or suppressive roles of tumor-associated elements, such as stroma, vasculature, and the immune response, in modulating cancer progression (31–33). Even after cancer cells arise, there are a number of population-level processes that need to be engaged to permit advancement through progression, including immune activity (34–37), invasive activity (38, 39), tumor angiogenesis (30, 40), and facilitation of stem cell proliferation (41, 42), including the need for recruited helper cells for stem cell expansion (43).

The failure of the cancer cell aggregate to mobilize host cooperation, or to evade host suppression, may cause the cancer to halt its progression, or even regress (44). In this regard, those key host processes that require modification to permit cancer progression amount to “bottlenecks,” the surmounting of which is essential for cancer incidence. The failure of cancer cells to recruit blood vessels and to overcome host immune surveillance constitute two common progression-level bottlenecks, either of which could hold a tumor in a dormant state for an extended period (45–47). Recently, we have shown in murine studies that even host age itself (48) can modulate tumor progression, with the greatest tumor growth curtailment coming from the oldest hosts. Yet, some of the most compelling data on progression-phase limitations to tumor growth come from human data from autopsies of individuals dying of noncancer causes. Once thought anomalous, “dormant” cancers were shown to be more the rule than the exception, with the discovery that most “cancer-free” individuals harbor small, indolent tumor lesions that will not progress to manifest disease (25–27, 49). These and other findings have raised the concern that as our ability to screen for cancer improves, so will our tendency to treat lesions that pose no threat (49, 50). Reinforcing this point, although lethal lung cancer is 10 times more common in smokers, spiral computed tomography screening is revealing nonsmokers have about the same number of lung lesions as smokers (51, 52). Suggested is a strong role for disturbance of the local tumor–host interaction in the advancement of dormant lung lesions toward frank disease.

The reach of host-derived progression modifiers may even extend to the provocation of transformation itself. A seminal study by Bhowmick and colleagues (28) confirmed the ability of host fibroblasts, in which the TGF-β type II receptor was inactivated, to elicit the transformation of adjacent normal epithelia. Carcinoma-associated fibroblasts have also been observed to convert genetically abnormal (initiated) but nontumorigenic epithelia to carcinomas (29). Together, these observations seem to support an augmented version of the mutational paradigm; one that includes disruptions to homeostatic tissue-level signaling as potential initiators of early cell-level carcinogenesis, with DNA lesions being at times a consequential action.

Lending credence to this notion, the immune response represents a tissue-level interaction that has been shown to influence both carcinogenesis progression and initiation. This can be accomplished by the secretion of certain cytokines and growth factors, which encourage proliferation of epithelia while at the same time generating DNA-damaging reactive oxygen species as part of their general infection-fighting function, further potentiating tumor development as per the classic mutational paradigm (53, 54).

Explaining the known increased cancer incidence with aging, one may be seeing a gradual compromise of natural immune homeostatic surveillance that rids the body of damaged cells potentiated to become cancerous (55). At the same time, there is evidence for the immune system playing a promoting role once cancers already exist. Mechanisms include dendritic cell depletion (56), recruitment of metastasis-promoting mesenchymal stem cells to the tumor site (57), induced accumulation of immunosuppressive myeloid-derived suppressor cells (58), the directed maturation of T cells from an antitumor Th1 phenotype toward a tumor-associated Th2 phenotype (59, 60), or generally through a process called immunoediting (61, 62). Reductions of these tumor-promoting aspects of immunity in older individuals may contribute to slow the progression of existing cancers. Taken together, these results help to explain how immune compromise with age can facilitate cancer incidence while inhibiting tumor progression.

The dichotomy of early suppression and late promotion exhibited by the immune system is also played out by TGF-β and cancer stem cell kinetics. Before a tumor arises, TGF-β inhibits the initial growth of epithelial populations (63) and prevents tumor formation (64), but after a tumor is fully established, higher levels of TGF-β in the blood (65) and the tumor microenvironment (66) are tumor-promoting, as is the comparatively higher host expression of TGF-β1 seen in younger hosts (A. Beheshti and colleagues, unpublished data). Analogously, increasing the death rate of nonstem cancer cells reduces tumor size initially through simple cell count reduction. The longer-term effect, though, is to free cancer stem cells from the contact-inhibiting influence of the nonstem cells (41), allowing them to cycle and produce both stem and nonstem progeny. The ultimate result of nonstem cell killing may thus be a paradoxical heightening of the overall tumor growth rate.

That mediation of tumor progression may even have a systemic basis is supported by the finding that methionine metabolism, done primarily by the liver, correlates strongly with prostate cancer recurrence after radical prostatectomy (67). Although a direct connection is not yet clear, that a fundamental metabolic pathway should be so intertwined with cancer progression offers highly suggestive evidence that cancer progression is a system-level process having system-level controls.

The emerging realization is that cancer development entails not just negotiating processes en route to producing a cancer cell population, but involves larger-scale population-, tissue-, and organism-scale influences that accelerate or suppress progression of the nascent cancer population. This realization has major ramifications for carcinogenesis risk estimation and the development of countermeasures. In particular, it contradicts the common presumption in risk analysis of a direct coupling between the molecular steps in cellular oncogenesis leading up to transformation, and the final epidemiologic endpoint of cancer incidence (left curve, Fig. 2). The critical roles of posttransformational cell–cell and cell–host interactions as “tissue-scale checkpoints”—potential barriers to tumor advancement—have begun to be appreciated in principle, but remain to be incorporated in quantitative modeling constructs as pivotal risk modifiers (right curve, Fig. 2). Of specific concern here is that, because the effects of any progression-level bottlenecks on tumor advancement are already folded into the epidemiological data, risk models that ignore these inputs by assuming cancer incidence is inevitable once cancer cells are created, cannot correctly back-infer the cancer's “natural history” (i.e., the mutational, population-level, and host-interactive steps leading up to cancer incidence). In such analytic models, any reduction in tumor incidence because of such bottlenecks to tumor advancement will necessarily be taken into account through an inappropriate inflation of the estimated number of pretransformation mutational events necessary to create the first cancer cell.

To avoid this disparity, it is clear there is a need to better interface biologically based mechanistic carcinogenesis models with both natural history models and their epidemiologic cancer incidence endpoints. Meanwhile, research into the role of posttransformational mechanisms in cancer development is proceeding at a rapid pace. With it, a broadened picture of tumor progression is emerging that awaits integration into existing carcinogenesis models to provide improved risk assessments. In particular, the recognition that dormant microscopic tumors are highly prevalent in the population is poised to add a clear and measurable progression-level component to the multistage mutational theory of carcinogenesis. Knowledge that intertissue interactions among cancer, stromal (68), immune (59), and endothelial (30, 69) cells can radically alter the development of existing cancer cell populations is revising our understanding of cancer staging. It is also extending our understanding of cancer cell genetics itself by showing how accommodation of microenvironmental cues may exert overriding control over the expression of a preexisting aggressive potential (70). This contrasts with tumor cell–centric thinking on this point, which has posited such potential is only acquired gradually through mutation over the course of cancer development (70). Rounding out this picture, the physical and molecular determinants of altered stromal signaling (28), a permissive angiogenic state (30), environmentally modified cancer stem cell kinetics (41, 42), the role of recruited cells in aiding cancer stem cells to drive tumor development (43), and aspects of the immune response that facilitate tumor progression (54, 55) are being elucidated.

Looking ahead, as the traditional considerations of intracellular, molecular-scale hallmarks of carcinogenesis are joined through modeling with consideration of the equally influential roles played by multiple tissue-scale bottlenecks—endpoints that tend also to be macroscopic and thus potentially easier to track—the relative roles of these intracellular molecular events will be more properly contextualized. Risk projections stemming from such multiscale modeling efforts would not only better inform policy, but help in the identification of additional therapeutic control points in cancer development.

No potential conflicts of interest were disclosed.

Conception and design: L. Hlatky, P. Hahnfeldt

Development of methodology: L. Hlatky, P. Hahnfeldt

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): L. Hlatky

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): L. Hlatky, P. Hahnfeldt

Writing, review, and/or revision of the manuscript: Lynn Hlatky, Philip Hahnfeldt

Study supervision: L. Hlatky

The authors thank C. Lamont for graphics.

This work was supported by DOE grant DE-SC002606 (L. Hlatky), DOE grant DE-SC001434 (P. Hahnfeldt), NASA grant NNX11AK26G (L. Hlatky) and NCI ICBP Grant U54CA149233 (L. Hlatky).

1.
NRC
. 
Health risks from exposure to low levels of ionizing radiation: BEIR VII phase 2
.
Washington, DC
:
National Academy of Sciences Report
.
Available from
: http://www.nap.edu.
2.
Feinendegen
LE
. 
Evidence for beneficial low level radiation effects and radiation hormesis
.
Br J Radiol
2005
;
78
:
3
7
.
3.
Durante
M
,
Cucinotta
FA
. 
Heavy ion carcinogenesis and human space exploration
.
Nat Rev Cancer
2008
;
8
:
465
72
.
4.
Cucinotta
FA
,
Chappell
LJ
,
Kim
MH
,
Wang
M
. 
Radiation carcinogenesis risk assessments for never-smokers
.
Health Phys
2012
;
103
:
643
51
.
5.
Coffey
C
,
Martin
S
,
Sahakian
N
. 
Health hazard evaluation report: HETA-2002-0343-2902, Fayette County Courthouse, Uniontown, PA, Appendix 1, p. 34
.
Washington, DC
:
National Institute for Occupational Safety and Health, Centers for Disease Control
; 
2003
.
Available from: (accessed Aug 1, 2013)
http://www.cdc.gov/niosh/hhe/reports/pdfs/2002-0343-2902.pdf.
6.
Moolgavkar
SH
. 
The multistage theory of carcinogenesis and the age distribution of cancer in man
.
J Natl Cancer Inst
1978
;
61
:
49
52
.
7.
Hahn
WC
,
Weinberg
RA
. 
Modelling the molecular circuitry of cancer
.
Nat Rev Cancer
2002
;
2
:
331
41
.
8.
Polyak
K
,
Hamilton
SR
,
Vogelstein
B
,
Kinzler
KW
. 
Early alteration of cell-cycle-regulated gene expression in colorectal neoplasia
.
Am J Pathol
1996
;
149
:
381
7
.
9.
Hanahan
D
,
Weinberg
RA
. 
The hallmarks of cancer
.
Cell
2000
;
100
:
57
70
.
10.
Lave
LB
. 
Hormesis: policy implications
.
J Appl Toxicol
2000
;
20
:
141
5
.
11.
BEIR VII: health risks from exposure to low levels of ionizing radiation
.
Washington, DC
:
National Academies Press
; 
2005
.
12.
Hall
EJ
. 
Henry S. Kaplan Distinguished Scientist Award 2003. The crooked shall be made straight; dose-response relationships for carcinogenesis
.
Int J Radiat Biol
2004
;
80
:
327
37
.
13.
Tubiana
M
. 
Dose-effect relationship and estimation of the carcinogenic effects of low doses of ionizing radiation: the joint report of the Acadamie des Sciences (Paris) and of the Acadamie Nationale de Medecine
.
Int J Radiat Oncol Biol Phys
2005
;
63
:
317
9
.
14.
Scott
BR
. 
Low-dose radiation risk extrapolation fallacy associated with the linear-no-threshold model
.
Hum Exp Toxicol
2008
;
27
:
163
8
.
15.
Kutanzi
K
,
Kovalchuk
O
. 
Exposure to estrogen and ionizing radiation causes epigenetic dysregulation, activation of mitogen-activated protein kinase pathways and genome instability in the mammary gland of ACI rats
.
Cancer Biol Ther
2013
;
14
:
564
73
.
16.
Rithidech
KN
,
Honikel
LM
,
Reungpatthanaphong
P
,
Tungjai
M
,
Golightly
M
,
Whorton
EB
. 
Effects of 100MeV protons delivered at 0.5 or 1cGy/min on the in vivo induction of early and delayed chromosomal damage
.
Mutat Res
2013
;
756
:
127
40
.
17.
Georgakilas
AG
,
O'Neill
P
,
Stewart
RD
. 
Induction and repair of clustered DNA lesions: what do we know so far?
Radiat Res
2013
;
180
:
100
9
.
18.
Whyte
S
,
Walsh
C
,
Chilcott
J
. 
Bayesian calibration of a natural history model with application to a population model for colorectal cancer
.
Med Decis Making
2011
;
31
:
625
41
.
19.
Barrett
JC
. 
Cell culture models of multistep carcinogenesis
.
IARC Sci Publ
1985
;
58
:
181
202
.
20.
Dekkers
F
,
Bijwaard
H
,
Bouffler
S
,
Ellender
M
,
Huiskamp
R
,
Kowalczuk
C
, et al
A two-mutation model of radiation-induced acute myeloid leukemia using historical mouse data
.
Radiat Environ Biophys
2011
;
50
:
37
45
.
21.
Little
MP
,
Hawkins
MM
,
Charles
MW
,
Hildreth
NG
. 
Fitting the Armitage-Doll model to radiation-exposed cohorts and implications for population cancer risks
.
Radiat Res
1992
;
132
:
207
21
.
22.
Chen
CW
. 
Armitage-Doll two-stage model: implications and extension
.
Risk Anal
1993
;
13
:
273
9
.
23.
Pierce
DA
,
Mendelsohn
ML
. 
A model for radiation-related cancer suggested by atomic bomb survivor data
.
Radiat Res
1999
;
152
:
642
54
.
24.
Black
WC
,
Welch
HG
. 
Advances in diagnostic imaging and overestimations of disease prevalence and the benefits of therapy
.
New Engl J Med
1993
;
328
:
1237
43
.
25.
Cunha
GR
,
Hayward
SW
,
Wang
YZ
,
Ricke
WA
. 
Role of the stromal microenvironment in carcinogenesis of the prostate
.
Int J Cancer
2003
;
107
:
1
10
.
26.
Zahl
P-H
,
Maehlen
J
,
Welch
HG
. 
The natural history of invasive breast cancers detected by screening mammography
.
Arch Intern Med
2008
;
168
:
2311
6
.
27.
Folkman
J
,
Kalluri
R
. 
Cancer without disease
.
Nature
2004
;
427
:
787
.
28.
Bhowmick
NA
,
Chytil
A
,
Plieth
D
,
Gorska
AE
,
Dumont
N
,
Shappell
S
, et al
TGF-β signaling in fibroblasts modulates the oncogenic potential of adjacent epithelia
.
Science
2004
;
303
:
848
51
.
29.
Hayward
SW
,
Wang
Y
,
Cao
M
,
Hom
YK
,
Zhang
B
,
Grossfeld
GD
, et al
Malignant transformation in a nontumorigenic human prostatic epithelial cell line
.
Cancer Res
2001
;
61
:
8135
42
.
30.
Hahnfeldt
P
,
Panigrahy
D
,
Folkman
J
,
Hlatky
L
. 
Tumor development under angiogenic signaling: a dynamical theory of tumor growth, treatment response, and postvascular dormancy
.
Cancer Res
1999
;
59
:
4770
5
.
31.
De Wever
O
,
Mareel
M
. 
Role of tissue stroma in cancer cell invasion
.
J Pathol
2003
;
200
:
429
47
.
32.
Bhowmick
NA
,
Moses
HL
. 
Tumor-stroma interactions
.
Curr Opin Genet Dev
2005
;
15
:
97
101
.
33.
Chouaib
S
,
Kieda
C
,
Benlalam
H
,
Noman
MZ
,
Mami-Chouaib
F
,
Rüegg
C
. 
Endothelial cells as key determinants of the tumor microenvironment: interaction with tumor cells, extracellular matrix and immune killer cells
.
Crit Rev Immunol
2010
;
30
:
529
45
.
34.
Prehn
RT
. 
The immune reaction as a stimulator of tumor growth
.
Science
1972
;
276
:
170
1
.
35.
Prendergast
GC
. 
Immune escape as a fundamental trait of cancer: focus on IDO
.
Oncogene
2008
;
27
:
3889
900
.
36.
Wilkie
KP
. 
A review of mathematical models of cancer-immune interactions in the context of tumor dormancy
.
Adv Exp Med Biol
2013
;
734
:
201
34
.
37.
d'Onofrio
A
. 
Multifaceted kinetics of immuno-evasion from tumor dormancy
.
Adv Exp Med Biol
2013
;
734
:
111
43
.
38.
Barcellos-Hoff
MH
,
Akhurst
RJ
. 
Transforming growth factor-beta in breast cancer: too much, too late
.
Breast Cancer Res
2009
;
11
:
202
.
39.
Gavert
N
,
Ben-Ze'ev
A
. 
Epithelial-mesenchymal transition and the invasive potential of tumors
.
Trends Mol Med
2008
;
14
:
199
209
.
40.
Folkman
J
. 
Tumor angiogenesis: a possible control point in tumor growth
.
Ann Intern Med
1975
;
82
:
96
100
.
41.
Enderling
H
,
Anderson
AR
,
Chaplain
MA
,
Beheshti
A
,
Hlatky
L
,
Hahnfeldt
P
. 
Paradoxical dependencies of tumor dormancy and progression on basic cell kinetics
.
Cancer Res
2009
;
69
:
8814
21
.
42.
Tysnes
BB
,
Bjerkvig
R
. 
Cancer initiation and progression: involvement of stem cells and the microenvironment
.
Biochim Biophys Acta
2007
;
1775
:
283
9
.
43.
Hanahan
D
,
Coussens
LM
. 
Accessories to the crime: functions of cells recruited to the tumor microenvironment
.
Cancer Cell
2012
;
21
:
309
22
.
44.
Serrano
M
. 
Cancer regression by senescence
.
N Engl J Med
2007
;
356
:
1996
7
.
45.
Ghajar
CM
,
Peinado
H
,
Mori
H
,
Matei
IR
,
Evason
KJ
,
Brazier
H
, et al
The perivascular niche regulates breast tumour dormancy
.
Nat Cell Biol
2013
;
15
:
807
17
.
46.
Hahnfeldt
P
. 
The host support niche as a control point for tumor dormancy: implications for tumor development and beyond
.
Adv Exp Med Biol
2013
;
734
:
19
35
.
47.
Wilkie
KP
,
Hahnfeldt
P
. 
Tumor-immune dynamics regulated in the microenvironment inform the transient nature of immune-induced tumor dormancy
.
Cancer Res
2013
;
73
:
3534
44
.
48.
Beheshti
A
,
Sachs
RK
,
Peluso
M
,
Rietman
E
,
Hahnfeldt
P
,
Hlatky
L
. 
Age and space irradiation modulate tumor progression: implications for carcinogenesis risk
.
Radiat Res
2013
;
179
:
208
20
.
49.
Welch
HG
,
Black
WC
. 
Overdiagnosis in cancer
.
J Natl Cancer Inst
2010
;
102
:
605
13
.
50.
Esserman
LJ
,
Thompson
IM
,
Reid
B
. 
Overdiagnosis and overtreatment in cancer: an opportunity for improvement
.
JAMA
2013
;
310
:
797
8
.
51.
Sone
S
,
Takashima
S
,
Li
F
,
Yang
Z
,
Honda
T
,
Maruyama
Y
, et al
Mass screening for lung cancer with mobile spiral computed tomography scanner
.
Lancet
1998
;
351
:
1242
5
.
52.
Li
F
,
Sone
S
,
Abe
H
,
MacMahon
H
,
Doi
K
. 
Low-dose computed tomography screening for lung cancer in a general population: characteristics of cancer in non-smokers versus smokers
.
Acad Radiol
2003
;
10
:
1013
20
.
53.
Maeda
H
,
Akaike
T
. 
Nitric oxide and oxygen radicals in infection, inflammation, and cancer
.
Biochemistry
1998
;
63
:
854
65
.
54.
Coussens
LM
,
Werb
Z
. 
Inflammation and cancer
.
Nature
2002
;
420
:
860
7
.
55.
Prendergast
GC
,
Metz
R
. 
A perspective on new immune adjuvant principles: reprogramming inflammatory states to permit clearance of cancer cells and other age-associated cellular pathologies
.
Oncoimmunology
2012
;
1
:
924
9
.
56.
Scarlett
UK
,
Rutkowski
MR
,
Rauwerdink
AM
,
Fields
J
,
Escovar-Fadul
X
,
Baird
J
, et al
Ovarian cancer progression is controlled by phenotypic changes in dendritic cells
.
J Exp Med
2012
;
209
:
495
506
.
57.
Jing
Y
,
Han
Z
,
Liu
Y
,
Sun
K
,
Zhang
S
,
Jiang
G
, et al
Mesenchymal stem cells in inflammation microenvironment accelerates hepatocellular carcinoma metastasis by inducing epithelial-mesenchymal transition
.
PLoS ONE
2012
;
7
:
e43272
.
58.
Gabrilovich
DI
,
Nagaraj
S
. 
Myeloid-derived suppressor cells as regulators of the immune system
.
Nat Rev Immunol
2009
;
9
:
162
74
.
59.
de Visser
KE
,
Eichten
A
,
Coussens
LM
. 
Paradoxical roles of the immune system during cancer development
.
Nat Rev Cancer
2006
;
6
:
24
37
.
60.
Kareva
I
,
Hahnfeldt
P
. 
The emerging “hallmarks” of metabolic reprogramming and immune evasion: distinct or linked?
Cancer Res
2013
;
73
:
2737
42
.
61.
Schreiber
RD
,
Old
LJ
,
Smyth
MJ
. 
Cancer immunoediting: integrating immunity's roles in cancer suppression and promotion
.
Science
2011
;
331
:
1565
70
.
62.
Jiang
X
,
Shapiro
DJ
. 
The immune system and inflammation in breast cancer
.
Mol Cell Endocrinol
2013 Jun 19
. [Epub ahead of print]
63.
Tucker
RF
,
Shipley
GD
,
Moses
HL
,
Holley
RW
. 
Growth inhibitor from BSC-1 cells closely related to platelet type beta transforming growth factor
.
Science
1984
;
226
:
705
7
.
64.
Pierce
DF
 Jr
,
Gorska
AE
,
Chytil
A
,
Meise
KS
,
Page
DL
,
Coffey
RJ
 Jr
, et al
Mammary tumor suppression by transforming growth factor beta 1 transgene expression
.
Proc Natl Acad Sci U S A
1995
;
92
:
4254
8
.
65.
Tsushima
H
,
Ito
N
,
Tamura
S
,
Matsuda
Y
,
Inada
M
,
Yabuuchi
I
, et al
Circulating transforming growth factor β1 as a predictor of liver metastasis after resection in colorectal cancer
.
Clin Cancer Res
2001
;
7
:
1258
62
.
66.
Calon
A
,
Espinet
E
,
Palomo-Ponce
S
,
Tauriello
DV
,
Iglesias
M
,
Céspedes
MV
, et al
Dependency of colorectal cancer on a TGF-β-driven program in stromal cells for metastasis initiation
.
Cancer Cell
2012
;
22
:
571
84
.
67.
Stabler
S
,
Koyama
T
,
Zhao
Z
,
Martinez-Ferrer
M
,
Allen
RH
,
Luka
Z
, et al
Serum methionine metabolites are risk factors for metastatic prostate cancer progression
.
PLoS ONE
2011
;
6
:
e22486
.
68.
Radisky
DC
,
Bissell
MJ
. 
Cancer. Respect thy neighbor!
Science
2004
;
303
:
775
7
.
69.
Folkman
J
,
Hahnfeldt
P
,
Hlatky
L
. 
Cancer: looking outside the genome
.
Nat Rev Molec Cell Biol
2000
;
1
:
76
9
.
70.
Schedin
P
,
Elias
A
. 
Multistep tumorigenesis and the microenvironment
.
Breast Cancer Res
2004
;
6
:
93
101
.