REVIEW

doi:10.1038/nature17038

Metastatic colonization by
circulating tumour cells
Joan Massagu1 & Anna C. Obenauf1

Metastasis is the main cause of death in people with cancer. To colonize distant organs, circulating tumour cells must
overcome many obstacles through mechanisms that we are only now starting to understand. These include infiltrating
distant tissue, evading immune defences, adapting to supportive niches, surviving as latent tumour-initiating seeds
and eventually breaking out to replace the host tissue. They make metastasis a highly inefficient process. However, once
metastases have been established, current treatments frequently fail to provide durable responses. An improved understanding of the mechanistic determinants of such colonization is needed to better prevent and treat metastatic cancer.

alignant tumours start early on the road to metastasis. Cancer

cells that are invasive and motile can enter the circulatory system long before a tumour is diagnosed. Most of these cells will
perish, but a small proportion manages to infiltrate distant organs and
survives as disseminated seeds for eventual relapse. Thus, at diagnosis, a
primary tumour might have already seeded distant organs with thousands
of cancer cells. These cells will face many obstacles before they can take
over their host organ and form clinically relevant lesions. Indeed, organ
colonization is the most complex and rate-limiting phase of metastasis.
Research on metastatic colonization has been hindered by the complexity of the biology and a lack of adequate experimental models. However,
the development of patient-derived and genetically engineered mouse
models of metastasis, enhanced imaging technologies and advanced
genomic sequencing, including the ability to analyse single cells, together
with improved access to clinically annotated tissue samples, has brought
fresh insight to our understanding of the molecular mechanisms that
allow circulating tumour cells (CTCs) to invade distant organs, settle in
supportive niches and eventually overtake their host tissue (Fig. 1). This
has allowed us to better conceptualize the metastatic process as a whole
and provides a basis for superior treatments. Such advances show that the
elucidation of metastatic colonization is a tractable problem with clinical
benefits. In this Review, we highlight current concepts and open questions
at the forefront of this field. Each aspect of metastatic colonization that
we explore is an individually rich area of research, and we cite specialized
articles that cover them in depth.

The inefficiency of metastatic colonization

Even small tumours can shed millions of cancer cells. However, many
people who have recovered from cancer never experience relapse or do so
only after a long period of latency without clinically manifest disease. The
number of CTCs cancer cells that are found in blood from patients
far exceeds the number of overt metastatic lesions that develop1. Cancer
cells that survive after infiltrating distant organs, known as disseminated
tumour cells (DTCs), can be present in the bone marrow of people with
cancer for years, yet only about half of these patients develop overt metastasis2. These clinical observations suggest that metastatic colonization is
a very inefficient process in which most cells die and only a minority of
those that survive form macrometastases.
Data from experimental mouse models are in line with the clinical
evidence. For example, intravenously injected cancer cells that reach the

lungs die in large numbers within 2 days3, as do arterially injected cancer

cells that lodge in the brain, liver or bone marrow4. And populations of
cells that are enriched for highly metastatic cells experience extensive
attrition after infiltration of distant organs5. The vast majority of melanoma cells that are injected in the portal vein fail to form micrometastases in the liver, and only 0.02% form macrometastases6,7. Similarly, most
cancer cells that infiltrate the brain will die810. Such inefficiencies cannot
be attributed simply to a scarcity of cancer stem cells with metastasisinitiating potential: the majority of breast cancer stem cells that reach the
lungs undergo apoptosis11 and colorectal cancer stem cells are cleared
quickly after infiltrating the liver parenchyma12. Observations such as
these, in mouse models and in the clinic, imply that the factors that influence the survival and tumour-initiating activity of DTCs are important
determinants of metastasis.

Organ infiltration

The early steps in the metastatic cascade, which include the invasion and
migration of cancer cells into tissues and the circulatory system, have
been extensively studied13,14. Cytoskeletal rearrangements within cancer
cells15, combined with adhesive interactions between cells and the secretion of extracellularmatrix-degrading metalloproteinases (MMPs) and
cathepsins16,17, drive their invasion and migration through the stroma, a
network of supportive, connective tissue cells. Cancer cells can migrate as
single cells boring a path through the extracellular matrix18, move along
collagen fibres19 or migrate collectively as ensembles that forge ahead
of the tumour invasion front20. In prostate cancer, invasion along nerve
fibres provides an additional route for dissemination21. In response to
transforming growth factor- (TGF-) and other cell-signalling proteins
that are released by stromal cells, carcinoma cells can undergo epithelialto-mesenchymal transition (EMT) a reversible phenotypic change in
which cells lose intercellular adhesion and epithelial polarization and
gain motility and invasiveness22. EMT has an important role in gastrulation and other morphogenic events that occur during development.
In carcinoma cells, EMT can promote cell entry into the vasculature,
known as intravasation, and support the induction of a stem-cell phenotype, whereas a reversal of this state after extravasation in which
cancer cells exit capillaries to enter organs can facilitate colonization23.
However, studies in models of breast and pancreatic cancer suggest that
EMT could be dispensable for the establishment of metastasis, despite
it contributing to the aggressiveness of cancer cells by increasing their

Cancer Biology and Genetics Program, Memorial Sloan Kettering Cancer Center, New York 10065, USA. Present address:Research Institute of Molecular Pathology, Vienna Biocenter,
1030 Vienna, Austria.
2 9 8 | N AT U R E | VO L 5 2 9 | 2 1 JA N UA RY 2 0 1 6

2016 Macmillan Publishers Limited. All rights reserved

REVIEW INSIGHT
Colonization

Pre-colonization (minutes to hours)

Local invasion

Primary tumour
Brain capillary wall
Bloodbrain barrier
Lung capillary wall
Tight endothelium

Circulating tumour cells

Intravasation

Platelet

Blood vessel

Liver and bone marrow capillary wall

Fenestrated endothelium

Circulation

Seeding and latency (days to years)

Arrest

Overt colonization
Overgrowth

Extravasation
Settlement

Latency and resistance

Cancer
therapy

Immune
resistance
Cell death

Supportive
niche

Micrometastasis
Single cell

Activation
and growth

Drug
resistance
Release of
survival signals

Figure 1 | Metastatic colonization. Metastasis proceeds through multiple

steps and restrictive bottlenecks. The pre-colonization phase of metastasis
comprises a series of events that occur on a timescale of minutes to hours.
Local invasion of the primary tumour by cancer cells is followed by their
intravasation into the tumour vasculature. The cancer cells then enter the
circulatory system as single cells or clusters that are coated with platelets.
Circulatory patterns, which move blood through the lungs and then on to other
organs, and the differing structure of the capillary walls in each organ influence
the dissemination of CTCs. On their arrest in capillaries at distant sites, the
cancer cells extravasate into the parenchyma of target organs to commence
colonization. Colonization can be parsed into many steps that occur on a
timescale of years. After extravasation, colonizing cancer cells must develop
resistance to immunity and other host-tissue defences to survive. Settlement in

supportive niches enables them to survive and retain their stem-like tumourinitiating capacity. The cancer cells then enter a latent state as single cells or
micrometastases. During latency, which can last from months to decades,
disseminated cells must achieve long-term survival. They might also acquire
traits that are required to overtake host tissue. When the cancer cells break
out of latency, they reinitiate overt outgrowth and overtake the local tissue
microenvironment. Therapeutic treatment can partially eliminate clinically
manifest metastases. However, under therapy-induced stress, cancer cells and
non-neoplastic stromal cells mobilize survival signals that nurture the residual
disease until minority drug-resistant clones emerge to lead the outgrowth of
a drug-resistant tumour. Different host-tissue microenvironments select for
cancer cells with distinct metastatic traits, which gives rise to organ-specific
populations of metastatic cells.

chemoresistance24,25. Therefore, the contribution of EMT to metastasis

might be more nuanced than thought previously.
Cancer cells can leave tumours as single cells or in clusters (Fig. 1).
There is growing evidence to indicate that distinct cancer-cell clones
show cooperative behaviour, which promotes their mutual survival and
metastatic ability2629. Polyclonal metastatic seeding, for example, has been
documented in patients with prostate cancer30, and in experimental models, polyclonal clusters of CTCs establish metastases more efficiently than
single cells31.
In the bloodstream, cancer cells are exposed to considerable shear
forces, the innate immune system and oxidative stress. To protect themselves during transit, they associate with platelets32 and undergo reversible metabolic changes that increase their ability to withstand oxidative
stress33. For instance, melanoma cells show an increased dependence on
nicotinamide adenine dinucleotide phosphate (NADPH)-generating
enzymes in the folate pathway, and inhibition of this pathway reduces
overt metastasis34.
CTCs become mechanically entrapped in capillaries before they exit the
blood and move into tissue. This is considered to be the main mechanism

for cancer-cell arrest. The first capillary bed that a CTC encounters is
determined by patterns of blood circulation in the body. In most organs,
the venous circulation leads to the right ventricle of the heart and on to the
lungs, although the venous circulation from the gut drains into the liver.
The resulting retention of CTCs in the lungs or liver, respectively, contributes to the high incidence of metastasis in these organs35. However, some
CTCs bypass these initial filters, perhaps through larger arteriovenous
shunts, to reach other organs through the arterial circulation.
CTCs that lodge in the microvasculature can initiate growth within the
lumen to form an embolus that eventually ruptures the vessel36 or extravasate by breaching vascular walls5,6. The composition of these walls differs
between organs and is another factor that influences where cancer cells
extravasate (Fig. 1). The capillaries in the liver and bone marrow, called
sinusoids, are lined with fenestrated endothelial cells and a discontinuous basal lamina37 gaps that might facilitate the extravasation of CTCs
and contribute to the high incidence of liver and bone metastasis13,38. In
contrast, the endothelium of lung capillaries has tight junctions and a
basement membrane, and the capillary walls of the brain are reinforced
by pericytes and the processes of astrocytes, which together constitute
2 1 JA N UA RY 2 0 1 6 | VO L 5 2 9 | N AT U R E | 2 9 9

2016 Macmillan Publishers Limited. All rights reserved

INSIGHT REVIEW
Systemic
signals
(LOX, VEGF, PlGF)

Mobilized
cancer cell

Blood vessel

Pre-metastatic
niche
Bone-marrow-derived cells
Neutrophils
Organ-resident cells
Extracellular matrix

Endothelium

Perivascular
niche
L1CAM

Pre-existing
microenvironment

TGF-

TNC, POSTN
WNT, PLOD2

Myeloid cells
Fibroblasts

CXCR4
Ad hoc
niche

CXCL12,
WNT, BMP

Native stem-cell
niche

Figure 2 | Metastatic niches. Cancer cells that infiltrate distant tissues

survive and retain their stem-cell potential by locating themselves in
supportive niches, which are akin to the niches that support normal adult
stem cells. A number of different niches have been proposed. Pre-metastatic
niches form before the arrival of cancer cells by systemic signals from the
primary tumour that recruit supportive stromal cells. Perivascular niches
support cancer cells that spread over the capillary basement membrane
after extravasation. The cancer cells remain close to cells of the endothelium
and their paracrine factors. Ad hoc niches are established by the secretory
products of cancer cells and act in an autocrine manner or recruit stromal
components as sources of supportive signals. Native stem-cell niches of the
host tissue are often invaded by the infiltrating cancer cells, which allows
the cells to occupy directly a supportive microenvironment. The location or
composition of these niches can overlap. For example, a native stem-cell niche
might be in a perivascular site, or pre-metastatic signals might combine with
those released by the cancer cell at an ad hoc niche.

the bloodbrain barrier13,37. Diverse genes have been identified in model

systems that mediate the extravasation of breast-cancer CTCs in the lungs
and are also associated with lung metastasis in the clinic. These include the
protein Fascin-1 and other components of invading protrusions known as
invadopodia39, autocrine enhancers of cancer-cell motility such as epiregulin and WNT ligands39,40, and mediators of endothelial disjunction and
vascular permeability that include angiopoietin-like 4 (ANGPTL4), vascular endothelial growth factor (VEGF), the cyclooxygenase COX2 (also
known as prostaglandin-endoperoxide synthase), MMP1 and osteonectin4144. Platelets that are associated with CTCs can stimulate extravasation by releasing TGF- and triggering EMT in the cancer cells32 or by
secreting adenine nucleotides, which relax endothelial cell junctions45.
Physical contacts between macrophages and CTCs help to pull CTCs
across capillary walls in the lungs46. Many of these factors also enhance
the extravasation of CTCs in the brain, which involves the cancer-cellderived sialyltransferase enzyme ST6GALNAC5 (ref. 47), cathepsin S48
and microRNAs mir-105 and mir-181c49,50. These mediators each provide
a finite increase in the probability of metastatic seeding and frequently
act in parallel. In sum, a combination of priming signals from the tumour
stroma, the composition of CTC clusters, blood-circulation patterns, the
structure of target-organ capillary walls and cancer-cell-autonomous
functions determine metastatic infiltration of specific organs.

Tissue defences against infiltrating cancer cells

In primary tumours, cancer cells develop in a co-evolving microenvironment that suppresses immune surveillance17,51. However, because support
is not available immediately to cancer cells as they infiltrate distant organs,
most of these cells will die57. The seed and soil hypothesis was based on
the observation that different cancers show a predilection for metastasis
in different organs and predicted that certain organs are more hospitable

to wandering cancer cells than others52. Seed and soil is an appealing

metaphor but can be misleading. To disseminating cancer cells, every
distant soil is deadly, although some soils can be less deadly than others.
In fact, the most welcoming of all soils could be the primary tumour itself.
The preferential reseeding of primary or metastatic tumours by CTCs,
rather than the seeding of tumour-free secondary sites, is called tumour
self-seeding53. In experimental systems, self-seeding can amplify the most
aggressive cancer-cell clones53 and disperse drug-resistant clones during
the treatment of metastatic melanoma with targeted therapy54.
In a new and challenging microenvironment, freshly disseminated cancer cells can be particularly vulnerable to immune surveillance (Fig. 1).
Major players in antimetastatic immune surveillance include cytotoxic
T cells and natural killer (NK) cells55. For instance, the depletion of cytotoxic T cells or NK cells has been shown to increase metastasis56,57 and
inhibition of the tyrosine kinase Mer, a negative regulator of NK cells,
suppresses metastasis58. Moreover, the specific immune-cell composition
of an organ can influence the organs susceptibility to overt metastasis. The
liver, for example, is rich in NK cells. Neutralization of pro-apoptotic NKcell-derived tumour necrosis factor (TNF)-related apoptosis-inducing
ligand (TRAIL), or the genetic depletion of NK cells, increases hepatic
metastasis in mice59,60. Advances in immunotherapy, most notably those
that use immune checkpoint inhibitors, have yielded striking results in
metastatic melanoma and other tumour types61,62. Thus, immunity is a
major defence against metastasis.
Other cell types can also mount a strong defence against metastatic
infiltration. Astrocytes, the most abundant type of cell in the brain, reject
extravasated cancer cells by releasing the serine protease plasminogen
activator (PA). PA converts the zymogen plasminogen to plasmin, which
then mobilizes the pro-apoptotic cytokine Fas ligand (FasL) to kill infiltrating cancer cells. To avert this fate, brain metastatic cells from breast
and lung adenocarcinomas produce the PA inhibitors neuroserpin and
serpin B2 (ref. 9).

Supportive niches

Adult stem cells reside in specialized niches that provide cues that help
to maintain a balance between stem-cell proliferation and quiescence
as well as self-renewal and differentiation. Stem-cell niches are rich in
developmental and self-renewal signals, such as hedgehog, Wnt, members of the TGF- family and the chemokine CXCL12 (refs 63, 64).
Tumours are thought to arise from mutant stem cells in their native
niches or from the progeny of cells that retain their tumour-initiating
capacity and benefit from these niche signals6568. After cancer stem cells
disperse to distant sites, their survival and tumour-initiating potential can benefit similarly from interactions with specialized niches69
(Fig. 2). Evidence suggests that prostate-carcinoma stem cells occupy
native haematopoietic stem-cell niches in the bone marrow70. The
CXCL12 receptor CXCR4 is a marker and mediator of breast-cancer
metastasis to CXCL12-rich bone-marrow sites71. Breast tumours that are
rich in a CXCL12-secreting mesenchymal stroma select for CXCL12responsive cancer-cell populations that are predisposed to survive in the
bone marrow72.
Small blood vessels are surrounded by a space that is rich in supportive signals and can favour cancer stem-cell growth and resistance to
anticancer therapy73,74 (Fig. 2). A striking example of the interaction of
metastatic cells with perivascular sites is observed in brain metastasis by
breast-cancer, lung-cancer and melanoma cells in which the extravasated
cancer cells remain closely associated with capillaries8,75. The cells spread
along the basal lamina that surrounds the capillaries and proliferate to
form a sheath that eventually engulfs and remodels the co-opted capillary
network a process that is mediated by expression of the cell-adhesion
molecule L1CAM in the metastatic cells9 (Fig. 2).
DTCs can set up an ad hoc niche by producing components of stem-cell
niches themselves (Fig. 2). Lung-metastatic breast cancer cells produce the
extracellular-matrix protein tenascin C, which is deposited in the incipient colony to amplify Notch and Wnt signalling76. Breast-cancer stem cells
can also secrete TGF-, which stimulates stromal fibroblasts to produce

3 0 0 | N AT U R E | VO L 5 2 9 | 2 1 JA N UA RY 2 0 1 6

2016 Macmillan Publishers Limited. All rights reserved

REVIEW INSIGHT
periostin, a binding partner of tenascin C that recruits Wnt factors11. The
secretion by cancer cells of the collagen-crosslinking enzymes LOX and
PLOD2, which stiffen the extracellular matrix, amplifies integrinfocal
adhesion signalling and also favours metastasis7779.
Experimental models have provided evidence that systemic signals
from primary tumours can influence the microenvironment of distant
organs by creating pre-metastatic niches before the arrival of CTCs80,81
(Fig. 2). Different classes of systemic mediators, such as tumour-derived
inflammatory cytokines, exosomes and extracellular-matrix-remodelling
enzymes, have been shown in breast-, lung- and gastrointestinal-tumour
models to recruit bone-marrow-derived cells and precondition the lung,
liver or bone marrow for infiltrating cancer cells8285. Melanoma cells
secrete exosomes that might induce vascular leakiness and inflammation during the formation of pre-metastatic niches86. Similarly, macrophage inhibitory factor (MIF)-containing exosomes that are released
by pancreatic cancer cells increase liver metastasis by inducing TGF-
secretion, stimulating the production of the glycoprotein fibronectin
in hepatic stellate cells and recruiting bone-marrow-derived cells to the
liver83. Integrins were proposed to target cancer-cell-derived exosomes to
specific organs to unload their cargo and prepare the organ for the arrival
of tumour cells87.
Observations from the clinic raise questions about the role of premetastatic niches in people with cancer. Most patients develop metastasis months or years after the removal of their primary tumour, until
which time tumour cells remain largely dormant. Yet in experimental
models, pre-metastatic niches support the immediate outgrowth of DTCs.
Research must address whether pre-metastatic niches remain primed for
years after the removal of a primary tumour or, alternatively, if their role
is to enhance the survival of infiltrating cells to increase their numbers
before they enter a latent state.

Growth and survival pathways

A plethora of genes and signals support metastatic cell growth and survival in experimental models, and the expression of these genes can
predict relapse in the clinic (Box 1). Many such pro-metastatic stromal
mediators ultimately activate stem-cell support pathways, such as the Wnt,
TGF-, bone morphogenetic protein (BMP), Notch and signal transducer
and activator of transcription (Stat)3 pathways. Others activate pathways

that integrate cell metabolism and survival, including the phosphoinositide 3-kinaseprotein kinase B (PI3KAKT), mitogen-activated
protein kinase (MAPK) and hypoxia-inducible factor (HIF) pathways.
Positional and mechanical pathways (Hedgehog and Hippo) and inflammatory pathways (NFB and Stat1) are also activated69. These pathways
also drive development and tissue regeneration, but what is distinctive in
the case of metastasis are the strategies that cancer cells employ to ensure
that sufficient pathway activation is achieved in microenvironments with
low levels of activating signals (Fig. 3). DTCs seem to be selected according to their ability to use whatever cues the host tissue offers.
By expressing autocrine pathway activators or by recruiting stromal
cells to produce them (Fig. 3), DTCs can stimulate a crucial pathway or
amplify the pathways signalling output69. For example, Stat3 stimulation
by autocrine interleukin (IL)-6 mediates metastasis in prostate cancer
cells, and PI3KAKT stimulation by autocrine insulin-like growth factor 2
(IGF2) mediates metastasis in oesophageal cancer cells88,89. The intracellular tyrosine kinase Src amplifies the ability of stromal CXCL12 to activate
PI3KAKT signalling in breast cancer cells that infiltrate the bone marrow90. Breast cancer cells produce colony-stimulating factor 1 (CSF1) to
recruit tumour-associated macrophages as a source of epidermal growth
factor (EGF)91 or secrete CXCL1 to recruit myeloid precursors as a source
of calprotectin (S100A8/9) for MAPK activation92. Colorectal-cancer
stem cells that reach the liver express TGF- to recruit mesenchymal cells
as a source of IL-11 for Stat3 activation in the cancer cells12.
Cancer cells can also obtain vital support through physical contact
with stromal cells (Fig. 3). Claudin-2-mediated cellcell interactions
between breast cancer cells and hepatocytes induce c-Met (hepatocyte
growth factor receptor) signalling and stimulate metastasis to the liver93.
Contact between membrane vascular cell adhesion molecule 1 (VCAM1)
on breast cancer cells that have infiltrated the lungs and 4 integrins on
stromal monocytes and macrophages activates PI3KAKT signalling
in the cancer cells94. By contrast, contact between VCAM1 from breast
cancer cells that exit dormancy in the bone marrow and 4 integrins on
monocytic precursors accelerates the differentiation of these precursors
into osteoclasts that mediate osteolytic metastasis95.
The activity of pro-metastatic pathways can be further increased by
epigenetic alterations. For example, von HippelLindau (VHL)-mutant
renal-cell-carcinoma cells gain metastatic activity in multiple organs

BOX 1

Origin of metastatic traits

Metastasis develops through genetic and epigenetic changes and
subsequent selection for favourable traits under the pressure of
successive bottlenecks138,139. Genomic comparisons show close
clonal relationships between primary tumours and their metastases.
Specific ancestors of metastatic clones can often be identified in the
primary tumour30,139,140, which supports the hypothesis that late clonal
expansion in the primary tumour gives rise to metastasis-competent
clones. These studies also provide evidence that metastases seed
further metastases30. Disseminated cancer cells remain dependent
on the oncogenic mutations that underlie the primary tumour,
which provides a basis for treating metastasis with drugs that target
these oncogenic drivers125. Oncogenic mutant alleles accumulate in
metastases, such as the gain of mutant KRAS in pancreatic cancer
metastasis141 and TP53 and androgen-receptor mutations in prostate
cancer metastasis142. However, no recurrent metastasis-specific
mutations have been identified so far, which suggests that epigenetic
alterations and other mechanisms of modifying gene expression are
the predominant source of selectable pro-metastatic traits during
clonal evolution in metastasis139,140.
The cell-autonomous traits that favour the dissemination of cancer
cells, their resistance in the circulation, extravasation and initial survival

in distant organs become important as soon as the cells leave the

primary tumour, and are pre-selected in the primary tumour. For
example, certain mediators of neoangiogenesis in breast tumours,
such as COX2, epiregulin, MMP1 and VEGF, are repurposed by cancer
cells for extravasation in the lungs and brain42,47,143. Stromal TGF-
in triple-negative breast carcinomas which do not express the
oestrogen receptor, progesterone receptor or HER2 induces the
expression of the protein angiopoietin-like 4 (ANGPTL4) and primes
the cancer cells for extravasation in the lungs41. These early metastatic
traits can be selected under the stresses of tissue invasion, immune
surveillance or hypoxia. The evidence favours a model in which a
considerable proportion of cancer cells in a primary tumour acquire
pro-metastatic traits that confer a finite probability of success in the
early steps of metastasis. Clones with the most effective combination
of pro-metastatic traits are most likely to give rise to metastatic lesions
and also to re-seed the primary tumour. Beyond these early steps,
cancer cells continue to evolve after their dissemination to distant
organs to acquire traits for overt colonization, as demonstrated in bone
metastasis114. Cancer cells that undergo early dissemination could
evolve in parallel with, but independently from, the primary tumour144.
The origin of metastatic traits remains a fertile area for future research.

2 1 JA N UA RY 2 0 1 6 | VO L 5 2 9 | N AT U R E | 3 0 1

2016 Macmillan Publishers Limited. All rights reserved

INSIGHT REVIEW
KEY

Activation

Stromal cell

Amplification

Soluble factors
and ECM molecules
Growth and
survival factors

Paracrine cytokines
and chemokines

Membrane-bound
ligand

Intracellular mediator

Cancer cell

Autocrine
activation

Autocrine and intracellular

amplification

Paracrine
activation

Paracrine
amplification

Figure 3 | The activation of growth and survival

pathways by disseminated cancer cells. During
colonization, metastasis-initiating cells require
the activity of a common set of pathways that
support the survival and growth of stem and
progenitor cells. After infiltrating distant tissues
that offer limiting levels of pathway activators,
the disseminated cancer cells express paracrine
mediators (such as IL-6 and IGF2), ECM
components (such as TNC, LOX and PLOD2) and
intracellular mediators (such as SRC and Ezrin)
that activate and amplify survival and growth
pathways. They also express paracrine factors
(such as CSF1, CXCL1 and TGF-) that help to
recruit stromal cells as a further source of soluble
activators and amplifiers that include the cytokines
IL-6, WNT, EGF and S100A8/9 and the ECM
components TNC and POSTN. Cancer cells can
also achieve pathway activation through cellcell
contact using receptorligand pairings such as
VCAM14 integrin and Jagged1Notch.

Cellcell contact
and activation

Stem-cell growth and survival pathways

through changes in DNA methylation and histone acetylation that expand

the range of HIF target genes the dominant oncogenic pathway in
these cells96. Additional inputs come from the expression of microRNAs
that promote or suppress metastasis by regulating multiple mediators of
tumourstroma interactions97100. These examples show that DTCs resort
to diverse mechanisms to procure vital inputs for survival and the retention of their tumour-initiating capacity.
Exactly when, where and how cancer cells turn to these various stromal
cues is unclear. Are these niches and pathways important for all stages of
metastatic colonization, at all times? Some might be crucial only after
extravasation, when cancer cells are challenged by tissue defences or during the latent phase of metastasis in which cancer cells subsist for years
without outgrowth. Yet others might be important only when DTCs exit
dormancy and enter a proliferative state. Such questions remain unanswered because most experimental models of metastasis do not incorporate a latency phase. This gap in knowledge is also of concern from a
translational perspective. Treating overt metastasis by targeting a survival
mechanism that was relevant only during the initial infiltration of distant
organs might have no clinical benefit. Likewise, targeting an oncogenic
driver pathway in latent DTCs could be futile. But targeting mechanisms that support the viability of latent DTCs could actually eradicate
residual disease.

Latent metastasis

The observation that patients relapse with metastatic disease months or

years after removal of their primary tumour, combined with the detection
of DTCs in the bone marrow of people with no evidence of metastatic
disease, demonstrates that cancer cells that disseminate before treatment
of the primary tumour retain the ability to initiate metastatic growth long
thereafter. Some organs are more permissive than others to the accumulation of latent DTCs. For example, people with colorectal or gastric
cancer can harbour DTCs in the bone marrow, yet the incidence of bone
metastasis in these patients is low101. The incidence of DTCs in the bone
marrow is a predictor of metastasis not only to bone, but also to the liver,
lungs and brain2.
Tumour dormancy is thought to occur through two modes (Fig. 1). In
cellular dormancy, isolated DTCs enter a state of proliferative quiescence.
Indeed, most DTCs in bone-marrow samples from people with cancer
are found as quiescent single cells101103. But in tumour mass dormancy,
micrometastases cease to grow because of insufficient vascularization or

constant culling by immune defences102. It is uncertain which mode most

frequently leads to overt metastases.
Despite the biological and clinical relevance of metastatic latency, little
is known about the mechanisms by which cancer cells enter a dormant
state and the signals that sustain it, the niches that dormant cancer cells
inhabit and what triggers the resumption of the cells aggressive growth.
The paucity of experimental model systems that incorporate a latent phase
and the cost of studying such a process over an extended period in animal
models have hindered progress. However, stromal signals that impose
tumour dormancy have been identified in mouse xenograft models.
TGF- and BMPs, members of the TGF- family, can enforce quiescence
and inhibit self-renewal in carcinoma DTCs104106. The perivascular niche
has also been implicated in the induction of cancer-cell dormancy107.
But environments that are rich in type I collagen108 or fibronectin109
inhibit dormancy.
A scarcity of stromal growth factors and an abundance of growthinhibitory cues can favour metastatic dormancy in experimental models.
On their own, however, these signals might not be able to sustain longterm metastatic latency. Tissues that host DTCs, such as the lungs, liver or
bone marrow, do not exist in a perpetual state of growth inhibition. On the
contrary, they support cell proliferation as part of normal tissue homeostasis and regeneration in a context that would regularly stimulate DTCs
to enter the cell cycle. Therefore, cancer-cell-autonomous mechanisms
that self-impose quiescence might be necessary to maintain DTCs in a
dormant state. It is also unclear how a continuously quiescent DTC population could evolve and acquire the necessary traits for overt metastasis.
Evidence that DTCs are kept in a latent state by the immune system
comes from the transmission of cancer during organ transplantation. Kidney, liver and lung transplants from donors who were cured of melanoma
or who had developed glioblastoma, which is generally considered to
be a nonmetastatic tumour, have resulted in the transmission of donorderived tumours to immunosuppressed recipients110,111. Such cases suggest that DTCs are maintained in a latent state by constant pressure from
the immune system. DTCs might enter the cell cycle intermittently, and
their progeny might undergo rapid elimination by the immune system
while evolving to acquire traits that facilitate metastatic outbreak.

Overt metastasis

In some organs, breaking down growth-inhibitory or immune barriers

can be sufficient to initiate the aggressive metastatic outgrowth of DTCs.

3 0 2 | N AT U R E | VO L 5 2 9 | 2 1 JA N UA RY 2 0 1 6

2016 Macmillan Publishers Limited. All rights reserved

REVIEW INSIGHT

Survival
Seeding from
primary tumour

Drug-sensitive
cancer cell

Protected niche

Therapy-induced secretome

Overt colonization

Latency

Cancer therapy

Stromal cell

Incomplete elimination
of tumour
Drug-resistant
mutant cell

Soluble factor
Accelerated growth

Relapse

Reactive stroma

Drug-resistant tumour

Figure 4 | The biology of metastasis, before and after cancer therapy.

Latent metastasis occurs owing to conditions that preserve the survival and
tumour-initiating ability of disseminated cancer cells. The elimination of latent
metastasis by targeting these survival mechanisms would prevent metastasis.
Cancer cells that exit latency form manifest metastases. This condition is
treated with combinations of conventional chemotherapy, targeted therapy
and immunotherapy. Although treatment can dramatically reduce the
metastatic burden, tumour elimination is frequently incomplete. A considerable

proportion of the tumour cell population will withstand treatment by adaptating

its intracellular pathways or by activating supportive paracrine inputs. The stress
of targeted therapy causes drug-sensitive cancer cells to express a large number
of secreted factors, known as a therapy-induced secretome, that can salvage
drug-sensitive cells and accelerate the growth of minority drug-resistant cancer
cells. This accelerated growth drives relapse as a drug-resistant tumour. The
survival and growth mechanisms that residual cancer cells use during cancer
therapy might resemble those used by their predecessors in the latent phase.

However, organs differ markedly in the structure and composition of

their tissues, and their overt colonization involves distinct organ-specific
metastatic traits112. This translates into patterns of metastasis distribution
that vary greatly depending on the tumour type. For example, prostate
cancer has a propensity to relapse in bone; uveal melanoma tends to recur
in the liver; and sarcomas often return in the lungs. In contrast, melanomas, breast carcinomas and lung adenocarcinomas tend to relapse in
multiple organs. The kinetics of relapse also varies. For example, whereas
recurrence in the brain frequently occurs at an early stage in metastatic
lung cancer, this is typically a late event in metastatic breast cancer. Certain oncogenic mutations seem to affect metastatic tropism. For example,
KRAS-mutant colon cancer advances to colonize the lungs from established liver metastases113.
Bone metastasis is the best-understood case of overt colonization
and offers clear examples of the organ-specific traits that determine this
final stage of metastasis. Osteolytic bone metastasis occurs when the
balance of bone-generating and bone-resorbing osteoclasts is altered
to favour the latter. Numerous mediators of osteoclast activation have
been implicated in this process114. Cancer-cell-derived parathyroid hormone-related protein (PTHrP), IL-11 and TNF- prompt osteoblasts
to release receptor activator of NFB ligand (RANKL), a protein that
stimulates osteoclast maturation114116. Bone metastatic cells also produce
MMPs, which increase RANKL activity117 and reduce the levels of the
RANKL antagonist osteoprotegerin118. Expression of the Notch ligand

Jagged1 and cell-adhesion molecules VCAM1 and soluble intercellular

adhesion molecule 1 (sICAM1) also contributes to the mobilization of
osteoclasts95,119,120. Bone-matrix degradation by hyperactivated osteoclasts releases TGF- that, in turn, augments the production of PTHrP,
IL-11 and Jagged1 in the cancer cells and drives a vicious cycle of bone
destruction114116.
Interestingly, prostate cancer cells that spread to bone alter the homeostatic balance in favour of osteoblastic activity, which stimulates the deposition of bone matrix and leads to eventual displacement of the bone
marrow. Cancer-cell factors that are implicated in osteoblastic metastasis include fibroblast growth factors (FGFs), insulin-like growth factors
(IGFs) and VEGFs, as well as endothelin 1, Wnt factors and BMPs114.
Thus, bone metastasis provides a compelling example of how cancer cells
engage their host microenvironment in overt metastasis. Similarly, specific stromal components can be engaged in other organs by metastatic
cells with the necessary organ-specific colonization traits. For example,
when breast and lung carcinomas spread to the brain, the cancer cells can
profitably engage astrocytes and microglia by expressing endothelin 1
(ref. 121). However, our knowledge of overt-colonization traits that are
specific to organs other than bone is limited, and these traits need further
investigation.
In certain groups of patients, metastasis is confined mainly to a
particular organ that is better able to resist therapy than others. An
example is the rise in the incidence of late brain and leptomeningeal
2 1 JA N UA RY 2 0 1 6 | VO L 5 2 9 | N AT U R E | 3 0 3

2016 Macmillan Publishers Limited. All rights reserved

INSIGHT REVIEW
metastasis in people with HER2+ breast cancer. Such patients can
benefit from advances in targeted therapies that suppress extracranial metastasis. However, this success is often short-lived owing to
the emergence of brain metastasis. Brain metastasis is a major cause
of morbidity and mortality, with an overall incidence that is ten times
higher than that of all primary brain tumours combined, and has few
therapeutic options. A better understanding of its underlying mechanisms is urgently needed.

After therapy

The surgical removal of a malignant tumour is often complemented with

radiotherapy and systemic chemotherapy to suppress relapse. If metastasis becomes clinically manifest, most systemic treatments are designed
to target metastasis irrespective of organ site. Treatments include classic
chemotherapy, targeted therapy against oncogenic drivers, immunotherapeutic agents that leverage the antitumour power of the immune
system and, increasingly, a combination of all of these. Treatments that
target metastasis in a particular organ by taking aim at cancer-cell
interactions with the host tissue would be indicated when metastasis
is confined to that organ, as is the case with bone metastasis in some
people with breast cancer. A meta-analysis suggests that adjuvant therapy
with osteoclast-inhibitory bisphosphonates suppresses bone metastasis
and prolongs survival in postmenopausal women with breast cancer122.
Denosumab, an antibody that targets RANKL, reduces the incidence of
bone fractures associated with metastasis in patients who are receiving
aromatase inhibitors123.
Despite these advances, therapies frequently achieve only partial
tumour shrinkage and leave behind substantial amounts of disease. Continued treatment can keep the residual tumour indolent for some time.
However, drug-resistant cancer-cell clones eventually emerge that drive
rapid relapse124,125. As a result, the cure rates of patients with metastasis
remain disappointingly low.
Research is beginning to focus on the biology of residual metastatic cells
after therapy, and is aiming to better suppress the re-emergence of cancer
cells (Fig. 4). The cancer-cell population can resist treatment through
alterations to negative-feedback signalling loops126 and supportive interactions with the tumour microenvironment. For example, DNA-damaging agents induce the secretion of trophic factors such as IL-6 and the
metallopeptidase inhibitor Timp-1 in normal cells of the thymus, which
creates a chemoprotective niche for the survival of residual cancer cells
that facilitates eventual relapse127. Similarly, stromal fibroblasts secrete
Wnt16b in response to chemotherapy, which promotes resistance to therapy in prostate cancer128. Chemotherapy-induced expression of TNF- in
tumour-associated endothelial and mesenchymal cells has been shown to
amplify the expression of the pro-metastatic cytokine CXCL1 in cancer
cells92. In BRAF-mutant melanomas that are treated with RAF inhibitors,
tumour-associated macrophages secrete TNF- and VEGF129,130. Tumourassociated fibroblasts secrete hepatocyte growth factors (HGFs)131, which
protect melanoma cells and limit the effectiveness of therapy.
Under the stress of therapy, the cancer cells themselves can be a source
of survival signals54,132,133. Targeted therapy with tyrosine-kinase inhibitors
that is directed against melanoma (vemurafenib and dabrafenib) or lung
adenocarcinoma (erlotinib and crizotinib) triggers the production of a
complex secretome the therapy-induced secretome that activates
multiple survival pathways in the remaining, drug-sensitive cancer cells54.
Furthermore, this secretome can stimulate the outgrowth, dissemination
and further metastatic seeding of clones with mutations that confer drug
resistance. Collectively, these findings reveal a complex biology in cancer
cell populations that remain after treatment of metastatic tumours and
contribute to tumour relapse.

Future directions

An important objective for current research is the identification of

mediators of metastasis that are common to different organs and types
of tumours. Although organ-specific metastasis has intrigued researchers for over a century, in reality many patients are affected by, or are at

risk of, metastasis in multiple organs. For these cases, it would be valuable to identify common mediators as potential therapeutic targets. For
example, checkpoint immunotherapy which has shown encouraging
success in the clinic is based on the premise that immune evasion is a
shared feature of metastatic disease, irrespective of the organ site. Deeper
understanding of the common mediators of metastasis and how tumours
regrow after therapy would help to improve strategies for the elimination
of residual disease.
The advent of single-cell analysis techniques, such as single-cell RNA
sequencing and signalling-pathway profiling, is allowing functional and
phenotypic analysis of heterogeneous cell populations in unprecedented
detail134137. The application of these techniques to residual disease and
overt metastases will enable researchers to define tumour heterogeneity, cell-population structures and evolution, and cell-type-specific
response patterns to stromal cues and therapeutic agents, as well as other
parameters, at a depth never before possible. Furthermore, the ability to
analyse circulating tumour DNA in the blood of people with cancer will
facilitate the monitoring of therapy responses, the emergence of distinct
resistant clones and patterns of early disease recurrence.
Preventing metastasis in high-risk patients would be far better than
having to treat it. The systemic nature of metastatic disease, the heterogeneity of metastatic tumours, the multitude of genes and pathways involved
in different organs and the many mechanisms of drug resistance paint a
sobering picture of the problems that must be overcome to address overt
metastatic disease. Ostensibly, the goal of systemic therapy that is delivered after removal of a primary tumour is to prevent relapse. However,
most agents used in adjuvant therapy are designed to target growing
cancer cells rather than the quiescent DTCs that predominate during
metastatic latency. A better understanding of the basis for metastatic colonization and its latent phase, in particular, is needed to develop superior
treatments. Research on the mechanisms that support the viability of
latent metastatic cells should yield clues for targeting residual disease,
with the goal of preventing metastasis.
Received 26 August; accepted 11 November 2015.
1.

Nagrath, S. et al. Isolation of rare circulating tumour cells in cancer patients by

microchip technology. Nature 450, 12351239 (2007).
2. Braun, S. et al. A pooled analysis of bone marrow micrometastasis in breast
cancer. N. Engl. J. Med. 353, 793802 (2005).
3. Wong, C. W. et al. Apoptosis: an early event in metastatic inefficiency. Cancer Res.
61, 333338 (2001).
4. Minn, A. J. et al. Distinct organ-specific metastatic potential of individual breast
cancer cells and primary tumors. J. Clin. Invest. 115, 4455 (2005).
5. Chambers, A. F., Groom, A. C. & MacDonald, I. C. Dissemination and growth of
cancer cells in metastatic sites. Nature Rev. Cancer 2, 563572 (2002).
6. Luzzi, K. J. et al. Multistep nature of metastatic inefficiency: dormancy of
solitary cells after successful extravasation and limited survival of early
micrometastases. Am. J. Pathol. 153, 865873 (1998).
7. Cameron, M. D. et al. Temporal progression of metastasis in lung: cell survival,
dormancy, and location dependence of metastatic inefficiency. Cancer Res. 60,
25412546 (2000).
8. Kienast, Y. et al. Real-time imaging reveals the single steps of brain metastasis
formation. Nature Med. 16, 116122 (2010).

This study used multiphoton laser-scanning microscopy to track the fate of
individual metastasizing cancer cells in the brain.
9. Valiente, M. et al. Serpins promote cancer cell survival and vascular co-option in
brain metastasis. Cell 156, 10021016 (2014).

This study identified that proteins called serpins promote metastasis in the
brain by shielding cancer cells from the tissue defences of the reactive brain
stroma and also promote vascular co-option mediated by L1CAM.
10. Heyn, C. et al. In vivo MRI of cancer cell fate at the single-cell level in a mouse
model of breast cancer metastasis to the brain. Magn. Reson. Med. 56, 1001
1010 (2006).
11. Malanchi, I. et al. Interactions between cancer stem cells and their niche govern
metastatic colonization. Nature 481, 8589 (2012).

This study demonstrated that breast cancer stem cells generate a permissive
metastatic niche by triggering the secretion of stromal periostin.
12. Calon, A. et al. Dependency of colorectal cancer on a TGF--driven program in
stromal cells for metastasis initiation. Cancer Cell 22, 571584 (2012).

This study found that TGF- signalling in stromal cells increases the efficiency
of organ colonization by colorectal cancer cells and shed light on the paradox
that high levels of TGF- are associated with poorer prognosis in colorectal
cancer, despite the fact that cancer cells frequently exhibit inactivating
mutations in TGF--pathway components.

3 0 4 | N AT U R E | VO L 5 2 9 | 2 1 JA N UA RY 2 0 1 6

2016 Macmillan Publishers Limited. All rights reserved

REVIEW INSIGHT
13. Nguyen, D. X., Bos, P. D. & Massague, J. Metastasis: from dissemination to organspecific colonization. Nature Rev. Cancer 9, 274284 (2009).
14. Chaffer, C. L. & Weinberg, R. A. A perspective on cancer cell metastasis. Science
331, 15591564 (2011).
15. Hall, A. The cytoskeleton and cancer. Cancer Metastasis Rev. 28, 514 (2009).
16. Kessenbrock, K., Plaks, V. & Werb, Z. Matrix metalloproteinases: regulators of the
tumor microenvironment. Cell 141, 5267 (2010).
17. Quail, D. F. & Joyce, J. A. Microenvironmental regulation of tumor progression
and metastasis. Nature Med. 19, 14231437 (2013).
18. Giampieri, S. et al. Localized and reversible TGF signalling switches breast
cancer cells from cohesive to single cell motility. Nature Cell Biol. 11, 12871296
(2009).
19. Roh-Johnson, M. et al. Macrophage contact induces RhoA GTPase signaling to
trigger tumor cell intravasation. Oncogene 33, 42034212 (2014).
20. Friedl, P. & Gilmour, D. Collective cell migration in morphogenesis, regeneration
and cancer. Nature Rev. Mol. Cell Biol. 10, 445457 (2009).
21. Magnon, C. et al. Autonomic nerve development contributes to prostate cancer
progression. Science 341, 1236361 (2013).
22. Thiery, J. P., Acloque, H., Huang, R. Y. & Nieto, M. A. Epithelial-mesenchymal
transitions in development and disease. Cell 139, 871890 (2009).
23. Tam, W. L. & Weinberg, R. A. The epigenetics of epithelial-mesenchymal plasticity
in cancer. Nature Med. 19, 14381449 (2013).
24. Fischer, K. R. et al. Epithelial-to-mesenchymal transition is not required for lung
metastasis but contributes to chemoresistance. Nature 527, 472476 (2015).
25. Zheng, X. et al. Epithelial-to-mesenchymal transition is dispensable for
metastasis but induces chemoresistance in pancreatic cancer. Nature 527,
525530 (2015).
26. Tabassum, D. P. & Polyak, K. Tumorigenesis: it takes a village. Nature Rev. Cancer
15, 473483 (2015).
27. Marusyk, A. et al. Non-cell-autonomous driving of tumour growth supports subclonal heterogeneity. Nature 514, 5458 (2014).

This study showed that a minority subclone in a heterogeneous tumour is
able to drive proliferation of the whole tumour by overcoming environmental
constraints.
28. Cleary, A. S., Leonard, T. L., Gestl, S. A. & Gunther, E. J. Tumour cell heterogeneity
maintained by cooperating subclones in Wnt-driven mammary cancers. Nature
508, 113117 (2014).
29. Calbo, J. et al. A functional role for tumor cell heterogeneity in a mouse model of
small cell lung cancer. Cancer Cell 19, 244256 (2011).
30. Gundem, G. et al. The evolutionary history of lethal metastatic prostate cancer.
Nature 520, 353357 (2015).

This study used whole-genome sequencing and computational analysis to
provide evidence for metastasis-to-metastasis spread and the transfer of
multiple tumour clones between metastatic sites in people with prostate
cancer.
31. Aceto, N. et al. Circulating tumor cell clusters are oligoclonal precursors of breast
cancer metastasis. Cell 158, 11101122 (2014).

This study showed that CTC clusters have increased metastatic potential
compared with single CTCs.
32. Labelle, M., Begum, S. & Hynes, R. O. Direct signaling between platelets and
cancer cells induces an epithelial-mesenchymal-like transition and promotes
metastasis. Cancer Cell 20, 576590 (2011).
33. Le Gal, K. et al. Antioxidants can increase melanoma metastasis in mice. Sci.
Transl. Med. 7, 308re8 (2015).
34. Piskounova, E. et al. Oxidative stress inhibits distant metastasis by human
melanoma cells. Nature 527, 186191 (2015).
35. Denve, E. et al. Capture of viable circulating tumor cells in the liver of colorectal
cancer patients. Clin. Chem. 59, 13841392 (2013).
36. Al-Mehdi, A. B. et al. Intravascular origin of metastasis from the proliferation of
endothelium-attached tumor cells: a new model for metastasis. Nature Med. 6,
100102 (2000).
37. Aird, W. C. Phenotypic heterogeneity of the endothelium: I. Structure, function,
and mechanisms. Circ. Res. 100, 158173 (2007).
38. Budczies, J. et al. The landscape of metastatic progression patterns across major
human cancers. Oncotarget 6, 570583 (2015).
39. Minn, A. J. et al. Genes that mediate breast cancer metastasis to lung. Nature
436, 518524 (2005).
40. Matsuda, Y., Schlange, T., Oakeley, E. J., Boulay, A. & Hynes, N. E. WNT signaling
enhances breast cancer cell motility and blockade of the WNT pathway by
sFRP1 suppresses MDA-MB-231 xenograft growth. Breast Cancer Res. 11, R32
(2009).
41. Padua, D. et al. TGF primes breast tumors for lung metastasis seeding through
angiopoietin-like 4. Cell 133, 6677 (2008).
42. Gupta, G. P. et al. Mediators of vascular remodelling co-opted for sequential
steps in lung metastasis. Nature 446, 765770 (2007).
43. Tichet, M. et al. Tumour-derived SPARC drives vascular permeability and
extravasation through endothelial VCAM1 signalling to promote metastasis.
Nature Commun. 6, 6993 (2015).
44. Weis, S., Cui, J., Barnes, L. & Cheresh, D. Endothelial barrier disruption by VEGFmediated Src activity potentiates tumor cell extravasation and metastasis. J. Cell
Biol. 167, 223229 (2004).
45. Schumacher, D., Strilic, B., Sivaraj, K. K., Wettschureck, N. & Offermanns, S.
Platelet-derived nucleotides promote tumor-cell transendothelial migration and
metastasis via P2Y2 receptor. Cancer Cell 24, 130137 (2013).
46. Qian, B. Z. et al. CCL2 recruits inflammatory monocytes to facilitate breasttumour metastasis. Nature 475, 222225 (2011).

47. Bos, P. D. et al. Genes that mediate breast cancer metastasis to the brain. Nature
459, 10051009 (2009).
48. Sevenich, L. et al. Analysis of tumour- and stroma-supplied proteolytic networks
reveals a brain-metastasis-promoting role for cathepsin S. Nature Cell Biol. 16,
876888 (2014).
49. Zhou, W. et al. Cancer-secreted miR-105 destroys vascular endothelial barriers
to promote metastasis. Cancer Cell 25, 501515 (2014).
50. Tominaga, N. et al. Brain metastatic cancer cells release microRNA-181ccontaining extracellular vesicles capable of destructing blood-brain barrier.
Nature Commun. 6, 6716 (2015).
51. Hanahan, D. & Coussens, L. M. Accessories to the crime: functions of cells
recruited to the tumor microenvironment. Cancer Cell 21, 309322 (2012).
52. Fidler, I. J. The pathogenesis of cancer metastasis: the seed and soil hypothesis
revisited. Nature Rev. Cancer 3, 453458 (2003).
53. Kim, M. Y. et al. Tumor self-seeding by circulating cancer cells. Cell 139,
13151326 (2009).
54. Obenauf, A. C. et al. Therapy-induced tumour secretomes promote resistance
and tumour progression. Nature 520, 368372 (2015).
55. Eyles, J. et al. Tumor cells disseminate early, but immunosurveillance limits
metastatic outgrowth, in a mouse model of melanoma. J. Clin. Invest. 120,
20302039 (2010).
56. Bidwell, B. N. et al. Silencing of Irf7 pathways in breast cancer cells promotes
bone metastasis through immune escape. Nature Med. 18, 12241231 (2012).
57. Smyth, M. J. et al. Perforin is a major contributor to NK cell control of tumor
metastasis. J. Immunol. 162, 66586662 (1999).
58. Paolino, M. et al. The E3 ligase Cbl-b and TAM receptors regulate cancer
metastasis via natural killer cells. Nature 507, 508512 (2014).
59. Milsom, C. C., Lee, C. R., Hackl, C., Man, S. & Kerbel, R. S. Differential post-surgical
metastasis and survival in SCID, NOD-SCID and NOD-SCID-IL-2Rnull mice with
parental and subline variants of human breast cancer: implications for host
defense mechanisms regulating metastasis. PLoS ONE 8, e71270 (2013).
60. Takeda, K. et al. Involvement of tumor necrosis factor-related apoptosis-inducing
ligand in surveillance of tumor metastasis by liver natural killer cells. Nature Med.
7, 94100 (2001).
61. Postow, M. A. et al. Nivolumab and ipilimumab versus ipilimumab in untreated
melanoma. N. Engl. J. Med. 372, 20062017 (2015).
62. Sharma, P. & Allison, J. P. Immune checkpoint targeting in cancer therapy:
toward combination strategies with curative potential. Cell 161, 205214
(2015).
63. Morrison, S. J. & Spradling, A. C. Stem cells and niches: mechanisms that
promote stem cell maintenance throughout life. Cell 132, 598611 (2008).
64. Hsu, Y. C., Li, L. & Fuchs, E. Emerging interactions between skin stem cells and
their niches. Nature Med. 20, 847856 (2014).
65. Schepers, A. G. et al. Lineage tracing reveals Lgr5+ stem cell activity in mouse
intestinal adenomas. Science 337, 730735 (2012).
66. Chen, J. et al. A restricted cell population propagates glioblastoma growth after
chemotherapy. Nature 488, 522526 (2012).
67. Driessens, G., Beck, B., Caauwe, A., Simons, B. D. & Blanpain, C. Defining the
mode of tumour growth by clonal analysis. Nature 488, 527530 (2012).
68. Kreso, A. & Dick, J. E. Evolution of the cancer stem cell model. Cell Stem Cell 14,
275291 (2014).
69. Oskarsson, T., Batlle, E. & Massague, J. Metastatic stem cells: sources, niches,
and vital pathways. Cell Stem Cell 14, 306321 (2014).
70. Shiozawa, Y. et al. Human prostate cancer metastases target the hematopoietic
stem cell niche to establish footholds in mouse bone marrow. J. Clin. Invest. 121,
12981312 (2011).
71. Mller, A. et al. Involvement of chemokine receptors in breast cancer metastasis.
Nature 410, 5056 (2001).
72. Zhang, X. H. et al. Selection of bone metastasis seeds by mesenchymal signals in
the primary tumor stroma. Cell 154, 10601073 (2013).

This study showed how stromal signals that resemble those in the bone
marrow can skew heterogeneous cancer cell populations in the primary
tumour towards a predominance of clones that are primed for bone
metastasis.
73. Hambardzumyan, D. et al. PI3K pathway regulates survival of cancer stem cells
residing in the perivascular niche following radiation in medulloblastoma in vivo.
Genes Dev. 22, 436448 (2008).
74. Cao, Z. et al. Angiocrine factors deployed by tumor vascular niche induce B cell
lymphoma invasiveness and chemoresistance. Cancer Cell 25, 350365 (2014).
75. Carbonell, W. S., Ansorge, O., Sibson, N. & Muschel, R. The vascular basement
membrane as soil in brain metastasis. PLoS ONE 4, e5857 (2009).
76. Oskarsson, T. et al. Breast cancer cells produce tenascin C as a metastatic niche
component to colonize the lungs. Nature Med. 17, 867874 (2011).
77. Erler, J. T. et al. Lysyl oxidase is essential for hypoxia-induced metastasis. Nature
440, 12221226 (2006).
78. Gilkes, D. M. et al. Procollagen lysyl hydroxylase 2 is essential for hypoxiainduced breast cancer metastasis. Mol. Cancer Res. 11, 456466 (2013).
79. Eisinger-Mathason, T. S. et al. Hypoxia-dependent modification of collagen
networks promotes sarcoma metastasis. Cancer Discov. 3, 11901205 (2013).
80. Kaplan, R. N. et al. VEGFR1-positive haematopoietic bone marrow progenitors
initiate the pre-metastatic niche. Nature 438, 820827 (2005).
81. McAllister, S. S. & Weinberg, R. A. The tumour-induced systemic environment
as a critical regulator of cancer progression and metastasis. Nature Cell Biol. 16,
717727 (2014).
82. Cox, T. R. et al. The hypoxic cancer secretome induces pre-metastatic bone
lesions through lysyl oxidase. Nature 522, 106110 (2015).
2 1 JA N UA RY 2 0 1 6 | VO L 5 2 9 | N AT U R E | 3 0 5

2016 Macmillan Publishers Limited. All rights reserved

INSIGHT REVIEW
83. Costa-Silva, B. et al. Pancreatic cancer exosomes initiate pre-metastatic niche
formation in the liver. Nature Cell Biol. 17, 816826 (2015).
84. Erler, J. T. et al. Hypoxia-induced lysyl oxidase is a critical mediator of bone
marrow cell recruitment to form the premetastatic niche. Cancer Cell 15, 3544
(2009).
85. Wculek, S. K. & Malanchi, I. Neutrophils support lung colonization of metastasisinitiating breast cancer cells. Nature http://dx.doi.org/10.1038/nature16140
(9 December 2015).
86. Peinado, H. et al. Melanoma exosomes educate bone marrow progenitor cells
toward a pro-metastatic phenotype through MET. Nature Med. 18, 883891
(2012).
87. Hoshino, A. et al. Tumour exosome integrins determine organotropic metastasis.
Nature 527, 329335 (2015).
88. Nowak, D. G. et al. MYC drives Pten/Trp53-deficient proliferation and metastasis
due to IL6 secretion and AKT suppression via PHLPP2. Cancer Discov. 5,
636651 (2015).
89. Li, B. et al. Id1-induced IGF-II and its autocrine/endocrine promotion of
esophageal cancer progression and chemoresistanceimplications for IGF-II
and IGF-IR-targeted therapy. Clin. Cancer Res. 20, 26512662 (2014).
90. Zhang, X. H. et al. Latent bone metastasis in breast cancer tied to Src-dependent
survival signals. Cancer Cell 16, 6778 (2009).
91. Wyckoff, J. et al. A paracrine loop between tumor cells and macrophages
is required for tumor cell migration in mammary tumors. Cancer Res. 64,
70227029 (2004).
92. Acharyya, S. et al. A CXCL1 paracrine network links cancer chemoresistance and
metastasis. Cell 150, 165178 (2012).
93. Tabaris, S. et al. Claudin-2 promotes breast cancer liver metastasis by
facilitating tumor cell interactions with hepatocytes. Mol. Cell. Biol. 32, 2979
2991 (2012).
94. Chen, Q., Zhang, X. H. & Massague, J. Macrophage binding to receptor VCAM-1
transmits survival signals in breast cancer cells that invade the lungs. Cancer Cell
20, 538549 (2011).
95. Lu, X. et al. VCAM-1 promotes osteolytic expansion of indolent bone
micrometastasis of breast cancer by engaging 41-positive osteoclast
progenitors. Cancer Cell 20, 701714 (2011).
96. Vanharanta, S. et al. Epigenetic expansion of VHL-HIF signal output drives
multiorgan metastasis in renal cancer. Nature Med. 19, 5056 (2013).
97. Tavazoie, S. F. et al. Endogenous human microRNAs that suppress breast cancer
metastasis. Nature 451, 147152 (2008).
98. Ma, L., Teruya-Feldstein, J. & Weinberg, R. A. Tumour invasion and metastasis
initiated by microRNA-10b in breast cancer. Nature 449, 682688 (2007).
99. Korpal, M. et al. Direct targeting of Sec23a by miR-200s influences cancer cell
secretome and promotes metastatic colonization. Nature Med. 17, 11011108
(2011).
100. Pencheva, N. et al. Convergent multi-miRNA targeting of ApoE drives
LRP1/LRP8-dependent melanoma metastasis and angiogenesis. Cell 151,
10681082 (2012).
101. Pantel, K. & Brakenhoff, R. H. Dissecting the metastatic cascade. Nature Rev.
Cancer 4, 448456 (2004).
102. Sosa, M. S., Bragado, P. & Aguirre-Ghiso, J. A. Mechanisms of disseminated
cancer cell dormancy: an awakening field. Nature Rev. Cancer 14, 611622
(2014).
103. Kang, Y. & Pantel, K. Tumor cell dissemination: emerging biological insights from
animal models and cancer patients. Cancer Cell 23, 573581 (2013).
104. Gao, H. et al. The BMP inhibitor Coco reactivates breast cancer cells at lung
metastatic sites. Cell 150, 764779 (2012).
105. Bragado, P. et al. TGF-2 dictates disseminated tumour cell fate in target organs
through TGF--RIII and p38/ signalling. Nature Cell Biol. 15, 13511361
(2013).
106. Kobayashi, A. et al. Bone morphogenetic protein 7 in dormancy and metastasis
of prostate cancer stem-like cells in bone. J. Exp. Med. 208, 26412655 (2011).
107. Ghajar, C. M. et al. The perivascular niche regulates breast tumour dormancy.
Nature Cell Biol. 15, 807817 (2013).
108. Barkan, D. et al. Metastatic growth from dormant cells induced by a col-Ienriched fibrotic environment. Cancer Res. 70, 57065716 (2010).
109. Aguirre-Ghiso, J. A., Liu, D., Mignatti, A., Kovalski, K. & Ossowski, L. Urokinase
receptor and fibronectin regulate the ERKMAPK to p38MAPK activity ratios that
determine carcinoma cell proliferation or dormancy in vivo. Mol. Biol. Cell 12,
863879 (2001).
110. Strauss, D. C. & Thomas, J. M. Transmission of donor melanoma by organ
transplantation. Lancet Oncol. 11, 790796 (2010).
111. Collignon, F. P., Holland, E. C. & Feng, S. Organ donors with malignant gliomas:
an update. Am. J. Transplant. 4, 1521 (2004).
112. Obenauf, A. C. & Massagu, J. Surviving at a distance: organ-specific metastasis.
Trends Cancer 1, 7691 (2015).
113. Urosevic, J. et al. Colon cancer cells colonize the lung from established liver
metastases through p38 MAPK signalling and PTHLH. Nature Cell Biol. 16,
685694 (2014).
114. Weilbaecher, K. N., Guise, T. A. & McCauley, L. K. Cancer to bone: a fatal
attraction. Nature Rev. Cancer 11, 411425 (2011).
115. Yin, J. J. et al. TGF- signaling blockade inhibits PTHrP secretion by breast
cancer cells and bone metastases development. J. Clin. Invest. 103, 197206
(1999).
116. Kang, Y. et al. A multigenic program mediating breast cancer metastasis to bone.
Cancer Cell 3, 537549 (2003).

117. Lynch, C. C. et al. MMP-7 promotes prostate cancer-induced osteolysis via the
solubilization of RANKL. Cancer Cell 7, 485496 (2005).
118. Lu, X. et al. ADAMTS1 and MMP1 proteolytically engage EGF-like ligands in an
osteolytic signaling cascade for bone metastasis. Genes Dev. 23, 18821894
(2009).
119. Sethi, N., Dai, X., Winter, C. G. & Kang, Y. Tumor-derived JAGGED1 promotes
osteolytic bone metastasis of breast cancer by engaging notch signaling in bone
cells. Cancer Cell 19, 192205 (2011).
120. Ell, B. et al. Tumor-induced osteoclast miRNA changes as regulators and
biomarkers of osteolytic bone metastasis. Cancer Cell 24, 542556 (2013).
121. Kim, S. W. et al. Role of the endothelin axis in astrocyte- and endothelial cellmediated chemoprotection of cancer cells. Neuro Oncol. 16, 15851598 (2014).
122. Early Breast Cancer Trialists Collaborative Group (EBCTCG) et al. Adjuvant
bisphosphonate treatment in early breast cancer: meta-analyses of individual
patient data from randomised trials. Lancet 386, 13531361 (2015).
123. Gnant, M. et al. Adjuvant denosumab in breast cancer (ABCSG-18): a
multicentre, randomised, double-blind, placebo-controlled trial. Lancet 386,
433443 (2015).
124. Holohan, C., Van Schaeybroeck, S., Longley, D. B. & Johnston, P. G. Cancer drug
resistance: an evolving paradigm. Nature Rev. Cancer 13, 714726 (2013).
125. Higgins, M. J. & Baselga, J. Targeted therapies for breast cancer. J. Clin. Invest.
121, 37973803 (2011).
126. Lito, P. et al. Relief of profound feedback inhibition of mitogenic signaling by RAF
inhibitors attenuates their activity in BRAFV600E melanomas. Cancer Cell 22,
668682 (2012).
127. Gilbert, L. A. & Hemann, M. T. DNA damage-mediated induction of a
chemoresistant niche. Cell 143, 355366 (2010).

This study shows that in response to DNA damage, paracrine factors are
released in the thymus, which creates a chemoresistant niche that promotes
the survival of a minimal residual tumour burden and serves as a reservoir for
eventual relapse in a mouse model of Burkitts lymphoma.
128. Sun, Y. et al. Treatment-induced damage to the tumor microenvironment
promotes prostate cancer therapy resistance through WNT16B. Nature Med. 18,
13591368 (2012).
129. Smith, M. P. et al. The immune microenvironment confers resistance to MAPK
pathway inhibitors through macrophage-derived TNF. Cancer Discov. 4,
12141229 (2014).
130. Wang, T. et al. BRAF inhibition stimulates melanoma-associated macrophages to
drive tumor growth. Clin. Cancer Res. 21, 16521664 (2015).
131. Straussman, R. et al. Tumour micro-environment elicits innate resistance to RAF
inhibitors through HGF secretion. Nature 487, 500504 (2012).
132. Kurtova, A. V. et al. Blocking PGE2-induced tumour repopulation abrogates
bladder cancer chemoresistance. Nature 517, 209213 (2015).
133. Huang, Q. et al. Caspase 3-mediated stimulation of tumor cell repopulation
during cancer radiotherapy. Nature Med. 17, 860866 (2011).
134. Macosko, E. Z. et al. Highly parallel genome-wide expression profiling of
individual cells using nanoliter droplets. Cell 161, 12021214 (2015).
135. Patel, A. P. et al. Single-cell RNA-seq highlights intratumoral heterogeneity in
primary glioblastoma. Science 344, 13961401 (2014).
136. Levine, J. H. et al. Data-driven phenotypic dissection of AML reveals progenitorlike cells that correlate with prognosis. Cell 162, 184197 (2015).
137. Lawson, D. A. et al. Single-cell analysis reveals a stem-cell program in human
metastatic breast cancer cells. Nature 526, 131135 (2015).

Using single-cell gene-expression profiling, this study found that early
metastatic cancer cells possess a stem-like gene expression signature and
give rise to heterogeneous tumours, which provides evidence to support a
hierarchical metastasis stem-cell model.
138. Greaves, M. & Maley, C. C. Clonal evolution in cancer. Nature 481, 306313
(2012).
139. Vanharanta, S. & Massagu, J. Origins of metastatic traits. Cancer Cell 24,
410421 (2013).
140. Naxerova, K. & Jain, R. K. Using tumour phylogenetics to identify the roots of
metastasis in humans. Nature Rev. Clin. Oncol. 12, 258272 (2015).
141. Campbell, P. J. et al. The patterns and dynamics of genomic instability in
metastatic pancreatic cancer. Nature 467, 11091113 (2010).
142. Robinson, D. et al. Integrative clinical genomics of advanced prostate cancer. Cell
161, 12151228 (2015).
143. Pencheva, N. & Tavazoie, S. F. Control of metastatic progression by microRNA
regulatory networks. Nature Cell Biol. 15, 546554 (2013).
144. Klein, C. A. Parallel progression of primary tumours and metastases. Nature Rev.
Cancer 9, 302312 (2009).
Acknowledgements The authors thank K. Ganesh and T. Wiesner for useful input.
J.M. was supported by US National Institutes of Health grants CA163167 and
CA129243, the Congressionally Directed Medical Research Programs of the US
Department of Defense, Cancer Center Support Grant P30 CA008748 and the
Center for Metastasis Research of the Memorial Sloan Kettering Cancer Center.
A.C.O. was an Erwin Schredinger Fellowship awardee (J3013, Austrian Science
Fund).
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial
interests. Readers are welcome to comment on the online version of this
paper at go.nature.com/fi9jfd. Correspondence should be addressed to J.M.
([email protected]).

3 0 6 | N AT U R E | VO L 5 2 9 | 2 1 JA N UA RY 2 0 1 6

2016 Macmillan Publishers Limited. All rights reserved