Metastases go with the flow

See allHide authors and affiliations

Science  30 Nov 2018:
Vol. 362, Issue 6418, pp. 999-1000
DOI: 10.1126/science.aat9100

To colonize distant organs and thus disseminate throughout the body, cancer cells and associated factors exploit several fluids for transport. Recently, circulating tumor cells (CTCs) were found to survive and exploit the inner biomechanics of the bloodstream to foster tumor metastasis (1, 2). Thus, in addition to using both the blood and lymphatic circulation as a means to travel throughout the body (35), the underlying forces allow CTCs to seed distant metastases. The contribution of fluids, particularly vascular flow mechanics, and physical constraints raises interesting questions about the biology of metastasis.

According to the “seed and soil” theory of metastasis, CTCs survive and establish growing colonies in distant organs within environments that are compatible with their growth. Successful metastatic outgrowth involves several steps, including organ infiltration, immune escape, growth, and survival in supportive niches (6). The metastatic potential of tumor cells is also tightly linked to body fluids that favor their journey to distant organs (1, 2). CTCs disseminate early through the lymphatic circulation to spread to tumor-draining lymph nodes, which often correlates with reduced survival. Although removal of these metastatic lymph nodes has shown no benefit on overall survival of patients with, for example, melanoma (7), as well as other types of cancer, lymph nodes were recently demonstrated to be intermediate steps for metastases in mice (35). Efficient CTC colonization of distant organs occurs mostly via the blood circulation (6) (see the figure).

On their way to blood vessels (intravasation), invading tumor cells encounter mechanical pressures imposed by architectural constraints of tissues. In particular, space constraints induce nuclear squeezing, which challenges the integrity of the nucleus and triggers genomic rearrangements that might foster metastatic potential (8). It is likely that such pressure also applies to CTCs during arrest in distant sites, extravasation (exiting vessels), and metastatic outgrowth.

Both individual and rare groups of invasive carcinoma cells enter the tumor-associated vasculature (9, 10). Although it is unclear whether clusters of CTCs can transfer from the lymphatic to the blood circulation (4, 5), a mixture of single and clumps of CTCs disseminate throughout the body before they reach a distant organ (10). In the circulation, CTCs face multiple physical constraints that will directly affect successful seeding. Single CTCs need to overcome the mechanical stress imposed by shear forces, likely to induce apoptosis (programmed cell death), before they lodge in a capillary in a distant organ. This mechanical stress can considerably reduce the ability of CTCs to successfully initiate the growth of a metastatic colony. When shed as clumps and traveling as clusters, CTCs are more resistant to shear forces and cytotoxic immune cells (10). They are also more likely to become lodged in microvessels in distant organs before seeding metastatic colonies (10). Indeed, mechanical constraints imposed by vessel architecture and size contribute to the intravascular arrest of CTCs (11), and clusters of CTCs are rapidly trapped in tiny capillaries. Intravital imaging in living mice recently demonstrated that CTC clusters could also form mechanically, as a consequence of the initial arrest of a single CTC (12). Nevertheless, some clusters of CTCs avoid lodging and are capable of squeezing through capillary-sized vessels as a group (13). Such clusters reduce their hydrodynamic resistance by forming single-file structures that rely on intercellular adhesive interactions. However, whether extravasation from blood vessels of clusters of CTCs is more efficient than single CTCs remains to be demonstrated.

In addition to mechanical trapping imposed by vessel architecture and size, single or clusters of CTCs also rely on blood flow, their adhesive potential, and blood components (such as platelets) to arrest successfully at distant sites (2). Indeed, although physical restriction can explain how CTCs become lodged in a capillary bed downstream of the tumor along the anatomical route of the blood circulation, it does not explain the sites of metastasis, which are specific for each cancer type (organotropism). Recent observations demonstrate that both the efficiency and location of CTCs becoming lodged in distant sites correlate with the presence of permissive flow regimes in some vascular regions (2). The transit of CTCs in the blood circulation is stopped when their adhesive capacity overcomes the shear forces imposed by blood flow. In vivo measurements revealed that CTCs and endothelial cells (which line vessels) rapidly engage adhesion forces that exceed 200 pN, where hemodynamic shear forces (≈100 pN) cannot dislodge them (2). Thus, CTCs mostly engage and stabilize adhesions with endothelial cells of vessel walls in regions with low hemodynamic flow. These regions might also favor the sequential formation of intravascular clusters from single CTCs (12). An evaluation of a cohort of 100 patients with brain metastasis demonstrated that metastases preferentially develop in anatomical regions with reduced cerebral blood flow dynamics (2). In addition to limiting survival of CTCs in transit, shear forces also control the location of their final arrest. Combined with the physical constraints imposed by vessel architecture, shear forces imposed by hemodynamic flow are thus essential factors of hematogenous metastasis.

The influence of fluid mechanics on metastasis

From the onset of tumorigenesis, fluid mechanics in blood and lymph drive tumor growth, invasion, and metastatic potential. Although metastatic tumor cells seed draining lymph nodes, metastases mostly form in distant organs via the circulatory system. Blood flow forces control the arrest and recruitment of myeloid cells and the efficiency of extravasation, which precede metastatic outgrowth.


Once lodged in microvessels, CTCs can become fragmented by blood flow, which generates immune-interacting intermediates that promote extravasation and the development of metastases from surviving CTCs (1). Intravital imaging of melanoma cell dissemination to the lungs revealed that shear forces caused the formation of microparticles from tumor cells that were lodged in lung capillaries (1). These tiny fragments, called cytoplasts, are devoid of nuclear components, are only formed in the presence of shear forces, and are not derived from apoptotic events. Cytoplasts that contain mitochondria adhere to vessel walls and display intravascular migration (crawling along the vascular wall), as has been shown for intact CTCs (2), that is favored by hemodynamic flow (1). Soon after they are produced, myeloid cells such as neutrophils, conventional and patrolling monocytes, nonalveolar macrophages, and dendritic cells are sequentially recruited and take up (phagocytose) cytoplasts. Phagocyte populations then compete to either stimulate or restrict metastatic outgrowth. Although most of the cytoplast-ingesting myeloid cells are known to facilitate successful metastases, conventional dendritic cells, although sparse and recruited last by the generation of cytoplasts, provide antimetastatic defense. Shear forces are thus essential for CTC fragmentation, yet the flow regimes that induce this phenomenon, if any, remain to be determined. It also remains unclear how flow-mediated fragmentation of a CTC facilitates the extravasation of its corresponding karyoplast (the parental nucleated cell).

Although it is not clear whether blood flow also stimulates CTC transmigration (active migration through the vascular wall) or protects them from immune recognition, endothelial cells respond to shear flow forces by producing cellular protrusions that are capable of engulfing single or clustered CTCs (2). This flow-dependent process, called endothelial remodeling, drives subsequent extravasation of arrested CTCs and reestablishes blood perfusion in the capillary. Interestingly, endothelial cells of cerebral microvessels similarly remove blood clots, preventing early embolus formation and reestablishing blood flow in vascular occlusive disorders (14).

Multiple challenges remain. For example, further work is needed to address whether hemodynamic shear forces modulate the formation and location of pre-metastatic niches (PMNs). Such niches provide pro-metastatic conditions in tissues that are otherwise hostile to tumor cells and are shaped by tumor-shed factors. Indeed, although several molecular and microenvironmental programs set the basis for metastatic traits and organ-specific tropism of cancer types (6), tumor-shed soluble factors and extracellular vesicles disseminate through body fluids and initiate PMN formation (15). Fluid-borne tumor-derived factors thus also determine metastatic potential of cancer cells, and it is likely that their target location and extravasation is influenced by fluid mechanics, both in the blood and the lymphatic circulation. The development of animal models, imaging approaches, and biomarkers could help clarify the formation of PMNs in patients. Similarly, how tumor dormancy—in which CTCs are lodged in distant sites but do not grow—is influenced by fluid mechanics (and whether dormancy results from specific PMNs) could be investigated. Organotropism could also be influenced by, in addition to other molecular factors, distinct flow profiles and vessel architecture within and between organs. Myeloid cells detect tumor-derived material, which is either transported or shed by fluid flows. A better understanding of how they compete and are recruited at metastatic or dormancy sites by fluid-borne material could lead to the design of antimetastatic immunotherapy. Similarly, perhaps antiangiogenesis compounds that could reduce flow-mediated pro-metastatic endothelial remodeling should be explored. Blood flow profiles, which can be assessed with contrast-enhanced computed tomography or magnetic resonance imaging, might be used to identify future sites of metastasis and could be used in diagnostics. Further understanding of the molecular mechanisms that underlie flow-mediated endothelial remodeling and subsequent extravasation of CTCs could help to better define interventions for metastasis prevention.

References and Notes

Acknowledgments: J.G.G. thanks the Goetz laboratory, F. Winkler, X. Trepat, and H. Peinado for their comments.

Navigate This Article