One major obstacle is the shortage of predictive human models that allow scientists to watch the process unfold and identify what makes particular organs vulnerable. Animal research has revealed much about how cancer spreads, but rodent biology cannot fully reproduce what happens in patients. Those differences may help explain why treatments that appear promising in laboratory animals often disappoint in clinical trials.
Now in a new study led by Columbia Engineering professor, Gordana VunjakNovakovic, and published in Science Translational Medicine, the researchers found first model of its kind, the chip containing separate compartments with millimeter-scale bone and lung tissues connected by a flowing vascular channel. This design does more than place human tissues beside cancer cells. It allows researchers to observe the continuous exchange of signals between cells in the bloodstream and the organs they may eventually colonise.
“The pressing need for developing human tissue models of metastasis has been a key motivation for our study,” said Vunjak-Novakovic, who is a university professor at Columbia University and the Mikati Foundation professor of Biomedical Engineering and professor of Medical Sciences (in Medicine). “Our objective was to probe the ability of cancer cells to adhere to and traverse across endothelium [inner lining of blood vessels] and to determine their capacity to survive in the tissues they are colonising through cell reprogramming and niche remodeling,” VunjakNovakovic said. Reaching another organ is only the beginning of metastasis.
Cancer cells must attach to the blood vessel lining, cross that barrier, evade local defenses, and adapt to unfamiliar tissue. Only then can they multiply and form a secondary tumor. This phase, known as organ colonisation, is exceptionally difficult to observe directly in patients or reproduce accurately in animals. Metastasis is also selective rather than random. Different cancer cells tend to favor particular organs, a pattern known as “organ tropism.” Breast cancer commonly spreads to sites, including bone and lung, making those tissues especially useful for testing whether the chip could reproduce behaviour seen in the human body. The platform also accepts patient cells and engineered human tissues, giving scientists a way to investigate individual differences in metastatic behavior.
Researchers can alter one component at a time, examine organ-specific interactions, and search for molecular pathways that could become new treatment targets. “Cancer is very smart, unfortunately. We learned how the cells cross barriers to get from blood circulation into the tissues. “We were also able to reproduce something that happens in patients, where cancer cells condition the target tissues, even before they colonise them, to make them more receptive,” said VunjakNovakovic.
The team produced bone, lung, and vascular endothelium from induced pluripotent stem cells (iPSCs), which can be guided to develop into many specialised cell types. Tissue-specific scaffolds and bioreactors helped each tissue mature, while separate compartments provided the conditions needed to preserve its function over time. Vascular circulation then linked the compartments. A selectively permeable endothelial barrier separated the flowing channel from the organ tissues, just as blood vessel walls do inside the body.
The researchers introduced breast cancer cells into the circulation and watched as they encountered the barrier, crossed into nearby tissue, and established distinct patterns of colonisation. The results closely reflected the cells’ known preferences. Breast cancer cells that typically spread to bone colonized the engineered bone more strongly and produced more extensive degeneration. Cells with an affinity for lung tissue caused greater damage there while colonising the bone only modestly.
Differences in tissue invasion and secreted molecules showed that the device captured key features of organ-specific metastasis. The chip also reproduced a more insidious feature of cancer spread. Before metastatic cells fully occupy a distant organ, they can release signals that alter its environment and make it easier to invade. Scientists call this preparation of future tumor sites pre-metastatic niche formation.
Analysis of both tissue compartments revealed evidence that the breast cancer cells were conditioning distant organs before colonisation. Recreating this early transformation could help researchers investigate opportunities to interrupt metastasis before a secondary tumor becomes established. “The key advantages of this advanced model of metastasis are that it is human and can be patient-specific,” VunjakNovakovic said, adding that “it faithfully mimics some of the key aspects of human metastasis that are otherwise largely inaccessible for direct study.”
