METASTASIS IS THE CLINICAL NAME FOR THE PROCESS in which cancer cells leave the primary tumor, enter the bloodstream, and take root in distant organs. Because metastasis causes the vast majority of cancer deaths, discovering the mechanics of where and how cancer cells travel could open new routes for treatment.

A study published in the journal Cell in 2014 began to address an important part of the “how.” For decades, cancer cells have been observed traveling in groups: clusters. The 2014 study demonstrated the importance of these groups. Breast and prostate cancer patients with clusters of cancer cells in their bloodstreams were more likely to develop metastases than those with only single cancer cells. In subsequent experiments, the same research team, led by Daniel Haber, director of the Massachusetts General Hospital Cancer Center, showed that successful clusters were made up of genetically different cells and that they tended to leave the site of the original tumor as a group, rather than forming in the bloodstream.

Another team, led by Andrew Ewald, an associate professor at Johns Hopkins University School of Medicine, had been working on the cellular and molecular mechanisms driving these clusters. In an article published in Proceedings of the National Academy of Sciences in February, they noted that when breast cancer cells were studied in a Petri dish, almost all single cells died, while cells that moved into clusters were estimated to be 100 times more likely to form successful colonies.

Haber and his team also found a striking difference between cell clusters and single circulating cancer cells: in particular, expression of the gene for a protein called plakoglobin was 200 times higher in the clusters. “Plakoglobin is like glue that keeps cells together,” Haber explains. “If you remove it in a mouse, the clusters fall apart—and when the clusters fall apart, you just don’t get lung metastases.” Ewald’s team had found that a similar protein, keratin 14, played a key role in the cancer cell clusters they studied.

This makes keratin 14 or plakoglobin a potential target for treatment, albeit a tricky one. “The clusters of cancer cells travel only for a few minutes, but that gives you a window to try and do something,” Haber says.

One aspect of these clusters posed a physiological conundrum. Wouldn’t single cells be smaller and nimbler? Much of the body is accessible only through capillaries, tiny passages that are roughly the width of a single cell. “The dominant idea in the literature has been that clusters are simply too big to make it through capillaries,” says Mehmet Toner, a biomedical engineer at Massachusetts General Hospital.

So Toner, along with Haber and other researchers, took a closer look. Toner used micro-etching techniques, commonly used to make tiny computer chips, to create synthetic capillaries in the laboratory. The researchers then introduced cell clusters and observed them through a high-precision microscope.

They were astonished by what they saw. Clusters arriving at a capillary would form a single-file line of cells to travel through the capillary and then reassemble into a cluster once they had made it through. “We realized that the clusters do go quite easily through these capillaries—clusters as big as 10, 20 or 30 cells,” Toner says. The team later studied the clusters in a strain of zebrafish that has transparent skin so they could watch the process in a living organism. The team will publish their results this month in PNAS.

All of these findings are helping physicians rethink cancer treatment. Knowing that metastases often come from a cluster of many different kinds of cancer cells is hardly welcome news. The more heterogeneous the clusters, the more difficult it will be to find a drug that can eliminate all of them.

But “it’s not all doom and gloom,” Ewald says. If you can attack molecular features common to traveling clusters of cancer cells—the proteins that hold them together, for example—that might put new chinks in cancer’s armor. “We could potentially use this to target all the metastatic cells,” he says. Targeting clusters on the move through the proteins that hold them together is a different approach from standard therapies, which focus on attacking rapidly proliferating cells in tumors.

There’s a lot yet to understand about metastasis, but the hope is that once researchers learn more about cancer’s travel plans—the when, where and how—they can disrupt its devastating itinerary.