Dormant cancer cells found hiding behind protective cell shields

Cancer cells can leave the organ where they first formed and travel to places like the bone marrow, liver, lungs, or brain, where some of them stop dividing and sit inactive for months, years, or even decades. Researchers call this state dormancy or quiescence. Because most existing cancer treatments are thought to only kill cells that are actively dividing, these sleeping cells can survive chemotherapy and later wake up to form new tumors, a process called metastatic relapse [1,5].

A researcher at the Fred Hutchinson Cancer Research Center described these cells as living "a bit like a time bomb," since there is no way to know when they will start growing again [1]. A separate review notes that metastatic relapse is thought to account for as much as 90% of cancer deaths, and that only a very small fraction of cells that spread through the body, roughly 0.02%, ever go on to form an overt tumor; most instead die off or fall dormant [5]. Late relapse, sometimes appearing years after treatment that seemed to have worked, remains the main cause of death in breast cancer specifically [2].

Several lines of research now point to the same basic idea: dormant cells are not surviving on their own, but with help from the normal cells around them, known as the tumor microenvironment. In mice, a team led by Cyrus Ghajar at Fred Hutchinson found that breast cancer cells that had spread to the bone tended to cluster around blood vessels, in a region called the perivascular niche. The cells lining those blood vessels, called endothelial cells, appeared to shield the tumor cells from three types of chemotherapy and one targeted therapy tested in the study, acting through proteins called integrins that normally help with blood vessel repair and wound healing [1].

A more recent mapping study from the MRC Laboratory of Medical Sciences, Imperial College London, and UCL found a similar pattern in human breast tumor tissue, using published data combined with single-cell and spatial gene-sequencing techniques that show which genes individual cells are using and where those cells sit within the tumor. Dormant cells were consistently found near two other cell types: a kind of immune cell called CXCL10-positive macrophages, and a kind of tumor-supporting cell called myofibroblastic cancer-associated fibroblasts. The researchers describe these surrounding cells as potentially forming a shield around the dormant cells, though they are careful to say they do not yet know whether the surrounding cells push cancer cells into dormancy, or whether the cancer cells recruit and alter their surroundings, or both [3].

Separately, work at Virginia Commonwealth University looked inside the dormant cells themselves rather than at their neighbors. In mouse and human breast cancer cell lines, chemotherapy pushed some cells into a dormant, drug-tolerant state that depended on a protein called Bcl-xL, part of a family of proteins that normally block a cell's internal self-destruct pathway. Blocking Bcl-xL, either by genetically reducing it or with a drug called A-1331852, was tested alongside low-dose chemotherapy in mice carrying triple-negative and HER2-positive breast tumors, two subtypes that behave differently in terms of drug sensitivity and relapse patterns [2].

When Ghajar's team combined chemotherapy with antibodies designed to block the protective integrins, they killed most of the dormant tumor cells hiding in their laboratory model of the perivascular niche, whereas chemotherapy or the antibodies alone did not work as well. In mice implanted with human breast cancer cells, chemotherapy alone led to bone metastases in almost 75% of animals; adding the anti-integrin antibody beforehand cut that to 22%. No side effects beyond those normally seen with chemotherapy were observed when the antibody was added, which the researchers called an unexpected and encouraging finding [1].

The MRC-led mapping study points toward a different kind of opportunity: it found increased activity of the complement pathway, part of the immune system, specifically within the dormant regions of tumors, suggesting these areas might be vulnerable to treatments that target that pathway, though the researchers stress this still needs to be tested experimentally [3]. A cancer biologist not involved in the Ghajar study, from the National Cancer Institute, noted a further reason this general approach is attractive: a treatment that kills dormant cells without first waking them up avoids the risk of leaving behind a population of cells that has been prompted to start growing again [1].

Each of these findings comes with real limits. The integrin-blocking work has so far only been done in mice and in laboratory cell cultures that mimic the bone environment; the researchers cautioned there is a long way to go before a version of the antibody suitable for people could even be tested, and they are still trying to understand why about a quarter of treated mice developed bone metastases anyway [1]. The Bcl-xL work found that a drug blocking that protein caused serious toxicity when given by mouth or into the vein, forcing the researchers to switch to injecting it under the skin near the tumor to avoid harming the animals, an issue that would need to be resolved before any human use [2].

The tumor-mapping study is explicitly described as generating predictions that still need to be tested experimentally, since it was built from existing datasets rather than new treatment trials, and the researchers do not yet know the direction of cause and effect between dormant cells and their protective neighbors [3]. A broader review of the tumor microenvironment adds that the surrounding tissue can have both tumor-suppressing and tumor-promoting effects depending on context, which is part of why deliberately disrupting it is described as a challenging undertaking rather than a straightforward one [4]. Taken together, the sources describe converging evidence for a survival mechanism behind dormancy, but each piece is at an early, largely preclinical stage.