For decades, cancer biologists have relied on so-called lines of cancer cells for their experiments. But these cultured cells often bear little resemblance to the tumor they came from. That’s because a piece of tumor tissue dropped into a petri dish doesn’t just start growing. Instead, researchers pull out a few cells in the tumor that happen to replicate well—often cells that don’t need the surrounding normal cells that nurture tumors inside the body. And the genetic makeup of cell lines can change over the years they multiply in labs. No wonder, then, that an experimental drug that kills a colon cancer cell line won’t necessarily help a patient with colon cancer.
Now, several U.S. and European funding agencies want to change that. Today, they’re launching the Human Cancer Models Initiative (HCMI), which aims to give the research community tumor cells that behave more like actual human tumors. The project involves four groups: the U.S. National Cancer Institute (NCI) in Bethesda, Maryland; Cancer Research UK in London; the Wellcome Trust Sanger Institute in Hinxton, U.K.; and the nonprofit Hubrecht Organoid Technology in Utrecht, the Netherlands, which was founded by Hans Clevers, a cancer researcher at the Hubrecht Institute.
The project will draw on new insights into how to make the mixture of cells from a human tumor grow outside the body. For example, Clevers adds certain growth factors and a gellike matrix to get cells isolated from a particular organ to grow into a similar miniorgan, or organoid. Others use a special bed of mouse cells to coax cells from tumors into growing. When samples of such cells were treated with known cancer drugs, they responded in a way that was remarkably similar to that of mice with tumors grown from these same cells.
HCMI will scale up production of these tissue-based human cancer models and share them with the community. NCI will fund the development of 600 models; the Sanger Institute and Cancer Research UK will create 200; and the Hubrecht Institute will produce 200 models as part of a 2- to 3-year pilot project. (The total funding level hasn’t yet been determined.) Although the focus will be largely on common cancers, such as colon and pancreatic, NCI will try to include rare and childhood cancers too, says Louis Staudt, director of NCI’s Center for Cancer Genomics.
Each model will come with a complete genetic analysis and clinical information about the cancer patient it came from, such as whether a specific drug helped them. The project will dovetail with an NCI effort to supplement its 60 workhorse human cancer cell lines with cells derived from patient tissue samples implanted in mice, Staudt says.
The first samples from the repository could be available to researchers this year. In the long run, says Staudt, he would like to see the project expand to 10,000 models. “I'm extremely excited about this. I think this is big opportunity to boost cancer research. It’s a chance to raise all boats.”