Understanding Cancer's Spread.

Where it goes, where it grows, and why

As all cancer patients know, the words from a doctor delivering the diagnosis can be devastating. Each patient quickly learns, however, that the tumor may be only the beginning.

In a woman's breast or a man's prostate gland, for instance, the tumor that was detected often can be eradicated with surgery and radiation. Usually more dangerous is the spread--metastasis--of the initial cancer to organs such as bone, brain, or liver. Cancer cells that have traveled to new locations in the body and grown there cause most cancer pain and cancer-related deaths.

Metastasis was first described in 1839 by the French gynecologist Joseph Recamier, and soon thereafter, physicians found that certain cancers were most likely to spread to certain organs. Breast and prostate cancer, for example, move to lymph nodes, bones, lung, and then the liver. Skin cancer tends to spread to the lungs, colon cancer targets the liver, and lung cancer typically moves to the adrenal glands and the brain.

In 1889, Stephen Paget proposed that cancer cells shed from an initial tumor were dispersed randomly throughout the body by the circulatory system. He called these circulating cancer cells "seeds" and proposed that only some seeds fall onto fertile soil--organs where they can grow.

About 30 years later, a researcher named James Ewing encountered Paget's theory by arguing that cancer cells don't spread randomly throughout the circulation in search of fertile soil. Rather, he suggested that circulating cancer cells become trapped in the first small blood vessels, or capillaries, they encounter and then grow in the surrounding organ.

Researchers today are still trying to understand why metastasized cancer cells, which are called metastases, grow where they do. In 1992, Leonard Weiss used autopsy studies of cancer patients and measurements of blood flow to organs to show that, on its own, neither Ewing's theory nor Paget's completely explains the pattern of cancer metastasis.

Weiss found that, supporting Ewing's hypothesis, most metastases seem to form in organs where cancer cells stick soon after entering the bloodstream. Favoring Paget's theory, however, some metastasis patterns can't be predicted by blood flow alone.

Now, researchers are taking these old ideas, combining them in new ways, and looking for molecular explanations of cancer's spread. Many of the new explanations contain elements of the original proposals, though none mirrors either one exactly.

Scientists today don't hold that cancer cells are dispersed randomly throughout the body, so that only growth conditions in different organs determine the distribution of metastases. The cells are guided by a variety of chemical signals, some researchers propose. Others agree with Ewing that cancer cells get stuck in the first capillaries they reach. In either case, some organs may be more likely to support a cancer's growth than others are.

Teasing out the cellular interactions involved in cell movement and growth is a crucial step in combating cancer, most scientists now concur.

If cancer cells can target specific organs, that capability isn't unique. White blood cells do likewise in the body. They follow a trail of signaling molecules called chemokines, which act as address labels produced by each organ in the body. Each white blood cell carries a receptor that guides the cell to areas with high concentrations of a particular chemokine.

"Like white blood cells, tumor cells move into blood and into tissues," says Jonathon Sedgwick, director of immunology at DNAX Research Institute, a biotechnology firm in Palo Alto, Calif. "Scientists here wondered if the tumor cells might be using the same signals."

Anja Muller and Albert Zlotnik of DNAX, in collaboration with the National Cancer Institute in Mexico City, have looked for chemokine receptors in cells from various cancers. Not only are the receptors present, but the researchers have shown that breast cancer cells move toward chemokines made in lymph nodes and in lung tissue--major sites of breast cancer spread. Skin cancer cells, which also spread to the lymph nodes and the lungs, respond to the same signals and also to a chemokine made by skin cells, the researchers reported in the March 1 NATURE (SN: 3/17/01, p. 175).

Muller and Zlotnik have shown that antibodies designed to block interactions between chemokines and their receptors on white blood cells reduce the spread of breast cancer cells to lymph nodes and lung tissue in mice.

"This is proof of principle that chemokines may be important in metastasis," says Sedgwick.

Cancer cells that follow chemical signals to target organs, however, isn't the only possible explanation for the results. Chemokines, which are chemically similar to natural growth factors in the body, could also be feeding cancer's growth in target organs, Muller says.

The DNAX team plans to study whether the antibodies that limit the interactions between chemokines and their receptors have any effect on the growth of metastases after the cancer cells have already spread.

That's an important question, says Lance A. Liotta of the National Cancer Institute in Bethesda, Md. "Imagine that a woman has breast cancer. She usually has the cancer many years before it's diagnosed, and during that time millions of [cancer] cells have probably already reached her circulation." So, a drug that prevents cancer cells from homing in on an organ might not benefit such a patient.

Despite researchers' progress in linking chemokines to metastasis, these might not be the only molecules that bring cells to fertile soil for growth. A variety of chemicals called adhesion molecules, found on the lining of blood vessels, act as hooks to halt circulating cancer cells in their target organs. In the healthy body, these adhesion molecules direct immune function.

Researchers have traditionally believed that many cancers spread to the lungs because the circulatory system delivers much blood there--and with it many circulating cancer cells. Erkki Ruoslahti of the Burnham Institute in La Jolla, Calif., suggests that selective adhesion molecules expressed in the lung may also account for some of these metastases.

In the December 2000 CLINICAL CANCER RESEARCH, Carlton R. Cooper and Kenneth J. Pienta of the University of Michigan Comprehensive Cancer Center in Ann Arbor showed that prostate cancer cells are more likely to adhere to blood vessels feeding bone than to blood vessels feeding other tissues. As-yet-unidentified adhesion molecules make cancer cells stick to the cells lining these blood vessel walls, Pienta and his colleagues suggest.

If cells stick within a blood vessel, how do they enter an organ to cause problems there? At the March meeting of the American Association for Cancer Research in New Orleans, Cooper reported that a cytokine called TGF-beta, which bone releases during the normal process of remodeling itself, actually reduces the adhesion of prostate cancer cells to blood vessel walls.

"We're not sure exactly what is going on," says Cooper, but he suggests that TGF-beta releases the stuck cancer cells to "sneak out into bone, where they then dig in."

"Bone metastases are incredibly painful," notes Cooper, recalling that his granduncle suffered for years from unrelenting bone pain from prostate cancer that had spread. He and his colleagues are working to develop compounds that might interfere with factors that help cancer cells stick in blood vessels and others that, like TGF-beta, encourage metastasis in organs.

Not everyone agrees that specific molecules direct the movement of cancer cells in the same way that they guide immune cells. "People forget that cancer cells are big and they aren't designed to circulate," says Ann F. Chambers of the London Regional Cancer Centre in London, Ontario. "They can't wander through the body until they find an organ they like."

Chambers and her colleagues have directly observed the movements of fluorescently labeled cancer cells in live animals. The researchers have shown that cancer cells generally stop in the first tangle of capillaries that they encounter, apparently when the vessels simply become too small for the cells to pass through. In contrast, the white blood cells circulate freely.

Chambers argues that adhesion molecules, rather than making cancer cells stick to blood vessel walls in particular organs, may help cancer cells slip from the bloodstream into the organ. Alternatively, she suggests, adhesion molecules might play a role in determining whether cancer cells grow in a new site.

In fact, Chambers' studies have convinced her that most cancer cells that get into the circulation and slip through vessel walls then die or fail to grow in their new locations.

Such cells lying in wait may explain why some breast cancer metastases, for example, show up as late as 2 decades after surgery and other treatments eradicated the initial cancer.

Of the cells that do grow, few divide often enough to produce clinically relevant metastases. Therefore, Chambers says, there must be something special about the way particular kinds of cancer react to the environment of different organs.

Deciphering what makes some organs hospitable to cancer cells has been difficult. More is known about cancer cell growth in bone than in any other tissue.

"There seems to be something special about bone that promotes tumor growth after the cancer cells arrive," notes Michael L. Cher of Wayne State University/Karmanos Cancer Institute in Detroit. Both breast and prostate cancer typically spread to bone before they spread to any other organ, he says.

Normal bone remodeling releases a variety of chemicals including TGF-beta and substances called metalloproteinases. At the March cancer meeting, Cher and his colleagues presented data showing that metalloproteinases are also released by prostate cancer cells and increase the rate at which bone remodels itself.

His team's new results are interesting, Cher says, because metalloproteinases had already been identified as agents that help cancer cells escape from their original tumor and enter the bloodstream. They eat away cartilage and other connective tissue around a tumor.

It's possible, Cher says, that metallo-proteinases also help cancer cells form metastases once they've reached bone by boosting the bone-remodeling process.

Other researchers, working with both breast and prostate cancer cells have implicated a hormone called parathyroid hormone-related protein, or PTHrP, in the growth of bone metastases. When made by the cancer cells, PTHrP carries out its normal role of stimulating bone growth. The growing bone then seems to release compounds that speed the division of cancer cells. This communication triggers a vicious cycle in which growing bone and cancer cells feed off each other.

"We're acquiring a wealth of information on why [cancer] cells go to bone and why they grow there," says Patricia S. Steeg of the National Cancer Institute in Bethesda, Md. "But for other cancers and sites of metastasis, we have very little information."

Understanding how cancer interacts with the organs it targets "is important for guiding development of antimetastasis treatments," says Chambers. "Metastases continue to be the primary source of cancer mortality."

A better understanding of the way tumor cells interact with their environment may also help physicians decide when to administer anticancer drugs, she says. Drugs that block the growth of cancer cells that have moved to a target organ would be potentially effective any time after a diagnosis. Drugs that block the homing of cancer cells to particular organs might have more limited use.

Although cancer treatments that block these interactions seem to offer great promise, a recent study alarmed some researchers in the field.

In the February CANCER RESEARCH, a team of German researchers reported a study of batimastat--a drug developed to block metalloproteinases and thereby prevent primary tumors from spreading into surrounding tissues. They found that, at least in mice with breast cancer, the drug increases the number of tumors forming in the liver while decreasing the number of metastases in the lungs.

"We've never seen these liver metastases before," Steeg says. "That implies [that batimastat] is tickling normal liver ... so it becomes extremely fertile ground." Other drugs might also alter the interaction between tumor cells and particular organs and cause unexpected patterns of metastasis, she concludes.

The German study and many others demonstrate that, however cancer cells reach their target organs, interactions between tumor cells and their immediate environment play a critical role in determining whether and where metastases will be present.

"I would predict ... that we may be able to design specific therapies to interrupt this process," Steeg says. "It's a vitally important area of research."