Originally Published:20091001.

Last June, researchers announced that patients with advanced colorectal cancer with a particular genetic mutation in their tumors should not receive two federally approved drugs as part of their treatment regimens. Studies had found that the costly medications weren't making a dent in the disease.

Less than six months later, an alliance of leading cancer centers added genetic testing of rumors for the mutation, known as KRAS, to its colorectal cancer treatment guidelines. In a study of one of the drugs, researchers found that targeting the medication only to patients who would benefit from it would not only spare thousands of patients from useless therapy but also would save more than $600 million a year.

Testing for the KRAS mutation became a standard of care almost overnight," says George Demetri, M. D., director of the Center for Sarcoma and Bone Oncology at the Dana-Farber Cancer Institute in Boston. "It's important for hospitals and our health care system because that information can be absolutely critical so we don't waste money, so that patients don't have false expectations and so mat drugs can he developed more cleverly."

Welcome to the fast-moving world of cancer genetics. It's a landscape where new discoveries are pushing personalized cancer treatment forward, where hospitals are grappling with how to adapt their oncology and pathology departments to keep up, and where genetic findings are changing researchers' and drugmakers' approach to clinical trials.

Cancer care always has been somewhat personalized because oncologists have long known that every cancer behaves differently and people react differently to drugs, explains Richard L Schilsky, M.D., immediate past president of the American Society of Clinical Oncology. What has changed in recent years is the depth of the molecular understanding thanks to advances in technology that enable researchers to peer into cancers' genes. What they're discovering are specific mutations that drive cancers.

These findings not only have allowed researchers to break cancer into smaller and smaller subtypes by genetic mutation, but also have spurred the development of new agents that shut down those mutations and slow or stop the disease.

"With the advent of the more targeted therapies, which are replacing chemotherapy in some diseases, we're really now at a point where we can study the tumor in depth, try to figure out what is the optimal targeted approach to that patient's tumor, and hopefully give them a therapy that would be effective and have minimal side effects," Schilsky says.

Hospitals that want to remain competitive in oncology need to make sure they offer targeted therapy, says Sarah Brandt, a specialist in cancer services at the consulting firm Sg2 in suburban Chicago. But each hospital will have to decide how far it will go in its personalized cancer care effort, several experts say.

"The first thing you'll see is different cancer centers across the country will be early adopters, midstage adopters or later adopters, and then the community hospitals will try to figure out, 'Do we do this ourselves or do we just work with a reference center close by?'" Demetri says.

Who Do You Test?

At Dana-Earber, hospital leaders are trying to figure out where their approach will fall on the spectrum, Demetri says. A philosophical debate is going on within the institution because at this point there are more known cancer-driving mutations than there are Food and Drug Administration-approved therapies to attack them.

On one side, people argue that genetic tests should be part of the standard cancer patient workup even when no approved drug is on the market The other camp says the tests should remain in the research realm unless an approved drug is available. "What we're trying to do at Dana-Farber is take a solid, judicious, positive approach and say that this is absolutely important to our field and we have to develop it rationally, but we cannot in the process over-promise to the public," Demetri says. "Otherwise, we'll run the risk that [patients] say, 'Gee, you told me everything about my tumor, but it didn't make a dam bit of difference.'"

Demetri says he falls on the "more enthusiastic side." He points to preliminary findings in melanoma research indicating that an agent targeting a mutation called. BRAF can shrink tumors. "If I were a patient with melanoma, I would dam well want to know if my tumor had that BRAF mutation."

Massachusetts General Hospital in Boston has taken a test-for-everything approach. The hospital spent two years developing a method to extract and test DNA from tumors collected through routine biopsies. Currently, the tumors of patients with lung cancer and certain other cancers are screened for 120 abnormalities on 13 known cancer genes.

The hospital began patient testing this spring and expects to be able to run the screening on all cancer types by the end of the year, says Leif Ellisen, MD., associate professor of medicine for Harvard Medical School and co-director of the Translational Research Lab.

The oncology field is struggling to figure out the threshold at which tumors should be tested for mutations, Ellisen says. Some cancer specialists argue for testing patients only for the most common mutations in their type of cancer, and some hospitals may go that route, he says.

Massachusetts General decided to test all cancers for all well-established cancer-driving mutations not only because some effective, approved therapies exist, but also because new agents are available through clinical trials. Only by testing patients can you match them with a clinical trial, Ellisen notes.

Plus, more and more targeted therapies will gain FDA approval in the coming years, he says. "Today's tumor mutation with no approved agent is tomorrow's 'you must test everybody' because there is a really great drug out there that is going to have a benefit for that patient."

Ellisen also notes that researchers are finding mutations common in one cancer are present at a lower frequency in other cancers. Studies are beginning to show that targeted therapies can have a beneficial impact in patients with a low frequency of the mutation. He points to the example of a patient with advanced lung cancer who had run out of FDA-approved treatment options. Testing at Massachusetts General showed that her lung cancer had a mutation called ALK, usually seen in a type of lymphoma. She entered a clinical trial at the hospital targeting that mutation, and her tumor has since disappeared, Ellisen says.

The ALK mutation is present in 5 percent or less of lung cancers, Ellisen says.

"If you think about the impact on this individual person's life, it's quite significant That's why we have the belief that the most effective way is to test for all these things in all these tumors because if you're one of that 5 percent, it makes a difference," he says.

Advances in the tumor genetics field also are starting to change the way clinical trials are done.

Ellisen uses the BRAF mutation as an example. The mutation made a splash in 2002 when its role in melanoma was discovered, and drug companies began to develop drugs to block the genetic defect.

The drugs blocking the mutation are still being tested, but results from one clinical trial found that patients lived for a median of at least six months without the disease getting worse; some even saw their tumors shrink.

The mutation also can be found in other types of cancer. Massachusetts General has partnered with a pharmaceutical company, which has a targeted drug in clinical trial for melanoma, to start a different trial to test the same agent in patients with BRAF mutations in other cancers.

"So instead of just testing random patients with a random inhibitor, we're saving we want to test anybody with any tumor that has that specific genetic abnormality with an inhibitor that targets that abnormality," Ellisen says.

Community Hospitals in the Mix

Although the nation's major cancer centers are on the front lines of the push toward personalized cancer care, community hospitals with oncology programs also need to develop strategies for offering these targeted therapies, cancer experts say.

"Having a plan for what a hospital or health network is going to do in cancer really requires an educated group of people having a strategic plan, often with external guidance from experts in the field," Demetri says. "You can either be overly optimistic and get conned quickly into wasting money or you can be overly negative and get left in the dust"

Testing for genetic mutations in tumors often requires sophisticated, expensive and rapidly evolving technology. Hospitals have to decide whether they'll invest in the technologies and if so, which ones. One practical consideration is whether enough volume exists to make the investment worthwhile, Brandt says.

Many facilities will likely opt to send the most sophisticated tests out to a reference lab. Quality control considerations and economies of scale argue for that approach, Demetri says, adding that it would avoid unnecessary duplication of technology among hospitals.

One big wild card is whether the technology will advance to a point at which if s cheap enough for any hospital to buy. Demetri sees a comparison with computers.

"Nobody thought everybody would have their own computer in their house," he says. "If we really see the technology develop in an exponential way, I could imagine that a lot of these things could be little boxes sitting on counters in pathology departments in the middle of tiny little towns."

Building the Right Team

As advances in genetic testing technologies and new discoveries push targeted therapies forward, hospital leaders no doubt will be inundated with information and vendors will pressure CEOs to buy their machines. Oncologists and pathologists will weigh in with their views on whether to take an aggressive or conservative approach to tumor genetic testing. Pathologists will have views on whether to test in-house.

"CEOs are going to need to integrate multiple lines of evidence and judgment into a plan that fits," Demetri says.

He suggests that hospitals assemble a multidisciplinary panel to work through the issues and arrive at informed decisions. "This is not just the purview of the pathologist, this is not just the purview of the surgeons or the medical oncologists or the radiologists. This is going to be a bit of a team effort."

To be competitive, hospitals will have to make sure they have a strong oncology team, including medical oncologists, Brandt says. Traditionally, hospitals have had an easier time forming alliances with surgical and radiation oncologists, while medical oncologists have been out in the community, she says.

Changes in Medicare payment for chemotherapy drugs over the past several years have caused a dip in medical oncologists' profits. As a result practices are merging and some are more open to partnering with hospitals or forming employment arrangements.

"The benefit to being independent just isn't there anymore," Brandt says.

Because medical oncologists often order the genetic testing on tumors and prescribe cancer medications, hospitals will want to make sure that the physicians with whom they have a relationship are well-versed in the latest tumor testing, FDA-approved targeted therapies and clinical trials.

One of the challenges will be physician education. "We have a huge educational mission now trying to make sure this molecular revolution doesn't leave practicing physicians behind because then they won't know how to use these tests judiciously," Demetri says.

Another issue is the shortage of oncologists that is expected to hit in the next decade. A 2007 ASCO study estimates that demands for visits will jump by 4$ percent by 2020, but the number of oncologists will fall 4,000 short. One shortterm solution is to hire more midlevel providers to ease physicians' workloads, while the longterm solution is to train more oncologists, Schilsky says.

Meanwhile, oncologists will have to take more time with patients, Demetri says. They may find themselves facing patients with reams of information from the Internet and have to explain the "difference between science fiction and science fact, and the difference between reasonable current state of the art and the rapid evolution where things can literally change overnight."

There won't be one cure for any cancer, Demetri says. "I would much rather say to the public, little by little every year if we can come up with some new test that is absolutely critical, this is the way the war on cancer is going to be won. We're going to pick it off one molecular subtype at a time, one drug at a time." - Geri Aston is a freelance writer in Chicago.