Byline: Sharon Begley
What treatable tumors can teach us about improving the odds in the deadliest cases.
Given a choice, no one would opt to get cancer. And it's cruelly insensitive to tell patients how fortunate they are to have a particular cancer. Yet there is no question that the 33-year-old man who walked into oncologist George Bosl's office at Memorial Sloan-Kettering Cancer Center in New York in 2001 was lucky. He had testicular cancer.
He was lucky because the vast majority of such men are cured, sometimes with surgery alone, sometimes with radiation or chemotherapy as well. By "cured," we don't mean that a patient has cancer cells scattered throughout his body that will need to be kept in check by a lifetime of chemotherapy. We mean cured: the cancer is gone. Even men whose testicular cancer has metastasized--in the case of Bosl's patient, it had spread to the lungs and abdomen--have at least a 70 percent chance of being rid of their cancer forever, which is what this man has every reason to expect: nine years after testicular surgery, 12 weeks of treatment with the chemotherapy agents cisplatin and etoposide (both of which are decades old, not new miracle drugs), and an operation to remove the metastases in his lungs and abdomen, he remains free of cancer. "No matter how widespread testicular cancer is at the time of diagnosis, a patient has some chance of being cured," says Bosl. "The cure rate for men diagnosed in 2010 will be 90 to 95 percent," and even if a patient relapses twice, which usually means the cancer has returned in a form resistant to treatment, he has a fighting chance.
"Cured" is not a word you hear much from oncologists. Indeed, the hoary phrase "a cure for cancer" now sounds bitterly ironic, since scientists discovered no such thing after President Nixon declared war on the disease in 1971. Metastatic melanoma, lung cancer, pancreatic cancer, and esophageal cancer are often death sentences, with the result that cancer will kill 569,490 people in the United States this year, projects the American Cancer Society. But there is a glimmer of hope in this bleak picture.
Some cancers are curable. Almost 90 percent of children with the most common form of pediatric leukemia will be cured; women whose estrogen-receptor-positive cancer responds to tamoxifen or aromatase inhibitors will often be truly rid of their disease, as will women with HER2-positive breast cancer, which responds to herceptin. Other cancers are treatable, in the sense that although patients have to take drugs for the rest of their lives (as diabetics must take insulin forever), at least they're alive and healthy. The best example of a treatable cancer is CML (chronic myelogenous leukemia), which can be held in check by Gleevec and related drugs. And now, says Bosl, cancer researchers are asking, "Can we use information about what makes some cancers curable to design treatments for the others?" The answer is an emphatic yes, says his Sloan-Kettering colleague Charles Sawyers, whose research was instrumental in developing Gleevec: "I feel like I've seen the future."
Pediatric oncologists saw it first. Although the search for a cancer cure has long centered on discovering new drugs, astonishing advances in pediatric leukemia have come with drugs that were developed from the 1950s to the 1970s. In the 1960s, the cure rate for ALL (acute lymphoblastic leukemia, the most common childhood form) was 5 percent, says oncologist Ching-Hon Pui of St. Jude Children's Research Hospital in Memphis. The reason it is now 90 percent lies in how those drugs are given and how patients are cared for.
Put simply, the kids are blasted with the highest doses of the most chemotherapy drugs they can stand--steroids and vinca alkaloids, asparaginase and anthracyclines and more. They are almost all treated at academic medical centers that specialize in pediatric cancer, not at community hospitals that set broken bones. At the former, oncologists take samples of leukemic cells from bone marrow to identify the genetic abnormalities, customizing treatment to the specific form of ALL a child has. "The lesson is that you need to precisely characterize the cancer cells," says Pui. "We apply personalized therapy to these kids; we don't treat them all the same." And unlike adults who might miss a chemo or radiation treatment, kids are 100 percent compliant. "You can damn well bet that Mom will get that kid to the chemo appointment and make sure he takes his pills," says oncologist Daniel DeAngelo of the Dana-Farber Cancer Institute in Boston.
All this make a difference. Adults with ALL--the same disease as in kids--are less likely to be cured (only 30 to 40 percent survive five years), in large part because they are not treated with the same intensive chemo and do not get the full-bore supportive care that children do to keep them from dying of complications, says oncologist James Downing of St. Jude. The life-or-death importance of that approach shows up starkly in adolescents and young adults. Some 16 to 20-year-olds are treated by a pediatric oncologist; others, by an adult oncologist. Seven years after diagnosis, scientists led by Wendy Stock of the University of Chicago Cancer Research Center reported in a 2008 study, 67 percent of those treated by a pediatric oncologist were alive; 46 percent of those treated as adults were. "They got the same drugs, but the doses were higher," says DeAngelo. "It shows what you can do with intense dosing."
How intense? When a 12-year-old boy from Illinois with ALL arrived at St. Jude in 2001, Pui administered everything but the kitchen sink: the drugs prednisone, vincristine, daunorubicin, and asparaginase, followed by cyclophosphamide, cytarabine, and mercaptopurine. Six weeks later, he gave the boy four courses of high-dose methotrexate, daily mercaptopurine, and four triple-drug treatments into the spinal fluid. For the next 20 weeks he got five more drugs, followed by two and a half years of treatment with three rotating drug pairs. It worked. The boy has been in remission ever since, and just started law school.
"In the past, we used one drug at a time," says Pui. "When one failed, we moved to another. But we learned that that just induces resistance, so now we hit the cancer with many drugs simultaneously. And we also used to use lower doses. But that also causes resistance to develop." High doses are less likely to. Now oncologists are studying whether slamming a cancer with multiple chemo agents will have the same success in adults, including preventing the development of resistance.
The other big object lesson in curing cancer comes from Gleevec, which is singlehandedly responsible for increasing the number of CML patients who survive at least eight years, from 20 percent in the past to 80 percent today. Gleevec must be taken forever, and so in that sense is a treatment more than a cure. (About 5 percent of CML patients per year develop resistance to Gleevec, but scientists figured out why, and developed two similar drugs that take the place of Gleevec when that happens.) Gleevec targets the single genetic change that causes CML, a "translocation" in which two pieces of DNA swap places and leave one gene stuck in the "on" position. The stuck gene makes a molecule called a kinase that, after a cascade of biochemical reactions, tells cells to divide or proliferate. Gleevec locks onto the kinase, incapacitating it. "The future," says Sawyers, "is about identifying such mutations in a tumor so we can offer individualized therapy. The lesson we learned from Gleevec should apply to other cancers that depend on kinases, including lung and melanoma."
That lesson--target the cancer-causing mutation--comes with a corollary, however. It is crucial to target not just any old mutation in a tumor, which can have hundreds, but to disable what are called "driver" mutations. Those are DNA changes that cause the malignancy, rather than "passenger" mutations, which are just along for the ride. It is also crucial that the cancer cells either not have a backup plan--that is, another way to proliferate out of control--or that drugs cripple that pathway, too. "To kill the cancer, you not only need a driver mutation, but the cancer cell also has to be truly addicted to it," says oncologist David Weinstock of Dana-Farber. Since 2008, he and colleagues have been searching for the driver mutations in individual tumor specimens. "We're asking, is there something the cancer is addicted to, like the translocation in CML?" he says.
It's the kind of high-risk, huge-payoff quest that the National Institutes of Health rarely funds; Weinstock got a $750,000 grant from Stand Up to Cancer, the two-year-old entertainment-industry initiative that has allotted $83.5 million for innovative research. (Its next star-packed telethon airs on Sept. 10 in 195 countries.) Weinstock started with cases of adult ALL, in which no such Achilles' heel had been discovered, and hit pay dirt: an obscure gene called CRLF2 drives some 10 percent of ALLs. His team has identified drugs that target the deadly, cancer-driving product of this gene, and plans to launch clinical trials next year.
The recognition that different tumors are powered by, and even addicted to, specific mutations is triggering a revolutionary change in how cancers are classified and treated. Treatment will be based not on the organ where the cancer began, such as the breast or colon or lung, but on the driver mutation--with luck, its Achilles' heel. For instance, lung cancer is actually many different diseases with different molecular drivers, explains Sawyers. The alphabet soup of drivers includes EGFR, BRAF, MEK, and HER2. The good news is that all these are kinases, like Gleevec's target. He estimates that some 200 drugs targeting driver mutations are in the pipeline. The bad news is "we don't routinely profile lung tumors to identify the genetic alteration that's driving them," says Sawyers. "It should be malpractice not to genotype cancer patients."
Genotyping is not even routine in the clinical trials that test experimental drugs. As a result, says Sawyers, "the development of kinase inhibitors for lung cancer has been a 10-year saga of missteps." For instance, a drug that cripples EGFR cancers was tested on patients whose cancer was not driven by that mutation. Although one EGFR inhibitor, Iressa, was approved in 2003, it was a close call, and the drug was almost pulled from the market when it failed to help most lung-cancer patients--those without the EGFR mutation, who should never have gotten it in the first place. The lesson of Gleevec and herceptin--identify the driver mutation and cripple it--is still being ignored, keeping what could be effective drugs from reaching the market. "You would be shocked at how primitive the molecular characterization of cancer is," says cancer researcher Tyler Jacks of MIT. "Companies run big dumb trials" rather than test drugs only on patients whose cancer is driven by the mutation the drug targets. "There has to be a complete reassessment of how we do this."
That's finally happening, with the result that drug companies are seeing hints of success against some of the worst cancers. Metastatic melanoma is one. It seems to shrug off standard chemotherapy, repairing the DNA damage the drugs cause and continuing to multiply. As a result, half the patients with metastatic melanoma--which has seen no therapeutic advances in the last 20 years--die within six months. But in 2002, scientists discovered that about half of melanomas are driven by a mutation in a gene called BRAF (pronounced bee-raff). A biotechnology company, Plexxikon, developed an oral drug that targets the product of that mutation. In a study published last month in The New England Journal of Medicine, scientists reported that the drug shrank the metastases (in bone, liver, and bowel) of 80 percent of patients with metastatic melanoma; in two of 32 patients, the cancer vanished. An 80 percent response rate in a solid tumor is unheard of. If the drug is approved by the FDA, it might do for metastatic melanoma what Gleevec did for CML.
Unfortunately, most of the patients relapse within a year. It's not clear why. Maybe other mutations kick in, keeping the drug from binding to BRAF, or the melanoma switches to a different driver mutation. If so, scientists will have to find backup drugs to keep the cancer in check after it becomes resistant to a drug. With CML, they have: the drugs Sprycel and Tasigna attack leukemic cells that become resistant to Gleevec. With melanoma, they're trying: a clinical trial now underway is testing a one-two punch: a drug that targets BRAF and one that targets a backup pathway. With pancreatic cancer, it may be a lost cause. These tumors are so rife with mutations, says Daniel Von Hoff of the Translational Genomics Research Institute, that "24 patients have 63 genetic changes, and the cancer probably has six or seven different drivers." (He is working on attacking pancreatic cancer by cutting off its fuel supply instead.) With other cancers, scientists haven't found these backup drugs, with the result that when lung and other recalcitrant cancers are hit with one drug they switch drivers as smoothly as a flight from L.A. to Sydney switches pilots over the Pacific. And go on proliferating as insidiously as ever.
In these cases, oncologists believe that combinations of several kinase-targeting drugs, covering all known resistance mutations and backup pathways, "could shut off all mechanisms of escape," says Sawyers. That would be applying the lesson of pediatric leukemia, where success comes from hitting a cancer with everything you've got, upfront, before resistant cells run amok. In that vein, oncologist Dennis Slamon of UCLA's Jonsson Comprehensive Cancer Center is studying whether treating breast cancers with herceptin (which is based on his discoveries), as well as a drug targeting a pathway that makes tumors resistant to herceptin, might save more women. "I don't think we'll be treating patients with 12 drugs," he says, "but it might be three. It's going to be hand-to-hand combat."
Kinase hunters are also seeing glimmers of hope against a form of lung cancer. In 2007 Japanese researchers discovered that about 3 to 5 percent of lung cancers are driven by a mutation in the ALK gene. An oral drug called crizotinib, from Pfizer, shrinks or stabilizes the tumor in 90 percent of patients, the company reported in June. Pfizer hopes to submit the drug for FDA approval next year. The success of another kinase-targeting drug "spurred us to think there are a lot more kinase mutations in cancers that haven't been found, especially in solid tumors," says William Pao of Vanderbilt University, who is using a $750,000 Stand Up to Cancer grant to search for them.
At a recent biology meeting, Jim Watson, codiscoverer of the double-helix structure of DNA, chided MIT's Jacks for saying the revolution in identifying driver mutations and tailoring drugs against them would yield results in 10 to 20 years. It simply has to be faster than that, Watson said: "People need miracles, and we're the only ones who can give it to them." Kids with leukemia, men with testicular cancer, and patients who are alive because of Gleevec or herceptin have their miracles. For the first time, people with cancers that have long outwitted science have a realistic chance of getting a miracle, too.…