Personalized medicine is changing the way cancer drugs are developed as clinical trials enroll patients based on the genetic flaws of their tumors, not just the organs where their disease originates.
Researchers at M.D. Anderson Cancer Center in Houston found that when therapy was tied to a specific molecular defect — whether the tumor originated in the lung, breast or other site — 27 percent of patients improved, versus 5 percent of those who didn’t get tailored treatment. Their report was presented Friday at the American Society of Clinical Oncology meeting in Chicago.
New technologies enabled researchers to peer deep into cancer cells, identify molecular changes that fuel malignancy and then assign patients to trials for drugs that target tumors’ specific flaws. The goal is to improve on the notoriously low response rate doctors get when experimental drugs are first tested in cancer patients, often on a one-size-fits-all basis.
“We are giving them an extra diagnosis, and we match that with a clinical trial that uses a drug that would be appropriate for the molecular abnormality,” said Razelle Kurzrock, chair of M.D. Anderson’s department of investigational therapies, in a telephone interview. “There are a lot of drugs designed to target a specific abnormality in cancer. The technology is here. We can individualize therapy at the molecular level.”
Linda Barth, a 60-year-old former magazine editor in Houston, was ready for a new approach, having exhausted all the approved drugs for her endometrial cancer diagnosed in 2006. Barth tried an aggressive combination of three medicines that shrank the tumors in her lungs and neck and made her so weak she couldn’t walk. A second experimental drug provided no benefit.
By honing in on the genetics of Barth’s cancer, researchers found she was one of 460 clinical trial patients among 1,140 screened who had a known molecular twist that helped the malignant cells grow.
Apostolia-Maria Tsimberidou, her doctor at M.D. Anderson, started Barth on a trial of two experimental drugs from GlaxoSmithKline in January that targeted a gene called KRAS, which when mutated has the potential to cause normal cells to become cancerous. Her tumors have since shrunk 68 percent, according to CT scan results in May, she said.
“They know my DNA — all of these treatments they have designed for me,” Barth said in a telephone interview. “Who knows if it will cure me? If it just stabilizes my cancer and prevents it from growing, I can live with that.”
The M.D. Anderson clinical trial program uses molecular tests to identify the biological pathways that trigger cancer, spur its growth or strengthen its resistance to treatment.
After patients are assigned to one of the center’s 120 studies using experimental medicines that target their individual cancer’s traits, they are compared to similar people who don’t get matched therapy. Reasons for not getting a matched therapy might include travel difficulties or lack of insurance.
The trials tend to be a last-ditch effort to find a therapy that works for each patient. Participants in the trials failed an average of four previous drugs.
Other cancer centers are doing similar work, often looking at one tumor type or for a specific mutation. At Memorial Sloan-Kettering Cancer Center in New York, everyone with metastatic lung or colon cancer undergoes tests for dozens of variations that may influence how they respond to treatment, said oncologist Leonard Saltz.
While the approach may benefit more patients who enter cancer studies and speed drug development, it narrows the number of people who qualify for experimental treatment — and the potential size of the market, said Saltz, who is head of the colorectal oncology section at Memorial Sloan-Kettering.
“Historically it has been difficult to get drug companies to be interested in this type of approach because it potentially limits their market,” he said in a telephone interview. “If you have a drug that works in 10 percent of the population but you can sell it to everybody, that has been a widely successful business model.”
The targeted approach may also save an experimental drug from failure.
Initial studies of Pfizer’s crizotinib included patients with any kind of cancer. Researchers quickly realized that the three patients who responded to treatment, out of 37, had a defect in a gene known as ALK that is key to the growth and survival of tumor cells, said Mace Rothenberg, senior vice president of clinical development and medical affairs in for New York-based Pfizer’s oncology unit.
The company decided in 2009 to focus solely on lung cancer patients with the ALK mutation, a group that makes up just 5 percent of the lung cancer population. The bet paid off. The company filed for regulatory approval in the U.S. based on its early trials, with data showing dramatic response to treatment in the properly targeted patients. Pfizer has two larger studies under way.
Rothenberg said he was concerned that a drug that could only benefit 5 percent of patients might not be commercially viable.
“Fortunately, my colleagues in commercial development immediately recognized that, given the high incidence of non- small-cell lung cancer, it still accounted for a considerable number of patients,” Rothenberg said.
Matching tumors to treatment isn’t a magic bullet. The treatment stopped working in the M.D. Anderson trials after 5.2 months for those getting targeted care, compared with 2.2 months for patients treated with unmatched drug therapy.
“It is going to take a lot more effort and investigation to get beyond this,” Otis Brawley, the chief medicalofficer at the American Cancer Society, in a telephone interview. “If the time to treatment failure is truly 5.2 months versus 2.2 months, I see it as progress, but it is still not satisfying.”
Eventually researchers will need to combine targeted treatments to cut off all the avenues cancer uses to grow, not just those researchers can find now using genetic and molecular tests, Brawley said. Advanced technology is identifying a dizzying array of nuances in cancer growth and development that suggest tumors within a specific type of cancer have many ways to evade treatment and flourish, he said.
Less than 5 percent of adult cancer patients are in clinical trials. Often the reason is the design, with researchers giving 1,000 patients an experimental drug that benefits only 10 percent of them, said John Marshall, chief of hematology and oncology and director of clinical research at Georgetown University’s Lombardi Comprehensive Cancer Center in Washington.
“When you are in the trenches in a clinic and you are afraid you are going to die, and the doctor wants to flip a coin to see what treatment you should get, that’s another level of uncertainty in a freaked-out world,” Marshall said in an interview. “We’re hoping to make it much easier. If we offer them a gene test that might inform a treatment decision, they are totally engaged.”
Most patients don’t have the option now. It’s easier to give chemotherapy that is unlikely to work than to do a genetic analysis, Marshall said.
“We are getting smarter, but not fast enough.”