Skip to Main Content

A huge federal trial of personalized cancer medicine has run into an unexpected roadblock: Many of the tumor samples aren’t robust enough to be put through genetic analysis.

The samples, taken from patients with advanced cancer, were collected by doctors in hundreds of clinics nationwide. When researchers checked them, they found as many as 1 in 5 didn’t have enough malignant cells to analyze, in most cases because the biopsy had been poorly done.

The glitch raises troubling questions about the new era of precision medicine.


Biopsies done in rural or community centers were more likely to be shoddy. Those patients’ tumors couldn’t be analyzed and they couldn’t be matched with medicines targeted at their specific genetic mutations.

And that suggests that access to the most advanced, customized cancer treatments might be highly unequal — and dependent on the quality of the local doctors performing biopsies.


“We’ve seen a real dynamic in the quality of the tissue samples,” said trial investigator Dr. Kurt Oettel, an oncologist at the Gundersen Health System, which serves counties in Wisconsin, Iowa, and Minnesota. “It’s posed a big challenge, and I think it’s a new learning curve for everybody.”

Precision medicine customizes treatments based largely on a patient's genetic profile. Hyacinth Empinado/STAT

The MATCH trial, launched last August by the National Cancer Institute, is billed as “the largest, most scientifically rigorous precision medicine cancer trial to date.” It’s part of the Obama administration’s $215 million Precision Medicine Initiative, aimed at driving a revolution in health care by customizing treatment to a patient’s genetic makeup.

The study started off with a bang, signing up close to 800 patients at breakneck speed on its way to a goal of 3,000 recruits. Then, in November, investigators halted the trial, which is expected to cost at least $30 million.

Researchers said they’d always planned the pause to take stock of how the study was going. But Dr. Keith Flaherty, the oncologist who is leading the trial, didn’t expect the interim analysis to reveal such stark and frequent quality concerns in the biopsy samples.

Shoddy biopsies leave patients scrambling

The protocol calls for tiny bits of tumor to be collected and shipped off to a central lab for DNA extraction. Doctors use a hollow needle to withdraw thin tubes of cancer tissue about the diameter of mechanical pencil lead and an inch or so in length.

The trial investigators are supposed to gather four such tissue “cores” from each patient. Many times, however, the technicians at the central lab at the MD Anderson Cancer Center in Houston would open tissue shipments to find they had received only one or two. And even when there were enough cores, the samples were often of poor quality, containing very few tumor cells.

Overall, 10 to 20 percent of the samples did not yield enough cancer tissue for DNA testing, Flaherty said. More often than not, he found, these subpar biopsies came from smaller, more rural treatment centers, though many did come from major academic research hospitals.

The slipshod biopsies had real implications for the patients.

To be eligible to enroll in the MATCH trial, cancer patients had to have exhausted all standard treatment options. So this trial was often their last chance.

But the bad biopsies meant they couldn’t be matched with a personalized therapy. With time ticking, they grasped at straws. Many tried more chemotherapy, usually in vain. Others scrambled to find other clinical studies testing experimental therapies. Some simply entered hospice care.

A paradigm shift in biopsy philosophy

The doctors who take biopsies from cancer patients are known as interventional radiologists. These specialists have pioneered many procedures in minimally invasive medicine — and for cancer patients, their job has traditionally been to collect as little tissue as needed for a pathologist to make a diagnosis.

The more tissue, the more the risk, because each time doctors go back with the needle elevates the chance of bleeding, infection, or worse.

But precision medicine has flipped that equation on its head: More tissue can now mean more reward, because only larger and more invasive biopsies can provide the type of genetic information needed for personalizing therapy.

“You have these competing issues of trying to minimize toxicity from the procedure and trying to get optimal tissue,” said Dr. Michael Thompson, a Wisconsin oncologist and the incoming chair of the Community Research Forum Council at the American Society of Clinical Oncology.

Additionally, personalized treatment regimens like the one in the MATCH trial require biopsies not just at the time of initial diagnosis, but oftentimes again and again as various drug options stop working each time a tumor evolves a new resistance mutation.

“You need to get tissue in situations where you never wanted to get tissue in the past,” Oettel said.

Flaherty hopes that better training of interventional radiologists and the medical oncologists they work with will improve biopsies.

“It’s a major component of what we’re learning in trying to execute a big national trial,” said Flaherty, who directs the Henri and Belinda Termeer Center for Targeted Therapies at Massachusetts General Hospital.

Some community health advocates, however, aren’t happy that trial leaders have called out their rural clinics as particular trouble spots.

“I can tell you: There’s variability in sample quality from any institution,” said Lynn Dressler, director of personalized medicine at Mission Health in Asheville, N.C. She called for more data to be made public before any conclusions are drawn.

“Lumping the differences into two major categories of academic versus community really does an injustice to the entire field and to the sites,” she said.

Mission Health is one of the many sites that will start registering patients in the MATCH study as soon as the enrollment pause lifts. That was originally slated to happen this month, but the date has been pushed back to April or May, while “a number of changes are being made to the trial,” the NCI announced on its website last week.