Lessons from the lab on de-escalating breast cancer treatment
There are tumors growing in a petri dish in Bloomfield, and researchers are studying every one of them.
Adrian Lee and his colleagues are studying these lab-grown tumors and mining through data to determine whether breast cancer patients could benefit from having less treatment rather than more.
The idea, known as deescalating or de-implementing treatment, is nothing new. These studies are part of an effort by researchers to determine which patients may benefit from what Lee – a professor at UPMC Hillman Cancer Center and the director of the Institute for Precision Medicine – calls “right sizing therapy.”
A study from the group, published last year in JAMA Network Open, looked at two treatments: sentinel lymph node biopsy, which is done by removing the lymph node to which cancer is most likely to spread, and radiation therapy to the breasts. What they found was some older patients don’t require as much therapy as a younger person would; regardless of how much treatment the patient received, the likelihood of the tumor coming back was still low.
“At baseline, a lot of the tumors in older women have a different biology,” said Neil Carleton, a graduate student at the University of Pittsburgh who works in the lab and researches de-implementation. Some of the tumors in elderly patients are slower growing, smaller or unlikely to kill a patient, he said.
After looking at 2,109 women over age 70 with early-stage breast cancer, the results from their study suggested that doctors should strongly consider de-escalating those procedures in older patients. The process involves stopping or scaling back certain practices because they may be more harmful to the patient, costly or simply ineffective.
These guidelines for not doing such treatments in older women have actually been recommended for many years, Lee said. The Society of Surgical Oncology says the sentinel node biopsy should not be routinely used in some breast cancer patients over 70 years of age. But he noted that it takes a lot of evidence to persuade people to stop doing something.
“You give therapy to try and reduce the risk of recurrence, for example,” he said. “Obviously, surgeons and patients, both of them would like to reduce the risk of recurrence as much as possible, so that leads to this potential-overtreatment.”
He compared it to having the flu and asking a doctor for antibiotics. “If you have the flu, that makes zero sense because it’s a viral infection, and antibiotics don’t do anything against viruses,” he said. “That’s a classic overtreatment.”
Other research has found similar results with different forms of cancer. In a study published in JAMA in March, researchers in New York found that one-quarter of the more than 600 cancer center websites they studied had recommendations that all men receive prostate cancer screenings. The U.S. Preventive Service Task Force, however, recommends against prostate screenings in men 70 and older.
Now, the team is looking at whether sentinel lymph node biopsy is appropriate for early-stage breast cancer patients younger than 70. They’re also using artificial intelligence to look through data from over 50,000 patients who were treated for breast cancer. The technology they’re using looks through mountains of patient data to identify key variables for their research.
“Say we want to find the size of a tumor, and we have a million pages of
text,” Lee said. The technology “can sift through that in minutes and find the relevant text that says the size of the tumor.”
By the time the data gets to Lee’s team, the patients’ identities have been removed from what they’re studying. All they see are several thousand variables, he said.
That’s the same with those tumors that are growing in petri dishes in their lab in The Assembly building, a former Ford plant that is now a research center. Looking through a microscope, researchers can see the cancer cells grown from samples taken from real patients and see how they respond to treatment to determine what types of therapy work best.
“It’s cool to have specimens from those same patients that you see in the clinic,” Carleton said. “We collected these prospectively. They’re all in the same patient population that would meet the criteria for these surgical de-escalations.”
They’ll also sequence RNA and DNA from the samples to try to understand the genetics of the cancer, using them as both research tools and what Lee envisions
as a way of “treating you outside of your body.” Part of what makes this type of research valuable, he said, is that elderly people tend to be less recruited than younger patients in clinical trials.
The reasons are complicated; Carleton said research has shown that some physicians don’t approach them, despite evidence that many elderly patients would be willing to contribute to such trials. But he pointed out that their research notes that estrogen receptor-positive breast cancer, which is the most common form of breast cancer, is found most often in elderly patients.
In addition to extending their recommendations on right-size breast cancer treatment to younger women, their team is also looking into tracking tumors through blood biopsies, which could help avoid surgical and radiation procedures, Carleton said. They’re also researching how surgeons talk about these procedures with patients during their first appointment.
That’s part of their work looking into how to strike a balance between what he called a historically “paternalistic” bedside manner from doctors who decades ago would have just told people what to do, and trying not to put the full burden of the decision making into the hands of the patient.
“I think it goes back to the shared decision making,” Carleton said. “You want the patient to be fully informed of what the spectrum of care is, and also have them have input on that.”