Measuring Ki-67 Levels to Inform HR+ Breast Cancer Treatment

Mark Pegram, MD: Despite the success of the legacy endocrine treatments we’ve had for adjuvant therapies, we now have a new class of drugs in the form of the CDK4/6 inhibitor, abemaciclib, based on the monarchE [NCT03155997] data, which showed an IDFS [invasive disease–free survival] benefit, particularly in patients with high Ki-67. Suddenly, we’ve taken an interest in Ki-67. What has been your experience with your pathologists in getting that measured? Were you routinely doing it? Or did you have to start ordering it?

Lee Schwartzberg, MD, FACP: We were routinely doing it for [approximately] the [past] 8 years. It has become common because of monarchE and the approval based on Ki-67 being important to measure. There has been a lot of debate in the literature about measuring it accurately. Using computer-assisted diagnosis techniques or a pathologist who gets used to measuring Ki-67, whether they’re only a breast pathologist or a pathologist who sees many diseases, Ki-67 has gotten more common in [several] diseases, [and] pathologists have gotten more comfortable with measuring it accurately. That said, there’s heterogeneity of Ki-67. One [must] be careful that an experienced pathologist is looking through the tumor and [ensuring] they get an accurate reflection of what that proliferation marker is.

Mark Pegram, MD: I hate that it has historically been scored with a single percentage number, because across the face of a breast tumor, there are areas around the periphery that are well vascularized that have high Ki-67 scores and rapid proliferation. Whereas the chronic centers of tumors tend to be slow proliferating, if at all, and have low levels of Ki-67. Giving that tumor a single percentage score when it’s a range isn’t correct from a scientific point of view. We [must] develop better ways of capturing the Ki-67 level, because there are instances in practice where you have a patient whose Ki-67 is 18% or 19% and you’re wondering about abemaciclib. Let’s say they have node-positive disease and high-risk factors. Should you go back and ask your pathologists, “Is it really 18%, or could it be 20%?”

Lee Schwartzberg, MD, FACP: You’re absolutely right.

Mark Pegram, MD: It’s probably between those 2 numbers. The error bar is around 20%, or at least plus or minus 5%. That’s a bit tricky.

Lee Schwartzberg, MD, FACP: I don’t like dichotomous variables in general for biologic processes that tend to be continuous, but we have to make a point. I agree with you completely. In your center, do you look at the Ki-67 at only the biopsy, at the final diagnosis—the excisional lumpectomy or mastectomy—or both? Because your point is well taken.

Mark Pegram, MD: At our tumor board, I tend to see it both on the biopsy and on the final surgical specimen.

Lee Schwartzberg, MD, FACP: That’s probably the right way to do it. There’s also grade. Over the years, I’ve gotten more impressed that [although] grade is also highly subjective—perhaps even more subjective than Ki-67, because you don’t get a number—particularly in the community, there can be heterogeneity in the ways different pathologists score it. A true grade 3 has a worse prognosis, and that correlates pretty well with Ki-67. Particularly for community oncologists, it’s important to have both of those things: an accurate grade and an accurate Ki-67 with the range.

Transcript edited for clarity.