Tools to Guide Treatment Decision-Making in HR+ Breast Cancer

Lee Schwartzberg, MD, FACP: How do you use biomarkers in that setting? We talked about the clinical features. More nodes obviously increase your risk of recurrence, larger tumors, and higher stage. But those are the anatomic features. What about biomarkers? How do you look at that?

Mark Pegram, MD: The degree of steroid receptor expression is another variable. Some ER [estrogen receptor]–positive tumors are 2% ER expression and PR [progesterone receptor] negative. That’s still considered ER+ under the current ASCO [American Society of Clinical Oncology]/CAP [College of American Pathologists] guidelines, but that isn’t really ER+ disease, is it? Most of us who are experienced and have been around for a while probably still lump those in with triple-negative [disease] and treat them in kind because based on studies of metrics, such as the Allred score of semi-quantitative steroid receptor expression based on intensity of staining and the percentage of nuclei that are stained, that the lower the percentage of stain, the less probability you have of responding to endocrine therapy. You have to take that into account. That’s 1 thing.

We already mentioned Ki-67 as a marker of proliferation that’s commonly obtained routinely in many clinics, if not the majority, in this day and age. Then there’s grade. We still look at grade in our tumor boards and take that into consideration, along with other clinical pathological parameters. But the real pay dirt in management of early ER+ disease are the multigene assays. At our institution, we do a lot of Oncotype testing. It has been around probably the longest. It’s part of the AJCC [American Joint Committee on Cancer] staging system. There are so much robust data, ranging from the NSABP [National Surgical Adjuvant Breast and Bowel Project] days to the RxPONDER trial and the TAILORx trial in between with thousands, if not tens of thousands, of patient subjects randomized in those large trials. It has a lot of very robust data behind it. That’s what we tend to use at our center.

That said, we occasionally see people come from other clinics for second opinions or are transferring care because they moved from out of state or what have you, and they may have a MammaPrint assay instead. We typically don’t repeat multigene assays at our center. If somebody has one of the other robust multigene assays, we’re usually fairly content with that. I don’t think everyone has to be forced into doing an Oncotype, so I’m comfortable with other options. How about you?

Lee Schwartzberg, MD, FACP: I agree. Throughout my career, I’ve mainly used Oncotype and some MammaPrint. It’s worth saying, and this is in the ASCO guidelines as well, that all the multigene classifiers are relatively good at prognostic features. In other words, they can tell the risk of recurrence. That’s true for Oncotype, MammaPrint, EndoPredict, Prosigna, and BCI [Breast Cancer Index]. They all will predict the prognosis within that 5- to 10-year parameter. But as you say, 1 reason we use Oncotype is to make a predictive decision on whether chemotherapy is indicated. Another way to say it is: do we get any additional benefit from chemotherapy over endocrine therapy using that as a baseline, where everyone is going to get it?

The population-based data and especially data from prospective trials that specifically asked, “Does chemotherapy benefit patients?” gave us the answer from TAILORx and RxPONDER. That’s 1 reason, besides it being the oldest test that has been evaluated for almost 20 years. It’s hard to believe, but it’s almost 20 years old, and it changed the way we think about breast cancer. These tests are all pretty good prognostically, and Oncotype is probably the easiest to use predictively, in my experience.

Transcript edited for clarity.