Genomic Testing for Risk of Late Recurrence/ASCO Guidelines

Transcript:Ruth O’Regan, MD: So, for estrogen receptor—positive breast cancer, we now have actually a number of different genomic assays that are available. The 21-gene recurrence score has obviously been around for a number of years and is very commonly used at the time of diagnosis. We need to estimate the benefit of chemotherapy. But, I think it also gives you very important prognostic information and also may give you a hint as to whether the patient is going to benefit significantly from endocrine therapy. I think it’s multifaceted with that approach. Most of the data with the 21-gene recurrence score are in patients with node-negative breast cancer, but there are some limited data in a fairly small number of patients with node-positive breast cancer. The ASCO guidelines support the use of the 21-gene recurrence score in node-negative breast cancers, but not currently in node-positive breast cancers, which is a little unfortunate because I do think there are older patients out there where it’s very useful to determine whether they would benefit from chemotherapy or not.

Another one is the 70-gene signature. And, just like the 21-gene recurrence score, it’s able to basically determine patients who have a very low risk of recurrence and probably do not need chemotherapy. However, it doesn’t have quite the same predictive value, I think, as the 21-gene recurrence score. It’s certainly very useful, prognostically, and the MINDACT study did include patients that had node-negative and had up to three positive lymph nodes. So, at least there are data basically showing that this test is prognostic in patients with node-positive breast cancer. Now, the ASCO guidelines didn’t support the 70-gene signature at the time the guidelines were issued. However, my understanding is that they’re relooking at the data now since the MINDACT results have become available.

The PAM50 is another assay that determines what type of breast cancer a patient has. There are two estrogen receptor-positive types—luminal A and luminal B. Again, probably useful for prognosis, but not probably as much for predictive value in terms of whether patients need chemotherapy and how they’re going to do with endocrine therapy at this point. That was not supported by the ASCO guidelines.

The last one to mention is the Breast Cancer Index, which is a combination of the Molecular Grade Index and the H/I ratio. That has predominantly been used to predict benefit, or at least predict prognosis in years 5 to 10 following a diagnosis of estrogen receptor-positive breast cancer, and also has some predictability based on the MA.17 study, being able to select patients who are likely to benefit from switching from tamoxifen to letrozole after five years versus not. That also, however, was not supported by the ASCO guidelines at this point. And, I think the issue with these genomic tests is that in a lot of cases, we’re just waiting for more data to really validate them in these different settings.

I think from our perspective, the biology of an estrogen receptor—positive breast cancer probably doesn’t change, whether you’ve got a positive node versus you don’t. So, you may have a higher risk of recurrence if you’ve got positive nodes, but it doesn’t mean that if you’ve got a low recurrence score, for example, on the 21-gene assay and you have two lymph nodes positive, that chemotherapy is going to majorly impact your risk of recurrence.

In terms of late recurrence, most of the genomic assays that are currently available have attempted to look at the question of what the risk of recurrence for an individual patient is in years 5 through 10 and following a diagnosis of estrogen receptor—positive breast cancer. There aren’t as much data beyond 10 years. Most of them haven’t looked at that at this time point. The one that probably has the best data is the Breast Cancer Index, which basically comprises the Molecular Grade Index and the H/I ratio. The two combined, the Breast Cancer Index, has been shown in a number of series to be able to predict the risk of late metastasis in patients with estrogen receptor–positive early stage breast cancer.

There are also some data where if you can include the Breast Cancer Index plus the tumor size and grade, they actually are able to determine a group of patients with 1 to 3 positive lymph nodes that have an incredibly low risk of recurrence in years 5 to 10. So, they have prognostic data, primarily in the node negative setting, but also, I think, in the node positive setting for years 5 through 10. The H/I ratio was also used in the MA.17 study to show that you could actually determine which patients were going to benefit from switching from tamoxifen to letrozole versus just stopping tamoxifen at 5 years. That has probably got the best data in terms of late recurrences. But, certainly, a number of other assays, including the EndoPredict, were able to show that if you use the estrogen regulated component of EndoPredict, that that also predicts for recurrences years 5 to 10. There was also recently an interesting paper that came out with the 21-gene recurrence score, basically showing that if you look at the single-gene estrogen analysis from the 21-gene recurrence score and they had a certain cut-off, and above that cut-off, you could actually use the recurrence score and that was predictive of recurrences in years 5 to 10, as well. So, that’s something that might be useful if you’ve already sent the 21-gene recurrence score to look back at 5 years and work out what the patient’s risk of recurrence is years 5 through 10.

I think, overall, as far as working out the risk of late recurrence, you can certainly do it at the time of diagnosis. I think a lot of people use the Breast Cancer Index towards the end of 5 years of treatment. Now, obviously, you’re looking at the primary cancer at that time point, the cancer at the time of diagnosis to estimate the risk. I think that’s a very interesting question, the idea of actually re-testing. It’s really like doing another molecular assay coming up to 5 years to estimate the risk of recurrence years 5 to 10. But, the caveat there is that you obviously are looking at the cancer at the time of diagnosis. So, I think you’ve based all your prognostic information on both early and late recurrences on what the cancer looked like at the time of diagnosis. It’s certainly possible that things may change over time. And, if a patient has micrometastatic disease, maybe there might be a change in those cells following aromatase inhibitor versus not. For example, they could acquire an ESR1 mutation. But, there are no data available right now to address that issue.

Transcript Edited for Clarity