Blood-Based Markers Improve Precision Medicine in Breast Cancer

July 11, 2019
Brandon Scalea

By better understanding the subtypes of breast cancer, clinicians can create more personalized and effective treatment plans for patients with breast cancer.

Blood-based markers are the key to better understanding the biology of various breast cancer subtypes, said Mark E. Burkard, MD, PhD.

One of the most prominent markers in breast cancer is circulating tumor DNA (ctDNA), which is being used to measure tumor burden and detect potentially targetable mutations.

“[ctDNA] is of interest because it's a way of doing a simple blood test to get a sense of how much tumor burden is in a person, especially with metastatic disease,” said Burkard. “It also can tell you what particular genetic characteristics that tumor has. ctDNA can be very helpful for a specific selection of treatment.”

Alterations such as NTRK, PIK3CA, AKT, and ESR1 are also helping to drive treatment decisions.

“There are definitely emerging targets, and some of these only happen at the time of recurrence,” he added. “You may not find [these targets] in the breast cancer tumor that the surgeon resected; you may only find them after in the metastatic setting.”

In an interview during the 2019 OncLive® State of the Science Summit™ on Breast Cancer, Burkard, an associate professor of medicine in the Hematology/Oncology section of the University of Wisconsin (UW) School of Medicine and Public Health, and a member of the UW Health Breast Center, discussed the significance of utilizing markers in the blood and highlighted emerging actionable mutations in breast cancer.

OncLive: What is the significance of ctDNA in breast cancer?

Burkard: In terms of measuring residual tumor burden, one area this is useful is in triple-negative breast cancer. We know that when we treat these patients with neoadjuvant chemotherapy, we can get a sense of how well that [approach] worked and whether it eliminated tumors based on their surgical response.

There are standard ways to assess for that. For the patients who failed to have a good response, there are some recent data from a group of researchers at Indiana University suggesting that you can also stratify patients [to determine which ones are] most likely to have recurrent breast cancer by doing blood biomarkers. Patients who have a high level of ctDNA in their blood are the ones who are most likely to have their cancer return. That's a nice way to potentially better identify patients who would benefit from more therapy.

Another use [for ctDNA] in the metastatic setting would be for genomic or precision oncology. We are selecting patients more and more in breast cancer based on the genes in an individual tumor. In the past, we focused on HER2 exclusively, but we now know more about how patients with BRCA mutations can benefit from PARP inhibitors. We know there are very rare patient cohorts who can benefit from FDA-approved drugs for genetic translocations of the TRKA/B/C genes or who have microsatellite instability-high (MSI-H) tumors.

We were just hearing about the new evidence that PI3K inhibitors may be useful in patients with PIK3CA-mutant breast cancer. Usually, the way we figure out if a patient has [these mutations] is by doing a biopsy on the tumor. Looking at the DNA in the tumor is another way to do this. In many of our patients—especially those with estrogen receptor (ER)—positive breast cancer—this isn't feasible, because they may have bone-only disease or the metastatic disease is in inaccessible sites. Cell-free DNA (cfDNA) is another way to determine [if there are molecular abnormalities]. It's a good adjunct for patients with metastatic breast cancer in whom you want to take a precision medicine approach.

What is the frequency of patients with MSI-H breast cancer?

[The prevalence is] very low. In one of the studies by Luis A. Diaz, Jr., MD, of Memorial Sloan Kettering Cancer Center, [the prevalence of those patients] was determined to be as high as 2% based on results from a retrospective analysis; that's on the high end.

However, with most of the MSI-H assays, it's far less than 1%. Other estimates are as low as 0.3%. Therefore, these [markers] are very rare, and NTRK is probably even more rare. Frankly, it may be hard to pull those out of blood tests, but those are [markers for] which you have actionable drugs, if you can find them.

Are there any other promising targets under investigation?

We have AKT mutations; A17K is the most common activating mutation. Although there are no FDA-approved AKT inhibitors today, there are several in clinical trials. There are also the PIK3CA mutations that I mentioned.

Another emerging target is FGFR, which are very rarely mutated or translocated; however, FGFR1 amplification is very common as a mechanism of resistance to hormone therapy in ER-positive breast cancer. The final one which we don't have a drug for are the ESR mutations—ESR1 is the gene; it’s mutated in almost 25% of breast cancers after they develop resistance to hormonal therapy. Therefore, there are many active drug developments but no FDA-approved drugs. People have tried high-dose fulvestrant (Faslodex), and there is some evidence suggesting that this may be promising, but there are newer drugs coming out. Many emerging targets are out there.

What are some trials evaluating the use of blood-based markers in breast cancer?

Many of these studies are observational; they are built on as adjuncts to the existing trials. What we've learned from them is more about the biology of the disease. For example, one thing we learned that's slightly frightening is that when patients develop resistance to hormonal therapies in ER-positive breast cancer, often the resistance is polyclonal. Therefore, if you biopsy a tumor and look for the ESR1 mutation, you find it. If you look for the FGFR1 mutation, you might find it in a different tumor. With a blood sample that is collecting DNA from every tumor in a person's body, sometimes you can find both of these resistance mechanisms happening simultaneously. This reveals that when resistance develops, it can develop at many places and in many cells simultaneously, which is very interesting biologically.

Practically, this poses a very unique challenge; you're not going to reverse that [resistance] by only targeting FGFR1 or only ESR1 mutations. You have to do something a little more strategic to get around that. The other place where this is being implemented into clinical practice is through measuring residual disease. We've had protein biomarkers like CA27-29 and CA15-3 for a long time. We also have circulating tumor cells, which are measured in cell-search assays and other tests. [With this approach,] people count the number of cells they find in the bloodstream that came from the cancer; as the tumor burden increases, [the number of these cells] increases, and as [the burden] decreases, they often go back down.

cfDNA is another way of accomplishing that, although it's not clear if that [approach] will be better. It poses some advantages and disadvantages compared with the other methods. In terms of advantages, [this approach] is potentially more sensitive because there are technologies for DNA amplification that could see as little as 1 molecule of DNA with a mutation.

There are also challenges [with this method] because people have other DNA floating in their bodies, and sometimes, we find out that those DNA are mutated for other reasons. For example, there's a group of people who have p53 in their bone marrow, which you can pick up this way. In this case, it doesn't come from the tumor at all, it's a [genetic alteration] in their bone marrow. It's hard to distinguish those things.

There are various other challenges with measuring disease burden by quantitating the amount of DNA harboring a particular mutation in the blood. That's another area of interest, and I suppose it has different advantages and disadvantages against the pre-existing measures of tumor burden.

Could you expand on the challenges faced with understanding these genetic alterations?

Almost every breast cancer is unique in its genetic characteristics. One of the major challenges is that the most common mutations in breast cancer are PIK3CA and p53, which occur in up to 25% of breast cancers. Everything else is less common. Capturing the diversity [of these genetic alterations] is a challenge. If you want to quantify the amount of tumor burden by measuring the DNA in the blood, you have to accept that diversity and say, "I don't even know what mutations I'll find in this particular patient." You need to just quantify the amount of DNA in each mutation. That is a special challenge, and it's particularly difficult when one is trying to measure deletions and amplifications of genes.

For example, HER2 amplification happens in 20% of breast cancer. Usually, you would capture that by doing a biopsy, taking the DNA out, and measuring the number of times you get DNA sequences from the HER2 gene versus everywhere else. If you had a lot more in the HER2 gene, it's amplified.

In the blood, that poses a challenge, because now you've mixed DNA from the tumor with DNA from many other blood and non-blood cells. It may be hard to capture that amplification in the setting of that deletion. What was a 200-fold amplification might only show up as maybe 10-fold, and what is a 10-fold amplification might not show up at all.