Ask AI
Back Back
MRD ctDNA Testing in Triple-Negative Breast Cancer: A Candid Conversation With Dr Marija Balic
Activity Information Ask a Question

Activity

Progress
1
Course Completed
Activity Information

Released: May 21, 2026

John Marshall
John Marshall, MD
Catch Up On All Episodes of Oncology Unscripted

Watch Now

Ep 30 Interview: MRD ctDNA Testing in Triple-Negative Breast Cancer: A Candid Conversation With Dr. Marija Balic

[00:00:06] John Marshall, MD: John Marshall back, episode 30, as promised, for Oncology Unscripted. I am always very, very lucky to get to interview the best and the brightest in our field across the country and around the world.

And today is no exception. I am joined by Marija Balic from UPMC, up in Pittsburgh, not too far from where I'm sitting right now. It's a kinda windy road to get there, but you can get there nonetheless. And we are so excited she's joined us because she has been on point of an NSABP study looking at neoadjuvant therapy in breast cancer, of course. But this is specific to triple-negative breast cancer and specific to MRD ctDNA testing. And her data has been so profound that she's gotten, of course, to present it at San Antonio and at other meetings along the way. And we are lucky enough that she said yes. So, Dr. Balic, welcome to Oncology Unscripted.

Tell us a little bit about yourself.

[00:01:12] Marija Balic, MD, PhD, MBA: So, thank you. Thank you so much for this enthusiastic introduction. A little bit about myself: I joined, 2 years ago, to lead the breast program here at UPMC at the University of Pittsburgh, and mainly also based on the NSABP that is very close here in Pittsburgh, and the role of NSABP that it has played in my life.

I joined from Austria. I led the breast center at the University of Graz and was vice president of the ABCSG, which was basically built after the model of the NSABP. And, since 2 years, I have joined both and am leading both the program at the university as well as the translational research program at the NSABP.

[00:01:56] John Marshall, MD: I'm a good friend of Norman Wolmark's and all the gang at NSABP. Soonmyung Paik used to be here at Georgetown a long time ago. That's a great group with some of the smartest people on the planet, and I know they're even smarter with your being there. So, welcome.

[00:02:12] Marija Balic, MD, PhD, MBA: Thank you. Thank you. Well, Bernie Fisher initiated the thoughts of early metastatic spread, so it's the source of all the research we are doing currently.

[00:02:20] John Marshall, MD: It is. It's the beginning of it all, isn't it? Right there. So, it's so great for you to be a part of that.

Listen, in GI cancers, and colon cancer in particular, we are in this debate of: should we be doing MRD testing or not? We haven't had the kind of innovations you guys have had in breast cancer, so we're taking anything we will take. So, we've been fairly early adopters.

When I talk to my breast cancer friends, it's more controversial about whether you do it or not, and are there certain kinds of breast cancer where you do it more than others. So, what's sort of your take today on MRD testing in earlier breast cancer?

[00:02:58] Marija Balic, MD, PhD, MBA: My take would be that we are closer to demonstrating the potential clinical utility, but we are not there yet compared to GI and colorectal cancers, where we've seen the stage II data and basically the treatment decision contribution based on ctDNA, whether to treat patients with adjuvant treatment or not. I think triple-negative breast cancer may be the first, where we will be able to demonstrate this utility based on our data, based on data that was represented from the I-SPY clinical t rial, and also the summary that was published by Mark Magnanua last year in Nature Medicine. In triple-negative breast cancer, we see a high frequency at baseline. In our own data, 96% of patients were positive.

[00:03:43] John Marshall, MD: That shocked me. I mean, that you had such a high percentage of positive. Yes, tumors in situ, right? In colon, we have a pretty high percentage that aren't positive even with tumor in situ. So, that's a difference right there, a biology difference.

[00:04:01] Marija Balic, MD, PhD, MBA: Definitely, and we see differences across the subtypes at baseline. Triple-negative breast cancer is probably with the highest percentage of positive cases. The greatest challenge is in hormone receptor-positive cases, both at baseline and throughout the treatment. And then, somewhere in the middle, based on its mixed biology, also HER2-positive breast cancer, depending on whether it's hormone receptor-positive or not.

[00:04:31] John Marshall, MD: What is your all sort of theory on that? The answer I give patients on this is that the faster your tumor is moving, the more likely the ctDNA will be positive. If it's slower, maybe—well, you know, for us it's peritoneal, but slow turnover, we see fewer uptake. Is that sort of your take on it too?

[00:04:49] Marija Balic, MD, PhD, MBA: There seems to be a correlation with the cell cycle and the speed of the cell cycle and the release of ctDNA, of course, and the high turnover of the triple-negative breast cancer seems to correlate with the ctDNA positivity at baseline. Which, we have also very nicely shown, is the clearance through the treatment with effective neoadjuvant treatments.

We do manage to clear 90% of those patients even before surgery.

[00:05:17] John Marshall, MD: We'll get back to the high-level results, I mean, which were just so impressive. But before we go there, a lot of us are being marketed the different assays that are out there. And so, there are 4, 5, 6 major sequencers that we have access to in this country. Some are tissue-based, some are blood-based, some offer both. An evolution in how many genes are picked and how much AI is used, et cetera. There's part of me that feels like, do I care whether my patient's CAT scan was done on a GE scanner or a Siemens scanner? I get the same CAT scan. Versus, no, these are fundamentally different assays and tests, and it is a little bit apples and oranges.

Do you have a sort of mindset about the different assays that are out there?

[00:06:08] Marija Balic, MD, PhD, MBA: So, I do believe that tumor-informed assays still allow for higher sensitivity, and we are currently even facing the first generation of ultralow or ultrasensitive assays. So, I am still convinced that the tumor information really generates more significant information. And then not only the way how you sequence tumors and how you get this tumor-informed information, but also how do you generate the approach to analyze blood and plasma? Between digital PCR for selected probes and large scale of selected probes, current assays are even using up to 1000 or 2000 probes. There is definitely a potential to increase sensitivity and thereby be much more prognostic and hopefully predictive as well.

[00:06:59] John Marshall, MD: Very helpful and very insightful. Let me give you a chance to tell us the punchline of what we're here to talk about. Your study, maybe a quick overview of the design and what are some of the high-level points that you found.

[00:07:12] Marija Balic, MD, PhD, MBA: So, basically what we have performed, that was a subanalysis in the NSABP B-59/GBG 96-GeparDouze study. It was a large study with more than 1500 patients who were randomized into adding immunotherapy to a maximum of chemotherapy backbone. And in a substudy of patients who were enrolled in the NSABP, a little more than 200 patients were available for tumor-informed sequencing, which was based on whole-exome sequencing, a very deep whole-exome sequencing between 400 and 800×.

And basically, after subtraction of the germline information, specific probe sets were identified, and then plasma hybridized to those specific probes. Per patient, we were able to identify between 40 and 200 specific alterations to be generated. And with that, and with appropriate dilatory algorithm to identify who are really positive cases, we were able to identify 96% of patients who were positive at baseline.

Then, before surgery, 10% of patients were still positive, but it went down to 6% after surgery, and that is basically identifying time points for de-escalation or escalation in treatment. And the primary endpoint was post-surgical endpoint and prediction of prognosis. And then, at our recent presentation at AACR, we included this post-neoadjuvant treatment, pre-surgery time point to really associate with long-term prognosis.

[00:08:53] John Marshall, MD: Just really, really impressive data. I was wondering, you know, we've been having discussions with regulatory agencies even now about MRD to negative as a regulatory endpoint. And I know you and others in the breast cancer field really helped lead the path CR quick regulatory endpoint in the adjuvant setting. Do you think this is starting to creep in on that? Maybe covariables or co-endpoints, you know, where you have path CR and MRD, that that's an even stronger regulatory endpoint?

[00:09:28] Marija Balic, MD, PhD, MBA: So, I think that's the way where this is developing. There will be some changes with sensitivity of the assays. I think that will be really important to define prior to designing clinical trials and really defining the endpoints. But, for example, in our case, when we take into account this post-neoadjuvant but pre-surgery time point, our positive predictive value for non-PCR was 85.7%, which is very high, but it is still not 100%.

So, the question is: can we improve this with better technologies to really be able, potentially at some point, to say, "Well, these patients don't need surgery"? We are close, but not there yet.

[00:10:14] John Marshall, MD: I, you know, over 40 years of doing this, I've watched everything from, you know—I didn't actually live during the radical mastectomy or practice during the radical mastectomy years, but we went from lots of surgery to lumpectomies to back to more surgery, prophylactic bilateral mastectomies. So, there's been this swing in the sort of mantra and community think around surgery and breast cancer.

Now, in GI cancers, particularly colorectal cancer, we're increasingly identifying patients who may not need their rectums removed and therefore need a permanent ostomy, you know, just a quality-of-life bonus. And so, we're using a lot of MRD testing to watch and wait those patients. Is this starting to be something you guys are folding in, or is there really not that same pressure for the mastectomy that we have for rectal removal?

[00:11:11] Marija Balic, MD, PhD, MBA: Oh, well, there is definitely a pressure, and it very much depends on the part of the world where you practice. The hypothesis that breast-conserving surgery is safe for the patients comes from Pittsburgh and from Bernie Fisher, basically, the first randomized trial that has demonstrated the safety.

I think the pre-surgery time point is also the most important time point for surgeons to really potentially identify patients who may omit surgery, and it may evolve there. But, as said, we are still not there yet. The drop of patients from 9% to 6% after surgery and the hazard ratio of 10 versus 30-fold increased risk after surgery if they remain ctDNA positive does show that there is certain role of ctDNA clearance just based on surgery.

So, we definitely have to work better on optimizing—

[00:12:07] John Marshall, MD: Is the answer for that? Is that a pathologist not seeing microscopic disease and therefore calling it a path CR, but it's not really a path CR?

[00:12:18] Marija Balic, MD, PhD, MBA: Potentially, potentially that's the reason, and it's definitely the reason. So, what we did see is basically also in a more granular analysis that patients who are ctDNA negative and still have remaining tumor are almost having as good prognosis as those who are ctDNA negative and achieve pathological complete response.

I do think that there is some resistant disease that treatment is still not eradicating that can then be removed by surgery, leading to additional curation of the patient.

[00:12:53] John Marshall, MD: Terrific. Dr. Balic, Marija, thank you so much for taking time, what I know is a very busy day, and thank you for sharing the data that you've got with me and with our audience out there. And my guess is we'll see more of this data in Chicago. So, maybe we'll cross paths in the big hallway somewhere along the way.

Thanks for joining us on Oncology Unscripted.

[00:13:12] Marija Balic, MD, PhD, MBA: Thank you for having me.

 

This transcript has been developed using AI and edited for clarity.