Using Treatment Response and Tumor Biology to Shape Breast Cancer Prognosis and Trial Design

David Winchester, MD, City of Hope, had an instrumental role in the creation of the breast cancer staging system published in the 8th edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual. In this interview, Dr. Winchester explains how a new breast cancer staging model developed using data from over 140,000 patients incorporates treatment response and tumor biology to better predict outcomes following neoadjuvant chemotherapy. He highlights how this model reclassifies survival expectations based on treatment response, especially in subtypes like triple-negative and HER2-positive disease, and discusses its implications for adjuvant therapy decision-making and clinical trial design. This marks a major step toward personalized prognostic staging in breast cancer.

Transcript

Your study results showed that initial clinical stage had a smaller impact on survival in patients achieving pCR compared to those with no response. How should this influence how providers think about clinical staging and risk stratification after NACT?

So the scheme is very complex. The way we’re going to apply this data to the next edition is really to assign, well, the model actually creates predicted survival for all these different combinations. And so now we have to construct a table that then matches their survival with their stage. So stage 1A is survival that’s better than 94% at 3 years. Stage 1B is survival that’s between 92 and 94% and so on and so forth.

So and, and those survival ranges are based upon survival ranges that have been historically used in the eighth edition and prior to that.

Were there any particular tumor subtypes, such as triple-negative or HER2-positive breast cancer, where the prognostic impact of response to NACT stood out more prominently?

Absolutely. You mentioned one that was dramatically different. If you look at patients with triple-negative breast cancer, for example, if they start out in stage 3, which means they have positive nodes and a fairly large tumor at least, or both, if they have a complete response, most of those patients fall into a stage 1 category, which is pretty dramatic.

On the other hand, if they stay in the same stage or if their T and their N categories remain the same, their prognosis is dramatically worse. Their survival drops by as much as 60% at 3 years. So you see a big range with the triple negatives, you see a big range with the HER2-positive tumors, the luminal A breast cancers, the ER positive, PR positive, HER2-negative tumors. Their responses are not as important for the prognosis. They still make a difference. The response still makes a big difference, but not as big.

And the other thing about the HER2-positive and the hormone-positive breast cancers is that they’re going to continue to undergo treatment longer, whereas a triple-negative breast cancer patients are essentially done after 6 months. But the ER-positive tumors especially for the stage 3 breast cancers, they’re going to see probably 10 years or longer of endocrine therapy. Their impact on survival is not as affected as it is for the chemotherapy-sensitive tumors, more sensitive tumors.

How might your proposed staging system improve decision-making for adjuvant therapy or clinical trial eligibility after neoadjuvant treatment?

Yes, that’s a great question. And that’s kind of one of the objectives of having the stage assignment, is that we can then look at patients who undergo neoadjuvant chemotherapy after surgery and understand who’s going to likely fail or relapse. And then we can invoke the adjuvant therapy for those patients after surgery too. And the same is true, you know, it’s just about being more precise in our treatment.

The study didn’t address the utility of adjuvant therapy after surgery, after neoadjuvant chemotherapy, and it didn’t address neoadjuvant endocrine therapy either. So we still have some gaps of treatments approaches and helping us to understand what comes next. But having a stage now that we can assign patients after neoadjuvant chemotherapy and surgery will allow stratification for better clinical trials, having that stage assignment.

The patients that are in this publication were treated from 2010 to 2018. A lot of those patients were treated with what we consider outdated therapy. Breast cancer therapy continues to improve rapidly, which is why we have to look at smaller groups of data that are more condensed in the time of their diagnosis to understand how to measure their outcomes. But I think our next generation is going to include adjuvant therapy too.

Looking ahead, what further validation steps or external studies would you like to see to support broader adoption of this postneoadjuvant staging system in clinical practice?

We’d like to validate this with another database outside of the United States. The trouble is that this study consists of 140,000 patients with complete records. There’s no such thing anywhere else in the world that we can do that with, but we’d hope to collaborate with other European databases to validate our work. As I mentioned, we don’t have any way to stage patients if they are treated with neoadjuvant endocrine therapy, which is not chemotherapy, but it’s drugs that specifically target the hormone receptors. And so that’s a growing percentage of our patients as well. And there’s been since 2018, 2010, you know, some exciting new endocrine therapies too that have become very important.

Are there any final thoughts you’d like to add?

The other point I would make is that this is not a perfect staging system. It’s an attempt to include, you know, roughly 60 to 80, or 100,000 patients a year that are treated with neoadjuvant chemotherapy. But we need to continue to look at important variables in future models and continue to adapt our models according to those important variables as they emerge and probably to be more specific in terms of categorizing the treatment for these patients too.

The point is, it’s very important to continue collecting data on a national basis because finding differences like we found in this study require a lot of patients, a lot of data, a lot of participants.