Clinical Pathway Development for Early-Stage and Metastatic HER2-Positive Breast Cancer

In this new series, we speak with cancer care practitioners about pathways being used in their practices, how they are being applied in a particular disease state, and what challenges persist for treatment decision-making.

In this first installment, we spoke with Lee Schwartzberg, MD, FACP, about clinical pathway development for early-stage and metastatic HER2-positive breast cancer. Dr Schwartzberg serves as executive director at West Cancer Center, professor of medicine and chief of the division of hematology/oncology at the University of Tennessee Health Science Center, and president/CMO for Vector Oncology. In addition, Dr Schwartzberg was the founding editor-in-chief of the journal Community Oncology and currently serves as the editor-in-chief of the Practice Update Oncology website. Dr Schwartzberg is also a member of the board of directors for the National Comprehensive Cancer Network. His major research interests are new therapeutic approaches to breast cancer, targeted therapy, and supportive care.

JCP: How are clinical pathways developed and used at your cancer center?

Dr Schwartzberg: I would say we are midway in the process between what we call care plans and full-blown clinical pathways. We started about 7 or 8 years ago with developing specific care plans for each disease, and, over time, what we did was create priority care plans and winnowed down the number of choices based on stage and different characteristics of each tumor.

With regard to breast cancer, we have a number of options for early-stage breast cancer based on the biomarkers, so whether a patient is ER-positive HER2-negative, triple negative, and HER2-positive. The same is also true for metastatic breast cancer. It’s a little broader there, particularly as it relates to HER2-positive, but again, it’s the same thing. And then, in other words, we define by phenotype which are the priority care plans.

What we have done over the last few years is try to create specific (one or sometimes two) priority care plans for every given disease state and, in breast cancer, for first- and second-line treatment. We have found it difficult without a full-blown algorithmic guideline to go beyond second-line in terms of priority, although we track that every quarter in terms of the compliance with the priority care plans. We will eventually move to an algorithmic approach, but at the moment it’s a narrow pick list. 

Currently, how are you approaching prioritization of therapies?

We have a monthly meeting of a clinical practice and quality committee. There are subcommittees that create care plans, and there are subcommittees that are disease-specific. The disease-specific subcommittees get together once or twice a year, or sometimes more frequently if there’s a practice-changing recommendation, and update the priority care plans in that fashion.

Within the pathway for HER2-positive patients, is there one pathway for HER2-positive or do you further stratify patients from there?

We [have separate care plans] for the adjuvant/neoadjuvant and for the metastatic patients. For the neoadjuvant or adjuvant, the regimens are essentially identical with potential caveats. But in general, for any HER2-positive patient that is stage T1B or greater, we will recommend adjuvant or neoadjuvant chemotherapy. We basically have 3 care plans that one can use in that setting. One is for low-risk patients. Those are patients with T1 lesions N0; that can either be clinical or pathologic. For those patients, the priority care plan is weekly paclitaxel times 12 and trastuzumab. For patients who don’t fit those criteria, we have 2 care plans: AC [adriamycin and cyclophosphamide] followed by paclitaxel, Herceptin [trastuzumab] and PERJETA [pertuzumab]; or Taxotere [docetaxel], carboplatin, trastuzumab, and PERJETA. And that’s up to the individual physician, so we don’t mandate one specific neoadjuvant or adjuvant protocol. We have those 3 choices.

Can you talk a little bit more about the benefits of neoadjuvant therapy?

Yes, absolutely. This has been a dynamic field. In general, the benefit of neoadjuvant chemotherapy is to shrink tumors and make larger tumors more amenable to breast-conserving therapy. That’s probably the most evidence-based reason to do it. But, of course, we also know that pathologic complete response [PCR] is a very good biomarker for long-term outcome. That is true in the HER2 subset as well as in some other subsets. And it is true in both the ER-positive HER2-positive group as well as the ER-negative HER2-positive group, and each of those are about 50% of the HER2 population.


We do see more pathologic complete responses to chemotherapy in the ER-negative HER2-positive group; roughly 60% to 70% of patients will achieve a PCR as opposed to 30% to 35% with the ER-positive HER2-positive or the so-called triple positive. Nonetheless, the outcome is superior for both groups if they achieve a PCR. So, that’s another reason to do it.

A third reason to do it is to make an individual patient decision for adjuvant therapy based on whether or not the patient receives a PCR. It’s a controversial area, and people are doing trials now about so-called de-escalation to determine whether a patient who has gotten a PCR needs to continue the full course of therapy. But, as of now, there is no data to support that. So, the standard approach in our care plans is that if one started with dual-antibody therapy for the neoadjuvant setting, then one would continue that for the adjuvant setting as well. In other words, the standard is to complete 1 year of trastuzumab or trastuzumab and pertuzumab. 

One other comment about the adjuvant and neoadjuvant: in the ER-positive patients, particularly the node-positive patients, we have also recommended the use of extended adjuvant anti-HER2 therapy. So, after one finished the year of trastuzumab, with or without pertuzumab, neratinib is an option as well.

That brings up the issue of how new therapeutic approaches are incorporated into the pathway. What level of evidence is required for you to make a determination that the clinical pathway needs to be updated?

I think in HER2-positive breast cancer, because there’s now a wealth of data, the evidence to change practice must be based on phase 3 trials. Even though the subset is 15% to 20% of patients, particularly in the metastatic setting, because we’re doing so much better in the adjuvant and neoadjuvant setting now, we really have options. I didn’t mention before the pathway for metastatic disease, but this is very well established. Based on the CLEOPATRA study, patients who are appropriate candidates should receive taxane or trastuzumab and pertuzumab in the first-line metastatic setting. We have data saying that’s the best regimen.

Now, the nuances there are if you had a patient that had previously been treated with trastuzumab and pertuzumab and relapsed relatively early, they might not be a candidate for the CLEOPATRA approach that I just outlined. But that’s a small group of patients. If patients had not previously been treated, or more likely had finished their adjuvant therapy 1 year or longer ago, the taxane and dual-antibody approach is appropriate.

Then, the second-line therapy is also well established for metastatic or HER2-positive breast cancer, which is the use of Kadcyla (T-DM1), the antibody drug conjugate. Kadcyla has been tested in the first-line setting and was not better with pertuzumab than taxane, so that remains the best second-line option. So, we have very good data for the first and second lines.

For the third line, it’s a little bit less clear, but the best data we have is with the combination of lapatinib and capecitabine. But many people also continue trastuzumab and use a different chemotherapy partner beyond second-line therapy, and that hasn’t been tested lately. So, in that setting, we have a larger list of options to select from, and we don’t have a priority regimen to recommend because there’s no phase 3 data to support one over the other. 

One of the challenges of using clinical pathways is the struggle between needing to standardize treatment decisions across the patient population, but also optimize care for individual patients. How do you balance these priorities in a disease like breast cancer?

It’s an ongoing challenge, and as molecular biomarkers continue to be defined, we are seeing more and more subgroups of patients that require specific therapy. It really hasn’t happened that much in breast cancer beyond the 3 biomarkers that we use: ER, HER2 and triple negative. Those are still the 3 large categories, but there are emerging biomarkers and they may inform some of the decisions. But, for example, in lung cancer now we have multiple groups that might be treated differently.

In my experience, the biggest issue between individualizing therapy and using a pathway is beyond first-line or sometimes second-line therapy where the data is pretty clear because clinical trials have specifically addressed that viewpoint. And as I mentioned, for HER2-positive breast cancer, we are very comfortable with first- and second-line therapy and even third-line because of the data that has been previously generated. However, breast cancer patients, particularly HER2-positive breast cancer patients, frequently get 7, 8, or more lines of therapy in the course of their disease. And there’s really a very limited data base, mostly small phase 2 trials, to support one over the other. In fact, there’s no data to support one over the other. So, at the moment we can’t recommend care plans for later lines of treatment. We do measure compliance with care plans, and we believe that a target of 80% to 90% compliance with first-line therapy is appropriate, but as the line of therapy gets deeper, by definition the compliance goes down and that’s probably appropriate.

In the future, I think what we will be doing in our own practice, and more broadly across the oncology community, is using real-world evidence to help refine these data. If we can aggregate big data and look at large groups of patients that are treated according to the standard practice and see what the outcomes are, that’s probably the only way we’re going to really get enough evidence to say that Drug X is better than Drug Y in third-line therapy, because we’re just not going to have those clinical trials invested in a more formal prospective fashion.

What are some other challenges that you face in optimizing decision-making in your practice?

There are many factors that go into therapeutic decision-making, and that’s why it’s important to make it a shared decision-making experience with the patient. For instance, many women would prefer an agent that doesn’t cause hair loss, so if it’s possible, even though the first-line therapy would be a taxane, many of them will say, “Is there anything else I can do as an alternative?” Capecitabine plus trastuzumab is a non-alopecia-causing regimen, and one could say that there might not be a difference, or patients would be willing to take a slight decrease in efficacy for that reason.

Another instance would be a patient who had preexisting neuropathy, and one would not want to use a drug that caused neuropathy, because we want to preserve quality of life. This is something that pathways don’t really consider. Pathways are usually defined on first, efficacy, second, toxicity, and third, cost. But toxicity is usually thought of in terms of acute toxicities, and there are other toxicities that the patient may be more interested in.

The cost is an increasing issue for us, particularly for the orals with the copays. While there are good assistance programs, they are very difficult to navigate, and sometimes the copays are still expensive. So, I personally am seeing more issues with cost as it relates to the individual patient obligation with regard to the orals, and that does play a role in metastatic HER2-positive breast cancer as well. But cost is only a factor in terms of a priority care plan after we consider efficacy and toxicity. If both [treatment options] are equal, then we will look for the lower cost alternative.