DE-ESCALATING TREATMENT FOR HER2-POSITIVE EARLY BREAST CANCER

Abstract

HER2-positive breast cancers are defined by amplification of the cerbB2 gene and overexpression of its protein HER2. This subtype of breast cancer is aggressive, grows rapidly, and metastasizes early. However, new treatments targeting HER2 combined with chemotherapy have dramatically altered the course of patients harboring this type of breast cancer. Traditionally, these patients have been treated with aggressive combination chemotherapy, combined with trastuzumab, an antibody that blocks HER2 signaling and also induces antibody-dependent cellular cytotoxicity. Trastuzumab combined with newer drugs that also block HER2 signaling has been shown to be effective even in the absence of chemotherapy. Significant pathologic complete response (pCR) rates have been observed with dual targeted therapy in preclinical studies and in neoadjuvant trials in patients without chemotherapy. Before this de-escalation strategy can become part of routine clinical care, biomarkers need to be developed that identify patients who either do or do not respond well to such therapies.

INTRODUCTION

Breast cancer mortality has been declining for the past 30 years due to earlier diagnosis and more effective treatment. For therapeutic purposes, breast cancers can be divided into three subtypes: estrogen receptor-positive (ER+), HER2-positive (HER2+), and triple negative breast cancer (TNBC). ER + breast cancer is the most common subtype and represents 70% of all cancers. The hallmark treatment of this subtype is endocrine therapy to either block ER or to reduce estrogen levels. About 15% of breast cancers are HER2+, and about 50 to 60% of these also express the estrogen receptor (1). Traditionally, HER2 + patients have been treated with chemotherapy and therapies that target HER2. Another 15% of breast cancers do not express estrogen receptors, progesterone receptors, or HER2, and they are called triple negative breast cancer, or TNBC. Chemotherapy remains the mainstay of treatment for this group of tumors.

HER2 is a member of a tyrosine kinase family of four membrane receptors (2). These receptors upon ligand binding act coordinately to stimulate downstream signaling through the PI3 kinase, MAP kinase, and other pathways. Trastuzumab is a monoclonal antibody that binds to the external domain of HER2, inhibits signaling, and activates antibody-dependent cellular cytotoxicity (ADCC) (3). The standard treatment for these tumors has been aggressive chemotherapy combined with trastuzumab, which has resulted in significant improvement in patient survival (4).

Oncogene addiction is a concept whereby the cancer is driven by a single powerful driver pathway such as HER2. Other redundant survival pathways become relatively inactive because they are not needed for cell survival. Potent inhibition of this driver pathway should kill the cell. We hypothesized that HER2 + breast cancers might be addicted to HER2 signaling and that complete blockade of the pathway with targeted therapy might kill the cell without any chemotherapy at all. We therefore began a series of studies to determine if therapies targeting various components of the HER family of receptors might completely block signaling and cause tumor cell death using both preclinical models and neoadjuvant trials in patients.

ACTIVATION OF THE HER PATHWAY

The four members of the HER signaling family include HER1, HER2, HER3, and HER4 (11). Ligands activate signaling by binding to HER1, HER3, and HER4 but not to HER2, which does not have a ligand (2,5). HER2 is then activated by heterodimerization with another member of the family or by homodimerization when abundant HER2 receptors are present. Receptor dimerization in the membrane then results in activation of a downstream signaling cascade including PI3K/AKT and RAS/MAPK among others. This series of phosphorylation steps transmits the signal from the membrane to the nucleus to activate transcription factors that enhance cell proliferation, migration, and cell survival. Trastuzumab binds to the external domain of HER2, blocks signaling caused by HER2 homodimers, and induces ADCC (3,6). While very effective, trastuzumab results in incomplete blockade of the receptor family because it is ineffective in blocking HER1/HER3 or HER1/HER4 heterodimers and is a weak inhibitor of HER1/HER2 or HER2/HER3 heterodimers. Thus, incomplete blockade by trastuzumab may cause resistance in some patients' tumors through an escape mechanism generated by other members of the family. Several other drugs have been developed to block HER2 signaling in different ways. Small molecule tyrosine kinase inhibitors such as lapatinib, neratinib, and afatinib bind to the tyrosine kinase domains of these receptors to block signaling (3,7). All three of these drugs not only block HER2 but also HER1; combined with trastuzumab, they would be expected to more completely block signaling. Pertuzumab is yet another agent that binds to the heterodimerization domain on HER2 and prevents homo- and heterodimerization (8,9). This drug is also effective when combined with trastuzumab to more completely block signaling through the pathway. We began to explore the possibility that combinations of these agents might be more effective than trastuzumab alone in causing more complete inhibition of the signaling pathway.

PRECLINICAL RESULTS

We used xenograft models of various HER2 overexpressing human breast cancer cell lines to investigate single agents and combinations of these HER2-targeted therapies. Gefitinib, targeting HER1, pertuzumab, trastuzumab, and lapatinib were used alone or in combination (10-12). In these models, combinations of these agents were much more effective than single agents and rapidly eradicated tumors in athymic mice. The most effective combinations were trastuzumab combined with pertuzumab, and gefitinib or trastuzumab combined with lapatinib (blocks HER1 and HER2). In tumors that were estrogen receptor-positive, endocrine therapy with the antiestrogen tamoxifen or with estrogen deprivation was required for maximal anti-tumor effects. In other studies, we showed that lapatinib, when given intermittently or in lower doses, was just as effective as continuous therapy at full dose (11). These studies demonstrated that in in vivo models of HER2-amplified breast cancer, dual or triple anti-HER2 therapy could rapidly eradicate tumors in mice suggesting that these combinations should be studied in patients.

CLINICAL TRIALS

We chose the neoadjuvant setting (therapy given for several months prior to surgery) to test the strategy that dual HER2-targeted therapy would be just as effective in patients as it was in mouse models without chemotherapy. We included endocrine therapy with aromatase inhibitors (AI) in postmenopausal patients or ovarian suppression plus an AI to induce estrogen deprivation if tumors were also ER+. We completed two trials in collaboration with the Translational Breast Cancer Research Consortium: TBCRC 006 and TBCRC 023 (13,14). The first trial of 64 patients with a median breast tumor size of more than 6 cm were given trastuzumab and lapitinib plus estrogen deprivation therapy for 12 weeks. The second trial was a randomized phase 2 trial of 94 patients in which one group was treated with the same regimen for 12 weeks and the other for 24 weeks hypothesizing that ER-positive tumors were slower to regress and would require longer therapy. Complete remission at the time of breast surgery was observed in 15–27% of the patients depending on the subtype with the lower rate in ER+/HER2 + tumors. The pCR rate was higher in the 24-week arm for the ER + subset but not in the ER- group. A third clinical trial conducted by the SOLTI Group in Spain, called PAMELA, was a 150-patient phase 2 trial of the same regimen given for 18 weeks (15). The complete response rate seen at surgery was 31% overall, and like the other trials, was higher in the ER-/HER2 + group than in the ER+/ HER2 + patients. A fourth trial called NeoSphere was a 107-patient phase 2 trial using pertuzinab in combination with trastuzumab but without endocrine therapy (16). Overall, the pCR rate was somewhat less at 17%, but again much higher in the ER- subset. These four trials of dual HER2-targeted therapy without chemotherapy showed similar pCR rates compared to older studies of trastuzumab alone with combination chemotherapy. Important questions that remain are: (1) Why don't all patients who overexpress HER2 respond well to targeted therapy alone?; (2) Can patients who will respond well be identified a priori?; and (3) Can the mechanisms of resistance to targeted therapy in the other HER2 + tumors be identified and strategies to block resistance be developed? Note that all the patients on these early phase trials went on to receive standard chemotherapy plus trastuzumab following surgery so no long-term outcome data are available.

BIOMARKERS OF SENSITIVITY AND RESISTANCE

Why aren't all tumors amplified for the HER2 gene and expressing high levels of the protein inhibited by HER2-targeted therapies? The definition of HER2 positivity is based on the ratio of HER2 gene copy number to the number of copies of chromosome 17 using a centromere probe. The cutoff ratio of 2 is somewhat arbitrary and is based on studies in which patients received chemotherapy in addition to trastuzumab therapy (17). We wondered whether low levels of HER2 amplification might identify patients whose tumors were not driven by HER2 and therefore would not respond as well to HER2-targeted therapy in the absence of chemotherapy. About 30% of HER2 + tumors also have mutations in the PIK3CA gene, which would then activate the pathway inside the cell rendering such tumors resistant to therapies targeting the HER receptor family in the membrane. Downregulation of the PTEN tumor suppressor gene would have a similar effect by relieving the block on PI3K. In TBCRC 006, we measured HER2 amplification by FISH, PIK3CA mutation status, and PTEN downregulation by immunohistochemistry in the baseline biopsies of the patients (18). Patients had to have a HER2 FISH ratio of 2 or greater to be entered into the trial; in our analysis, no patient had a pCR with a HER2 ratio less than 4 or a HER2 gene copy number less than 10. Still, many patients who did have a HER2 FISH ratio greater than 4 failed to have a pCR. However, the majority of these patients had dysregulation of the PI3 kinase pathway via activating mutations in PIK3CA or downregulation of PTEN (18). These studies suggest that the level of HER2 amplification, assays for PIK3CA mutations, and PTEN status might help to select patients for future studies of this de-escalation approach. Similar data were collected when patients' tumors were subjected to transcriptional profiling to identify intrinsic subtypes. The HER2-enriched subtype or tumors expressing high levels of HER2 mRNA resulted in a much greater pCR rate (15). These results are currently being validated in a larger subset of patients via collaboration between our group and Aleix Prat in the SOLTI Group in Spain.

CONCLUSIONS

These cumulative data from the preclinical studies and neoadjuvant clinical trials suggest that a significant subgroup of patients with HER2 + breast cancer might be safely treated with dual HER2-targeted therapies in the absence of chemotherapy thereby avoiding considerable toxicity. These HER2-targeted agents are associated with significantly less toxicity than chemotherapy and would be welcomed by patients. We are currently developing an algorithm that includes assessment of HER2, PIK3CA mutation, and PTEN status, and the presence or absence of tumor-infiltrating lymphocytes, which have also been shown to predict response to HER2-targeted therapy. The algorithm will be prospectively tested in a randomized neoadjuvant clinical trial of target therapy alone versus the standard approach of chemotherapy plus trastuzumab in relatively low-risk HER2 + breast cancer.

DISCUSSION

Moore, New York: Thank you for this spectacular demonstration that we can cure a very aggressive breast cancer now without chemotherapy. You know our patients will be enormously happy with this news. Our patients also ask us so much about the immune system and how they can activate their own immune systems. What do you think of these tumor-infiltrating lymphocytes that may predict for a better response?

Osborne, Houston: Well, first of all, I love to see patients whose tumors have tumor-infiltrating lymphocytes. When our pathologist calls me and says this patient has lots of lymphocytes in the tumor, I am happy because those patients tend to respond to chemotherapy or HER2-targeted therapy much better. I think tumor-infiltrating lymphocytes will add to the predictive ability of PI3CA mutations and the HER2-level, and I think that further activation of the immune system with checkpoint inhibitors will allow us to de-escalate the treatment even more. So I'm very excited about that.

Mackowiak, Baltimore: I probably missed this so forgive me—did you mention toxicity?

Osborne, Houston: I didn't, but it's very well tolerated. Trastuzumb by itself has almost no side effects that impact the patient's quality of life. It can suppress myocardial contractility, but when you temporarily stop it, the contractility comes right back. When you combine it with doxorubicin chemotherapy, a drug that also affects the heart, it causes problems so we avoid that combination. Lapatinib, another drug that we combined with trastuzumab, is well tolerated even in elderly patients I have treated. The side effects are mainly a skin rash on the face and sometimes diarrhea, so you have to lower the dose a little bit. Both of those are very well tolerated, especially compared to chemotherapy.