The survival and growth of HER2 + breast cancer is highly reliant upon signaling through the HER2/PI3K/AKT pathway. Significant preclinical evidence points to the role of the related receptor HER3 in mediating resistance to HER2 inhibitors. Shortly following the initial inhibition of HER2 there is dynamic upregulation of HER3 expression through a variety of pre and post-translational mechanisms. The increased HER3 expression allows HER2:HER3 heterodimer formation that in turn allows for persistent signaling through the HER2/PI3K/AKT pathway. This resistance mechanism is so robust that in one in vitro study, 100x the clinical concentration of the HER2 inhibitor lapatinib is required to completely eradicate HER2 pathway signaling (16), and dual treatment with trastuzumab and lapatinib was unable to completely eradicate HER3 signaling in another(30).
In this preclinical study, we sought to demonstrate that HER3 PET imaging with Ga-68 HER3P1 could image the changes in HER3 expression that occur rapidly during treatment with the HER2 inhibitor lapatinib. These data show that HER3 PET imaging can serve as a non-invasive method of assessing dynamic changes in HER3 expression, and thereby has the ability guide therapeutic regimen choice. We had previously reported on antibody-based HER3 PET imaging, showing that the upregulation of HER3 in triple negative breast cancers in response to AKT-inhibition can be non-invasively assessed (31). We have also previously reported on HER3P1 imaging, showing that imaged Ga68 HER3P1 uptake correlates closely with HER3 expression as assessed by Western blot (23).
We extend on our prior findings by comparing Ga-68 HER3P1 imaging in two HER2 + breast cancer mouse models known to be resistant to HER2 inhibition with lapatinib (MDA-MB-453 and HCC-1569). While both HER2 + cell lines we selected for imaging are resistant to HER2 inhibition, only MDA-MB-453 demonstrates significant increase in HER3 expression. With Ga-68 HER3P1 imaging, we find that Ga-68 HER3P1 uptake increases in the MDA-MB-453 tumors nearly three fold after two days of treatment with lapatinib relative to pre-treatment imaging but remains the same in HCC-1569 tumors.
Of note, this observed increased in Ga-68 HER3P1 uptake in the MDA-MB-453 model in vivo was significantly greater than the increase in HER3 expression observed via Western blot in the same cell line. These data may speak to the multifactorial handling of HER3 cycling and processing – whereby increased trafficking and recycling of the HER3 receptor could contribute as much or more to increased signaling through HER3 as increased expression(18). This difference in HER3 increase between our two studied cell lines in response to HER2 inhibition suggests different mechanisms of HER2 inhibitor resistance, and indeed HCC-1569 resistance to HER2 inhibition is rather mediated by PTEN loss(28), such that an increase in HER3 expression would not be expected. While the sum of our imaging data highly suggests that increased Ga68-HER3P1 uptake in MDA-MD-453 is reflective of increased HER3 expression and trafficking, we cannot entirely exclude the possibility that an increase in permeability also contributes to increased uptake, although were this the case would anticipate that HCC-1569 tumors would also demonstrate increased uptake post lapatinib.
We additionally hypothesized that tumor cell lines that upregulate HER3 expression in response to HER2 inhibition would be sensitive to additional HER3 inhibition, whereas tumor cell lines that do not increase HER3 expression under similar treatment conditions would not. In this context we suggest the potential therapeutic relevance of Ga68-HER3P1 PET imaging by showing that the HER3 upregulating MDA-MB-453 cell line is sensitive to the combination of HER3-downregulating siRNA and lapatinib, whereas HCC-1569 cells are insensitive. We suggest with these findings that HER3P1 PET imaging could be used to identify patients whose tumors utilize dynamic HER3 upregulation to resist HER2 inhibition and thereby those patients most likely to benefit by the addition of a HER3 inhibitor to their therapeutic regimen.
It is not known what percentage of HER2 + breast cancers upregulate HER3 in response to HER2 inhibition. A study of patients treated with two weeks of neo-adjuvant lapatinib prior to surgery found a wide-range of changes in HER3 mRNA post treatment compared to pre-treatment biopsy, from − 28% to + 78% across 17 patients with HER2 + breast cancers(32). This study found that HER3 mRNA increased non-significantly with response as measured by Ki-67 staining of resected specimens. Recognizing that mRNA is not the only mechanism by which cell-surface HER3 expression is increased, this small study sample provides indirect evidence that changes in HER3 expression may be a resistance mechanism employed by a subset of tumors to resist HER2 inhibition. In an additional study matched pre- and post-treatment biopsies in 8 patients with HER2 + tumors treated with 2 weeks of lapatinib showed upregulation of HER3 post-treatment by 135% as assessed by IHC (20).
The strong preclinical evidence for the role of HER3 in mediating resistance to HER2 treatment led to the development of several HER3-specific antibodies that have been evaluated preclinically and entered clinical trials(33–35). These antibodies have been evaluated as either single agents, in combination with chemotherapy, or with HER2 or EGFR inhibitors in several different oncologic indications. Most of these studies have had disappointing clinical results, although a recent Phase I study of U3-1402, a HER3 antibody – topoisomerase inhibitor conjugate, has reported positive preliminary data(36).
That the clinical trial results appear so at odds with the preclinical evidence may speak to two critical points that argue for the value of a noninvasive method of assessing tumoral HER3 expression: A) Dynamic increase in HER3 expression likely mediates HER2 inhibitor resistance in only a subset of HER2 + breast cancer patients; B) It has been very challenging to identify this subset of patients in whom this resistance process is active – as the dynamic increase in HER3 is seen only by comparing pre-treatment HER3 expression to on-treatment expression. To identify this process in action requires both pre-treatment and on-treatment biopsies, which can be challenging to acquire. HER3 PET imaging may significantly improve our ability to observe dynamic upregulation of HER3, acting as a noninvasive predictive biomarker in future trials.
The role of antibody-based HER3 imaging agents have already undergone initial exploration. In a small clinical study, a Zr89 labelled anti-HER3 mAB (GSK2849330) was used to assess biodistribution and target engagement of antibody (37). Imaging with antibody based agents coupled with long-lived radionuclides such as Zr89 or Cu64 limits the ability to assess changes in HER3 expression due to the agents long half-lives. By contrast, HER3 PET imaging with a short-lived radiopharmaceutical such as Ga68-HER3P1 allows for repeat non-invasive assessment of tumoral HER3 expression over a short time scale. We can assess total HER3 expression across all tumor sites prior to therapy and then identify changes in HER3 expression that occur with targeted inhibition of HER2 or other targeted inhibitors within days of therapy initiation. The ability to noninvasively assess changes in HER3 expression across all tumor sites could allow for routine assessment of changes in HER3 without subjecting patients to repeat invasive biopsy, helping to readily identify those patients in whom a HER3 inhibitor may be of real clinical benefit. Our in vitro data demonstrating the benefit of additional HER3 knockdown only in cell lines that upregulate HER3 in response to HER2 inhibition with lapatinib supports this point. Of note, as we have only evaluated the role of Ga-68 HER3P1 imaging with lapatinib, it remains unknown if it would have a similar role in evaluated dynamic in increase in HER3 expression mediated by other HER2 inhibitor therapies such as trastuzumab or pertuzumab, especially considering the possible role of pertuzumab interfering with HER family heterodimerization(38). Additionally, it is worth noting that we have only evaluated changes in HER3 that occur rapidly with HER2 inhibitor initiation. This study does not evaluate the role that HER3 PET imaging may have in understanding mechanisms of resistance that occur with long-term treatment, only intrinsic resistance mechanisms.
We believe that Ga68-HER3P1 PET imaging could enable a paradigm shift in how HER3 inhibitor therapies are evaluated. The efficacy of HER3 inhibitor addition to targeted therapy should be evaluated in the context of a pre- and on-treatment HER3 PET imaging. Our preclinical study provides evidence that patients demonstrating an increase in on-treatment HER3 expression are those in whom the HER3 up-regulating resistance mechanism is most active and identifies those patients most likely to benefit by HER3 inhibitor addition.