PI3K-Targeting Agents May Help Overcome Resistance in Breast Cancer Treatment

Sara Hurvitz, MD

Director, Hematology/Oncology Breast Cancer Program

Assistant Professor, Department of Medicine

UCLA Jonsson Comprehensive Cancer Center

Los Angeles, CA

Through a variety of research roles, Sara Hurvitz, MD, is exploring novel targeted therapies for the treatment of breast cancer, including developing and implementing clinical trials. Her interests include the phosphatidylinositol 3-kinase (PI3K) signaling pathway.

1. Please describe your research as it relates to PI3K signaling.

Multiple inhibitors of the PI3K pathway currently are in preclinical and clinical development. We have had the opportunity to evaluate several of these inhibitors in our laboratory, where we have tested their in vitro activity utilizing our cell line panels (which include over 50 breast cancer cell lines). This in vitro work allows us to gauge whether particular subtypes of breast cancer respond differentially to inhibition of the PI3K pathway, whether combinations of PI3K inhibitors with other targeted agents are synergistic, and whether there are any biomarkers that may predict response to these therapies.

In addition, tumor-bearing mouse models are used to evaluate the effectiveness of these therapies among the breast cancer subtypes. As an example, Neil O’Brien, PhD, in our lab evaluated the activity of several inhibitors of the PI3K pathway in human epidermal growth factor receptor 2 (HER2)-positive, trastuzumab-resistant tumors using xenograft models and demonstrated that inhibiting PI3K in combination with HER2-inhibition overcomes trastuzumab resistance independent of feedback activation of Akt (O’Brien NA et al. Cancer Res. 2012;72[24 suppl 3; abstr P4-08-01]; manuscript submitted). This type of preclinical work done by our lab and many others is critically important since it leads to the rational design of clinical trials in the appropriate patient populations to evaluate these agents.

I am interested in evaluating PI3K inhibitors in combination with other biologically targeted agents. While clinical trials are already under way (and some have been reported) that evaluate the activity of combined PI3K-pathway inhibitors with endocrine therapy or HER2-targeted agents, I am particularly intrigued by the potential activity of PI3K inhibitors in combination with cyclin D kinase (CDK) inhibitors or PARP-inhibitors based on promising preclinical activity observed in certain breast cancer subtypes.

2. What is the most exciting recent development in PI3K pathway targeting?

One of the major challenges to improving the long-term outcomes of patients with breast cancer is treatment resistance. The PI3K pathway is frequently altered in breast cancers and has been implicated in the development of resistance to hormonal approaches and anti-HER2 therapies. The potential for mammalian target of rapamycin (mTOR) blockade in combination with antiestrogen therapy to reverse resistance to endocrine therapy has already been validated in the BOLERO-2 study in which everolimus was combined with exemestane and shown to significantly improve progression-free survival (PFS) for postmenopausal women with estrogen receptor (ER)-positive aromatase inhibitor-pretreated metastatic disease, compared with exemestane alone.

Similarly, in the BOLERO-3 study, the addition of everolimus to chemotherapy and trastuzumab improved PFS in heavily pretreated HER2-positive metastatic disease, again supporting the hypothesis that inhibition of the PI3K/Akt/ mTOR pathway can reverse treatment resistance.

An exciting possibility is that some PI3K pathway inhibitors appear to cross the blood-brain barrier. As such, they may be promising in the prevention or treatment of brain and leptomeningeal metastases. In addition, new combinations of PI3K pathway inhibitors with other targeted therapies (such as PARP-inhibitors or CDK4/6 inhibitors) may provide synergistic benefit based on preclinical studies. The potential role of PI3K pathway inhibition in these settings remains to be validated clinically.

3. What are the relative merits of pan-isoforms versus isoform-specific PI3K inhibitors?

At this point, it is unclear whether pan-isoform or isoform- selective inhibitors will be associated with a better therapeutic index (ie, efficacy-to-safety ratio). If we were able to identify which, if any, part of the PI3K pathway was driving the growth and survival of a particular tumor, it would be theoretically possible to choose the best inhibitor for that tumor.

For example, studies have demonstrated that activating mutations in the genes coding for p110α (including activating mutations in exon 9 and 20 of PIK3CA) are common in breast cancer. Therefore, in those cancers, targeting p110α may be more important. Other evidence points to the importance of p110β in tumors where phosphatase and tensin homolog (PTEN) is downregulated. In those tumors, p110β inhibitors may be more effective.

At this point, however, there has been no clinical validation of these theories. Thus, our ability to select the best inhibitor for a particular patient is not yet a reality. Another possible advantage of isoform-selective PI3K inhibitors is that they may be better tolerated. The p110α isoform has been shown to play more of a role in glucose metabolism; thus, it is possible that inhibiting p110α may lead to more toxicity in terms of insulin resistance and hyperglycemia. Conversely, inhibiting p110β may lead to less toxicity in terms of insulin resistance. Again, these are theoretical advantages that require validation clinically.

4. What are the most significant hurdles that must be overcome in order to allow optimal use of these agents?

Those agents that target a molecular abnormality unique to cancer cells logically have the potential for the least toxicity and most benefit. Agents that target a pathway (such as the PI3K pathway) that is essential for normal cellular functioning, however, may be associated with important toxicities and, as such, need to be carefully evaluated.

In addition, as more and more targeted agents are developed, it will be critical to evaluate their relative benefits in different types of cancer. This will be best achieved by evaluating these drugs in neoadjuvant trials that allow serial tissue testing pre- and post-treatment, or in the metastatic setting where tumor tissue from metastatic sites is sampled and evaluated for molecular alterations that may predict benefit from these agents.