Propranolol: What is BLOCKing Its Clinical Investigation in Breast Cancer?

In this issue of Clinical Cancer Research, Hiller and colleagues (1) report the results of a phase II, randomized, placebo-controlled clinical trial demonstrating that 1 week of oral propranolol administration in the preoperative period reduces biomarkers associated with breast cancer metastases. In patients with newly diagnosed breast cancer, a favorable effect on the breast cancer tumor microenvironment is identified with nonselective β-adrenergic blockade. The authors present genome-wide transcriptional profiling data and demonstrate a reduction in mesenchymal polarization in propranolol-treated patients. The investigators also report a downregulation of central inflammatory and progrowth transcription factors NF-ĸB/Rel and AP-1, as well as a reduction in CREB, a downstream mediator of catecholamine signaling through the cAMP response element. Changes in neutrophil infiltration and macrophage polarization suggest propranolol may alter the tumor microenvironment in a way that could ultimately reduce breast cancer metastases. These data highlight potential antitumoral effects and offer intriguing mechanisms for repurposing an inexpensive and relatively well-tolerated medication used for various nononcologic indications, such as cardiovascular disease, migraines, and anxiety.

Previously, in vitro studies have demonstrated that β-agonists (terbutaline) result in increased breast cancer cell invasion and proliferation compared with β-antagonists (propranolol; ref. 2). Other in vitro studies have similarly shown that propranolol inhibits breast cancer cell proliferation, invasion, migration, and neoangiogenesis promoted by adrenergic stimulation (3). Breast cancer xenografts treated with propranolol result in a reduction in metastasis formation and growth, and decreases in lymphatic vessel density (4). In terms of clinical data in breast cancer, a few observational cohort studies have identified associations with the use of β-blockers and improvement in recurrence-free and overall survival. These benefits are observed primarily with nonselective β-antagonists, as opposed to selective β-antagonists. However, these retrospective studies are hypothesis-generating and have a number of limitations, limiting their applicability and generalizability.

On a mechanistic level, high levels of catecholamines have been identified in the tumor microenvironment. Sources of the catecholamines norepinephrine and epinephrine include innervating nerve fibers from the sympathetic nervous system, leakage from vascular neuromuscular junctions, and release from chromaffin cells of the adrenal medulla. Prolonged exposure to catecholamines may facilitate breast cancer development and progression through stimulation of the β-adrenergic receptors (β-AR). The β2-adrenergic receptor (β2-AR), a member of the G protein–coupled receptor superfamily, is a major mediator of catecholamine-induced effects on the malignant behaviors of cancer cells. β-AR to catecholamines activates cytosolic Gαs guanine nucleotide–binding protein to stimulate adenylyl cyclase synthesis of cAMP and downstream effector signaling. Persistent catecholamine stimulation not only impacts tumor cells but also mesenchymal and immune cells, which further amplify stimuli for invasion and metastasis, such as release of proteases, angiogenic factors, chemokines, and adhesion molecules. One could hypothesize that blockade of β-AR signaling could reduce the development of metastatic disease through modulation of epithelial-to-mesenchymal transition, as a result of reduction in Snail/Slug and Smad transcription, for example, and immune cell activity, due to M1 macrophage polarization, and an increase in dendritic cell recruitment and CD8⁺ cytotoxic T cells (Fig. 1).

In the article by Hiller and colleagues, there was a mixture of breast cancer subtypes (hormone receptor and her2 status). Much of the in vivo data previously described and retrospective clinical studies have been in triple-negative models, with relatively less supportive data in hormone receptor–positive breast cancer. However, there are also some data supporting propranolol's potential in HER2-positive (HER2⁺) disease. The β2-AR signaling pathway induces upregulation of HER2 expression and signaling, through activation of ERK. This activation can lead to the synthesis of catecholamines, and, in both in vitro and in vivo studies, catecholamines antagonize the antiproliferative effects of trastuzumab. Propranolol enhances the antitumor activities of trastuzumab and resensitizes mice injected with trastuzumab-resistant cells to trastuzumab. In accordance, high levels of β2-AR expression associate with worse responses to trastuzumab in the neoadjuvant setting (5). While the data presented in the Hiller and colleagues' study reflect a mix of various breast cancer subtypes in a relatively small patient population, it is notable that a mechanistic basis for the potential clinical activity of propranolol is observed even with a short duration of administration.

In patients with newly diagnosed breast cancer, there has been a trend toward administering neoadjuvant chemotherapy in an effort to determine the response to chemotherapy by measuring pathologic complete response (pCR) rates and residual cancer burden (RCB), particularly in patients with high-risk cancers such as triple negative, HER2⁺, or highly proliferative estrogen-driven breast tumors. Given ongoing research demonstrating that pCR and RCB can serve as a surrogate for disease-free survival in high-risk breast cancers, assessing the impact of propranolol on RCB and pCR rates in the neoadjuvant setting in patients with these tumors could be a potential next step. Given the immune modulation described in the study by Hiller and colleagues, it would be intriguing to combine propranolol with immunotherapeutic agents, such as checkpoint inhibitors, along with chemotherapy that may induce immunogenic cell death. In preclinical investigations, it would be worthwhile to test whether sequencing of these agents matter. On the other hand, the potential benefit for propranolol appears to be its antimetastatic potential, as opposed to reduction in tumor initiation or tumor size. The recommended clinical endpoint for propranolol in a phase III randomized trial would be invasive disease-free survival. However, embarking on a large, randomized trial with propranolol in an unselected population would be cost prohibitive and logistically challenging. This could be mitigated by narrowing the study cohort to a high-risk population, such as those with residual disease after neoadjuvant chemotherapy; although, the best timing of propranolol utilization still remains unclear, such as when the tumor is intact or after it has been removed.

While the results from Hiller and colleagues are thought provoking, there a number of unanswered questions. Further research in breast cancer will need to focus on tumor subtype, as well as optimal sequencing with other agents including chemotherapy and/or immunotherapy. While β2-AR expression is identified on the breast cancer cell membrane and also on cells in the tumor microenvironment, such as macrophages and monocytes, it is unclear whether β2-AR expression is a reliable predictor of clinical benefit from propranolol. Measurement of β2-AR expression should be integrated into future prospective clinical trials. The utility of selective versus nonselective β-antagonists also needs further exploration, given that clinical practice is now largely dominated by selective β-antagonist usage in patients with cardiovascular disease. The ideal dosing schedule, such as the escalated strategy in this preoperative trial based on symptom presentation, and the optimal duration of propranolol have not been established. The data presented by Hiller and colleagues highlights the biological relevance of β-adrenergic signaling in tumors, as well as the need for further clinical investigation of propranolol and its potential antitumor activity.

K. Kalinsky is a paid consultant for Biotheranostics, Genentech, Eli-Lilly, Pfizer, Novartis, AstraZeneca, and Eisai, reports receiving commercial research grants from Incyte, Acetylon, Novartis, Seattle Genetics, Cytomx, Genentech, Eli-Lilly, Pfizer, Calithera, Amgen, and Zeno, speakers bureau honoraria from Eli-Lilly, and reports immediate family members who are/have been employees of Array and Pfizer. No potential conflicts of interest were disclosed by the other authors.

This work was supported by funding from Irving Scholar Award (to K. Kalinsky).