Background: Resistance to endocrine therapies in HR-positive breast cancer is a significant challenge. The steroidal aromatase inhibitor (AI) exemestane (EXE) has demonstrated short-term efficacy in metastatic HR-positive HER2-negative breast cancer (mHR+BC) that has progressed during treatment with a non-steroidal AI. Combination strategies have not shown a survival benefit. Immunotherapy represents a promising approach as it may increase durability of responses. Low dose cyclophosphamide (CTX) has demonstrated efficacy in combination with neoadjuvant letrozole in HR+BC, conceivably by enhancing anti-tumor immune responses. Here we investigated whether EXE combined with immunomodulatory CTX could provide durable responses in heavily pretreated patients and assessed immunological profiles (NCT01963481).

Methods: Phase II trial of EXE (25mg PO daily) with CTX (50 mg PO daily) enrolled postmenopausal women (n=23) with mHR+BC who had progressed on prior endocrine therapy (including nonsteroidal AI, tamoxifen, and/or fulvestrant); prior chemotherapy was allowed. The primary endpoint was PFS (per RECIST 1.1) at 3 months; secondary endpoints were response rate, tolerability, and immune correlates. Detailed functional immune profiling of peripheral T cell subsets were performed by flow cytometry at baseline, 1, 3, 6, 9 & 12 months, with healthy donors available as controls.

Results: All 23 patients have been enrolled, and 21 are evaluable for response. Median age was 54 (range 31-77), median prior lines of endocrine therapy was 2 (1-3) and chemotherapy was 1 (0-5). The majority (15/23) had visceral organ involvement. Combination treatment was well tolerated with one grade 3 urinary tract infection but no grade 4 or 5 toxicity. An objective response was observed in 19% of patients (4/21, 1 CR and 3 PR) and an additional 33% (7/21) had SD, resulting in a 3-month-PFS of 48.5% (95% CI, 30.5-77.1). Responses were durable in all patients, lasting =/> 9 months and included patients with liver metastases.

Comparison of peripheral immune cell subsets of patients (n=16) at baseline to age/sex-matched healthy controls demonstrated an increased proportion of CD4+ memory T cells with central memory phenotype (CD45RO+CD27+, p<0.0001). When patients were stratified based on PFS at 3 months, the proportion of naïve Tregs (CD4+CD45RO-FOXP3+Helios+) at baseline was significantly lower (p=0.003) in the non-progressor group compared to patients with progression. Remarkably, when these patient groups were compared for changes in T cell subsets during treatment, the proportion of both naïve and memory Treg subsets increased from baseline to 3 months (p<0.01), but only in the non-progressor patient group. While preliminary, these findings are possibly indicative of novel predictive biomarkers.

Conclusion: EXE and CTX had a favorable safety profile with evidence of clinical activity in patients with heavily pretreated mHR+BC, including durable responses in liver and bone. Correlative studies are ongoing to identify potential biomarkers of response or resistance to therapy.

Citation Format: Kwa M, Novik Y, Oratz R, Jhaveri K, Wu J, Gu P, Meyers M, Muggia F, Bonakdar M, Abidoglu C, Kozhaya L, Li X, Joseph B, Iwano A, Friedman K, Goldberg JD, Unutmaz D, Adams S. Phase II trial of exemestane with immunomodulatory oral cyclophosphamide in metastatic hormone receptor (HR)-positive breast cancer: Prolonged progression-free survival (PFS) in patients with distinct T regulatory cell (Treg) profile. [abstract]. In: Proceedings of the Thirty-Eighth Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2015 Dec 8-12; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2016;76(4 Suppl):Abstract nr P2-11-11.