RE: Systemic or vaginal hormone therapy after early breast cancer: a Danish observational cohort study

We express concern over the conclusion drawn based on this retrospective study of breast cancer (BC) patients treated from 1997 to 2004 that those on aromatase inhibitor (AI) therapy who use vaginal estrogen treatment (VET) have higher recurrence rates, especially because this was not seen in women receiving systemic menopausal hormone therapy (MHT) (1). Specifically, we address risk stratification of the patient population, changes in oncology practice, and changes in available formulations of VET. Major advances have occurred over the nearly 2 decades since the study period. Vaginal dryness and dyspareunia plague many BC survivors, adversely affecting quality of life and jeopardizing adherence to AI therapy.

Since 2005, reporting of HER2 expression and the use of molecular signature technology have become standard practice. Before the availability of targeted therapies such as Trastuzumab, which has been shown to reduce recurrence risk by 50% and mortality by one-third (2), recurrence in this group was high. Current molecular signature tests also predict the risk of recurrence in this population. Potential confounders such as physical activity and body mass index were also not considered. Additionally, the type and timing of recurrence were not defined in this study. An early local recurrence, for example, would more likely be due to residual disease than the effect of VET.

The use and type of endocrine therapies have also changed significantly. Twenty-four percent of the postmenopausal women in the study received tamoxifen (TAM) rather than an AI. The first ATAC results (Arimidex, TAM, alone or in combination) published in 2002 showed superiority of AI over TAM in this setting, and current guidelines support AI use (3,4). Additionally, 37% of the study population received no endocrine therapy (ET), which is inconsistent with current practice for hormone receptor–positive breast cancer and increases risk of recurrence. It is likely that patients thought to be at higher risk for recurrence were selected to receive AI, creating bias. Also, adherence to ET was reported at 88%-90%, which is likely an overestimate, particularly in women with vaginal symptomatology. A recent review showed that one-third of women were noncompliant with ET at 5 years (5).

The formulation and dosage of VET were undefined. Vaginal estrogen creams (often used in high doses) and the 25-mcg vaginal estradiol tablet available at the time have demonstrated significant systemic absorption (6). Available since 2010, the options of low-dose and ultra-low-dose vaginal estradiol inserts (10 mcg and 4 mcg) and the low-dose estrogen ring demonstrate minimal systemic absorption (6). Patients at low risk of recurrence often receive these low-dose formulations today with the approval of their medical oncologists and support from The North American Menopause Society. Further, adherence to VET is low, so it should not be assumed that women who filled 2 prescriptions reflected long-term adherence (7).

Given these limitations, practice changes limiting use of VET in BC survivors on AIs cannot be recommended. Large prospective studies are needed using current BC risk stratification and treatments, monitoring of adherence, and use of newer lower-dose vaginal estrogen formulations.

Funding

We received no funding in the writing of this article.

Notes

Role of the funder: Not applicable.

Disclosures: Dr Holly J. Pederson serves as a consultant for Myriad Genetics, Inc. Drs Faubion and Pruthi have no financial disclosures. Dr Goldfarb has the following disclosures: Advisory Boards: Sermonix Pharmaceuticals LLC, Ms. Medicine, and Revision Skincare. Consulting: NanOlogy, Spectrum Pharmaceuticals LLC, and Shook, Hardy and Bacon. Grant Funding: Sprout Pharmaceuticals Inc and Paxman Cooling Ltd.

Author contributions: HJP—Writing—original draft, review and editing. SSF—Writing—review and editing. SP—Writing—review and editing. SG—Writing—review and editing.

Data availability

No new data were generated or analyzed in support of this research.