Association of PET-based estradiol-challenge test for breast cancer progesterone receptors with response to endocrine therapy

Abstract

Estrogen receptor (ER) testing of breast cancer imperfectly predicts response to endocrine therapy (ET). We hypothesize that a brief estradiol challenge will increase tumor progesterone receptor (PgR) levels only in tumors with functional ER. In this prospective, phase 2, single-center, single-arm trial (NCT02455453), we report the association of response to ET with change in tumor uptake of the progestin analog, 21-[18F]fluorofuranylnorprogesterone (FFNP), before and after a one-day estradiol challenge. In 43 postmenopausal women with advanced ER+ breast cancer, we show a post-challenge increase in tumor FFNP uptake only in 28 subjects with clinical benefit from ET (responders), but not in 15 without clinical benefit (nonresponders) (p < 0.0001), indicating 100% sensitivity and specificity. We further show significantly longer survival (p < 0.0001) in the responding subjects. Our results demonstrate that change in tumor FFNP uptake after estradiol challenge is highly predictive of response to ET in women with ER+ breast cancer.

Conflict of interest statement

F.D. reports grants from the National Cancer Institute, grants from Siteman Cancer Center/Barnes-Jewish Hospital Foundation during the conduct of the study, and grants from the ECOG-ACRIN Medical Research Foundation and the National Cancer Institute, and grants from NIH/NIBIB outside the submitted work. C.M. reports research funding from Pfizer, consulting fees from Pfizer, AstraZeneca, Eli Lilly, and Novartis. M.N. reports being a speaker for Merck and Esai. J.A.K. was a founder and has an equity interest in Radius Health, Inc., which is developing Elacestrant (RAD 1901), an antiestrogen for breast cancer. B.A.S. reports funding by the ECOG-ACRIN Medical Research Foundation and American College of Radiology for his contribution in administrative roles and also is a consultant for Avid Radiopharmaceutical, Inc., Capella Imaging, LLC., Curium Pharma, General Electric Healthcare, Imaginab, Inc., Progenics Pharmaceuticals, American College of Radiology, American Medical Foundation for Peer Review and Education and Siemens Healthineers.

Figures

Fig. 1. Responder.

A woman with newly diagnosed ER+/PR+/HER2− invasive ductal carcinoma with metastatic disease at diagnosis in left axillary lymph nodes and a rib. She was treated with an aromatase inhibitor and palbociclib after the estradiol challenge test and had marked improvement of all lesions. Selected fused transaxial 21-[18F]fluorofuranyl-norprogesterone [FFNP]-PET/CT (top) and CT (bottom) images at baseline show intense FFNP uptake in the primary left breast cancer (arrows). One day after estradiol, tumor FFNP uptake measured as the maximum standardized uptake value (SUVmax) (arrows) was increased by 78% from baseline.

Fig. 2. Nonresponder.

A woman with ER+/PR−/HER2− invasive ductal carcinoma post neoadjuvant ET and breast conserving surgery, who developed metastatic disease to left prevascular and aortopulmonary lymph nodes. She was treated with an aromatase inhibitor after the estradiol challenge test but developed progressive disease. Selected fused transaxial FFNP-PET/CT (top) and CT (bottom) images at baseline show minimal FFNP uptake in the prevascular lymph node metastasis (arrows). One day after estradiol, tumor FFNP uptake measured as the maximum standardized uptake value (SUVmax) (arrows) is unchanged.

Fig. 3. Tumor FFNP uptake in responders and nonresponders.

a Baseline SUV, b Percent change in SUV after estradiol challenge.

Fig. 4. Percent change in tumor FFNP uptake with estradiol challenge based on prior therapies.

Relationship of percent change in standardized uptake value (SUV) for FFNP after estradiol challenge to previous subject therapy.

Fig. 5. Overall survival results.

a Dichotomized based on percent change in standardized uptake value (SUV) for FFNP after estradiol challenge (≥7% vs. <7%). The overall survival results based on clinical response (responders versus nonresponders) were identical to those based on percent change in FFNP SUV. b Dichotomized based on the optimal cut-off value of the baseline FFNP SUV (≥4.16 vs. <4.16).