Evaluation of pharmacodynamic (PD) and biologic activity in a preoperative window-of-opportunity (WOO) study of giredestrant (GDC-9545) in postmenopausal patients (pts) with estrogen receptor-positive, HER2-negative (ER+/HER2–) operable breast cancer (BC).

Abstract

577 Background: Modulation of ER activity and/or estrogen synthesis is the mainstay therapeutic strategy in ER+ BC treatment. Giredestrant is a highly potent, nonsteroidal oral selective ER degrader (SERD) that achieves robust ER occupancy and is effective regardless of ESR1 mutation status. The first short-term preoperative WOO study (NCT03916744) of giredestrant in ER+/HER2– operable BC was designed for dose selection, while providing an early readout of PD as measured by traditional immunohistochemistry (IHC) and transcriptional profiling by assessing treatment effects in paired tumor tissue pre/posttreatment. We present an interim analysis. Methods: Pts were assigned to 14 days’ preoperative treatment with 10, 30, or 100 mg PO giredestrant QD. Pts had newly diagnosed, stage I–III operable, ER+/HER2– untreated BC ≥1.5 cm in diameter (by ultrasound). Modulation of ER signaling and cell proliferation were assessed using paired formalin-fixed paraffin-embedded tumor specimens collected before and after ̃14 days of study treatment. ER, progesterone receptor (PR), and Ki67 protein levels were analyzed by IHC. Change from baseline in tumor cell proliferation by Ki67 was the primary endpoint. Gene expression analysis was performed using the Illumina TruSeq RNA Access method. Results: From Jul 26, 2019 to Oct 15, 2020, 46/75 biomarker-evaluable pts were enrolled across three dose cohorts (10 mg: n = 15; 30 mg: n = 18; 100 mg: n = 13). Pt demographics and tumor characteristics were similar across cohorts. Baseline PAM50 analysis classified tumors as Luminal A (77%) or B (23%). Giredestrant treatment resulted in robust and indistinguishable PD and biologic activity at all doses. Geometric mean posttreatment proportional reduction of Ki67 was 79% (95% CI: 69–89; 10 mg: 80%; 30 mg: 76%; 100 mg: 80%), and 51% of tumors exhibited complete cell cycle arrest, defined as Ki67 ≤2.7%. Mean posttreatment proportional reductions of ER and PR H-scores were 71% (95% CI: 67–75) and 60% (95% CI: 51–70), respectively. An analysis of a predefined, experimentally derived set of 38 ER target genes (the ‘ER activity signature’), was completed for 42 paired tumor specimens. Forty-one of 42 pts (98%) showed a posttreatment reduction in ER activity with a mean proportional decrease of 79% (95% CI: 70–88). A wide range of baseline ER activity was observed with no correlation to baseline ER or PR H-score, or Ki67. There were no discontinuations due to adverse events (AEs). A single grade 3 serious AE was reported in each cohort (all assessed as unrelated to giredestrant). No grade 4 or 5 AEs were reported. Conclusions: Giredestrant was well tolerated in the preoperative setting in ER+/HER2– operable BC, and PDs were consistent with the 30 mg dose achieving maximal ER inhibition. Clinical trial information: NCT03916744.