Circulating tumor (ct)DNA monitoring of ER+/HER2- high-risk breast cancer during adjuvant endocrine therapy.

1010 Background: ctDNA monitoring during adjuvant endocrine therapy is an opportunity to detect molecular relapse before clinically apparent recurrence. ctDNA positivity rates, dynamics and the frequency of asymptomatic imaging-detectable metastatic disease at the time of ctDNA detection remain unknown in high-risk ER+/HER2- BCs. We present ctDNA results from a prospective, multicenter, randomized ctDNA interventional trial, DARE (NCT04567420). Methods: Patients receiving adjuvant endocrine therapy for >6 months but <7 years, with either recurrence risk >15% (PREDICT, RSPC, CTS5), >4 positive axillary lymph nodes, (primary tumor >5 cm, or 1-3 positive nodes with grade 3 histology, or >3 cm tumor, or high molecular risk (Oncotype Dx RS >26, MammaPrint high risk, EndoPredict >4, Prosigna score >60) were eligible for ctDNA surveillance with the Signatera assay (Natera, Inc.) every 6 months. ctDNA+ patients had systemic staging with imaging and if there was no evidence of metastatic disease patients were randomized to switching to fulvestrant + palbociclib (Arm A) or to continuation of adjuvant therapy (Arm B). Negative predictive value (NPV) was calculated for recurrence in the screening group after each ctDNA- test. In randomized patients, early ctDNA dynamics were correlated with recurrence-free survival (RFS) and ctDNA clearance rates were calculated by trial arm. Results: 552 patients had tissue sent for assay design; 494 had ctDNA results; 52 failed WES and/or had incomplete tumor/normal/blood sets; 6 had pending reports. Among patients not randomized, 432 were ctDNA-, of these N=43 had one time point and 389 had >2 ctDNA- result, overall median screening time 27.4 months (0-45.5), 4 ctDNA- patients had recurrence (NPV 100% at 6 months and 99% at 12 months post-testing). Forty patients were randomized, 34 had post-randomization ctDNA result. Randomization rates were 53% and 76% for patients who tested ctDNA-positive on the first screening (N=19) versus those who turned positive in follow up testing (N=15). At any time post-randomization, ctDNA clearance rates were 63% (10/16) in Arm A and 22% (4/18) in Arm B. Among randomized patients, 6 of 9 patients with increased ctDNA levels from the pre-randomization to the 3-month on-treatment recurred (median time to recurrence 4.8 months, range: 3.3-24.3), among those with a decrease in ctDNA post-treatment only 1 of 6 experienced recurrence at 10.3 months (HR: 5.3, 95% CI: 1.1-53, p=0.04). Conclusions: This study demonstrates the ability of ctDNA to identify breast cancer patients at high risk of relapse for randomization in a prospective, multicenter, randomized clinical trial. Patients with serially ctDNA- results during surveillance had 99% RFS after a median f/u of 27.4 months. Interim analysis revealed higher clearance rates in Arm A compared to patients randomized to Arm B. Early on treatment ctDNA dynamics is prognostic of patient outcomes. Clinical trial information: NCT04567420 .