Dynamics and type of ESR1 mutations under aromatase inhibitor or fulvestrant combined with palbociclib after randomization in the PADA-1 trial.

1002 Background: In PADA-1, ER+ HER2- metastatic breast cancer patients who displayed a rising ESR1 mutation in blood (b ESR1mut) during a first-line therapy with aromatase inhibitor (AI) and palbociclib were randomized between keeping the same treatment or switching to fulvestrant (FUL) and palbociclib (PAL) (Bidard, Lancet Oncol 2022). In this analysis, we investigated the kinetics of b ESR1mut after randomization, under AI+PAL or FUL+PAL. Methods: Patients who had a rising b ESR1mut detected and no synchronous disease progression underwent further serial ctDNA analyses at randomization and then every 2 months until disease progression. ctDNA analysis at randomization, i.e. before any change in endocrine therapy, was intended to study the impact of sampling fluctuations – since rising b ESR1mut levels were often close to the limit of detection of the assay. b ESR1mut detection was performed with droplet digital PCR (Jeannot, Oncogene 2020), while left over samples were subjected to next-generation sequencing, which allowed for b ESR1mut typing and clonality assessment (Callens, Anal Chem 2022). Results: 172 patients were randomized in PADA-1 after having a rising b ESR1mut and no synchronous disease progression. In these patients, b ESR1mut had a median mutant allelic frequency of 0.8 % (range 0.1-25.8%) and a median copy number of 14 copies/ml of plasma (4-1033) on the “rising” sample, with no imbalance between randomization arms. Among them, N=75 (46.6%) had no b ESR1mut detected at the repeat blood sample at randomization. Of note, these patients had a lower level of bESR1mut at “rising” compared to those who remained b ESR1mut+ at randomization (p=0.01). After treatment start, patients who were switched to FUL+PAL experienced a higher rate of b ESR1mut clearance at 2 months, compared to those remaining on AI (70.9% vs 32.8%, p<0.001). b ESR1mut clearance at 2 months was associated with longer PFS (HR=0.36 95%CI=[0.25;0.52], p<0.001). The length of b ESR1mut clearance was also longer in patients randomized to FUL (median: 7.3 mo 95%CI=[3.7;11.2] vs 1.9 mo 95%CI=[1.8;2.3]; p<0.001). At clinical disease progression, N=62 (83%) and N=49 (73%) patients tested positive for b ESR1mut in the AI and FUL arms, respectively. Mutation typing in 95 patients with available material revealed that the Y537S mutation was the most prevalent (N=36, 37.9%), while N=25 (26.3%) and N=18 (69.2%) had polyclonal b ESR1mut at time of rising b ESR1mut and progression, respectively. The mutation type -including Y537S- and the presence of polyclonal b ESR1mut at time of rising b ESR1mut did not influence patients’ survival and hazard ratio between arms. Conclusions: b ESR1mut ctDNA kinetics support the clinical benefit observed in the FUL+PAL arm over the AI+PAL arm. ESR1 mutation type and clonality did not impact the benefit of the treatment switch. Clinical trial information: NCT03079011 .