BACKGROUND

A prior pooled analysis of the MONALEESA-2, -3, and -7 trials identified baseline markers predictive of sensitivity or resistance to ribociclib plus endocrine therapy (ET). We report the results of an analysis of paired baseline and end-of-treatment (EOT) circulating tumor DNA (ctDNA) samples across the MONALEESA trials.

PATIENTS AND METHODS

Paired baseline and EOT ctDNA samples from MONALEESA-2, -3, and -7 were sequenced using a targeted next-generation sequencing panel. Genes with an EOT alteration prevalence of >5% were included. A McNemar test was carried out on paired samples and adjusted for multiple testing to control the false discovery rate. A Bayesian mixed-effects model was used to adjust for ctDNA fraction at both time points and for study differences.

RESULTS

The analysis included 523 paired samples. At EOT, 21 genes had a >5% alteration prevalence. A trend for higher ctDNA fraction at EOT versus baseline (P = 0.08) was observed. Prevalence of alterations was higher at EOT versus baseline in RB1, SPEN, TPR, PCDH15, and FGFR2 in the ribociclib arm; PBRM1 in the placebo arm; and ESR1 in both arms. The mixed-effects model demonstrated that the same trends for increased prevalence of these alterations at EOT were observed after adjusting for ctDNA fraction and that the increased rate of RB1 and SPEN alterations at EOT were specific to ribociclib plus ET. Analysis of ESR1 indicated a similar increase at EOT in both arms. The most common acquired ESR1 mutations at EOT included Y537C/N/S/D, D538G, E380Q, and L536H/R/P/LC. The prevalence of PIK3CA hotspot mutations at baseline and EOT was similar.

CONCLUSIONS

This analysis identified acquired gene alterations in patients with hormone receptor-positive/human epidermal growth factor receptor-2 negative advanced breast cancer treated with ribociclib plus ET or placebo plus ET. These data may support further studies on acquired resistance mechanisms and inform future systemic interventions in the post-cyclin-dependent kinase 4/6 inhibitor setting.