ESR1 mutations are frequent in newly diagnosed metastatic and loco-regional recurrence of endocrine-treated breast cancer and carry worse prognosis

Abstract

Background: Emerging mutations in the ESR1 gene that encodes for the estrogen receptor (ER) are associated with resistance to endocrine therapy. ESR1 mutations rarely exist in primary tumors (~ 1%) but are relatively common (10-50%) in metastatic, endocrine therapy-resistant cancers and are associated with a shorter progression-free survival. Little is known about the incidence and clinical implication of these mutations in early recurrence events, such as local recurrences or newly diagnosed metastatic disease.

Methods: We collected 130 archival tumor samples from 103 breast cancer patients treated with endocrine therapy prior to their local/metastatic recurrence. The cohort consisted of 41 patients having at least 1 sample from local/loco-regional recurrence and 62 patients with metastatic disease (of whom 41 newly diagnosed and 28 with advanced disease). The 5 most common ESR1 hotspot mutations (D538G, L536R, Y537S/N/C) were analyzed either by targeted sequencing or by droplet digital PCR. Progression-free survival (PFS), disease-free survival (DFS), and distant recurrence-free survival (DRFS) were statistically tested by Kaplan-Meier analysis.

Results: The prevalence of ESR1 mutations was 5/41 (12%) in newly diagnosed metastatic patients and 5/28 (18%) for advanced metastases, detected at allele frequency > 1%. All mutations in advanced metastases were detected in patients previously treated with both tamoxifen (TAM) and aromatase inhibitors (AI). However, in newly diagnosed metastatic patients, 4/5 mutations occurred in patients treated with TAM alone. PFS on AI treatment in metastatic patients was significantly shorter for ESR1 mutation carriers (p = 0.017). In the local recurrence cohort, ESR1 mutations were identified in 15/41 (36%) patients but only 4/41 (10%) were detected at allele frequency > 1%. Again, most mutations (3/4) were detected under TAM monotherapy. Notably, 1 patient developed ESR1 mutation while on neoadjuvant endocrine therapy. DFS and DRFS were significantly shorter (p = 0.04 and p = 0.017, respectively) in patients that had ESR1 mutations (> 1%) in their loco-regional recurrence tumor.

Conclusions: Clinically relevant ESR1 mutations are prevalent in newly diagnosed metastatic and local recurrence of endocrine-treated breast cancer. Since local recurrences are amenable to curative therapy, these mutations may inform the selection of subsequent endocrine therapies.

Keywords: Breast cancer; ESR1 mutation; Endocrine treatment; Loco-regional/local recurrence; Metastasis.

Fig. 1

ESR1 mutation analysis in the metastatic cohort and its clinical significance. a Analysis of matched samples from the metastatic cohort through the course of disease: primary tumor, newly diagnosed metastases, and advanced metastases. Samples are colored according to their mutation type. Red indicates ESR1 Mut. Green indicates ESR1 WT. ESR1 mutations at an allele frequency of > 1% are marked by an asterisk. Dark gray indicates that a tumor was present at this time point but a sample was not available. Lower bars represent the treatments given for each patient pre-biopsy, either at the adjuvant phase before the metastatic disease or at the advanced phase before the advanced metastatic biopsy. TAM, tamoxifen, light blue; AI, aromatase inhibitor, blue. b Prevalence of ESR1 mutations divided according to the metastatic disease stage and the type of treatment prior to biopsy. c Kaplan-Meier plots of progression-free survival calculated from the start of AI treatment at the metastatic setting

Fig. 2

ESR1 mutation analysis in the loco-regional cohort. a Analysis of matched samples from the loco-regional cohort through the course of the disease: primary tumor, all local recurrences, and advanced metastatic recurrence. Samples are colored according to their mutation type. Red indicates ESR1 Mut. Green indicates ESR1 WT. ESR1 mutations at an allele frequency of > 1% are marked by an asterisk. Dark gray indicates that a tumor was present at this time point but a sample was not available. Lower bars represent the treatments given for each patient prior to the tested local recurrence sample. TAM, tamoxifen, light blue; AI, aromatase inhibitor, blue. b Prevalence of ESR1 mutations in the loco-regional cohort divided according to the type of treatment prior to biopsy. ESR1 mutations at an allele frequency of > 1% are colored with dark red

Fig. 3

Survival analysis for patients with loco-regional recurrence. a Kaplan-Meier plots of distant recurrence-free survival from the primary tumor comparing ESR1 mutations at an allele frequency higher versus lower than 1%. b Kaplan-Meier plots of disease-free survival from the tested local recurrence comparing ESR1 mutations at an allele frequency higher versus lower than 1%. c Kaplan-Meier plots of distant recurrence-free survival from the tested local recurrence comparing ESR1 mutations at an allele frequency higher versus lower than 1%. d–f Univariate Cox regression analysis for the same comparisons as in a–c represented by forest plots showing the hazard ratio for the various clinical parameters. LNDx, lymph node status at diagnosis; LR type, local or regional; error lines represent the 95% confidence interval