P4: Clinical Behavior, Management and Treatment Response of Estrogen Receptor Low (1-10%) Breast Cancer by PR expression

Introduction: Estrogen and progesterone receptor (ER/PR) status guide management and impact outcomes of breast cancer (BC) patients. While the College of American Pathologists guidelines recommend a positivity threshold of ≥1%, several studies note that ER-low (1-10%) tumors represent a more aggressive phenotype. This study investigates the clinical significance of PR expression in ER-low disease by comparing ER-negative ( < 1%), ER-low/PR-, ER-low/PR+ and ER-positive ( >10%).


Methods: All patients (pts) with Her2 negative invasive BC at diagnosis were identified from the National Cancer Database 2018-2019. ER status was defined as negative if < 1% of cells stained positive, ER-low if 1-10%, and positive if >10%. PR status was defined as negative if < 1% and positive if ≥1%. For patients treated with neoadjuvant therapy, pathological complete response (pCR) was defined as ypT0/ypTis, ypN0/ypN0i+. Groups were compared using chi-square tests.


Results: We identified 232,762 pts with Her2- invasive BC, of which 2% (4584) were ER-low, 13.8% ER-, and 84.2% ER+. Chemo was given in 83.9% of ER- and 82.3% of ER-low/PR- disease. ER-low/PR+ disease was less frequently treated with chemo (58.9%), and ER+ disease had the lowest rate of chemo (22.9%). Neoadjuvant chemo (NAC) use was also highest in ER- (36.6%) and ER-low/PR- (32.0%) and was less frequent in ER-low/PR+ (19.2%) and lowest in ER+ (4.2%). Adjuvant endocrine therapy in the ER-low subgroup varied by PR status (47.7% in PR- and 69.6% in PR+, p< 0.01). Similarly, ER-low pts differed in their Oncotype DX Recurrence Score (RS) by PR status (median RS of 49 in ER-low/PR-, vs 22.5 in ER-low/PR+, p< 0.01) and in their Ki67 scores (median 60% and 36%, respectively, p< 0.01). ER-low patients selected for NAC were younger, had higher T category, tumor grade, ki67, and were more likely to be clinically node positive. With NAC, pCR rates were similar between ER-low/PR- and ER-low/PR+ (39.5% and 38.1%, p=0.67), and response rates were closer to those of the ER- group (39.7%) than of the ER+ group (8.4%).


Conclusions: Our study suggests that the tumor features and clinical management of ER-low tumors vary significantly by PR expression. Overall ER-low/PR- more closely resembles ER- BC, and ER-low/PR+ acts as an intermediate between ER- and ER+ disease. In ER-low disease selected for treatment with NAC, response is similar to ER- regardless of PR status.