Abstract
Aims: Multigene assay is recommended currently for prognostic stratification of the clinically indeterminate group of breast cancer (BC) patients defined as lymph node (LN)-negative, oestrogen receptor (ER)-positive, HER2-negative (LN-/ER+/HER2(-)) to determine the use of chemotherapy. However, this cohort, comprising approximately 40% of BC, is not a homogeneous group and shows variable outcome. This study aims to determine the prognostic value of routinely assessed variables, singly and in combination, in LN-/ER+/HER2(-) BC patients.
Methods and results: A total of 830 LN-/ER+/HER2(-) chemotherapy-naive BCs were investigated. The prognostic value of histological grade, tumour size, lymphovascular invasion (LVI), progesterone receptor (PgR) and Ki67 labelling index (Ki67LI) was assessed. In this series, only 25% of patients received hormone therapy. Median follow-up was 172 months. In the whole cohort, tumour grade, size, LVI, PgR and Ki67LI were correlated highly with outcome in a time-dependent manner. The outcome of this group varied widely from 97% (20% of cases) to 50% survival rate after 10-year follow-up using a combination of these markers. A prognostic index (Nottingham Px) incorporating grade, size, PgR and Ki67LI, was developed. The index can stratify the whole cohort robustly as well as the higher-risk subgroup (NPI score >3.4) into distinct prognostic classes.
Conclusion: Current routinely assessed variables can provide additional prognostic information in LN-/ER+/HER2(-) BC. The proposed (Nottingham Px) index can stratify the BC clinically indeterminate group of patients into excellent and poor prognostic subgroups and can be used to identify reliably patients for systemic chemotherapy or further multigene prognostic testing. Performance of prognostic variables in these tumours is time-dependent, and should be considered in future studies.