Prognostic significance of different molecular typing methods and immune status based on RNA sequencing in HR-positive and HER2-negative early-stage breast cancer

Abstract

Background: This study was conducted to evaluate the prognostic significance of different molecular typing methods and immune status based on RNA sequencing (RNA-seq) in hormone receptor (HR)-positive and human epidermal growth factor receptor 2 (HER2)-negative (HR + /HER2-) early-stage breast cancer and develop a modified immunohistochemistry (IHC)-based surrogate for intrinsic subtype analysis.

Methods: The gene expression profiles of samples from 87 HR + /HER2- early-stage breast cancer patients were evaluated using the RNA-seq of Oncotype Dx recurrence score (RS), PAM50 risk of recurrence (ROR), and immune score. Intrinsic tumor subtypes were determined using both PAM50- and IHC-based detection of estrogen receptor, progesterone receptor, Ki-67, epidermal growth factor receptor, and cytokeratins 14 and 5/6. Prognostic variables were analyzed through Cox regression analysis of disease-free survival (DFS) and distant metastasis-free survival (DMFS).

Results: Survival analysis showed that ROR better predicted recurrence and distant metastasis compared to RS (for DFS: ROR, P = 0.000; RS, P = 0.027; for DMFS, ROR, P = 0.047; RS, P = 0.621). Patients with HR + /HER2- early-stage breast cancer was classified into the luminal A, luminal B, HER2-enriched, and basal-like subtypes by PAM50. Basal-like subgroups showed the shortest DFS and DMFS. A modified IHC-based surrogate for intrinsic subtype analysis improved the concordance with PAM50 from 66.7% to 73.6%, particularly for basal-like subtype identification. High level of TILs and high expression of immune genes predicted poor prognosis. Multi-factor Cox analysis showed that IHC-based basal-like markers were the only independent factors affecting DMFS.

Conclusions: Prognosis is better evaluated by PAM50 ROR in early-stage HR + /HER2- breast cancer and significantly differs among intrinsic subtypes. The modified IHC-based subtype can improve the basal-like subtype identification of PAM50. High immunity status and IHC-based basal-like markers are negative prognostic factors.

Keywords: HR-positive/HER2-negative breast cancer; Immune rescore; Intrinsic subtype; Risk of recurrence; Tumor-infiltrating lymphocyte.

Fig. 1

Survival analysis based on RS and ROR risk categories. Disease-free survival analysis by RS (a) and ROR (b); Distant metastasis-free survival analysis by RS (c) and ROR (d)

Fig. 2

Intrinsic subtype distribution and prognosis evaluation. a Intrinsic subtypes classified by PAM50; Disease-free survival analysis by modified IHC surrogate (b); Distant metastasis-free survival analysis by PAM50 (c)

Fig. 3

Basal-like subtype case defined by both PAM50 and modified IHC-based assay. PAM50 heatmap a; HE staining b; ER + , 10% c; PR + , 1% d; HER2, 1 + e; f. CK5/6 + f; EGFR + g; CK14 + h; Ki-67 index 80% i

Fig. 4

Prognosis evaluation of modified IHC surrogate and IHC basal-like marker. Intrinsic subtypes classified by St. Gallen IHC (a) and a modified IHC surrogate (b) Disease-free survival analysis by modified IHC surrogate (c); Distant metastasis-free survival analysis by modified IHC surrogate (d); Disease-free survival analysis of all cases expressing basal-like markers regardless of Ki-67 index (e); Distant metastasis-free survival analysis of all cases expressing basal-like markers regardless of Ki-67 index (f)

Fig. 5

Immune status and prognosis. Disease-free survival analysis based on TILs subgroups (a) and immune score (c); Distant metastasis-free survival analysis based on TILs subgroups (b) and immune score (d)

Fig.6

Forest plots of univariate analysis for DFS (a) and DMFS (b)

Fig. 7

Forest plots of multivariate analysis for DFS (a) and DMFS (b)