A functional signal profiling test for identifying a subset of HER2-negative breast cancers with abnormally amplified HER2 signaling activity

Abstract

The results of clinical trials evaluating the efficacy of HER2 inhibitors in patients with breast cancer indicate that the correlation between HER2 receptor levels and patient outcomes is as low as 50%. The relatively weak correlation between HER2 status and response to HER2-targeting drugs suggests that measurement of HER2 signaling activity, rather than absolute HER2 levels, may more accurately diagnose HER2-driven breast cancer. A new diagnostic test, the CELx HER2 Signaling Profile (CELx HSP) test, is demonstrated to measure real-time HER2 signaling function in live primary cells. In the present study, epithelial cells extracted fresh from breast cancer patient tumors classified as HER2 negative (HER2-, n = 34 of which 33 were estrogen receptor positive) and healthy subjects (n = 16) were evaluated along with reference breast cancer cell lines (n = 19). Live cell response to specific HER2 agonists (NRG1b and EGF) and antagonist (pertuzumab) was measured. Of the HER2- breast tumor cell samples tested, 7 of 34 patients (20.5%; 95% CI = 10%-37%) had HER2 signaling activity that was characterized as abnormally high. Amongst the tumor samples there was no correlation between HER2 protein status (by cell cytometry) and HER2 signaling activity (hyperactive or normal) (Regression analysis P = 0.144, R2 = 0.068). One conclusion is that measurement of HER2 signaling activity can identify a subset of breast cancers with normal HER2 receptor levels with abnormally high levels of HER2 signaling. This result constitutes a new subtype of breast cancer that should be considered for treatment with HER2 pathway inhibitors.

Keywords: CELx HSP test; HER2 negative; HER2 signaling; targeted therapy; tumor primary cells.

Conflict of interest statement

CONFLICTS OF INTEREST

YH, DJB, BER, IAM, AD, SMS, SM, BFS, LGL are all employed by and have membership interest in Celcuity. LF has membership interest in Celcuity. CAL and SAH have no conflicts to declare.

Figures

Figure 1. Characteristics of primary epithelial cells derived from patient tissue specimens

(A) A representative culture of primary cells from a digested tumor biopsy at day 1 (a) and day 5 (b) of culture, respectively. (B) Flow cytometric analysis of luminal (EpCAM+, Claudin4+) and basal (CD49f+, CD10+) markers on four representative tumor primary cells harvested at the time of CELx HSP test. (C) Plot showing the Mean Fluorescence Channel (MFC) of the luminal marker EpCAM (x-axis) and the basal/progenitor biomarker CD49f (y-axis) for all 34 tumor samples tested (filled circles). For comparison, 2 healthy samples are displayed (empty circles). (D) Comparison of expression levels of HER2, ERα, and PR between primary cells and cell lines (HER2+ and HER2−), which were measured by flow cytometry. (E) Histogram plot of HER2 expression measured by flow cytometry of a representative tumor (shaded peak) compared to HER2+ cell line SKBR3 (solid line) and HER2- cell line MDA231 (dashed line) that is coincidentally in the same range as the healthy samples. The bar below the graph represents the range of means from all the primary tumors and healthy tissue tested (MFC range 31-210).

Figure 2. Optimization and specificity of CELx HSP test in primary cells

(A) R37 primary cells seeded at different densities (8000, 10000, 12000, 15000 cells per well) in a sensor plate were stimulated with NRG1b (3 nM). CELx curves are displayed using Delta CI values to demonstrate the relative signals normalized to the time point (arrow) when NRG1b was added. Positive correlation between cell number and NRG1-driven CELx signal is shown in the inset. (B and C) Dose-response curves of NRG1b and EGF stimulation of CELx signals in R39 primary cells. (D) Dose-response curve of pertuzumab showing its specific inhibitory effect on NRG1b-driven CELx signal.

Figure 3. Identification of subgroups of HER2− tumor-derived primary cells by CELx HSP test

(A) Representative CELx time-course curves representing a high, abnormal HER2 signaling activity in a high responder (R39) and a low HER2 signaling activity in a non-responder (R58). In this display, curves of NRG1 stimulation in the absence versus presence of pertuzumab (10 μg/mL) are presented. (B) Box-and-whisker plots of the CELx HSP test scores for four cell sample groups (HER2− patient-derived tumor cells and cells derived from healthy tissue are plotted with HER2+ cell lines, HER2− cell lines [11].