Putative Clinical Potential of ERBB2 Amplification Assessment by ddPCR in FFPE-DNA and cfDNA of Gastroesophageal Adenocarcinoma Patients

Abstract

Anti-HER2 monoclonal antibody trastuzumab improves the survival of those patients with advanced gastroesophageal adenocarcinoma (GEA) exhibiting HER2/ERBB2 overexpression/amplification. The current gold standard methods used to diagnose the HER2 status in GEA are immunohistochemistry (IHC) and silver or fluorescence in situ hybridization (SISH or FISH). However, they do not permit spatial and temporal tumor monitoring, nor do they overcome intra-cancer heterogeneity. Droplet digital PCR (ddPCR) was used to implement the assessment of HER2 status in formalin-fixed paraffin-embedded (FFPE) tumor DNA from a retrospective cohort (86 patients) and in cell-free DNA (cfDNA) samples from a prospective cohort (28 patients). In comparison to IHC/SISH, ddPCR assay revealed ERBB2 amplification in a larger patient fraction, including HER2 2+ and 0-1+ of the retrospective cohort (45.3% vs. 15.1%). In addition, a considerable number of HER2 2+ and 0-1+ prospective patients who were negative in FFPE by both IHC/SISH and ddPCR, showed ERBB2 amplification in the cfDNA collected just before surgery. cfDNA analysis in a few longitudinal cases revealed an increasing ERBB2 trend at progression. In conclusion, ddPCR in liquid biopsy may improve the detection rate of HER2 positive patients, preventing those patients who could benefit from targeted therapy from being incorrectly excluded.

Keywords: ERBB2; HER2; cell-free DNA (cfDNA); droplet digital PCR (ddPCR); gastroesophageal adenocarcinoma; liquid biopsy.

Figure A1

Quality assessment of cfDNA using Agilent Tape Station 2200 (Cell-free DNA screen tape assay).

Figure A2

IHC/SISH staining for HER2/ERBB2 expression/amplification evaluation. Representative images of HER2 3+, HER2 2+, HER2 1+, and HER2 0+ samples showing a progressive reduction in the HER2 expression intensity (brown color; IHC 20× and 40× magnification). Pictures of SISH analysis, performed in the equivocal cases (HER2 2+) to discriminate for the presence of tumor cell nuclei with ERBB2 amplification, were also included (black color; SISH 40× magnification).

Figure 1

The box plot shows the ERBB2 CNV positive value distribution in the FFPE-DNA of HER2 3+, HER2 2+, and HER2 0–1+ GEA subgroups (retrospective cohort). Statistically significant differences were present between HER2 3+ and HER2 2+ or HER2 0–1+ subgroups. No difference was found when comparing HER2 2+ and HER2 0–1+ patients. The p values were calculated using a two-tailed Wilcoxon Mann-Whitney test; a p-value < 0.05 was considered statistically significant. CNV, copy number variation.

Figure 2

Receiver Operating Characteristics (ROC) curve obtained by comparing ERBB2 CNV values of the HER2 positive patients vs. CNV values of the HER2 negative ones. HER2 positive or negative status was determined using IHC/SISH staining as “standard method”. Area under the curve (AUC) 0.909, Std. Error 0.04091, CI 95% 0.8288–0.9892, p-value < 0.0001.

Figure 3

ERBB2 CNV values in serial dilutions of tumor DNA with increasing amount of its normal counterpart. For each dilution, the absolute quantity (ng) of tumor and normal DNA as well as dilutions were reported. The positivity for ERBB2 amplification was detected up to a 1:64 dilution.

Figure 4

The box plot shows ERBB2 CNV value distribution in FFPE-DNA and time-matched cfDNA samples in the HER2 positive subgroup (A) and in the HER2 negative subgroup (B) in the prospective cohort. (A) No difference was observed when comparing the FFPE-DNA and cfDNA of HER2 positive patients. The p-value was calculated using a two-tailed Student’s t-test. (B) A statistically significant difference was present between the FFPE-DNA and the cfDNA of the HER2 negative group. The p-value was calculated using a two-tailed Wilcoxon Mann-Whitney test. A p-value < 0.05 was considered statistically significant. FFPE, formalin-fixed paraffin-embedded; cfDNA, cell-free DNA.

Figure 5

Longitudinal analysis of four “progressor” GEA patients (G016, G025, G030, and A564). ERBB2 CNV values and tumor markers S-CEA and S-CA19.9 are reported for each patient at different time points. The administration of neoadjuvant or adjuvant therapy, surgery, and the time of disease progression are indicated. A representative TAC image is shown on the right-hand side. A red arrow marks the site of metastasis. S-CEA, soluble carcinoembryonic antigen; S-CA19.9, soluble carbohydrate antigen.

Figure 6

Longitudinal analysis of three “non-progressor” GEA patients (G005, G013, and G015). ERBB2 CNV values are reported for each patient at different time points. The administration of adjuvant therapy and surgery is indicated.