Eight-Gene Profile Predicts Benefit of Adjuvant Trastuzumab in HER2-Negative Disease
The National Surgical Adjuvant Breast and Bowel Project (NSABP) B-31 trial suggested that efficacy of adjuvant trastuzumab (Herceptin) extended to patients with HER2-negative breast cancer. In a study reported in the Journal of the National Cancer Institute, Pogue-Geile et al identified and confirmed an eight-gene predictive model for trastuzumab benefit in the B-31 population that categorized patients into no-, moderate-, and large-benefit subgroups. Analysis showed that tumors in the no-benefit subgroup are characterized by high ESR1 mRNA and intermediate ERBB2 mRNA expression and that HER2-negative tumors are predominantly in the moderate-benefit group. These findings provide justification for the evaluation of trastuzumab in HER2-negative patients in the NSABP B-47 trial.
Study Details
In the study, tumor samples from patients in the B-31 trial were randomly divided into discovery (n = 588) and confirmation (n = 991) cohorts. A predictive model for trastuzumab response was constructed through gene-expression profiling of 462 genes with the nCounter assay in discovery cohort samples. Gene-by-treatment interaction was tested using Cox models and correlations between variables were assessed by Spearman correlation. A predefined cutpoint for the predictive model was tested in the confirmation cohort.
Identification of Eight-Gene Profile
In the discovery cohort analysis, it was observed that the top predictive genes for trastuzumab response included several estrogen receptor–associated genes and HER2 amplicon ERBB2 genes. Eight predictive genes associated with HER2 (ERBB2, c17orf37, GRB7) or estrogen receptor (ESR1, NAT1, GATA3, CA12, IGF1R) were selected for model construction on the basis of having a Spearman correlation coefficient greater than 0.7 and a minimum interaction P value < .10. Cutpoints were derived for two principal components, which defined three patient subsets based on three-dimensional subset treatment effect pattern plot and the disease-free survival event rate in each subgroup. Disease-free survival hazard ratios for the subgroups were 1.56 in the no-benefit group and 0.27 in the large-benefit group, with the remaining patients being classified as the moderate-benefit group with a hazard ratio of 0.56.
ESR1 and ERBB2 mRNA Levels in Subgroups
In the confirmation set, the predefined cutpoints for the model classified patients into three subgroups with disease-free survival hazard ratios of 1.58 (P = .29, n = 100) in the no–trastuzumab benefit group, 0.60 (P = .01, n = 449) in the moderate-benefit group, and 0.28 (P < .001, n = 442, P < .001 for interaction between model and trastuzumab) in the large-benefit group.
A correlation analysis between ERBB2 and ESR1 mRNA levels in each subgroup defined by the eight-gene prediction model showed that the subgroup with no benefit expressed high levels of ESR1 mRNA and intermediate levels of ERBB2 mRNA rather than the lowest ERBB2 mRNA levels in both the discovery and confirmation cohorts. Investigation of the distribution of central HER2 assay–negative cases among the benefit categories defined by the predictive model showed that only a few HER2-negative patients belonged to the subgroup with no benefit, with the majority belonging to the moderate benefit subgroup.
The investigators concluded, “We developed a gene expression–based predictive model for degree of benefit from trastuzumab and demonstrated that HER2-negative tumors belong to the moderate benefit group, thus providing justification for testing trastuzumab in HER2-negative patients (NSABP B-47).”
Soonmyung Paik, MD, of the NSABP, is the corresponding author for the Journal of the National Cancer Institute article.
This study was supported by grants from the National Cancer Institute and Pennsylvania Department of Health. The study authors reported no potential conflicts of interest.
The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.
