Genetic testing during treatment can improve accuracy of response and outcome prediction compared to other prognostic tests, according to results from a study assessing on-treatment changes in gene expression in patients receiving chemotherapy. Researchers from Oikonomidou O Institute of Genetics and Molecular Medicine in Edinburgh, United Kingdom, presented the findings at the San Antonio Breast Cancer Symposium on December 6, 2018.

Researchers sequenced 97 samples from 50 patients receiving neoadjuvant chemotherapy for primary breast cancer taken at pretreatment, two weeks on treatment, mid-therapy, and resection. Patient age at diagnosis ranged from 29–76 years, and most (80%) had human epidermal growth factor receptor 2-positive disease.

Ion AmpliSeq transcriptome sequencing yielded expression values for 12,635 genes. On-treatment testing uncovered significant changes in genes commonly associated with proliferation—including PCNA, AURKA, and MKI67—regardless of response class.

Researchers identified AAGAB as an on-treatment marker of response, which was associated with a testing accuracy of 100% and a validation accuracy of 78% in the I-SPY 1 trial—a series of trials assessing neoadjuvant treatment of locally advanced breast cancer. AAGAB was predictive of long-term survival in both chemotherapy cohorts at the same expression level as defined for treatment response (p = 0.048 testing; p = 0.0031 validation).

AAGAB was more predictive of response to chemotherapy than established prognostic tests, including MammaPrint and Pam50 RORS. However, AAGAB was not predictive of response or survival in patients treated with endocrine therapy.

“Changes in gene expression of on-treatment chemotherapy breast cancer resulted in the identification of a novel gene marker that was as effective in predicting prognostic status as established prognostic tests,” the researchers concluded.