Identification of a Gene Panel Predictive of Triple-Negative Breast Cancer Response to Neoadjuvant Chemotherapy Employing Transcriptomic and Functional Validation

• Published in: International journal of molecular sciences Vol. 23; no. 18; p. 10901
• Main Authors: Vishnubalaji, Radhakrishnan, Abdel-Razeq, Hikmat, Gehani, Salahddin, Albagha, Omar M. E., Alajez, Nehad M.
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 17.09.2022; MDPI
• Subjects: Breast cancer; Cancer therapies; Cell culture; Cell cycle; Chemotherapy; Clinical medicine; CRISPR; Depletion; Discriminant analysis; Doxorubicin; ErbB-2 protein; Gene expression; Genes; Medical prognosis; Metastases; neoadjuvant chemotherapy; Network analysis; pathological complete response; Patients; predictive biomarkers; Proteins; residual disease; Survival; Survival analysis; TNBC; Transcriptomics; Tumors
• ISSN: 1422-0067; 1661-6596; 1422-0067
• DOI: 10.3390/ijms231810901

Triple-negative breast cancer (TNBC) patients exhibiting pathological complete response (pCR) have better clinical outcomes compared to those with residual disease (RD). Therefore, robust biomarkers that can predict pCR may help with triage and resource prioritization in patients with TNBC. Herein, we identified a gene panel predictive of RD and pCR in TNBC from the discovery (n = 90) treatment-naive tumor transcriptomic data. Eight RD-derived genes were identified as TNBC-essential genes, which were highly predicative of overall survival (OS) and relapse-free survival (RFS) in an additional cohort of basal breast cancer (n = 442). Mechanistically, targeted depletion of the eight genes reduced the proliferation potential of TNBC cell models, while most remarkable effects were for combined SLC39A7, TIMM13, BANF1, and MVD knockdown in conjunction with doxorubicin. Orthogonal partial least squares-discriminant analysis (OPLS-DA) and receiver operating characteristic curve (ROC) analyses revealed significant predictive power for the identified gene panels with an area under the curve (AUC) of 0.75 for the validation cohort (n = 50) to discriminate RD from pCR. Protein–Protein Interaction (PPI) network analysis of the pCR-derived gene signature identified an 87-immune gene signature highly predictive of pCR, which correlated with better OS, RFS, and distant-metastasis-free survival (DMFS) in an independent cohort of basal and, to a lesser extent, HER2+ breast cancer. Our data have identified gene signatures predicative of RD and pCR in TNBC with potential clinical implications.