Abstract 271P
Background
Human endogenous retroviruses (HERVs) are carcinogenic due to genomic instability and epigenetic induced transcription. In reaction, the host activates an interferon (IFN)-based innate response. In breast cancer (BC), HERVs are associated with poor prognosis. A high level of HERV expression also activates an antitumour immune response. Using data from The Cancer Genome Atlas, we searched for HERVs expressed in early BC that were related to both anticancer immunity and good prognosis.
Methods
Patients were classified into BC subtypes by immunohistochemistry (IHC), based on hormone receptor (HR) and human epidermal growth factor 2 (HER2) status, i.e. HR+/HER2-, HR-/HER2+, triple-negative (TN). HERV sequences were detected using Telescope software on RNA sequencing data. Only HERVs overexpressed in tumor tissue were kept for further analysis. We computed an overall HERV expression per case by summing HERV counts. Expression was compared across BC subtypes with the Wilcoxon test. HERVs were further selected based on their correlation with enrichment scores of three IFN I pathways. Average expression generates a signature that was assessed for its impact on 5-year Disease-Free Interval (DFI) by Cox regression. Correlations with tumour-infiltrating lymphocytes (TILs), programmed death ligand 1 (PDL1) and estrogen receptor 1 (ESR 1) were evaluated with Pearson’s correlation coefficient.
Results
We analyzed 824 BC tissue samples; 608 (73.8%) HR+/HER2-, 46 (5.6%) HR-/HER2+, 170 (20.6%) TN. HERV expression did not differ across IHC subtypes. In total, 23 HERVs were significantly correlated with the IFN I pathway and selected for further analysis. Average expression differed across IHC subtypes (highest in TN, lowest in HR+/HER2-). HERV signature was significantly correlated with TIL and PD-L1 scores (respectively, Pearson r=0.66 and r=0.53, p <0.001 each) and negatively correlated with ESR1 (r=-0.36, p<0.001). HERV signature was significantly associated with longer DFI in HR+/HER2- cases only (hazard ratio 0.54 [0.33-0.88]; p=0.01).
Conclusions
We identified 23 HERVs with anti-tumor potential, whose signature was associated with better survival in HR+/HER2 BC.
Clinical trial identification
Editorial acknowledgement
We would like to thank Ms Fiona Ecarnot, PhD, University Hospital Besancon France for providing the medical writing and editorial assistance.
Legal entity responsible for the study
Georges Francois Leclerc Cancer Center, Dijon, France.
Funding
Has not received any funding.
Disclosure
All authors have declared no conflicts of interest.
Resources from the same session
302P - Perceptions of women with HER2+ breast cancer on the risk of recurrence and disease management: Results from the ASKHER survey
Presenter: Matteo Lambertini
Session: Poster session 14
303P - Predicting quality of life trajectories in young women with breast cancer: 5-year results from a large prospective cohort
Presenter: Bryan Vaca-Cartagena
Session: Poster session 14
304P - Impact of estrogen receptor positivity for adjuvant endocrine therapy in luminal T1a/bN0M0 breast cancer: A multi-institutional retrospective observational study
Presenter: Shinsuke Sasada
Session: Poster session 14
305P - Prognosis of isolated locoregional recurrence after early breast cancer with immediate breast reconstruction surgery: A retrospective multi-institutional study
Presenter: Hirohito Seki
Session: Poster session 14
306P - Patient-reported symptoms in early breast cancer and future cardiovascular events: A province-wide administrative database study
Presenter: Edith Pituskin
Session: Poster session 14
307P - Exposure to Di-2-ethylhexyl phthalate and breast cancer incidence: A cohort study
Presenter: Lijuan Tang
Session: Poster session 14
308P - Impact of the COVID-19 (C19) pandemic on breast cancer (BC) treatment patterns in the US
Presenter: Mariana Chavez Mac Gregor
Session: Poster session 14
310P - Identification of racial disparities across MammaPrint and BluePrint subtypes in HR+HER2- breast cancer
Presenter: Sonya Reid
Session: Poster session 14
311P - Optimizing breast cancer staging: Redefining tumor size classification using big data analytics
Presenter: Bin Feng
Session: Poster session 14