Associations between DNA damage and PDL‐1 expression in ovarian cancer, a potential biomarker for clinical response

Abstract

Despite the majority of ovarian cancer patients achieving a complete clinical response with first-line therapies, nearly 80% will recur within 24 months. There is an urgent need for novel therapeutic strategies to treat ovarian cancer effectively. Renewed interest in immunotherapy in ovarian cancer has shown that tumor-infiltrating lymphocytes are associated with high-grade serous, clear cell, and endometrioid histology. As a result, immune checkpoint inhibitors have emerged as promising treatments for ovarian cancer. Programmed death ligand-1 (PD-L1) is expressed by cancer cells to evade host surveillance. Its interactions with Programmed death-1 (PD-1) negatively regulate CD8+ T-cells to control inflammatory responses and maintain self-tolerance. Blockade of PD-1/PDL-1 restores anti-tumor cell immunity and has been successfully used to treat some cancers, including non-small cell lung cancer and melanoma. The use of PD-1/PDL-1 blockade in ovarian cancer has been limited by our lack of understanding of the mechanisms that control PD-L1 expression. Emerging evidence has suggested that DNA repair defects and genomic instability influence the expression of PDL-1. However, little is known about the relationship between genomic instability and PDL-1 expression in ovarian cancer, despite the high prevalence of DNA repair defects reported. We hypothesized that genomic instability increases the expression of PD-L1; therefore, PD-L1 expression levels will correlate with high levels of DNA damage in the tumor genome. To test this hypothesis, we have measured DNA repair defects and genomic instability in ovarian tumors using Repair Assisted Damage Detection (RADD). RADD measures persistent DNA lesion and strand breaks within tumor cells using a cocktail of DNA repair enzymes. Lesion sites are then tagged with nucleotide analog and fluorescently labeled for detection. DNA repair defects are revealed by the presence of DNA damage within the tumor genome. In parallel, we also measured PDL-1 expression within these tumor cells using immunohistochemistry. Tumors were categorized and stratified based on the basal levels of DNA damage, and PDL-1 expression was correlated to the type and quantity of DNA damage present in each tumor sample. Using this method, we have established that differences in DNA damage and PD-L1 expression exist among ovarian tumor grade, stage, and pathology. Using different enzyme cocktails in our RADD protocol, we were also able to establish that oxidative lesions are primarily retained within the genome of ovarian tumors. Correlations in DNA damage and PD-L1 expression suggest that oxidative lesions influence the expression of PD-L1 in ovarian cancers. This study establishes a potential biomarker for assessing the utility of PD-L1 inhibitors in ovarian tumors.