Gene expression predicts dormant metastatic breast cancer cell phenotype

Background: Breast cancer can recur months to decades after an initial diagnosis and treatment. The mechanisms that control tumor cell dormancy remain poorly understood, making it difficult to predict which patients will recur and thus benefit from more rigorous screening and treatments. Unfortunately, the extreme rarity of dormant DTCs has been a major obstacle to their study. Methods: To overcome this challenge, we developed an efficient system to isolate and study rare dormant breast cancer cells from metastatic organs including bones, which represent a major site of metastasis. After isolation of cells from the long bones, we used single cell RNA-sequencing (scRNA-seq) to profile proliferative and dormant PyMT-Bo1 breast cancer cells. We also compared this signature to dormant versus proliferative tumor cells isolated from the lungs. Finally, we compared our dormant signature to human datasets. Results: We identified a group of genes including Cfh, Gas6, Mme and Ogn that were highly expressed in dormant breast cancer cells present in the bone and lung. Expression of these genes had no impact on dormancy in murine models, but their expression correlated with disease-free survival in primary human breast cancer tumors, suggesting that these genes have predictive value in determining which patients are likely to recur. Conclusions: Dormant breast cancer cells exhibit a distinct gene expression signature regardless of metastatic site. Genes enriched in dormant breast cancer cells correlate with recurrence-free survival in breast cancer patients.