Abstract NG02: Residual breast cancer cells co-opt SOX5-driven endochondral ossification to maintain dormancy

Abstract Breast cancer remains the leading cause of cancer-related mortality among women globally. While significant advances in diagnostics and targeted treatment strategies have rendered primary tumors largely curable, approximately 30% of patients develop recurrent disease. These relapses, often emerging after prolonged periods of clinical remission spanning years to decades, are responsible for the majority of breast cancer-associated deaths. Recurrence originates from minimal residual disease (MRD), which is composed of residual tumor cells (RTCs) capable of evading primary tumor-oriented therapy and persisting in a quiescent, non-proliferative state. Therefore, therapeutic strategies aimed at eliminating MRD by targeting the specific mechanisms that enable RTC survival and dormancy escape hold promise for preventing recurrence and reducing mortality. However, such interventions have yet to be integrated into routine clinical practice, largely due to limited understanding of the biological mechanisms that govern RTC dormancy and reactivation. To identify potential drivers of the transition from dormant MRD to actively proliferating recurrent tumors, we interrogated data from a focused in vivo CRISPR-Cas9 screen that we performed using MMTV-rtTA;TetO-Her2/neu (MTB/TAN)-derived cells. In this clinically relevant genetically engineered mouse (GEM) model, doxycycline-induced HER2 expression promotes primary tumor formation, while doxycycline withdrawal mimics targeted therapy by suppressing HER2, leading to tumor regression and establishment of dormant MRD. Recurrences arise spontaneously and independently of HER2 signaling, mirroring the latency and oncogene-independence observed in human breast cancer. Through CRISPR-Cas9 screening, we identified the transcription factors SOX5 and its paralog SOX6 as candidate regulators of tumor relapse. Functional validation via recurrence-free survival assays at both orthotopic and metastatic (lung) sites revealed that SOX5 depletion accelerated recurrence by promoting early exit from dormancy. SOX5 demonstrates spatiotemporally restricted expression during embryogenesis, with established roles in neural differentiation and cartilage specification. However, its function within the mammary epithelium remains unexplored. Intriguingly, our findings demonstrate that SOX5 induces dormant RTCs to activate a cartilage-mediated bone development program known as endochondral ossification. This transdifferentiation from mammary epithelial identity to mesodermal cartilage and bone lineages was confirmed by progressive, SOX5-dependent uptake of the bone-seeking tracer [18F]NaF, as visualized by PET/CT imaging in dormant MRD, which was reversed upon tumor recurrence. In line with our prediction that SOX5-dependent osteochondrogenic differentiation promotes dormancy, we found that the expression of osteochondrogenic genes in primary breast tumors or their enrichment in residual disease following neoadjuvant therapy correlated with improved recurrence-free survival in patient cohorts. In conclusion, our study reports a previously unexplored role for SOX5-mediated osteochondrogenic differentiation in sustaining RTC dormancy and highlights its potential utility in predicting recurrence risk among breast cancer patients. Clinical studies incorporating [18F]NaF-PET imaging post-neoadjuvant therapy may enable prospective identification of patients whose residual disease exhibits osteochondrogenic differentiation, indicative of a lower risk for relapse. Such stratification could inform personalized therapeutic approaches, guiding the application of emerging MRD-targeting adjuvant treatments and ultimately reshaping the clinical management landscape for breast cancer survivors. Citation Format: Amulya Sreekumar. Residual breast cancer cells co-opt SOX5-driven endochondral ossification to maintain dormancy [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(8_Suppl):Abstract nr NG02.