Abstract 865: Single-cell sequencing identification of malignant cell subpopulations linked to recurrence in hepatocellular carcinoma

Abstract Introduction: Hepatocellular Carcinoma (HCC) is a highly malignant and challenging-to-treat tumor, with early recurrence affecting up to 35-50% of patients within two years after surgery, significantly impacting prognosis. Identifying high-risk patients for recurrence is critical for guiding postoperative adjuvant therapy. The rapidly advancing field of single-cell RNA sequencing technology offers an unprecedented opportunity to discover clinically relevant cell subpopulations. Here, we harnessed single-cell techniques to unveil a cell subpopulation closely linked with HCC recurrence, highlighting its clinical significance. Methods and Results: We conducted single-cell RNA sequencing on four HCC samples and harmonized data from three public datasets (GSE112271, GSE125449, CNP0000650) to elucidate the heterogeneity of malignant cells in HCC. Our investigation revealed a distinctive HCC subpopulation enriched in patients with early recurrence (11. 4% vs. 3. 8%, P=0. 02). This subpopulation exhibited up-regulated STMN1 and EZH2, signifying high proliferative potential and a propensity for epithelial-mesenchymal transition (EMT). Applying the CIBERSORTx algorithm to the TCGA-LIHC dataset (N=323 patients), we inferred the proportion of this recurrence subpopulation among all malignant cells. Notably, patients with a higher proportion of this subpopulation (N=160) demonstrated significantly shorter recurrence-free survival (31. 2 vs. 42. 9 months, P=0. 002), reduced median survival (45. 9 vs 55. 7 months, P=0. 03) and a higher rate of microvascular invasion (35% vs. 23%, P=0. 01). To validate this recurrence-related subpopulation in clinical samples, we conducted multiplex immunohistochemistry staining on paraffin-embedded (FFPE) HCC samples (including CD45, CD31, ARG1, STMN1, EZH2) and used machine learning for its identification. Our findings revealed that patients with a higher proportion of this recurrence subpopulation (N=40) experienced significantly shorter median survival (38. 5 vs. 51. 0 months, P0. 005) and progression-free survival (25. 1 vs. 33. 4 months, P=0. 01) compared to a lower proportion (N=40). Multivariate Cox regression analysis confirmed that the proportion of this recurrence subpopulation is an independent predictor of worse prognosis (P=0. 003). Conclusion: Through the power of single-cell RNA sequencing technology, we have unearthed and characterized a malignant cell subpopulation intricately connected with HCC recurrence. Importantly, this subpopulation can be identified in FFPE tissue samples and correlates robustly with patient prognosis. Our study underscores the transformative possibility of single-cell techniques in developing cell subpopulation-based biomarkers, emphasizing their clinical relevance and potential utility in guiding postoperative adjuvant therapy for HCC patients. Citation Format: Yucheng Dong, Xin Guo, Xuezhu Wang, Fan Yang, Yongchang Zheng, Yang Chen. Single-cell sequencing identification of malignant cell subpopulations linked to recurrence in hepatocellular carcinoma abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts) ; 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84 (6Suppl): Abstract nr 865.