Abstract P62: Liposarcoma Clonal Complexity: A Multi-Modal Single-Cell and Genomic Analysis

Abstract Introduction: Liposarcomas are rare mesenchymal malignancies that remain poorly characterized at single-cell resolution, limiting insights into tumor identity and heterogeneity. While MDM2 and CDK4 amplifications are hallmark diagnostic features, current clinical tools such as fluorescence in situ hybridization (FISH) lack the resolution to determine whether these aberrations are uniformly distributed across malignant cells. Methods: We constructed a single-cell transcriptomic atlas of liposarcoma, profiling over 200,000 cells from 37 tumor resections and 6 matched patient-derived sarcoma cell (PDSC) cultures to investigate tumor heterogeneity and in vitro evolution. Using inferred copy number variation (CNV) analysis, we mapped MDM2 and CDK4 amplifications at single-cell resolution and integrated these findings with matched whole-genome sequencing (WGS) and FISH to validate genomic architecture. We further assessed conservation of transcriptional and genomic programs between tumors and in-vitro models. Results: Using single-cell CNV analysis, we identified malignant cells that exhibited strong transcriptional similarity to adipocyte stem progenitor cells (ASPCs), yet were distinguished by consistent upregulation of oncogenic programs, including MYC target genes. We defined four malignant clonal profiles based on MDM2 and/or CDK4 amplification patterns with specific transcriptomic profiles. Amplifications on chromosome 12 frequently co-occurred with gains on chromosomes 1 and 6, contributing to substantial inter-patient variability. Interestingly, the intra-patient variability of the clonal profiles at single cell resolution also varied markedly between patients, highlighting the extent of clonal heterogeneity and potential complexities associated with treatment response. CNV profiles inferred from scRNA-seq data showed high concordance with matched FISH and whole genome sequencing (WGS) results. Notably, patient-derived sarcoma cell (PDSC) models preserved key CNVs and somatic mutations from their respective tumors, supporting their utility as faithful, patient-specific models of the disease. Conclusions: This single-cell atlas reveals new insights into clonal heterogeneity and transcriptional plasticity in liposarcoma. Our findings deepen the understanding of tumor diversity and evolution, support the fidelity of our PDSC models, and provide a molecular framework for stratifying malignant liposarcoma clonal profiles and support the use of PDSCs for preclinical evaluation of targeted therapies. Citation Format: Camino RUANO SM, Brandon Xuan Ming PHUA, Nur Nadiah ISMAIL, Zheng Xi YAP, Dexter Kai Hao THNG, Benjamin Jieming CHEN, Vivian Yujing LIM, Nur Fazlin Bte Mohamed Noor, Lee Yeah, Haopeng Tong, Tse Hui LIM, Xin Xiu SAM, Chin-Ann Johnny ONG, Claramae Shulyn CHIA, Timothy Kwang Yong TAY, Melissa Ng, Dean HO, Tan Boon TOH, Valerie Shiwen YANG, Xing Yi WOO. Liposarcoma Clonal Complexity: A Multi-Modal Single-Cell and Genomic Analysis [abstract]. In: Proceedings of Frontiers in Cancer Science 2025; 2025 Nov 5-7; Singapore. Philadelphia (PA): AACR; Cancer Res 2026;86(13_Suppl):Abstract nr P62.