Abstract 3478: Feasibility of single-cell transcriptomic profiling of pleural effusions from advanced-stage cancer patients

Abstract The pleura is a frequent metastatic site during the evolution of cancers, which can lead to a symptomatic accumulation of pleural fluid that contains tumor and immune cells. To improve patients’ comfort this fluid is removed via chest draining, and the recovered liquid is generally considered as biological waste. Here, we aim to develop an optimized protocol for single-cell sequencing of floating cells in pleural effusion from advanced-stage cancer patients. To date, pleural effusion samples have been obtained from treatment-naïve, advanced-stage cancer patients (lung n= 8, breast n= 3; ovary n=4, others n=4). After red blood cell lysis, 768 cells/sample were isolated using the image-based cellenONE technology. Single-cell RNA sequencing was performed using SORT-seq. Bioinformatic analysis was done using an in-house pipeline on cells passing these filters: >200 detected genes per cell, >500 UMIs per cell, <25% of UMIs mapping on mitochondrial genes and <20% of spike-in reads. Pleural samples exhibited very heterogeneous data quality, which was reflected in a marked difference in the quality control (Qc) metrics between the samples but was not influenced by tumor type. After Qc filtering, the median number of analyzable cells/sample (mNAC) was 153 [5 - 460], for which the median number of expressed genes/cell per sample (mNEG) was 2,431 [1,022 - 4,026]. The time elapsed between patient sample collection and the initiation of single-cell sorting likely impacted data quality (>3h: mNAC: 99; mNEG: 1,553) compared to samples processed more rapidly (<3h: mNAC: 169; mNEG: 2,492), although not reaching statistical significance. A higher isolation frequency and recovery rate during single-cell sorting resulted in the capture of higher quality cells. The duration of single-cell isolation varied between 11’ and 100’ (median: 21’) and was dependent on the initial cell concentration. No significant differences were observed in terms of cell diameter or elongation with regards to cancer type. However, cell diameter was correlated with mNAC. For cell type inference, we used SingleR with the Human Pan-Cancer Atlas database as reference. Epithelial cells were the most predominant cell type in these samples, followed by subpopulations of monocytes and macrophages. A combination of the “mean variance” and “rank based” methods was used to obtain the list of the most variable genes, to define expression clusters and discern samples by cancer type. Calculation of the enrichment of transcriptomic signatures and therapy response signatures (i.e., IFN-gamma) was used to apprehend the potential biological differences between samples and cell subpopulations. Results will be presented at the conference. This protocol is currently applicable to analyze floating cells from pleural effusions and is expected to provide a complementary source for the molecular profiling of advanced-stage cancers. Citation Format: Sandra Ortiz-Cuaran, Lucas Michon, Marion Godefroy, Joyce Levy, Osman Osman, Maxime Boussageon, Aurélie Swalduz, Maurice Pérol, Bruno Russias, François Monjaret, Pierre Saintigny. Feasibility of single-cell transcriptomic profiling of pleural effusions from advanced-stage cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 3478.