Abstract 4898: Highly accurate profiling of leukocyte composition from bulk peripheral blood with targeted digital cytometry

Abstract

Abstract Background: Cell profiling methods such as flow/mass cytometry and single-cell RNA-sequencing are powerful tools for quantifying immune composition from healthy and neoplastic tissues. However, only a modest number of markers can be interrogated by the former and the latter remains cost prohibitive for large-scale analysis. Here we demonstrate robust, accurate, and reproducible enumeration of immune cell subsets from 36 whole blood samples using SureSelect XT HS2 RNA sequencing combined with CiberMed’s iSortTM digital cytometry solution. Enrichment of genes in the LM22 signature matrix, a well-established collection of reference profiles for deconvolving 22 human immune subsets, was achieved with two new targeted sequencing panels—Agilent SureSelect CD CiberMed Heme and Agilent SureSelect CD CiberMed Heme + HiRes. Both panels were assessed for their ability to profile leukocyte subsets with CiberMed’s iSortTM Fractions software, an optimized and standardized version of CIBERSORTx with novel enhancements. Methods: Whole blood samples were freshly collected from 36 healthy donors and split into two fractions—one was immediately processed for complete blood count (CBC) and flow cytometric enumeration of major leukocyte populations; the other was stored in PAXgene Blood RNA tubes for subsequent RNA sequencing. A Sysmex system was employed for CBC quantification and a Becton Dickinson 6-color TBNK MultiTest in vitro diagnostic (IVD) assay was employed for enumerating B cells, CD8 T cells, CD4 T cells, NK cells. These data were used as ground truth to assess iSortTM deconvolution performance from targeted and whole-transcriptome bulk RNA sequencing data. RNA-seq expression values were used as input to iSortTM Fractions to deconvolve 22 immune subsets in each sample. Results: Across 7 major populations, cell type fractions determined by iSortTM Fractions were highly concordant with ground truth fractions determined by clinical grade standards (r≥0.96) and exhibited strong reproducibility across technical replicates (r≥0.98). Furthermore, targeted enrichment using the SureSelect CD CiberMed Heme panel reduced the sequencing requirement by nearly 50-fold compared to whole-transcriptome sequencing, while also improving accuracy. Conclusion: These new panels for digital cytometry are being released through the Agilent Community Design program to enable focused, reliable, and high-throughput analysis of cell type composition from peripheral blood samples. Citation Format: Aki Nakao, Aaron M. Newman, Ash A. Alizadeh, Maximilian Diehn, Mistuni Ghosh, Manjula Aliminati, Jayati Ghosh, Kristi Stephenson, Ashutosh Ashutosh. Highly accurate profiling of leukocyte composition from bulk peripheral blood with targeted digital cytometry [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(6_Suppl):Abstract nr 4898.