Retrieving single-cells and cell clusters from cytology direct smear slides for genomic evaluation

Abstract

Abstract Introduction/Objective The key role of cytopathologists is to generate diagnoses based on cytologically stained samples. However, diagnoses increasingly need genomic data in addition to morphologic information. By current methods, morphologically diagnostic samples may be inadequate to perform comprehensive ancillary testing. Therefore, there is a great need for a workflow that can reliably identify and retrieve target cells based on their morphological features. This workflow should also enable highly sensitive molecular profiling from individual cells or small clusters of cells, allowing accurate detection of genomic alterations like copy number variations (CNVs) and single nucleotide variations (SNVs). Methods/Case Report Previously, we established a workflow that combines the RareCyte (Seattle, WA) cell-retrieval platform with primary template-directed amplification (PTA) to generate whole genome libraries from Propidium iodide (PI) stained effusion tumor single-cells or cell clusters and detected copy number variation. In this study, we modified and expanded the assay to identify and retrieve malignant or suspicious cells from Diff-Quik direct smear slides from patient pleural effusion specimens. The samples were further subjected to PTA amplification and library preparation for future copy number variation (CNV) evaluation. In parallel, patient cells from PI-stained smear slides were also obtained and processed to serve as controls. Results (if a Case Study enter NA) Direct smear slides were made from patient pleural fluid (N=5). In contrast to the previous PI staining, which only highlighted nuclei under fluorescence microscopy, Diff-Quik staining allowed nuclear and cytoplasmic morphology evaluation to identify target cells via light microscope. The XYZ coordinates of each target cell or cluster were further documented in the RareCyte Imager software to facilitate cell retrieval. We have successfully collected 39 single-cell and 18 bulk-cell samples from 5 different cases. PTA amplification has been completed and DNA yields of amplified libraries are promising. Conclusion We demonstrated that our modified cell retrieval platform can reliably obtain both single-cell or cluster samples based on morphology from cytologic smear slides made from patient effusion specimens. These collected samples were shown to be compatible with downstream amplification and library preparation enabling CNV evaluation and future deep sequencing for diagnostic, prognostic, and predictive pathogenic mutation identification.