单细胞RNAseq分析评估前列腺穿刺活检标本中癌细胞的新方法

Background : The core biopsy of prostate for detection of prostate cancer (PCa) remain heavily reliant on histopathology, varying by both surgeon and pathologist but carries the risk of misdiagnosis. This can result in missed opportunities for timely intervention or unnecessary treatment strategies. Our study leverages single-cell RNA (scRNA) seq analysis to enhance diagnostic accuracy, aiming to identify PCa cells by a novel precision oncology approach at the cellular and molecular level. Methods : We investigated 23 biopsy cases from patients with potential PCa. Core fine needle biopsies were performed on patients undergoing MRI-fusion image-guided prostate biopsy prior to decision to perform robotic-assisted radical prostatectomy if biopsy is diagnosed PCa positive. We used the pico-well-based HIVE™ (HoneyComb Biotech, MA) to isolate barcoded sc from the cell suspension of prostate tissue biopsy. This method allows specific physical single-cell analysis followed by in-depth scRNAseq. Each sample underwent traditional pathology diagnosis at post biopsy and prostatectomy. Initially, cell type markers were used to classify each clusters cell type from the scRNA seq data. For known PCa gene markers, we assessed KLK3, FOLH1, PCA3, KRT34, AMACR, and TP63. To improve the accuracy of the subset identified as PCa cells, Single Cell Variational Aneuploidy analysis (SCEVAN) classification was used to categorize subsets of cells based on calculated aneuploidy level. This algorithm accurately performs a variational deconvolution to unravel the clonal substructure of tumors using the scRNAseq data. It employs a multichannel segmentation approach based on the premise that cells within a particular copy number clone have similar breakpoints. Results : Our study included 23 patients categorized into 5 benign cases versus specific PCa Gleason grades: 4 GG1, 5 GG2, 3 GG3, 3 GG4, and 3 GG5 based on standard histopathology. The subsets of tumors classified using 2 approaches; deconvolution with cell mRNA markers and deconvolution of marker expression profiles with SCEVAN classification, were consistent. These tumor subsets analysis by the 2 analytic approaches identified PCa cells. Notably, in cases initially diagnosed as benign, scRNAseq data showed variability; some cases were benign with no detectable PCa cells, while others displayed PCa cells through SCEVAN classification, highlighting the potential of the approach uncovering hidden molecular malignant profiles. Conclusion : Our study demonstrates the efficiency of scRNAseq data in detecting PCa cells, which can improve and further validate conventional histopathology diagnosis. This novel approach allows one to uncover occult PCa cells by molecular profiling, enhancing diagnostic status prior to therapy, and potentially triaging patients for surgery, radiation, or observation.