A Correlation Study Between RNA Sequencing and Immunophenotyping in Hematopoietic Malignancies. (21st September 2018)

• Record Type: Journal Article
• Authors: Mo, Huan; Cox, Brian; Wang, Jun; Rowsell, Edward
• Is Part Of: American journal of clinical pathology. Volume 150(2018)Supplement 1
• Page Start: S104
• Page End: S104
• Publication Date: 2018-09-21
• Subjects: Diagnosis, Laboratory -- Periodicals; Pathology -- Periodicals; 616.07
• DOI: 10.1093/ajcp/aqy097.251
• Languages: English
• ISSNs: 0002-9173

Abstract

Abstract: Objectives: RNA sequencing (RNA-seq) provides transcriptomic information. However, whether the RNA transcript levels correlate with immunophenotypes has not yet been fully established. Methods: We evaluated the data sets of the acute myeloid leukemia (LAML) and diffuse large B-cell lymphoma (DLBCL) projects of the Cancer Genome Atlas (TCGA) and extracted the immunophenotype and RSEM-normalized transcript logarithm scores of RNA-seq data. Results: In the LAML data set, the mean RSEM score of CD34 immunophenotypically positive cases is 12.7 (12.3–13.0, n = 108), compared to 5.7 (5.0–6.4, n = 37) of negative cases with P < .01. Other available markers, including CD56/NCAM1, CD14, MPO, and HLA-DR, all show significant associations between immunophenotypes and transcript levels. Similarly, in the DLBCL data set, CD10/MME (10.1 [9.3–10.9], n = 11 compared to 7.0 [6.0–8.0], n = 26), BCL2, MUM1/IRF4, and CD30/TNFRSF8 all demonstrate significant associations ( P < .01). BCL6 is the only exception, where positive (n = 26) and negative (n = 4) cases showed RSEM of ~11.2 each ( P = .96). We also evaluated the lineage markers of admixed tumor and nontumor cell populations in the DLBCL data set (n = 48); the markers and respective RSEM are as follows: CD19 (B cells/tumor) at 12.4 (10.3–14.6), CD3 (T cells) at 10.4 (7.2–13.6), CD13/ANPEP (myeloid) at 8.5 (5.7–11.3), CD163 (histiocytes) at 10.6 (6.2–15.0), CD34 (stromal cells) at 9.0 (7.6–10.3), and TDT/DNTT (negative control) at 0.52 (0.0–2.0). Conclusion: With the exception of BCL6, all immunophenotypic markers show significant associations between transcript levels and immunophenotypes in both data sets. In the DLBCL data set, we also demonstrate that the transcript-level data may be used to evaluate the admixed composition of tumor and nontumor cells.