Explorando o panorama imunofenotípico das discrasias de células plasmáticas: análise computacional com linguagem R em citometria de fluxo

Abstract

The present study analyzed multiparametric flow cytometry (MFC) data from 145 patients treated at the Dalton Cunha Blood Center between 2023 and 2025, with a suspected diagnosis of multiple myeloma (MM). The main objective was to investigate the immunophenotypic characteristics of plasma cells and explore technological advances for the diagnosis and management of the disease. The samples included both cases with a confirmed diagnosis of MM and inconclusive results as controls, and were processed with standardized and proven clinical protocols using the DxFLEX eight-core cytometer (Beckman Coulter). The data were processed with advanced bioinformatics tools based on the R language, mainly using the flowCore, tidyverse and ggplot2 packages. The application of the FlowSOM algorithm allowed the identification of six main metaclusters, with high reliability in the separation of cell populations, highlighting metacluster 1 as representative of plasma cells. These were characterized by maturational asynchrony and specific phenotypic aberrations, including high expression of CD38 and CD138, combined with the absence of CD19 and CD45, in addition to alterations in CD56 and CD117 markers. For the most presented analyses, fluorescence intensity media (FIM) was used to evaluate expression patterns of monoclonal markers in plasma cells, organized by diagnostic groups. The results showed greater dispersion of FIM in the MM group, particularly for the CD38, Kappa and Lambda markers, reflecting maturational asynchrony and plasmacytic clonality. In addition, markers such as CD56 showed higher expression in MM, consolidating itself as one of the main aberrations associated with a worse prognosis. The brightness analysis between the protectors revealed diagnostic relationships, such as coexpression between CD38/CD138 and positivity for CD56 and CD117, characterizing aberrations frequently observed in MM. Furthermore, the absence or decrease of CD45 and CD3 are also observed in MM cases. The use of heatmaps and barplots provided detailed visualizations of the differences between the MM and inconclusive groups, offering additional support for interpretations of the basic biology of cellular data. This study highlights the relevance of the integration between flow cytometry and bioinformatics in the diagnosis and management of MM. The use of robust pipelines allowed in-depth analyses of cellular characteristics and immunophenotypic patterns, contributing to diagnostic decision making and the evolution of precision medicine. Despite limitations, such as the need for a larger number of data and applied machine learning studies, the results reinforce the potential of these approaches in the advancement of oncohematology. Keywords: Multiple myeloma; Plasmocytes; Flow cytometry; Bioinformatics.