Computational immunophenotypic profiling of multiple myeloma by flow cytometry and bioinformatics integration

Abstract

Abstract Introduction: Multiple myeloma (MM) is a hematologic malignancy characterized by clonal proliferation of plasma cells in the bone marrow and monoclonal immunoglobulin production (Figure 1). It is often preceded by asymptomatic precursor conditions such as monoclonal gammopathy of undetermined significance (MGUS) and smoldering myeloma (SMM). Multiparameter flow cytometry (MFC) plays a central role in immunophenotypic characterization, identifying maturational asynchrony and aberrant antigen expression, such as CD56 and CD117. This study, conducted at Hemocentro Dalton Cunha (Hemonorte), integrates R-based computational analysis to better understand the immunophenotypic landscape of MM, thereby aiming to enhance diagnostic accuracy and precision medicine strategies.Objective: To explore the immunophenotypic profile of plasma cell dyscrasias using R-based computational analysis applied to flow cytometry data.Methods: MFC data from 145 patients (2023–2025) with either confirmed MM or inconclusive diagnoses were analyzed. Samples were processed using standardized protocols and acquired on a DxFLEX cytometer (Beckman Coulter). Data preprocessing in R encompassed compensation, log transformation, and plasma cell gating (CD38⁺/CD138⁺). FlowSOM was used to identify cell clusters. Statistical comparisons (Wilcoxon test, correlation analysis) and heatmaps (using the ComplexHeatmap package) were employed to assess immunophenotypic differences between groups.Results: Demographic analysis revealed a slight female predominance (54%) and an overall average age of 66 years, with MM cases specifically averaging 67 years. FlowSOM confirmed an increase in plasma cell populations in MM. Variability in median fluorescence intensity (MFI) for CD19, CD20, CD45, CD38, Kappa, and Lambda light chains indicated maturational asynchrony and clonal expansion. Increased expression of CD56 and CD117 reflected prognostic aberrations, while CD138 and CD28 remained stable. Correlation analysis revealed strong associations between CD38/CD138 and aberrant markers (CD56/CD117), alongside a loss of correlation with CD45 and CD3, collectively highlighting distinct immunophenotypic remodeling in MM.Discussion: Our findings support the value of MFC combined with computational tools for MM profiling, aligning with existing literature that highlights aberrant CD56/CD117 expression and maturation defects. Bioinformatics significantly improved the identification of clonal patterns and aberrant cell subpopulations, further reinforcing this integrated approach for personalized diagnostics. However, validation in larger cohorts and rigorous methodological standardization remain essential for clinical translation.Conclusion: This study effectively underscores the utility of multiparametric flow cytometry integrated with computational analysis in characterizing the immunophenotypic complexity of multiple myeloma (MM). It identifies key maturation and phenotypic aberrations relevant for diagnosis and prognosis, thereby supporting precision medicine approaches in onco-hematology, while simultaneously highlighting the need for broader validation.