Photo Credit: Ipopba

The following is a summary of “Computational Flow Cytometry Accurately Identifies Sezary Cells Based on Simplified Aberrancy and Clonality Features,” published in the January 2024 issue of Dermatology by Seheult et al.

Accurate identification of circulating neoplastic cells, specifically Sezary cells, is pivotal in diagnosing, staging, and prognosticating patients with mycosis fungoides (MF) and Sezary syndrome (SS). Despite recent improvements in flow cytometry, the complex immunophenotype of Sezary cells, combined with potential overlap with reactive T cells, requires advanced analytical expertise.

Researchers conducted a retrospective study investigating the power of machine learning to streamline the analysis, employing just two pre-defined Sezary cell-gating plots. 

They examined 114 samples from 59 SS/MF patients and 66 samples from unique patients with inflammatory dermatoses, the single dimensionality reduction plot effectively highlighted clonal CD3+/CD4+ T-cells.

The results showed receiver operator curve analysis (ROC), an aberrancy scale feature computed through comparisons with controls (area under the curve = 0.98), showcased superior discriminatory performance compared to the loss of CD2 (0.76), CD3 (0.83), CD7 (0.77), and CD26 (0.82) in distinguishing Sezary cells from reactive CD4+ T cells. Their results demonstrated a close alignment with exhaustive expert analysis, showing impeccable positive and negative percent agreements (100% and 99%, respectively) for event classification, as well as precise Sezary cell quantitation (regression slope = 1.003, R squared = 0.9996). 

They concluded the potential of machine learning in simplifying and enhancing the accuracy of Sezary cell identification, offering promising prospects for refining diagnostic and prognostic approaches in patients with MF and SS.

Source: sciencedirect.com/science/article/abs/pii/S0022202X24000198