Cluster analysis of lymphocyte subset from peripheral blood in newly diagnosed idiopathic aplastic anaemia patients

Cluster analysis classified patients into four distinctive subgroups: Cluster 1 (CD4⁺ T cells dominant), Cluster 2 (CD8⁺ T cells dominant), Cluster 3 (NK cells dominant) and Cluster 4 (B cells dominant). Patients in Clusters 1 and 4 suffered more severe disease status than ones in Clusters 2 and 3 (p = .013). And with it, patients in Cluster 2 had the highest white blood cell count, haemoglobin level, reticulocyte count and reticulocyte percentage, while patients in Cluster 3 had the lowest lymphocyte percentage and the highest neutrophil count (all p < .05). Unexpectedly, patients in Cluster 3 tended to have superior curative effect than ones in other clusters, an ordinal logistic regression analysis further confirmed the independent correlation between Cluster 3 and good response to treatment. Lymphocyte subset clustering may serve as a biomarker for assessing disease severity and treatment efficacy in newly diagnosed IAA patients.Key MessagesNewly diagnosed IAA patients could be classified into 4 distinctive subgroups with similar immune patterns by using cluster analysis of lymphocyte subset.Clusters of lymphocyte subset were closely correlated with disease severity and treatment response of IAA.Lymphocyte subset clustering may serve as a promising tool for assessing disease severity and treatment efficacy in newly diagnosed IAA patients.