Using cluster analysis with principal component analysis to study the iron metabolism in polycythemia vera

Abstract

Background. Iron deficiency is a common complication in patients with polycythemia vera (PV). Unfortunately, little is known about the pathomechanisms of iron deficiency in PV. There have been no studies in the last decade documenting iron disorders in PV, despite progress in understanding the iron metabolism and new laboratory techniques measuring iron parameters.

Objectives. The aim of this study was to assess the relationships between iron metabolism parameters, haematological and biochemical factors and clinical attributes in polycythemia vera patients with the use of cluster analysis (CA) and principal component analysis (PCA).

Patients and methods. The study was performed on 60 patients (F/M 26/34) aged 38–84 (66 ± 10) years. The following parameters were determined in blood samples: hepcidin, prohepcidin, iron, total iron binding capacity (TIBC), unsaturated iron binding capacity (UIBC), ferritin, soluble transferrin receptor (sTfR), transferrin saturation (TfS), complete blood cell count, erythropoietin (Epo), uric acid, lactate dehydrogenase (LDH).

Results. The CA divided all the 17 parameters into three clusters and showed that hepcidin concentrationis related to the duration of hydroxyurea therapy. PCA also revealed a positive correlation between hepcidin and therapy duration.

Conclusions. We demonstrated that CA and PCA are efficacious methods for assessing the relationship between iron metabolism parameters and clinical attributes in PV patients.