Classification models based on the level of metals in hair and nails of laryngeal cancer patients: diagnosis support or rather speculation?

Abstract

The etiology of cancer is complex, and the disturbances in toxic and essential metals homeostasis are among many of the factors that lead to the development of malignancy. The aim of this study is to investigate the relationship between cancer risk and element status as well as cancer risk and external factors, such as diet, smoking and drinking habits, in order to support diagnosis of cancer. The samples of hair and nails obtained from patients with larynx cancer and healthy subjects were analyzed. Essential elements (Ca, Cr, Mg, Zn, Cu, Mn, and Fe), besides toxic metals (Cd, Co, and Pb), were determined using inductively coupled plasma atomic emission spectrometry (ICP-OES) and mass spectrometry (ICP-MS) techniques. The concentration of essential elements was from 1.5- (Zn) to 4.7-fold (Fe) higher in hair and from 2.4- to 3.3-fold higher in the nails of the control group compared to the patients, while the opposite trend was observed for the heavy metals. The differences between two groups in the level of metals (except for Zn) were statistically significant (p < 0.05). The association of cancer with metals and other factors was evaluated using various statistical methods, for which the best predictions were obtained using logistic regression, artificial neural networks and canonical discriminant analysis. The classifiers constructed using the data from a survey of diet and lifestyle, and analysis of elements in hair and nails, can be useful tools for estimating cancer risk and early screening of the disease.