Talanta.2016 Oct.159:425-432 

Thermogravimetric analysis coupled with chemometrics as a powerful predictive tool for ß-thalassemia screening

Roberta Risoluti1*, Stefano Materazzi1, Francesco Sorrentino2, Laura Maffei2, Patrizia Caprari3 


1Department of Chemistry, Sapienza – University of Rome, Rome, Italy;

2Thalassemia Unit, S. Eugenio Hospital, Rome, Italy;

3Department of Hematology, Oncology and Molecular Medicine, Istituto Superiore di Sanità, Rome, Italy

* Corresponding author at: Department of Chemistry, Sapienza – University of Rome, Piazzale Aldo Moro 5, 00185 Rome, Italy; Tel +390649913616; fax: +390649387137; e-mail address: roberta.risoluti@uniroma1.it



β-Thalassemia is a hemoglobin genetic disorder characterized by the absence or reduced β-globin chain synthesis, one of the constituents of the adult hemoglobin tetramer. In this study the possibility of using thermogravimetric analysis (TGA) followed by chemometrics as a new approach for β-thalassemia detection is proposed. Blood samples from patients with β-thalassemia were analyzed by the TG7 thermobalance and the resulting curves were compared to those typical of healthy individuals. Principal Component Analysis (PCA) was used to evaluate the correlation between the hematological parameters and the thermogravimetric results. The thermogravimetric profiles of blood samples from β-thalassemia patients were clearly distinct from those of healthy individuals as result of the different quantities of water content and corpuscular fraction. The hematological overview showed significant decreases in the values of red blood cell indices and an increase in red cell distribution width value in thalassemia subjects when compared with those of healthy subjects. The implementation of a predictive model based on Partial Least Square Discriminant Analysis (PLS-DA) for β-thalassemia diagnosis, was performed and validated. This model permitted the discrimination of anemic patients and healthy individuals and was able to detect thalassemia in clinically heterogeneous patients as in the presence of dβ-thalassemia and β-thalassemia combined with Hb Lepore. TGA and Chemometrics are capable of predicting ß-thalassemia syndromes using only a few microliters of blood without any pretreatment and with an hour of analysis time. A fast, rapid and cost-effective diagnostic tool for the β-thalassemia screening is proposed.

Keywords: Thalassemia, Screening, Blood, Thermogravimetric Analysis, Chemometrics, Class-modelling

Abbreviations: TGA, thermogravimetric analysis; PCA, Principal Component Analysis; PLS-DA, Partial Least Square Discriminant Analysis; Hb, hemoglobin; MCV, mean corpuscular volume; MCH, mean corpuscular hemoglobin; HbF, fetal hemoglobin; TG, thermogravimetry;; RBC, red blood cell; Hct, hematocrit; RDW, red cell distribution width.



β-Thalassemia is worldwide recognized as hereditary hemoglobin disorder characterized by a deficiency in the β-globin chain production.

The β-thalassemia carrier state, β-thalassemia intermedia, and β-thalassemia major are three clinical and hematological conditions of increasing severity [1,2] and the hematological features of β-thalassemia are microcytosis, hypochromia, and anemia. The first level tests for β-thalassemia screening consist of the complete blood count and the osmotic fragility while the confirmatory diagnosis requires the extensive analysis techniques based on genetic investigations.

The aim of this study was the application of the thermogravimetric analysis (TGA) associated with chemometrics as a new method for the detection of β-thalassemia disease.



Blood samples were collected in K2EDTA from β-thalassemia patients and healthy individuals analyzed by a Perkin Elmer TGA7 Thermobalance (Massachusetts, USA) where thirty µl of blood sample were heated from 20°C to 800°C in oxidizing conditions, with a 10°C/min heating rate.

In order to characterize the severity of anemia, the hematological parameters were determined by the ADVIA 120 hematology analyzer.

The experimental design of the study is reported in figure 1.




Fig. 1.      Analytical strategy followed in the study



The thermogravimetric profiles of blood samples from β-thalassemic patients are found to be distinct from those of healthy individuals resulting in different amounts of water content and corpuscular fraction. In particular, the calculation of the first derivative curve of the thermograms reveals a different distribution of the bulk and bound water. The associated chemometric study permitted to correlate the thermogravimetric curves of investigated samples to a typical blood composition of the two populations. The method was fully validated and successfully applied for the prediction of thalassemia with a 100 % of correct classification rate, compared to confirmatory tests.

The parallel hematological overview confirms significant decrease in Hb, Hct, MCH, and MCV values and increase in RBC counts and RDW values in thalassemic subjects as compared with healthy subjects.



This study provides several important advances in hematological screening frontiers: a) authors developed a novel diagnostic tool that requires no pretreatment of a reduced amount of blood sample (few microliters); b) the bulk/bound water ratio demonstrated to be significant in β-thalassemia diagnosis, suggesting to be considered also in different hereditary hematological disease screening; the validated model is capable of detecting thalassemia in patients with heterogeneous severity including the association of dβ-thalassemia and β-thalassemia combined with Hb Lepore, usually requiring the molecular analysis for diagnosis and in those cases involving recently transfused patients, a circumstance in which the first level test are not performed but delayed in time; c) once validated, the model can be easily applied by not high-skilled personnel, thus resulting in real rapid and cost-effective tool for biomedical applications.


[1] D.J. Weatherall, J.B. Clegg. The Thalassemia Syndromes. Oxford (U.K.): Blackwell Science, 2001.

[2] A. Cao, R. Galanello. Beta-thalassemia, Gen. Med. 12 (2010) 61-76.


Multiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier SchönmannMultiselect Ultimate Query Plugin by InoPlugs Web Design Vienna | Webdesign Wien and Juwelier Schönmann