Abstract
The main cause of mortality in the thalassemia population remains iron-induced cardiac failure (Borga-Pignatti et al Ann N Y Acad Sci 2005); in addition iron overload in the liver, pancreas and other organs causes progressive damage . Iron overload in human tissues can be treated by chelation therapy. Thus, early detection of iron overload is crucial. Nowdays liver iron overload in human tissues can be monitored noninvasively by magnetic resonance imaging (MRI) by two techniques, T2* and R2 MRI (Ferriscan®). There is not too much literature that compares the two methods in hemoglobinopathies. Our center instituted a network, LICNET (Liver Iron Cutino Network), promoted from Piera Cutino partnership and addressed to the diagnostics of iron overload in liver by R2 MRI in patients with hemoglobinopathies.
Patients with thalassemia Major (TM), thalassemia intermedia (TI) and Sickle-Cell/b-thalassemia (S/b-T)), were retrospectively considered for this study. Primary endpoint was to evaluate agreement between T2* and R2 MRI measures of liver iron concentration (LIC) using a Bland-Altman (B-A) method that compares differences between observations on the same patient made with the two methods (Bland & Altman Lancet 1986). Secondary endpoints were to evaluate: 1) hepatic iron overload in our population; 2) difference in R2 LIC in patients with different chelation regimen; 3) relation between hepatic iron overload versus transfusion requirements. LIC was measured by calculating T2* and by measuring R2 using commercial Ferriscan® technique (St Pierre TG et al Blood 2005). To convert liver T2* to LIC a regression equation was used: LIC T2*=0.0254×R2*+0.202 (where R2*=1000/T2*) (Wood JC et al Blood 2005).
LICNET involves 14 Italian thalassemia and radiology centers. Overall 301 adult patients with hemoglobinopathies (TM (177), TI (74) and S/b-T (50)) underwent to iron evaluation from 2012 to 2014. The mean age at R2 MRI evaluation was 33.2±10.7, 41.2±13.8 and 38.7±13.9, respectively in TM, TI and S/b-T. Iron overload was assessed in patients where most of the patients have been treated with deferasirox (DFX) therapy (TM (28.8%), TI (25.7%) and S/b-T (26.0%)), the remaining cohorts were treated with deferoxamine (DFO), deferiprone (DFP) chelation both alone and in combination or sequential administration.
One hundred and twelve observations were measured both for T2* and R2. Concerning the primary endpoint, in the B-A plot it was observed that T2* method yielded a higher LIC than Ferriscan (differences>0), the estimated bias (estimated mean difference) was 2.6 (95% LoA – 17.8; 22.9), and this difference increased at high levels of iron content (Estim. Diff= -1.18+0.32Average mg/g/dw, p=0.0001) (Fig. 1). Secondary endpoints showed that hepatic iron overload determined by T2* was not statistically different among 3 cohorts of patients while it was border line by LIC-R2 (p=0.2608 and p=0.0672). Furthermore, DFX treated patients showed lower LIC-R2 determination in comparison with other treatment (Table 1). Finally, the increase of transfusion requirements was not associated with more severe iron overload in patients with TI and S/b-T. This may be in relation with compliance and type of chelation treatment. These findings show that LIC-R2 (Ferriscan®) is crucial to have accurate and reliable measures for iron body burden control in hemoglobinopathies.
Table 1. Liver iron concentration determined by Ferriscan (R2) in patients with hemoglobinopathies treated by different chelation regimens. TM TI S/ b -T Chelation Therapy LIC R2 (mean±sd) LIC R2 (mean±sd) LIC R2 (mean±sd) DFO 5.3±5.7 8.5±7.7 20.9±19.9 DFP 12.9±12.3 12.5±8.1 12.7±20.2 DFX 7.6±9.2 6.1±7.1 3.7±3.2 Combined DFO+DFP 10.1±12.1 17.8 (n=1) --- Sequential DFO-DFP 4.3±3.1 --- --- Combined DFO+DFX --- 9.7 (n=1) ---
Figure 1. Bland- Altman plot of Liver iron concentration: difference LIC T2* and LIC-R2 versus average of values measured by T2* and Ferriscan Figure 1. Bland- Altman plot of Liver iron concentration: difference LIC T2* and LIC-R2 versus average of values measured by T2* and Ferriscan
Disclosures
No relevant conflicts of interest to declare.
Proteomic analysis of hepatic iron overload in mice suggests dysregulation of urea cycle, impairment of fatty acid oxidation, and changes in the methylation cycle