Ferritin thresholds for cardiac and liver hemosiderosis in β-thalassemia patients: a diagnostic accuracy study

Abstract

Ferritin is frequently used to screen some dire consequences of iron overload in β-thalassemia patients. The study aimed to define the best cutoff point of ferritin to screen for cardiac and liver hemosiderosis in these cases. This was a registry-based study on β-thalassemia patients living throughout Mazandaran province, Iran (n = 1959). In this diagnostic research, the index test was ferritin levels measured by a chemiluminescent immunoassay. As a reference test, T2*-weighted magnetic resonance imaging (T2*-weighted MRI) was applied to determine cardiac and liver hemosiderosis. A cutoff point of 2027 ng/mL for ferritin showed a sensitivity of 50%, specificity 77.4%, PPV 42.1%, and NPV 82.5% for cardiac hemosiderosis (area under curve [AUC] 0.66, 95% CI 0.60–0.71, adjusted odds ratio [OR] 2.05, 95% CI 1.05–4.01). At an optimum cutoff point of 1090 ng/mL, sensitivity 66.7%, specificity 68%, PPV 82.9%, and NPV 46.8% for liver hemosiderosis were estimated (AUC 0.68, 95% CI 0.63–0.73, adjusted OR 3.93, 95% CI 2.02–7.64. The likelihood of cardiac hemosiderosis serum ferritin levels below 2027 ng/mL is 17.5%. Moreover, 82.9% of β-thalassemia patients with serum ferritin levels above 1090 ng/mL may suffer from liver hemosiderosis, regardless of the grades.