PF790 RELATIONSHIP BETWEEN LIVER IRON CONCENTRATION MEASURED BY MRI T2∗ AND SERUM FERRITIN IN HEMOGLOBIN H DISEASE

Abstract

Background:Patients with non‐transfusion dependent thalassemia (NTDT) are at risk of developing liver iron overload. T2∗‐weighted magnetic resonance imaging (MRI T2∗) is an accurate and reproducible method to determine liver iron concentration (LIC) but it is not always readily available especially in resources limited setting. Serum ferritin (SF) level remains important in the quantification of body iron. Past studies demonstrated (1) a positive correlation between LIC and SF in β‐thalassemia intermedia (TI) and transfusion dependent thalassemia (TDT), and (2) lower SF values in TI patients than in TDT patients with comparable LICs because of preferential iron accumulation in the liver in TI.Aims:The relationship between LIC measured by MRI T2∗ and SF in hemoglobin H (HbH) disease, the most common type of NTDT in Southern China, was determined in this study.Methods:Data was obtained from 82 patients with HbH disease (male 34%, mean age 51 ± 15 years) identified in a cross‐sectional single‐center study performed during Jan 2017 to Jun 2018 to evaluate the complication profile of Southern Chinese adults with NTDT. Data collection included review of medical charts, blood samples tested for steady‐state SF, and MRI T2∗ (SIEMENS 1.5T Aera) to determine LIC. SF and LIC data from 27 TDT patients (male 52%, mean age 32 ± 8 years) managed in the center during the same period were collected for comparison. Nonparametric analysis of covariance (ANCOVA) was used to evaluate for any difference in the LIC‐SF relationship in patients with HbH disease and TDT. The LIC‐SF relationship was subsequently studied by regression analysis. Transformation of variables was adopted to handle nonlinearity of residuals. Robust regression with robust sandwich variance estimator was used to address the problem of outliers, influential observations and heteroscedasticity.Results:Among patients with HbH disease, 63 (77%) and 19 (23%) were of deletional vs. non‐deletional type respectively, 45 (55%) received blood transfusion before, 10 (12%) had history of iron chelation therapy, 8 (10%) were splenectomized, 3 (4%) and 2 (2%) were hepatitis B and C virus carriers respectively. All TDT patients received blood transfusions every 3–4 weeks and iron chelation. Median SF was 394 (range 19–2,159) μg/L and median LIC was 3.14 (range 0.23–14.24) mg Fe/g dry weight (dw) in HbH disease patients. Median SF was 1,325 (range 79–5,409) μg/L and median LIC was 3.94 (range 1.12–9.27) mg Fe/g dw in TDT patients. ANCOVA showed that the relationship between LIC and SF was different in HbH disease and TDT (p < 0.001) and thus regression analysis was performed separately in the two groups of patients. The models best predicted LIC when SF was raised to the power of 0.5 and 0.3 in HbH disease and TDT respectively. LIC was positively correlated with SF0.5 in HbH disease (p < 0.001) and SF0.3 in TDT (p < 0.001). The estimated regression equation for HbH disease was LIC = 0.21 × SF0.5 – 0.81 (R2 = 0.52; 95% CI for the slope: 0.17–0.25), and for TDT was LIC = 0.97 × SF0.3 – 3.81 (R2 = 0.57; 95% CI for the slope: 0.73–1.20) (Figure 1).Summary/Conclusion:The study demonstrated a positive correlation between LIC and SF. Patients with HbH disease had lower SF values than in TDT patients when their LICs were comparable. SF had moderate predictive ability of LIC in both HbH disease and TDT, suggesting that it is an acceptable marker to estimate liver iron deposition. However, result has to be interpreted cautiously because the relationship between LIC and SF alters with the type of thalassemia syndrome and SF could also be affected by other factors.image