The central role of liver stifness measurement in adults with transfusion dependent thalassemia (TDT) beyond what could be informed by liver iron concentration

Abstract

Abstract Clinical management of transfusion-dependent β-thalassemia (TDT) is still primarily based on chronic transfusion and iron chelation therapy. Long-term exposure to iron overload and/or hepatitis C virus infection can lead to liver fibrosis, cirrhosis, and hepatocellular carcinoma. Studies in thalassemia and other chronic liver disease established measurement of liver stiffness as a key non-invasive tool to identify patients at high risk of liver fibrosis (EASL guidelines). Such assessment remains extremely important in the setting of eligibility for curative stem/cell therapies. Here, we conducted a retrospective cohort study of 223 TDT patients from 4 Italian center centers using the Webthal® computerized medical record who had at least 1 measurement of liver stiffness (transient elastography, Fibroscan®), during a follow up period from 2010 to 2019. Patients with HIV or hepatocellular carcinoma were excluded. The median age at start of observation was 32 years (Q1-Q3:25-36; 108/223 [52%] female). The median value of liver stiffness over the entire observation period was 5.6 KPa (Q1-Q3: 4.6-7.5), with 48 patients (22%) having values over the cut off of 8 kPa (indicating advanced fibrosis) and the remaining 175 (78%) below. Iron and hepatic indices were as follows (n; median over the observation period, Q1-Q3 interval): serum ferritin (223; 1343 ng/ml, 602-2387), liver iron concentration (LIC) (189; 3.86 mg/g dw, 2.09-6.01), MRI-cardiac T2* (189; 39.5 ms, 30.2-36.5), GGT (223; 20.72 U/L, 14.53-30.33), ALT (223; 31.9 U/L, 20.6-42.5), AST (223; 30.9 U/L, 24.0-37.4), total bilirubin (218; 1.63 mg/dL, 1.15-1.92), indirect bilirubin (218; 1.04 mg/dL, 0.72-1.34). A total of 143 (64%) patients were HCV positive, 133 (60%) underwent splenectomy, 135 (61%) had osteoporosis, 32 (14%) had diabetes, 44 (20%) had hypothyroidism, 11 (5%) had hypoparathyroidism, while there were no cases of pulmonary hypertension or hypogonadism. Bivariate analyses showed a positive association between average liver stiffness over the observation period and age (p<0.001), sex (M vs F, p=0.017), LIC (p=0.11), ferritin (p=0.005), HCV status (positive vs negative, p=0.008) and splenectomy (yes vs no, p=0.017). Of these, only age (p=0.001), gender (p=0.009), and ferritin (p=0.046) remained significant on multivariate analysis. Moreover, evidence (p=0.029) of an association with LIC depending on HCV status was found. Considering the subgroup of patients with liver stiffness <8 KPa (n=175, no advanced liver fibrosis) bivariate analysis showed a positive association between average liver stiffness and age (p=0.014), gender (M vs F, p=0.072), LIC (p=0.027), ferritin (p=0.068), and HCV status (positive vs negative, p=0.004). With the exception of age (p=0.3) all these associations reached or were near the significance threshold at multivariate analysis. In HCV negative patients (n=62) only the association between liver stiffness and LIC remained significant (p=0.013, age p=0.4, gender p=0.2, ferritin p=0.5) while in HCV positive patients, only ferritin approximated the significance threshold (p=0.068, age p=0.5, gender p=0.2, LIC p=0.5). In the subgroup with liver stiffness ≥8 KPa (n=48, 83.3% HCV positive, median LIC: 3.9 mg/g dw, Q1-Q3:2.9-5.2, min-max:1.3-32) no significant association with covariates were found at bivariate analysis. Take together, our data indicate that liver stiffness ≥8 kPa could reflect chronic liver disease history even when LIC is low – further highlighting progressive changes overtime compared with LIC which primarily reflects current/spot hepatic iron status. These data are extremely important in both clinical management and selecting TDT patients for curative therapies.