The Emerging Role of Liver Stiffness Measurement in Transfusion Dependent Thalassemia

β-thalassemia is a globally distributed hereditary red blood cell disorder whose clinical management still relies on chronic red blood cell transfusions combined with iron chelation therapy. In transfusion-dependent β-thalassemia (TDT), long-term monitoring of liver and cardiac iron overload is essential to prevent severe organ damage. Prolonged iron exposure, particularly when combined with virus-related chronic hepatitis, acts synergistically to promote liver fibrosis and progression to cirrhosis [1]. Previous studies have suggested that liver fibrosis may reflect years of sustained oxidative stress related to iron overload affecting multiple organs [2]. Consistently, experience in the allogeneic transplant setting shows that iron-related organ damage negatively impacts transplant-related mortality [3]. To investigate the relationship between liver iron overload and liver fibrosis in TDT using non-invasive approaches, we analyzed data from a retrospective cohort of 831 TDT patients followed between 2010 and 2019 at four Italian reference centers for hemoglobinopathies using the Webthal computerized medical record system [4]. From this cohort, we selected 165 patients, who underwent liver stiffness measurement (LSM) and liver iron concentration (LIC) assessment within the same year (Figure S1). Patients with a diagnosis of hepatocellular carcinoma, HIV or HBV infection were excluded. No intrahepatic mass-like of extramedullar hematopoiesis was present in our cohort. When multiple paired evaluations were available, the most recent concurrent assessment was considered. Concerning HCV Ab positive patients, the evaluation after direct-acting anti-viral drugs treatment was considered (Figure S2) [5]. Liver stiffness (LSM) was assessed by transient elastography (FibroScan) according to European Association for the Study of the Liver (EASL) guidelines [6]. Based on EASL guidelines, a cutoff of 8 kPa was used since LSM below 8 kPa is associated with a very low probability of advanced liver fibrosis (bridging fibrosis) [6]. In our cohort, the median age at evaluation was 38.5 years (Q1–Q3: 30.5–43.0) and 51% were male (Table S1). The median LSM was 5.2 kPa (Q1–Q3: 4.3–6.3). A total of 151 patients (91.5%) had no advanced liver fibrosis (LSM < 8 kPa; median 5.1 kPa, Q1–Q3: 4.2–5.9), whereas 14 patients (8.5%) had LSM ≥ 8 kPa (median 10.35 kPa, Q1–Q3: 8.8–12.3). The median serum ferritin level was 1207 ng/mL (Q1–Q3: 467–2522) (Table S1). Overall, 100 patients (61%) were Hepatitis C Virus Antibodies (HCV Ab) positive (Figure 1A), 101 (61%) had undergone splenectomy, and 23 (14%) had diabetes. The median LIC [7-9] was 3.6 mg/g dw (Q1–Q3: 1.9–8.3) distributed as follows: 17 patients (10%) had severe overload (LIC ≥ 15 mg/g), 34 (21%) moderate overload (LIC ≥ 7 and < 15 mg/g dw), and 38 (23%) mild (LIC ≥ 3 and < 7 mg/g dw) and 76 (46%) no overload (LIC < 3 mg/g dw) (Figure 1B). Twenty patients (12%) had abnormal liver function tests (Figure 1C). The univariate analyses showed significant associations between LSM and most liver biochemical parameters (aspartate aminotransferase [AST] p < 0.001, alanine aminotransferase [ALT] p = 0.005, gamma GT p = 0.022) and a negative, not significant association with total (p = 0.2) and indirect bilirubin (p = 0.15). A positive, not significant association was observed between LSM and sex (male vs. female, p = 0.08), LIC (p = 0.09), HCV Ab status (positive vs. negative, p = 0.12), splenectomy (yes vs. no, p = 0.14), and hypoparathyroidism (p = 0.08). The positive associations with age and male sex are consistent with findings in other chronic liver disease models, including non-alcoholic steatohepatitis [10, 11]. The positive association between liver stiffness and splenectomy in TDT (Table S2) contrasts with reports of reduced liver fibrosis after splenectomy in non-thalassemic populations [12, 13], underscoring the unique pathophysiology of TDT and the detrimental impact of prolonged exposure to cytotoxic factors such as iron, which can be aggravated with splenectomy in view of its scavenging role for non-transferrin-bound iron species. Overall, compared to HCV Ab negative patients (n = 65), HCV Ab positive patients (n = 100) had lower LIC values (2.8 mg/g dw (Q1–Q3: 1.8–5.9) vs. 6.5 mg/g dw (Q1–Q3: 2.6–12.2), p < 0.001) and higher liver stiffness (5.5 KPa (Q1–Q3: 4.3–6.8) vs. 5.1 KPa (Q1–Q3: 4.2–5.8), p = 0.1—Table S3). Among patients with LSM < 8 kPa (n = 151, no advanced liver fibrosis), the LIC was distributed as follows: (Table S4, Figure 1B): 14 patients (9.3%) had severe overload, 34 (23%) moderate overload, 33 (22%) mild overload and 70 (46%) no overload. Considering the model that investigates the association between LSM and LIC (Table S5, Figure 1D,E), adjusted for age and sex and stratifying by LSM groups and HCV Ab Status, the association between LSM and LIC remained significant only in HCV-negative patients. This highlights how the cumulative effect of different liver-toxic factors is reflected by LSM more than LIC. This is also supported by our observation on patients with LSM ≥ 8 kPa (n = 14). Among them 10/14 (71%) were HCV positive, the median LIC was 4.7 mg/g dw (Q1–Q3: 1.7–6.5): 3 patients (21%) had severe overload, 5 (36%) mild and 6 (43%) no overload (Table 1, Figure 1F,G). No association among LSM and LIC was found in these patients, who were on a stable dose of oral iron chelation therapy at least in the 6 months before the analysis, allowing us to exclude the possible confounding factor related to intensive iron chelation. Notably, in a cross-sectional study, Di Marco et al. previously reported that liver stiffness increases independently of LIC and proportionally to liver fibrosis in a small TDT cohort (Table S6) [14-20]. Our literature review revealed substantial heterogeneity in the reported correlation between LIC and liver stiffness in TDT populations (Table S6). This variability likely reflects small sample sizes, younger population and higher serum ferritin and LIC levels compared with our cohort (Table S6). Additionally, our study involved 2.4% of patients < 12 years of age for whom there is still a lack of international validation on liver stiffness evaluation by FibroScan. Our findings highlight the effect of liver iron overload and HCV infection on liver stiffness and support the added value of LSM over LIC as an indicator of cumulative oxidative liver injury in TDT patients. Thus, the early recognition of liver fibrosis by non-invasive approach plays a key role in the identification of TDT patients requiring the removal/treatment of liver-toxic factors such as HCV, alcohol, smoke, or metabolic disorders such as diabetes and/or the revision of the intensity/quality of iron chelation strategy. Our data suggest that LSM ≥ 8 kPa better reflects the chronic hepatic disease burden in TDT than LIC, which represents a snapshot. This is extremely important in the perspective of aging in TDT patients and of curative treatments such as HSCT or gene therapy. Indeed, the loss of LSM/LIC correlation when LSM is > 8 kPa strongly supports the use of LSM into annual follow-up alongside LIC in TDT patients. Study conception: G.L.F., L.D.F. Data collection: G.B.F., F.L., S.B., R.O., R.L., A.P. Statistical analysis: B.G. Writing – original draft: B.G., L.D.F., G.L.F.; Review and interpretation: L.D.F., G.L.F., B.G, G.B.F., F.L., S.B., R.O., R.L., A.P., L.B. The authors would like to thank Antonia Gigante for her support. The authors have nothing to report. The study was approved by the local Ethics Committees of participating centers, and all patients provided written informed consent for data collection and use. The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request. Figure S1: Patient selection flow diagram. With concurrent evaluation it is intended evaluation of LSM and LIC in the same year. Figure S2: Longitudinal trend of LSM during the period of observation in HCV Ab negative and positive patients. Table S1: Demographic and clinical characteristics of the study population. Table S2: Demographic, biochemical and histological features in patients with and without spleen. Table S3: Demographic, biochemical and histological features in patients with HCV Ab positive vs. HCV Ab negative. Table S4: Characteristics of patients by LSM category. Table S5: Association of age, sex, and liver iron concentration (LIC) with liver stiffness measurements (LSM), stratified by HCV antibody status. Table S6: Comparison of main characteristics of studies on liver stiffness and iron overload in TDT patients. Means and standard deviations (SD) or their estimation are reported. Supporting Information: is available online and includes additional details on data collection, variable categorization and statistical analysis and supporting tables and figures. Specifically, Figure S1 provides a patient selection flow-diagram, Figure S2 provides the longitudinal trend of LSM during the period of observation in HCV Ab negative and positive patients. Table S1 provides demographic and clinical characteristics of patients, Tables S2–S4 summarize analyses by splenectomy, HCV Ab Status and LSM category; Table S5 reports the association of age, sex, and liver iron concentration (LIC) with LSM, stratified by HCV Ab status; Table S6 reports the comparison of main characteristics of studies on liver stiffness and iron overload in TDT patients. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.