Factor analysis of hepcidin on cardiac iron overload and fibrosis among thalassemia major children

Background Cardiac iron overload is a common cause of death in thalassemia major and is associated with hepcidin, which is primary iron homeostasis. Therapeutics that could increase hepcidin levels or act as hepcidin agonists might help treat the abnormal iron absorption in individuals with β-thalassemia and related disorders. We aimed to verify the structure and study the relationship of release variables, hepcidin, cardiac iron overload, or fibrosis. Methods A cross-sectional study was conducted among thalassemia majors, aged 6–18 years at Rumah Sakit Anak dan Bunda Harapan Kita Indonesia, between January 2019 and May 2020. Clinical data, hepcidin-25, interacting variables laboratory test, MRIT2* used to assess cardiac iron overload, late gadolinium enhancement (LGE), and soluble suppression of tumoregenicity2 (ST2) to assess fibrosis were studied. The correlation test was performed with SPSS version 20, Amos 22 was used to assess confirmatory factor analysis (CFA), and squared multiple correlation (SMC) was used to determine the proportion of total variation explained by the model. Results We recruited 80 patients, of those 8 (10%) were cardiac iron overload, 5 (5.25%) were fibrosis, and 3 (3,75%) were ST2 > 35 mg/dL. CFA showed variables that interact with hepcidin release were Hb, reticulocyte-He (Re-He), HIF-1α, Immature granulocyte (IG), hs-CRP, IL-6, ferritin, and transferrin saturation. There was no direct hepcidin role in cardiac iron overload and fibrosis. No interacting variable role on hepcidin. SMC of hepcidin on cardiac iron overload was 20%. Conclusion Factor analysis of hepcidin was Hb, Ret-He, HIF-1α, IG, hs-CRP, IL-6, ferritin, and transferrin saturation. No direct hepcidin role in cardiac iron overload and fibrosis. Hepcidin explains approximately 20% of the total variation in cardiac iron overload.


Background
Cardiac iron overload and heart failure a common causes of death in thalassemia major due to repeated transfusion and increased iron absorption [1].Hepcidin plays a crucial role in regulating iron metabolism [2].Its expression is influenced by various pathways, including hypoxia/ erythropoietic (stimulated by anemia and hypoxia), STAT/inflammatory (stimulated by IL-6), and bone morphogenetic protein/BMP-SMAD (stimulated by ferritin, transferrin saturation, and transferrin) [3,4].
Thalassemia β major demonstrated a varying degree of correlation between hepcidin with anemia/hypoxia, infection/inflammation, and iron deposition.Before transfusion, hepcidin might show low, normal, or higher results [5,6], but its expression is still debated [7,8].Severe hepcidin deficiency leads to iron accumulation in the myocardium, which can ultimately result in heart failure and death [9,10].Postmortem examination showed the presence of iron deposits, increased chamber dimensions and mass, and cardiac fibrosis [11].However, a direct correlation between cardiac iron overload and cardiac fibrosis could not be established [12].Therapeutics that could increase hepcidin levels or act as hepcidin agonists might help treat the abnormal iron absorption in individuals with β-thalassemia and related disorders [5].As far as we are aware, there are limited studies on factor analysis of hepcidin on cardiac iron overload and fibrosis.Accordingly, this study aimed to illuminate the direct role of hepcidin and its interaction release variables specifically in terms of cardiac iron overload and cardiac fibrosis.

Methods
A cross-sectional study was conducted at Rumah Sakit Anak dan Bunda Harapan Kita Indonesia, between January 2019 and May 2020, involving a thalassemia major, aged 6-18 years, and ferritin level > 1000 ng/dL.
The study collected clinical findings, hepcidin-25, interacting release variables, MRIT2*, late gadolinium enhancement (LGE), and soluble suppression of tumor-egenicity2 (ST2).Variables of Hb, reticulocytes, Ret-He, and immature granulocytes (IG) were examined using a Sysmex XN, and ferritin was checked using Vidas.Furthermore, the Assay Kit was used to check IL-6, and HIF-1α was examined with ELISA Cloud-Clone Corp, USA [13,14].Hepcidin-25 (bioactive) HS ELISA RE54261 was measured using the ELISA method with Kit Cloud-Clone Corp, USA [15].MRIT2* and LGE were performed one week before transfusion in the Department of Radiology Dr. Cipto Mangunkusumo Hospital and Premier Hospital, Jakarta.Cardiac iron overload was defined as MRIT2* < 20 ms while cardiac fibrosis was defined as enhanced contrast in myocardium during LGE.ST2 was also measured as a fibrosis marker.ST2 was analyzed with a commercially available ELISA (Medical and Biological Laboratories Company, Woburn, MA, USA) [16].
The correlation between variables is considered significant with a p < 0.05.The confirmatory factor analysis (CFA) was conducted on Amos 22 to verify and investigate the pattern of effect within a system structure of observed variables and to test the relationship between variables and myocardial iron overload or fibrosis.The squared multiple correlation (SMC) was used to determine the proportion of total variation explained by mode (r).The SMC model was categorized as fit when it fulfilled the goodness of fit criteria.Ethical clearance was provided by the Ethical Research Committee (Faculty of Medicine University of Indonesia S-996/UN2.F1/ETIK/ PPM.00.02/2019), and informed consent was obtained from the parents or guardians of participants.

Results
A total of 80 thalassemia majors confirmed by Hb analysis, aged 6-18 years, were recruited.Clinical findings showed no difference in sex distribution.Baseline characteristics can be seen in Table 1.Furthermore, the range of upper arm circumference (UAC) was 1.4 − 24.4 cm and the UAC age ratio was classified into three, namely good nutritional status (29 subjects/33.3%),malnutrition (24 subjects/27.3%),and poor nutrition (27 subjects/30,7%).
The CFA on path analysis confirmed the fitted interacting variables of anemia/hypoxia, infection/inflammation,  and iron storage to be Hb, Ret-He, HIF-1, hs-CRP, IL-6, ferritin, and transferrin saturation.LGE did not provide a better fit on path analysis, as shown in Fig. 1.SEM shows that the regression weight (r) of hepcidin is 0.11, MRIT2* 0.20, and ST2 0.37.The proportion of total variation explained by interacting release variables on hepcidin was 11%, 20% on myocardial, and the total variation cardiac iron overload on cardiac fibrosis was 37%.Hepcidin did not show a direct role in MRTI2*, as indicated by p = 0.95 and r = 0.01.Similarly, MRIT2* did not show a direct role in ST2, as indicated by p = 0.88 and r = − 0.02.

Discussion
In this study, the mean age range was 12.3 ± 2.8 years, which was not significantly different from other studies [17].Based on the analysis, the baseline data might be considered a general characteristic among thalassemia majors aged 6-18 years in Indonesia as approximately 30% of patients originated from provinces other than their current location.
Subjects between the ages of 6 to 10 years may be suitable for the MRIT2* procedure without sedation.This procedure is mainly recommended for individuals over 8 years of age [18,19].The MRIT2* examination is suggested at the age of 6, while LGE (late gadolinium enhancement) is recommended at 13 years of age [20].In this study, patients who had ferritin levels > 1000 ng/m Fig. 1 SEM model analysis interacting hepcidin release variables, hepcidin, cardiac iron overload, and ST2 were recruited, as this condition often accompanies left ventricular dilatation, tricuspid regurgitation, and pulmonary hypertension [21].
The following variables, Hb, Ret-He, HIF-1, IG, hs-CRP, IL-6, ferritin, and transferrin saturation, were included in the SEM model analysis to determine their interaction with hepcidin release.The mean Hb level was 9.2 ± 1.2 g/ dL, indicating the absence of severe anemia, which can lead to hepcidin deficiency.This explains the lack of a relationship between Hb and hepcidin.Furthermore, severe anemia and depleted hepcidin can lead to the accumulation of iron in hepatic parenchymal tissue [9].
We found 19 subjects, accounting for 24%, with an upper normal limit of reticulocyte.A significant correlation was observed between reticulocytes (r = − 0.43, p = 0.01) and Ret-He with ferritin (r = 0.29, p = 0.01).This is consistent with the increased reticulocyte reflects increased erythropoiesis [22].
A total of 9 subjects (11.25%) had a white blood count of > 10,000/µL and 14.8% had upper normal IL-6 value with no sign of infection.Based on normal clinical results, 10 subjects, accounting for 12.5% had thrombocytopenia, which is caused by hypersplenism, use of DFP, increased platelet destructions, and reduced thrombopoietin.Furthermore, elongated PT and aPTT are more often present in subjects whose hepatic iron deposits are within a moderate-severe degree compared to a normal-mild degree [23].
The increase of AST, ALT, SI, and TIBC is consistent with the results of other studies.Furthermore, a correlation was observed between AST and ferritin (r = 0.27, p = 0.02), and the increase in AST and ALT have significantly higher ferritin than subjects with normal AST and ALT (p = 0.01) [24].
The level of hepcidin is not significantly different from the result of a study by Jagadishkumar using Cloud Clown Corp reagent [25].In this study, a deficiency of hepcidin was observed in 37.5% of the subjects.Additionally, the hepcidin-ferritin ratio indicated a deficiency of hepcidin in all subjects.This result suggests that the hepcidinferritin ratio was unable to increase alongside ferritin.A Low hepcidin-to-ferritin ratio indicates post-transfusion partial erythropoiesis correction [26].
The homeostasis of hepcidin remains unclear, and several studies have shown contradicting results [7].The expression of hepcidin is intricate and involves the coordination of interacting release variables [7,8].A previous study showed that subjects with transfusion-related iron deposits had a varying hepcidin increase [15].Lower hepcidin level on thalassemia variants has a characteristic of decreased Ret-He [27].In animals, hepcidin regulation occurs due to a single stimulation, but it is complex with unknown precipitating factors in humans.Furthermore, no correlation was found between ferritin and MRIT2* (r = − 0.21 p = 0.06), which is consistent with the report of another study [28].
In this study, hepcidin was found to be negatively correlated with hs-CRP, which contradicts previous studies [28,29].Hepcidin increased in normal subjects, due to infection, inflammation, and increased liver iron.Parischa et al. [28] reported no correlation between Hepcidin and CRP, while Ganz et al. [29] suggested increased hepcidin levels in subjects whose CRP level is > 10 mg/ dL.Iron accumulation increases ROS and ROS, which is associated with inflammation and is primarily caused by ROS [30].The decline of hepcidin resulted in increased ROS, triggering an increase in IL-6.
One of the variables that have a direct role in MRIT2* is Ret-He.Higher reticulocyte represents higher erythropoiesis activities, reducing hepcidin and affecting myocardial iron deposits.Conversely, hb is a variable that is directly correlated with ST2.A previous study showed that myocardial injury is caused by abnormal perfusion, genetic disorder, and an increase in volume and pressure [31].
Neither hs-CRP nor IL-6 plays a direct role in hepcidin, but they have a combined effect on hepcidin levels.A plausible explanation would consider hepcidin as an acute-phase reactant, where infection/inflammation stimulates the formation of IL-6 [32].In return, IL-6 strongly correlated with hs-CRP, but these two variables do not play a direct role in MRIT2*.This is because the myocardial iron deposit received much influence from NTBI due to iron deposition overload rather than infection or inflammation.
Currently, there is a lack of published studies exploring the involvement of hs-CRP and IL-6 in myocardial fibrosis in thalassemia.Some of the factors affecting myocardial fibrosis are cytokines, such as IL-6, IL-1β, TNFα, MCP-1, TGF-β, angiotensin II, angiotensin II type I, and ECM [34].
The combined effect, which plays a role in MRIT2* is Ret-He and transferrin.These results collaborate with the theory stating that periodic transfusion in anemia increases ferritin, transferrin, NTBI, iron deposit, and cardiac iron overload.In previous studies, there was no correlation between MRIT2* and ST2, and this result is consistent with other studies showing no significant relationship between fibrosis with LGE nor MRIT2* [35,36].
The variables with the strongest association with ST2 are hs-CRP, transferrin saturation, and Hb.Increasing ST2 is associated with myocardial fibrosis.ST2 is an IL-1 and IL-33 member, influenced by inflammation.CRP was positively correlated with an increase in ST2 and severe cardiac fibrosis [37].Increased levels of IL-6 and CRP are associated with myocardial fibrosis [33].Furthermore, the combined effects that played a role in hepcidin were hs-CRP and IL-6.This is consistent with the report of previous studies that there was an increase in hs-CRP, IL-6, IL-8, and TNFα in thalassemia major.Increased hepcidin decreases the production of ROS, reducing inflammation [38].

Conclusion
In the CFA path analysis, several variables, including Hb, Ret-He, HIF-1, hs-CRP, IL-6, ferritin, and transferrin saturation were examined in relation to hepcidin, cardiac iron overload, and fibrosis.The result indicated that hepcidin does not have a direct role in cardiac iron overload and fibrosis.The proportion of total variation explained by the interacting release variables on hepcidin was found to be 11%.Furthermore, hepcidin amounted to 20% of the total variation in myocardial iron overload, while myocardial iron overload explain 37% of the total variation in myocardial fibrosis.

Table 2
Laboratory results, MRIT2*, TDI, and percentage of abnormal values SD standard deviation, median (min-max) * percentage of below normal values *** percentage of exceeds normal values

Table 3
Correlation of interacting variables of hepcidin release, MRIT2*, and ST2 t independent samples t test, χ 2 chi-square test for independence * Significant p value < 0.05