Beta-thalassemia: renal complications and mechanisms: a narrative review.

ABSTRACT Objectives: Beta-thalassemias are a group of recessively autosomal inherited disorders of hemoglobin synthesis, which, due to mutations of the beta-globin gene, lead to various degrees of defective beta-chain production, an imbalance in alpha/beta-globin chain synthesis, ineffective erythropoiesis, and anemia. Improved survival in thalassemic patients has led to the emergence of previously unrecognized complications, such as renal disease. Methods: A comprehensive literature review through PubMed was undertaken to summarize the published evidence on the epidemiology and pathophysiology of renal disease in thalassemia. Literature sources published in English since 1990 were searched, using the terms beta-thalassemia, renal disease. Results: Renal disease is considered to be the 4th cause of morbidity among patients with transfusion dependent thalassemia. Chronic anemia, hypoxia and iron overload are the main mechanisms implicated in development of renal injury, whereas several studies also suggested a contributive role of iron chelators. Discussion and Conclusion: Kidney disease may develop through progressive renal tubular and glomerular damage; thus, its early recognition is important in order to prevent and/or reverse deterioration. This review will provide an insight on the involved mechanisms implicated in kidney disease in thalassemic patients and will discuss the updates on diagnosis and prevention of renal complications in thalassemia.


Introduction
Beta-thalassemia syndromes are the most common inherited monogenic disorders worldwide. They are heterogeneous disorders caused by reduced or absent beta-globin synthesis, a major component of adult hemoglobin A (HbA, α2β2), leading to an imbalance of the globin chains [1]. Consequently, beta-thalassemia leads to reduced hemoglobin production and accumulation of α-globins which form insoluble hemichromes [2]. The former causes microcytic, hypochromic anemia; the latter is associated with oxidative stress, ineffective erythropoiesis and hemolytic anemia [3,4]. The clinical and hematological spectrum of betathalassemia disease ranges from mild to clinically overt conditions, including transfusion dependent (TDT) beta-thalassemia major (TM) and non-transfusion dependent (NTDT) beta-thalassemia intermedia (TI) or thalassemia minor (TMin) [5].
Thalassemia syndromes are prevalent in the Mediterranean, Africa and Southeast Asia [6]. Although they present a significant public health concern for the developing countries, population movements from countries where thalassemia is prevalent have also led to an increased number of affected patients in the developed countries [7]. At least 20.7% of the world population carry a beta-thalassemia variant, whereas annually approximately 40000 children are born with beta-thalassemia [8].
In recent years, mainly in developed countries, the prognosis of beta-thalassemia has improved due to blood transfusion, iron chelation therapy and the advances in the knowledge of the disease. In addition, the numbers of newborns with beta-thalassemia have been limited in Western societies due to the introduction of screening programs.
Despite the improved survival noted for beta-thalassemia, many patients experience complications in several systems, including cardiopulmonary disorders, endocrine organ diseases, liver impairment and thromboses in different vascular beds [9]. Furthermore, the improved patient survival has allowed previously unrecognized renal complications to emerge. The effect of thalassemia on the kidney has not been extensively evaluated. The aim of this narrative review is to summarize information regarding the renal complications and their pathophysiologic mechanisms in beta-thalassemia patients. Literature sources published in English since 1990 were searched through PubMed, using the terms beta-thalassemia and renal disease or manifestations.

Epidemiology
Knowledge on the epidemiology of renal complications in beta-thalassemia is limited. Cross-sectional studies in various thalassemia groups from five thalassemia centers in North America have shown reduced creatinine clearance in 7.8%, and albuminuria in up to 59% of patients [10]. More recent studies, found renal dysfunction in 1.8% of TDT patients [11], whereas renal problems were classified as the fourth most common cause of morbidity (4%) after endocrine (44.7%), cardiovascular (41.3%) and hepatic (40.5%) disease in the same patient population [12].
However, there is currently no data from large observational cohorts on the prevalence of chronic kidney disease (CKD) following contemporary definitions in patients with beta-thalassemia.

Renal manifestations of beta-thalassemia
The study of renal function abnormalities in thalassemia is now timely, because of the increasing use of deferasirox and the patients' improved survival. Several authors have reported abnormalities of renal tubular function in patients with thalassemia major and intermedia and others have suggested that renal hyperfiltration is common in patients with thalassemia.
The etiology, the pathogenetic mechanisms involved and the possible evaluation, including emerging biomarkers are presented in Table 1. The most common pathophysiologic and clinical manifestations of renal disease in beta-thalassemia patients are:

Tubular dysfunction
Evidence of tubular dysfunction among patients with beta-thalassemia was initially described about 2 decades ago by Ong-ajyooth et al [13]. Since then, several studies confirmed and evaluated further the  na Plasma NGAL with PCR and ELISA -In all adult cases, except one sample, NGAL protein was expressed more compared to the controls -Positive correlation with ferritin -Negative correlation with sex, age -NGAL upregulation was not found in pediatric beta-thalassemia patients. Iron overload and oxidative status in beta-thalassemia patients induce NGAL/Lcn2 expression. Upregulation of NGAL in this disorder may play a beneficial role in decreasing ROS or chelating iron. Obviously, chelating of iron is one of the major therapeutic goals in b-thalassemia. tubular function in beta-thalassemia (Table 2). Up to 60% of patients with TDT have been reported to develop signs of tubular dysfunction [14]. Common signs of tubulopathy, such as proteinuria (8.6%), hypercalciuria (12.9%), phosphaturia (9.2%), hyperuricosuria (38%), magnesiumuria (8.6%), and increased excretion of β2-microglobin (β2Μ) (13.5%) were demonstrated in a TDT Iranian population [15]. Early recognition of renal dysfunction is of great importance in terms of intervening in a timely manner and improving prognosis. Therefore, in the recent years, several molecules produced and released by proximal tubular cells as measurable proteins, have been tested as possible valid biomarkers of renal injury. Such proteins include N-Acetyl-beta-D-glucosaminidase (NAG), neutrophil gelatinase-associated lipocalin (NGAL), Kidney Injury Molecule-1 (KIM-1), Liver-type Fatty Acid-Binding Protein (L-FABP) and Interleukin-8 (IL-8), all of which were studied mainly in children and younger patients with TM [16][17][18]. However, further research is needed to evaluate the significance of these biomarkers as predictors of renal disease in thalassemia patients.
Tubular dysfunction among patients with beta-thalassemia has been related to iron overload, chronic anemia, as well as desferioxamine (DFO) toxicity [19][20][21][22]. The mechanism of iron overload-associated renal damage has not been fully elucidated. Non-transferrin iron can lead to organelle membrane dysfunction and subsequent cell injury/death. Iron-catalyzed generation of reactive oxygen species (ROS) is responsible for initiating the peroxidative reaction. The possible association of oxidative stress and its impact on nitric oxide (NO) metabolism in iron overload associated renal injury was investigated through a randomized study on rats. In this study, animals were assigned into Feloaded (given 500 mg elemental iron/kg body weight as iron dextran; IV), Fe-depleted (given an iron-free diet for 20 weeks), and control groups. Chronic iron deposition in proximal tubules, glomeruli and interstitium was associated with significant glomerulosclerosis, tubular atrophy, and interstitial fibrosis [23]. Iron staining in the distal tubules was negligible, but iron overload resulted in iron deposition in the glomeruli and proximal tubules with various renal lesions and evidence of increased ROS activity, enhanced ROSmediated inactivation and sequestration of NO. Iron depletion was associated with reduced malondialdehyde (MDA) and tissue nitrotyrosine abundance, increased urinary NO metabolite (NOx) excretion, normal NO synthase (NOS) expression and absence of renal injury. The iron overload-induced toxicity is supported by clinical studies demonstrating concordance between serum ferritin levels and markers of tubular injury. At the same time, chelation therapy was associated with a reversal of toxicity. An initial study from Greece including 36 patients aged 5-22 with TM showed a significantly higher urinary excretion of NAG in patients with poor compliance to DFO and serum ferritin >3000 ng/mL compared to regularly chelated patients with serum ferritin <3000 ng/mL (2.86 ± 0.287 vs. 2.37 ± 0.32 nmol/mg creatinine/h, p < .05) [24]. Similar results were observed in a second Greek study with 91 TM patients, aged 5-65 and on DFO chelation. However, urine markers did not correlate with age, indicating that DFO treatment was not related to renal dysfunction [25].
Kidney biopsy would offer many information about iron deposits but it is an invasive method with many risks and significant patient discomfort. Nowadays, kidney magnetic resonance imaging (MRI) is essential for noninvasive evaluation of renal iron overload. Eighty-three patients suffered from TM (mean age 24.7 years) and 37 patients with TI (mean age 28.8 years) were enrolled in an interesting study performed in Tehran, Iran. The mean transfusion duration was 21.8 days, while 57.5% of patients were splenectomized. All patients received iron chelation therapy (mean therapy duration 21.8 years) with variable compliance. The results showed statistically significant but moderate correlation between serum ferritin and kidney T2* relaxation (correlation coefficient = −0.446, p < .001). The non-significant correlation of liver or heart MRI T2* relaxation time with serum ferritin highlights the important role of kidney MRI on iron load assessment [26].
Anemia and potential associated chronic hypoxia could also lead to activation of the oxidative stress, lipid peroxidation and tubular dysfunction [26,27]. Moreover, hypoxia per se induces differentiation of tubular cells to myofibroblasts [28], while at the same time, activates fibroblasts, changes the extracellular matrix metabolism of renal cells and leads to the obliteration of peritubular capillaries [29][30][31], hypertransfusion program and DFO treatment, and moderate anemia (Hct > 25%). Patients with severe anemia had significantly higher low-molecular weight proteinuria and malondialdehyde (MDA, an indicator of lipid peroxidation), and lower urine osmolarity than those with moderate anemia. The abnormalities were the least severe in patients on hypertransfusion and DFO therapy [22].
Interestingly, improved renal function was illustrated in patients with TDT after curative hematopoietic stem-cell transplantation (HSCT). When comparing 29 children with TM who had undergone HSCT to 39 children of the same age and similar disease severity but who had not experienced HSCT, parameters of tubule function were better in patients that had undergone HSCT, as demonstrated by urine protein level (0.36 mg/mg creatinine vs 3.03 mg/mg creatinine, p < .001), osmolality (712 mosmol/kg vs 573 mosmol/kg, p = .006), NAG (17.7 U/g creatinine vs 42.9 U/g creatinine, p = .045), and β2M (0.09 microg/mg creatinine vs 0.13 microg/mg creatinine, p = .029) [33].

Glomerular dysfunction
Changes in glomerular function are also evident in multitransfused patients with beta-thalassemia. Chronic anemia is thought to reduce systemic vascular resistance leading to hyperdynamic circulation and subsequent increased renal plasma flow and glomerular filtration rate (GFR) [34]. Glomerular hyperfiltration appears to be deleterious to the mesangial compartment, causing it to increase matrix volume and cellularity, initiating a sclerotic process [35,36]. In the longterm, such modifications may theoretically lead to a progressive decline in GFR through the typical pathway of hyperfiltrationalbuminuriaprogressive renal damage. Furthermore, renal tubular cells that are subject to iron overload, induce injury into the interstitium by releasing cytokines and growth factors that may cause tubulointerstitial scarring and glomerular sclerosis [37].
As it has already mentioned, when iron exceeds the binding capacity of transferrin, as it occurs in iron overload, it accumulates as a non-transferrin-bound iron which accelerates the generation of radical oxygen species and cellular damage.
Additional contributing independent factors to renal injury may be glomerulonephritis induced by HIV, hepatitis B and hepatitis C infections, as well as noniron related liver and heart disease [38].
In studies performed in children, adolescents [10,39] and adults with TDT receiving iron chelation therapy [40], calculation of GFR based on serum creatinine concentration or creatinine clearance demonstrated glomerular hyperfiltration. It is worth mentioning that Milo et al demonstrated reduced GFR in 9 TDT patients through the use of inulin clearance [41]. Individuals were on chelation therapy with DFX for at least 1 year. All patients had lower inulin clearance and GFR than normal healthy young adults, while 4 out of 9 patients had mild to moderate chronic kidney disease (CKD). By using Cockcroft-Gault (CG) equation and serum creatinine, median creatinine clearance in the same study was 134.9 (mean 133.3) mL/min per 1.73 m 2 , significantly higher than the GFR estimated by inulin clearance (p < .008) [41].
Economou et al, who used cystatin C (Cys C) in 42 pediatric and adult patients with TM, also described a reduction in GFR compared to normal controls (p ≤ .001). On the contrary, compared to normal levels, elevated Cys C levels, were also significantly accompanied by other indexes of glomerular dysfunction, such as proteinuria (223.28 ± 125.84 and 79.34 ± 29.1 mg/m 2 /day, respectively, p < .0001). All patients of the study were chronically transfused and under a regular chelation program, which consisted of either DFX or combination of DFP and DFO [42]. Similarly, elevated Cys C levels were also visualized in TM patients not receiving chelation therapy [43]. Another study found a very weak inverse linear correlation between serum ferritin and Cys C eGFR, which, however, was not observed when concomitant use of chelation therapy was considered [44]. Prospective and larger studies are needed to validate these findings.
Quinn et al applied 24-h urine collection and reported similar levels of creatinine clearance (mean 136.1 mL/min per 1.73 m 2 ) in 106 adult patients with TM, the majority of whom were treated with DFP or DFO. When they assessed GFR using this method, they predicted hyperfiltration [10]. On the other hand, Piga et al reported normal GFR measured by plasma sampling of 51Cr EDTA in patients with TM [45].
Taking everything into consideration, the estimations regarding GFR are frequently varying and inconsistent. Most of the times these variations apparently derive from the different parameters applied in each study. Patient demographics, clinical characteristics, comorbidities and iron chelation therapy, in addition to the study design and data analysis, are important factors that might influence the outcome of each study.
Milo et colleagues propose inulin clearance as a more accurate method to the early detection of reduced GFR and deceleration of the progressive deterioration toward CKD. According to the authors, it is important to note that the same study has some limitations such as the fact that it involves a relatively small number of participants. Additionally, it is a cross-sectional study and all participants were adults (median age was 30, range 21-37 years) [41].
Reduced estimated GFR based on serum creatinine (MDRD study simplified equation), has also been demonstrated in a cohort of TM adult patients regularly followed-up for 10 years [46]. As in the previous study, all patients were transfusion-dependent, treated with iron chelation, and they were adults at the time of the enrollment. Thus, it is unclear whether hyperfiltration preceded the demonstrated reduced GFR or it never existed. This theory could be supported by similar studies conducted on children with TM, but to our knowledge there is no current reference presenting such data.
Other factors that favor glomerular hyperfiltration are the male gender and history of splenectomy, according to data derived from univariate and multivariate analysis among adult, multitransfused patients with TM and TI [47].
The variability of GFR estimations might, indeed, be related to the method used for GFR calculation. Equations based on serum creatinine (e.g. CKD-EPI, MDRD) overestimate GFR due to the low serum creatinine levels observed in patients with TM, which might be related to their muscular atrophy and reduced muscle mass [48,49], and, therefore, decreased generation. On the contrary, urinary creatinine excretion in healthy people is expected to equal creatinine generation [50], most of which is generated by muscle metabolism [50]. Other possible factors contributing to low serum creatinine in thalassemia patients are compensated by increased tubular creatinine secretion which prevents its appropriate rise in serum [51], or extrarenal creatinine metabolism, especially when kidney function is reduced [52]. Lastly, variability in serum creatine measurements among individuals may additionally contribute to overestimation of GFR [50].
Calculation of GFR based on creatinine clearance (e.g. Cockcroft-Gault Equation) might also result in overestimation of GFR, which may be encountered as hyperfiltration. In this case, increased creatinine clearance is, largely, due to the proximal tubular secretion of creatinine. Urinary creatinine excretion consists of filtered creatinine and proximal tubular secretion of creatinine. As glomerular disease progresses, the remnant tubular cells hyper secrete creatinine, resulting in increased creatinine excretion and, thus, elevated creatine clearance [51].
Measurement of the area under the plasma clearance curve (AUC,) following a single intravenous injection of 51Cr-EDTA, is another procedure for determining GFR. Similarly, to the aforementioned methods, it has been proven that measurement of GFR by 51Cr-EDTA overestimates the true renal clearance by approximately 10% [53].
Cys C is a low molecular weight proteinase inhibitor that is produced in all nucleated cells and maintains a constant appearance rate in blood. Opposed to serum creatinine, serum Cys C concentration is independent of gender and muscle mass [54]. Cys C is freely filtered through renal glomeruli and then totally reabsorbed and catabolized in the proximal renal tubule and does not return to the blood. Therefore, serum concentration is mainly determined by GFR [55]. Previous studies have demonstrated the superiority of serum Cys C compared to Cr in the evaluation of GFR, especially when there is a minor reduction in GFR [56,57].
Inulin clearance is the most widely accepted method for estimating GFR [58]. However, this is a method that, due to its complexity, cannot be consistently used in clinical practice.
Considering the above it is rather obvious that the use of an accurate method to evaluate the GFR method may lead to early detection of reduced GFR and delay the progressive deterioration toward chronic kidney disease.

Hematuria
An Iranian study involving a comparative evaluation of 108 children with beta-thalassemia TDT or NTDT demonstrated microscopic hematuria in 19 patients. Interestingly, hematuria was more common in NTDT than TDT (90% vs. 10%) [59]. A subsequent study, revealed the presence of microscopic hematuria in 10.6% from a total of 500 patients with TDT, while, at the same time, its frequency was significantly increased in ages over 20 years (9.8% vs. 20%, p = .04) [60]. The presence of hematuria might be related to either hypercalciuria or hyperuricosuria and nephrolithiasis [59].

Nephrolithiasis
The prevalence of nephrolithiasis in TDT population might reach 59%, as demonstrated in a study involving 27 subjects with TM by non-iodinated contrast renal tract CT. Among the affected subjects, 69% had multiple stones, whereas 56% of the affected patients had stones of variable composition. The majority of renal calculi were struvite (33%), followed by calcium oxalate (31%) and cystine (22%). The development of struvite stones may warrant an evaluation of an underlying infection [61].
Through a retrospective study of 166 participants with TDT, Wong et al showed that nephrolithiasis is associated with reduced bone mineral density of the femoral neck (odds ratio (OR) = 5.59, 95% CI 1. 16-27.03) and increased risk of fracture in males (OR = 2.13) [62].
It is worth mentioning that stone formers in this study had elevated creatinine and lower ferritin serum levels suggesting that DFX is involved as a causative factor [62]. The impact of hypercalciuria after vitamin D and calcium supplementation on renal stone development in the general population remains controversial [63,64]. Similarly, calcium or vitamin D supplementation was not found to be a risk factor for kidney stones in individuals with thalassemia [61]. As we already know, hyperuricosuria due to tubular dysfunction and increased cell turnover has been associated with uric acid urolithiasis in the general population. However, in TDT patients, uric acid stones represented only 2% of all stones, which might be reflected by the increased frequency of blood transfusion, which limits cell turnover [61]. Splenectomy was proved to be an independent risk factor for the development of hyperuricemia and nephrolithiasis in patients with TI, by further increasing erythrocyte turnover and number. Despite xanthine oxidase inhibition, patients receiving allopurinol had both significantly higher level of serum urate and higher prevalence of urolithiasis, compared to those not receiving it (p < .01) [65].

The role of iron chelating agents in renal disease
The advent of iron chelators improved patient survival in beta-thalassemia. However, this achievement has allowed previously unrecognized complications to emerge.
Currently, three iron chelators are available: oral agents deferasirox (DFX), deferiprone (DFP), as well as parenteral deferoxamine mesylate (DFO). Renal manifestations attributed to chelating agents are rare. Although serum ferritin levels have been inversely correlated with GFR [66], iron chelation therapy has also been related to renal function deterioration [67]. Oral agent DFP lacks support of large randomized controlled trials. About two decades ago, Hoffbrand et al studied the efficacy and safety of DFP in 51 transfusion dependent iron overloaded patients.
They reported that only 1 patient discontinued the drug after 4 months of treatment because of renal failure. However, that patient was not thalassemic [68]. El Alfy et al demonstrated a modest non-significant and non-progressive elevation in serum creatinine from a baseline mean of 29.2 ± 12-37.1 ± 10 micromol/ L over a 6-month period, in a study of 100 children under 10 years old (mean 5.1 years) with TDT (91 patients with TM), to whom were prescribed 50-100 mg/kg DFP. Two patients had baseline serum creatinine values of 71 and 62 micromol/L, which resolved by the end of the study, while two others had a single but transient episode of elevation of serum creatinine above the upper limit of normal, which also resolved by the end of the study, without a need for interruption of therapy or dose adjustment [69]. The etiology for these fluctuations is unclear but has not been reported in other currently available data. Moreover, a case report has been published by Bragadesh et al. suggesting that the DFP may be an alternative iron chelator when renal impairment is obvious [70]. Although DFP is not considered nephrotoxic, it is characterized by a narrow therapeutic window and serious adverse effects, such as agranulocytosis, neutropenia and disabling arthropathy [71]. Thus, it is approved specifically for patients with TDT, when treatment with other chelating agents is contraindicated or inadequate. DFX is well absorbed by the gastrointestinal tract, forms a complex with the plasma iron and is excreted via the hepatobiliary route. It also binds intracellular iron inducing ferritin degradation through proteasome mediated mechanism [72]. DFX is generally well tolerated. However due to its lipophilicity, DFX enters the tubular cells and forms a highly charged complex with iron. This triple negative charged DFX-iron complex does not penetrate membranes easily but accumulates and may result in proximal tubulopathy and Fanconi syndrome (FS) [73]. FS is associated with non-anion gap metabolic acidosis, hypophosphatemia due to phosphaturia, aminoaciduria, proteinuria and glucosuria (with normal serum glucose concentrations). Despite its importance, there is limited data about the risk factors, incidence and clinical picture of FS in patients receiving DFX. A retrospective study conducted by Chuang et al indicated a younger age of treatment initiation as the sole factor that was significantly associated with FS (7.8 vs. 19.2 years, p = .008) [74].
However, this increase was sometimes transient, never exceeded two times the upper limit of normal (ULN) and was generally within the normal range [85]. Later studies on adult and pediatric patients with TDT and normal renal function, confirmed absence of progressive increase in serum creatinine over a 5-year follow up after treatment. An increase over 33% above the value at the start of DFX and greater than the ULN was reported in consecutive serum creatinine level, in 8.8% of the patients, mostly in doses 25-35 mg/kg/day. Nonetheless, this increase was manageable and did not lead to drug interruption [86]. On the other hand, a retrospective study of 72 patients (mean age: 20.3 ± 0.9 yrs; 36 male, 36 female) with thalassemia major or intermedia treated at Sultan Qaboos University Hospital found that the renal side effects related to deferasirox appear to be higher than those reported in published clinical trials [92]. Reassuringly, DFX proved to be safe in managing iron overload in  thalassemia patients with end stage renal disease [93] or diabetes mellitus [94]. DFO chelates iron is located in plasma and ferritin by forming a metabolically inactive complex, which is renally excreted. DFO is poorly absorbed by the oral route. Consequently, intravenous or subcutaneous administration is required [95]. Acute renal failure necessitating dialysis following intravenous DFO overdose was described in patients who received 10times the recommended dose due to administration pump failure [96] or with inadequate monitoring [67]. Similarly to DFX, transient and within the normal range serum creatinine increase has been documented in 14% of patients receiving DFO [85], whereas the same agent was also associated with increased level of NAG and tubular dysfunction in patients with TM [97]. It is important to note that DFO poses a risk of Yersinia, mucormycosis (zygomycosis) and Vibrio vulnificus infection and sepsis [98][99][100], an additional factor that causes acute kidney injury if left untreated.
The mechanisms involved in GFR alteration concurrently with iron depletion include impaired mitochondrial function and consequent production of adenosine and adenosine triphosphate, that leads to the activation of the tubulo-glomerular feedback and vasoconstriction of the afferent glomerular arterioles. Interference with prostaglandin production and imbalance between vasodilating and vasoconstrictive substances is also implicated in GFR reduction [67].
The management of serum creatinine elevation should be individualized based on the magnitude of increase and the presence of additional risk factors for renal disease or comorbid conditions. Recommendations for dose modifications and monitoring in patients who experience DFX-associated adverse events are available [101].
In conclusion, this narrative review highlights the most common pathophysiologic and clinical manifestations of renal disease in beta-thalassemia patients. There is a need for close monitoring and follow up of renal function both in NTDT and TDT patients as their life expectancy has increased and this puts them potentially at increased risk of sever renal disease. More longitudinal data is required to fully portray any possible differences in renal disease in TDT and NTDT patients, as well as the current prevalence of this comorbidity in the era of new iron chelation and blood transfusion guidelines.

Disclosure statement
No potential conflict of interest was reported by the authors.