Introduction

Friedreich’s ataxia (FRDA) is a neurodegenerative disorder which represents the most common inherited ataxia in the Caucasian race, with an estimated prevalence of 1:50,000 individuals. FDRA is inherited in an autosomal recessive pattern and its mayor clinical manifestations are neurologic dysfunction, cardiomyopathy and diabetes mellitus [1–3]. The most representative clinical feature is a progressive mixed sensory and cerebellar ataxia of all four limbs and trunk that manifests as gait instability. Cerebellar dysarthria and nystamus are also commonly present. Other neurological signs that can be found are optic atrophy and neurosensorial deafness [3, 4]. Heart involvement is present in 63% of patients with FDRA [3]. The most common echocardiographic abnormality is hypertrophic cardiomyopathy which can be concentric, asymmetrical or just a thickening of the papillary muscles, although the range of abnormalities appears to be wide [5]. FDRA cardiomyopathy is often asymptomatic but at the end stages of the disease, ventricular arrhythmia is the most frequent cause of death, together with bulbar dysfunction [3].

The hypothesis that iron-mediated oxidative damage has a main pathophysiological role in FDRA has led to the use of antioxidant molecules as a possible treatment. Therapies such as idebenone have shown, in several preclinical and phase two studies, promising results, particularly in FDRA cardiomyopathy [6–16]. Moreover, studies performed in tissue culture models of FDRA with different iron chelators [17–19] as well as clinical studies in patients with FDRA have shown positive results [7]. The first attempt to remove an excess of mitochondrial iron in patients with FDRA was published by Boaddert et al. in 2007 [7]; the iron chelator employed was deferiprone (3-hydroxy-1, 2-dimethylpyridin-4-one (DFP)). DFP was chosen from among other molecules because it is able to cross the blood-brain barrier and cellular and subcellular membranes, it has the ability to enter cell organelles, it chelates the accumulated iron pools, and it transfers chelated iron to plasmatic transferrin [7, 20–26]. Doses of DFP of 20 and 30 mg/kg/day showed no differences of tolerance and efficacy but the recommended 80 mg/kg/day dose employed in thalassemic patients produced serious adverse reactions. The treatment was continued for 6 months in nine FRDA patients, and the results showed a significant improvement in the neurological signs and symptoms, especially in the younger patients. These results were in consonance with the measures of iron concentrations in the dentate nuclei made by RMI techniques that demonstrated a reduction in the iron deposits at this level. Agranulocytosis is considered the most serious side effect of deferiprone with an incidence of 0.4–2% patients per year but it’s reversible after discontinuation of the drug, thought some patients may require treatment with granulocyte-colony-stimulating factor [27, 28]. Others side effects frequently associated with deferiprone are nausea, vomiting, arthralgia, arthritis, zinc deficiency and fluctuating liver functions test [29]. Side effects other than neutropenia rarely require discontinuation of therapy and they usually appear during within the first year. Although this data concerns patients with thalasemia, which are treated with higher doses of deferiprone (75 mg/kg/day or more), important attention was focused on the possible side effects of DFP, specially in haematological and iron metabolism parameters.

The objective of the present prospective open-label single-arm study was to evaluate the effects of the combined therapy of idebenone (20 mg/kg/day) and low oral doses of DFP (20 mg/kg/day) on the neurological signs and cardiac function parameters and to assess, using MRI, the effects of DFP on the brain iron deposit in a group of FRDA patients.

Materials and Methods

Patients

We prospectively studied 20 patients (age range, 8–25 years; average, 16.5 years). They were eight males (age range, 10–22 years and average, 15.6 years) and 12 females (age range, 9–26 years and average, 13.0 years) with genetic diagnosis of FDRA (range of GAA expansion size, 810–1,800; mean = 1,305). Fifteen patients (75%) had previous echocardiographic signs of mild cardiomyopathy. All patients were under idebenone treatment (20 mg/kg/day) before the inclusion in the study. No consent was obtained from families to discontinue idebenone treatment and we could not asses a population of FDRA patients without idebenone treatment.

DFP was given orally (20 mg/kg/day/12 h) over a period of 11 months. Therapy was provided every 3 months by the hospital pharmacy and final drug administration relayed on parents, care-givers or patients. One patient was excluded from the study after 6 months of DFP therapy because of severe neutropenia. Exclusion criteria included iron deficiency, defined as ferritin and haemoglobin levels below the reference range for age and sex, abnormal serum transaminases greater than two times the upper limit of reference ranges in two consecutive analyses, serum creatinine outside the normal range, and history or evidence of neutropenia. Young women of childbearing age were required to produce a negative pregnancy test before inclusion in the study and had also to agree to use effective methods of contraception. Clinical DFP side effects such as vomiting, abdominal pain, diarrhoea and arthropathy were controlled as well. Written informed consent was obtained before enrolment from parents and adult patients. The study protocol was made with the approval of the Sant Joan de Déu Hospital ethics committee and the Spanish National Medication Agency of the National Health Service.

Clinical Evaluation

Neurological evolution of our FRDA patients was evaluated using the ICARS [30]: posture and gait (0–34 points), kinetic functions (0–52 points), dysarthria-speech (0–8 points) and oculomotor movement disorders (0–6 points). Higher scores indicate a more severe disease (maximum score = 100 points). Neurological evaluations, score calculations and video recording were always performed by the same investigator every 6 months.

Cardiological evaluation included a clinical examination, standard 12-lead electrocardiogram and echocardiography. Echocardiographic measurements of cardiac chambers and systolic and diastolic functions were studied using echocardiography before and after the treatment. Left ventricular end-diastolic and end-systolic diameters, fractional shortening (FS), ejection fraction (EF), septum (IVS) and posterior wall (PW) thickness and left ventricular mass index (LVMI) were measured using two-dimensional M-mode imaging according to the criteria of the Society of American Echocardiography. Mitral valve-pulsed Doppler flow (E´ wave, A´ wave and isovolumetric relaxation time (IVRT)) and Doppler tissue imagine of the mitral (Em wave, Am wave, Sm wave) and tricuspid annulus (Et wave, At wave, St wave) were taken from apical four-chamber views. All the measures were performed by the same cardiologist and echocardiographic equipment (Philips iE33).

Magnetic Resonance Imaging Measurements

Iron clusters tend to be paramagnetic and therefore cause local inhomogeneities in the magnetic field. These inhomogeneities lead to an increase in the magnetic resonance signal decay time (T2*) which is an index used to quantify magnetic resonance imaging (MRI) signal. If a homogeneous external magnetic field is assumed, then the changes in T2* mainly reflect variations in the local iron concentration in the absence of blood vessels or haemorrhages [7, 31–33]. Brain MRI examinations were performed on a 1.5 Tesla MRI instrument (SignaHD, General Electric, Milwaukee, WI, USA). A voxel encompassing the left and right nuclei (dimension 9 mm2) was positioned in the largest section of the dentate nuclei. For each selected section, iron monitoring was performed using a single-slice multigradient echo sequence. Volume shimming and single-slice multigradient echo was repeated at the level of the putamen and thalamus nuclei. A high-resolution image of local T2* values was built and the mean value of T2* was calculated in various regions of interest (ROIs) by the same radiologist. As internal control, several circular ROIs of 11 m2 were drawn in the white matter of the brain hemispheres where iron concentration is assumed to be low.

Biochemical Analysis

Patients were biochemically evaluated every 6 months (blood cell count, ions, glucose, insulin, C-peptide, hepatic and renal function, lipid and iron metabolism, vitamins A and E, lactic acid and idebenone concentrations). Blood samples were taken in baseline conditions and after idebenone and DFP treatment. All samples were collected in the fasting state and were measured as previously reported [25, 34]. DFP-possible chelator effect on other metals (Se, Mn, Co, Cu, Zn and Mo) was monitored every 6 months using the inductively coupled plasma mass spectrometry (ICP-MS-Agilent Technologies). Basic periodic biochemical controls were made of iron, hepatic and renal metabolism every 3 months. During the first 2 months, blood cell count was done every 10 days in order to closely monitor the risk of neutropenia or agranulocytosis.

Statistical Analysis

Wilcoxon’s test was applied to compare the paired data (ICARS scores, MRI and echocardiographic data before and after the start of the therapy). Spearman’s test was applied to search for correlation between the different variables of the study. Mann–Whitney U test was applied to compare values between the different groups. Statistical significance was considered as p < 0.05. Calculations were performed with the SPSS 17.0 programme.

Results

Neurological Parameters (ICARS Score)

No significant differences were observed in total ICARS scores when comparing baseline status and the end of the study in the whole group of patients. In nine patients, the total ICARS score increased while in ten cases the score improved. When we analysed posture and gait scores, they increased significantly after 11 months of therapy (Wilcoxon’s test, p = 0.04; in five patients improved, in ten worsened and in four remained unchanged). Other kinetic function improved significantly (Wilcoxon’s test, p = 0.015; in 13 patients improved, in four worsened and in two remained stable) (Table 1, Fig. 1). These results were similar to those for patients with no wheelchair dependence (n = 14), showing an improvement in kinetic functions (Wilcoxon’s test, p = 0.045) and a worsening in posture and gait scores (Wilcoxon’s test, p = 0.045).

Table 1 Individual results of ICARS scores in all the FDRA patients under baseline conditions and after 11 months of idebenone and DFP therapy
Full size table
Fig. 1
figure 1

ICARS scale values in all the FDRA patients under baseline conditions and after 11 months of idebenone and DFP therapy. The boxes represent the mean values with the vertical lines representing the standard error of the mean. **p value under 0.05

Full size image
Fig. 2
figure 2

Echocardiographic values of IVS and LVMI in all the FDRA patients under baseline conditions and after 11 months of combined idebenone and DFP therapy. The boxes represent the mean values with the vertical lines representing the standard error of the mean. **p value under 0.05

Full size image

Echocardiografic Parameters

Echocardiogaphic data showed a significant reduction of the IVS thickness (Wilcoxon’s test, p = 0.04; in 11 patients decreased, in four increased and in four remained stable) and in LVMI (Wilcoxon’s test, p = 0.038; in 15 patients decreased and in four increased) when comparing baseline conditions and the end of the study (Table 2, Fig. 2). No other significant differences were observed in other echocardiografic parameters of heart hypertrophy, systolic (FS and EF) and diastolic functions (IVRT and E´/Em index) when comparing baseline conditions and the end of the study.

Table 2 Results of echocardiographic data in all the FDRA patients, in patients with FDRA cardiomyopathy, and patients without cardiomyopathy under baseline conditions and after 11 months of combined idebenone and DFP therapy
Full size table

In the group of patients diagnosed with cardiomyopathy (n = 15) statistical analyses showed similar results with a significant reduction of the IVS thickness (Wilcoxon’s test, p = 0.048) and a significant reduction of the LVMI (Wilcoxon’s test, p = 0.02).

Brain Area Iron Levels Assessed by RMI Tools

The MRI T2* values in the dentate nucleus showed a statistically significant reduction (Wilcoxon’s test p = 0.007; in 14 patients decreased). No other brain areas (thalamus and putamen) underwent any significant change in T2* values.

Biochemical and Hematologic Parameters

After the treatment period, a significant decrease in serum ferritin levels was observed in all patients (Wilcoxon’s test, p < 0.0001), but ferritin levels decreased below the normal range for age and sex in only six patients. Significant decreased haemoglobin (Wilcoxon’s test, p = 0.001), hematocrite (Wilcoxon’s test, p = 0.003) and increased transferrin values (Wilcoxon’s text, p < 0.0001) were also noted.

Concerning the biochemical monitoring made from idebenone concentrations, hepatic and renal metabolism and serum concentration of Se, Mn, Co, Cu, Zn and Mo, no statistically significant differences were found between the several points of analyses (data not shown). Plasma reference values for Zn were 628–1,200 μg/L. Zn deficiency was observed in two patients (plasma Zn values of 326 and 452 μg/L) before the inclusion in the study and oral Zn supplementation was started.

Deferiprone Treatment Adverse Effects

The most frequent adverse effects were related to gastrointestinal discomfort such as nausea, vomiting and unspecific abdominal pain (25%). They all were of mild intensity, appearing the first days after the introduction of treatment and disappearing within 2 days. One patient complained of knee arthralgia which was relieved by ibuprofen within 2 days. Two patients presented neutropenia but in only one case it was severe (neutrophils account 450 cells/mm3), and therapy was interrupted definitively after 6 months of treatment. In both patients neutropenia resolved without complications after discontinuation of DFP. In the patient with mild neutropenia, there was no relapse after reintroduction of DFP.

Discussion

FDRA pathogenesis is still unclear. The first attempt to use iron chelators to remove mitochondrial iron pools together with idebenone in patients with FRDA was done by Boaddert et al., with promising results. Moreover, previous studies treating patients with idebenone indicated that antioxidant therapy with low doses (5 and 10 mg/kg/day) reduced cardiac hypertrophy, but definitive improvement in neurological and heart function has not been demonstrated [9–16]. Studies with high doses of idebenone (20 mg/kg/day) indicated a dose-dependent trend to improvement of the ICARS. When data was analyzed excluding patients who required wheelchair assistance, it showed a significant improvement in ICARS, and this improvement in the ICARS was dose dependent [6, 16–26, 35].

Other potential therapeutic approaches focused on different cellular pathophysiology pathways in FDRA are also been developed. Erythropoietin increase frataxin concentration through a posttranscriptional mechanism and is currently being test in pilot clinical trials, though its important mid and long-term side effects may limit its use [36, 37], especially in children. Pioglitazone, an oral antidiabetic drug, stimulates mitochondrial function and antioxidant response by acting on the peroxisome proliferator-activated receptor gamma pathway and it will soon be tested in clinical trials [38]. Another promising group of molecules are the histone deacetylase inhibitors. They have probed to increase frataxin expression in lymphocytes of patients with FDRA in vitro, however, many important questions still remain to be answer, specially about their effects on the transcription process of other genes [39, 40].

Limited data is available on the natural course of FRDA. The progression rate on the ICARS in untreated patients is estimated to be of 4.4–5 points per year [12, 41], but it is important to note that the population studied was older than our group (mean age 31 years), and in younger patients this rate could be even greater. Eleven months of combined therapy with high idebenone and DFP showed a stabilization of the neurological functions as the total ICARS scores remained stable, and this result was independent of patient age. In the overall group, the stabilization of the ICARS score was due to a significant recovery of kinetic functions but with a worsening of gait and posture scores. These data support the subjective reports made by parents and patients of an improvement in fine skills and manipulative dexterity. These findings can be extrapolated to those observed by Boaddert et al. where the patients exhibited improvements in fine skills and manipulative dexterity, although in his case this improvement corresponded with a modest but significant reduction of the ICARS scored in younger patients not seen in our study.

The results in the ICARS score were not influenced by the severity of the neurologic parameters, as we found the same significant results when patients with wheelchair dependence were excluded.

Although the natural history of cardiac lesions in FDRA has scarcely been reported [42], idebenone treatment has demonstrated an initial reduction in cardiac hypertrophy in FRDA patients [6–16]. This reduction is only seen, mainly, in the first years after initiation of therapy with a trend to progress over time. The echocardiographic studies with combined idebenone and DFP therapy demonstrated an improvement in the variables that measure cardiac hypertrophy (IVS and IMV) in patients with cardiomyopathy. Our results showed a decrease in the degree of heart hypertrophy which might be due to the combined therapy of idebenone and DFP. No changes were observed in diastolic or systolic cardiac function parameters. However, none of our patients had cardiac function impairment before the study and in all cases cardiac function remained within the normal reference ranges.

FDRA patients had increased T2* values in dentate nucleus measured by MRI techniques [7, 31]. The analysis of the T2* values after 11 months of idebenone and DFP treatment revealed a significant reduction with no other significant reduction in T2* values in other brain areas. These results indicated that low oral doses of DFP are able to chelate labile iron that is specifically accumulated in dentate nuclei of FDRA patients.

Combined therapy with idebenone and DFP was well tolerated with no important adverse effects, with the exception of one case of severe neutropenia that resolved after discontinuation of the therapy, with no complications. Mild adverse gastrointestinal effects or arthralgia appeared during the first days of treatment and improved in a short period of time. The appearance of neutropenia occurred after several months of treatment and no triggering factor was found. During biochemical follow-up, patients showed a significant steady evolution to a ferropnic state with a decrease in blood iron-associated parameters which may restrict the use of DFP for prolonged periods of time. Moreover, DFP has been recently shown experimentally to be cytotoxic and inhibitory to cell aconitase activity when used for extended periods of time at concentrations that are comparable to the doses used in this study [43]. However, DFP in vitro inhibition of aconitase activity seems to recover short after cells exposure to a free DFP medium [25]. This issue raises the question of whether DFP should be administrated in alternating periods, considering moreover that Zn, an essential cofactor for Superoxide dismutase activity, might also be decreased in some cases under DFP therapy. In fact, none of our cases showed decreased Zn values during DFP therapy. In two patients, Zn supplementation was required before DFP was started, but Zn values remained within the normal range during DFP administration in all patients. The other metals studied remained within our reference values after DFP therapy.

In conclusion, combined therapy with idebenone and DFP at low doses in patients with FDRA indicates a stabilizing effect in neurologic dysfunctions due to an improvement in the kinetic functions with no effect on gait and posture. Heart hypertrophy parameters and iron deposits in dentate nucleus improved significantly. Combined therapy was well tolerated with mild side effects, excluding the risk of neutropenia which should be closely controlled. Progressive reduction of plasma iron parameters may limit long-term DFP therapy in patients with FDRA, and the question of which is the best DFP administration protocol still needs to be resolved.