European Heart Journal – Cardiovascular Imaging (2015) 16, 325–334 doi:10.1093/ehjci/jeu191

Improvement of heart iron with preserved patterns of iron store by CMR-guided chelation therapy Antonella Meloni 1, Vincenzo Positano 1, Giovan Battista Ruffo 2, Anna Spasiano 3, Domenico Giuseppe D’Ascola 4, Angelo Peluso 5, Petra Keilberg 1, Gennaro Restaino 6, Gianluca Valeri 7, Stefania Renne8, Massimo Midiri9, and Alessia Pepe 1* 1 CMR Unit, Fondazione G. Monasterio CNR-Regione Toscana, Area della Ricerca S. Cataldo, Via Moruzzi, 1, Pisa 56124, Italy; 2U.O.C. Ematologia con Talassemia, Ospedale Civico, Palermo, Italy; 3UOSD Centro per le Microcitemie, AORN Cardarelli, Napoli, Italy; 4U.O. Microcitemie, A.O. ‘Bianchi-Melacrino-Morelli’, Reggio Calabria, Italy; 5Microcitemia-Azienda Unita` Sanitaria Locale TA/1, Presidio Ospedaliero Centrale, Taranto, Italy; 6Centro di Ricerca e Formazione ad Alta Tecnologia nelle Scienze Biomediche ‘Giovanni Paolo II’, Universita` Cattolica del Sacro Cuore, Campobasso, Italy; 7Dipartimento di Radiologia, Azienda Ospedaliero-Universitaria Ospedali Riuniti ‘Umberto I-Lancisi-Salesi’, Ancona, Italy; 8Struttura Complessa di Cardioradiologia-UTIC, P.O. ‘Giovanni Paolo II’, Lamezia Terme, Italy; and 9Istituto di Radiologia, Policlinico ‘Paolo Giaccone’, Palermo, Italy

Received 11 June 2014; accepted after revision 26 August 2014; online publish-ahead-of-print 22 September 2014

T2∗ multislice multiecho cardiac magnetic resonance (CMR) allows quantification of the segmental distribution of myocardial iron overload (MIO). We evaluated whether a preferential pattern MIO was preserved between two CMR scans in regularly chelated thalassaemia major (TM) patients. ..................................................................................................................................................................................... Methods We evaluated prospectively 259 TM patients enrolled in the MIO in Thalassaemia (MIOT) network with a CMR follow-up and results (FU) study at 18 + 3 months and significant MIO at baseline. The T2∗ in the 16 segments and the global value were calculated. Four main circumferential regions (anterior, septal, inferior and lateral) were defined. We identified two groups: severe (n ¼ 80, global T2∗ ,10 ms) and mild –moderate MIO (n ¼ 179, global T2∗ ¼ 10 –26 ms). Based on the CMR reports, 56.4% of patients changed the chelation regimen. For each group, there was a significant improvement in the global heart as well as in regional T2∗ values (P , 0.0001). At the baseline, the mean T2∗ value over the anterior region was significantly lower than the values over the other regions, and the mean T2∗ over the inferior region was significantly lower than the values over the septal and the lateral regions. The same pattern was present at the FU, with a little difference for patients with mild– moderate MIO. ..................................................................................................................................................................................... Conclusion A preferential pattern of iron store in anterior and inferior regions was present at both CMRs, with an increment of T2∗ values at FU due to a baseline CMR-guided chelation therapy. The anterior region seems the region in which the iron accumulates first and is removed later.

Aims

----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords

Cardiac magnetic resonance † Myocardial iron overload † Preferential pattern † Thalassaemia major

Introduction Thalassaemia major (TM) is an inherited disorder of haemoglobin synthesis that results in chronic haemolytic anaemia.1 Regular blood transfusions are mandatory for long-term survival, but they can cause a secondary state of tissue iron overload since humans have no physiological mechanism for active elimination of excess iron. The iron-induced cardiomyopathy is treatable and reversible

if intensive chelation treatment, aimed to remove the iron excess, is instituted in time.2 Three chelators with differentiated mechanism of action and tissue sensitivity are currently available worldwide: deferoxamine (DFO), deferiprone (DFP), and deferasirox (DFX). To improve compliance, minimize/reduce toxicity, and enhance iron removal if an additive or synergistic effect occurs, two chelators can be used either sequentially (on different days) or in combination (both given on the same day).3 However, drug toxicity from

* Corresponding author. Tel: +39 050 315 2824; Fax: +39 050 315 3535. Email: [email protected] Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2014. For permissions please email: [email protected].

326 overdosing of chelation therapy has to be avoided. Therefore, the direct monitoring of cardiac iron overload is the nodal point in the management of TM patients. To date, cardiac magnetic resonance (CMR) is the only technique that quantitatively and non-invasively assesses cardiac iron burden.4,5 Cardiac iron is indicated by a decrease in relaxation time values, determined by the paramagnetic proprieties of iron compounds.6 The most relevant relaxation parameter is the T2∗ , and two validated approaches generally used in the clinical arena are the single-slice7 and the multislice approach.8 In the single-slice approach, the T2∗ value evaluated in a single region of interest drawn in the mid-ventricular septum is taken as representative of the T2∗ value for the whole heart.9 The multislice approach allows a global analysis of the left ventricle (LV)8 and consequently the identification of a heterogeneous iron distribution, demonstrated in primis by histological studies performed on haemochromatotic hearts.10 Importantly, it was shown that the heterogeneity of T2∗ values represented true heterogeneous iron density and could not have been generated by geometric and susceptibility artefacts11 or by myocardial fibrosis and blood oxygenation.12 So, the multislice, multiecho T2∗ CMR is ideally suited to non-invasively detect preferential patterns of myocardial iron deposits. A preferential pattern of iron store in the anterior and inferior regions was previously detected in TM patients with severe and mild–moderate iron overload.13 The prospective evolution of the iron store distribution has never been investigated in patients under chelation therapy. The multislice multiecho T2∗ approach showed a good reproducibility,8 making it valuable for serial monitoring over time. This study aimed to determine whether a preferential pattern of myocardial iron overload (MIO) was preserved between two CMR scans in regularly chelated TM patients.

Methods Study population The Myocardial Iron Overload in Thalassaemia (MIOT) network was built in Italy in 2006 and has gradually grown, reaching the size of 69 thalassaemia centres and 8 CMR centres where exams are performed using homogeneous, standardized, and validated procedures.14 An online database was developed to allow the centres to collect and share patients’ data.15 The database contains clinical and instrumental data as well as information about the chelation regimen followed by the patient from the beginning to the date of the last T2∗ CMR. Of the 1380 TM patients enrolled, 812 had a CMR follow-up (FU) study performed at 18 + 3 months, according to the protocol. All patients had been regularly transfused since early childhood and started chelation therapy from the mid-to-late 1970s, while patients born after the 1970s have received chelation therapy since early childhood. All records were reviewed to assess for changes in chelation regimen based on the results of the baseline CMR. Based on the correspondence between the time history of drug administration and the prescribed regimen, the compliance was defined excellent (.80%), good (60 – 80%), and insufficient (,60%). Each CMR scanning was performed within 1 week before the regularly scheduled blood transfusion. The study complied with the Declaration of Helsinki. All patients gave written informed consent to the protocol. The institutional review board approved this study.

A. Meloni et al.

Cardiac magnetic resonance CMR exams were performed on 1.5-T scanners (GE Signa/Excite HD, Milwaukee, WI, USA) using an eight-element cardiac phased-array receiver surface coil for signal reception. Three parallel short-axis views (basal, medium, and apical) of the LV were obtained by a T2∗ gradient-echo multiecho sequence with electrocardiogram triggering. Each single short-axis view was acquired at nine echo times (TEs, 2.0 –22 ms with an echo spacing of 2.26 ms) in a single end-expiratory breath-hold.16 Acquired images were analysed by expert CMR operators in each of eight CMR centres using a previously validated, custom-written software (Hippo-MIOTw).16 The software is able to map myocardial T2∗ distribution in a 16-segment LV model according to the American Heart Association standardized segmentation.17 Six equiangular segments were used in the basal and medium slices and four in the apical slice (Figure 1A). For each segment, the mean value of the signal intensity along all the TE values was calculated, and the assessed decay curve was fit to the single exponential model. In heavily iron-overloaded hearts, a truncation model was applied to delete the late points with a low signal-to-noise ratio.18 An appropriate correction map built using the data of healthy volunteers scanned in the CoreLab of the MIOT Network was implemented within the software to correct for susceptibility artefacts.16 The inter-site, inter-study, intra-observer, and inter-observer variabilities of the proposed methodology had been previously assessed.8,14 The global T2∗ value averaged over all 16 segmental T2∗ values and the T2∗ value in the mid-ventricular septum evaluated by averaging the T2∗ values in segments 8 and 9 were automatically provided. Four different main circumferential regions, such as anterior, septal, inferior, and lateral, were defined by averaging the correspondent T2∗ values (Figure 1B). Cardiac iron concentration (CIC) was derived from T2∗ values using the formula described by Carpenter et al. 4 The lower limit of normal was taken as 26 ms for global heart and 24.4 ms for mid-ventricular septum T2∗ value.14 MIO was considered severe when T2∗ value was ,10 ms.19

Statistical analysis All data were analysed using SPSS v.13.0 (Chicago, IL, USA) statistical package. All continuous variables are expressed as the mean and standard deviation. Correlation analysis was performed using Spearman’s test. Comparisons between groups were made by an independent sample t-test or the Mann–Whitney test. A x2 test was performed for non-continuous variables. Reproducibility of the T2∗ technique was evaluated using the coefficient of variation (CoV) and the intra-class correlation coefficient (ICC). Within-group comparisons between baseline and FU T2∗ data were performed by the paired t-test. The analysis of variance (ANOVA) or the Kruskall– Wallis test in case of no-normal distribution was used to detect the presence of a significant difference among regional T2∗ changes. One-way repeated-measures ANOVA was used to evaluate whether there was a significant difference between different T2∗ /CIC measurements. Mauchly’s test was used to test assumption of sphericity, and when sphericity could not be assumed, Greenhouse–Geisser-corrected results were taken. The Bonferroni adjustment was used in all pairwise comparisons. A two-tailed P-value of ,0.05 was considered statistically significant.

Results Characterization of the study population The analysis was focused on the 259 patients with significant MIO at baseline (global T2∗ ,26 ms). One-hundred and forty-nine patients

327

Improvement of heart iron by CMR-guided chelation therapy

Figure 1 Heart multislice multiecho T2∗ approach. (A) Segmentation of each slice after the definition of epicardial and endocardial contours. (B) Bull’s-eye representation of the 16 myocardial standard segments and definition of the four different main circumferential regions.

were females, and mean age was 29.5 + 7.8 years. Baseline clinical data of the analysed patients are summarized in Table 1. The selected patient population was divided into two groups: severe (n ¼ 80, baseline global T2∗ ,10 ms) and mild –moderate MIO (n ¼ 179, baseline global T2∗ between 10 and 26 ms). The two groups were not different regarding age (severe MIO 29.1 + 6.7 years vs. mild – moderate MIO 29.6 + 8.2 years; P ¼ 0.613) and sex (severe MIO M/F 39/41 vs. mild– moderate MIO 71/108; P ¼ 0.172). Table 2 reports the chelation treatment performed at both CMRs. With only one exception, all patients were under chelation treatment at both the scans. The exception was a patient who at the first CMR had temporarily interrupted the DFO therapy due to a cataract. The percentage of patients who maintained the same therapy between the two scans was significantly different among the regimens [none: 0/1 (0%), DFO: 30/97 (30.9%), DFP: 12/19 (63.2%), DFX: 49/65 (75.4%), sequential DFO/DFP: 12/17 (70.6%), combination DFO + DFP: 39/58 (67.2%), sequential DFP/DFX: 1/1 (100%), and combination DFO + DFX: 0/1 (0%); P , 0.0001]. Although the therapy was the same, 30 patients (21%) changed dosages/frequency after the baseline CMR scan. Among the 116 (45%) patients who changed chelation therapy according to the baseline CMR exam, 61 (52.6%) switched to a dual therapy (i.e. sequential or combined). Globally, based on the CMR reports, 56.4% of patients changed the chelation regimen (dosages/ frequency or drug). Of the 80 patients with severe MIO at baseline, 56 patients (70%) changed the chelation regimen.

Table 1

Baseline clinical data of the analysed patients

Variable

Value

................................................................................ Regular transfusion starting age (years) Chelation starting age (years) Pre-transfusion haemoglobin (g/dL)

1.70 + 2.37 4.47 + 4.09 9.71 + 0.49

Mean serum ferritin (ng/L) Diabetes mellitus (%)

2010 + 1659 12.8

MRI liver iron concentration (mg/g/dw)

11.36 + 11.18

HCV-RNA positivity (%) Global heart T2∗ (ms)

34.9 14.45 + 6.40

Number of segments with T2∗ ,20 ms

12.19 + 4.69

Left ventricular (LV) ejection fraction (%) Right ventricular (RV) ejection fraction (%)

59.09 + 7.36 59.97 + 7.42

LV end-diastolic volume index (mL/m2)

88.41 + 19.12

RV end-diastolic volume index (mL/m2) Myocardial fibrosis by CMR (%)

83.39 + 19.94 18.5

Left atrial area (cm2/m2)

18.07 + 3.99

Right atrial area (cm2/m2)

16.83 + 3.23

The percentage of patients with excellent/good levels of compliance to the active chelation treatment was 87.7% at the baseline and 93.7% at the FU (P ¼ 0.030).

328

A. Meloni et al.

Table 2

Chelation treatments at the baseline and FU CMRs Chelation at the FU

............................................................................................................................... None

DFO

DFP

DFX

Sequential DFO/DFP

Combined DFO 1 DFP

Sequential DFP/DFX

Combined DFO 1 DFX

............................................................................................................................................................................. Chelation at the baseline None (n ¼ 1) DFO (n ¼ 97)

0 0

0 30

0 8

1 18

0 15

0 25

0 1

0 0

DFP (n ¼ 19)

0

0

12

1

1

5

0

0

DFX (n ¼ 65) Sequential DFO/DFP (n ¼ 17)

0 0

7 1

1 0

49 1

0 12

8 3

0 0

0 0

Combined DFO + DFP (n ¼ 58)

0

8

1

7

3

39

0

0

Sequential DFP/DFX (n ¼ 1) Combined DFO + DFX (n ¼ 1)

0 0

0 1

0 0

0 0

0 0

0 0

1 0

0 0

0

47

22

77

31

80

2

0

Total at the FU

Numbers in italic represent the number of patients who maintained the same therapy.

T2∗ changes The mean intra-observer variability assessed by CoV from one welltrained operator in each of eight sites was 7.37 + 4.42%. There was high correlation between the T2∗ values calculated from images obtained in Pisa and at the other seven MRI sites concerning the global heart (R ¼ 0.95). The CoV and ICC for the global heart considering all the T2∗ values independently from the sites were, respectively, 8% and 0.97. Table 3 summarizes the global heart and the regional T2∗ and CIC values at both baseline and FU CMRs for the whole selected patient population. A significant increase was detected for all T2∗ values (P , 0.0001 for all the pairwise comparisons). Accordingly, there was a significant reduction in all the regional CIC values (P , 0.0001 for all the pairwise comparisons).The T2∗ improvement was not different among the four regions (P ¼ 0.398). A significant correlation between the global T2∗ value and the T2∗ value in the mid-ventricular septum was present at both basal and FU CMRs (r ¼ 0.955 and 0.962, respectively, P , 0.0001). At the baseline, 29 (11.2%) of the 259 patients with significant global heart iron overload had a normal mid-ventricular septum T2∗ value. For these patients, the mean global heart T2∗ was 24.18 + 1.79 ms, and the mean mid-ventricular septum T2∗ was 28.15 + 2.87 ms (P , 0.0001). Of the 29 patients with significant global heart iron overload and a normal mid-septum T2∗ value, the 37.9% changed the chelation regimen, and while three patients had an insufficient compliance at the baseline, at the FU only one had an insufficient compliance. At the FU, 182 of 259 patients (70%) had still a pathological global heart T2∗ value and out of them, 12 (7%) had a normal mid-ventricular septum T2∗ value. For each group, cardiac segments were sorted by a mean T2∗ value (Table 4). At the baseline as well as at the FU in both groups, the lowest T2∗ value was detected in the medium anterior segment and the highest in the basal inferoseptal segment. For each group, the segment order was significantly preserved between the two scans (severe MIO: r ¼ 1.00, P , 0.0001 and mild– moderate MIO: r ¼ 0.935, P , 0.0001).

For each group, there was a significant improvement in the global heart as well as in regional T2∗ values (P , 0.0001 for all the pairwise comparisons; Table 3). Accordingly, for each group, there was a significant reduction in all the regional CIC values (P , 0.0001 for all the pairwise comparisons). Of the patients with mild –moderate iron at baseline, 73 (41%) had a normal global heart T2∗ at the FU. The T2∗ improvement was not statistically different among the regions in either the severe MIO or mild –moderate MIO group (P ¼ 0.263 and 0.719, respectively). The decrease in CIC was significantly higher in the group with severe MIO than in the group with mild –moderate MIO (P , 0.0001 for all the regions).

Patterns of MIO At the baseline, as well as the FU CMR, a significant circumferential variability was detected in the whole patient population (P , 0.0001 for both CMRs) and the pattern was the same. Specifically, the mean T2∗ value over the anterior region was significantly lower than the mean T2∗ values over the other regions, and the mean T2∗ over the inferior region was significantly lower than the T2∗ values over the septal and the lateral regions. A significant circumferential variability was also preserved within each single slice at both the CMRs (P , 0.0001). A significant slice-to-slice variability was found at both the CMRs (P , 0.0001). A T2∗ value was significantly lower in the medium and in the apical slices than in the basal slice. A significant circumferential variability was present at the baseline and at the FU CMR in both groups (P , 0.0001; Figures 2A and 3A). At baseline, the mean T2∗ value over the anterior region was significantly lower than the mean T2∗ values over the other regions, and the mean T2∗ over the inferior region was significantly lower than the T2∗ values over the septal and the lateral regions. Exactly the same pattern was present at the FU for the group with severe MIO. The group with mild –moderate MIO showed a little difference: the mean T2∗ value over the inferior region was lower than that one over the lateral region, but statistical significance was not reached. In each single slice, the significant circumferential variability was preserved in

329

Improvement of heart iron by CMR-guided chelation therapy

Table 3 Comparison between basal and FU T2*/CIC values for the whole patient population with MIO at the baseline and for the two identified groups Region

Baseline values T2∗ (ms) CIC (mg/g/dw)

FU values T ∗2 (ms) CIC (mg/g/dw)

Difference: FU 2 baseline T ∗2 (ms) CIC (mg/g/dw)

P (paired)

............................................................................................................................................................................... Whole patient population with MIO at baseline Global

T2∗ : 14.45 + 6.40 CIC: 2.42 + 1.72

T2∗ : 19.52 + 9.81 CIC: 1.87 + 1.64

T2∗ : 5.07 + 6.5 CIC: 20.55 + 1.08

T2∗ : ,0.0001 CIC: ,0.0001

Anterior

T2∗ : 14 + 5.72 CIC: 3.09 + 2.46

T2∗ : 16.69 + 9.18 CIC: 2.36 + 2.19

T2∗ : 4.54 + 5.57 CIC: 20.73 + 1.85

T2∗ : ,0.0001 CIC: ,0.0001

Septal

T2∗ : 15.49 + 7.12 CIC: 2.24 + 1.62

T2∗ : 20.64 + 10.84 CIC: 1.75 + 1.53

T2∗ : 5.15 + 7.56 CIC: 20.49 + 1.03

T2∗ : ,0.0001 CIC: ,0.0001

T2∗ : 19.08 + 10.86 CIC: 1.76 + 1.62 T2∗ : 20.35 + 9.85 CIC: 2.09 + 1.95

T2∗ : 5.06 + 7.45 CIC: 20.52 + 1.18 T2∗ : 5.31 + 6.91 CIC: 20.59 + 1.29

T2∗ : ,0.0001 CIC: ,0.0001 T2∗ : ,0.0001 CIC: ,0.0001

T2∗ : 14.02 + 7.02 CIC: 2.28 + 1.58 Lateral T2∗ : 15.04 + 6.69 CIC: 2.68 + 2.14 Patient population with severe MIO at baseline Inferior

Global

T2∗ : 7.29 + 1.74 CIC: 4.41 + 1.77

T2∗ : 10.71 + 5.92 CIC: 3.34 + 2.12

T2∗ : 3.41 + 5.5 CIC: 21.07 + 1.60

T2∗ : ,0.0001 CIC: ,0.0001

Anterior

T2∗ : 6.00 + 1.57 CIC: 5.73 + 2.87

T2∗ : 8.92 + 5.56 CIC: 4.27 + 2.85

T2∗ : 2.91 + 5.37 CIC: 21.45 + 2.90

T2∗ : ,0.0001 CIC: ,0.0001

Septal

T2∗ : 7.91 + 2.04 CIC: 4.05 + 1.73

T2∗ : 11.48 + 6.75 CIC: 3.10 + 1.97

T2∗ : 3.57 + 6.36 CIC: 20.95 + 1.52

T2∗ : ,0.0001 CIC: ,0.0001

T2∗ : 9.70 + 5.73 CIC: 3.84 + 2.48 T2∗ : 11.58 + 6.02 CIC: 3.11 + 2.20

T2∗ : 3.01 + 5.24 CIC: 21.21 + 1.89 T2∗ : 3.77 + 5.55 CIC: 20.97 + 1.85

T2∗ : ,0.0001 CIC: ,0.0001 T2∗ : ,0.0001 CIC: ,0.0001

T2∗ : 6.69 + 1.89 CIC: 5.05 + 2.38 Lateral T2∗ : 7.81 + 1.99 CIC: 4.08 + 1.61 Patient population with mild –moderate MIO at baseline Inferior

Global

T2∗ : 17.65 + 4.98 CIC: 1.52 + 0.57

T2∗ : 23.46 + 8.57 CIC: 1.21 + 0.72

T2∗ : 5.81 + 6.87 CIC: 20.32 + 0.62

T2∗ : ,0.0001 CIC: ,0.0001

Anterior

T2∗ : 14.89 + 4.68 CIC: 1.92 + 0.78

T2∗ : 20.16 + 8.32 CIC: 1.51 + 0.97

T2∗ : 5.27 + 6.93 CIC: 20.41 + 0.93

T2∗ : ,0.0001 CIC: ,0.0001

Septal

T2∗ : 18.89 + 5.85 CIC: 1.43 + 0.58

T2∗ : 24.74 + 9.77 CIC: 1.15 + 0.69

T2∗ : 5.85 + 7.96 CIC: 20.28 + 0.61

T2∗ : ,0.0001 CIC: ,0.0001

Inferior

T2∗ : 17.29 + 5.92 CIC: 1.63 + 0.69

T2∗ : 23.28 + 9.95 CIC: 1.31 + 0.89

T2∗ : 5.98 + 8.08 CIC: 20.32 + 0.76

T2∗ : ,0.0001 CIC: ,0.0001

Lateral

T2∗ : 18.26 + 5.39 CIC: 1.47 + 0.57

T2∗ : 24.27 + 8.63 CIC: 1.15 + 0.69

T2∗ : 6.01 + 7.35 CIC: 20.32 + 0.61

T2∗ : ,0.0001 CIC: ,0.0001

both groups at both the scans (P , 0.0001 for all the comparisons; Figures 2B and 3B). A significant slice-to-slice variability was present. For patients with severe MIO, the T2∗ value was significantly lower in the medium and apical slices than in the basal slice at both the scans (P , 0.0001; Figure 2C). For patients with mild– moderate MIO at baseline, only the T2∗ value in the medium slice was significantly lower than that in the basal slice (P ¼ 0.019), while at the FU, both the mean and the apical slices presented a significantly lower T2∗ value than the basal slice (P ¼ 0.001; Figure 3C).

Discussion The introduction of the multislice multiecho T2∗ CMR has made it possible to extend myocardial iron evaluation from the midventricular septum to the entire LV by a segmental approach.8,20 This approach has been demonstrated to have a good intra-observer, inter-observer, and inter-study reproducibility,20 and it has been

shown to be transferable among the first six MRI sites enrolled in the MIOT network.14 Also the intra- and inter-observer variability tested in the eight MRI centres that enrolled patients in this study were demonstrated to be good and concordant with the previously published data. Since possible bias due to susceptibility artefacts can be corrected by applying one standardized T2∗ map of normal human hearts16 and since anyway the artefacts are additive in the R2 ∗ (1/T2∗ ) domain and vanish with the progression of iron overload,21 the multislice approach can account for the heterogeneous myocardial iron distribution. Positano et al. 22 investigated the heterogeneity of compensated T2∗ values in a cohort of 230 TM patients, finding a significant increase in T2∗ value heterogeneity moving from a normal state to a condition of iron overload. In the wake of this study, Meloni et al. looked for the presence of regions with higher iron deposition. A preferential pattern of iron store in the anterior and inferior regions was detected in TM patients with severe and mild–moderate iron overload

330

A. Meloni et al.

Table 4 Order (ascending) of segments based on T2* values at both basal and FU CMR in the group with severe MIO at baseline and in the group with mild –moderate MIO at baseline Segment

Severe MIO at baseline

..................................................................

Baseline

FU

T ∗2

6.44 + 1.96 8.7 + 2.98

.............................

Mild–moderate MIO at baseline

.....................................................................

Baseline

FU

..............................

..............................

Order

T ∗2

................................

Order

T ∗2

Order

T ∗2 ss

Order

4 15

9.46 + 6.02 13.12 + 9.24

4 15

15.31 + 5.75 19.30 + 7.29

3 14

21.24 + 9.65 26.27 + 12.39

3 15

............................................................................................................................................................................... Basal anterior Basal anteroseptal Basal inferoseptal

8.91 + 2.67

16

12.92 + 8.19

16

20.69 + 7.29

16

26.97 + 11.34

16

Basal inferior Basal inferolateral

6.74 + 2.62 8.60 + 2.99

5 14

10.24 + 6.57 12.58 + 6.58

5 14

16.87 + 6.30 18.17 + 6.32

5 9

23.47 + 11.21 22.73 + 8.95

8 5

Basal anterolateral

8.37 + 2.49

13

11.52 + 5.68

13

19.03 + 6.63

13

25.90 + 9.89

14

Medium anterior Medium anteroseptal

5.56 + 1.92 6.93 + 2.15

1 6

8.20 + 5.21 9.99 + 6.34

1 6

14.27 + 5.32 16.63 + 5.78

1 4

19.07 + 9.13 21.92 + 9.92

1 4

Medium inferoseptal

7.73 + 2.17

12

11.05 + 6.40

12

19.70 + 6.91

15

24.91 + 10.28

12

Medium inferior Medium inferolateral

6.22 + 2.07 7.69 + 2.35

3 11

9.29 + 5.93 11.72 + 6.74

3 11

17.28 + 6.90 18.61 + 6.79

7 12

23.10 + 10.44 24.52 + 9.87

6 11

Medium anterolateral

6.96 + 1.93

7

10.29 + 5.02

7

17.17 + 6.43

6

23.41 + 10.14

7

Apical anterior

6.02 + 1.76

2

9.19 + 6.19

2

15.18 + 6.05

2

20.07 + 9.08

2

Apical septal Apical inferior

7.25 + 2.15 7.10 + 2.32

9 8

10.21 + 5.96 9.73 + 5.33

9 8

18.25 + 7.66 17.83 + 8.31

10 8

23.78 + 11.02 23.53 + 11.58

10 9

Apical lateral

7.35 + 2.18

10

11.89 + 8.41

10

18.42 + 6.86

11

25.02 + 10.93

13

at their first CMR.13 On the basis of these findings, the aim of this study was to determine whether a preferential pattern of MIO was preserved between two CMR scans in TM patients regularly transfused and chelated. According to our protocol, the FU CMR was performed after 18 + 3 months from the baseline CMR. This timing was a compromise between the monitoring guidelines according to the basal cardiac iron status5 and the real availability of CMR scans in Italy.15 Our study prospectively demonstrated a significant increment of global and regional T2∗ values at the FU. Based on the CMR using a segmental approach report, 56.4% of the patients changed chelation regimen (drug or frequency or dosage). Moreover, the discrepancy between global and mid-septum T2∗ values present in one-tenth of the patients strengthens the added value of the multislice approach. In fact, 37.9% of the patients with a normal mid-septum T2∗ value, but a pathological global heart T2∗ value and heterogeneous MIO have changed the chelation regimen, and while three patients had an insufficient compliance at the baseline, at the FU only one had an insufficient compliance. At the FU, we registered a significantly higher level of compliance to the chelation therapy, probably thanks to the opportunity for the patients to ‘touch’ directly the iron status of their hearts. Thus, the significant reduction in the cardiac iron burden was likely due to a baseline CMR-guided chelation therapy and to the improvement of the compliance to the chelation therapy. As cardiac complications in well-chelated, well-monitored TM patients are significantly rarer than previously reported,23 the pattern of MIO can emerge as the sensitive predictor of cardiac complications in well-chelated TM populations to detect iron overload early and to identify borderline patients, who could particularly

take advantage of tailoring chelation therapy for preventing ironrelated heart damage. DFP has been proved to be superior to DFO at removing cardiac iron,24,25 and clinical data support the dose-dependent efficacy of DFX in removing and preventing myocardial iron accumulation.26,27 This explains while the percentage of patients who remained on DFO was lower than that of patients who continued the two oral monotherapies between scans. It should be pointed out that several patients refused to change the DFO therapy. To our knowledge, low cardiac T2∗ assessed at the first CMR triggered intensification of chelation regimens in 20% of patients. The association between DFP and DFO has been shown to be the most effective means of reducing cardiac iron loading,28 – 30 and, in fact, after the first CMR 61 patients switched to a combined therapy. In the whole selected patient population as well as in both groups, the anterior region showed the lowest improvement, although a significant difference in the T2∗ increase among the four regions was not detected. Converting the T2∗ values to CIC, the decrease in heart iron burden was significantly higher in the group with severe MIO than in the group with mild –moderate MIO (P , 0.0001 for all the regions). Due to the inverse exponential relationship between T2∗ and heart iron, a small improvement in T2∗ values in severely iron-loaded hearts corresponds to a significant clearance of cardiac iron. The finding indirectly reflects also the higher percentage of changes in the chelation regimen based on the CMR reports among the patients with more severe iron deposition. In this study, we reconfirmed at the baseline the same pattern of MIO detected in a previous study.13 Most importantly, we found out that the circumferential variability was preserved at the FU,

Improvement of heart iron by CMR-guided chelation therapy

331

Figure 2 Baseline and FU CMR results in the patients with severe MIO at baseline. (A) Circumferential T2∗ variability. (B) Mean circumferential T2∗ values in each slice. (C) Slice-to-slice variability.

despite the T2∗ increase. To be more precise, for the group with mild – moderate MIO at the FU, the difference between the inferior region and the lateral region lost the statistical significance. It should be pointed out that many patients with mild– moderate MIO at baseline had at the FU normal global and segmental T2∗ values. This finding

strongly supports the independence of the patterns from artifactual sources. Moreover, at both the scans, the circumferential pattern was highly conserved across slices. In a histological study performed on one freshly deceased, unfixed, iron-loaded heart, Ghugre et al. 31 also

332

A. Meloni et al.

Figure 3 Baseline and FU CMR results in the patients with mild –moderate MIO at baseline. (A) Circumferential T2∗ variability. (B) Mean circumferential T2∗ values in each slice. (C) Slice-to-slice variability.

reported a slice-to-slice reproducibility. This conservation of the relative relationship among regional T2∗ values also within each single slice seems to indicate the absence of a variable regional response to chelation treatment. Carpenter et al.4 reported no systematic variation in mean iron concentration between different LV segments or from base to

apex in their T2∗ CMR study against human hearts. Among the patients with severe MIO, the significantly higher myocardial iron burden in their 10 vs. our 80 patients (5.26 + 0.87 vs. 7.29 + 1.74, P ¼ 0.0005) may explain the differences found in our study and in Ghugre’s data.31 We found a significantly lower iron content in the basal slice vs. the medium and apical slices at both basal and FU CMRS.

Improvement of heart iron by CMR-guided chelation therapy

Limitations It was not possible to compare CMR findings with other imaging techniques (i.e. echocardiography), because they were performed with different time intervals in several patients. Due to the lower rate of cardiac complications in the last years in the thalassaemia population,23 prospective studies with a significant longer follow-up times are recommended to focalize about the clinical relevance of the T2∗ segmental approach in management of the chelation therapy.

Conclusions A preferential pattern of iron store in anterior and inferior regions was present at both basal and FU CMRs, with an increment of T2∗ values at FU due to both a baseline CMR-guided chelation regimen and the improvement of the compliance to the chelation therapy. The anterior region seems to be the region in which the iron accumulates first and is removed later. Our data seem to suggest that the segmental T2∗ cardiac MR approach could be useful for identifying early iron deposit and for tailoring chelation therapy in each patient to better prevent cardiac complications.

Acknowledgements We thank the MIOT haematologists: V. Caruso (Catania), A. Filosa (Napoli), S. Campisi (Siracusa), F. Sorrentino (Roma), M.R. Gamberini (Ferrara), L. Pitrolo (Palermo), R. Rosso (Catania), M. Santodirocco (San Giovanni Rotondo), T. Casini (Firenze), M. Lendini (Olbia), M.C. Putti (Padova), C. Cirotto (Sassari), A. Quarta (Brindisi), A. Maggio (Palermo), M.E. Lai (Cagliari), M.G. Bisconte (Cosenza), C. Gerardi (Sciacca), C. Borgna-Pignatti (Ferrara), S. Grimaldi (Crotone), R. Renni (Casarano), M.P. Smacchia (Roma), B. Piraino (Messina), C. Tassi (Bologna), P.P. Bitti (Nuoro), C. Fidone (Ragusa), F.V. Commendatore (Lentini), S. Armari (Legnago), G. Giuffrida (Catania), D. Maddaloni (Fabriano), A. Pietrapertosa (Bari), G. Palazzi (Modena), A. Pietrangelo (Modena), A. Ciancio (Matera), S. Pulini (Pescara), S. Macchi (Ravenna), G. Roccamo (S. Agata di Militello), G. Serra (Copertino), G. Filati (Piacenza), M. Furbetta (Perugia), M.L. Boffa (Napoli), R. Biguzzi (Cesena), M. Rizzo (Caltanisetta), M. Centra (Foggia), and A. Quota (Gela). We also thank the MIOT cardioradiologists E. Chiodi (Ferrara) and A. Vallone (Catania), Claudia Santarlasci for skillful secretarial work, and all patients for their cooperation. Conflict of interest: A.P. received speaker’s honoraria from Chiesi, ApoPharma, Inc., and Novartis.

Funding The MIOT project receives ‘no-profit support’ from industrial sponsorships (Chiesi Farmaceutici S.p.A. and ApoPharma, Inc.).

References 1. Weatherall DJ, Clegg JB, Higgs DR, Wood WG. The hemoglobinopathies. In: Sriver CR, Beadet AL, eds. The Metabolic and Molecular Bases of Inherited Disease. 7th ed. New York: McGraw-Hill; 1995. p. 3417 –84. 2. Anderson LJ, Westwood MA, Holden S, Davis B, Prescott E, Wonke B et al. Myocardial iron clearance during reversal of siderotic cardiomyopathy with intravenous desferrioxamine: a prospective study using T2* cardiovascular magnetic resonance. Br J Haematol 2004;127:348–55.

333 3. Kwiatkowski JL. Real-world use of iron chelators. Hematol Am Soc Hematol Educ Program 2011;2011:451 –8. 4. Carpenter JP, He T, Kirk P, Roughton M, Anderson LJ, de Noronha SV et al. On T2* magnetic resonance and cardiac iron. Circulation 2011;123:1519 –28. 5. Pennell DJ, Udelson JE, Arai AE, Bozkurt B, Cohen AR, Galanello R et al. Cardiovascular function and treatment in beta-thalassemia major: a consensus statement from the American Heart Association. Circulation 2013;128:281 –308. 6. Gossuin Y, Roch A, Muller RN, Gillis P. Relaxation induced by ferritin and ferritin-like magnetic particles: the role of proton exchange. Magn Reson Med 2000;43:237 – 43. 7. Anderson LJ, Holden S, Davis B, Prescott E, Charrier CC, Bunce NH et al. Cardiovascular T2-star (T2*) magnetic resonance for the early diagnosis of myocardial iron overload. Eur Heart J 2001;22:2171 –9. 8. Pepe A, Positano V, Santarelli F, Sorrentino F, Cracolici E, De Marchi D et al. Multislice multiecho T2* cardiovascular magnetic resonance for detection of the heterogeneous distribution of myocardial iron overload. J Magn Reson Imaging 2006;23: 662 –8. 9. Westwood M, Anderson LJ, Firmin DN, Gatehouse PD, Charrier CC, Wonke B et al. A single breath-hold multiecho T2* cardiovascular magnetic resonance technique for diagnosis of myocardial iron overload. J Magn Reson Imaging 2003;18:33 –9. 10. Olson LJ, Edwards WD, McCall JT, Ilstrup DM, Gersh BJ. Cardiac iron deposition in idiopathic hemochromatosis: histologic and analytic assessment of 14 hearts from autopsy. J Am Coll Cardiol 1987;10:1239 –43. 11. Positano V, Pepe A, Santarelli MF, Ramazzotti A, Meloni A, De Marchi D et al. Multislice multiecho T2* cardiac magnetic resonance for the detection of heterogeneous myocardial iron distribution in thalassaemia patients. NMR Biomed 2009;22:707 –15. 12. Meloni A, Pepe A, Positano V, Favilli B, Maggio A, Capra M et al. Influence of myocardial fibrosis and blood oxygenation on heart T2* values in thalassemia patients. J Magn Reson Imaging 2009;29:832 –7. 13. Meloni A, Positano V, Pepe A, Rossi G, Dell’Amico M, Salvatori C et al. Preferential patterns of myocardial iron overload by multislice multiecho T2* CMR in thalassemia major patients. Magn Reson Med 2010;64:211 –9. 14. Ramazzotti A, Pepe A, Positano V, Rossi G, De Marchi D, Brizi MG et al. Multicenter validation of the magnetic resonance T2* technique for segmental and global quantification of myocardial iron. J Magn Reson Imaging 2009;30:62 –8. 15. Meloni A, Ramazzotti A, Positano V, Salvatori C, Mangione M, Marcheschi P et al. Evaluation of a web-based network for reproducible T2* MRI assessment of iron overload in thalassemia. Int J Med Inform 2009;78:503–12. 16. Positano V, Pepe A, Santarelli MF, Scattini B, De Marchi D, Ramazzotti A et al. Standardized T2* map of normal human heart in vivo to correct T2* segmental artefacts. NMR Biomed 2007;20:578 –90. 17. Cerqueira MD, Weissman NJ, Dilsizian V, Jacobs AK, Kaul S, Laskey WK et al. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart: a statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation 2002;105:539 –42. 18. He T, Gatehouse PD, Smith GC, Mohiaddin RH, Pennell DJ, Firmin DN. Myocardial T2* measurements in iron-overloaded thalassemia: an in vivo study to investigate optimal methods of quantification. Magn Reson Med 2008;60:1082 –9. 19. Kirk P, Roughton M, Porter JB, Walker JM, Tanner MA, Patel J et al. Cardiac T2* magnetic resonance for prediction of cardiac complications in thalassemia major. Circulation 2009;120:1961 –8. 20. Pepe A, Lombardi M, Positano V, Cracolici E, Capra M, Malizia R et al. Evaluation of the efficacy of oral deferiprone in beta-thalassemia major by multislice multiecho T2*. Eur J Haematol 2006;76:183 –92. 21. Anderson DJ, Dendy JM, Paschal CB. Simulation study of susceptibility gradients leading to focal myocardial signal loss. J Magn Reson Imaging 2008;28:1402 –8. 22. Positano V, Salani B, Pepe A, Santarelli MF, De Marchi D, Ramazzotti A et al. Improved T2* assessment in liver iron overload by magnetic resonance imaging. Magn Reson Imaging 2009;27:188 –97. 23. Modell B, Khan M, Darlison M, Westwood MA, Ingram D, Pennell DJ. Improved survival of thalassaemia major in the UK and relation to T2* cardiovascular magnetic resonance. J Cardiovasc Magn Reson 2008;10:42. 24. Pennell DJ, Berdoukas V, Karagiorga M, Ladis V, Piga A, Aessopos A et al. Randomized controlled trial of deferiprone or deferoxamine in beta-thalassemia major patients with asymptomatic myocardial siderosis. Blood 2006;107:3738 –44. 25. Pepe A, Meloni A, Capra M, Cianciulli P, Prossomariti L, Malaventura C et al. Deferasirox, deferiprone and desferrioxamine treatment in thalassemia major patients: cardiac iron and function comparison determined by quantitative magnetic resonance imaging. Haematologica 2011;96:41– 7. 26. Pathare A, Taher A, Daar S. Deferasirox (Exjade(R)) significantly improves cardiac T2* in heavily iron-overloaded patients with beta-thalassemia major. Ann Hematol 2010;89:405 –9. 27. Pennell DJ, Porter JB, Cappellini MD, El-Beshlawy A, Chan LL, Aydinok Y et al. Efficacy of deferasirox in reducing and preventing cardiac iron overload in beta-thalassemia. Blood 2010;115:2364 –71.

334 28. Berdoukas V, Chouliaras G, Moraitis P, Zannikos K, Berdoussi E, Ladis V. The efficacy of iron chelator regimes in reducing cardiac and hepatic iron in patients with thalassaemia major: a clinical observational study. J Cardiovasc Magn Reson 2009;11:20. 29. Lai ME, Grady RW, Vacquer S, Pepe A, Carta MP, Bina P et al. Increased survival and reversion of iron-induced cardiac disease in patients with thalassemia major receiving intensive combined chelation therapy as compared to desferoxamine alone. Blood Cells Mol Dis 2010;45:136–9.

A. Meloni et al.

30. Pepe A, Meloni A, Rossi G, Cuccia L, D’Ascola GD, Santodirocco M et al. Cardiac and hepatic iron and ejection fraction in thalassemia major: multicentre prospective comparison of combined deferiprone and deferoxamine therapy against deferiprone or deferoxamine monotherapy. J Cardiovasc Magn Reson 2013; 15:1. 31. Ghugre NR, Enriquez CM, Gonzalez I, Nelson MD Jr, Coates TD, Wood JC. MRI detects myocardial iron in the human heart. Magn Reson Med 2006;56:681 – 6.

IMAGE FOCUS

doi:10.1093/ehjci/jeu222 Online publish-ahead-of-print 11 November 2014

.............................................................................................................................................................................

Giant superior vena cava aneurysm after Fontan operation Ibrahim Cansaran Tanidir1*, Aysel Turkvatan2, Taner Kasar1, and Alper Guzeltas1 1 ¨ zal Bulvarı No: 11, Department of Pediatric Cardiology, Istanbul Mehmet Akif Ersoy Thoracic and Cardiovascular Surgery Center and Research Hospital, ˙Istasyon Mah.Turgut O 34303 Ku¨c¸u¨kc¸ekmece Istanbul, Turkey and 2Department of Radiology, Istanbul Mehmet Akif Ersoy Thoracic and Cardiovascular Surgery Center and Research Hospital, Istanbul, Turkey

* Corresponding author. Tel: +90 212 692 20 00; Fax: +90 212 471 94 94, Email: [email protected]

A 23-year-old female, as diagnosed with complex cyanotic congenital heart disease in infancy, admitted to the emergency department with acute chest pain and shortness of breath on exertion. At the age of 3 and 7 years, she underwent right bidirectional Glenn anastomosis and extracardiac Fontan operation, respectively, because of the diagnosis of double inlet left ventricle, right ventricular hypoplasia, and ventriculoarterial discordance. Since then she was lost in follow-up. The initial chest X-ray showed an oval, welldefined opacity with smooth contours causing dilatation of the mediastinal shadow. A chest computed tomography scan demonstrated a contrast-enhancing fusiform sac, suggesting a giant superior vena cava (SVC) aneurysm measuring 5.7 × 8.1 cm with no significant stasis and no evidence of thrombus or filling defects in the pulmonary arteries (Panels A–E, and see Supplementary data online, Video S1). A catheterization study revealed a large SVC aneurysm measuring 5.7 cm (Panel F, and see Supplementary data online, Video S2). Mean superior-inferior caval and right-left pulmonary artery pressures were ,6 mmHg and no clinically significant fistula was detected. Medical therapy was recommended. Aneurysmal dilation of the SVC is a rare anomaly. To the best of our knowledge, there has been only two cases following Glenn anastomosis, and there have been no reports after Fontan operation. The exact pathogenesis of SVC aneurysm is not known and the possibility of the deficiency of longitudinal muscle layer of the adventitia has been reported. The common feature in all reported patients was the region of the aneurysm. (Panel A) Chest X-ray, arrow shows a mediastinal shadow, (Panel B) axial multiple detector computed tomography (MDCT) image, (Panel C) oblique coronal thin maximum intensity projection MDCT image, (Panel D) anterior volume rendering MDCT image, (Panel E) posterior volume rendering MDCT image, and (Panel F ) the angiogram demonstrating the fusiform aneurysm of superior vena cava. Ao, aorta; LIV, left innominate vein; LPA, left pulmonary artery; RIV, right innominate vein; RPA, right pulmonary artery; SVC, superior vena cava. Supplementary data are available at European Heart Journal – Cardiovascular Imaging online. Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2014. For permissions please email: [email protected].