ARTICLES

Comparison of effects of oral deferiprone and subcutaneous

desferrioxamine on myocardial iron concentrations and ventricular
function in beta-thalassaemia

Lisa J Anderson, Beatrix Wonke, Emma Prescott, Sally Holden, J Malcolm Walker, Dudley J Pennell

Summary Introduction
Heart failure due to iron overload can develop either as a
Background Despite the introduction of the parenteral iron result of excess dietary absorption (hereditary
chelator desferrioxamine more than 30 years ago, 50% of haemochromatosis) or from repeated blood transfusions.
patients with thalassaemia major die before the age of The most striking model of cardiac iron overload is seen
35 years, predominantly from iron-induced heart failure. The in thalassaemia major, in which heart failure remains the
only alternative treatment is oral deferiprone, but its long- major cause of death (60%), greatly exceeding deaths
term efficacy on myocardial iron concentrations is unknown. from infection (13%) and liver disease (6%).1 Despite the
introduction of the iron-chelating agent desferrioxamine
Methods We compared myocardial iron content and cardiac more than 30 years ago in the UK, 50% of patients still
function in 15 patients receiving long-term deferiprone die before reaching the age of 35 years.2 This high
treatment with 30 matched thalassaemia major controls who mortality is partly the result of difficulties with
were on long-term treatment with desferrioxamine. administration of desferrioxamine. This drug requires
Myocardial iron concentrations were measured by a new long subcutaneous or intravenous infusions on at least
magnetic-resonance T2* technique, which shows values 4 days a week; compliance with treatment is inadequate in
inversely related to tissue iron concentration. many cases. The need for an effective alternative approach
with an oral iron chelator has long been acknowledged,3
Findings The deferiprone group had significantly less and medium-term results from prospective trials of the
myocardial iron (median 34·0 ms vs 11·4 ms, p=0·02) and oral chelator deferiprone (1,2-dimethyl-3-hydroxypyridin-
higher ejection fractions (mean 70% [SD 6·5] vs 63% [6·9], 4-one) seemed promising;4–6 however, the long-term
p=0·004) than the desferrioxamine controls. Excess effectiveness of this drug has been questioned because
myocardial iron (T2* <20 ms) was less common in the liver iron content is high in some patients.7,8 Since the
deferiprone group than in the desferrioxamine controls (four primary objective of iron-chelation therapy is to prevent
[27%] vs 20 [67%], p=0·025), as was severe (T2* <10 ms) the lethal cardiac complications from myocardial iron
iron overload (one [7%] vs 11 [37%], p=0·04). The odds ratio deposition, myocardial iron and ventricular function
for excess myocardial iron in the desferrioxamine controls should also be taken into account in assessment of the
versus the deferiprone group was 5·5 (95% CI 1·2–28·8). effectiveness of chelating agents. New magnetic resonance
techniques can assess both myocardial iron and
Interpretation Conventional chelation treatment with ventricular function in the same study.9 Our aim,
subcutaneous desferrioxamine does not prevent excess therefore, was to investigate whether deferiprone is
cardiac iron deposition in two-thirds of patients with effective in controlling myocardial iron.
thalassaemia major, placing them at risk of heart failure and
its complications. Oral deferiprone is more effective than Methods
desferrioxamine in removal of myocardial iron. Participants
We included all patients based at the Whittington
Lancet 2002; 360: 516–20 Hospital, London, UK, who received chelation with
See Commentary page 501 deferiprone alone for longer than 3 years (mean duration
5·7 years [SD 1·8]), between May, 1999, and December,
2000. For each deferiprone patient, we assigned two
controls with thalassaemia major, matched for age, sex,
and current ferritin concentration, who were receiving
standard subcutaneous desferrioxamine. Controls were
chosen from the thalassaemic population also treated at
this centre. The criteria for matching were a maximum
age difference of 5 years (mean difference 2·4 years) and
a maximum ferritin difference of 1000 g/L (mean
difference 447 g/L). When more than two controls were
identified, those with the smallest age differences were
chosen. All patients received transfusions every 2–3 weeks
Cardiovascular Magnetic Resonance Unit, Royal Brompton to maintain the pretransfusion haemoglobin concentration
Hospital, London SW3 6NP, UK (L J Anderson MB, at 90–95 g/L, and all had received iron-chelation therapy
Prof D J Pennell MD); University College Hospital, London since the late 1970s or from early childhood in patients
(S Holden BA, J M Walker MD); and Whittington Hospital, London born after this time. Exclusion criteria were the inability to
(B Wonke MD, E Prescott BSc) undergo magnetic-resonance scanning (claustrophobia,
Correspondence to: Prof Dudley J Pennell pregnancy, or pacemaker fitted). The reason for starting
(e-mail: [Link]@[Link]) deferiprone was refusal or inability to comply with the

516 THE LANCET • Vol 360 • August 17, 2002 • [Link]

For personal use. Only reproduce with permission from The Lancet Publishing Group.
 ARTICLES

A Normal myocardial iron ventricular slice was acquired at nine separate echo times.
After subtraction of background noise, the signal intensity
of the liver or myocardial parenchyma was plotted against
the echo time for the image. A trendline was fitted to the
resulting exponential decay curve, with an equation of the
form y=Ke-TE/T2* (where K is a constant, TE the echo
time, and y the signal intensity). Myocardial T2* values
measured in healthy volunteers showed a normal
distribution with a mean value of 52 ms and a lower 95%
confidence limit of 20 ms. We measured ventricular
B Severe myocardial iron overload
volumes, mass, and ejection fraction by standard
cardiovascular magnetic-resonance techniques,10 which
are highly reproducible,11 with published normal ranges.12
Serum ferritin was measured by enzyme immunoassay,
and all values are referable to the WHO Ferritin 80/602
First International Standard (normal range 15–300 g/L).
Follow-up echocardiographic data in the form of
M-mode left-ventricular dimensions and shortening
fraction (defined as the difference between left-ventricular
end-diastolic and end-systolic dimensions as a percentage
Figure 1: Magnetic-resonance scans in patients with of end-diastolic dimension) were available for nine of
thalassaemia 15 patients who received deferiprone and 20 of
Left scans are horizontal long axis, the right ones mid-short axis. A: low 30 desferrioxamine controls. These echocardiograms were
myocardial iron deposition. The left-ventricular volumes are normal, and taken for clinical reasons for annual or biannual review.
myocardial signal intensity (long arrow) is similar to that arising from
skeletal muscle (short arrow). Left-ventricular ejection fraction was 70%.
In this case, the liver is very dark (dotted arrow), indicating heavy hepatic Statistical analysis
iron deposition despite the normal myocardial appearances. B: severe We compared patients’ characteristics by means of
myocardial iron overload. The myocardial signal intensity is dark (long Student’s t test (age and serum ferritin) or Fisher’s exact
arrow) compared with skeletal muscle (short arrow). The ventricle is
dilated and thickened. Cine imaging showed greatly reduced systolic test (presence or absence of diabetes mellitus, hypo-
function (left-ventricular ejection fraction 39%) with a restrictive filling pituitarism, hypothyroidism, hepatitis C). Myocardial
pattern. Liver signal in this case is well preserved (dotted arrow). T2* and liver iron values were positively skewed in all
groups, and non-parametric analyses were used for
subcutaneous regimen in 12 patients and toxic effects of comparison of these variables. Paired comparisons were
desferrioxamine in three (auditory in two and growth made between the deferiprone and desferrioxamine
impairment in one). The mean administered dose of groups with each deferiprone patient paired to the mean
deferiprone was 80·5 mg/kg bodyweight (SD 10·1), value of the two matched desferrioxamine patients.
divided into three doses per day. The drug was Wilcoxon’s signed-rank test was used for myocardial T2*
manufactured by Pfertec Pharmaceuticals, Essex, UK, and liver iron concentrations, and Student’s t test for
under licence from Apotex, Toronto, Canada. The mean indices of left-ventricular function. Since left-ventricular
dose of desferrioxamine was 37·4 mg/kg bodyweight (7·9) volumes and mass vary with the height and weight of a
on 5·1 days (0·8) per week; desferrioxamine treatment patient, we normalised these indices to body surface area.
had been started a mean of 18·3 years (2·8) earlier. At the We assessed comparisons of proportions of patients with
time of magnetic-resonance assessment, the drug was excess myocardial iron deposition in the groups with
administered via 24 h subcutaneous infusions in Fisher’s exact test. Unpaired comparisons between the
13 patients and via overnight subcutaneous infusions 30 desferrioxamine control patients and a larger
in 17. population of 160 patients receiving subcutaneous
At the time of this analysis, 224 patients with desferrioxamine were made with the Mann-Whitney U
thalassaemia major had been referred for magnetic- test (myocardial T2* and liver iron) and the unpaired
resonance scanning (of a total UK population of around Student’s t test (age, left-ventricular ejection fraction, and
820 patients). Of these, 160 patients had received long- serum ferritin). The odds ratio for the prevalence of
term chelation therapy with subcutaneous desferrio- excess myocardial iron in the deferiprone group
xamine. We compared this group with the desferrio- compared with the desferrioxamine group was calculated
xamine control group to assess whether the controls were with a 95% CI.
representative of the general thalassaemic population.
Role of the funding source
Protocol The sponsors of the study had no role in study design,
To quantify iron loading in the heart and the liver (figure data collection, data analysis, data interpretation, or
1), we measured T2*, a magnetic-resonance variable that writing of the report.
is inversely related to tissue iron concentration, by a
Deferiprone Desferrioxamine
previously validated method.9 The technique has high group (n=15) controls (n=30)
reproducibility both in the liver (coefficient of variation
Age (mean, SD) (years) 29·0 (6·3) 29·4 (7·1)
3·3%) and in the heart (5·0%).9 All patients were
M/F 12/3 24/6
scanned, using the same sequence, with a Picker 1·5 T Serum ferritin (mean, SD) (g/L) 1236 (651) 1250 (508)
Edge Scanner (Marconi Medical Systems, Cleveland, Diabetes mellitus 6 (40%) 11 (37%)
OH, USA). Each scan included the measurement of liver Hypopituitarism 7 (47%) 20 (67%)
and heart iron by means of T2*. For the calculation of Hypothyroidism 2 (13%) 2 (7%)
liver T2* a single transverse slice through the liver was Hepatitis C 4 (27%) 9 (30%)
acquired at eight different echo times, and for the Data are number (%) unless otherwise indicated.
measurement of myocardial T2* a single short-axis mid- Table 1: Characteristics of patients

THE LANCET • Vol 360 • August 17, 2002 • [Link] 517

For personal use. Only reproduce with permission from The Lancet Publishing Group.
ARTICLES

Normal range Deferiprone group Desferrioxamine controls Mean difference (95% CI) p
Variable
Myocardial T2* (median, IQR) (ms) >20 34·0 (18·0–56·0) 11·4 (7·0–25·0) ·· 0·02
Left-ventricular measurements (mean, SD)
Ejection fraction (%) 56–78 70 (6·5) 63 (6·9) 7·5 (2·8 to 12·2) 0·004
End-systolic volume index (mL/m2) 12–32 24 (10) 36 (11) –11 (–17 to –5) 0·03
End-diastolic volume index (mL/m2) 44–89 81 (19) 94 (13) –12 (–23 to –1) 0·01
Mass index (g/m2) 64–109 87 (18) 94 (11) –6·9 (–15 to 1·3) 0·09
Liver iron (median, IQR) (mg/g liver dry 0·35–1·36 5·1 (2·8–10·0) 3·5 (2·7–4·6) ·· 0·03
weight)
All measurements made by magnetic resonance. Normal ranges are those derived from cardiovascular magnetic resonance for the measured variables.12
Table 2: Paired comparisons of heart and liver iron content and indices of left-ventricular function

Results receiving standard subcutaneous treatment with

Clinically relevant characteristics for patients in the two desferrioxamine. The two sets of desferrioxamine-treated
study groups are compared in table 1. patients were similar in age, and there was no significant
The deferiprone-treated group had significantly less difference in myocardial iron or left-ventricular ejection
myocardial iron than the desferrioxamine-treated group fraction, although serum ferritin was higher in the large
(median myocardial T2* 34·0 vs 11·4 ms, p=0·02, table 2). group (table 3). The proportion of patients with excess
The deferiprone group also had a higher mean left- myocardial iron deposition (T2* <20 ms) was similar in
ventricular ejection fraction (p=0·004) and less left- the two desferrioxamine groups, as was the proportion
ventricular dilatation in systole (p=0·03) and diastole with severe iron overload.
(p=0·01) than the control group. The left-ventricular mass The median follow-up times with echocardiography in
index was lower, but not significantly so, in the deferiprone the deferiprone (nine patients) and desferrioxamine
patients (p=0·09). Excess myocardial iron (myocardial T2* (20 patients) groups were 2·5 years (range 1·1–5·9) and
<20 ms) was noted in four (27%) deferiprone-treated 2·1 years (1·4–6·6). The initial and final shortening
patients compared with 20 (67%) desferrioxamine-treated fractions in the deferiprone group were 33% (SD 9) and
patients (p=0·025), and severe iron overload (T2* <10 ms) 36% (6); the mean improvement was 3·1% (p=0·1), with
was seen in one (7% ) and 11 (37%), respectively (p=0·04). great improvement seen in two patients (figure 2).
The odds ratio for excess myocardial iron in the The initial and final shortening fractions in the des-
desferrioxamine versus the deferiprone group was 5·5 ferrioxamine group were 32% (7) and 33% (5); the mean
(95% CI 1·2–28·8). improvement was 0·7% (p=0·6) (figure 3). Although
Liver T2* measurements were converted into dry-weight there was no significant difference in end-systolic
liver iron measurements as previously described.9 Despite dimensions between the groups at the initial scan
the lower myocardial iron and improved left-ventricular (deferiprone patients 3·4 cm, desferrioxamine patients
function in the deferiprone-treated patients (table 2), this 3·5 cm, p=0·4), end-systolic dimensions were significantly
group had significantly higher liver iron content than did the smaller in the deferiprone group at the final scan (3·1 vs
desferrioxamine group (median 5·1 vs 3·5 mg/g liver dry 3·6 cm, p=0·02).
weight, p=0·03).
Patient 1
To ensure that the myocardial iron and ventricular
function in the desferrioxamine group were representative 1992 1993 1998
of these features in the wider population with thalassaemia
major, we also compared the results of the desferrio-
xamine control group with those of a further 160 patients
Normal Desferrioxamine Desferrioxamine p
range controls (n=30) population
LVEDD 5·5 cm LVEDD 5·3 cm LVEDD 4·6 cm
(n=160)
LVESD 4·7 cm LVESD 4·4 cm LVESD 2·8 cm
Variable
Age (mean, .. 29 (7·1) 27 (8·1) 0·2
Patient 2
SD) (years)
Myocardial T2* >20 11·4 (7·0–25·0) 14·8 (9·2–30·0) 0·2 LVEDD 6·2 cm LVEDD 5·7 cm
(median, IQR) (ms) LVESD 5·1 cm LVESD 4·5 cm
Left-ventricular 56–78 63 (6·9) 65 (9·3) 0·2
ejection fraction
(mean, SD) (%)
Serum ferritin 15–300 1250 (508) 2034 (1252) 0·0004
(mean, SD)
(g/L) 1996 2000
Liver iron 0·35–1·36 3·5 (2·7–4·6) 4·8 (2·7–10·0) 0·07
(median, IQR) Figure 2: M-mode echocardiography in patients treated with
(mg/g dry weight) deferiprone
Patients with .. 20 (67%) 101 (63%) 0·8 Patient 1: images of left ventricle from a 24-year-old man at the start of
treatment with deferiprone alone and after 1 year and 6 years of
excess myocardial
treatment. The initial scan showed severely impaired ventricular function
iron
with a dilated ventricle. There was some improvement by 1 year. After
Patients with .. 11 (37%) 46 (29%) 0·6 6 years, ventricular dimensions and systolic function have normalised.
severe myocardial Patient 2: images of left ventricle from a 21-year-old man at the start of
iron treatment with deferiprone alone and after 4 years of treatment. The
Table 3: Comparison of the desferrioxamine control group with initial image shows a dilated, impaired left ventricle. After 4 years the
ventricular dimensions have improved but have not yet normalised.
a larger population of patients with thalassaemia major LVEDD=left-ventricular end-diastolic dimension; LVESD=left-ventricular
treated with subcutaneous desferrioxamine end-systolic dimension.

518 THE LANCET • Vol 360 • August 17, 2002 • [Link]

For personal use. Only reproduce with permission from The Lancet Publishing Group.
 ARTICLES

iron in a one-to-one ratio. Side-effects of deferiprone

include discoloured urine (40%), nausea (24%),
LVEDD 5·0 cm arthropathy (11%), and reversible agranulocytosis (0·5%).
LVESD 3·8 cm Although results of formal dose-response studies have
shown that iron excretion is worse with deferiprone than
with desferrioxamine, findings of medium-term clinical
trials have shown that it is effective,4,6,25,26 possibly because
compliance is better. However, in a longer trial of
18 patients over 4·6 years,7 eight patients had hepatic iron
concentrations above 80 mol/g liver wet weight (a value
previously believed to be associated with increased risk of
cardiac disease, and about equivalent to 15 mg/g dry
weight). The investigators concluded that deferiprone
“does not adequately control body iron burden”. This
conclusion is, however, arguable. First, all patients had
previously been treated with desferrioxamine, yet ten had
1996
hepatic iron concentrations of more than 80 mol/g liver
wet weight at the beginning of the trial. Second, during
treatment with deferiprone, the mean liver iron content
fell from a mean of 88·7 mol/g (SD 12·1) to 65·5 mol/g
LVEDD 4·9 cm
LVESD 3·3 cm
(7·9) liver wet weight. Third, myocardial iron content was
not measured. We aimed to assess the effect of
deferiprone on myocardial iron, because heart failure is
the main cause of death in thalassaemia, greatly exceeding
deaths from liver disease.1 Thus, the primary role of iron-
chelation therapy is to prevent premature death from
myocardial iron overload.
The emergence of advanced magnetic-resonance
techniques has made possible accurate assessment of both
liver and cardiac iron in the same study. Such
measurements have shown that myocardial iron cannot be
predicted from liver iron concentration, and that left-
2000 ventricular ejection fraction is unrelated to liver iron or
serum ferritin concentrations in thalassaemic patients.9,27
Figure 3: M-mode echocardiology in patient treated with Thus, direct myocardial iron measurements are essential.
intravenous and subcutaneous desferrioxamine Our results indicate significantly lower myocardial iron
Images of left ventricle from a 28-year-old man with poor compliance to content and a lower proportion of patients with excess
subcutaneous desferrioxamine treatment and who presented with poor myocardial iron in the deferiprone group than in the
left ventricular function and was treated for 1 year with intensive
intravenous desferrioxamine. The first image (upper) was acquired in desferrioxamine controls, combined with better left-
1996 when the patient was switched back from intravenous to ventricular ejection fractions. These findings suggest a
subcutaneous desferrioxamine. Subsequent to this change in regimen, cardioprotective effect of deferiprone, arising despite the
compliance to standard subcutaneous treatment was good, and by 2000 higher liver iron contents in the deferiprone group. These
(lower) echocardiographic appearances had improved.
results show that deferiprone is an effective chelator for
myocardial iron, and emphasise the importance of the
Discussion variation between organs in iron concentrations and most
The iron chelator desferrioxamine was introduced more notably the poor correlation between liver and myocardial
than 30 years ago13,14 and remains the only chelator iron.9,27 A possible mechanism for better cardioprotection
approved for regular use in North America and the only from deferiprone than from desferrioxamine is its greater
first-line agent approved for use in Europe. Desferrio- ability to penetrate myocardial cells, where excess iron is
xamine improves hepatic, cardiac, and endocrine stored in lysosomes as ferritin and haemosiderin.
dysfunction and lengthens survival in patients with iron Deferiprone can cross cell membranes more effectively
overload.15–17 The disadvantages include high cost,18 the than desferrioxamine because it is of lower molecular
requirement for daily parenteral administration, and local weight and is lipophilic.28,29 Conversely, in the liver,
and systemic toxic effects, which include visual19 and desferrioxamine has the advantage of facilitated transport
auditory neurotoxic effects,20 skeletal abnormalities,21 and into cells via an active uptake mechanism.30 Although our
growth retardation.22 Toxicity is increased in the presence results indicate better myocardial iron concentrations with
of low hepatic iron,23 so the risk of side-effects could be deferiprone and better hepatic iron concentrations with
lessened by reducing the dose when hepatic iron desferrioxamine, there is much variability among patients
concentrations are low; however, patients should be that remains to be explained. An individualised approach
regularly monitored for such adverse effects.23 Despite the to iron-chelation therapy, with assessment of iron in each
administration of this arduous regimen for more than target organ might, therefore, be necessary to optimise
20 years, or from early childhood in younger patients, care of patients.
iron-induced heart failure remains the principal cause of Our study had some limitations. The retrospective
premature death in these patients.15,24 analysis used a magnetic-resonance technique that was
To date, only deferiprone has been introduced for not available when treatment with deferiprone was
clinical use as an orally active alternative to instituted. Ideally, the controls would be matched for
desferrioxamine. Deferiprone is a bidentate chelator, baseline myocardial iron deposition, but this information
binding iron in a ratio of three to one, whereas was not available for either group. Although all our
desferrioxamine is a larger hexadentate molecule, binding patients on long-term deferiprone were included, the

THE LANCET • Vol 360 • August 17, 2002 • [Link] 519

For personal use. Only reproduce with permission from The Lancet Publishing Group.
ARTICLES

number is small; a larger, randomised prospective study, and effectiveness of iron-chelation therapy with deferiprone for
assessing tissue iron and ventricular function, is needed to thalassemia major. N Engl J Med 1998; 339: 417–23.
confirm our findings. In this small group, myocardial iron 8 Hoffbrand AV, AL-Refaie F, Davis B, et al. Long term trial of
deferiprone in 51 transfusion dependent iron overloaded patients.
could have been low before deferiprone treatment started. Blood 1998; 91: 295–300.
However, the echocardiographic data that are available 9 Anderson LJ, Holden S, Davis B, et al. Cardiovascular T2* magnetic
show striking improvement in end-systolic dimension resonance for the early diagnosis of myocardial iron overload.
after the introduction of deferiprone, which would argue Eur Heart J 2001; 22: 2171–79.
against this possibility. Alternatively, the desferrioxamine 10 Bellenger NG, Francis JM, Davies CL, Coats A, Pennell DJ.
Establishment and performance of a magnetic resonance cardiac
controls could have had unusually high myocardial iron function clinic. J Cardiovasc Magn Reson 2000; 2: 15–22.
concentrations, but the comparison with the larger group 11 Bellenger NG, Davies LC, Francis JM, Coats AJS, Pennell DJ.
of 160 patients shows that myocardial iron concentrations Reduction in sample size for studies of remodelling in heart failure by
and left-ventricular ejection fractions in the controls are the use of cardiovascular magnetic resonance. J Cardiovasc Magn
Reson 2000; 2: 271–78.
representative.
12 Lorenz CH, Walker ES, Morgan VL, et al. Normal human right and
Excess myocardial iron deposition happens in more left ventricular mass, systolic function and gender differences by cine
than half of patients with thalassaemia major treated with magnetic resonance imaging. J Cardiovasc Magn Reson 1999; 1: 7–21.
desferrioxamine long term. Oral deferiprone seems to be 13 Sephton-Smith R. Iron excretion in thalassaemia major after
more effective than subcutaneous desferrioxamine in administration of chelating agents. BMJ 1962; 2: 1577–80.
removing iron from the myocardium, but a larger 14 Barry M, Flynn D, Letsky E, Risdon RA. Long term chelation therapy
in thalassaemia major: effect on liver iron concentration, liver histology
prospective trial is needed to confirm our results. Since and clinical progress. BMJ 1974; 2: 16–20.
heart failure remains the most frequent cause of death in 15 Brittenham GM, Griffith PM, Nienhuis AW, et al. Efficacy of
thalassaemia, the effectiveness of iron-chelation therapy desferrioxamine in preventing complication of iron overload in
should be assessed by monitoring of both cardiac iron and patients with thalassaemia major. N Engl J Med 1994; 331: 567.
liver iron content, and these measurements can be made 16 Modell B, Letsky EA, Flynn DM, Peto R, Weatherall DJ. Survival and
desferrioxamine in thalassaemia major. BMJ 1982; 284: 1081.
by the T2* magnetic-resonance technique. 17 Wolfe L, Olivieri N, Sallan D, et al. Prevention of cardiac disease by
subcutaneous deferoxamine in patients with thalassaemia major.
Contributors N Engl J Med 1985; 312: 1600–03.
S Holden and E Prescott managed the cardiology and haematology 18 Karnon J, Zeuner D, Brown J, Ades AE, Wonke B, Modell B.
clinics, co-designed the study, and entered patient data. B Wonke and Lifetime treatment costs of beta-thalassaemia major.
J M Walker managed the patients in clinics, set up the databases, Clin Lab Haematol 1999; 21: 377–85.
co-designed the study, and did the research investigations. L Anderson 19 Arden GB, Wonke B, Kennedy C, Huehns ER. Ocular changes in
co-designed the study, did the CMR scans, and coordinated the study, patients undergoing long-term desferrioxamine treatment.
and co-wrote the paper. D Pennell co-designed the study, managed the Br J Ophthalmol 1984; 68: 873–77.
research, and co-wrote the paper.
20 Olivieri NF, Buncic JR, Chew E, et al. Visual and auditory
neurotoxicity in patients receiving subcutaneous deferoxamine
Conflict of interest statement infusions. N Engl J Med 1986; 314: 869–73.
B Wonke received £2000 from Apotex (manufacturers of deferiprone) for 21 Brill PW, Winchester P, Giardina PJ, Cunningham-Rundles S.
speaking at a conference. J M Walker received £1500 from Novartis Deferoxamine-induced bone dysplasia in patients with thalassemia
(manufacturers of desferrioxamine) as a research equipment grant. The major. AJR Am J Roentgenol 1991; 156: 561–65.
other authors have no conflict of interest to declare. 22 De Virgiliis S, Congia M, Frau F, et al. Deferoxamine-induced growth
retardation in patients with thalassemia major. J Pediatr 1988; 113:
Acknowledgments 661–69.
LJA was supported by a British Heart Foundation Junior Fellowship 23 Porter JB. A risk-benefit assessment of iron-chelation therapy.
Grant (FS/98064). This work was also supported by the Wellcome Trust, Drug Saf 1997; 17: 407–21.
and CORDA, the heart charity. 24 Olivieri NF, Nathan DG, MacMillan JH, et al. Survival in medically
treated patients with homozygous beta-thalassemia. N Engl J Med
1994; 331: 574–78.
25 Tondury P, Kontoghiorghes GJ, Ridolfi-Luthy A, et al.
References
L1 (1,2-dimethyl-3-hydroxypyrid-4-one) for oral iron chelation in
1 Zurlo MG, De Stefano P, Borgna-Pignatti C, et al. Survival and patients with beta-thalassaemia major. Br J Haematol 1990; 76: 550–53.
causes of death in thalassaemia major. Lancet 1989; 2: 27–30. 26 Al-Refaie FN, Wonke B, Hoffbrand AV, Wickens DG, Nortey P,
2 Modell B, Khan M, Darlison M. Survival in beta thalassaemia major Kontoghiorghes GJ. Efficacy and possible adverse effects of the oral
in the UK: data from the UK Thalassaemia Register. Lancet 2000; iron chelator 1,2-dimethyl-3-hydroxypyrid-4-one (L1) in thalassemia
355: 2051–52. major. Blood 1992; 80: 593–99.
3 Weatherall DJ, Pippard MJ, Callender ST. Iron loading in 27 Anderson L, Porter JB, Wonke B, et al. Cardiac iron deposition is not
thalassemia: five years with the pump. N Engl J Med 1983; 308: predicted by conventional markers of iron overload in homozygous
456–58. beta-thalassaemia. Blood 2000; 96 (suppl I): 606a (abstr).
4 Olivieri NF, Brittenham GM, Matsui D, et al. Iron-chelation therapy 28 Shalev O, Repka T, Goldfarb A, et al. Deferiprone (L1) chelates
with oral deferiprone in patients with thalassemia major. N Engl J Med pathologic iron deposits from membranes of intact thalassemic and
1995; 332: 918–22. sickle red blood cells both in vitro and in vivo. Blood 1995; 86:
5 Kontoghiorghes GJ, Bartlett AN, Hoffbrand AV, et al. Long-term trial 2008–13.
with the oral iron chelator 1,2-dimethyl-3-hydroxypyrid-4-one (L1): 29 de Franceschi L, Shalev O, Piga A, et al. Deferiprone therapy in
1, iron chelation and metabolic studies. Br J Haematol 1990; 76: homozygous human beta-thalassemia removes erythrocyte membrane
295–300. free iron and reduces KCl cotransport activity. J Lab Clin Med 1999;
6 Agarwal MB, Gupte SS, Viswanathan C, et al. Long-term assessment 133: 64–69.
of efficacy and safety of L1, an oral iron chelator, in transfusion 30 Porter JB. Development of primary hepatocyte cultures for the
dependent thalassaemia: Indian trial. Br J Haematol 1992; 82: 460–66. preclinical evaluation of iron chelators. MD Thesis, University of
7 Olivieri NF, Brittenham GM, McLaren CE, et al. Long-term safety Cambridge, 1999.

520 THE LANCET • Vol 360 • August 17, 2002 • [Link]

For personal use. Only reproduce with permission from The Lancet Publishing Group.