Magnetic Resonance Materials in Biology, Physics and Medicine 6 (1998) 7712
`
`
`
`MAGMA
`tic Resonance Materials in
`Physics, Biology and Medicine
`
`
`T2 relaxation time study of iron overload in b-thalassemia
`
`S.I. Mavrogeni a-*, E.D. Gotsis 1’, V. Markussis b, N. Tsekos C, C. Politis d, E. Vretou d,
`D. Kremastinos i‘
`
`" Onassis Cardiac Surgery Center, 50 Experuu Street, 175-6] P.Fa/ir0. Athens. Greece
`b Ent'ephalor Institute, Athens, Greece
`C University ofA/Iinnesota.
`iMitmeapoliS. USA
`“ Drakepauiirm Hospital, Athens. Greece
`
`Received 25 September 1997; received in revised form l5 December 1997; accepted 21 January 1998
`
`Abstract
`
`Myocardial iron deposition occurs as a result of blood transfusion therapy in b-thalassemia major patients. Since this deposition
`causes various cardiac complications, it is of interest to assess the iron content of the myocardium in relation to the clinical picture
`of the patients. Two different MR1 indices were used to achieve this purpose: the T2 relaxation time and the heart/skeletal muscle
`signal intensity ratio. ECG gated spin echo images were obtained from 54 adult thalassemic patients, with a mean age of 26
`(18—44) years, at TE = 22 ms and 60 ms, using a 1.5 T system. Patients were divided into 2 groups (A and B), according to their
`serum ferritin lcvcls (> or< 2000 ng ml 1). Results were compared with nine controls, with a mean age of 25 (18e43) years.
`Heart T2 relaxation time in controls (44.3i3.5 ms) was higher than in group A (29.9:57 ms, P<0.001) and group B
`(33.4 i 6.8 ms, P < 0.01). T2 was measurable in 66% of group A and 83% of group B patients. The heart/muscle signal intensity
`ratio in group A (0.45 i 0.27) was lower than in group B (0.82 i 0.33. P < 0.001) and the controls (1.15 i 0.20, P < 0.001). The
`heart/muscle signal intensity ratio was measurable in 94“ 0 of the patients and demonstrated an inverse relationship with the serum
`ferritin levels (r : — 0.52, P < 0.01). This study indicates that the heart/muscle ratio is a sensitive index of iron overload and it
`can be measured in the majority of patients, irrespective of tissue iron concentration, thereby offering an advantage over the use
`of T2 relaxation time. © 1998 Elsevier Science B.V. All rights reserved.
`
`Keywords: b-Thalassemia; Heart; T2 relaxation time
`
`1. Introduction
`
`b-Thalassemia major is a genetic hemoglobinopathy
`which is especially prevalent
`in Mediterranean coun-
`tries. It is characterised by various degrees of ineffective
`hemopoiesis and intramedullary hemolysis. Current
`therapy includes regular blood transfusions with simul-
`taneous iron Chelatjon therapy, mainly employing des-
`ferrioxamine,
`in
`order
`to
`avoid
`secondary
`hemosiderosis. Bone marrow transplantation Offers an
`effective alternative to constant blood transfusions,
`
`
`,_,_
`* Corresponding author. Tel: + 30
`9882797; e-mail: som@aias.net.gr
`
`1 9882797;
`
`fax: + 30 l
`
`to be
`where the circumstances are appropriate for it
`performed. Gene therapy offers a bright outlook to the
`future treatment and annulment of the disease and its
`complications. Despite the benefits 0f iron-chelation
`therapy, iron deposition occurs in various organs, as a
`result of inadvertent iron overloading from the trans—
`fused blOOd- Iron deposition in the heart and other
`organs is the causative factor of the main complications
`in b-thalassemia major [l~5]. Excessive iron is primar-
`ily retained in the reticuloendothelial system and, when
`the capacity Of this system is exceeded, secondary depo-
`sition in parenchymal organs follows [6].
`The estimation of iron stores in each individual or-
`.
`.
`.
`.
`gan and the total 1ron burden in b-thalassemic patlents
`
`1352—8661/98/$ — see front matter © 1998 Elsevier Science B.Vi All rights reserved.
`PIISl352-8661(98)00003-9
`
`Apotex Tech.
`Ex. 2033
`
`Apotex Tech.
`Ex. 2033
`
`

`

`8
`
`SJ. Mavrogeni el al. / Magnetic Resonance Materials in Biology, Physicr and Medicine 6 (1998) 7712
`
`is useful in evaluating the efficacy of chelation therapy.
`This is a difficult task since hemosiderin and ferritin,
`the iron storage proteins, are mainly intracellular. On
`the other hand, serum ferritin levels are highly corre-
`lated with the amount of iron deposited in the tissues,
`and are hence considered a very sensitive index [7].
`However, the levels may be affected by factors such as
`fever,
`inflammation, etc.
`[8,9]. At present.
`the most
`accurate way to estimate the total body iron deposition
`is the direct measurement of iron content in liver biopsy
`specimens [10], an invasive procedure which cannot be
`applied during routine follow—up, and which does not
`offer any information concerning iron deposition in the
`heart.
`
`The ability of stored intracellular tissue iron to en-
`hance magnetic susceptibility provides the basis by
`which it can be detected by magnetic resonance imaging
`(MRI). Recent studies in experimental animals have
`shown that the T2 relaxation time is linearly correlated
`with the total iron content for all organs, including the
`heart [1 1]. However, T2 relaxation time calculation may
`prove to be impossible in severely iron-overloaded pa—
`tients, due to a very low signal intensity which is similar
`to background noise [12].
`The aim of this study was to assess heart, liver and
`skeletal muscle iron content
`in b-thalassemia major
`patients using two different indices (T2 relaxation time
`of the heart, liver and skeletal muscles, and the heart/
`muscle and liver/muscle MRI signal intensity (SI) ra-
`tios), and to correlate these parameters with the
`corresponding serum ferritin levels.
`
`2. Patients and methods
`
`2.]. Patient population (Table I).
`
`Fifty four thalassemic patients, 24 males and 30
`females, with a mean age of 26 (18744) years, were
`studied. Thirty nine of them had no symptoms of heart
`failure (NYHA 1) whereas 15 were diagnosed with
`
`Table 1
`
`Study group stratification (values as x i SD or range)
`
`
`
`
`
`Group BGroup A Group C
`
`Subjects (n)
`Age (yrs)
`Ferritin (ng ml")
`Blood units transfused
`(n)
`
`13.3 i 4.4 12.7 i 5Chelation (years) 7
`
`30
`26 (18-—36)
`4102 i 1541
`1025 i 274
`
`9
`24
`25 (18743)
`26 (18744)
`11504—381“ u
`965 i 429
`7
`
`
`
`
`
`Group A: high ferritin population.
`Group B: low ferritin population.
`Group C: controls.
`“ P< 0.01.
`
`heart failure (NYHA 11—111). All patients were iron-
`chelated regularly and 40 of them were splenectomized.
`A serum ferritin level below 2000 ng ml’l was consid-
`ered to be the target value for a patient successfully
`treated with deferoxamine [27]. Patients were stratified
`into two groups, according to their average serum
`ferritin values ( > or < 2000 ng ml “ I) over the previous
`5 years: Group A (n = 30) with mean ferritin levels of
`4150 i 1653 ng ml“ and group B (n = 24) with mean
`ferritin levels of 1240 i 366 ng ml’ 1. The total number
`of transfusions and years of iron-chelation therapy were
`similar in both groups. The results from thalassemic
`patients were compared with those of nine normal
`volunteers (Group C; controls) with a mean age of 25
`(18443) years. An informed consent was obtained from
`all participants.
`
`2.2. MRI techniques
`
`studies were conducted using a 1.5 T
`All MRI
`Siemens system (Magnetom, SP), with an ECG gated
`spin echo sequence (acquisition matrix=256 x 256,
`FOV = 40 x 40 cm2, number of slices, eight, TR equal
`to RR interval, and TE = 22 and 60 ms). The oblique
`orientation of the imaging slices was determined from
`scout coronal anatomical images (parameters as above
`with TE 2 22 ms), in order to depict a short axis view
`of the heart. These images included the liver and a
`section of skeletal muscle (latissimus dorsi). The signal
`intensities of the heart (81”),
`the liver (SIL) and the
`skeletal muscle (SIM) were determined using circular
`regions of interest (ROIs). Two indices were employed:
`(a) The T2 relaxation time of the heart (T2“),
`liver
`(T2“) and skeletal muscle (TZM), calculated from the
`images collected at
`the two TE values [13];
`(b) The
`heart/muscle (SIB/SIM) and liver/muscle (SIL/SIM) sig-
`nal intensity ratios at TE = 22 ms. The skeletal muscle
`was used as an internal standard to form this ratio since
`iron deposition is minimal in skeletal muscle [12].
`
`2.3. Statistical analysis
`
`The results are expressed as x: SD and were com-
`pared by means of the unpaired two—tailed Student 1
`test. The Chi square test was used to compare percent-
`ages between groups. Correlations between various
`parameters were sought by employing Pearson’s corre-
`lation coefficient. Statistical significance was considered
`for P < 0.05.
`
`3. Results
`
`The T2 relaxation time of the left ventricle of the
`heart (TZH) in the high ferritin group (group A) was
`29.9 i 5.7 ms. In the low ferritin group (group B), the
`Apotex Tech.
`Ex. 2033
`
`Apotex Tech.
`Ex. 2033
`
`

`

`SJ. Mavrogem' er a1. /Magnetic Resonance Materials in Biology, Physics and Medicine 6 (I998) 7~12
`
`9
`
`Table 2
`T2 relaxation time in the study groups (values expressed as x i SD)
`
`Group A
`
`Group B
`
`Group C
`
`Subjects (n)
`T2 heart (ms)
`r2 liver (ms)
`
`30
`299 i 5.7 (20)
`24.3 t 1.8 (3)
`
`T2 muscle
`(ms)
`
`28.4 i 2.7 (30)
`
`24
`33.4 J_r6.8 (20)
`29.0 i 10.3
`(12)
`29.2 i 2.4 (24)
`
`9
`44.3 i 3.5 (9)"b
`30.0 i 7.7 (9)
`
`30.0 i 0.9 (9)
`
`The numbers in the brackets denote the number of subjects in whom
`the T2 relaxation time was measurable.
`Group A: high ferritin population.
`Group B: low ferritin population.
`Group C: controls.
`“ P<0.001, compared to group A.
`b P<0.()l. compared to group B.
`
`T2 relaxation time was 33.45 + 6.8 ms and in the
`
`the time was 44.3 i3.5 ms. The
`controls (group C),
`T2H in group C was higher compared both to group A
`(P<0.001) and group B (P<0.01). There was no
`statistically significant difference between groups A and
`B (Table 2).
`intensity of the myocardium (SIH) was
`The signal
`only measurable at both echo times in 20/30 (66.7%) of
`the group A patients and in 20/24 (83.3%) of the group
`B patients (PzNS) (Fig. 1(a, b)Fig. 2(a, b)). As a result
`the T2H could only be calculated in these patients.
`Seven additional group A patients and the remaining
`four group B patients had measurable signal intensities
`at TE=22 ms but not at TE=60 ms. The signal to
`noise ratio was too low at both echo times to permit T2
`and signal
`intensity ratio calculation in the final
`re—
`maining three group A patients. Twelve group A pa-
`tients
`(40%) and three group B patients
`(12.5%)
`manifested heart failure (NYHA 117111). Out of these
`patients, T2H was measurable only in four out of 12 of
`the group A population and in one out of three patients
`belonging to group B.
`There was no difference in the T2 relaxation time of
`
`the liver (T2L) between groups A, B and C. The signal
`intensities of the liver (SIL) at both echo times and,
`consequently, the calculated T2L, were only measurable
`in three out of 30 (10%) group A patients and 12 out of
`24 (50%) group B patients (P<0.01, chi square test).
`Twelve additional patients of group A and nine group
`B subjects had measurable signal intensities of the liver
`at TE = 22 ms, but not at TE 2 60 ms.
`The T2 relaxation time of the skeletal muscle (T2M)
`was measurable in all subjects and was similar in the
`three groups.
`In group A, no correlation was docu—
`mented between T2“, T2L or TZM and ferritin levels,
`nor between the number of transfusions or iron-chela—
`
`tion time. In group B, an inverse correlation was found
`between the average ferritin levels and T2H (r =2 — 0.49,
`P < 0.05).
`
`The heart/muscle signal intensity ratio (SIH/SIM) in
`group A (0.45 i 0.27) was lower
`than in group B
`(0.82 i 0.33, P < 0.001) and group C (1.15 i 0.2, P <
`0.001) (Table 3). Similarly, the liver/muscle signal inten-
`sity ratio (SIL/SIM) in group A (0.17 i 0.13) was lower
`than in group B (0.37i0.3l, P<0.01) and group C
`(1.28 -l_~ 0.16, P < 0.01). Patients with heart failure, as a
`subgroup, had a lower SIH/SIM ratio (0.33 :1; 0.19) com—
`pared to the mean group A (P<0.001) and group B
`(P < 0.001) values. Ferritin levels were inversely corre-
`lated with both SIH/SIM (r = ~—0.52, P<0.01) (Fig. 3)
`and SIL/SIM ratios (r = —— 0.41, P < 0.01). In group B, a
`positive correlation (r= 0.55, P<0.01) was found be-
`tween the SIB/SIM ratio and T2H values.
`
`
`Mwmwfi“ . “a ,.
`
`t: 5r
`
`l. (a and b) Short axis view of the heart ofa thalassemic patient
`Fig.
`with mild iron overload, using TE = 22 ms (3) and TE = 60 ms (b).
`Apotex Tech.
`Ex. 2033
`
`Apotex Tech.
`Ex. 2033
`
`

`

`10
`
`5.1. Mavrogeni el 01. /Magnetic Resonance Materials in Biology, Physics and Medicine 6 (1998) 7712
`
`
`
`Fig. 2. (a and b) Short axis view of the heart of a thalassemic patient
`with severe iron overload and impaired LV function, using TE = 22
`ms (a) and TE 2 60 ms (b). In the latter image, the myocardial signal
`intensity is equal to the background noise.
`
`4. Discussion
`
`The presence of iron affects the tissue T1 and T2
`relaxation times [14,15]. However,
`the effect on T1
`relaxation time is not as significant
`[15], and most
`studies use the T2 time for characterization of iron
`
`deposition. This effect is proportional to the concentra-
`tion of iron in the tissue and depends on the applied
`magnetic field (B0) [16]. This concept was employed to
`evaluate a group of adult thalassemic patients.
`In this study, the heart and liver T2 relaxation times
`were significantly reduced, with higher iron concentra—
`tions resulting in lower T2 values. The signal intensities
`(SI) of different tissues demonstrated similar trends. No
`
`significant difference was found in the heart T2 relax-
`ation time between the high and low serum ferritin
`groups. Serum ferritin levels were only inversely corre-
`lated to heart T2 relaxation times (T2H)
`in the low
`ferritin group. These observations result from the fact
`that the T2 of several patients in the high ferritin group
`could not be measured due to low signal to noise ratio.
`On the contrary, signal
`intensity measurements were
`feasible in all patients,
`irrespective of iron burden.
`Significant differences were recorded between the two
`patient groups, and an inverse correlation was found
`between heart/muscle and liver/muscle SI
`ratio and
`serum ferritin levels
`in the thalassemic population.
`Since there is a significant error involved in the calcula-
`tion of the T2 relaxation time when the tissue signal
`intensity (SI) is low, heart SI (normalized in proportion
`to skeletal muscle employed as the control standard) is
`a more appropriate index for correlation with iron
`deposits.
`serious complication of b-tha-
`a
`failure,
`Heart
`lassemia, was present in both ferritin groups, although
`it was more prevalent in the severely iron—overloaded
`group. When heart failure patients from groups A and
`B were analysed as a separate sub-group, the T2 relax-
`ation time of the heart was measurable only in one
`third of the population, and the heart/muscle signal
`intensity ratio was even lower than the values observed
`in the high ferritin group. Similarly, the T2 relaxation
`time of the liver was measurable in half of the patients
`in the low ferritin group, whereas in the majority of the
`patients
`in the high ferritin group, measurements
`proved to be impossible. When the liver/muscle signal
`intensity ratio was employed, however, measurements
`in all patients were achieved and differences were man—
`ifested between the high and low ferritin groups, and
`the control subjects.
`
`Table 3
`
`Signal intensity ratios (SI) in the study groups (values expressed as
`x i SD)
`
`
`
` Group B Group C Group A
`
`Subjects (n)
`SIM/SIM
`
`SIL/SIM
`
`
`24
`30
`0.45 i 0.27 (27) 0.82 i 0.33
`(24)a
`0.17i0,13 (15) 0.37i0.3l
`(21)“
`
`9
`1.15 i 0.20 (9)b
`
`1.28i0.l6 (9)”
`
`
`intensities measured at TE=22 ms (H: heart, M: skeletal
`Signal
`muscle, L: liver). The numbers in the brackets denote the number of
`subjects in whom the signal intensity ratios were measurable.
`Group A: high ferritin population.
`Group B: low ferritin population.
`Group C: controls.
`“ P < 0.001.
`
`hP<0.01, compared to group A.
`C P<0.0()l, compared to groups A and B.
`
`Apotex Tech.
`Ex. 2033
`
`Apotex Tech.
`Ex. 2033
`
`

`

`SJ. Mavrogeni er al. / Magnetic Resonance Materials in Biology, Physics and Medicine 6 (1998) 77712
`
`11
`
`1.60 '7
`
`1.40 e»
`
`1.20 +
`
`1.00 +
`
`0.80 ~—
`
`
`
`heart/muscleSIratio
`
`0.60 ~
`
`0.40 T
`
`0.20 r
`
`.
`
`‘
`
`‘o
`
`0
`
`o
`
`o
`
`.
`
`9.
`
`.
`
`.
`
`.
`
`.
`
`o
`
` l l l i» l 0.00
`
`
`
`I
`h
`i
`l
`
`0
`1000
`2000
`3000
`4000
`5000
`6000
`7000
`8000
`9000
`10000
`
`Ferrltln (nglml)
`
`Inverse relationship between the heart/muscle signal
`Fig. 3.
`(r r v (3:5)}; F<933Hr
`
`intensity (SI) ratios and the average ferritin levels of the thalassemic patients
`
`Serum ferritin is currrently considered the most accu—
`rate index of body iron content. Variations in serum
`ferritin mainly correspond to changes in the reticuloen-
`dothelial system storage iron levels and not to changes
`in the parenchymal
`iron content
`[17]. Liver disease,
`inflammation.
`infection and assay mmlicafiems
`[8,9,18] have been reported to influence measurements.
`In this study, in an attempt to overcome these limita-
`tions,
`the average ferritin levels of the preceding five
`years were employed.
`
`There is a controversy over the precise relationship
`between the quantity of iron present in the myocardium
`and the degree of heart dysfunction. Correlations be—
`tween heart T2 relaxation time and cardiac biopsy are
`absent since right ventricular biopsies are subject
`to
`serious sampling errors and iron deposition is patchy
`(not uniform) [)9]. A good contention was found be-
`tween iron content and T2 relaxation time only when
`whole rat hearts were sampled [11]. Under these cir—
`cumstances, MRI appears to be useful
`in addressing
`this issue,
`
`Recently, studies of thalassemic patients at 0.5 T
`intliicateb that cardiac comphhcabons are rehateb to )0»!
`T2 times [20]. The heart/muscle signal
`intensity ratio
`was also found to be sensitive to iron content
`in a
`
`group of patients treated by mmtiple mood transfusions
`[12]: and‘a t‘rend‘tbr worse tieart tunctron was ev1dent
`in the more heavily-transfused patients, even though
`there was no correbatbon with serum 331x215]: Dave-)5.
`)2?
`
`[21], employing a new in vivo,
`addition, Zaino et al.
`non-invasive method for measuring iron utilizing nu-
`clear resonance scattering (NRS), detected symptomatic
`cardiac disease in the patients with the more elevated
`cardiac iron levels.
`
`Liver iron load, as studied by MRI, has been consid-
`ered as an alternative to liver biopsy for the assessment
`
`of total iron burden [22]. In vitro measurements of T2
`relaxation time of liver samples, from iron-overloaded
`rats [l4] and spleen samples from thalassemic patients
`[15], demonstrated a linear correlation between relax-
`ation rate (1/T2) and iron content. The reticuloendothe-
`‘nal system preferentially accumulates iron from the
`breakdown of transfused red blood cells before iron
`
`deposition occurs in other parenchymal organs such as
`the heart. The data presented here are consistent with
`the study by Buja et al.
`[23], which indicates that
`cardiac iron deposition is accompanied by excessive
`hepatic iron deposits.
`Although the T2 relaxation time is generally a reli-
`able index of iron deposition, being well-correlated with
`liver biopsy information,
`it may be unmeasurable in
`severely iron—overloaded patients, due to low signal
`intensity equal to background noise. For this reason,
`some authors consider the application of the T2 relax-
`ation time to assess heart
`iron deposition as being
`inaccurate [12], The use of a 0.5 T machine, where the
`magnetic susceptibility phenomenon is less prominent
`[16],
`the use of shorter echo times (TE) [15] or the
`application of MR spectroscopy [22],, could be more
`promising in the study of severe iron-overloaded pa—
`tients. Other non-invasive techniques such as dual en-
`ergy computed tomography [24,251,
`superconducted
`quantum. interference ram: (\SQULDL agaplirzatirmi [310‘
`and nuclear resonance scattering (NRS) can overcome
`this limitation, but are not available on a wide scale.
`This study suggests that M'Rl' could prove to be an
`effective non—invasive method in performing tissue
`characterization. MRI,
`in particular, provides
`the
`means for the simultaneous examination of heart iron
`deposition and heart
`function. Moreover, measure—
`ments are quantifiable, enabling the repeated examina-
`tions of thalassemic patients and evaluation of the
`iron-che'la‘tion therapy efficacy.
`Apotex Tech
`Ex. 2033
`
`Apotex Tech.
`Ex. 2033
`
`

`

`12
`
`SJ. Mavrogeni et al. / Magnetic Resonance Materials in Biology, Physics and Medicine 6 (1998) 7712
`
`In conclusion, T2 relaxation time is a very useful
`index in the monitoring of tissue iron deposition. How-
`ever,
`it
`is not always measurable using commercially
`available systems. As a consequence, the use of signal
`intensity ratio appears to be a valuable alternative for
`patient screening and follow-up.
`
`Acknowledgements
`
`We are indebted to Sonja Phillips, BSc. MSc., for
`proofing the manuscript.
`
`References
`
`[1] Kattamis C, Touliatos N, Haidas S, Matsaniotis N. Growth of
`children with thalassemia: effect of different
`transfusion regi-
`mens. Arch Dis Childhood 1970;45:502—6.
`[2] Propper RD, Cooper B, Rufo RR, et a]. Continuous subcuta-
`neous administration of deferoxamine in patients with iron over—
`load. N Engl J Med 1977;297:418723.
`lron state and
`[3] Aldouri MA, Wonke B, Hoffbrand AV, et a1.
`hepatic disease in patients with thalassemia major treated with
`long—term subcutaneous
`desferrioxamine.
`J Clin
`Pathol
`1987;40:135379.
`[4] Engle MA. Cardiac involvement in Cooley's anemia. Ann NY
`Acad Sci 1969;] 1926947702.
`[5] Kremastinos DT, Tsiapras D, Tsetsos G, Rentoukas E. Vretou
`H, Toutouzas P. Left ventricular diastolic Doppler characteris-
`tics in bthalassemia major. Circulation 1993;88:1127 V35.
`[6] Jacobs A. The pathology of iron overload. 1n: Jacobs A, Wor—
`wood M, editors. Iron in biochemistry and medicine, 11. New
`York: Academic Press, 19802439752.
`[7] Jacobs A, Miller F, Worwood M, Beamisch R, Wardrop C.
`Ferritin in the serum of normal subjects with iron def~ciency
`and iron overload. Br Med J 1972;XX:206721].
`[8] Crosby WH. Serum ferritin fails to indicate hemochromatosis:
`nothing gold can stay. N Engl J Med 1976;294:33374.
`[9] Green R, Watson LR, Saab GA, Crosby WH. ‘Normal’ serum
`ferritin: a caution. Blood 1977;50:54577.
`[10] Ridson RA, Barry M, Flynn DN. Transfusiona] iron overload:
`the relationship between tissue iron concentration and hepatic
`fibrosis in thalassemia. J Pathol 1975;] 16:83795.
`[11] Liu P, Henkelman M, Joshi J, Waien S, Butany J, Olivieri N.
`Quantification of myocardial
`tissue iron contents using NMR
`relaxation. Validation in a novel murine thalassemia model.
`JACC 1992;19(3):187A (abstract 935—976).
`
`[12] Johnston DL, Rice L, Vick W, Hedrick TD, Rokey R. Assess-
`ment of tissue iron overload by nuclear magnetic resonance
`imaging. Am J Med 1989;87:4077.
`[l3] Sechtem U, Higgins C. Relaxation measurements and contrast
`media.
`In: Underwood R, Firmin D, editors. Magnetic Reso~
`nance of
`the
`cardiovascular
`system. London: Blackwell,
`199128977101.
`
`[14] Stark DD, Moseley ME, Bacon BR, et a]. Magnetic resonance
`imaging and spectroscopy of hepatic iron overload. Radiology
`1985;154:137742.
`[15] Gomori JM, Grossman R]. Drott HR. MR relaxation times and
`iron content of thalassemic spleens: an in vitro study. AJR
`1988;150:567—79.
`[16] Vymazal J, Brooks RA, Zak O, McRill C, Shen C, Di Chiro G.
`T] and T2 of ferritin at different field strengths: Effect on MRI.
`Magn Res Med 1992;27:368774.
`[17] Gomori JM, Horev G, Tamary H, et a1. Hepatic iron overload:
`Quantitative MR imaging. Radiology 1991;179:36779.
`[18] Worwood M, Cragg SJ, Jacobs A, McLaren C, Ricketts C,
`Economidou J. Binding of serum ferritin to concanavallin A:
`Patients with homozygous betaithalassemia and transfusional
`iron overload. Br J Haematol 19801462409716.
`[l9] Olson LJ, Edwards WD, McCall JT, Ilstrup DM, Gersh BJ.
`Cardiac iron deposition in idiopathic hemochromatosis: histo»
`logic and analytic assessment of 14 hearts from autopsy. JACC
`1987;10:1239743.
`[20] Liu P, Olivieri N, Sul1ivan H, Henkelman M. Magnetic reso—
`nance imaging of cardiac iron overload in thalassemia: Detection
`of iron content and association with cardiac complications.
`JACC 1993:21(2):409A (abstract 9487119).
`[21] Zaino EC, Wielopolski L.
`in vivo measurement of hepatic and
`cardiac iron in Cooleys anemia by nuclear resonance scattering.
`Ann NYAcad Sci 1990;612:57376.
`[22] Dixon RM. Styles P. Al-Refaie F, et a1. Assessment of hepatic
`iron overload in thalassemic patients by magnetic resonance
`spectroscopy. Hepatology 1994;19:904—710.
`[23] Buja LM, Roberts WC. Iron in the heart: Etiology and clinical
`significance. Am J Med 1971;51:209721.
`[24] Houang MT, Skalicka A, Arozena X. Huehns ER, Shaw DG.
`Correlation between computed tomographic values and liver
`iron content in thalassemia major with iron overload. Lancet
`1979;]:1322733.
`[25] Olivieri NF, Grisaru D. Daneman A, Martin DJ, Rose V,
`Freedman MH. Computed tomography scanning of the liver to
`determine eff~cacy of iron chelation therapy in thalassemia
`major. J Pediatrics 1989;114:427—30.
`[26] Brittenham G, Griffith PM, Nienhuis AW, et al. Efficacy of
`deferoxamine in preventing complications of iron overload in
`patients with thalassemia major. N Engl J Med 1994;331:5677?
`73.
`
`[27] Olivieri NF, Nathan DG, MacMillan JH, Wayne AS, Liu PP,
`McGee A. et a]. Survival
`in medically treated patients with
`homozygous b—thalassemia. N Engl J Med 1994;331:574—8.
`
`Apotex Tech.
`Ex. 2033
`
`Apotex Tech.
`Ex. 2033
`
`