A Distante and C VecchioF Lattanzi, P Bellotti, E Picano, F Chiarella, A Mazzarisi, C Melevendi, G Forni, L Landini,

and iron overloadQuantitative ultrasonic analysis of myocardium in patients with thalassemia major

1524-4539 Copyright © 1993 American Heart Association. All rights reserved. Print ISSN: 0009-7322. Online ISSN:Circulation is published by the American Heart Association. 7272 Greenville Avenue, Dallas, TX 72514

1993, 87:748-754Circulation 

http://circ.ahajournals.org/content/87/3/748the World Wide Web at:

The online version of this article, along with updated information and services, is located on

http://www.lww.com/reprintsReprints: Information about reprints can be found online at   journalpermissions@lww.com410-528-8550. E-mail: Health, 351 West Camden Street, Baltimore, MD 21202-2436. Phone: 410-528-4050. Fax: Permissions: Permissions & Rights Desk, Lippincott Williams & Wilkins, a division of Wolters Kluwer  http://circ.ahajournals.org//subscriptions/Subscriptions: Information about subscribing to Circulation is online at

by guest on July 13, 2011http://circ.ahajournals.org/Downloaded from

748

Quantitative Ultrasonic Analysis of Myocardiumin Patients With Thalassemia Major

and Iron OverloadFabio Lattanzi, MD; Paolo Bellotti, MD; Eugenio Picano, MD; Francesco Chiarella, MD;Alessandro Mazzarisi; Caterina Melevendi, MD; Gianluca Forni, MD; Luigi Landini, PhD;

Alessandro Distante, MD; and Carlo Vecchio, MD

Background. Patients with f-thalassemia major present with severe anemia and need continuoustransfusion therapy. The consequent iron overload leads to hemochromatosis. Initial cardiac dysfunctionhas been documented even in thalassemics without clinical manifestations of heart failure as well as byconventional echocardiographic-Doppler techniques. The purpose of this study was to assess the acousticquantitative properties of myocardium in patients with iron overload.Methods and Results. Thirty-eight patients with (3-thalassemia major, without clinical signs of cardiac

failure, and 20 age- and sex-matched young controls were studied by echocardiography. An on-lineanalysis of the ultrasonic radiofrequency signal was performed to obtain quantitative operator-indepen-dent measurements of the integrated backscatter (IB) signal of the ventricular septum and the posteriorwall. The integrated values of the radiofrequency signal were normalized for the pericardial interface andexpressed in percent (IB%). Thalassemic patients had been receiving transfusion therapy for 16±5 yearsand had received 313±+138 transfusion units; they all had received chelation treatment (desferroxiamine)for 9±2 years. Patients and controls showed comparable values of echocardiographically assessed percentfractional shortening (32±3% versus 36±4%, p=NS), whereas thalassemics showed higher values of leftventricular mass index (118±30 versus 98± 15 g/m2,p<0.05). The IB% values were higher in patients withthalassemia major than in controls for both septum (35±14% versus 21+6%, p<0.001) and posterior wall(16±6% versus 11±3%, p<0.001). In thalassemic patients, no significant correlation was found betweenthe septum IB% value and hematological parameters, such as the total number of transfusions (r=0.2,p=NS) or the mean ferritin value (r=0.1, p=NS). No significant correlation was also found between theseptum IB% value and the echocardiographically assessed left ventricular mass index (r=0.2, p=NS).

Conclusions. These data demonstrate that myocardial reflectivity is abnormally increased in patientswith thalassemia major under transfusion treatment, probably due to myocardial iron deposits and/orsecondary structural changes. These quantitatively assessed abnormalities in regional reflectivity can bedetected when conventional echocardiographic parameters of systolic left ventricular function areundistinguishable from normal controls. (Circulation 1993;87:748-754)KEY WORDs * echocardiography * tissue * thalassemia * hemochromatosis

P atients with f8-thalassemia major present withsevere anemia and need continuous transfusiontherapy. The consequent iron overload, also

caused by extravasal hemolysis and increased ironabsorption, leads to hemochromatosis.' Cardiac disor-ders related to biventricular failure are the mostfrequent cause of death in this syndrome.2 A fewstudies have documented initial cardiac dysfunctioneven in thalassemics without clinical manifestations ofheart failure,3,4 also by means of conventional echocar-diographic-Doppler techniques.5-7 However, the qual-itatively assessed echocardiographic reflectivity of the

From the Institute of Clinical Physiology, C.N.R. (F.L., E.P.,A.M., L.L., A.D.), Pisa, and the Cardiology (P.B., F.C., C.V.) andPediatric (C.M., G.F.) Divisions, Galliera Hospital, Genova, Italy.

Presented in part at the 40th Annual Scientific Session of theAmerican College of Cardiology, Atlanta, Ga., March 3-7, 1991.Address for correspondence: Dr. Fabio Lattanzi, Istituto di

Fisiologia Clinica, C.N.R., Via Paolo Savi 8, 56100 Pisa, Italy.Received July 1, 1991; revision accepted November 3, 1992.

myocardium remains unchanged in the early phase ofthe iron-overload disease and becomes clearly abnor-mal (with increased echogenicity of myocardial walls)only in a few thalassemic patients in the advancedstages of the disease (usually represented by dilatedcardiomyopathy).8

Nonconventional quantitative ultrasound methodsfor tissue characterization have recently proved to bereliable in identifying progressive degrees of histopath-ological impairment of cardiovascular structures, also inhumans.9'10 In particular, a significant correlation hasbeen found in in vitro1l-'s and in vivo models16"17between radiofrequency backscattered ultrasound sig-nal increase and the fibrotic and/or calcium content ofthe target. Moreover, abnormalities in both absoluteand cyclic systolic-to-diastolic variations of the quanti-tatively evaluated radiofrequency reflections of myocar-dial walls are present in patients with cardiac diseasesassociated with histomorphological tissue abnormali-ties, such as cardiomyopathies.17-19

by guest on July 13, 2011http://circ.ahajournals.org/Downloaded from

Lattanzi et al Quantitative Analysis of Myocardial Reflectivity in Thalassemia Major 749

TABLE 1. Demographic Data of Patients With ThalassemiaMajor and Control Subjects

Patients withthalassemia major Controls

No. of subjects 38 20Age (years) 17.7+5.1 17.6±6.6Sex (M/F) 21/17 12/8Weight (kg) 47± 14 57+12*Body surface (M2) 1.40±0.27 1.61±0.31*Systolic blood pressure (mm Hg) 110±10 116±8Diastolic blood pressure (mm Hg) 71±9 76±6

*p<0.05 vs. patients' data. Values are mean±SD.

The purpose of this study is to assess whether abnor-mal myocardial wall reflectivity is present in patientswith 83-thalassemia major who are receiving transfusiontreatment but are free from symptoms due to cardiacdysfunction by means of a simple ultrasound system foron-line quantitative evaluation of the radiofrequencybackscattered signal.

MethodsPatient Population

Forty-two patients with 83-thalassemia major (ho-mozygous form) and no clinical signs of cardiac dysfunc-tion were initially considered for the study. After con-ventional echocardiographic examination, four patientswere excluded for instrumental signs of left ventriculardysfunction (end-diastolic diameter .56 mm and/orfractional shortening <30%). The remaining 38 patientswere enrolled for conventional and quantitative ultra-sound myocardial analysis. No patients were excludedbecause of a poor-quality acoustic window. Their meanage was 18 years (age range, 7-26 years); 21 of themwere male (55%). Demographic data of patients arereported in Table 1.Each patient was receiving blood transfusions every 2-3

weeks to maintain hemoglobin levels between 10.5 and13.5 g/dL. They all were also receiving desferroxiamine aschelation therapy (25-50 mg/kg body wt s.c. infused 4-6days a week to maintain serum ferritin level below 1,300ng/mL) to prevent as well as to treat the iron toxicity.20Hematological data of the patients at the time of ultra-sound evaluation are summarized in Table 2.

Control GroupTwenty age- and sex-matched normal subjects were

also studied. They had no personal and/or family historyof cardiac or hematological disease and had normalclinical, ECG, and echocardiographic findings. Theirmean age was 17 years (age range, 6-30 years); 12 of

TABLE 2. Hematological Profile of the 38 Patients WithThalassemia Major at the Time of Echocardiographic StudyTransfusions (years) 16.2±5.0

Transfusions (total no.) 314±138Mean serum ferritin (ng/mL) 2,054+1,734Maximal serum ferritin (ng/mL) 4,031+2,677Total transfused iron burden (g) 60.2±22.8Chelation (years) 8.9±2.3

them were male (60%). Demographic data of patientsare reported in Table 1.Each patient and control underwent on the same day

a conventional two-dimensional echocardiographic andtransmitral Doppler flow evaluation and a quantitativeradiofrequency analysis.

Two-dimensional EchocardiographyM-mode and two-dimensional echocardiographic

tracings were obtained in each subject using a Hewlett-Packard 77020AI phased-array sector scanner (An-dover, Mass.) with 3.5- and 2.5-MHz transducers. Two-dimensional images were obtained in the parasternallong- and short-axis views and apical two- and four-chamber views. Left ventricular diameters and wallthicknesses were measured from the two-dimensionaltargeted M-mode echocardiographic tracings accordingto the criteria of the American Society of Echocardiog-raphy.21 Left ventricular percent fractional shorteningwas calculated as end-diastolic diameter minus end-systolic diameter divided by end-diastolic diameter andmultiplied by 100. Left ventricular mass was calculatedby the Penn convention method22 and normalized forbody surface (left ventricular mass index).The wall echoreflectivity was visually and qualita-

tively assessed by two independent observers as eithernormal or increased. In case of split decisions (whichoccurred in six cases), a third observer reviewed thestudy, and his judgment was binding.The Doppler transmitral flow velocity profile was

obtained from the apical four-chamber view. In eachsubject, four to six consecutive cardiac cycles with thehighest diastolic flow velocity as well as the best signal-to-noise ratios were chosen for analysis. The ratio be-tween the early and late peaks of flow velocity (E/A) wasmeasured in each patient. This Doppler diastolic indexhas been shown to have satisfactory reproducibility andto be an early and consistent sign of left ventricular fillingabnormalities in patients with thalassemia major.7

Ultrasonic Tissue CharacterizationThe quantitative analysis system was developed in our

institution and has been used in a different studypopulation.19An ESAOTE Biomedica SIM 3000 two-dimensional

mechanical sector scanner echocardiograph was usedfor spatial orientation of the ultrasound beam; quanti-tative analysis of ultrasonic reflectivity was performed inthe regions of interest, i.e., the ventricular septum andthe posterior wall of left ventricle. These regions werevisualized in the parasternal long-axis view. The acqui-sition of the backscattered signal was performed at enddiastole since a systematic variation in backscatter am-plitude occurs during the cardiac cycle.23A 3.5-MHz frequency transducer (focal distance, 7

cm; -3 dB focal region, 6 cm) was used.The "native" radiofrequency signal was sampled be-

fore the processing chain of the two-dimensional instru-ment. The radiofrequency signal underwent preampli-fication, bypassing the receiving circuits of theultrasonic equipment. The analog signal was fed to anamplifier, and the gain sweep of the amplifier (2-60 dB)was accomplished in 30 steps. This allows full use of theinput dynamic range of the analog-to-digital converter.Sampling was performed by a flash converter with 8 bits

by guest on July 13, 2011http://circ.ahajournals.org/Downloaded from

750 Circulation Vol 87, No 3 March 1993

FIGURE 1. Two-dimensional echocardiographic imagesfrom quantitative examination oftwo differentpatients with thalassemiamajor. On the two-dimensional display, the quantitative acquisition gate (delimited by two parallel solid white lines) is positionedon septum (first patient, left panel) and on posterior wall (second patient, right panel).

of amplitude resolution at a rate of 40 MHz. Thedigitized signal from analog-to-digital converter wasanalyzed in real time by a hardware prototype devel-oped in our electronics laboratory.24 The two-dimen-sional acquisition gate was visualized on the two-dimen-sional image so as to ensure its proper positioning(Figure 1).For analysis of the myocardium, the gate width was

kept at 3 ,usec, which corresponds to 2.35 mm (for 64points), given the velocity of ultrasound in biologicaltissues of 1.57 mm/nisec. This allowed for sampling ofthe radiofrequency signal in the mid subendocardiallayers of the myocardium, thus excluding epicardial andendocardial specular reflections. The acquisition gatewas placed immediately behind the specular echo of theendocardium (left endocardium for the septum) tominimize the transmural variations in backscatter thatare due to the position from which the signal is acquiredwithin the wall. For evaluation of the pericardial echo, a1.5 -,sec gate length was used (which corresponds to 1.2mm, for 32 points). The acquisition gate was centeredon the strongest pericardial reflections, immediatelybehind the mitral leaflets. The representative value ofreflectivity for both myocardial walls and pericardiumwas calculated as the mean of three measurements.The hardware analysis involved the measurement of

the integrated amplitude of the rectified radiofrequencysignal corresponding to the two-dimensional area se-lected from the echocardiographic image.More analytically, the two-dimensional integrated

backscatter index (2D-IB) was calculated over a tissuearea, i.e., corresponding to a (n-m) segment in depthand a (r-l) segment in lateral displacement, as follows:

12DIB- -Wj=rITB (Xj)xr-xl

where

IB (Xj) - mi=niS(xj, Yi)Yn Ym

represents the processing over the depth of the interro-gated tissue; S(xe, yj) is the sequence of the digitizedbackscattered echoes over the selected two-dimensionalarea and is expressed in millivolts.The system provides a simultaneous display of con-

ventional information together with tissue characteriza-tion parameters (the 2D-IB alphanumeric index and thelateral displacement profile averaged over the selecteddepth). Alphanumeric 2D-IB data values are trans-ferred on-line to a personal computer (model AT) forstatistical analysis.

Ultrasonic Quantitative Data AnalysisPrimary 2D-IB values are highly influenced by inter-

patient differences in chest morphology, heart structuredepth, and ultrasonic impedance. Therefore, to quanti-tatively assess the reflectivity of the interventricularseptum and posterior free wall, we used the percent2D-IB (as related to pericardial interface).

In particular, 2D-IB results for each heart structure,initially displayed in millivolts, were expressed as per-cent values, assuming the pericardial interface (fromwhich the peak echo intensity was consistently recordedin each patient) to be 100%. In this manner, theindividual pericardial signal strength was used to nor-malize myocardial signals in each patient following thisprocedure: The primary 2D-IB measurement (in milli-volts) of myocardial wall (septum or posterior wall)reflectivity is divided by the primary 2D-IB measure-ment (in millivolts) of pericardial reflectivity and theresult is multiplied by 100 to obtain the percent inte-grated backscatter index.18,25,26

by guest on July 13, 2011http://circ.ahajournals.org/Downloaded from

Lattanzi et al Quantitative Analysis of Myocardial Reflectivity in Thalassemia Major 751

TABLE 3. Conventional Echocardiographic and DopplerFindings in Patients With Thalassemia Major and inControl Subjects

Patients withthalassemia major Controls

VS (mm) 9.3±1.7 8.7±1.7PW (mm) 8.6±1.6 8.5±1.2LVEDD (mm) 44.9±5.9 48.0±3.9LVMI (g/m2) 118.4±29.9 98.4±15.1*FSh (%) 32.4±2.5 36.0±3.5E wave (cm/sec) 93±16 86±9*A wave (cm/sec) 42+11 44±8E/A ratio 2.4±0.7 2.0±0.5*RR (msec) 797±124 824±105

VS, ventricular septum thickness; PW, posterior wall thickness;LVEDD, left ventricular end-diastolic diameter; LVMI, left ven-tricular mass index; FSh, left ventricular percentual fractionalshortening; E wave, peak late diastolic flow velocity; A wave, peakearly diastolic flow velocity.

*p<0.05 vs. patients' data.

Together with the percent IB, the following threequantitative measurements were displayed: 1) the pri-mary 2D-IB values (expressed in millivolts) for eachstructure; 2) the normalized 2D-IB values, expressed indB (that is, a relative unit measurement of soundenergy, conventionally used), of primary measurementsreferred to an external specular reflection (previouslyacquired, and with a very high reflectivity) (calculatedas 20 log V/Vr, where V [value corresponding to themyocardial wall-septum or posterior wall] and Vr[value corresponding to the reference external reflec-tion] are both primary 2D-IB values and expressed inmillivolts); 3) the normalized 2D-IB values, expressed indB, of primary measurements are referred in eachpatient to the pericardial reflection and calculated withthe above-mentioned formula, replacing the externalreflection with the internal pericardial interface foreach subject.

Statistical AnalysisData are given as mean±SD. Intergroup differences

were tested for significance using the unpaired Stu-dent's t test. Relations between radiofrequency andtwo-dimensional echocardiographic measurementswere expressed in terms of linear regression analysis. Avalue ofp<0.05 was considered statistically significant.

ResultsIn all subjects, the conventional echocardiographic

and radiofrequency quantitative ultrasound measure-ments were obtained.

Echocardiographic FindingsEchocardiographic and Doppler findings of patients

and control subjects are shown in Table 3.No abnormal values of left ventricular diameter, left

ventricular mass index, or percent fractional shorteningwere found. No epipericardial junction abnormalitieswere recorded.No significant differences in two-dimensional-echo-

cardiographic parameters were present between pa-tients and control groups, although patients tended to

TABLE 4. Anatomic and Quantitative Radiofrequency Data inPatients With Thalassemia Major and in Control Subjects

Patients withthalassemia major Controls

Interventricular septum distance(mm)Posterior wall distance (mm)Pericardium distance (mm)Primary pericardial 2D-IB (mV)Primary septal 2D-IB (mV)Primary posterior wall 2D-IB(mV)Normalized ER pericardial2D-IB (dB)Normalized ER septal 2D-IB(dB)Normalized ER posterior wall2D-IB (dB)Normalized IR septal 2D-IB(dB)Normalized IR posterior wall2D-IB (dB)Percent septal 2D-IB (%)

Percent posterior wall 2D-IB(%)

44.5±7.7 46.5+6.991.8±13.5 92.6±+13.2101.0±13.9 101.5+13.196.9+15.2 100.2±17.832.6± 12.7 20.9±6.9*

15.5+5.3t 10.5+2.6*t

-2.0±1.4 -1.7±1.7

-12.0±3.5 -15.8+3.3*

- 18.3+3.Ot -21.4+2.2*t

-9.9±3.6 -14.0+2.6*

- 16.3+3.Ot - 19.7+2.4*t34.8+13.7 20.9+6.3*

16.3+5.5t 10.8+3.2*t

Distance, from the ultrasonic transducer to the acquisition gateon the structure; ER, external (specular reflector) reference; IR,internal (pericardium) reference; 2D-IB, two-dimensional inte-grated backscatter index.

*p<0.001 compared with the value in the patients' group.tp<0.001 compared with the septal value within the same group.

have a higher left ventricular mass index -probably dueto reduced body surface with respect to age-matchedcontrols. The patients show a significantly higher, al-though normal, value of transmitral Doppler E/A ratiocompared with controls, mainly due to an increasedearly diastolic flow velocity (E wave).The qualitatively assessed echocardiographic reflec-

tivity of myocardial walls was considered increased in 10thalassemic patients (interventricular septum in three,posterior wall in one, and both in six) and in none of thecontrols (26% versus 0%,p<0.01). Pericardial interfaceechocardiographic appearance was qualitatively judgedto be in the normal range in each study subject.

Quantitative Ultrasonic Backscatter Data2D-IB primary, normalized, and percent values of

myocardial wall in patients and controls are summarizedin Table 4. To evaluate differences on analyzed heartstructures depth in the chest wall, data on the distancefrom the ultrasound transducer of the acquisition gateon septum, posterior wall, and pericardium, in bothpatients and control groups, are also presented in Table4. The analysis of these data demonstrates that heartstructures of patients and controls have no significantdifferences in the depth value.

Percent backscatter values of both ventricular septumand posterior wall in patients with thalassemia majorshowed highly significant differences compared withcontrol subjects, exhibiting a diffuse increase in quanti-tatively assessed myocardial reflectivity (Figure 2).

In both patients and controls, the ventricular septumshowed higher reflectivity than the posterior wall. This

by guest on July 13, 2011http://circ.ahajournals.org/Downloaded from

752 Circulation Vol 87, No 3 March 1993

-p<.01-Y

m6camC4

,.

m0

0CL

I.V. septum Posteriorwall

FIGURE 2. Bargraphs representingpercent two-dimensionalintegrated backscatter index (2D-IB) values of septum andposterior wall in patients with thalassemia major (m) and incontrol (o) subjects. Quantitative reflectivity of myocardialwalls appears to be significantly higher in thalassemia patients.

finding is consistent with the result obtained in otherstudies where the same analysis system was used and isvery likely due to the different distance of the insonatedstructures from the focal point of the ultrasonic beam aswell as to the different attenuation of the signal.

Relation Between Conventional Echocardiographicand Quantitative Backscatter Results

In patients with thalassemia major, the relation be-tween percent 2D-IB values of the myocardial walls andconventional echocardiographic measurements was an-alyzed. No significant correlation was found betweenquantitative reflectivity and thickness of ventricularseptum and posterior wall (r=0.1 and 0.1, respectively;p=NS for both). No significant correlation was foundwhen comparing septal and posterior wall percent2D-IB and left ventricular end-diastolic diameter (r=0.1and 0.2, p=NS) or left ventricular mass index (r=0.2and 0.2, p=NS) or percentual fractional shortening(r= -0.2 and 0.1, p=NS) or E/A ratio (r=0.3 and 0.1,p=NS).

The 10 thalassemic patients with qualitatively as-sessed increased echocardiographic appearance of sep-tum and posterior wall did not exhibit higher values ofparietal 2D-IB compared with thalassemic patients withnormally appearing reflectivity (for septum: n=9,42.0+16.3% versus 32.6±12.2%, p=NS; for posteriorwall: n=7, 17.0+6.5% versus 16.1±5.3%,p=NS).Relation Between Hematological Profile andQuantitative Backscatter Results

Hematological data of patients with 13-thalassemiamajor were compared with septal and posterior wallpercent 2D-IB values for correlation. No significantcorrelation was identified when considering the years oftransfusion therapy (r=0.1 compared with septal per-cent 2D-IB and 0.04 compared with posterior wallpercent 2D-IB, p=NS for both), the total number oftransfusions (r=0.2 and 0.1, respectively; p=NS), theyears of chelation therapy (r=0.1 and 0.1,p=NS), or themean ferritin value (r=0.1 and 0.1, p=NS) (Figure 3).

DiscussionThe results of this study clearly demonstrate that in

patients with 3-thalassemia major, even without clinicalsigns of cardiac functional involvement, there are ab-

80

70

60

50

40

30

20

R=0.1p=NS

* *. *.

* .**

* *

j ,_- ' .

4 6 8 10 12 14 16 18 20 22 24 26

Transfusion time (years)

adcm

.0

CL0c.

80g

70.

60

50

40'

30

20

R-O.2p=NS C

* S

* * C *n-% *-. * ** *

101 - -

0

CD

00

c

100 200 30 400 500 600Total number of transfusions

700

R=0.1p=NS

* 9

C: *

* 0A

* *

2 4 6 8 10 12

Chelation time (years)

FIGURE 3. Scatterplots of relation between ultrasoundquantitative data (here considered as percent integrated back-scatter of the interventricular septum [2D-IBJ) and hemato-logical data (transfusion therapy duration, total number oftransfusions, and chelation therapy duration) of the studypatients. No significant correlations were found consideringeach of these parameters.

normalities in myocardial acoustic properties, probablyrelated to histopathological alterations. The increasedreflectivity of myocardial walls found in these patientsmight be technically explained by several morphologicalsubstrates that have been documented by histologicaland biochemical studies in cardiac hemochromatosis:fibrosis, hypertrophy, and iron deposition.1-3 Histopath-ological changes of heart structures due to secondaryiron overload are characterized mainly by the presenceof small iron deposits, predominantly within cells ofventricular myocardium and conduction tissue. These

in, -.1 Li

by guest on July 13, 2011http://circ.ahajournals.org/Downloaded from

Lattanzi et al Quantitative Analysis of Myocardial Reflectivity in Thalassemia Major 753

deposits are more diffuse in the subepicardial than inthe subendocardial regions; flogosis and fibrosis areminimal to mild. Macroscopic alterations consist of leftventricular hypertrophy and, in the late stages of thedisease, dilation of heart chambers. Pericardial effusioncan be observed in the advanced stages.The collagen content is a major determinant of

myocardial backscatter, as shown by theoretical evi-dence as well as by experimental12'l5 and clinical stud-ies.18 In humans, a significant relation has been demon-strated between the amount of connective tissue-morphometrically evaluated from endomyocardialbiopsies- and the intensity of regional myocardial back-scatter.18 In hemochromatosis, connective tissue may beincreased, which may account for the observed increasein myocardial ultrasonic reflectivity. However, in thispathology, the entity of myocardial fibrosis is usuallymild,2 and therefore other causes are likely to contrib-ute to increased echo density.

Left ventricular hypertrophy might be another causeof abnormal myocardial acoustic characteristics. In pre-vious studies, we have found an increased echoreflec-tivity in patients with hypertrophic cardiomyopathy19but not in the physiological myocardial hypertrophy ofathletes,27 suggesting that increased echoreflectivity inhypertrophic cardiomyopathy is probably related tofiber disarray as well as to fibrosis and necrosis foci.28The myocardial hypertrophy at times present in hemo-chromatosis is not associated with fiber disarray; there-fore, such morphological changes are not likely to be amajor determinant of regional echodensity-partly be-cause left ventricular wall thickness was normal in ourstudy population.

Finally, intracellular iron deposition in hemochroma-tosis might be responsible for the increased reflectivity ofmyocardial walls since multiple foci of intracellular irondeposition might create discontinuities in acoustic im-pedance and generate multiple interfaces, eventuallyincreasing the ultrasonic backscatter. Iron depositionmost likely occurs in the heart of patients undergoingfrequent blood transfusions, such as thalassemics. In thepatients of the present study, however, no significantcorrelation was found between an acoustic quantitativeparameter (septal backscatter) and hematological data(years and number of blood transfusions, years of chela-tion therapy, and mean serum ferritin value). This find-ing, although seemingly surprising, is in accordance withreports by others720 who did not find a correlationbetween hematological parameters and myocardial in-volvement. The mechanism by which iron overload pro-duces tissue damage has not been well established, but itis generally accepted that iron toxicity begins when theiron load exceeds the tissue- or blood-binding capacityand joins a free pool defined as nontransferritin iron.29The amount of iron that can be stored in a tissue dependson the capacity of the tissue to generate storage proteins;as we are not able to evaluate the amount of free pool,only a rough correlation can be derived between storagelevels and cardiac toxicity. Therefore, myocardial in-volvement may occur in the absence of excessive cardiacstores, partly because heart cells have a relatively smallamount of storage proteins and are very sensitive to freeiron-induced oxygen radicals that are directly responsi-

Furthermore, it is likely that the system used forquantitative ultrasound analysis can detect the increasein echoreflectivity due to abnormal iron deposition inthe myocardium but is not able to identify little varia-tions in the entity of these deposits. Moreover, theentity and progression of fibrotic reaction, anotherimportant cause of increased echoreflectivity, can beindependent (not correlated) to the entity of irondeposition, which may represent the initial trigger ofthat reaction.

Thus, the increased echoreflectivity detected in themyocardium of thalassemia major patients might bemultifactorial in origin and related to fibrosis, ironstores, or both. The relative role of each factor isdifficult to establish in our patients, particularly becauseno histopathological correlation was available.The only reliable method to obtain such data in a

clinical study, such as this one, is by the use of endo-myocardial biopsy, whose results would have allowed adirect correlative analysis between ultrasonic and histo-logical findings. However, this invasive procedure car-ries a definite risk and is not routinely performed inpatients with thalassemia major since no significantbeneficial effect on the clinical management of thesepatients can be obtained through the analysis of itsresults.Although echocardiographic indexes of regional and

global systolic function were normal in our thalassemicpatients, an increased E/A ratio versus normals wasdetected by Doppler echocardiography. This finding isconsistent with previous observations by Spirito et al,7who showed restrictive diastolic abnormalities by trans-mitral Doppler in these patients-consisting of in-creased E wave and reduced flow-velocity decelerationtime-in an early phase of cardiac involvement, whensymptoms of heart failure are absent and systolic func-tion is normal.Of interest, the qualitatively assessed reflectivity on

two-dimensional echocardiograms did not clearly differ-entiate patients with markedly increased regional back-scatter. This is, however, not surprising in view of thebasically different nature of the information provided byradiofrequency backscatter compared with the imagebrightness of standard two-dimensional images. In fact,the electronic processing in commercially available two-dimensional imaging systems heavily manipulates theradiofrequency native signal to optimize the display ofspecular reflectors, such as endocardial borders. Theprocesses of differentiation, thresholding, nonlinear am-plification, and overall and regional gain are aimed atgenerating a clearly defined, pleasant image but cer-tainly disrupt the linear relation between the signalreceived by the transducer and the signal displayed inthe echo monitor.30 A sampling of the backscatteredmyocardial signal, although much more complex, allowsa more accurate characterization of myocardial acousticproperties.

In conclusion, although further studies are needed toconfirm our observation, we emphasize that myocardialtissue characterization with a real-time integrated back-scatter imaging system is a useful tool for detectingmyocardial abnormalities in patients with thalassemiamajor before clinical signs of cardiac damage andechocardiographic evidence of left ventricular regional

ble for cellular damage. and global dysfunction are detectable.

by guest on July 13, 2011http://circ.ahajournals.org/Downloaded from

754 Circulation Vol 87, No 3 March 1993

References1. Bunn HF: Disorders of hemoglobin-Thalassemias, in Wilson JD,

Braunwald E, Isselbacher KJ, Petersdorf RG, Martin JB, FauciAS, Root RK (eds): Harrison's Principles of Medicine. New York,McGraw-Hill, 1991, pp 1542-1552

2. Engle MA: Cardiac involvement in Cooley's anemia. Ann NYAcadSci 1969;119:694-702

3. Leon MB, Borer JS, Bacharach SL, Green MV, Benz EJ, GriffithP, Nienhuis AW: Detection of early cardiac dysfunction in patientswith severe beta-thalassemia and chronic iron overload. N Engi JMed 1979;301:1143-1148

4. Freeman AP, Giles RW, Berdoukas VA, Walsh WF, Choy D,Murray PC: Early left ventricular dysfunction and chelation ther-apy in thalassemia major. Ann Intern Med 1983;99:450-454

5. Henry WL, Nienhuis AW, Wiener M, Miller DR, Canale VC,Piomelli S: Echocardiographic abnormalities in patients with trans-fusion-dependent anemia and secondary myocardial iron deposit.Am J Med 1978;64:547-555

6. Valdez-Cruz LM, Reinecke C, Rutkowski M, Dudell GG, Gold-berg SJ, Allen HD, Sahn DJ, Piomelli S: Preclinical abnormalsegmental cardiac manifestations of thalassemia major in childrenon transfusion-chelation therapy: Echocardiographic alterations ofleft ventricular posterior wall contraction and relaxation patterns.Am Heart J 1982;103:505 -511

7. Spirito P, Lupi G, Melevendi C, Vecchio C: Restrictive diastolicabnormalities identified by Doppler echocardiography in patientswith thalassemia major. Circulation 1990;82:88-94

8. Engle MA, Erlandson M, Smith CH: Late cardiac complications ofchronic, severe, refractory anemia with hemochromatosis. Circula-tion 1964;30:698-705

9. Miller JG, Perez JE, Sobel BE: Ultrasonic characterization ofmyocardium. Prog Cardiovasc Dis 1985;28:85-110

10. Skorton DJ, Miller JG, Wickline SA, Barzilai B, Collins SM, PerezJE: Ultrasonic characterization of cardiovascular tissue, in MarcusML, Schelbert HR, Skorton DJ, Wolf GL (eds): Cardiac Imaging.Philadelphia, WB Saunders, 1991, pp 538-556

11. O'Donnell M, Mimbs JW, Miller JG: The relationship betweencollagen and ultrasonic backscatter in myocardial tissue. J AcoustSocAm 1981;69:580-588

12. Mimbs JW, O'Donnell M, Bauwens D, Miller JG, Sobel BE: Thedependence of ultrasonic attenuation and backscatter on collagencontent in dog and rabbit hearts. Circ Res 1980;47:48-58

13. Picano E, Landini L, Lattanzi F, Mazzarisi A, Sarnelli R, DistanteA, Benassi A, L'Abbate A: The use of frequency histograms ofultrasonic backscatter amplitudes for detection of atherosclerosisin vitro. Circulation 1986;74:1093-1098

14. Lattanzi F, Picano E, Mazzarisi A, Aratari C, Pelosi G, PozzoliniA, Salvatore L, Landini L, Distante A, L'Abbate A: In vitro iden-tification of different degrees of mitral valve disease by onlineevaluation of radiofrequency ultrasound signal. Cardiovasc Res1988;21:841-846

15. Hoyt RH, Collins SM, Skorton DJ, Ericksen EF, Conyers D:Assessment of fibrosis in infarcted human hearts by analysis ofultrasonic backscatter. Circulation 1985;71:740-744

16. Lattanzi F, Picano E, Landini L, Mazzarisi A, Pelosi G, Benassi A,Salvatore L, Distante A, L'Abbate A: In vivo identification ofmitral valve fibrosis and calcium by real time quantitative ultra-sonic analysis. Am J Cardiol 1990;65:355-359

17. Vered Z, Barzilai B, Mohr GA, Thomas II U, Genton R, SobelBE, Shoup TA, Melton HE, Miller JG, Perez JE: Quantitativeultrasonic tissue characterization with real-time integrated back-scatter imaging in normal human subjects and in patients withdilated cardiomyopathy. Circulation 1987;76:1067-1073

18. Picano E, Pelosi G, Marzilli M, Lattanzi F, Benassi A, Landini L,L'Abbate A: In vivo quantitative ultrasonic evaluation of myocar-dial fibrosis in man. Circulation 1990;81:58-64

19. Lattanzi F, Spirito P, Picano E, Mazzarisi A, Landini L, DistanteA, Vecchio C: Quantitative assessment of ultrasonic myocardialreflectivity in hypertrophic cardiomyopathy. J Am Coll Cardiol1991;17:1085-1090

20. Fosburg MT, Nathan DG: Treatment of Cooley's anemia. Blood1990;76:435-444

21. Sahn DJ, DeMaria A, Kisslo J, Weyman A: Recommendationsregarding quantitation in M-mode echocardiography: Results of asurvey of echocardiographic measurements. Circulation 1978;58:1072-1083

22. Devereux RB: Detection of left ventricular hypertrophy byM-mode echocardiography: Anatomic validation, standardization,and comparison to other methods. Hypertension 1987;9(suppl II):II-19-II-26

23. Barzilai B, Madaras EI, Sobel BE, Miller JG, Perez JE: Effects ofmyocardial contraction on ultrasonic backscatter before and afterischemia. Am J Physiol 1984;247:H478-H483

24. Landini L, Salvadori M, Mazzarisi A, Benassi A: On-line evalua-tion of ultrasonic integrated backscatter. J Biomed Eng 1985;7:301-305

25. Logan-Sinclair R, Wong CM, Gibson DG: Clinical application ofamplitude processing of echocardiographic images. Br Heart J1981;45:621-627

26. Lattanzi F, Picano E, Mazzarisi A, Landini L, Benassi A, Masini A,Distante A, L'Abbate A: In vivo radio-frequency ultrasound anal-ysis of normal human heart structures. J Clin Ultrasound 1987;15:371-375

27. Lattanzi F, Di Bello V, Picano E, Caputo MT, Talarico L, Di MuroC, Landini L, Santoro C, Giusti C, Distante A: Normal ultrasonicmyocardial reflectivity in athletes with increased left ventricularmass: A tissue characterization study. Circulation 1992;85:1828-1834

28. Perez JE: Ultrasound characterization of myocardial hypertrophy.(editorial) JAm Coll Cardiol 1991;17:1091-1093

29. Gutteridge J, Halliwell B: Iron toxicity and oxygen radicals. Bail-here's Clin Hematol 1989;2:195

30. Skorton DJ, Collins SM: Digital computer image analysis in echo-cardiography. Echocardiography 1984;1:15-43

by guest on July 13, 2011http://circ.ahajournals.org/Downloaded from