diagnostic methods congenital heart disease · diagnoses that included valvular regurgitant...

DIAGNOSTIC METHODSCONGENITAL HEART DISEASE

Digital angiography in the pediatric patient withcongenital heart disease: comparison with standardmethodsAARON R. LEVIN, M.D., F.R.C.P. (EDIN), HARVEY L. GOLDBERG, M.D., JEFFREY S. BORER, M.D.,LAWRENCE N. ROTHENBERG, PH.D.. FLORENCE A. NOLAN, M.D., MARY ALLEN ENGLE, M.D.,BARRY COHEN, B.A., NANCY T. SKELLY, B.S., AND JOHN CARTER, M.S.

ABSTRACT Digital subtraction angiography (DSA) permits high-resolution cardiac imaging withrelatively low doses of contrast medium and reduced radiation exposure. These are potential advan-tages in children with congenital heart disease. Computer-based DSA (30 frames/sec) and conventionalcutfilm angiography (6 frames/sec) or cineangiography (60 frames/sec) were compared in 42 patients,ages 2 months to 18 years (mean 7.8 years) and weighing 3.4 to 78.5 kg (mean 28.2 kg). There were 29diagnoses that included valvular regurgitant lesions, obstructive lesions, various shunt abnormalities,and a group of miscellaneous anomalies. For injections made at a site distant from the lesion and on theright side of the circulation, the mean dose of contrast medium was 60% to 100% of the conventionaldose given during standard angiography. With injections made close to the lesion and on the left side ofthe circulation, the mean dose of contrast medium was 27.5%. to 42% of the conventional dose.Radiation exposure for each technique was markedly reduced in all age groups. A total of 92 digitalsubtraction angiograms were performed. Five studies were suboptimal because too little contrastmedium was injected; in the remaining 87 injections, DSA and conventional studies resulted inidentical diagnoses in 81 instances (p < .001 vs chance). The remaining six injections made duringDSA failed to confirm diagnoses made angiographically by standard cutfilm angiography or cineangi-ography. We conclude that DSA usually provides diagnostic information equivalent to that availablefrom cutfilm angiography and cineangiography, but DSA requires considerably lower doses of contrastmedium and less radiation exposure than standard conventional methods.Circulation 68, No. 2, 374-384, 1983.

DIGITAL SUBTRACTION ANGIOGRAPHY (DSA)has recently been shown to be a valuable tool in theevaluation of great-vessel anomalies'-9 and of heartdisease2 10, in the adult patient. Not only does thetechnique permit high-resolution imaging when injec-tions of contrast medium are made at sites distant fromthe heart or lesion, but such imaging can also beachieved with low loads of contrast medium and re-duced radiation exposure.9' 10 These factors offer po-

From the Divisions of Pediatric and Adult Cardiology, the NewYork-Cornell Medical Center, New York.

This project was supported in part by BRSG S07 RR 05396, awardedby the Biochemical Research Support Grant Program, Division of Re-search Resources, National Institutes of Health.

Address for correspondence: Aaron R. Levin, M.D., F.R.C.P., Divi-sion of Pediatric Cardiology, The New York Hospital-Cornell MedicalCenter, 525 East 68 St., Room F-468, New York, NY 10021.

Received Feb. 1, 1983; revision accepted April 21, 1983.Dr.Borer is an Established Investigator of the American Heart Asso-

ciation, and is supported in part by the AHA, Dallas.Dr. Rothenberg is a member of the Dept. of Medical Physics, Memo-

rial-Sloan Kettering Cancer Center, New York.

374

tential advantages in the diagnosis of congenital cardi-ac lesions in infants, children, and adolescents, andespecially in those patients with complex, cyanoticheart anomalies in whom use of moderate-to-highdoses of contrast medium during diagnostic cardiaccatheterization may constitute a particularly high risk.Few studies that use DSA have been undertaken in thepediatric age group,2' 9, 12 and the diagnostic accuracyof this technique is still unclear. At the New YorkHospital-Cornell Medical Center, we have evaluatedDSA with a single-mask mode in 42 pediatric patientswith congenital cardiac lesions.

MethodsComputer-based DSA was performed in 42 pediatric patients

with congenital cardiac disease who were undergoing conven-tional cardiac catheterization and angiographic evaluation. Thestudy group consisted of 18 male and 24 female patients (ages 2months to 18 years, mean 7.8), and 12 of these patients wereunder 3 years old (table 1). The weight distribution was 3.4 to78.5 kg, with a mean weight of 28.2 kg; 16 patients weighed

CIRCULATION

by guest on August 17, 2017

http://circ.ahajournals.org/D

ownloaded from

http://circ.ahajournals.org/

DIAGNOSTIC METHODS-CONGENITAL HEART DISEASE

TABLE 1Patient population (n = 42)

Age distribution (yr)

0-3 3-6 6-9 9-12 12-15 15-18 Total

Males 7 1 1 1 2 6 18Females 5 8 4 4 2 1 24

Mean 7.8 yr 42

Weight distribution (kg)

0-15 15-30 30-45 45-60 60-75 75-90 Total

Males 7 2 2 3 3 1 18Females 9 8 3 4 24

Mean 28.2 kg 42

under 15 kg (table 1). The body surface area ranged from 0.2 to1.97 m2 (mean 0.94). None of the patients had congestive cardi-ac failure at the time of study.As determined by standard film-based angiography and/or

subsequent assessment at cardiac surgery, a total of 29 anoma-lies or combinations of lesions was present in our patients (table2). Included were valvular lesions, intracardiac and extracar-diac shunt lesions, and' a group of miscellaneous con-ditions.

Routine right and left heart cardiac catheterization was per-formed with patients in the nonabsorptive state after sedationwith 2 mg/kg meperidine (Demerol), 1 mg/kg promethazineHCl (Phenergan), and 1 mg/kg chlorpromazine (Thora-zine).13 14 Venous and arterial catheters were inserted percuta-neously into the right femoral vessels. For femoral vein punc-tures, No. 6F to 8F Berman balloon angiocatheters were used,and No. 6F to 7F UMI pigtail catheters were used for the arterialapproach. When standard cutfilm angiography or cineangiog-raphy were performed, patients received a maximum of 1.0 to1.25 ml/kg body weight at each injection of meglumine diatri-zoate (Renografin 76). No patient received more than 3.0 ml/kgof contrast medium during the standard angiographic study. Themaximum dose of contrast medium did not exceed 50 ml for anysingle injection. Contrast medium was delivered at a rate of 16to 32 ml/sec, depending on catheter size and length.

At completion of the routine diagnostic catheterization and

TABLE 2Lesions studied

Regurgitant lesionsAortic valveMitral valve and prolapseTricuspid valveRuptured sinus of valsalva

Shunt lesionsAtrial septal defectCoronary A-V fistulaPatent ductus arteriosusTetralogy of FallotVentricular septal defect

Obstructive lesionsAortic stenosis Valvular

SubvalvularCoarctation of aorta

Pulmonic atresiaPulmonic stenosis Valvular

InfundibularPulmonic vein stenosisTricuspid atresia

Miscellaneous lesionsBlalock-Taussig shuntEccentric LVH1-TranspositionPotts' anastomosisRight ventricular outflow aneurysm

Septal aneurysm

angiographic procedure, and only after the definite diagnosishad been established from the processed films and hemodynam-ic data, digital angiograms were obtained. For DSA, patientswere positioned in the orientation that best defined the lesionduring standard angiographic assessments. Informed consentfrom each patient and approval of the Cornell University Medi-cal College'Committee on Human Rights in Research were

obtained before each DSA study. An American Edwards Labo-ratory CARDIAC 1,000 Digital Subtraction Angiography Com-puter was used. The computer was programmed to sample 30frames/sec; for diagnostic accuracy and radiographic definition,these data were later compared in the same patient with dataobtained by conventional biplane cutfilm angiography (2 to 6frames/sec) or by cineangiography (60 frames/sec).'For DSA,the amount of contrast medium injected depended on the site ofinjection. When injections were made at a site distant from thelesion, such as the venous side of the heart for lesions on the leftside, the mean amount of contrast medium delivered variedbetween 60% and 100% of the conventional dose administeredduring standard angiography (figure 1). When injections were

performed at central sites close to the lesion, or at the site of thelesion, the mean dose of contrast medium delivered ranged from27.5% to 42% of the conventional dose (figure 1). For DSA, therate of injection, which ranged between 16 and 32 ml/sec, was

maximal for the catheter size and length. No patient receivedmore than 1.0 to 1.5 ml/kg body weight of meglumine diatri-zoate for digital angiography, and DSA added no more than 10to 15 min to the total catheterization time. No patient received a

total dose of contrast medium in excess of 4 ml/kg body weightduring the entire procedure.

Radiation settings for infants and smaller children were 60kVp for cutfilm angiography and 70 kVp for both cineangiog-raphy and DSA (figure 2). A total setting of 800 mA was usedfor biplane cutfilm angiography compared with settings of 100

a)cn0

a)

c

0

ca)

C

0(-)

0

CL)

0-

RangeMean

PA LA-Injection site

FIGURE 1. Percent of conventional dose of contrast medium used

during DSA. For injections made at a site distant from the lesion and on

the right side of the circulation, the mean dose of contrast medium was

between 60% and 100% of the conventional dose given during standard

angiography. With injections made close to the lesion and on the left

side of the circulation, the mean dose of contrast medium ranged from

27.5% to 42% of the conventional dose. ACG = angiogram; IVC =

inferior vena cava; RA = right atrium; RV = right ventricle; PA =

pulmonary artery; LA = left atrium; LV = left ventricle; AO = aorta.

375

A-V = atrioventricular; LVH = left ventricular hypertrophy.

Vol. 68, No. 2, August 1983



ownloaded from


LEVIN et al.

RADIATION EXPOSUREInfant 3.4 kg

80r e7 CUt film03 CineM Digital70 V

60 F

50 I

40 I

30 F

20L

1OF0

rKvp

800

700

600

500

400

300

200

100

0mA

10,000

8,000

6,000

4,000

2,000

0mR

FIGURE 2. Radiation settings and exposure in the smallest patientstudied, weighing 3.4 kg. For each technique tube potential was similar.For cineangiography and DSA, tube current was reduced but the radi-ation exposure (mR) was markedly reduced for DSA.

mA for cineangiography and 1.0 mA for DSA (figure 2). Theexposure time for each biplane cutfilm angiogram was set at 6.3msec and the pulse width for each cineangiogram frame was 1 .0msec. Equivalent settings for older patients were 80 kVp and1200 mA for biplane cutfilm angiography, 70 kVp and 100 mAfor cineangiography, and 70 kVp and 5 mA for DSA (figure 3).The exposure time for each cutfilm angiogram was set at 16.0msec and that for each cineangiogram frame was 5.2 msec.Exposure measurements were performed with the mdh Model1015C x-ray dosimetry system with the Model IOX5-6 6 cm3ionization chamber. For each view or technique, the chamberwas centered in the x-ray beam at the position of the entrysurface in each patient. For cutfilm techniques, the tube poten-tial (kVp), tube current (mA), and exposure time (msec) wereset to the values used clinically. For fluoroscopy, cineangiog-raphy, and DSA techniques, which use the automatic brightnesscontrol system, we simulated the presence of patients by usingaluminum attenuation of sufficient thickness to drive the x-raygenerator to the tube potential, tube current, and pulse widthsthat had been recorded from the meters when the patients under-went dosimetry. Exposure times for fluoroscopy, cineangiog-raphy, or DSA were determined both by a measurement ofbeam-on time with a stopwatch while patients underwent do-simetry and by a review of videotapes at standard speed or by acount of the cineangiographic frames that had been exposed at60 frames/sec. The total exposure time for biplane cutfilm an-giography depended on the weight of the patient and variedfrom 6.3 msec per exposure in the smallest patient to 16.0 msecper exposure in the heaviest patient. Since 16 exposures wereobtained in each plane, the total exposure times varied from 208msec for infants to 512 msec for patients weighing more than 60kg (exposure time x 16 msec x 2). For cineangiography, totalexposure time varied similarly, depending on patient size andheart rate. With cineangiography performed at 60 frames persec, the total number of exposed frames averaged 396 framesper cineangiogram in infants and resulted in a total exposuretime of 396 msec. Similar values for older patients were an

376

average of 600 frames per cineangiogram for 3.12 sec totalexposure time.To minimize diaphragm movement and obscuring of the

heart, older patients who were able to cooperate were instructedto inspire fully and to maintain full inspiration during initialfluoroscopy in preparation of the mask (see below) for DSA,again just before injection of contrast medium, and until com-pletion of the fluoroscopic sequence. This procedure usuallytook no more than 10 sec. However, smaller children and in-fants breathed at their normal rates throughout the fluoroscopicprocedure. After the study was completed, images recorded bycutfilm angiography, cineangiography, and DSA were re-viewed by three independent observers in blinded fashion todetermine the extent to which diagnoses made from DSA repro-duced those made from cutfilm angiographic, cineangiograph-ic, and hemodynamic data. The observers also provided inde-pendent assessment of the quality of the angiograms obtained.To obtain the DSA, the area of interest was imaged by fluo-

roscopy to produce a picture on the image intensifier (figure 4).This picture was converted into an analog electric signal by thetelevision camera. The analog output from the television camerawas then converted by the computer into a digital format at a rateof 30 frames/sec, by division of the image into a 512 x 512pixel (picture element) matrix. The x-ray photon density, orbrightness, of each pixel was quantified and the quantity thusdetermined was proportional to the number of photons incidenton the image intensifier. The photon density of each pixel wasrecorded and stored digitally in the microprocessor memory. Aninitial set of 16 frames was averaged and stored in the memoryto define the background radiation transmitted through the pa-tient before injection of contrast medium. This background radi-ation was termed the "mask.

After determination of the mask, the contrast medium bolus

RADIATION EXPOSUREAdult 78.5 kg

90 r

80 F

70 I

601-

50 F

40I

30F

20Flo1

0

KVp

1,200

1,100

1,000

900

800

700

600

500

400

300

200

100

0 ImA

P Cut filmCM Cine

Digital16,000

14,000

12,000-

10,000

8,000

6,000

4,000-

2,000

0mR

FIGURE 3. Radiation settings and exposure in the oldest and largestpatient studied, weighing 78.5 kg. With each technique tube potentialwas similar. The reduction in tube current for DSA was greater than forcineangiography compared with cutfilm angiography. Radiation expo-

sure (mR) for DSA was markedly reduced.

CIRCULATION

U-L



ownloaded from



FIGURE 4. To obtain digital subtraction angiocar-diograms, the area of interest was imaged fluoro-scopically before angiocardiography, which pro-duced a picture on the image intensifier (II). Thepicture was converted to an analog electric signal (A)by the television camera (TV). The analog signal wasthen digitized (D) by the computer at 30 frames/secand the initial 16 frames averaged and stored in themicroprocesser memory as the "mask." Fluoros-copy was then repeated after administration of con-trast medium and the resulting images were digitizedas before. The mask was then electronically subtract-ed from the new image and the resultant image wasconverted back to an analog signal for viewing on astandard television monitor.

was administered and fluoroscopy was repeated; the resultingimages were digitized as before at a rate of 30 frames/sec. Thesingle mask was then electronically subtracted from all subse-quent images by the computer on a pixel-by-pixel basis, whichresulted in subtraction or marked attenuation of all structures inthe chest and in the relative enhancement of the image of thecontrast medium within the field of view. This processed digitalimage was then converted back to an analog signal for viewingon a standard television monitor.

Chi-square analyses were used to compare the frequency ofcorrect diagnoses vs chance.

ResultsDiagnostic accuracy. We obtained 92 digital subtrac-

tion angiograms in 42 subjects. Early in our experi-ence, five studies in children (ages 2 months to 23/4years) could not be interpreted because too little con-trast medium (less than 20% of the conventional doseor less than 3 ml per injection) had been injected,which resulted in suboptimal opacification (figure 5).In the remaining 87 injections, digital and convention-al studies resulted in identical diagnoses in 81 in-stances (p < .001 vs chance). In 16 of these, bothconventional cutfilm angiograms or cineangiogramsand digital subtraction angiograms were read as nor-mal when injections were performed to rule out associ-ated lesions in patients with other congenital cardiacdefects. There were 50 injections in which convention-al and DSA studies each indicated the presence of asingle lesion, nine injections in which two lesions wereidentified, and three instances in which three lesionswere identified by both methods. In three additionalinjections, both conventional cutfilm angiograms anddigital subtraction angiograms were falsely read asnormal when hemodynamic studies indicated the pres-ence of a lesion. Two of these injections were per-formed in a 2-month-old infant weighing 3.5 kg whohad pulmonary vein stenosis. One injection was madeinto the right ventricle and the second into the pulmo-nary artery in an unsuccessful attempt to define the siteof pulmonary venous obstruction. In the remainingVol. 68, No. 2, August 1983

false-negative injection, both cutfilm angiograms anddigital subtraction angiograms were read as normal in a6½/2 year old patient who had an insignificant smallatrial septal defect; in this patient, no oxygen step-upwas found at atrial level, but diagnosis was made bycatheter passage and by hydrogen gas-dilution studiesthat were positive in the right atrium and negative inthe venae cavae.

In the remaining six of the 87 satisfactory injections

CONVENTIONAL ANGIOGRAPHY

I0:

4

a.:

cn

(c0(9z

4

C,)

F-

C)

Normal Lesion2

Lesions3 No

Lesions Lesions

Normal 16

1 Lesion 50

2 Lesions 9

3 Lesions 3

Mild ASl MIldA)1PLSNo Lesions MiIdPS2 5 Eccild A ASPD. St 2

ASD 2iEc V S

Unreadable VSD 31V(Too little PDA 2J|contrast)

FIGURE 5. Comparison of results of conventional angiography anddigital angiography in 92 patients. Identical diagnoses were made witheach technique in 81 instances. In six patients, DSA failed to confirmdiagnoses made angiographically by standard techniques. Five patientshad suboptimal injections of contrast medium. See text. AS = aorticstenosis; PS = pulmonary stenosis; ASD = atrial septal defect; Ecc.LVH = eccentric left ventricular hypertrophy; Pul. V. St. pulmo-nary valve stenosis; VSD = ventricular septal defect; PDA patentductus arteriosus.

377



ownloaded from


LEVIN et al.

(figure 5), DSA failed to confirm diagnoses made an-giographically by standard cutfilm angiography orcineangiography. These patients ranged in age from 2to 177/2years. Five of the false-negative digital studieswere in patients with a single lesion: mild aortic valvu-lar stenosis in one child 16 years old, mild pulmonicvalvular stenosis in two children, 2 and 3 years old,and atrial septal defect in two patients 11 years and 17years old. In the two patients with atrial septal defect,failure to identify the defect resulted when digital an-giograms were obtained from the inferior vena cavawith the patient positioned in the left anterior obliqueprojection. During the levophase of those injections,no visualization of contrast medium was achieved inthe right atrium after left atrial opacification. The diag-nosis in one additional patient with two lesions (mildaortic stenosis and eccentric hypertrophy of the leftventricle) was not confirmed by the digital technique.

Doses of contrast medium. When lesions were as-sessed by intravenous administration of contrast medi-um, larger amounts were used than when contrast me-dium was administered directly into the left heart orinto an area close to the site of the lesion. However,irrespective of the site of contrast medium administra-tion (in all cases but one), doses for DSA were equal toor less than those used to image the same lesion withcutfilm angiography or cineangiography. Doses ofcontrast medium for DSA varied from 30% to 100%(average 79%) of the conventional dose in the 22 caseswhere injections for DSA were made into the inferiorvena cava (doses of contrast medium for DSA, 10 to 45ml, mean 21.6 ml, figure 1). Doses of contrast medi-um for DSA in the five right ventricular injectionsranged from 25% to 140% (average 60%, doses ofcontrast medium for DSA, 5 to 16 ml, mean 7.4 ml) ofthat used for conventional angiography in the samepatients; in six pulmonary artery injections, dosagewas from 40% to 140% (average 76%, doses of con-trast medium for DSA, 5 to 45 ml, mean 21 ml). In the2-month-old patient weighing 3.5 kg with pulmonaryvein stenosis, two digital angiograms were obtained,one in the right ventricle and the second in the pulmo-nary artery, with 140% (5.0 ml) of the conventionaldose of contrast medium, but the site of the pulmonaryvenous obstruction was not defined. When this patientis excluded, the mean amount of contrast medium in-jected into the right ventricle and pulmonary artery was40.5% (range 25% to 64%, 5 to 16 ml, mean 8 ml) and63.6% (range 40% to 90%, 8 to 45 ml, mean 24.5 ml),respectively, of the dose administered for conventionalangiography. Doses of contrast medium for DSA ofleft ventricular injections were 20% to 100% (average

378

41%, 3 to 35 ml, mean 9.4 ml) of conventional dosage,and comparisons to other sites of injection yieldedsimilar values (figure 1). The absolute amount of con-trast medium used in these studies ranged between 3and 45 ml.

Image quality. When three observers subjectivelycompared the quality of radiographic definition amongthe three methods of angiography, definition by DSAwas invariably deemed adequate for diagnostic pur-poses if no less than 3.0 ml of contrast medium wasdelivered per injection at central sites close to or at thelesion and if no less than 60% of the conventional dosewas injected from distant sites.

Radiation exposure. The measured radiation exposurerates at entry surface in each patient revealed that toobtain diagnostic images, DSA involved markedly lessradiation exposure than cutfilm angiography or cinean-giography; these were approximately 10% of the expo-sure rate required for cutfilm angiography and about4% of that required for cineangiography. Detailed ex-posure comparisons for individual patients dependedon the total number of cutfilm angiographic films ex-posed and on the total time of the cineangiography andDSA runs. For example, in the smallest infant, weigh-ing 3.4 kg, the radiation exposure with DSA was 140millirads, while comparable radiation exposure forcutfilm angiography and cineangiography was 1120and 6000 millirads, respectively (figure 2). For theoldest and heaviest patient weighing 78.5 kg, expo-sures were 700 millirads for DSA, 14,944 millirads forcutfilm angiography, and 15,600 millirads for cinean-giography (figure 3).

DiscussionDSA in adult patients with great-vessel or cardiac

diseases 1-2 has been shown to be highly accurate inreproducing information available with more conven-tional methods. Few studies, however, have been per-formed in pediatric patients with congenital heart dis-ease in whom the particular advantages of DSA, whichinclude reduced doses of contrast medium and reducedradiation exposure, might be most important. Al-though the role of DSA in the pediatric patient withcongenital heart disease is not yet established, encour-aging reports have appeared. Weinstein et al.2 obtaineddigitally subtracted radiographic images at 6 frames/sec on a 256 x 256 x 8 bit matrix and were able toadequately identify cardiac chambers and to note therecirculation of contrast medium through septal de-fects in four patients. In addition, these investigatorsidentified complicated congenital cardiac lesions,which included single-ventricle, double-chambered

CIRCULATION



ownloaded from



heart and anomalous pulmonary venous return; theseinvestigators estimated the ejection fraction, relativeshunt flow, and cardiac output with densitometricanalysis. Yiannikas et al.12 also used densitometry toquantitate shunt size. In a preliminary report of theirexperience with 100 patients with congenital heart dis-ease (40 with left-to-right shunts), injections of con-

trast medium for DSA were made from a peripheralarm vein. They concluded that DSA compared favor-ably with cardiac catheterization in localizing intracar-diac defects and in estimating pulmonary-to-systemicflow ratio, as well as right and left ventricular ejectionfractions. In addition, they were able to identify asso-

ciated lesions with a high degree of accuracy; in 85%of the patients, the diagnosis by DSA was in accordwith the final diagnosis obtained by conventionalmethods.Our data confirm these preliminary findings, which

indicate that DSA can be applied successfully in thediagnosis of congenital heart disease in pediatric pa-

tients spanning a wide range of ages and weights.However, our method of contrast medium injection forDSA differed from that of previous studies in thatinjections were made in the inferior vena cava, or closeto or at the site of the lesion, rather than in peripheralveins. With injections made centrally, we were able toreduce greatly the dose of contrast medium, as com-

pared with earlier studies, without apparent loss ofdiagnostic accuracy. Furthermore, we were able toimage the area of interest at 30 frames/sec with a 512x 512 pixel matrix, which potentially permits bothaccurate assessment of cardiac chamber function andsophisticated computer-based image processing. Al-though we made no attempt to quantitate flows or todefine right or left ventricular function in this study,we and others have shown previously that in adults,left ventricular volumes and regional and global systol-ic function as determined by intravenous DSA are

closely correlated with results obtained by convention-

al intraventricular cineangiography.'5 16The reduction in doses of contrast medium afforded

by DSA is a major advantage in the pediatric patient.The dose-related toxic and potentially lethal effects ofcontrast medium in the pediatric age group are welldocumented, especially in those newborn infants withcyanotic lesions or cardiac failure during the first fewhours or days of life.17-21 The usual dose of contrastmedium necessary for images of diagnostic qualitywith cutfilm angiography and cineangiography is 1.0to 1.5 ml/kg body weight. 13' 1 Loading of contrast me-dium with these doses has been associated with rapidfall in pH with acidosis,'71-9 which potentially can re-

Vol. 68, No. 2, August 1983

sult in cardiac arrhythmias and deterioration in ventric-ular function. The fall in pH tends to be cumulative if arecovery period of 10 to 15 min between injections isnot permitted, and it may be particularly marked inseverely cyanotic newborn infants who often exhibitmetabolic acidosis as a presenting feature of their ill-ness.A rapid rise in osmolality also results from contrast

medium administration and may be significantly miti-gated by the lower doses of contrast medium necessarywith DSA. Increased blood osmolality may result in asudden increase in intravascular volume with worsen-ing of cardiac failure or pulmonary edema.2022 It hasbeen estimated that every 1.0 ml of hyperosmolar con-trast medium results in a short term plasma expansionof 8 to 10 ml. 18. 20 Thus, infants and children withincipient or overt cardiac failure may be placed sud-denly at increased risk immediately after angiography.The present study indicates that adequate visualizationof cardiac anomalies in pediatric patients may be at-tained with DSA after injections as small as 20% ofconventional doses of contrast medium are made at orclose to the site of the lesion, and injections as small as60% of the conventional dose when made distant fromthe lesion or in the venae cavae; this reduces the risk ofcomplications. Furthermore, when lesions are com-plex and added injections are required for diagnosis,multiple angiograms may be performed with DSAwithout exceeding acceptably safe limits of contrastmedium doses.

Although DSA almost uniformly permits reductionin conventional doses of contrast medium necessaryfor diagnostic imaging, a limitation exists in that, withcurrent equipment, the smallest bolus of contrast medi-um must exceed 3.0 ml and must be no less than 20%of the conventional dose delivered under high pressurefor DSA to provide images of diagnostic quality. Inour first five patients, three with ventricular septaldefect and two with patent ductus arteriosus, doses ofcontrast medium were too small to permit an accurateimage assessment, and in each instance the amount ofcontrast medium injected was less than 3.0 ml (figure5). The smallest amount of contrast medium injected toachieve excellent visualization of the lesion was ina child with a patent ductus arteriosus who received3.0 ml or 36% of the conventional dose injected intothe left ventricle, which revealed an intact ventricularseptum and confirmed the patent ductus arteriosus(figure 6). Other examples are depicted in figures 7through 11.

In six patients (figure 5), DSA failed to define alesion that had been confirmed by either cutfilm angi-

379



ownloaded from


LEVIN et al.

FIGURE 6. Comparison of standard aortogram (left) showing a patent ductus arteriosus, with digital left ventriculogram (right)showing an intact ventricular septum and confirming the patent ductus arteriosus. The patient, age 11/4 years and weighing 8.2kg. received 36% (3.0 ml) of thc conventional contrast medium dose for the digital ventriculogram.

ography or cineangiography. In five of these patients asingle lesion was present (one aortic valve stenosis,two pulmonic stenoses, and two atrial septal defects).The remaining patient had two definable lesions, mildaortic stenosis, and eccentric left ventricular hyper-trophy. Mild stenotic lesions of the semilunar valvesmay be difficult to define by conventional angiographyas well; diagnosis often depends on the catheter-mea-sured pressure differential across the valve. In bothpatients with the small atrial septal defect, the diagno-sis was missed when assessment was attempted during

380

the levophase of injections made in the inferior venacava. In each instance 60% to 70% of the conventionaldose of contrast medium was used. Angiographic diag-nosis of an atrial septal defect may require convention-al doses of contrast medium during DSA when per-formed from the inferior vena cava.

In addition to the reduction in doses of the contrastmedium, another advantage associated with DSA isthe marked reduction of radiation exposure to the pa-tient as well as to the angiographer. High doses of ra-diation during cardiac catheterization and angiography

FIGURE 7. Tetralogy of Fallot in a 1 12-year-old patient, weighing 12.4 kg. The digital left ventriculogram (-ight) was obtainedwith 3.0 ml or 25% of the conventional contrast medium dose used for the standard left ventriculogram (1 0.0 ml, ljftl. Note theventricular septal defect and the overriding aorta.

CIRCULATION



ownloaded from


DIAGNOSTIC METHODS CONGENITAL HEART DISEASE

FIGURE 8. Right subclavian-to-pulmonary artery anastomosis in a 5-year-old patient weighing 15 kg with severe tetralogy ofFallot. The digital angiogram (right) was made by hand injection with 336/c (5 ml) of the contrast medium dose used for theconventional angiogram (15 ml, left).

have been shown to be potentially deleterious.'3-27 Be-cause of their small size, infants and children receivean increased radiation exposure from internal scatterduring conventional angiography, which results inhigher doses of radiation to both the gonads and thethyroid gland.2t Since children can be expected to livelonger after exposure than adults, the overall conse-quences of radiation exposure are greater than inadults. The significance of direct gonad exposure haslong been recognized, but the radiation exposure dur-

ing angiography is relatively small.27 The conse-quences of thyroid sensitivity have recently been ap-preciated, since radiation exposure has been linked tothyroid carcinoma. Previous studies28 29 indicate an in-cidence of thyroid cancer of between 1.6 and 9.3 casesper million per year per rad. Martin et al.'6 found thatwith conventional angiography, the average skin radi-ation exposure is 17. 1 R and the average thyroid expo-sure is 23 R. From this information, they calculatedthat in pediatric patients with a normal life expectancy,

FIGURE 9. Standiard aortogram (levft) in a 9-year-old patient weighing 27 kg with coarctation of the aorta. We used 5 ml or18.5% of the conventional dose of contrast medium (27 ml) for the digital angiogram (tight).

Vol. 68, No. 2, August 1983 381



ownloaded from


LEVIN et al.

FIGURE 10. Aortograms from the patient in figure 9, but later during the angiogram. Note the descending aorta and the well-dletined collateral vessels in the digital angiogram (rilght) compared with the standard angiogram (left).

the risk of developing thyroid carcinoma as a conse-quence of conventional angiography is about 20 timesgreater than the national incidence (0.7 per thousandcompared with 0.04 per thousand), although a latentperiod in excess of 40 years between irradiation andclinical presentation may be present."' Other investiga-tors have expressed similar tindings.24 '

Radiation exposure may produce injury to other or-

gan systems as well. Adams et al. 2' in evaluatingblood samples taken before and after cardiac catheter-ization with cineangiography in 20 children, foundchromosomal damage in all samples taken after thestudy. In a subsequent study, the dose effect wasshown to be enhanced by the use of a contrast agentduring angiography.) In addition, radiosensitivity isproportional to cell reproduction rate, thus making the

FIGURE I. Digital aortogram in a 14-year-old pa-tient weighing 49 kg, after surgery for aortic valvularstenosis. Note the domed aortic valve and negativejet. We injected 16 ml or 32/c4 of the standard con-trast medium dose (49 ml).

CIRCULATION382



ownloaded from



hematopoietic system and bone marrow highly radio-sensitive. While there are no reports in the literature ofan increased incidence of leukemia in patients whounderwent cardiac catheterization and angiography inchildhood, periodic checkups and continued vigilancewould appear prudent in these patients,27 given thepotential for prolonged latency between radiation ex-posure and resulting neoplasia.Numerous authors have reported damage to thoracic

organs after therapeutic radiation. These include cardi-ac inflammation,33-3 constrictive pericarditis,34 andmyocardial fibrosis.35 36 Since many children may re-quire multiple cardiac catheterizations to manage theirdisease, consideration should be given to the possibleaccumulation of radiation effects during conventionalimaging techniques and to subclinical damage thatmay not become evident until years later.37 38

Digital angiograms are performed with image-inten-sification fluoroscopy. While the tube potential set-tings for DSA, cutfilm angiography, and cineangio-graphy are similar, the tube current settings are muchhigher for cutfilm angiography or cineangiography,varying from 100- to 250-fold greater than that forDSA; as a result, the actual measured radiation expo-sures were far less with DSA than with the two conven-tional angiographic methods (figures 2 and 3) and var-ied from approximately 4% to 10% of the conventionalexposure. Thus, DSA provides greatly reduced radi-ation exposure without loss of required image resolu-tion, thereby reducing possible hazards from radiationduring catheterization.

While the pixel-by-pixel background subtractionmethod of DSA requires that patient movement beminimized to avoid misregistration of pixels and lossof resolution, our results indicate that this may beachieved by adequate sedation'13 14 in preparation of thepediatric patient for cardiac catheterization; in adultsand older children, diaphragmatic movement is re-duced by breath-holding during the preparation of themask and the subsequent angiographic sequence of thearea of interest.

In summary, our data indicate that DSA, when per-formed with an injection either at a distant site such asthe inferior vena cava, or centrally, close to the lesion,usually provides diagnostic information equivalent tothat available from cutfilm angiography and cinean-giography, while considerably less contrast mediumand radiation exposure are required than in the conven-tional methods. Thus, DSA may be used as a screeningtest for patients with complicated lesions or in sickpatients with contraindications for conventional angi-ography; as experience with DSA accrues, it may per-

Vol. 68, No. 2. August 1983

mit increasingly more accurate diagnostic imaging inthe pediatric patient.

We gratefully acknowledge the invaluable technical assis-tance of Margaret Ratigan, R. N., and Sheila Reams, and thesecretarial assistance of Alice Kramer.

References1. Stieghorst MF, Strother CM, Mistretta CA, Crummy AB, Sackett

JF, Lieberman R, Turski P: Digital subtraction angiography. Aclinical overview. AppI Radiol 10: 43, 1981

2. Weinstein MA, Modic MT, Buonocore E, Meaney TF: Digitalsubtraction angiography. Clinical experience at The ClevelandClinic Foundation. Appl Radiol 10: 53, 1981

3. Pond ED, Smith JRL, Hillman BJ, Ovitt TW, Capp MP: Currentclinical applications of digital subtraction angiography. Appl Ra-diol 10: 71, 1981

4. Buonocore E, Meaney TF, Borkowski GP, Pavlicek W, GallagherJ: Digital subtraction angiography of the abdominal aorta and renalarteries. Radiology 139: 281, 1981

5. Christenson PG, Ovitt TW, Fisher DH, Frost MM, Nudelman S,Roehrig H: Intravenous angiography using digital video subtrac-tion: intravenous cervicocerebrovascular angiography. Am J Neu-roradiol 1: 379, 1980

6. Crummy AB, Strother CM, Sackett JF, Ergun DL, Shaw CG,Kruger RA, Mistretta CA, Tumipseed WD, Lieberman RP,Myerowitz PD, Ruzicka FF: Computerized fluoroscopy: digitalsubtraction for intravenous angiocardiography and arteriography.Am J Radiol 135: 1131, 1980

7. Houk TL, Kruger RA, Mistretta CA, Riederer SJ, Shaw CG,Lancaster JC, Fleming DC: Real-time digital K-edge subtractionfluoroscopy. Invest Radiol 14: 270, 1979

8. Meaney TF, Weinstein MA, Buonocore E, Pavlicek W, Bor-kowski GP, Gallagher JH, Sufka B, Maclntyre WJ: Digital sub-traction angiography of the human cardiovascular system. Am JRadiol 135: 1153, 1980

9. Moodie DS, Yiannikas J, Gill CC, Buonocore E, Pavlicek W:Intravenous digital subtraction angiography in the evaluation ofcongenital abnormalities of the aorta and aortic arch. Am Heart J104: 628, 1982

10. Vas R, Diamond GA, Levisman JA, Nakano FH, Neidorf BS,Rose RM, Whiting JS, Forrester JS: Computer enhanced digitalangiography. Clin Cardiol 5: 318, 1982

11. Vas R, Diamond GA, Forrester JS, Whiting JS, Swan HJC: Com-puter enhancement of direct and venous-injected left ventricularcontrast angiography. Am Heart J 102: 719, 1981

12. Yiannikas J, Moodie DS, Buonocore E, Pavlicek W: Digital sub-traction angiography in patients with left-to-right shunts: evalua-tion of structure, function and shunt size. Am J Cardiol 49: 975,1982 (abst)

13. Levin AR: The science of cardiac catheterization in the diagnosis ofcongenital heart disease. Cardiovasc Clin 4: 3, 235, 1972

14. Levin AR: Cardiac catheterization: indications, techniques, inter-pretation, complications. In Radnay PA, editor: Anesthetic consid-erations for pediatric cardiac surgery. Int Anesthesiol Clin 18: 1, p33, 1980

15. Goldberg HL, Borer JS, Moses JW, Fisher J, Cohen B, Skelly N:Digital subtraction intravenous left ventricular angiography: com-parison with intraventricular angiography. J Am Coll Cardiol 1:858, 1983

16. Tobis J, Nacioglu 0, Johnston WD, Seibert A, Iseri LT, Roeck W,Elkayam U, Henry WL: Left ventricular imaging with digital sub-traction angiography using intravenous contrast injection and fluo-roscopic exposure levels. Am Heart J 104: 20, 1982

17. Gootman N, Rudolph AM, Buckley NM: Effects of angiographiccontrast media on cardiac function. Am J Cardiol 25: 59, 1970

18. Levin AR, Grossman H, Schubert ET, Winchester P, Gilladoga A:Effect of angiocardiography on fluid and electrolyte balance. Am JRoentgenol 105: 777, 1969

19. Tay DJ, Levin AR, Schubert E, Baltaxe HA: Contrast media withand without calcium for angiocardiography in children. Acta Ra-diol 17: 449, 1976

383



ownloaded from


LEVIN et al.

20. Shehadi WH: Adverse reactions to intravascularly administeredcontrast media: a comprehensive study based on a prospectivesurvey. Am J Roentgenol 124: 145, 1975

21. Bourassa ME, Noble J: Complication rate of coronary arteriog-raphy: a review of 5,250 cases studied by a percutaneous femoraltechnique. Circulation 53: 106, 1976

22. Morissette M, Gagnon RM, Lamoureux J, Lemire J: Effects ofangiographic contrast media on colloid oncotic pressure. Am HeartJ 100: 319, 1980

23. Fajaro LF, Stewart JR: Experimental radiation-induced heart dis-ease. Am J Pathol 59: 299, 1970

24. Gustafsson M, Mortensson W: Irradiation to the thyroid gland atcardiac catheterization and angiocardiography in children. Br JRadiol 49: 686, 1976

25. Martin EC, Olson A: Radiation exposure to the paediatric patientfrom cardiac catheterization and angiocardiography. Br J Radiol53: 100, 1980

26. Martin EC, Olson AP, Steeg CN, Casarella WJ: Radiation expo-sure to the pediatric patient during cardiac catheterization and an-giocardiography. Emphasis on the thyroid gland. Circulation 64:153, 1981

27. Waldman JD, Rummerfield PS, Gilpin EA, Kirkpatrick SE: Radi-ation exposure to the child during cardiac catheterization. Circula-tion 64: 158, 1981

28. Hempelmann L: Risk of thyroid neoplasms after irradiation inchildren. Science 169: 159, 1968

29. Hall E: Radiobiology for the radiologist, ed 2. New York, 1978,Harper & Row, pp 371-372

30. Maxon HR, Thomas SR, Saenger EL, Buncher CR, Kereiakes JG:Ionizing irradiation and the induction of clinically significant dis-ease in the human thyroid gland. Am J Med 63: 967, 1977

31. Adams FH, Norman A, Bass D, Oku G: Chromosome damage ininfants and children after cardiac catheterization and angiography.Pediatrics 62: 312, 1978

32. Callisen HH, Norman A, Adams FH: Absorbed dose in the pres-ence of contrast agents during pediatric cardiac catheterization.Med Phys 6: 504, 1979

33. Stewart JR, Cohn KE, Fajaro LF, Hancock EW, Kaplan HS: Radi-ation-induced heart disease. Radiology 89: 302, 1967

34. Greenwood RD, Rosenthal A, Cassday R, Jaffe N, Nadas AS:Constrictive pericarditis in childhood due to mediastinal irradia-tion. Circulation 50: 1033, 1974

35. Rubin E, Camara J, Grayzel DM, Zak FT: Radiation-induced car-diac fibrosis. Am J Med 34: 71, 1963

36. Fajardo LF, Stewart JR: Capillary injury preceding radiation-in-duced myocardial fibrosis. Radiology 101: 429, 1971

37. Beach SA, Dolphin GW: A study of the relationship between x-raydose delivered to the thyroid of children and the subsequent devel-opment of malignant tumors. Phys Med Biol 6: 583, 1961

38. Hempelmann LH, Hall WJ, Phillips M, Cooper RA, Ames WR:Neoplasms in persons treated with x-rays in infancy: fourth surveyin twenty years. JNCI 55: 519, 1975

CIRCULATION384



ownloaded from


Skelly and J CarterA R Levin, H L Goldberg, J S Borer, L N Rothenberg, F A Nolan, M A Engle, B Cohen, N T

with standard methods.Digital angiography in the pediatric patient with congenital heart disease: comparison

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1983 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.68.2.374

1983;68:374-384Circulation.

http://circ.ahajournals.org/content/68/2/374the World Wide Web at:

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

http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions:

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

document. Permissions and Rights Question and Answer information about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. FurtherEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:



ownloaded from

http://circ.ahajournals.org/content/68/2/374

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://circ.ahajournals.org//subscriptions/


diagnostic methods congenital heart disease · diagnoses that included valvular regurgitant...

Documents