Volume Number

17.0 1

LV diastolic function by ambulatory scintigraphy

JW, Glaeser DH, Quinones MA, Miller RR. Left ventricular diastolic performance at rest and during exercise in patients with coronary artery disease. Circulation 1981;63:1228-37. 18. Bonaduce D, Morgan0 G, Petretta M, Arrichiello P, Confort G, Betocchi S, Salvatore M, Chiariello M. Diastolic function in acute myocardial infarction: a radionuclide study. J Nucl Med 198&29:1786-g. 19. Bonow RO, Bacharach SL, Green MV, Kent KM, Rosing DR, Lipson LC, Leon MB, Epstein SE. Impaired left ventricular

diastolic filling in patients with coronary artery disease; assessment with radionuclide angiography. Circulation 1981; 64315-23. 20. Fouad FM, Soliminski MJ, Tarazi RC, Gallagher JH. Alterations in left ventricular filling with beta-adrenergenic blockade. Am J Cardiol 1983;51:161-4. 21. Bonow RO. Effects of calcium-channel blocking agents on left ventricular diastolic function in hypertrophic cardiomyopathy and in coronary artery disease. Am J Cardiol1985;55:172B-8B.

Altered systolic and diastolic function in children after “successful” repair of coarctation of the aorta We investigated whether left ventricular (LV) structural or functional abnormalities persist in children on long-term follow-up after successful correction of coarctation of the aorta. Two-dimensional directed M-mode and Doppler echocardlographic examinations were performed on 11 such subjects and 22 age-matched control subjects. Digftized tracings were made from M-mode recordings of the LV and Doppler mitral valve inflow recordings to measure septal, posterior wall, and LV dimensions, LV mass, shortening fraction, peak shortening and lengthening velocities, diastolic filling time, peak E velocity, peak A velocity, and velocity time integrals. Despite group simllarlties in age, body size, and systolic blood pressure, greater fractional shortening (p = O.OOOl), indexed peak shortening velocity fp < O.OOl), and greater LV mass index (p < 0.05) were seen in the coarctation group in the face of lower LV wall stress (p = 0.0001). LV mass index correlated with the resting arm-leg gradient, which ranged from -4 to +lO mm Hg. The coarctation group had decreased early filling (p < 0.006) with compensatory increased late diastolic filling (p < 0.05). Diastolic filling abnormallties were prominent in the older coarctation subjects and were related to both systolic blood pressure (p < 0.001) and LV mass index (p < 0.01). Despite apparently successful repair of coarctation of the aorta, persistent alterations in both systolic and diastolic LV function and LV mass are present in children at long-term follow-up, which are related to the resting arm-leg gradient. We speculate that these small measured arm-leg gradients represent persistent alterations in flow, which may result in LV hypertrophy and hyperkinesia. (AM HEART J 1990;120: 103.)

William Rodrigo

B. Moskowitz, MD, Richard M. Schieken, Bossano, MD. Richmond, Va.

MD, Michael

Increased rates of survival for children with coarctation of the aorta have resulted from both earlier intervention and improved surgical procedures.‘* 2 However, late cardiovascular morbidity and mortality continue to remain significant.3-5 The residual

From the Children’s Medical Medical College of Virginia. Received Reprint ology, 23298. 411120438

for publication

Center,

Oct. 30, 1989;

requests: William B. Moskowitz, Medical College of Virginia, Box

Division of Pediatric Cardiology, accepted

Feb.

20, 1990.

MD, Division of Pediatric Cardi543 MCV Station, Richmond, VA

Mosteller,

PhD, and

cardiovascular abnormalities include hypertension, recurrent coarctation, intracerebral aneurysm, and progressive calcific stenosis of an associated bicuspid aortic valve. Despite repair of coarctation of the aorta, there is a higher incidence of premature ischemic heart disease and sudden death. Left ventricular (LV) diastolic dysfunction has been described preoperatively in children with coarctation of the aorta.6 Results of echocardiographic and radionuclide studies show increased LV performance and persistent myocardial hypertrophy years after coarctation repair. 7, sThe purpose of the present study was to determine whether these preoperative 703

July

104

Moskowitz

et al.

Table

1. Coarctation subjects

Am&can

Subject Sex

NO.

Age (yri

F

1 2 3 4 5 6

16 a 11 5 18 9

M F M M F F

7 8 9 10 11

F

6 7

M

9

6

M F

17 10.3 * 4.5

Mean? SD DPA, Dacron patch angioplasty;

Table

ETE, end-to-end

anastomosis;

SFA, sub&vim

Age at repair (yr)

Repair

10 3 5 4 6 2

ETE ETE SFA ETE ETE DPA

0.8 5 Newborn 4.5 5

1990

Heart Journal

Gradient (mm Hd 0 4 0 -3 10

-4 8

SFA SFA

2 10

ETE ETE ETE

-3 0

4.1 f 2.7

2.2 + 5.1

flap angioplasty.

II. Responsesto exercise stresstesting

Subject

No.

Baseline heart (beatslmin)

1 2 3 5 6 9

102 74 88 80 80 88

rate

Baseline

Duration of exercise (min)

Peak heart rate

(mm Hg) 124/N 136171 107172 144170 106167 108172

9 11 9 10 9 7

192 178 152 202 182 166

blood pressure

LV functional disturbances persist after apparently successful repair, and if so which mechanisms appear likely to be responsible for these abnormalities. METHODS Study population.

Children with surgically repaired coarctation of the aorta who met the following criteria were evaluated: (1) isolated coarctation, (2) at least 1 year since surgicalrepair, (3) no transvalvular aortic gradient, (4) no aortic insufficiency, (5) normal mitral valve anatomy and function, (6) resting arm-leg systolic gradient lessthan or equal to 10 mm Hg, (7) normal resting arm blood pressure, and (8) not receiving any medications. Healthy normotensive children who had normal physical examination findings and normal cardiac structures by two-dimensional echocardiography served as control subjects. This study was approved by our institutional review board, and informed parental consentwas obtained before the study. The 11 coarctation subjects (six girls and five boys) are describedin Table I. The meanageat evaluation was10.3 years (range 5 to 18 years). Coarctation repair was performed at a mean age of 4.1 years (range newborn to 10 years). Subjects were studied after a mean duration of 6 years (range 1 to 12 years) after surgical repair. Surgical procedures consistedof excision and end-to-end anastomosisin seven, subclavian flap angioplasty in three, and Dacron patch angioplasty in one. The meanarm-leggradient was 2.2 mm Hg and ranged from -4 to 10 mm Hg. No

(beatslmin)

Peak

blood pressure (mm f&9 180/74 183/72 216/232/108 156168 160/-

relationship was seenbetween the type of repair and the resting arm-leg gradient. Six of the coarctation subjects(Nos. 1,2,3,5,6,and 9) had undergone recent exercise stress testing by means of a standardized treadmill or bicycle protocol. Heart rate and blood pressureresponsesto exerciseare shownin Table II. No overt systolic hypertensive responseto maximal exercisewasseenin any of the subjects.Four subjectshad positive arm-leg gradients (mean 29.0 mm Hg, range 15 to 56 mm Hg) measuredimmediately after exercise.One subject had diastolic hypertension. Anthropometrics and blood pressure. Height and weight of eachsubject in stocking feet were measuredwith a stadiometer and digital scale, respectively. Replicate blood pressuremeasurementswereobtained with the subject in the sitting position by means of a mercury sphygmomanometer and the appropriately sized compression cuff. The fourth Korotkoff phasewasrecorded asthe diastolic blood pressure.Supine arm-leg systolic gradient was determined by blood pressurecuff and vascular Doppler probe. Echocardiography. Supine M-mode echocardiograms of the left ventricle were generated from two-dimensional images(with a 2.5 MHz transducer) with the cursor line positioned at the free edgesof the mitral valve leaflets in the parasternal long-axis view. Tracings recorded at a paper speedof 50 mm/set were digitized with a commercially available hand-held cross-wirecursor and electronic coor-

Volume

120

Number

1

dinate digitizer. Multiple (three to five) cardiac cycles were traced for each subject, and only cycles with well-defined continuous surfaces were used. The heart rate, LV minor axis dimension, and rate of change of this dimension were obtained. The septum, posterior wall, and LV dimensions were measured at end diastole at the onset of the QRS wave and at end systole. The LV dimension at end systole was measured as the smallest systolic dimension. LV mass was calculated and indexed for body size.g End-systolic LV wall stress was calculated by means of a simplified formula.iO The fractional shortening and peak velocities of LV dimension shortening and lengthening were determined. To compare instantaneous velocities for ventricles of different sizes, the peak shortening velocity was normalized for chamber size by dividing it by end-diastolic dimension. Peak lengthening velocity was normalized by dividing it by the change in chamber dimension from diastole to systole.il Immediately after M-mode examination, phased-array pulsed Doppler examination of the LV inflow tract was performed. By means of the apical four-chamber view, transmitral flow velocities were recorded at the level of the mitral valve anulus. The sample volume position was adjusted to record the maximal velocity through the mitral valve. Spectral blood flow velocities were recorded at a paper speed of 100 mm/set. A Doppler tracing was considered adequate for analysis only when the mitral valve opening and closure points and peak velocities at rapid filling (E) and atria1 contraction (A) could be identified for at least three cardiac cycles. Diastolic filling time was defined as the number of milliseconds from the opening to the closing points. From digitized Doppler spectral tracings, peak E velocity, peak A velocity, and velocity time integrals (VTI), the areas under the Doppler tracings, were determined.12 Data analysis. All M-mode and Doppler echocardiographic measurementswere the average of at least three cardiac cycles, and values are presentedasmean 2 1 standard deviation. Becausethe variancesfor all variables were similar in both groups, statistical comparisonsbetween groups were made by meansof an unpaired t test. A twotailed test of significancewith p < 0.05wasusedto indicate a significant difference between groups and a significant correlation betweenvariables. Data wereadjusted for heart rate by meansof SAS’s General Linear Models procedure (SAS Institute, Cary, N.C.) with subsequentcomparisonof least-squaremeans.In addition, partial correlation coefficients with heart rate held constant were usedto describe the strength of the associationof resting arm-leg gradient and echocardiographic variables Stepwise regressionproceduresand linear modelswere used to assess the proportion of the variation in LV massindex and VT1 of the A wave that could be explained by blood pressureand other measuredvariables. The beat-to-beat and intraobserver variation coefficients for Doppler diastolic measurementswere calculated from three successivecardiac cycles of five studies (three normal subjects and two coarctation subjects). The beatto-beat and intraobserver variation coefficients were: peak E velocity, 2.0% and 1.9% ; peak A velocity, 3.3% and

Diastolic

Table

in aortic

coarctation

105

Ill. Baseline characteristics

Characteristics Age (~4 Weight (kg)

BSA (m2) SBP DBP

dysfunction

(mm Hg) (mm Hg)

Normal (n = 22) 11.7 44.5 1.36 105.2 59.3

f f + f +

0.4 0.5 .14 4.4 9.3

Coarctation (n = 11) 10.3 36.5 1.13 109.6 69.4

* k i * k

4.5 21.3 .40 8.9 5.4

p Value NS NS NS NS