Comparison of the right and left ventricle as a systemic ventricle during exercise in patients with congenital heart disease

Valvular and Congenital Heart Disease

Comparison of the right and left ventricle as a systemic ventricle during exercise in patients with congenital heart disease Hideo Ohuchi, MD,a Yoshimi Hiraumi, MD,a Hiroshi Tasato, MD,a Atsushi Kuwahara, MD,a Hiroshi Chado, MD,a Keiko Toyohara, MD,a Yoshio Arakaki, MD,a Toshikatsu Yagihara, MD,b and Tetsuro Kamiya, MDa Osaka, Japan

Background Few studies have investigated the clinical advantages of surgical correction with the morphologic left ventricle (MLV) instead of the morphologic right ventricle as a systemic ventricle (SV) in patients with congenital heart disease.

Methods Twenty-four healthy control subjects (group A1), 6 patients with isolated congenitally corrected transposition of the great arteries (TGA) (group A2) , 16 patients with TGA who had undergone an arterial switch operation (group B1), 18 patients with TGA who had undergone a venous switch operation (group B2), 9 patients with atrioventricular and ventriculoarterial discordance who had undergone a double switch operation (group C1), and 6 patients with atrioventricular and ventriculoarterial discordance who had undergone a conventional external conduit operation from the MLV to the pul· ), and oxygen monary artery (group C2), performed treadmill exercise testing. Their heart rate (HR), oxygen uptake (VO 2 pulse (O2 pulse), which reflects individual stroke volume, were measured, and contractile function was assessed by echocardiography.

Results The peak HR for the patients after a definitive operation were significantly lower than that in group A1 and was

· · correlated with peakVO 2 (r = .67, P < .0001). The peakVO2 and peak O2 pulse for the groups A2 and B2 were significantly lower than those for the groups A1 and B1, respectively. The peak O2 pulse data were strongly correlated with those · of peakVO 2 (r = 0.91, P < .0001). The left ventricular ejection fraction was significantly lower in groups B1 and C1 than in · · group A1 and was correlated with peakVO 2 (r = .50, P < .01). No significant differences inVO2, HR, and O2 pulse at peak exercise were observed between groups C1 and C2.

Conclusions Chronotropic incompetence and an impaired response of the stroke volume of the MRV during exercise are partly responsible for the reduced exercise capacity in groups A2 and B2 compared with groups with the MLV as an SV, and the SV function at rest is also related to exercise capacity. Superiority of the double-switch operation compared with the conventional conduit operation was not observed. A longer-term follow-up is necessary before the advantages of these 2 operations can be compared. (Am Heart J 1999;137:1185-94.)

Disadvantages of the morphologic right ventricle (MRV) as a systemic ventricle (SV) have been reported, including myocyte, geometry, and the function of the tricuspid valve.1 Because progressive dysfunction of the MRV and tricuspid valve as an SV and a systemic valve, respectively, and an increasing incidence of arrhythmias have also been reported,2 a definitive operation with the morphologic left ventricle (MLV) has been developed for patients with transposition of the great arteries (TGA) or atrioventricular and ventriculoarterial discorFrom the aDepartment of Pediatrics and the bDepartment of Thoracic Surgery, National Cardiovascular Center. Submitted January 20, 1998; accepted July 23, 1998. Reprint requests: Hideo Ohuchi, MD, Department of Pediatrics, National Cardiovascular Center, Fujishiro-dai, Suita, Osaka 565, Japan. Copyright © 1999 by Mosby, Inc. 0002-8703/99/$8.00 +0 4/1/94539

dance (AVD).3 However, few studies have evaluated whether the procedures using MLV as an SV affect the cardiorespiratory response to exercise compared with the procedures using the MRV as an SV.4 Therefore we evaluated whether the different types of SV, that is, MRV and MLV, influence the individual cardiorespiratory response and exercise capacity in a variety of patients with congenital heart disease, those with isolated congenitally corrected TGA, those with TGA who had an atrial or arterial switch operation, and those with AVD who had a double-switch operation or conventional functional conduit repair.

Methods Subjects

In this investigation of the influence of MRV and MLV as SVs on cardiorespiratory responses during exercise, patients with a

American Heart Journal June 1999

1186 Ohuchi et al

Table I. Clinical characteristics of the study groups No surgery

SV Diagnosis or surgery Cases Age (y) Height (cm) Weight (kg) Follow-Up (y) Physiologic PS

After definitive repair

A1

A2

B1

B2

MLV Control 24 12 ± 4 143 ± 14 39 ± 11 — —

MRV c-TGA 6 13 ± 7 140 ± 18 34 ± 11 — —

MLV ASO 16 11 ± 2 138 ± 12 33 ± 8 10 ± 2 27 ± 15

MRV VSO 18 11 ± 4 138 ± 18 35 ± 15 10 ± 4 —

C1 MLV DSO 9 10 ± 2 134 ± 10 29 ± 8 4±3 24 ± 0

C2 MRV ECR 6 14 ± 3 152 ± 15 41 ± 14 7±3 26 ± 16

Values are mean ± SD. c-TGA, Isolated congenitally corrected TGA; DSO, double-switch operation; ECR, extracardiac conduit repair from the left ventricle to the pulmonary artery; PS, pulmonary stenosis.

pacemaker, patients with significant physiologic pulmonary stenosis (pressure gradient ≥30 mm Hg during cardiac catheterization or transpulmonary valve flow velocity ≥3.0 m/s on Doppler echocardiography), and those taking a βblocker were excluded from the study. Systemic ventricular valve regurgitation (SVVR; tricuspid valve regurgitation in patients with MRV and mitral valve regurgitation in patients with MLV) was assessed by color flow mapping, and was graded qualitatively as no-to-slight SVVR, moderate SVVR, and severe SVVR. SVVR was considered no-to-slight if the regurgitant jet crossed less than one third of the systemic atrium from the SVV orifice, moderate if it projected from one third to two thirds of the depth of the systemic atrium, and severe if it reached beyond two thirds of the systemic atrium with a significantly wide jet.Aortic valve regurgitation was also assessed, as was the SVVR, by color flow mapping. Patients with severe SVVR, moderate-to-severe aortic valve regurgitation, and who had received a valve replacement were also excluded.The grade of the SVVR (tricuspid regurgitation) in 6 patients with isolated congenitally corrected TGA (group A2) was no-toslight SVVR. Sixteen of 42 patients with TGA who had undergone an arterial switch operation (ASO) (group B1), and 18 of 23 patients who had undergone a venous switch operation (VSO) (group B2) at our hospital were included in this study. The ages of the group B1 patients at the time of ASO ranged from 0.3 to 2.4 years, with a mean of 0.8 years, except for 2 patients.These 2 patients each required a reoperation for pulmonary stenosis after ASO at 3.7 and 5.9 years of age. The ages of the group B2 patients at the time of VSO ranged from 0.2 to 1.1 years, with a mean age of 0.7 years, except for one patient who required a reoperation for stenosis of the superior vena cava after VSO. Fourteen of these 18 patients had undergone a Mustard operation, and the other 4 patients had undergone a Senning operation. Nine of 18 patients with AVD who had undergone a double-switch operation (group C1) and 6 of 14 patients with AVD who had undergone an external conduit operation from the MLV to the pulmonary artery (group C2) at our hospital were included in this study. The ages at the time of definitive operation in the group C1 patients ranged from 3.3 to 8.6 years, with a mean of 5.7 years, and that in the group C2 patients ranged from 1.8 to 12.7 years, with a mean of 7.0 years.The follow-up periods are shown in Table I.Although there was no significant difference in the follow-up periods between groups B1 and B2, the

period was significantly longer in group C2 than in group C1.All 55 of these patients were in New York Heart Association functional class I.Twenty-four age- and body-size–matched patients with a history of Kawasaki disease who had no coronary arterial stenotic lesions served as control subjects (group A1) because they had similar cardiorespiratory responses to those of healthy control subjects during exercise testing.5 The magnitude of the physiologic pulmonary stenosis was not significantly different among groups B1, C1, and C2 as shown in Table I. No patients in groups A2 and B1 were taking any medicines; one patient of group B2 was taking diuretics. In group C1, 2 patients were taking digoxin, 3 were taking diuretics, 2 were taking antiplatelet agents, and 1 was taking an antiarrhythmic agent. In group C2, 3 patients were taking digoxin, 4 were taking diuretics, and 2 were taking antiplatelet agents.

Assessment of cardiac function by echocardiogram In all 79 patients, systemic ventricular contraction, left ventricular diastolic dimension (LVDd) if the SV was an MLV, SVVR, aortic valve regurgitation, pulmonary valve stenosis, and regurgitation were evaluated by echocardiography.The ejection fraction (EF) of the MLV in groups A1, B1, and C1 was calculated by Pombo’s method.6 Because of the difficulty in evaluating the ventricular function of the MRV, the contractility of the MRV was assessed retrospectively based on the echocardiographic reports and graded qualitatively as good, reduced, or poor. SVVR was evaluated as mentioned.

Pulmonary function tests Thirty-nine of the 79 patients (group A1 = 10, group A2 = 5, group B1 = 10, group B2 = 5, group C1 = 5, and group C2 = 4) underwent pulmonary function tests.The vital capacity (VC) and the forced expired volume in 1.0 second (FEV1) (Spirosift, SP-600, Fukuda Denshi,Tokyo, Japan) were measured.The VC values were expressed as a sex-matched percentage of the normal body height predicted value (%VC).

Exercise protocol All patients performed a ramp-like progressive exercise test on a treadmill (Q-5000 system, Quinton, Seattle,Wash). This exercise protocol has been described extensively in a previous study.7 In brief, we demonstrated a high correlation between the ventilatory anaerobic threshold (AT) and the lactate threshold using this treadmill test and estab-

American Heart Journal Volume 137, Number 6

Ohuchi et al 1187

Table II. Pulmonary function and echocardiographic findings of the study groups A1 Pulmonary function VC (mL) Pred-VC (%) FEV1 (%) Echocardiography LVDd (mm) Pred-LVDd (%) LVEF (%) RV contraction Good Reduced Poor SVVR No-slight Moderate Severe

2822 ± 266 100 ± 4 87 ± 3

A2

1948 ± 159 96 ± 14 89 ± 2

B1

B2

C1

1899 ± 160 86 ± 4 92 ± 2

2170 ± 215 95 ± 14 90 ± 2

1748 ± 287 78 ± 6 90 ± 2 42 ± 2 103 ± 6 63 ± 5*

C2

2385 ± 510 69 ± 7 89 ± 2

40 ± 1 94 ± 2 78 ± 2

— — —

42 ± 2 100 ± 4 65 ± 2*

— — —

— — —

— — —

6 0 0

— — —

8 10 0

— — —

2 4 0

24 0 0

6 0 0

15 1 0

14 4 0

9 0 0

3 3 0

Values are mean ± SE. FEV1, Forced expired volume in 1 second; Pred-LVDd, percent of normal for predicted LVDd value; Pred-VC, percent of normal for predicted VC value. *P < .05 vs group A1.

lished its clinical usefulness for determining the AT and evaluating cardiorespiratory tolerance in patients with congenital heart disease.The exercise intensity was increased by 0.7 metabolic units every 30 seconds, with completion of the incremental part of the exercise test in approximately 10 minutes.After a 4-minute rest, the patients performed a 3minute warm-up walk at a speed of 1.5 km/hr and then exercised with progressive intensity until exhaustion. Endurance time (ET) (in minutes) was defined as the duration of ramp exercise according to the protocol.

Heart rate and blood pressure measurements Twelve standard electrocardiogram leads were placed to monitor the heart rate (HR), and systolic blood pressure (SBP) was measured every 2 minutes during the exercise testing. Because it is difficult to measure blood pressure with a mercury sphygmomanometer during dynamic exercise, especially in younger patients, we measured SBP by the palpation method. In our preliminary study, the SBP obtained by this method in 12 patients with cardiac disease during treadmill exercise testing correlated well with the measurements obtained with a mercury sphygmomanometer (r = 0.98, P < .0001, Ohuchi et al, unpublished data, 1996).

Gas exchange measurements Ventilation and gas exchange were measured breath by breath. Subjects breathed through a mask connected to both a hot-wire anemometer (Riko AS500, Minato Medical Science, Osaka, Japan) for the continuous measurement of inspired and expired volume, and to a mass spectrometer (MG-300, Perkin Elimer), for the continuous measurement of oxygen and carbon dioxide partial pressures.We used 2 sizes of masks, one for children whose height was less than 150 cm that had a dead space of 80 mL, and the other for children whose height was more than 150 cm that had a dead space of 100 mL. Derived respiratory parameters, including the minute ventila-

· ventilatory equivalent for oxygen and carbon dioxide tion (VE), · VCO · · VO · ,VE/ (VE/ 2 2), and respiratory gas exchange ratio were computed in real time and displayed with the HR and oxygen · ) on a monitor during the exercise testing with the uptake (VO 2 use of a personal computer (PC-9801, NEC,Tokyo, Japan).A 30second moving average for breath-by-breath data was applied to provide smooth data during the exercise testing, and these averaged data were used for the analysis.The delay times and response characteristics of the oxygen and carbon dioxide analyzers were carefully checked before each session of exercise testing by sampling the square wave change in the oxygen and carbon dioxide concentration between calibrating gases.

Calculations The oxygen pulse (O2 pulse; mL/min/beats) was calculated · by dividing the VO 2 by the HR and is equal to the product of the stroke volume and the arterial-venous oxygen difference. Thus this value is related to stroke volume and oxygen extraction. Because the magnitude of oxygen extraction at peak exercise varies little in patients and normal subjects8 and the magnitude of increase in stroke volume more than 60% of the individual maximal exercise capacity is small,9,10 an O2 pulse at above moderate exercise intensity, especially at peak exercise, mainly reflects an individual stroke volume.According to the data of a report by Ensing et al,11 the correlation coefficient between the peak O2 pulse and stroke volume at peak exercise is high (r = 0.834, P = .0001).The stroke volume is an important determinant of individual exercise capacity.12 However, it is not easy to measure this variable during exercise. Conversely, it is relatively easy to measure the changes in O2 pulse during exercise noninvasively. Because of the only slight increase in stroke volume during moderate-to-severe exercise, changes in the O2 pulse during the period from AT to peak exercise mainly reflected the change in stroke volume during the same period. · VCO · · ,VE/ The values for SBP,VO 2 2, and O2 pulse obtained in

American Heart Journal June 1999

1188 Ohuchi et al

Table III. Cardiorespiratory variables at peak exercise of the study groups

ET (min) (% of normal) HR (beats/min) (% of normal) SBP (mm Hg) (% of normal) Oxygen uptake (mL/kg/min) (% of normal) Oxygen pulse (mL/beats) (% of normal) · VCO · VE/ 2 (% of normal)

A1 (n = 24)

A2 (n = 6)

B1 (n = 16)

B2 (n = 18)

C1 (n = 9)

C2 (n = 6)

9.4 ± 1.1 (99 ± 13) 194 ± 6 (102 ± 3) 150 ± 18 (100 ± 9) 45 ± 7 (101 ± 14) 8.7 ± 2.2 (100 ± 14) 36 ± 4 (101 ± 14)

9.2 ± 1.4 (101 ± 17) 186 ± 9† (98 ± 6) 149 ± 16 (104 ± 6) 39 ± 4† (86 ± 13)† 6.9 ± 2.1ll (89 ± 13)ll 42 ± 5† (113 ± 12)†

8.9 ± 0.8 (95 ± 8) 166 ± 10 (87 ± 6) 149 ± 16 (103 ± 9) 36 ± 4 (78 ± 9) 7.0 ± 1.4 (90 ± 8) 38 ± 5 (102 ± 14)

8.0 ± 1.4* (84 ± 16)* 170 ± 18 (89 ± 10) 140 ± 19 (96 ± 13)‡ 33 ± 6 (70 ± 12)* 6.7 ± 2.4 (79 ± 13)# 41 ± 6‡ (112 ± 12)*

8.4 ± 1.0 (87 ± 9) 163 ± 9 (85 ± 4) 135 ± 9 (96 ± 4) 32 ± 5 (65 ± 8) 5.6 ± 1.3 (75 ± 7) 44 ± 7 (118 ± 20)

7.5 ± 2.0 (78 ± 22) 166 ± 15 (88 ± 8) 153 ± 15§ (101 ± 8) 31 ± 7 (69 ± 16) 7.2 ± 1.8¶ (79 ± 16) 38 ± 10 (111 ± 21)

Values are mean ± SD. *P < .05 vs group B1. †P < .05 vs group A1. ‡P < .1 vs group B1. §P < .1 vs group C1. ll< .1 vs group A1. ¶P < .05 vs group C1. #P < .01 vs group B1.

this protocol are expressed as a sex-matched percentage of the · VCO · · , %VE/ normal body weight–predicted value (%SBP, %VO 2 2, and %O2 pulse).The HR values are expressed as a sex-matched percentage of the normal age-predicted value (%HR).The normal values were obtained from 125 healthy subjects.The AT was · VO · and end-tidal pressure · at which the VE/ defined as the VO 2 2 · VCO · for oxygen increase without a rise in either VE/ 2 or endtidal pressure for carbon dioxide,13 or defined by the V-slope method.14 This threshold corresponds to the metabolic rate above which anaerobic metabolism supplements the production of aerobic energy production and leads to lactic acidosis. Informed consent was obtained from all patients and their parents.This protocol was approved by the Ethics Committee of the National Cardiovascular Center.

Statistical analysis Differences between pairs of groups in mean cardiorespiratory values during exercise were assessed by a 2-way repeated measure analysis of variance. If the differences were significant, individual comparisons were made with the unpaired t test.A simple regression analysis was used to correlate the HR and O2 · pulse values at peak exercise with the peak VO 2 and when appropriate. Differences in frequencies were analyzed with the chi-square or Fisher’s exact test. Data are expressed as mean ± SD or SE. P values < .05 were considered significant.

Results Echocardiographic findings and pulmonary function at rest The echocardiograhic findings are summarized in Table II.Although there was no significant difference in LVDd among groups A1, B1, and C1, the LVEF was significantly less in groups B1 and C1 than in group A1 (P < .05). Even though the assessment was not

quantitative analysis, the percentage of patients with reduced function of the MRV was greater in groups B2 and C2 than in group A2, in which no patients with reduced MRV function were observed (group A2, 0 of 6; group B2 and C2, 14 of 24, P < .001).There were no patients with moderate SVVR in groups A1 and A2; however, 8 (16%) of the 49 patients who had undergone a definitive operation showed moderate SVVR in patients (P < .001). The data of pulmonary function at rest are also summarized in Table II. In the comparison of the 3 groups with the MLV as an SV, %VC was significantly smaller in groups B1 and C1 than in group A1 (P < .05).Among the 3 groups with the MRV as an SV, the %VC in group C2 was smallest; however, there was no significant difference among the groups, probably because of the small numbers of patients. No significant difference in FEV1 was observed among the study groups. •

ET and peak VO2 The cardiorespiratory data at peak exercise are shown in Table III.There was no significant difference in ET between groups A1 and A2.The ET was significantly shorter in group B2 than in group B1 (P < .05), and the ET was shorter in group C2 than in group C1, but this difference did not reach significance. · · Both peak VO 2 and %peak VO2 were significantly lower in group A2 than in group A1 (P < .05).Although · there was no significant difference in peak VO 2 · between groups B1 and B2, the %peak VO2 was significantly lower in group B2 than in group B1. However, no significant difference in either parameter was observed between groups C1 and C2.

American Heart Journal Volume 137, Number 6

Ohuchi et al 1189

Figure 1

Relations between percentage of normal predicted value for HR (%pred-HR) and exercise intensity in study groups. Open circle, Group A1; closed circle, group A2; open square, group B1; closed square, group B2; open triangle, group C1; closed triangle, group C2.

Cardiopulmonary function at rest and exercise capacity

· We observed that VC (in milliliters) was correlated with VO 2 · (mL/min) at the AT and peak exercise (AT; r = 0.79, peak VO2; r = 0.83, P = .0001).Although no correlation between LVDd and · · peak VO 2 was observed, LVEF was correlated with peak VO2 · and %peak VO2 (mL/kg/min) in groups A1, B1, and C1 (peak · ; r = 0.44, P < .01).When the 30 · ; r = 0.40, %peak VO VO 2 2 patients with the MRV as an SV were analyzed separately, the · · peak VO2 and %peak VO2 were significantly greater in the 16 patients whose MRV function was evaluated as good compared with the other 14 patients with reduced MRV function · : · : good, 36.2 ± 5.2; reduced, 30.8 ± 6.4; %peak VO (peak VO 2 2 good, 78.8 ± 16.5; reduced, 66.4 ± 8.0; P < .05, respectively). The SVVR in 4 of the 18 patients of group B2 was evaluated as · moderate, and these patients showed low mean peak VO 2 compared with the other 14 patients of group B2 patients with good systemic ventricular valvular function; however, no significant difference was observed between these 2 groups.

HR and SBP The %HR at rest, at the AT, and at peak exercise in all groups are shown in Figure 1. Peak HR was significantly lower in group A2 than in group A1, and %HR tended to be lower in group A2 than in group A1 (P < .1). Although there were no significant differences in HR and %HR during exercise between groups B1 and B2, both values at each exercise intensity were significantly lower in groups B1 and B2 than in group A1 (rest, P < .05;AT and peak, P < .001).There was also no significant difference in HR between groups C1 and C2, but both HR and %HR were significantly lower in groups C1 and C2 than in group A1 (rest, P < .01;AT and peak, P < .001). In addition, the magnitude of increase in HR was also impaired and the %HR decreased significantly as the exercise intensity increased.

There was no significant difference in SBP between the comparative groups of A1 and A2, B1 and B2, and C1 and C2.

Ventilatory equivalent for carbon dioxide • • ( VE/ VCO · 2)

· VCO · · · Both VE/ 2 and %VE/ VCO2 at peak exercise were significantly lower in group A1 than in group A2 (P < .05); that is, the ventilatory efficiency was greater in group A1 than in group A2.Although there was no sig· VCO · nificant difference in VE/ 2 between groups B1 and · · B2, %VE/VCO2 was significantly higher in group B2 than in group B1 (P < .05); that is, the ventilatory efficiency was impaired in group B2 compared with group B1. However, no significant difference in either value was observed between groups C1 and C2. · VCO · In all patients,VE/ 2 at the AT and peak exercise were inversely correlated with VC (AT level: r = –0.57, P < .01; peak level: –0.40, P = .05) and were also inversely correlated with LVEF in the 49 patients with MLV as an SV (r = –0.42, P < .01).

O2 pulse The %O2 pulse at rest, at the AT, and at peak exercise in all groups are shown in Figure 2.The peak O2 pulse and %O2 pulse tended to be lower in group A2 than in group A1 (P < .1).The %O2 pulse was significantly lower at peak exercise than at the AT in group A2. Although there was no significant difference in O2 pulse at each exercise intensity between groups B1 and B2, the O2 pulse at peak exercise was significantly lower in group B2 than in group A1. No significant change in %O2 pulse was observed between groups B1 and A1. However, the patients of group B2 showed a sig-

American Heart Journal June 1999

1190 Ohuchi et al

Figure 2

Relations between percentage of normal predicted value for O2 pulse (%pred-O2 pulse) and exercise intensity in study groups. Open circle, Group A1; closed circle, group A2; open square, group B1; closed square, group B2; open triangle, group C1; closed triangle, group C2.

nificant progressive decrease in %O2 pulse as the exercise intensity increased, and as a result the %O2 pulse at peak exercise was significantly lower in group B2 than in group B1 (P < .05), and was significantly lower in groups B1 and B2 than in group A1 (P < .05).The increase in O2 pulse during exercise was impaired in groups of C1 and C2 compared with group A1, and the peak O2 pulse was significantly lower in group C1 than in group A1 (P < .01).There was no significant difference in the change in %O2 pulse during exercise between groups C1 and C2, and the %O2 pulse of both groups decreased as the exercise intensity increased; as a result, the peak value was significantly lower in both C groups than in group A1 (P < .001). •

Relations between peak VCO · 2, peak HR, and peak O2 pulse The %HR and %O2 pulse at peak exercise were corre· lated with %peak VO 2 (%peak HR, r = 0.67; %peak O2 pulse: r = 0.91, P < .0001 for both; Figure 3). Because there was a strong correlation between the %O2 pulse · , we suggest that the at peak exercise and %peak VO 2 stroke volume at peak exercise had a marked effect on the individual exercise capacity. •· Age at definitive operation and peak VCO · 2 and peak O2 pulse

Although age at the time of definitive operation did not · in any of the patients of groups correlate with peak VO 2 B1 and B2 except for the 3 who required reoperation, the peak O2 pulse was inversely correlated with age at the

time of definitive operation (r = –0.47, P < .01). In the 2 · was groups of patients with AVD, peak HR and peak VO 2 correlated with age at the time of definitive operation · : r = –0.67, P < .01; peak HR: r = -0.60; P < .05). (peak VO 2

Arrhythmias during exercise testing All patients were in sinus rhythm during exercise. Two patients of group A2 had arrhythmias; one had rare monofocal premature ventricular contractions (PVCs); the other, aged 27 years, had both supraventricular premature contractions (SVPCs) and PVCs with a few couplets and triplets during exercise, but these disappeared immediately after exercise. Four patients of group B1 had arrhythmias; one had SVPCs and 3 had PVCs.These PVCs decreased in frequency during exercise in 2 patients and appeared after exercise in 1. Four patients of group B2 had both SVPCs and PVCs during exercise; 2 had SVPCs and 4 had PVCs. One of these 4 patients had multiform PVCs during exercise, but short run was not observed. Intermittent sinus rhythm with junctional escape was observed in 2 patients of group B2, but stable sinus rhythm began the minute exercise testing started. In group C1, one patient had SVPCs with a short run after exercise.Two patients had PVCs and 1 of these 2 patients had ventricular bigeminy after exercise. In group C2, 2 patients had SVPCs and 1 had PVCs. Despite the variable arrhythmias, the exercise testing did not need to be terminated because of arrhythmia in any patients.

Discussion A very important issue for pediatric cardiologists and cardiac surgeons is the question of how long the MRV

American Heart Journal Volume 137, Number 6

Ohuchi et al 1191

Figure 3

Correlations between percentage of normal predicted value for peak oxygen uptake (%pred-peak VO2) and percentage of the normal predicted value for HR (%pred-HR) and O2 pulse (%pred-O2 pulse) at peak exercise. Open circle, Group A1; closed circle, group A2; open square, group B1; closed square, group B2; open triangle, group C1; closed triangle, group C2.

can function as an SV.An experimental study demonstrated that the canine MRV has mechanical and energetic properties similar to those of the MLV.15 In addition to the morphologic aspects, it is likely that the properties of the tricuspid valve and conduction system also influence the cardiac function of the MRV over a long time period as an SV. Because of a geometric problem, it is clinically difficult to evaluate the function of the MRV in human beings; radionuclide angiographic assessments of the MRV function, that is, ventricular volume and EF of the ventricle, have been applied for this problem. Because preoperative myocardial hypoxia or intraoperative myocardial ischemia may contribute to abnormal systemic ventricular function in patients with TGA,16-19 investigating patients with isolated congenitally corrected TGA may enable us to estimate, to some extent, the potential of the MRV as an SV over a long period of time. Graham et al demonstrated the maintained systemic ventricular function of the MRV in infants and children with congenitally corrected TGA, and suggested the possibility of a progressive decrease in systemic ventricular EF with age.17 Taking the property of the larger volume of the MRV compared with the MLV into consideration,17 it is open to discussion whether the relatively low EF of the MRV compared with the MLV truly indicates a decreased function of the MRV in such patients with TGA.20 According to the report by Benson et al,21 although the EF of the MRV at rest was slightly lower in patients with congenitally corrected TGA than in normal subjects, data from these patients showed similar increases in the EF of the MRV (20% to 30%) compared with those in control subjects.

Thus Benson et al21 considered that the MRV functioned appropriately as an SV. However, Peterson et al22 demonstrated no increase in the EF of the MRV during exercise in patients with congenitally corrected TGA. Some investigators have reported that, although their subjects were patients who had undergone a VSO, this relatively low EF of the MRV was compensated by the relatively large volume of the MRV,22,23 whereas other investigators showed no difference in the volume of the SV between patients after a VSO and normal subjects.11 In our study, although the contractile function of the MRV was relatively maintained and significant SVVR was not observed in congenitally corrected TGA, we postulated that the stroke volume decreased from the AT to peak exercise in patients with congenitally corrected TGA because the %O2 pulse significantly decreased.Thus although our subjects were young (mean age of 13 years), the MRV had a certain disadvan· tage as an SV over the MLV because the peak VO 2 was significantly lower in the patients with congenitally corrected TGA than in the control subjects. The introduction of the VSO, that is, the Senning and Mustard procedures, dramatically improved the prognosis of patients with TGA24,25; however, the first choice of surgical procedure has been an ASO because of concerns about postoperative morbidity, including baffle obstruction, tricuspid regurgitation, arrhythmias, and MRV dysfunction.Although an advantage of the ASO over VSO was reported,26 there have been no clinical studies of the superiority of the procedure with an MLV as an SV to that with an MRV as an SV in terms of cardiorespiratory response to exercise. Minamisawa et al4 observed no significant difference in cardiovascular

1192 Ohuchi et al

response to exercise between patients who had undergone an ASO and those who had undergone a VSO. However, it is possible that there was no significant difference because their subjects were very young (mean age of 6 years). Moderate chronotropic incompetence and impaired exercise capacity were observed in both the ASO and VSO patients in that study, and the study authors speculated that postoperative morbidity, including cardiac dysfunction during exercise, and being unfamiliar with exercise might be responsible for impaired exercise capacity in such patients.There are many reports about abnormal cardiorespiratory responses to exercise in patients after a VSO.4,18,19,22,23,27-33 Parrish et al18 speculated that intermittent episodes of myocardial hypoxia, including intraoperative hypoxia or ischemia, and the intrinsic geometry of the MRV might contribute to an abnormal response of the EF of the MRV in patients after a Mustard operation. However, because there was no correlation between age at surgical correction and MRV dysfunction, Murphy et al34 suggested that the duration of hypoxia is not the sole explanation of MRV dysfunction in patients after a VSO. In our study, the %O2 pulse was greater in the patients who underwent a VSO than in the patients after an ASO at rest and decreased gradually as the exercise intensity increased; as a result, this change during exercise was significantly different from that in the ASO patients and the value at peak exercise was also significantly lower in the VSO patients than in the ASO patients.The %O2 pulse was significantly lower in the ASO patients than in the control subjects, and the progressive decrease in %O2 pulse during exercise in the VSO patients, which was relatively maintained at peak exercise in the patients with congenitally corrected TGA, suggests that some kinds of surgical damage to the SV and progressive MRV dysfunction during exercise is marked in patients after a VSO compared with patients with congenitally corrected TGA. In fact, the LVEF in the ASO patients was reduced compared with that of the control subjects, and a significant relation between LVEF and exercise capacity was demonstrated when patients had the MLV as an SV, and superior exercise capacity was also observed in the patients with good ventricular function when patients who had the MRV as an SV were analyzed separately.Thus it is likely that ventricular function at rest, whether in the MRV or MLV, is, to some extent, related to exercise capacity in these patients. Our results also suggest that the duration of hypoxia, in addition to preoperative and intraoperative hypoxia, contributes to the reduced HR response to exercise because the magnitude of chronotropic incompetence is related to age at the time of definitive operation in AVD groups and is greater in patients with TGA with dextro-great artery situs than in patients with congenitally corrected TGA. SVVR must also be an important factor in the exercise

American Heart Journal June 1999

capacity of these patients; however, its effect on exercise capacity is unclear because patients with significant SVVR were excluded from this study. Because of similar concerns about the MRV as an SV,2 the double-switch operation has been introduced for patients with AVD.3 The ET was slightly longer in our patients who underwent a double-switch operation compared with the patients who underwent a conventional repair; however, no significant differences in cardiorespiratory response during exercise were observed between these groups.A marked decrease in the %O2 pulse during the period from AT to peak exercise as well as moderate chronotropic incompetence were demonstrated; we thus also suggest that the dysfunction of the SV is so marked that the stroke volume is significantly decreased in both present groups, as demonstrated in this study.The MRV as an SV may be responsible for this decrease in stroke volume in group C2, whereas the similar decrease in stroke volume in group C1 is probably caused by the decreased EF of the MLV because the end-diastolic volume is rather large in patients after a double-switch operation.3 In addition, pulmonary valve regurgitation, which is inevitable in such patients after reconstruction of the right ventricular outflow tract, is probably responsible in part for the reduced exercise capacity in these AVD patients.35 The time-consuming surgical procedure of a double-switch operation may result in myocardial damage, which is probably responsible for the abnormal response of the MLV during exercise. Because the DSO procedure includes a VSO, myocardial dysfunction and arrhythmias may be future problems, as demonstrated in this study.A long follow-up period is necessary before it can be determined whether there are any advantages of the DSO procedure over the conventional functional repair. Definitive operation at relatively older age was related to lower peak HR and resulted in reduced exercise capacity in our patients with AVD, and it has been reported that the earlier repair of TGA with VSO is beneficial to exercise capacity.19 Because of the small number of patients and relatively narrow range of ages at the definitive operation in our study, exercise capacity was not correlated with age at the time of definitive operation in the patients with TGA. · Because VC was correlated with peak VO 2 and the VC in groups B1, C1, and C2 was low compared with that of the control group, small lung volume must be one of the determinant factors in the reduction of exercise · VCO · capacity.VE/ 2, which mainly represents ventilatory efficiency, was correlated with VC and LVEF in our patients; therefore this parameter seems to be a useful index representing overall cardiorespiratory function during exercise. Although it has been reported that there is not a strong correlation between exercise capacity and systemic ventricular function at rest;36 and no clinical

American Heart Journal Volume 137, Number 6

symptoms are noted in these TGA patients after a successful definitive operation, it is evident from our data that surgical correction with the MLV as an SV has a certain advantage over that with the MRV as an SV. In addition, Presbitero et al37 observed a significant progression of tricuspid valve regurgitation, unavoidable supraventricular arrhythmia, and the development of congestive heart failure in patients with congenitally corrected TGA compared with normal subjects and suggested that these patients have a shorter life expectancy.Therefore, taking the life span and superior cardiorespiratory response during exercise into consideration, we suggest that the MRV as an SV cannot function as a substitute for the MLV and that better preoperative and intraoperative protection against myocardial damage and modifications of surgical procedure to prevent postoperative arrhythmias are needed to obtain a better prognosis for patients who have an ASO and a double-switch operation.

References 1. Hagler DJ, Ritter DG, Mair DD, Tajik AJ, Seward JB, Fulton RE, et al. Right and left ventricular function after the Mustard procedure in transposition of the great arteries. Am J Cardiol 1979;44:276-83. 2. Sano T, Riesenfeld T, Karl TR, Wilkinson JM. Intermediate-term outcome after intracardiac repair of associated cardiac defects in patients with atrioventricular and ventriculoarterial discordance. Circulation 1995;92:(suppl II) II-272-8. 3. Yagihara T, Kishimoto H, Isobe F, Yamamoto F, Nishigaki K, Matsuki O, et al. Double switch operation in cardiac anomalies with atrioventricular and ventriculoarterial discordance. J Thorac Cardiovasc Surg 1994;107:351-8. 4. Minamisawa S, Nakazawa M, Momma K. Treadmill exercise testing for patients of transposition of the great arteries after surgical repair—a comparison between arterial switch operation and atrial switch operation (in Japanese). Shinjo 1993;25:765-9. 5. Rhodes J, Hijazi ZM, Marx GR, Fulton DR. Aerobic exercise function of patients with persistent coronary artery aneurysms secondary to Kawasaki disease. Pediatr Cardiol 1996;17:226-30. 6. Pombo JF, Troy BL, Russell RO Jr. Left ventricular volumes and ejection fraction by echocardiography. Circulation 1971;43:480-90. 7. Ohuchi H, Nakajima T, Kawade M, Matsuda H, Kamiya T. Measurement and validity of the ventilatory threshold in patients with congenital heart disease. Pediatr Cardiol 1996;17:7-14. 8. Jones NL. Clinical exercise testing, 3rd ed. Philadelphia, Saunders; 1988. p.165-74. 9. Weber KT, Janicki JS. Cardiopulmonary exercise testing. Philadelphia, Saunders; 1986. p.180-3. 10. Bar-Or O, Shephard RJ, Allen CL. Cardiac output of 10-to 13-yearold boys and girls during submaximal exercise. J Appl Physiol 1971; 30:219-23. 11. Ensing GJ, Heise CT, Driscoll DJ. Cardiovascular response to exercise after the Mustard operation for simple and complex transposition of the great vessels. Am J Cardiol 1988;62:617-22. 12. Ekblom B. Effect of physical training in adolesent boys. J Appl Physiol 1969;27:350-5. 13. Wasserman K, Whipp BJ, Koyal SN, Beaver WL. Anaerobic threshold and respiratory gas exchange during exercise. J Appl Physiol 1973;35:236-43.

Ohuchi et al 1193

14. Beaver WL, Wasserman K, Whipp BJ. A new method for detecting anaerobic threshold by gas exchange. J Appl Physiol 1986;60:2020-7. 15. Yamada O, Kamiya T, Suga H. Right ventricular mechanical and energetic properties. Jpn Circ J 1989;53:1260-8. 16. Graham TP, Atwood GF, Boucek RJ, Boerth RC, Bender HW Jr. Abnormalities of right ventricular function following Mustard’s operation for transposition of the great arteries. Circulation 1975;52:678-84. 17. Graham TP Jr, Parrish MD, Boucek RJ Jr, Boerth RC, Breiwese JA, Thompson S, et al. Assessment of ventricular size and function in congenitally corrected transposition of the great arteries. Am J Cardiol 1983;5:244-51. 18. Parrish MD, Graham TP Jr, Bender HW, Jones JP, Patton J, Partain CL. Radionuclide angiographic evaluation of right and left ventricular function during exercise after repair of transposition of the great arteries. Comparison with normal subjects and patients with congenitally corrected transposition. Circulation 1983;67:178-83. 19. Hochreiter C, Snyder MS, Borer JS, Engle MA. Right and left ventricular performance 10 years after Mustard repair of transposition of the great arteries. Am J Cardiol 1994;74:478-2. 20. Dimas AP, Moodie DS, Sterba R, Gill CC. Long-term function of the morphologic right ventricle in adult patients with corrected transposition of the great arteries. Am Heart J 1989;118:526-30. 21. Benson LN, Burns R, Schwaiger M, Schelbert HR, Lewis AB, Freedom RM, et al. Radionuclide angiographic evaluation of ventricular function in isolated congenitally corrected transposition of the great arteries. Am J Cardiol 1986;58:319-24. 22. Peterson RJ, Franch RH, Fajiman WA, Jones RH. Comparison of cardiac function in surgically corrected and congenitally corrected transposition of the great arteries. J Thorac Cardiovasc Surg 1988;96:227-36. 23. Paridon SM, Humes RA, Pinsky WW. The role of chronotropic impairment during exercise after the Mustard operation. J Am Coll Cardiol 1991;17:729-32. 24. Senning A. Surgical correction of transposition of the great vessels. Surgery 1959;45:966-80. 25. Mustard WT. Successful two-stage correction of transposition of the great vessels. Surgery 1964;55:469-72. 26. Menahem S, Ranjit MS, Stewart C, Brawn WJ, Mee RBB, Wilkinson JL. Cardiac conduction abnormalities and rhythm changes after neonatal anatomical correction of transposition of the great arteries. Br Heart J 1992;67:246-9. 27. Mathews RA, Fricker FJ, Beerman LB, Stephnson RJ, Fischer DR, Neches WH, et al. Exercise studies after the Mustard operation in transposition of the great arteries. Am J Cardiol 1983;51:1526-9. 28. Ramsay JM, Venables AW, Kelly MJ, Kalff V. Right and left ventricular function at rest and with exercise after the Mustard operation for transposition of the great arteries. Br Heart J 1984;51:364-70. 29. Reybrouck T, Dumoulin M, Van der Hauwaert LG. Cardiorespiratory exercise testing after venous switch operation in children with complete transposition of the great arteries. Am J Cardiol 1988;6: 861-5. 30. Musewe NN, Reisman J, Benson LN, Wilkes D, Levison H, Freedom RM, et al. Cardiopulmonary adaptation at rest and during exercise 10 years after Mustard atrial repair for transposition of the great arteries. Circulation 1988;77:1055-61. 31. Bowyer JJ, Busst CM, Till JA, Lincoln C, Shinebourne EA. Exercise ability after Mustard’s operation. Arch Dis Child 1990;65:865-70. 32. Reybrouck T, Gewilling M, Dumoulin M, van der Hauwaert LG. Cardiorespiratory exercise performance after Senning operation for transposition of the great arteries. Br Heart J 1993;70:175-9. 33. Myridarkis DJ, Ehlers KH, Engle MA. Late follow-up after venous

American Heart Journal June 1999

1194 Ohuchi et al

switch operation (Mustard procedure) for simple and complex transposition of the great arteries. Am J Cardiol 1994;74:1030-6. 34. Murphy JH, Barlai-Kovach MM, Mathews RA, Beerman LB, Park SC, Neches WH, et al. Rest and exercise right and left ventricular function late after the Mustard operation: assessment by radionuclide ventriculography. Am J Cardiol 1983;51:1520-6. 35. Carvalho JS, Shinebourne EA, Busst C, Rigby ML, Redington AN. Exercise capacity after complete repair of tetralogy of Fallot: delete-

rious effects of residual pulmonary regurgitation. Br Heart J 1992;67: 470-3. 36. Szlachcic J, Massie BM, Kramer BL, Topic N, Tubau J. Correlates and prognostic implication of exercise capacity in chronic congestive heart failure. Am J Cardiol 1985;55:1037-42. 37. Presbitero P, Somerville J, Rabajoli F, Stone S, Conte MR. Corrected transposition of the great arteries without associated defects in adult patients: clinical profile and follow up. Br Heart J 1995;74:57-9.

AVAILABILITY OF JOURNAL BACK ISSUES As a service to our subscribers, copies of back issues of the American Heart Journal for the preceding 5 years are maintained and are available for purchase from Mosby until inventory is depleted, at a cost of $15.00 per issue. The following quantity discounts are available: 25% off on quantities of 12 to 23, and one third off on quantities of 24 or more. Please write to Mosby, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, or call (800)453-4351 or (314)4534351 for information on availability of particular issues. If unavailable from the publisher, photocopies of complete issues may be purchased from UMI, 300 N. Zeeb Rd., Ann Arbor, MI 48106, (313)761-4700.