Right Ventricular Performance and Mass by Use of Cine MRI Late After Atrial Repair of Transposition of the Great Arteries
Background The long-term adaptation of the right ventricle after atrial repair of transposition of the great arteries (TGA) remains a subject of major concern. Cine magnetic resonance imaging (MRI), with its tomographic capabilities, allows unique quantitative evaluation of both right and left ventricular function and mass. Our purpose was to use MRI and an age-matched normal population to examine the typical late adaptation of the right and left ventricles after atrial repair of TGA.
Methods and Results Cine MRI was used to study ventricular function and mass in 22 patients after atrial repair of TGA. Images were obtained in short-axis sections from base to apex to derive normalized right and left ventricular mass (RVM and LVM, g/m2), interventricular septal mass (IVSM, g/m2), RV and LV end-diastolic volumes (EDV, mL/m2), and ejection fractions (EF). Results 8 to 23 years after repair were compared with analysis of 24 age- and sex-matched normal volunteers and revealed markedly elevated RVM, decreased LVM and IVSM, normal RV size, and only mildly depressed RVEF. Only 1 of 22 patients had clinical RV dysfunction, and this patient had increased RVM.
Conclusions Cine MRI allows quantitative evaluation of both RV and LV mass and function late after atrial repair of TGA. Longitudinal studies that include these measurements should prove useful in determining the mechanism of late RV failure in these patients. On the basis of these early data, inadequate hypertrophy does not appear to be the cause of late dysfunction in this patient group.
Atrial repair of TGA introduced by Senning1 and Mustard2 has provided outstanding clinical results, but RV dysfunction has been a consistent problem in a variable number of late survivors.3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Cine MRI is an emerging cardiovascular imaging modality that provides theoretically ideal tomographic capabilities for comprehensive RV and LV volume and mass quantification. The purpose of this study was to use cine MRI to measure RV and LV volumes, ejection fractions, outputs, and mass in an unselected group of late survivors of atrial repair of TGA.
The patient group consisted of 22 patients who had undergone either a Senning (n=12) or a Mustard (n=10) procedure for atrial repair of TGA. The patients were selected only by being at least 8 years postrepair, agreeing to return to Vanderbilt for study, and not having any contraindications for MRI, including significant atrial arrythmias. The study was approved by the Vanderbilt University Institutional Review Board, and all subjects gave informed consent. The mean age for the entire group at surgery was 0.96 years and mean age at the time of the MRI study was 15.7 years. Mean age at Senning repair was 4.5 months (range, 6 days to 2.3 years), whereas mean age at Mustard repair was 1.48 years (range, 6 months to 4 years). The mean interval between repair and MRI was 12.0 years in the Senning group and 19.6 years in the Mustard group (Table 1⇓).
The operative technique has been described previously.18 In infants weighing <10 kg who were <1 year of age, profound hypothermia and circulatory arrest were used. In all other patients, cardiopulmonary bypass with moderate hypothermia (25°C) was utilized. All patients received cold crystalloid cardioplegic solution. All but four patients were operated on at Vanderbilt.
The control group consisted of 24 sex-matched volunteers with no history of heart disease. Age was matched as closely as possible. The mean age of the control group was 21.3±7.0 years, and the mean BSA of this group was 1.67±0.29 m2.
In the Senning group, a VSD was present in 5 of 12 patients; in 2 of these 5 patients, the VSD was small and not repaired, and in the remaining 3 patients, the VSD was small and closed by interrupted sutures by use of an atrial approach. The 2 patients with VSDs not closed at surgery had Doppler echocardiographic evidence of a small residual defect at the time of MRI with estimated normal LV pressure. Residual mild pulmonary stenosis with Doppler gradients of 25 and 41 mm Hg was present in 2 of the 12 patients. No patients had pulmonary hypertension by Doppler echocardiography, and none had baffle obstruction.
In the Mustard group, a VSD was present in 2 of 10 patients; of these 2 patients, 1 had spontaneous closure of a small VSD and 1 had VSD closure by use of an atrial approach. Residual pulmonary stenosis was present in 1 patient (Doppler gradient 70 mm Hg). Only 1 patient in the entire group had tricuspid regurgitation by MRI and by clinical exam, and this regurgitation was considered mild by both modalities. No patients had pulmonary hypertension. Superior vena cava partial obstruction was present in 2 patients with a widely patent inferior vena cava; 1 patient had partial pulmonary venous obstruction.
Symptoms, Rhythm, and Medications
In the Senning group, 11 of 12 patients had no symptoms and 1 of 12 had easy fatigability (ability index 223 ) and was the only patient in this group receiving medical therapy (digoxin and enalapril). Sinus rhythm was predominant in 7 of 12 patients, junctional rhythm in 2 of 12, sinus/junctional rhythm alternating in 2 of 12, and an undetermined rhythm in 1 patient.
In the Mustard group, 7 of 10 patients were asymptomatic. One patient had moderate symptoms, and 2 of 10 had questionable easy fatigability. Digoxin was used in 5 of 10 patients, and diuretics and enalapril in 1 patient each.
MRI acquisition involved a standardized protocol that results in a series of short-axis cine loops covering the LV and RV from the atrioventricular valve plane to the apex. All studies were performed with a Siemens SP 4000 (Siemens Medical Systems, Inc) MR scanner. A series of scout images was acquired to localize the short-axis plane. The protocol began with a sagittal image to locate the position of the heart in the chest. A set of transverse images was then obtained to visualize the interventricular septum. Images parallel to the interventricular septum in the LV were then acquired, yielding a vertical long-axis view. Images acquired through the long axis of this image resulted in a horizontal long-axis view. From this view, the tricuspid and mitral valve planes were defined. An ECG-triggered gradient echo cine sequence (TR 50 ms, TE 12 ms, flip angle 60 degrees) was then used to acquire images in the short-axis plane in contiguous 7-mm locations from the valve plane to the apex of the heart. Fig 1⇓ shows the definition of the short-axis planes from the horizontal long-axis view in a healthy volunteer. This method of defining the short-axis view has been found in our laboratory to be reproducible within a few degrees in each axis (C.H. Lorenz, PhD, et al, unpublished data, 1994). This approach to data acquisition allows reproducible acquisition of cine images of the heart in the short-axis view for comparison between patients and for serial studies of the same patient.
The images were transferred to an independent computer workstation (MaxiView, Dimensional Medicine, Inc) for analysis. By use of visual inspection of the cine loops, the end-diastolic and end-systolic frames were chosen. The endocardial borders of the LV and RV were outlined at end-diastole and at end-systole at all levels from base to apex. The epicardial borders of the LV and RV were outlined at end-diastole for determination of ventricular mass. In addition, the borders of the interventricular septum were determined separately. All borders were marked by the same observer (C.H.L.).
The contours were then stacked to yield ventricular volume. The density of the myocardium was assumed to be 1.05 g/cm3. Ventricular mass was calculated as the volume of tissue between the epicardial and endocardial borders multiplied by the assumed density of the tissue.
The following parameters were calculated for each subject: LV and RV end-diastolic volume, LV and RV end-systolic volume, LV and RV stroke volume, LV and RV ejection fraction, LV free wall mass, interventricular septal mass, LV total mass, RV free wall mass, RV end-diastolic volume to mass ratio, LV end-diastolic volume to mass ratio, and LV total mass to RV mass ratio. Parameters were normalized for BSA.
The mean±SD of each parameter for the entire TGA group, the Senning group, the Mustard group, and the control population was calculated. The entire TGA group was compared with the control population by use of ANOVA for each parameter. The Senning and Mustard groups were also compared with each other and separately with the control group. A value of P<.05 was considered significant.
Fig 2⇓ shows an end-diastolic midventricular image of a patient who had undergone Senning repair (top) and the end-systolic image at the same level (bottom). The RV is heavily trabeculated, hypertrophied, and round in shape; the septum is flat; and the LV has decreased wall thickness compared with normal. At end-systole, the RV is round and the septum is concave toward the RV. This geometry was found consistently in all patients.
Fig 3⇓ shows RVEF for all patient groups compared with the control group. This variable was moderately depressed in the composite group, from 62±5% to 51±9% (P<.01) Table 2⇓ summarizes the differences between groups. The Senning and Mustard patients did not differ in RVEF. There was a weak but significant decrease in RVEF as a function of time since surgery for the Senning group (r=.51).
Normalized RV volume (RV end-diastolic volume/BSA) is shown in Fig 4⇓; this variable did not differ among any of the patient groups and was well within normal limits.
RV free wall mass normalized for BSA in g/m2 is depicted in Fig 5⇓. There is a very tight normal value of 27±4 g/m2. All patient groups were significantly increased from the control group; Senning and Mustard patients did not differ from one another.
RV volume to mass ratio is shown in Fig 6⇓ and averaged 2.3±0.9 in the control group. All patient groups showed a significantly depressed value due to the marked increase in RV mass with normal RV end-diastolic volume.
In Fig 7⇓, interventricular septal mass/BSA is shown with a normal value of 30±4 g/m2. For all patient groups, the value is less than for the control group. Interventricular septal mass/BSA is smaller in the Senning group than in the Mustard group (P<.01).
In Fig 8⇓, LVEF is shown with a normal value of 67±5%. The composite group and the Senning group are mildly decreased compared with the control group, but the Mustard group is not significantly different from the control group. There were no differences between the Senning and Mustard groups.
LV end-diastolic volume/BSA is shown in Fig 9⇓ and is less than normal for the Senning group and the composite group. The Mustard group is not significantly different from the control group or the Senning group.
Cardiac index, calculated as RV stroke volume divided by BSA multiplied by heart rate, averaged 2.5±0.8 L · min−1 · m−2 in the composite TGA group. The variable was equal to LV output calculated in the same manner and was significantly less than the MRI cardiac index of 3.0±0.6 min−1 · m−2 in the control group.
Fig 10⇓ shows LV free wall mass in g/m2. This variable is significantly depressed in all patient groups. The LV volume to mass ratio is shown in Fig 11⇓. This variable is increased in all patient groups when compared with the control group.
In Fig 12⇓, LV total mass to RV mass ratio is shown. This variable demonstrates a marked decrease from normal in all groups due to the marked increase in RV mass coupled with the mild decrease in LV mass.
Although laboratory evidence of RV dysfunction in late survivors of atrial repair of TGA has been demonstrated by a number of investigators, clinical myocardial failure is relatively uncommon. Indeed, a number of longitudinal studies of small numbers of patients have not shown progressive deterioration of cardiac function.16 22 Our study is unique in providing not only RV volume and functional data in late atrial repair patients but also in measuring for the first time both RV and LV mass in this patient group.
Cine MRI is a technique well suited for RV mass measurement, but the methodology at present is extremely tedious and time-consuming. Thus, few data are available for either normal or abnormal patient groups. We have shown marked RV hypertrophy in our TGA patients, as might be expected with this ventricle connected to the aorta. We did not show any evidence for inadequate myocardial hypertrophy as a potential cause of RV dysfunction.
An additional unique finding of the study was the modest decrease in LV mass and volume in our patients. An unexpected finding was the lack of septal hypertrophy in the TGA groups unless pulmonary hypertension was present. Our presupposition was that the septum would hypertrophy in a similar fashion to the RV free wall in TGA patients. Our data and data from patients with pulmonary hypertension24 25 indicate that the septum responds more to LV than RV loading in terms of a stimulus to hypertrophy. The coupling of volume or pressure overload of either ventricle to the molecular processes that initiate hypertrophy is obviously a complex process. Geometrical considerations suggest that myocardial cells can respond to increases in LV wall tension (a product of pressure and radius) with an increase in oxygen consumption26 and in all probability an increase in myocardial mass with such a chronic stimulus. Why the septum hypertrophies predominately with the LV rather than the RV remains open to speculation.
MRI total exam time is relatively long (approximately 1 hour), although the time for acquisition of a single cine loop is short (about 3 minutes).
The analysis technique used assumes that the subject’s hemodynamic state does not change significantly over the examination period. The functional parameters calculated therefore represent average values over the 1-hour acquisition.
In several cases, image quality was suboptimal due to respiratory motion. Both the control group and the TGA group were equally affected by respiratory artifacts. In no case did the degraded image quality prevent delineation of the ventricular borders.
All studies were analyzed by a single observer in this study. Previously, we have examined interobserver variability, intraobserver variability, and interstudy reproducibility using the acquisition and analysis techniques presented here. We found interobserver and intraobserver variability to be low (mean difference, 1±5 mL and 1±5 mL, respectively) and interstudy reproducibility to be good (mean difference between studies, 1±4 mL) (C.H. Lorenz, PhD, et al, unpublished data, 1994).
RV mass measurement currently is possible only with ultrafast computed tomography27 and MRI.24 28 29 30 Our study presents the largest normal database for RV mass to date, and the data are comparable to previously published results. It will be important to establish such normal values that are age and sex specific to allow appropriate controls for longitudinal studies.
Abnormalities of RV function either at rest or with exercise have been well documented in a variable number of survivors of atrial repair of TGA,3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 reaching as high as 50% or greater in some series.3 4 5 6 12 16 18 21 Despite these abnormalities, clinical RV failure is considerably less common. In our own patient population, without significant arrythmias, congestive heart failure in this group late after atrial repair is approximately 10% and appears to be slowly increasing with advancing age of our survivors. With the onset of acquired cardiovascular disease, including hypertension and coronary artery disease, this number should show a marked increase. It is our hope that the use of MRI to accurately measure both RV and LV mass and volume will prove useful in determining the cause of congestive heart failure in this and other patient groups in whom the RV is the systemic pumping chamber or in whom pulmonary hypertension is part of their clinical condition.
Selected Abbreviations and Acronyms
|BSA||=||body surface area|
|LV||=||left ventricle (ventricular)|
|LVEF||=||left ventricular ejection fraction|
|MRI||=||magnetic resonance imaging|
|RV||=||right ventricle (ventricular)|
|RVEF||=||right ventricular ejection fraction|
|TGA||=||transposition of the great arteries|
|VSD||=||ventricular septal defect|
This study was funded in part by the American Heart Association, Tennessee Affiliate. The authors would like to thank Yvonne Bernard, RN, for help with patient recruitment and John Bobbitt for preparation of the illustrations.
Presented in part at the 67th Scientific Sessions of the American Heart Association, Dallas, Tex, November 14-17, 1994, and published in abstract form (Circulation. 1994;90[pt 2]:I-98).
- Copyright © 1995 by American Heart Association
Graham TP Jr, 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;52:678-684.
Hagler DJ, Ritter DG, Mair DD, Tagik AJ, Steward JB, Fulton RE, Ritman ED. Right and left ventricular function after the Mustard procedure in transposition of the great arteries. Am J Cardiol. 1979:44:276-283.
Bender HW Jr, Graham TP Jr, Boucek RJ Jr, Walker WE, Boerth RG. Comparative operative results of the Senning and Mustard procedures for transposition of the great arteries. Circulation. 1980;62(pt 2):I-197-I-203.
Borow KM, Keane JF, Castaneda AR, Freed MD. Systemic ventricular function in patients with tetralogy of Fallot, ventricular septal defect and transposition of the great arteries repaired during infancy. Circulation. 1981;64:878-885.
Benson L, Bonet J, McLaughlin P, Olley P, Feiglin D, Druck M, Trusler G, Rowe R, Morch J. Assessment of right ventricular function during supine bicycle exercise after Mustard’s operation. Circulation. 1982;65:1052-1059.
Parrish M, Graham T, Bender H, Jones J, Patton J, Partain L. Radionuclide angiographic evaluation of right and left ventricular function during exercise after repair of transposition of the great arteries. Circulation. 1983;67:178-183.
Ramsay J, Venables A, Kelly M, 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;51:364-370.
Graham TP, Burger J, Bender HW, Hammon JW, Boucek RJ, Appleton S. Improved right ventricular function after intra-atrial repair of transposition of the great arteries. Circulation. 1985;72(suppl II):II-1145-II-1151.
Musewe NN, Reisman J, Benson LN, Wilkes D, Levison H, Freedom RM, Trusler GA, Canny GJ. Cardiopulmonary adaptation at rest and during exercise 10 years after Mustard atrial repair for transposition of the great arteries. Circulation. 1988;77:1055-1061.
Wong K, Venables A, Kelly M, Kalff V. Longitudinal study of ventricular function after the Mustard operation for transposition of the great arteries: a long term follow up. Br Heart J. 1988;60:316-323.
Buch J, Wennevoid A, Jacobsen J, Hvid-Jacobsen K, Lauridsen P. Long-term follow-up of right ventricular function after Mustard operation for transposition of the great arteries. J Thorac Cardiovasc Surg. 1988;22:197-202.
Warnes CA, Somerville J. Transposition of the great arteries: late results in adolescents and adults after the Mustard procedure. Br Heart J. 1987;58:148-155.
Redington A, Rigby M, Oldershaw P, Gibson D, Shinebourne E. Right ventricular function 10 years after the Mustard operation for transposition of the great arteries: analysis of size, shape, and wall motion. Br Heart J. 1989;62:455-461.
Warnes CA, Somerville J. Tricuspid atresia in adolescents and adults: current state and late complications. Br Heart J. 1986;53:535-543.
Frist WH, Lorenz CH, Walker ES, Loyd JE, Stewart JR, Graham TP, Pearlstein DP, Key SP, Merrill WH. MRI complements standard assessment of right ventricular function after lung transplantation. Ann Thorac Surg. In press.
Lorenz CH, Walker ES, Powers TA, Loyd JE, Newman JH, Frist WH, Graham TP. MR evaluation of right ventricular function and mass pre and post single lung transplantation. Radiology. 1993;189(P):364. Abstract.
Graham TP Jr, Covell JW, Sonnenblick EH, Ross J Jr, Braunwald E. Control of myocardial oxygen consumption: relative influence of contractile state and tension development. J Clin Invest. 1968;47:375-385.
Hadjuczok ZD, Weiss RM, Stanford W, Marcus ML. Determination of right ventricular mass in humans and dogs with ultrafast computed tomography. Circulation. 1990;82:202-212.