Long-Term, Randomized Comparison of Balloon Angioplasty and Surgery for Native Coarctation of the Aorta in Childhood
Background— The purpose of this study was to compare the long-term outcomes of children randomized to surgery or balloon angioplasty (BA) for native coarctation (CoA). A prior randomized, short-term comparison of BA and surgery for native CoA in 36 children demonstrated equivalent relief of obstruction. The risk of aneurysm formation and possibly restenosis was higher among patients treated with BA.
Methods and Results— Blood pressure, residual aortic obstruction, and exercise performance were evaluated. Need for repeat intervention was reviewed. Aortic arch anatomy was assessed with magnetic resonance angiography. For subjects who were not available to return for evaluation, the most recent clinical record was utilized. Among the 36 subjects initially randomized, 21 returned for evaluation (11 BA, 10 surgery). The average time since initial intervention to evaluation for all subjects was 10.6±4.7 years for BA subjects and 11.3±3.7 years for surgical subjects. Resting blood pressure, CoA gradient, exercise performance, MRI analysis of the aortic arch, and need for repeat interventions were not different for the 2 treatment strategies. There was a higher incidence of aneurysm formation (35% versus 0%) and a greater difference in blood pressure between the right and left legs with exercise among BA subjects. Some aneurysms developed late, first being detected more than 5 years after the initial intervention. Only 50% of BA subjects remained free of both aneurysm formation and repeat intervention compared with 87.5% of surgical subjects (P=0.03).
Conclusions— BA for the treatment of childhood CoA is associated with a higher incidence of aneurysm formation and iliofemoral artery injury than surgery. These differences should be considered when undertaking treatment for native CoA during childhood.
Received September 29, 2004; revision received February 9, 2005; accepted March 2, 2005.
Despite more than 20 years of experience, balloon angioplasty (BA) as treatment for native coarctation of the aorta (CoA) during childhood remains controversial. Surgical intervention for native CoA during childhood is currently the preferred method of treatment in many centers. Although stent implantation has become a common means of treating native and recurrent CoA in adolescents and adults, this modality is not used routinely in children because of issues related to vessel size and the need for repeated stent dilation to accommodate for somatic growth. We previously reported the results of a short-term, randomized comparison of BA and surgery for native CoA in children. Equivalent relief of obstruction was demonstrated, but the risk of aneurysm formation and possibly restenosis was higher among patients treated with BA.1 Despite this apparent difference, the authors of more recent studies continue to advocate BA for native CoA.2,3 To better understand the long-term implications of these differing treatment strategies, we compared the outcomes of 2 groups of children previously randomized to surgery or BA for the treatment of native CoA.
See p 3347
Between August 1985 and November 1990, all patients between the ages of 3 and 10 years seen at Primary Children’s Medical Center with a hemodynamically significant native CoA were invited to participate in an Institutional Review Board–approved study. Participants were randomly assigned to BA or surgery with equal probability. For subjects randomized to BA, the balloon diameter was selected to match the diameter of the uninvolved aorta proximal to the coarctation. If the residual peak systolic gradient across the CoA after balloon dilation exceeded 8 mm Hg, the next-largest balloon diameter was selected and angioplasty repeated.
Institutional Review Board approval was again obtained for this follow-up study. Subjects who returned for long-term follow-up were classified as group I. The subject’s history since initial intervention was reviewed carefully with respect to any cardiovascular complications or additional interventions performed. Subjects were examined, and heart rate and blood pressure (BP) were measured in the right arm with the subject in a sitting position. Right and left lower-extremity limb length was measured from the anterior superior iliac crest to the plantar surface of the heel. The sites of maximum Doppler signal for the radial and posterior tibial arteries were located and marked. Supine systolic blood pressure in all 4 extremities was measured with a Doppler probe and an appropriately sized cuff on the proximal arm or thigh. The residual resting systolic coarctation gradient was calculated as the difference in systolic BP between the right arm and the lower extremity with the highest measured systolic BP.
Subjects underwent exercise treadmill testing with a standard Bruce protocol with continuous ECG monitoring. Once fatigued, subjects immediately assumed a supine position. BP cuffs were applied simultaneously to each extremity, and systolic BP was determined with a Doppler probe over the previously marked pulse site. BP was measured in succession: right arm, right leg, left leg, and finally left arm. A large-volume air pump allowed rapid cuff inflation such that all 4 measurements were complete within 1 minute. Resting heart rate and right arm BP were again recorded with the subject in a sitting position 10 minutes after exercise termination.
Magnetic resonance angiography was performed and interpreted by a pediatric radiologist who was blinded to the patient’s treatment history. Measurements of the aorta were obtained in both anteroposterior and lateral views just distal to the left common carotid artery, at the isthmus, at the coarctation repair site, and at the diaphragm. An aneurysm was defined as an area of dilation that was 150% of the aortic diameter at the diaphragm or a discrete saccular dilation at the site that was not present before intervention.4
Subjects who could not be reached or were unable to return for follow-up evaluation as part of the present study were classified as group II. For these subjects, the most recent clinical records were reviewed for aortic arch imaging results and any additional interventions performed since initial BA or surgery.
Comparison of treatment strategies with regard to BP, exercise performance, and magnetic resonance angiography measurements was performed with a 2-tailed Student t test. Categorical data comparisons were performed with a Fisher exact test. Nonparametric data were compared by the Mann-Whitney U test. Differences were considered statistically significant for P<0.05. All data are expressed as mean±SD unless otherwise stated.
Among the 36 original participants randomized to BA (n=20) or surgery (n=16), 21 subjects returned for evaluation (11 BA, 10 surgery) between April 2001 and November 2002. The demographics for these group I subjects are summarized in Table 1. At the time of follow-up evaluation, there were no significant differences in height, weight, age, time since initial intervention, or right versus left leg length between treatment groups. Information gathered from the most recent clinical records for group II subjects included data for the remaining 9 BA and 6 surgical subjects (Table 2). The combined demographics for all subjects initially enrolled are summarized in Table 3 (groups I and II combined).
Group I Baseline BP Measurements
None of the group I subjects were taking antihypertensive or other cardiovascular medications at the time of evaluation. Average resting BPs and coarctation gradients were not statistically different for either treatment strategy. Relief of obstruction was achieved with both treatment modalities with no significant systolic BP gradient measured at rest.
Group I Exercise Testing
Exercise performance for group I subjects is summarized in Table 4. The range in systolic BP gradient with exercise was 14 to 128 mm Hg for the BA group and −17 to 152 mm Hg for the surgical group. The mean difference in BP between upper and lower extremity with exercise exceeded 50 mm Hg irrespective of initial treatment (P<0.001). BA subjects developed a greater difference in BP between the lower extremities than was seen among surgical subjects, which was not present at rest. Heart rate and BP measurements 10 minutes after exercise for group I subjects are also summarized in Table 4. The mean diastolic BP 10 minutes after exercise was statistically higher for the BA subjects (P<0.05).
Group I Magnetic Resonance Angiography
MRI of the aortic arch was declined by 1 group I surgical subject because of claustrophobia. Quantitative, noninvasive imaging of aortic arch dimensions among group I subjects demonstrated no statistically significant differences between treatment strategies. The ratio of the aortic diameter at the site of initial intervention to the diameter of the aorta at the diaphragm was 0.9±0.3 for the BA group and 0.9±0.2 for the surgical group (P=NS).
Among the 20 subjects initially treated with BA, 3 underwent repeat BA, and 3 underwent surgical repair (2 for aneurysm resection). Two subjects initially treated surgically underwent BA. One of these subjects later returned for repeat surgery. Fisher exact test analysis of repeat interventions demonstrated no difference between treatment groups (P>0.05).
Aneurysms were detected in 7 (35%) of the 20 BA subjects (4 diffuse, 3 discrete). No aneurysms were detected among the 16 surgical subjects (P=0.011). Aneurysms were detected among 4 BA subjects at a median of 1.03 (range 0.17 to 1.44) years after their initial intervention. One of these subjects was referred for elective surgical repair 6.8 years after initial BA at the discretion of his cardiologist, at which time an interposition graft was placed. No repeat interventions have been performed on the other 3 subjects in whom aneurysms were detected early. Late aneurysms developed in 3 additional BA subjects despite negative imaging performed ≈1 year after the initial procedure. An aneurysm was first detected in 1 subject 6.4 years after initial intervention. This subject was electively referred for surgical intervention 4.9 years later, at which time the aneurysm was resected and patch aortoplasty performed. One subject who initially underwent BA and subsequently underwent surgical resection and end-to-end anastomosis for recurrent CoA 0.8 years after BA had no evidence of aneurysm formation 8.4 years after initial intervention. Repeat imaging performed 6.1 years later (14.5 years after initial BA) demonstrated an aneurysm at the former repair site. This subject has not yet been referred for additional intervention. A third late aneurysm was detected 8.3 years after initial BA. No additional interventions have been performed given a stable appearance for the past 8 years.
Surgical technique for CoA has evolved to overcome observed problems with persistent or recurrent obstruction or aneurysm formation. An extended end-to-end anastomosis is currently the preferred treatment for CoA during childhood in some centers, with BA reserved only for recurrent obstruction. The report of BA in the treatment of native CoA in an infant with congestive heart failure in 19835 initiated a controversy that persists today. Although BA for CoA during infancy is now rarely performed because of the high incidence of recurrent obstruction, this alternative to surgery in the treatment of CoA in children remains in widespread practice. Acute and 1-year follow-up results of a prospective comparison of BA and surgery for native CoA showed a similar immediate gradient reduction but a higher risk of aneurysm formation and possibly restenosis with BA among 36 patients aged 3 to 10 years.1 Because of the continued controversy about the selection of BA or surgery for the treatment of native CoA in children, we sought to examine the long-term outcomes among this previously randomized group of 36 subjects.
Effective relief of obstruction nearly 14 years after initial treatment was demonstrated in the present study with both BA and surgery. Baseline BP measurements and exercise performance were also equivalent between the 2 groups, perhaps as a result of the effective relief of aortic arch obstruction. A statistically significant gradient with exercise in subjects from both groups was noted and is most likely related to the inherent vascular dysfunction of the upper segment known to occur in association with coarctation.6–9 The more rapid normalization of diastolic BP among the surgical subjects 10 minutes after exercise termination is interesting but of unknown significance.
The potential development of aneurysms has long been considered a major limitation to BA. A strikingly higher incidence of aneurysm formation after BA (35% versus 0% for surgery) was seen in the present study, with the development of 3 aneurysms more than 5 years after the initial intervention. The 35% incidence of aneurysm formation among BA patients in the present series is similar to some early reports in which aneurysm formation occurred in 36% to 43% of children treated with BA.10,11 More recently, however, Walhout et al12 found no aneurysms among 32 children with native CoA treated with BA, but the duration of follow-up was shorter than in the present study. The late development of aneurysms in the present series demonstrates the importance of long-term imaging in patients who have undergone BA. The recently reported 8% incidence of aneurysms among adolescents and adults treated with BA for native coarctation is much lower than that observed in the present study, but this may be due to a variety of factors, including greater vessel wall thickness in older patients.13 Aneurysm formation after surgery in subjects initially treated with BA, as occurred in 1 subject in the present study, has not been widely reported. On the contrary, successful surgical repair after BA in subjects who develop recurrent stenosis has been described.14 However, the presence of an aneurysm may complicate surgical intervention, in some cases necessitating the use of interposition grafts or patches.
In addition to the higher incidence of aneurysm formation, a greater discrepancy between right and left lower-extremity BP with exercise was observed among subjects initially treated with BA. This difference in lower-extremity BP with exercise is most likely due to femoral artery injury sustained at the time of intervention. Previous studies have reported the incidence of iliofemoral stenosis or occlusion after aortic valvotomy or coarctation BA to be 22% to 58%.15–17 Although we found no associated higher incidence of lower-limb-length discrepancy in the BA group, and no subjects complained of asymmetric claudication, the potential for progressive arterial insufficiency as previously reported may prove problematic for these subjects.17 The current availability of lower-profile balloon catheters may decrease the incidence of this complication.
These data demonstrate good relief of obstruction and an equivalent need for repeat interventions for both surgery and BA in the treatment of CoA in childhood; however, a higher incidence of aneurysm formation was observed among BA subjects, with some aneurysms first detected more than 5 years after initial intervention. Although some clinicians may consider these to be acceptable risks, knowing that future surgery can still be performed among patients initially referred for BA, only 50% of subjects initially treated with BA were free of detected aneurysms and had not been referred for additional interventions compared with 87.5% of surgical subjects (P=0.03). Although caution is warranted in extrapolating the results of procedures initially performed on this cohort of children nearly 20 years ago to the current era, the relatively uncontrolled vascular tearing inherent to BA appears to be associated with the development of aneurysms, some of which may develop late. Clinicians should consider these differences in outcomes and the need for long-term imaging studies among patients treated with BA when recommending treatment for native CoA during childhood.
This investigation was supported by Public Health Services research grant number M01-RR00064 from the National Center for Research Resources.
Shaddy RE, Boucek MM, Sturtevant JE, Ruttenberg HD, Jaffe RB, Tani LY, Judd VE, Veasy LG, McGough EC, Orsmond GS. Comparison of angioplasty and surgery for unoperated coarctation of the aorta. Circulation. 1993; 87: 793–799.
Guenthard J, Zumsteg U, Wyler F. Arm-leg pressure gradients on late follow-up after coarctation repair: possible causes and implications. Eur Heart J. 1996; 17: 1572–1575.
Cooper RS, Ritter SB, Rothe WB, Chen CK, Griepp R, Golinko RJ. Angioplasty for coarctation of the aorta: long-term results. Circulation. 1987; 75: 600–604.
Barlow A, Menahem S, Wilkinson J. Arterial morbidity following interventional balloon dilation procedures. Cardiol Young. 1996; 6: 54–58.
Burrows PE, Benson LN, Babyn P, MacDonald C. Magnetic resonance imaging of the iliofemoral arteries after balloon dilation angioplasty of aortic arch obstructions in children. Circulation. 1994; 90: 915–920.