Coarctation of the aorta is a narrowing that happens either discretely, typically at the isthmus of the aorta, or more diffusely across the transverse arch. It occurs in four out of 10,000 live births,Reference Reller, Strickland and Riehle-Colarusso ¹ comprising 4–8% of congenital cardiac defects. Its outpatient management can vary, depending on the mode of presentation. The more common, and thus more familiar, as well as more dramatic presentation occurs in the neonatal period when the arterial duct closes. In this case, the infant presents either with acute decompensation or with hypertension, with left-sided obstruction to outflow. The clinical findings are typically quite evident, with pallor and grunting, shock, and decreased lower extremity pulses. These patients are acutely ill and require parenteral infusion of prostaglandin E1 to open the arterial duct and to re-establish lower extremity blood flow, further stabilisation and support as needed, and then surgical repair when the child is a more suitable surgical candidate.

Less frequently, coarctation can occur in a more subtle manner, such that the child is found to have incidental hypertension, and the aortic obstruction is found during the evaluation. Other types of presentation that can be seen at this later stage include the finding of a murmur, the symptoms and findings of a cardiomyopathy, or decreased lower extremity pulses with lag between upper and lower extremity pulses. The management of these patients is quite different, with gentle therapy to reduce the hypertension as the first intervention. This is then followed by invasive intervention with either primary surgical repair or interventional catheterisation to include balloon angioplasty of the aorta; this may also include placement of a stent.

These two disparate means of presentation influence the initial evaluation of the patient. In infants appearing with acute closure of the arterial duct, signs of congestive heart failure can be noted. The parents may report dyspnoea and diaphoresis with feeds. They may also report that their child has tachypnoea as well as retractions, and may appear startlingly grey or pale. The examination in these patients can show an infant in extremis, with dyspnoea, diaphoresis, tachycardia, and pallor. Occasionally, differential cyanosis is seen, in which there is a demarcation across the chest such that the arms, shoulders, upper chest, and head appear pink, and the remainder of the lower portion of the body appears blue. A harsh systolic ejection murmur, associated with aortic valve obstruction, may be heard at the left sternal border, and a longer systolic decrescendo murmur, associated with the aortic obstruction, may be noted in the left axilla or back in the interscapular or infrascapular space. Lower extremity pulses may be either decreased in volume or entirely absent. There may also be a time delay between the onset of the right-sided radial pulse and that of one of the femoral arteries. Of note, these arteries are chosen to be assessed simultaneously, as both arteries are felt to be at the same approximate distance from the site of coarctation, with the right radial artery arising before the coarctation site and the femoral artery being located downstream of the obstruction.

In patients who present later, the history may be notable for headache, chest pain, exercise intolerance, leg pain or leg weakness, or, most commonly, a complete lack of symptoms. Other findings on physical examination may include a systolic ejection click at the left sternal border, which can be associated with the bicuspid aortic valve, and an S4 gallop at the apex, if an associated cardiomyopathy is diagnosed secondary to prolonged hypertension.

In the assessment for hypertension, it is important to ensure that the blood pressure is measured correctly; one of the most frequent aetiologies of spurious hypertension in the clinic is the incorrect assessment of this important piece of data. According to the guidelines of the National Heart, Lung, and Blood Institute, ² blood pressure is measured as follows: the patient should be seated for 5 minutes with their feet on the floor and their back supported. The blood pressure should then be measured manually with a sphygmomanometer. Ideally, the sphygmomanometer would be calibrated every 6 months. The cuff that is used should be of appropriate size. Historically, this has meant that the cuff should cover anywhere from 70 to 80% of the upper extremity. The latest recommendations, however, suggest that the cuff should just be the largest cuff that will fit over the upper extremity. There is often concern that the cuff can be “too large”; yet, a cuff that is larger will be less likely to underestimate systolic pressure, whereas a cuff that is too small has a much higher likelihood of overestimating systolic pressure. Errors in measurement occur when patients are brought in directly from the waiting area, seated, and have the cuff placed immediately for measurement. In addition, another source of error is taking the measurement with the patient sitting on the edge of the examination table, with legs dangling over the side. For expediency, blood pressure is obtained by using an electronic oscillometric device. Such devices have been demonstrated to sometimes overestimate systolic blood pressure. Therefore, although they are reasonable for use as a screening device, a manual determination should be performed if an elevated reading is found.

Blood pressure should be measured routinely in the right upper extremity. This is chosen by convention, as its brachial artery arises before the onset of coarctation in the large majority of cases. The one exception here occurs when there is an aberrant origin of the right brachial artery branching below the site of the coarctation, such that the hypertension associated with coarctation could be missed. This reinforces the need to assess pulses in both right and left radial arteries, at least one femoral artery, plus a carotid artery to compare the amplitude of the respective systolic impulses. The expectation in the case of aberrant origin of the right brachial artery is that the carotid pulse would be either normal or elevated, whereas the radial pulses would be diminished. Femoral pulses may be diminished or normal, and may not demonstrate pulse lag as compared with the right radial artery. Measurement of leg blood pressure can be helpful in determining whether there is a blood pressure difference between the upper and the lower extremities. The normal leg blood pressure is often elevated or equal to that of the right arm pressure, with the blood pressure in the leg being decreased in the presence of coarctation. A frequent technique utilised is the measurement of “four-quadrant” blood pressures, which is assessment of blood pressures in all four extremities. This is often used because it is easier to remember and to instruct students; however, evaluation of the right arm and either leg pressures is typically sufficient to evaluate for evidence of native or residual coarctation. The exception here occurs in patients with William’s syndrome, who can develop obstruction throughout their vasculature. Classically, besides coarctation, they can have supravalvar aortic stenosis. This can lead to a Coanda effect, in which the jet of blood adheres to the wall of the aorta and preferentially streams down the right subclavian artery. This leads to the finding of an artificially elevated blood pressure in the right arm. Therefore, blood pressure measurement should be evaluated in the left arm as well, unless it can be obtained directly by ultrasound measurement of the aorta.

The use of radiography of the thorax is more of historic significance in the diagnosis of chronic coarctation, as it is much less frequently utilised in the era of ultrasound evaluation. The classic finding is that of the “3” sign, found in the aortic knob, in which there is pre-stenotic dilation, followed by the inward “tuck” of the coarctation, and then a post-stenotic dilation of the aorta. Barium swallow, which is not necessarily a diagnostic study routinely used for patients with coarctation but is certainly used in patients with concern for vascular rings and other anomalies of the head and neck vessels of the aortic arch, can demonstrate an “E” sign, which is sometimes referred to as the “reverse 3 sign”. This results from the opacification of the oesophagus rightward of the aorta and shows the mirror image of the “3” sign described above. The chest radiograph can also show notching of the midpoints of the ribs in patients with prolonged blood flow bypassing the coarctation through collateral vessels. The intercostal arteries dilate and impinge upon the ribs, resulting in abnormal local growth and development of the rib margins.

Electrocardiography, although frequently used in the outpatient setting, can be insensitive and non-specific in the case of coarctation of the aorta. Evidence of right ventricular hypertrophy may be seen in the newborn with acute coarctation, due to the history of patency of the arterial duct and exposure of the right ventricle to the systemic circulation before delivery. In older children with prolonged hypertension associated with coarctation, signs of left ventricular hypertrophy can be present. Associated findings suggestive of ischaemia, such as inversion of inferior and/or lateral T waves, may be present if there is also cardiomyopathy associated with prolonged hypertension. A sign of left ventricular hypertrophy that is often missed is inversion of the T wave in lead V6 in the newborn. Before 24 hours of age, the inflection of the T wave in this lead can be either downward or upward; however, after 24 hours of age, the T wave should be upright. Persistent downward inflection of this T wave is associated with severe left ventricular hypertrophy. In the author’s experience, it has been seen in the setting of complete interruption of the aortic arch.

Transthoracic echocardiography continues to be the most important diagnostic test in the ambulatory setting for evaluation of both the anatomy and the haemodynamic effects of coarctation of the aorta. Its qualities of being non-invasive, radiation-free, limited in its discomfort, portable, limited in its need for sedation, relatively inexpensive, relatively rapid, and essentially complete in the ability to assess the various anatomical and physiological aspects of congenital cardiac defects make it the most routinely obtained investigatory device in the clinic; two-dimensional echocardiography, which is typically the first modality of echocardiography utilised in an evaluation, can demonstrate whether aortic narrowing is discrete or diffuse. It can also display derangements in the size of the chambers, including hypertrophy or dilation of the left ventricle and hypertrophy of the right ventricle. In addition, typically associated defects such as bicuspid aortic valve are easily found. The addition of color Doppler allows for visualisation of the acceleration of flow across the site of coarctation, including the direction and degree of increased velocity. The patency of the arterial duct can also be confirmed. Spectral Doppler can further define the degree of arch obstruction, with the estimation of the peak instantaneous gradient. The velocity across the patent arterial duct can give clues to the degree of dependence of the heart on the flow through the duct, with a minimally obstructed arch in conjunction with a restrictive duct suggesting that the requirement of persistent patency of the arterial duct is unnecessary.

When the diagnosis of coarctation is made, evaluation for other associated defects or disease processes should be considered. Bicuspid aortic valve is the most frequently noted associated defect, seen in 30 to 70% of patients with coarctation. Other associated defects include posterior malalignment ventricular septal defect, anomalies of the mitral valve, and other complex cardiac defects including d-transposition of the great arteries, Taussig–Bing malformation, double-inlet left ventricle, Shone’s complex, and tricuspid atresia. In addition, patients with more complex defects would likely generate suspicion, because of other clinical examination findings as well.

In patients diagnosed later in life with coarctation, Takayasu arteritis should be considered as a part of the differential diagnosis, especially if the aortic obstruction is found to be in a location other than the isthmus or transverse arch. Seen in females in their teens or 20s, but described in patients between 6 months of age and adulthood, it is also referred to as “pulseless disease”, as these patients often have diffuse obstruction in the arteries throughout the body. It is often preceded by fever, and is accompanied by fatigue, night sweats, weight loss, diffuse or local pain, and rashes. On examination, pulses may be either grossly asymmetric or absent. A rash may be noted. Hypertension is found, and various bruits can be auscultated. Laboratory studies can demonstrate anaemia as well as evidence of inflammation, including an elevated rate of erythrocyte sedimentation and elevated levels of immunoglobulins G, M, and A. Imaging often shows long-segment obstructions of arteries that are fusiform and irregular in nature. Renovascular obstruction is often seen as well. Therapy involves a combination of immunosuppression and vascular intervention, as appropriate.

Concurrent recognition of genetic syndromes associated with coarctation of the aorta is as important as the recognition of the main lesion itself. These can be determined by the combinations of clinical diagnoses of associated physical findings, imaging of other specific anomalies, and genetic testing. Selected genetic diseases associated with coarctation of the aorta include Turner syndrome (monosomy XO), DiGeorge syndrome (del22q11), William’s syndrome (del7q11), neurofibromatosis, PHACE – malformations of the brain in the posterior fossa, large or complex haemangiomas of the face, arterial anomalies, cardiac anomalies, and eye abnormalities – and Sturge–Weber syndrome, to name a few. Each of these has its own specific set of short- and long-term concerns for which prompt recognition allows for appropriate counselling, surveillance, and intervention – for example, 13% of all girls born with coarctation have Turner syndrome; conversely 14–20% of patients with Turner syndrome have coarctation, either in isolation or with other congenital cardiac defects.Reference Eckhauser, South and Meyers ³ These patients have a high association with incidence of bicuspid aortic valve as well as dissection of the aorta.

Increasingly, there has been interest in trying to find candidate genes that may code for congenital cardiac defects in patients who do not have associated syndromes, but may have sporadic or familial presentation of these defects. Although this search is in its relative infancy, several potential target genes for coarctation have already arisen, including the NOTCH-1 gene on chromosome 9,Reference Freylikhman, Tatarinova and Smolina ⁴ the MTHFR gene on chromosome 1,Reference Kuehl, Loffredo and Lammer ⁵ the TRPM2 gene on chromosome 21,Reference Moosmann, Uebe and Dittrich ⁶ and the FOXC1 gene on chromosome 6.Reference Sanchez-Castro, Eldjouzi and Charpentier ⁷ At this point, it remains to be seen which, if any, of these various genetic anomalies is specifically the cause for coarctation of the aorta. Consequently, it is still premature to consider routine genetic testing in sporadic cases of coarctation at present, unless it is for research purposes.

After the repair or intervention to improve the obstructed aorta, various near- and long-term complications have been recognised. Hemiparesis can occur secondary to spinal cord ischaemia, which can happen if there is a prolonged cross-clamp time in conjunction with a few available arterial collaterals to bypass the segment of the aorta being repaired. Chylothorax after damage to the thoracic duct has been noted. In addition, post-coarctectomy syndrome has been seen, in which the return of pulsatile abdominal flow can lead to mesenteric arteritis with a reflex vasoconstriction. This can manifest anywhere from mild abdominal pain up to presentation of an acute abdomen.

More frequently, residual hypertension occurs, with an incidence between 30 and 75%. Hypertension at rest is associated with residual aortic obstruction, a history of prolonged obstruction of more than 5 years,Reference Liberthson, Pennington and Jacobs ⁸ or patients who have undergone Norwood procedure for hypoplastic left heart syndrome.Reference Murdison, Baffa and Farrell ⁹ It can also be unmasked with exercise stress testing, with occasionally profound hypertensive response. Correia et alReference Correia, Gonçalves and Paiva ¹⁰ found that, although only three of 65 adult patients with a history of coarctation repair may be hypertensive at rest, 11 of the same 65 patients can demonstrate hypertension with stress testing. Predictors of this hypertensive response included the need for therapy with angiotensin-converting enzyme inhibition as well as the magnitude of the peak instantaneous gradient in the descending aorta by transthoracic echocardiography. This residual hypertension is thought to be due to several factors. The aortic pressure wave in the ascending and transverse aorta can exhibit enhanced reflection backwards from the site of repair, even in neonates who have undergone successful repair.Reference Murakami and Takeda ^{11 ,} Reference Murakami, Takeda and Yamazawa ¹² The aorta also shows decreased distensibility and compliance, with increased stiffness.Reference Ou, Celermaier and Joliver ¹³ Another factor that has been reviewed in several studies is the actual geometry of the aortic arch, in which an arch that has the appearance of “Gothic” geometry, or elevated height-to-width ratio, demonstrates higher stiffness and lower carotid distensibility as compared with an arch with a more “Romanesque” appearance, typical of a normal arch.Reference Ou, Celermaier and Mousseaux ¹⁴

Medical therapy for persistent hypertension typically consists of β-blockade. Angiotensin-converting enzyme inhibitors and angiotensin receptor blockers often have good results as well, as long as there is no evidence of significant residual arch obstruction or renal hypoperfusion. There has been some belief that these specific agents may help with vascular re-modelling,Reference Brili, Tousoulis and Antoniades ¹⁵ although this has not been studied completely. Calcium channel blockade can also be efficacious, especially if more than one agent is needed. Significant re-coarctation, although associated with residual hypertension, should undergo re-intervention. The indications for this include a peak-to-peak gradient of >20 mmHg, significant hypertension, or congestive heart failure.

If exercise-induced hypertension is noted, the patient has hypertension at rest, there is significant re-coarctation, or if there is evidence of aneurysm formation, isometric exercise is proscribed. If these do not exist, the Bethesda conference recommendations state that there are no restrictions on exercise.Reference Graham, Driscoll and Gersony ¹⁶ There are some providers, however, who feel that significant isometric activity should be avoided anyway, because of the abnormal vascular properties of the aorta and the higher risk of hypertension.

Over longer-term follow-up, other complications can arise. These include the development of cardiomyopathy, recurrence of the coarctation, or aneurysm of the aorta at the site of repair. This last development is typically associated with repairs involving patch augmentation of the aorta. Cerebral aneurysms have been seen in as many as 10% of patients, as compared with an incidence of 2–3% in the general population.Reference Curtis, Bradley and Wilde ^{17 –} Reference Cook, Hickey and Maul ¹⁹ They occur typically at the circle of Willis, and are seen with greater frequency in correlation with advancing age. At present, a preliminary recommendation is that screening imaging of the cerebral vasculature should start during the fourth decade.Reference Cook, Hickey and Maul ¹⁹ In coarctations that have not undergone repair, endarteritis can occur, because of progressive denudation of the endothelium and exposure of basement membrane proteins in the face of chronic turbulence of the blood flow. There is an increased incidence of coronary artery disease, which may be related to persistent hypertension as well as the previously mentioned abnormality of the reflected pressure wave in the aorta. Along those lines, flow-mediated dilation as well as dilation secondary to nitroglycerine can be impaired, which occurs with increased vascular stiffness.Reference De Divitiis, Pilla and Kattenhorn ²⁰

As with other congenital cardiac defects, education and empowerment of patients and families with the knowledge of both the concerns associated with coarctation and the need for lifelong follow-up are paramount. Efforts at involving the patient in this process are recommended to start when the patient is 12 years of age, although it is ultimately based on the emotional maturity and developmental level of the patient.Reference Sable, Foster and Uzark ²¹ Transfer to an adult CHD specialist, or to a combined programme that also includes adult cardiac care, would then occur at a time the patient is felt to have completed the appropriate transition curriculum as well as the individual institution’s target age.

Recently, the Adult Congenital and Pediatric Cardiology section of the American College of Cardiology published indicators of quality in the care of patients with adult CHD;Reference Gurvitz, Marelli and Mangione-Smith ²² one of the disease processes included was coarctation of the aorta. Their recommendations are summarised in Table 1, including numerous process indicators and a single structure indicator.

Table 1

American College of cardiology indicators of quality in the care of patients with adult congenital heart disease 7.

ACHD=adult congenital heart disease; COA=coarctation of the aorta; CoA=coarctation of the aorta 8

For female patients with coarctation reaching adulthood, one specific issue in which there are increasing data is the aspect of pregnancy in this group of patients. Beauchesne et alReference Beauchesne, Connolly and Ammash ²³ reported on 50 patients with a history of coarctation, with 118 pregnancies among the group. Only one patient had a significant complication of dissection of the aorta; it was notable that this patient had both Turner syndrome as well as a bicuspid aortic valve. A later, but smaller study, showed that there were nine cardiac events in 28 patients with coarctation across 30 pregnancies.Reference Jimenez-Juan, Krieger and Valente ²⁴ These included hypertension in five, cerebrovascular accidents in two, and arrhythmia in two. A newer technique of assessing outcomes in specific patients is the review of large administrative data sets; one such review evaluated 697 deliveries in patients with a history of coarctation of the aorta.Reference Krieger, Landzberg and Economy ²⁵ There were several significant findings in this study. Complications due to hypertension were noted in 24% of the study population, as compared with 8% in the general population. Caesarean section was performed in 42% of the patients with a history of coarctation, as compared with a rate of only 26% in the general population. The study population demonstrated adverse cardiovascular outcomes at an incidence of 4.8%, as compared with a rate of 2.2% in the general population. In addition, patients with coarctation were noted to have both longer hospital stays and higher hospital charges, which would certainly follow the findings of both greater rates of complications plus greater rates of caesarean section.

As with other congenital cardiac defects, the longer we follow-up patients into adulthood, the more we realise that “fixed” is not really fixed. These patients require lifelong follow-up with both screening and therapy for complications such as hypertension, cerebral aneurysms, cardiomyopathy, and aneurysms of the aorta. During pregnancy, the surveillance should be increased further; however, since the first reported repair of coarctation of the aorta by Crafoord in 1944,Reference Crafoord and Nylin ²⁶ data continue to accumulate about the long-term outcomes of these patients. It will be more helpful to be able to care for these patients into older adulthood as we continue to compile information about how these patients do. Although surgical outcomes were measured in 30-day short-term results, both providers and patients need to know about what could be called “very long-term” outcomes, clinical findings, and complications noted in middle and late adulthood; yet, another approach to try to change, and to improve, these outcomes may end up in the actual aortas of the patients themselves. What does the abnormal function of the vasculature mean? Can it be modified or improved? Can it be changed such that the risk of complications presently seen in adulthood can be decreased or eliminated?