Author + information
- Received November 14, 2002
- Revision received March 4, 2003
- Accepted March 20, 2003
- Published online June 18, 2003.
- Marcello de Divitiis, MD*,
- Carlo Pilla, MD*,
- Mia Kattenhorn, BSc*,
- Ann Donald, AVT*,
- Mariutzka Zadinello, MD*,
- Sharon Wallace, BA*,
- Andrew Redington, FRCP* and
- John Deanfield, FRCP*,* ()
- ↵*Reprint requests and correspondence:
Prof. John E. Deanfield, Vascular Physiology Unit, Institute of Child Health, 30 Guildford Street, London WC1N 1EH, United Kingdom.
Objectives We sought to evaluate the determinants of hypertension during daily life and left ventricular (LV) hypertrophy in patients with successfully repaired coarctation of the aorta (CoA), as well as their relationship to abnormalities of arterial function.
Background Arterial hypertension may recur late after repair of CoA, which is related to a more adverse outcome. Furthermore, patients with normal resting blood pressure (BP) may have hypertension during daily life and LV hypertrophy. The determinants of these two adverse prognostic factors have not been investigated.
Methods We studied 72 patients (9 to 58 years of age) who underwent coarctation repair at age 0.1 to 480 months (42 [60%] at <1 year) and had been followed up for 155 ± 76 months. They underwent ambulatory BP monitoring, echocardiography for LV mass, studies of brachial artery responses to flow (i.e., flow-mediated dilation [FMD]) and glyceryl trinitrate (GTN), and determination of pulse wave velocity (PWV) and measures of arterial reactivity and stiffness. Findings were compared with those of 53 healthy volunteers.
Results Patients had higher 24-h systolic BP and LV mass than controls. Both endothelium-dependent FMD and the response to the smooth muscle dilator GTN were reduced, and PWV was increased. There was a negative independent correlation between GTN response and 24-h systolic BP in both patients and control subjects. Systolic BP at 24 h was an independent predictor of LV mass, having an accentuated impact in coarctation subjects as compared with controls.
Conclusions In patients with repaired coarctation, reduced vascular reactivity is associated with hypertension during daily life and with increased LV mass, both of which are important predictors for late morbidity and mortality.
Despite successful repair, patients with coarctation of the aorta (CoA) have an increased risk of late morbidity and mortality from heart failure and coronary and cerebrovascular disease (1,2). Arterial hypertension, which occurs in up to one-third of these subjects, even in the absence of re-coarctation, is a major adverse prognostic factor (1–4). Patients with normal blood pressure (BP) at rest often show an abnormal systolic BP response to physical exercise and normal daily life activities (5–7). This phenomenon may have clinical relevance, as ambulatory systolic BP has a stronger impact than resting BP on both end-organ damage and cardiovascular outcome (8,9). Left ventricular (LV) mass is another independent predictor of cardiovascular events (10)and may be increased even in normotensive patients late after coarctation repair (6,11).
The determinants of hypertension and LV mass, as well as their relationship, have not been studied during long-term follow-up after coarctation repair. We have previously shown abnormalities of vascular function in upper limb conduit arteries late after successful surgery, with impaired brachial arterial reactivity and increased stiffness (12,13). We now examine the relationship between BP during daily life, LV mass, and the influence of conduit artery function on these two outcome variables.
We identified 261 patients (9 to 60 years of age) who had survived repair of isolated CoA at the Great Ormond Street Hospital for Children between 1970 and 1998, without major associated cardiovascular abnormalities such as ventricular septal defect or aortic and mitral valve functional abnormalities and with no evidence of re-coarctation at the last outpatient visit (defined as >3.5 m/s velocity by continuous wave Doppler in the aortic arch with a diastolic tail) in the past three years, and invited them to participate in the study by letter. Of the 80 patients who replied, 72 (50 males and 22 females) agreed to participate. The patients who took part in the study were comparable, in terms of age, gender distribution, and age at the time of repair, to the patients who did not reply to the letter. We also studied 53 healthy age-matched volunteers (37 males and 16 females), 25 of whom were siblings of the patients and 28 who were hospital employees. Coarctation repair had been performed at a median age of 3.1 months (range 0.1 to 480 months), and 42 subjects (60%) had undergone surgery during the first year of life. Forty-nine patients had repair by end-to-end anastomosis (68%), 14 by a subclavian flap (19.4%), 5 by patch angioplasty (6.9%), 1 by repair with graft interposition (1.4%), 1 by tube bypass (1.4%), and 2 by primary percutaneous balloon aortoplasty (2.8%). Four patients (three after end-to-end anastomosis and one after subclavian flap repair) had subsequently undergone balloon aortoplasty, and three (two with end-to-end anastomosis and one with patch angioplasty) had undergone re-operation because of re-coarctation >1 year before the study. Only two patients (2.8%) were receiving antihypertensive medication (beta-blockers). These were discontinued at least four days before the study, and data on these two patients were not included in the LV mass analysis. Thirty-five patients (49%) and one control subject had a bicuspid aortic valve. The mean duration of follow up was 155 ± 76 months.
Patients were ready for the procedure in the morning after having fasted for 12 h. After recording height and weight, a venous blood sample was taken for routine biochemistry measurements (e.g., glucose, cholesterol, triglycerides, and lipoproteins). Supine BP was measured in the right arm and right leg by an automatic oscillometric device (Accutor 4, Datascope Corp., Montvale, New Jersey) following a 5-min rest. Arterial hypertension was defined in adults as systolic BP >140 mm Hg and/or diastolic BP >90 mm Hg, and in children as systolic and/or diastolic BP above the 95th percentile for age, according to Rosner et al. (14).
All subjects underwent transthoracic echocardiography for the assessment of LV mass and for confirmation of the absence of re-coarctation, aortic aneurysm, and functional abnormalities of the mitral and aortic valves. Noninvasive assessment of endothelium-dependent flow-mediated dilation (FMD) and endothelium-independent dilation to glyceryl trinitrate (GTN) of the right brachial artery was then performed, and in 66 coarctation patients and 50 control subjects, vascular stiffness was assessed by measurement of the brachio-radial pulse wave velocity (PWV). Sixty-five patients and 44 controls volunteered to perform 24-h ambulatory BP monitoring. The institutional ethics committee approved the study, and written informed consent was obtained from each subject and/or family.
Brachial artery FMD and GTN-induced dilation (GTND)
All subjects were studied in a supine position in a quiet, air-conditioned room (22°C to 25°C). Vascular ultrasound assessment of endothelium-dependent and -independent dilation of the brachial artery was performed using a technique developed by our group (15), which has been shown to be accurate and reproducible (16). A B-mode scan of the right brachial artery was obtained in a longitudinal section between 5 and 10 cm above the elbow, and the responses to increased flow (FMD) (reactive hyperemia induced by inflation and deflation of a pneumatic cuff around the forearm) and to sublingual GTND were evaluated. The dose of GTN (25 μg) administered was lower than that used in some of our previous studies (12,15)and was selected to achieve a dilation more closely related to that obtained for FMD (13,17).
Brachial artery diameter was measured off-line by a semi-automatic edge-detection system (Information Integrity, Boston, Massachusetts), as previously described (13,15,18). Volume blood flow velocity (the stimulus to the vessel dilation) was calculated by measuring the brachial flow velocity–time integral, off-line, from video-recorded tapes at baseline and from 5 to 120 s after cuff release.
The PWV (the velocity of propagation of the arterial pulse from the brachial to radial probe) was measured as the index of arterial stiffness by using a well-validated photo-plethysmographic technique (19,20). Signals from two photo-probes placed over the right brachial and radial arteries were converted into pulse waveforms, and the transit time was determined from the time delay between the foot of the corresponding brachial and radial pulse waves. Thus, PWV was calculated as the ratio of the distance between the two probes and the transit time.
Patients were studied in the left supine position, using an Acuson Aspen system (Mountain View, California) with a multifrequency probe (2.5 to 4 MHz). The ascending aorta and aortic arch were visualized using a high, long-axis view and the suprasternal view to exclude the presence of an aortic aneurysm. Left ventricular measurements were taken from two-dimensional guided M-mode tracings, as recommended by the American Society of Echocardiography (21). Measurements were performed on 4 to 6 beats and then averaged. Left ventricular mass was calculated using the Penn Convention and was then adjusted for body surface area (22). Left ventricular hypertrophy was considered present if the LV mass index (LV mass/body surface area) exceeded 131 g/m2in men and 110 g/m2in women. Relative wall thickness was calculated as: (2 × posterior wall thickness)/end-diastolic diameter. Fractional shortening, ejection fraction, cardiac output, and total peripheral resistance were calculated by using standard formulae.
Ambulatory BP monitoring
Studies were performed using a validated oscillometric device (TM-2430, A&D Medical, Milpitas, California) (23)with an appropriate cuff size on the right arm. Blood pressure measurements were recorded automatically every 15 min from 7:00 amto 11:00 pm(daytime BP) and every 30 min from 11:00 pmto 7:00 am(night-time BP). Measurements were excluded if there was an inconsistent >20 mm Hg increase or decrease in systolic and/or diastolic BP from one measurement to another. Studies were excluded if there was an interval of >2 h of invalid or absent measurements. Adults were considered hypertensive if their daytime systolic and/or diastolic BP was >132/85 mm Hg in men and >129/82 mm Hg in women, according to the reference values of the studio della Pressioni AMbulatoriali E Loro Associazioni (PAMELA) (24). Children were considered hypertensive if their 24-h systolic and/or diastolic BP exceeded the 95th percentile of the reference values provided by Soergel et al. (25).
All data were stored and analyzed using the SPSS statistical package (SPSS Inc., Chicago, Illinois). The normality Kolmogorov-Smirnov test was performed to determine whether continuous variables were normally distributed.
Data are presented as the mean value ± SD. Age and age at the time of surgical repair, which presented a skewed distribution, are presented as the median value (range). A comparison between groups of continuous variables was performed by using the Student ttest, whereas age was compared by using the rank Mann-Whitney Utest. Categorical variables were compared by using the chi-square test. The relationships between clinical and functional variables were analyzed with the Pearson correlation test. Multilinear regression analysis was used to study the independent predictors of ambulatory BP in a model with 24-h systolic BP as the dependent variable and age, height, weight, study group, brachial GTND, and mean resting BP as independent variables. Average 24-h systolic BP was chosen as the target variable because of its well-known prognostic impact (9). Mean resting BP, which is normally used as a reference value for BP damage to the vascular wall (20), was included in the model to study the impact of GTN over the range of BPs. The determinants of LV mass were also studied using multilinear regression analysis, with LV mass as the dependent variable and age, height, weight, study group, and 24-h systolic BP as independent variables. The same analysis was then repeated after replacing 24-h systolic BP with daytime and then night-time systolic BP as independent variables. Parameters not showing a normal distribution (such as age) were log-transformed before this analysis. A p value <0.05 was considered to be statistically significant (two-tailed).
Age was similar between the two groups (median age 19.8 years [range 9 to 60] in coarctation patients vs. 19 years [range 9 to 53] in control subjects, p = 0.8). There was also no difference between patients and controls in body surface area (1.6 ± 0.3 vs. 1.6 ± 0.3 m2respectively; p = 0.9) and body mass index (22.2 ± 3.7 vs. 21.5 ± 3.2, respectively; p = 0.2).
Blood glucose and lipids (total cholesterol, triglycerides, and high-density lipoprotein) were also comparable between the two groups.
BP (Table 1)
All BP measures were significantly higher in patients than in control subjects. Fifteen patients with repaired coarctation (21%) had systolic hypertension at rest. One of these also had diastolic hypertension.
Adequate 24-h BP recordings were obtained in 57 (87%) of 65 patients and 37 (84%) of 44 control subjects, with no age or anthropometric differences between the patients and controls. Ambulatory systolic hypertension was detected in one control subject and in 31 patients with repaired coarctation (54%), 22 of whom (71% and 33%, respectively, of the total group) had normal BP at rest. Five coarctation patients had both high systolic and diastolic BP. One patient with repaired coarctation had hypertension at rest but normal ambulatory BP findings.
Echocardiography (Table 2)
The echocardiographic study was adequate for LV mass measurements in 62 (86%) of 72 patients and 52 (98%) of 53 control subjects. Left ventricular mass was higher in the coarctation patients. Six patients with repaired coarctation (9.7%) had LV hypertrophy, all with an eccentric geometric pattern (relative wall thickness <0.45).
Vascular function (Table 3)
The coarctation patients had a higher PWV and reduced FMD and GTND. We found no relationship between the surgical technique and vascular function, BP, and LV mass in our population, but we cannot exclude a possible type II error.
Determinants of ambulatory BP
In coarctation patients, 24-h and daytime systolic BP and pulse pressure were related to age and body mass index and were inversely related to GTND (Fig. 1). In the control subjects, 24-h and daytime systolic BP were inversely related to GTND (Fig. 1). No correlation was found between PWV or FMD and either resting BP or any of the ambulatory BP measures in both patients and controls. On multilinear regression analysis, 24-h BP was predicted by GTND, weight, and the presence of repaired coarctation (r2= 0.48) (Table 4).
Determinants of LV mass
In patients with repaired CoA, LV mass index was related to age, age at repair, body mass index, resting brachial artery diameter, resting systolic BP, 24-h and daytime systolic BP (Fig. 2), and 24-h and daytime pulse pressure. Similar correlations were found when LV mass was indexed for height or height2.7.
There was no correlation between LV mass and measures of vascular function or ambulatory diastolic BP measurements and night-time BP measurements.
In the control group, the LV mass index was related to body mass index, resting systolic BP, and 24-h and daytime systolic BP (Fig. 2). Furthermore, the LV mass index was related to resting brachial artery diameter and inversely related to GTND.
On multilinear regression analysis, weight, 24-h systolic BP, and the study group were independent predictors of LV mass (r2= 0.59) (Table 5). The same analysis performed after replacing 24-h systolic BP with daytime and then night-time systolic BP as the independent variable showed that daytime, but not night-time, systolic BP also significantly predicted LV mass (r2= 0.60).
This study has confirmed the presence of marked abnormalities of ambulatory BP regulation late after successful repair of CoA, which are not present at rest in more than one-third of patients. The novel findings in our study include the demonstration of an important role of abnormal vascular responses in the development of hypertension during daily life and the accentuated hypertrophic adaptation of the LV to afterload after coarctation repair. These findings are of potential relevance for the development of clinical complications in later life, even in these patients with an excellent anatomic result from surgery.
Preoperative hypertension normally regresses after successful repair of coarctation, but the proportion of patients with normal BP falls progressively during long-term follow-up (2). This is related to adverse outcomes (1–4). The characteristic findings in our patients included systolic hypertension during both day and night, with an accentuated systolic BP and pulse pressure response during normal daytime activities, even in those with normal resting BP. This pattern is known to be associated with more rapidly progressive secondary organ damage in the general population (9,26).
The pathophysiology of this late hypertension is at least partly related to upper limb conduit artery dysfunction. We have previously demonstrated abnormalities of both elastic properties and reactivity in the conduit arteries of the upper limb, which are not present in lower limb arteries (12,13). These include increased PWV and reduced FMD and GTND (12,13). We now demonstrate that the dynamic response of the brachial artery to GTN is a strong predictor of systolic BP and pulse pressure over 24 h and in particular during the active part of the day. Interestingly, these relationships were found in both coarctation patients and control subjects who had a lower range of BPs. We have previously reported a similar relationship between brachial artery GTND and peak systolic BP during physical exercise in the same type of patients (12). Interestingly, in our multilinear regression analysis, for a given arterial response to GTN, patients had a higher systolic BP than controls. The blunted GTND may be due to abnormal smooth muscle cell relaxation or structural abnormalities of the arterial wall, such as increased collagen and reduced elastin content (27). The finding of increased PWV after coarctation repair supports the latter explanation, although a significant correlation between PWV, a resting measure of vascular stiffness, and ambulatory BP was not observed. The cause-and-effect relationship of vascular dysfunction and hypertension is still controversial. The finding of vascular abnormalities in the normotensive offspring of patients with essential hypertension (28)and the predictive nature of arterial stiffness in early life for later hypertension (29)argues in favor of a contribution by vascular abnormalities to the development of clinical hypertension. Other potential mechanisms that might influence ambulatory BP, including baroreflex sensitivity (27)and sympathetic hyperactivity, still need further investigation in coarctation patients.
Left ventricular mass is another important predictor of cardiovascular outcome in the general population, independent of BP (9,10). It is known to be strongly influenced by age, body weight, and ambulatory systolic BP in both healthy subjects and those with hypertension (8,9). In this study, systolic BP during daily-life activities, rather than resting systolic BP, was strongly related to LV mass. Furthermore, multilinear regression analysis demonstrated that ambulatory systolic BP, body height and weight, and the presence of repaired coarctation had a significant influence on LV mass, independent of age and resting BP, suggesting that patients with repaired coarctation have a greater LV response to a given BP and weight than healthy subjects, as also seen for the relationship between GTN response and BP. There are several possible explanations for this observation. There might be a genetically determined hypertrophic response in coarctation, or LV mass may be determined by hemodynamic or hormonal differences that persist even after successful repair. These include changes in arterial pressure wave propagation in the aorta and/or an increase in aortic stiffness, as well as resetting of the sympathetic nervous and/or renin-angiotensin system. Doppler velocities in the aortic arch were not related to LV mass, suggesting that minor degrees of residual or re-coarctation were not responsible for these findings. Furthermore, no association was found between the type of repair and any measurement of vascular function and LV mass.
It is noteworthy that approximately 10% of the coarctation patients had LV mass measures that could be considered to represent pathologic hypertrophy, with a clear adverse prognostic impact. The increase in LV mass seen in many of the coarctation patients remains less than that associated with the increased risk in adult population studies (10). However, this significant increase in LV mass in our young coarctation patients is likely to progress with age, and its potential to evolve to a high-risk level will require serial evaluations.
Our patients had a wide age range and age at repair. Unlike many previous reports, more than one-half had undergone repair within the first year of life, reflecting current surgical practice. Despite early surgery, BP, LV mass, and vascular function were often abnormal, emphasizing the need for close surveillance of all patients during long-term follow-up.
We excluded patients who had significant associated structural cardiac anomalies but included patients with bicuspid aortic valves without functional abnormalities. It is now recognized that such patients may have diffuse vascular abnormalities with abnormal composition of the aortic wall (30). In this study, however, the presence of a bicuspid aortic valve had no independent effect on BP, LV mass, or conduit artery function. The echocardiographic criteria used to evaluate the anatomy and physiology of the aorta may still have enabled classification of some subjects.
We have not attempted to link the pathophysiologic abnormalities to cardiovascular outcome. Nevertheless, our findings suggest that the well-known adverse prognostic influence of BP elevation and LV mass increase in the general population are likely to be amplified in coarctation patients.
Our control group consisted of both siblings and unrelated individuals. If there were an inherited contribution to vascular and/or ventricular structure and function, this might have reduced the differences observed in these measures between patients and control subjects. Furthermore, we did not find a difference in vascular function between siblings and unrelated controls. We used echocardiography to exclude the presence of aortic aneurysm and a Doppler measurement to evaluate flow around the aortic arch. It is possible that our exclusion criteria may have resulted in the inclusion of some subjects with mild abnormalities of the repaired segment, which might have influenced flow propagation.
We used M-mode measurement of LV mass, as reported in numerous previous publications. These measurements have been linked to an adverse outcome in large epidemiologic studies (8–10). New techniques, including three-dimensional echocardiography or magnetic resonance imaging, might provide better imaging of global LV structure and shape and could be used in future studies that may result in a more reproducible measurement.
Using our strict quality criteria, LV mass measurements were acceptable in 86% of the coarctation patients. This figure is somewhat lower than those reported in patients with essential hypertension and is likely due to the scanning difficulties in patients who had undergone previous cardiac surgery.
Impairment of vascular mechanics contributes to the abnormalities of BP regulation and indirectly to increases in LV mass late after successful coarctation repair. Therefore, studies of interventions such as drug therapy on vascular function should be considered in these subjects (31). Our findings support the need for careful and continued follow-up and assessment in patients who have undergone “successful” repair of CoA.
☆ Research at the Institute of Child Health and Great Ormond Street Hospital for Children National Health Service Trust was supported by R&D funding received from the NHS Executive. The Coronary Artery Disease Research Association (CORDA) funded Mrs. Ann Donald, through a legacy from the late Marian Silcock.
- blood pressure
- coarctation of the aorta
- flow-mediated dilation
- glyceryl trinitrate
- glyceryl trinitrate-induced dilation
- left ventricle/ventricular
- pulse wave velocity
- Received November 14, 2002.
- Revision received March 4, 2003.
- Accepted March 20, 2003.
- American College of Cardiology Foundation
- Cohen M.,
- Fuster V.,
- Steele P.M.,
- et al.
- Clarkson P.M.,
- Nicholson M.R.,
- Barrat-Boyes B.G.,
- et al.
- Leandro J.,
- Smallhorn J.F.,
- Benson L.,
- et al.
- Mancia G.,
- Zanchetti A.,
- Agabiti-Rosei E.,
- et al.
- Gardiner H.M.,
- Celermajer D.S.,
- Sorensen K.E.,
- et al.
- Celermajer D.S.,
- Sorensen K.E.,
- Gooch V.M.,
- et al.
- Sorensen K.E.,
- Celermajer D.S.,
- Spiegelhalter D.J.,
- et al.
- Leeson P.,
- Thorne S.,
- Donald A.,
- et al.
- Greenwald S.E.,
- Denyer H.T.,
- Sobeh M.S.
- Sahn D.J.,
- DeMaria A.,
- Kisslo J.,
- et al.
- Devereux R.B.,
- Reichek N.
- Sehested J.,
- Baandrup U.,
- Mikkelsen E.
- Taddei S.,
- Virdis A.,
- Mattei P.,
- et al.
- Niiwa K.,
- Perloff J.K.,
- Bhuta S.M.,
- et al.