| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Endocrinology (J.E.O., C.M., G.S.C.), Imaging (J.A.S.B.), and Cardiology (J.H.), University College London Hospitals, London, United Kingdom W1T 3AA
Address all correspondence and requests for reprints to: Dr. Gerard Conway, Department of Endocrinology, The Middlesex Hospital, London, United Kingdom W1T 3AA. E-mail g.conway{at}ucl.ac.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Echocardiography is widely available, but is limited by two factors: chest wall anomalies in TS can lead to a suboptimal image, and the most appropriate correction for the mean height deficit of 20 cm in women with TS (9) has not been defined. Traditionally, criteria used to define ARD on echocardiography have been derived by correcting aortic measurements for body surface area (Bsa) (10, 11). When applied to women with TS, however, this correction method may result in comparisons being made with children in normal population tables, potentially overestimating the prevalence of ARD in TS.
Magnetic resonance imaging (MRI) can detect degrees of dilatation and coarctation that are not apparent on echocardiography (12), but is limited by its expense and poor tolerability due to claustrophobia, particularly for women with TS. MRI benefits from the intrinsic height adjustment achieved by comparing the dimensions of the ascending and descending aorta (Asc:Desc ratio), with a ratio of greater than 1.5 being considered abnormal in both children and adults (13, 14). This means of intrasubject correction for size is itself prone to error if anomalies occur in the descending aorta.
This study was designed to compare measurements of aortic dimensions obtained by echocardiography and MRI in an adult TS population with controls, to explore their relationships with clinical parameters and their sources of error. We sought to define the distribution of aortic dimensions found in an asymptomatic adult TS population as a reference source for future risk intervention studies.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
One hundred twenty-eight unselected women with TS, albeit from within a specialist adult TS clinic (most of the referrals were from the local pediatric clinic or primary care), were included in the study (median age, 29.2 yr; range, 18.7–55.5 yr). They were offered transthoracic echocardiography and MRI of the aorta, and successful images were obtained in 107 women using both modalities. This study was approved by the University College London Hospitals ethics committee, and all participants gave written informed consent. Clinical history and case notes were reviewed for previous cardiac surgery, history of estrogen replacement, previous GH administration, smoking, and exercise. Clinical parameters were recorded, including anthropometric measurements and a resting supine blood pressure at the right brachial artery. A fasting blood sample was collected for lipid profile, glucose, plasma renin activity, and karyotype. Thirty-six age-matched normal control women (median age, 32.4 yr; range, 20.5–54.0 yr) were recruited, all of whom completed echocardiography; 20 had MRI of the aorta.
Echocardiography
M-Mode and two-dimensional echocardiography were completed in 120 of 128 (94%) women with TS (Acuson Aspen echocardiography machine; Acuson, Mountain View, CA). Echocardiography data were technically inadequate in eight of 128 (6%) women. Measurements of aortic root diameter (ARdm) at the level of the annulus, estimates of left ventricular mass, and the presence of BAV were recorded.
MRI of the aorta
Axial and sagittal oblique MR images of the aorta, as well as cine and phase contrast sequences were obtained (Magnetom plus 1.5T scanner, Siemens, New York, NY) in 115 of 128 (89.8%) women with TS. Thirteen of 128 women (10.2%) were unable to tolerate MRI. Ascending (AAdm) and descending (DAdm) aortic diameters were measured at the level of the bifurcation of the pulmonary artery with calipers on hard copy images and were corrected for the appropriate scale. The presence of ARD on MRI was defined as a ratio (Asc:Desc ratio) of more than 1.5 (13, 14, 15). The presence and degree of coarctation and flow disturbance at the classical aortic coarctation site immediately distal to the origin of the left subclavian artery (ligamentum arteriosum site) were noted.
Statistical analysis
Associations between variables were assessed using Pearson’s correlation coefficient and ANOVA for continuous variables using log transformation where appropriate and controlling for height and body surface area as cofactors when indicated. Spearman’s correlation coefficient and 
2 test were used for categorical variables. Linear regression analysis was performed to assess independent associations of continuous variables, and binary logistic regression analysis was used for categorical variables. Linear regressions and 95% prediction intervals in Fig. 1
were plotted using SigmaPlot 8.0; the differences between these regression lines were assessed by comparing 95% confidence intervals.
|
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Women with TS were shorter, had greater body mass index and systolic and diastolic blood pressure, and were less likely to smoke or perform exercise compared with controls (Table 1). Diabetes mellitus was present in seven women with TS and in no controls. Plasma renin activity and triglyceride concentrations were greater in women with TS compared with controls.
|
. Aortic regurgitation was not exclusively associated with BAV or ARD, occurring in 11 of 22 (50%) women with BAV, 11 of 98 (11%) women with tricuspid AV, and seven of 19 (37%) women with ARD. A normal MRI (i.e. no coarctation, ARD, or other anomaly) was found in 39 of 115 (34%) women with TS compared with 19 of 20 (95%) controls. Entirely normal appearances on both echocardiography and MRI were found in 26 of 107 (24%) women with TS who underwent both studies. Representative MRI scans in Fig. 2 show the two major patterns of aortic pathology: a dilated aortic root and coarctation site narrowing. Women with karyotype 45,X were more likely to have a cardiovascular imaging abnormality compared with other karyotypes (data not shown), but variations in the use of GH in childhood, the timing of estrogen use, use of antihypertensives, blood pressure, lipids, plasma renin activity, smoking, and exercise were not associated with cardiovascular abnormality of any type.
|
|
) was immediately referred for surgery. She had a BAV, but no other obvious predictive features: she was not hypertensive, her karyotype was mosaic 45,X/46,XX, and she had only 1 yr of estrogen deficiency. The inclusion of data from this individual was found to have no major influence on any of the findings when calculations were repeated after exclusion (data not shown). Inclusive data are therefore used throughout, except in Fig. 2, where they were considered inappropriate. Comparison of TS and control women: aortic dimensions
With regard to echocardiography, absolute measurements of ARdm were similar in TS patients and controls (Table 3) despite the slightly higher blood pressure of TS patients. When controlling for differences in Bsa and height between TS and controls, however, ARdm was significantly greater in TS, with height adjustment having the greatest effect (Table 3). To apply height correction for ARdm, we used the equation: adjusted ARdm = ARdm + (1.6 – height) x 1.74, with 1.6 being the mean height of controls and 1.74 being the slope of the regression line of ARdm and height in women with TS. Using this adjustment and applying the common criterion for ARD on echocardiography of 3.4 cm (incidentally the upper limit of ARdm in our control group), 19 of 120 (16%) women qualified for ARD. Absolute measurements of ARdm were used to derive TS-specific reference ranges shown in Fig. 1A.
|
). Coarctation site abnormalities may be associated with pre- and poststenotic dilatation, potentially confounding the Asc:Desc ratio (Fig. 2C). This aspect was investigated in 106 women with TS, excluding the nine who had had previous coarctation surgery. Both AAdm and DAdm were significantly greater in 40 women with coarctation site abnormalities than in 66 women with TS who had no coarctation, but no alteration to the Asc:Desc ratio (Table 3). Individual AAdm were used to derive a TS-specific normal distribution (Fig. 1B).
|
). Determinants of ARD
Age and BAV were the only two variables that had significant associations with aortic measurements. The strength of association between age and aortic dimensions was more pronounced in women with TS than in controls (Fig. 1). Aortic dimensions were greater in those women with TS who had a BAV than in those without [ARdm, 3.11 ± 0.71 vs. 2.80 ± 0.36 cm (P = 0.003); Asc:Desc ratio, 1.61 ± 0.29 vs. 1.40 ± 0.20 (P < 0.001)]. There was also a positive association between ARdm and left ventricular mass (r = 0.317; P = 0.004), which persisted when women with a BAV were excluded (r = 0.286; P = 0.016), but was diminished when controlled for blood pressure.
TS and control women: coarctation site abnormalities on MRI
Of the 106 women with no previous cardiac surgery, 40 (38%) had previously undiagnosed abnormalities at the classical coarctation site, compared with one of the 20 control women (by 2 test, P < 0.001). The abnormalities in the TS women were defined as a visible flow disturbance on the jet vortex at this site in 24 of these 40 women (or 24 of 106; 22.6%) and a visible narrowing of the aortic lumen in 23 women (23 of 106; 21.7%). In eight TS women (eight of 106; 7.5%), the impingement on the aortic lumen was at least 30%, and flow disturbance was sufficiently great to warrant classification as occult coarctation (Fig. 2C). One control (one of 20; 5%) had a small notch at the coarctation site with no associated flow disturbance, which was deemed to be hemodynamically insignificant.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This is the largest study to date evaluating the cardiovascular system by both imaging modalities in adults with TS, and for the first time the relative merits of each can be characterized. The two modalities differ in the site of measurement of the aorta; echocardiography usually measures the ARdm at the annulus, whereas AAdm on MRI is measured slightly distal, at the level of bifurcation of the pulmonary arteries. It is therefore not surprising that the two imaging modalities identify different individuals qualifying for ARD. With long-term follow-up of this cohort we may be able to determine which group has the greatest risk of dissection. In this regard it is notable that the effect of age on the aorta was a particular feature of the TS women.
Among the few studies that have evaluated the use of cardiovascular MRI in TS, the prevalence of ARD on MRI has varied between 12.5% and 33.3% depending on the exact criteria used (12, 16), compared with 33% in this study. Data from our control group confirmed the computed tomography-derived definition of Asc:Desc ratio greater than 1.5 as a reliable criterion for ARD on MRI. The utility of this ratio is hampered by the fact that the DAdm is slightly increased in women with TS, eradicating a significant difference in the mean ratio compared with controls. This would imply that the wall of the entire aorta might be defective in TS, and the use of the ratio might be falsely reassuring. Also, coarctation site defects are associated with increases in both AAdm and DAdm, increases that may cancel each other out on calculation of the ratio. In women with any degree of coarctation, therefore, absolute values of AAdm are likely to be better predictors of the risk of dissection.
MRI has not been advocated as the investigation of first choice for evaluation of the aorta in TS in view of its limited availability and greater cost and the fact that a significant number of women with TS (10.2% in this study) are unable to tolerate the scan because of claustrophobia. MRI is of greatest advantage in the assessment of the coarctation site, which is not always visible on echocardiography because of the shape of the chest wall in women with TS. There is clearly a spectrum of pathology at the coarctation site, and there is no consensus as to what constitutes a significant defect worthy of intervention. Our finding of 8% of women with an important occult coarctation agrees with similar studies quoting 6% (16) and 12% (12). The goal now will be to develop a strategy to define risk in terms of flow disturbance.
In conclusion, we have shown that aortic imaging by echocardiography and MRI provides complementary information, but it is only with longitudinal study of this population that we will determine how each relates to future morbidity. Ideally, all women with TS will routinely have cardiovascular imaging by both techniques, and certainly echocardiography should be universal. Accepting that MRI may not be available for all women with TS, a practical way forward would be to target high risk subgroups, such as those with a BAV, those pursuing ovum donation, or those in whom clear echocardiographic imaging is unobtainable. Given the difficulties of correcting for short stature, we have defined TS-specific reference ranges for absolute measurements of both ARdm on echocardiography and AAdm on MRI in the hope that they will aid future interpretation of these measurements and help define thresholds for risk management strategies.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: AAdm, Ascending aortic diameter; ARD, aortic root dilatation; ARdm, aortic root diameter; Asc:Desc, ascending to descending aorta ratio; BAV, bicuspid aortic valve; Bsa, body surface area; DAdm, descending aortic diameter; MRI, magnetic resonance imaging; TS, Turner syndrome.
Received June 9, 2004.
Accepted August 17, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B. E. Hjerrild, K. H. Mortensen, and C. H. Gravholt Turner syndrome and clinical treatment Br. Med. Bull., June 1, 2008; 86(1): 77 - 93. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Lopez, K. L. Arheart, S. D. Colan, N. S. Stein, G. Lopez-Mitnik, A. E. Lin, M. D. Reller, R. Ventura, and M. Silberbach Turner Syndrome Is an Independent Risk Factor for Aortic Dilation in the Young Pediatrics, June 1, 2008; 121(6): e1622 - e1627. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Sachdev, L. A. Matura, S. Sidenko, V. B. Ho, A. E. Arai, D. R. Rosing, and C. A. Bondy Aortic Valve Disease in Turner Syndrome J. Am. Coll. Cardiol., May 13, 2008; 51(19): 1904 - 1909. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. A. Matura, V. B. Ho, D. R. Rosing, and C. A. Bondy Aortic Dilatation and Dissection in Turner Syndrome Circulation, October 9, 2007; 116(15): 1663 - 1670. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Bondy and for The Turner Syndrome Consensus Study Group Care of Girls and Women with Turner Syndrome: A Guideline of the Turner Syndrome Study Group J. Clin. Endocrinol. Metab., January 1, 2007; 92(1): 10 - 25. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Bondy, V. K. Bakalov, E. D. Lange, and I. Ceniceros Deficient Medical Care for Adults with the Turner Syndrome Ann Intern Med, December 5, 2006; 145(11): 866 - 867. [Full Text] [PDF]
|
|
|
|
 |
|
 K. J BEE, D. WILKES, R. B DEVEREUX, B. B LERMAN, H. C DIETZ, and C. T BASSON Structural and Functional Genetic Disorders of the Great Vessels and Outflow Tracts Ann. N.Y. Acad. Sci., November 1, 2006; 1085(1): 256 - 269. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N H Andersen, B E Hjerrild, K Sorensen, E M Pedersen, K Stochholm, L C Gormsen, A Horlyck, J S Christiansen, and C H Gravholt Subclinical left ventricular dysfunction in normotensive women with Turner's syndrome. Heart, October 1, 2006; 92(10): 1516 - 1517. [Full Text] [PDF]
|
|
|
|
 |
|
 M D C Donaldson, E J Gault, K W Tan, and D B Dunger Optimising management in Turner syndrome: from infancy to adult transfer Arch. Dis. Child., June 1, 2006; 91(6): 513 - 520. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. A. Bondy, P. L. Van, V. K. Bakalov, and V. B. Ho Growth Hormone Treatment and Aortic Dimensions in Turner Syndrome J. Clin. Endocrinol. Metab., May 1, 2006; 91(5): 1785 - 1788. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Ostberg, A. E. Donald, J. P. J. Halcox, C. Storry, C. McCarthy, and G. S. Conway Vasculopathy in Turner Syndrome: Arterial Dilatation and Intimal Thickening without Endothelial Dysfunction J. Clin. Endocrinol. Metab., September 1, 2005; 90(9): 5161 - 5166. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
, controls. Lines represent the linear regression plot with 95% prediction lines for the distribution. Solid lines, TS; dashed lines, controls.
