| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Department of Pediatrics, Divisions of Cardiology (J.v.d.B., L.v.O.-G., W.A.H.) and Endocrinology (E.M.N.B., S.M.P.F.d.M.K.-S.), Erasmus Medical Center, Sophia Childrens Hospital, 3015 GJ Rotterdam, The Netherlands; and Departments of Radiology (J.v.d.B., P.A.W., P.M.T.P., W.A.H.) and Epidemiology and Biostatistics (W.C.J.H.), Erasmus Medical Center, 3015 GE Rotterdam, The Netherlands
Address all correspondence and requests for reprints to: W. A. Helbing, Department of Pediatric Cardiology, Erasmus Medical Center, Sophia Childrens Hospital, Sp-429, Dr Molewaterplein 60, 3015 GJ Rotterdam, The Netherlands. E-mail: w.a.helbing{at}erasmusmc.nl.
| Abstract |
|---|
|
|
|---|
Methods: TS patients and healthy women prospectively underwent cardiac magnetic resonance imaging. Ventricular two-dimensional tomographic cine data were acquired to obtain biventricular volume, mass, and ejection fraction. Atrioventricular valve flow measurements were performed using a two-dimensional flow-sensitized sequence. Flow velocity curves were calculated and indices of biventricular diastolic filling were derived.
Results: Thirty-one patients [mean (SD) age 20 (2) yr, body surface area 1.75 (0.15) m2, 5 (2) yr after GH discontinuation] and 23 normal control women [age 21 (2) yr, body surface area 1.80 (0.13) m2] were included. Compared with controls, patients had smaller mean end-diastolic volumes [right ventricle (RV), 84 (11) ml/m2 vs. 79 (10), P = 0.02; left ventricle (LV), 81 (10) vs. 72 (9), P < 0.001], end-systolic volumes [RV 38 (7) ml/m2 vs. 36 (6), P = 0.04; LV 34 (5) vs. 29 (4), P < 0.001], and stroke volumes [RV 46 (6) ml/m2 vs. 43 (6), P = 0.03; LV, 47 (7) vs. 44 (4), P = 0.02]. Patients had a higher mean heart rate [79 (13) beats/min vs. 71 (10), P < 0.05]. Biventricular ejection fraction, mass, cardiac output, and diastolic filling pattern were comparable.
Conclusion: After discontinuation of GH treatment TS patients showed no myocardial hypertrophy and well-preserved biventricular function. Ventricular volumes were smaller in Turner patients, compared with controls, whereas mean heart rate was higher. These last observations may be part of the natural development in TS and not linked to GH treatment, which at this point we consider safe.
| Introduction |
|---|
|
|
|---|
The GH dose normally used for treatment of short stature in TS is relatively high [1.4 mg/m2 body surface area (BSA) per day in TS, compared with 0.7 mg/m2 BSA per day in GH deficiency]. This raises concerns on possible cardiovascular side effects. Several papers addressed these concerns during GH treatment and reported normal LV volumes without hypertrophy in TS, compared with controls (7, 8). However, cardiac data collected after discontinuation of GH treatment are scarce. Furthermore, previous studies focused on LV size and function only. We hypothesize that myocardial effects of systemic GH treatment if present are biventricular. As such the aim of this study was to prospectively assess biventricular size and function in TS at least 6 months after discontinuation of GH treatment and compare results with healthy controls. For this purpose cardiac magnetic resonance imaging (CMR), the gold standard for determination of biventricular volumes, function, and wall mass, was used.
| Subjects and Methods |
|---|
|
|
|---|
Because GH treatment has been an accepted treatment in TS for years, age-matched untreated TS patients are not available at our institution. Therefore, 23 healthy age-, gender-, and BSA-matched controls were included. Height, weight, diastolic blood pressure, and systolic blood pressure were measured. BSA, body mass index (BMI), and mean blood pressure were calculated. Besides the generally accepted contraindications for magnetic resonance imaging (MRI), clinically important left-sided pressure overload (aortic valve stenosis gradient > 35 mm Hg and a gradient > 25 mm Hg due to aortic coarctation) was used as an exclusion criterion. We previously published data from this population on aortic distensibility and aortic size (9).
CMR protocol
Imaging was performed on a 1.5-T Signa CV/i scanner using software releases (versions 8.4 and 9.1; GE Healthcare, Milwaukee, WI) with a four-channel phased-array coil. All images were acquired during breath holds at end expiration. A standardized multiphase, multislice volumetric ventricular data set in short axis direction was acquired. Imaging was performed using two-dimensional (2D) fast imaging using steady-state free precession (FIESTA) with the following imaging parameters: flip angle 45°; echo time set at minutes full; repetition time 3.4–3.6 msec; 8–9 mm slice thickness; 0–1 mm interslice gap; 12 views/segment; readout bandwidth 111Khz; a square field of view 30–34 cm); and a scanning matrix of 160 x 128. Twenty-four phases per cardiac cycle were reconstructed retrospectively.
Mitral valve and tricuspid valve flow measurements were performed perpendicular to flow using a standard 2D flow-sensitized (phase velocity encoded) scan. Flow sensitivity of the sequence was set at 120 cm/sec. The flow sensitivity of the sequence was increased with phase aliasing. Scans were retrospectively gated. Temporal resolution was approximately 35–50 msec per cardiac phase. Thirty phases were reconstructed retrospectively. Imaging parameters were: 2D fast spoiled gradient-echo, repetition time 6–7 msec, echo time 3 msec, flip angle 20 degree, readout bandwidth 90 KHz, slice thickness 6 mm, six views/segment, a rectangular field of view (75% in phase encoding direction), scanning matrix of 256 x 128. Heart rate (beats per minute) was assessed during each flow measurement.
CMR image analysis
CMR studies were analyzed on a commercially available Advanced Windows workstation (GE Healthcare). Volumetric data were quantitatively analyzed using the Mass analysis software package (version 3.1; Medis Medical Imaging Systems, Leiden, The Netherlands). End-diastolic and end-systolic time frames from every slice were used for the assessment of biventricular end-diastolic and end-systolic volumes. Stroke volumes, ejection fractions, and cardiac mass were assessed according to common analysis techniques (10).
Flow images were quantitatively analyzed using the Flow analysis software package (version 2.0; Medis Medical Imaging Systems). This package allows calculation of flow velocity curves by multiplying valve area (drawn on each of the 30 time frames) and spatial average flow velocities per time frame. From the resulting flow-velocity curves the following indices of biventricular diastolic filling were derived (Fig. 1
): 1) peak early filling rate (PeFR) (milliliters per second); 2) time to PeFR (milliseconds) measured from end systole; 3) early filling fraction (FF) defined as volume increase (milliliters) during the first one third of diastole normalized for ventricular stroke volume; 4) deceleration time (milliseconds), measuring the time from PeFR to the extrapolation point of deceleration of flow to the baseline; 5) peak atrial filling rate (PaFR) (milliliters per second); 6) atrial filling fraction (AFF) defined as the increase in ventricular volume after atrial contraction normalized for ventricular stroke volume; and 7) the ratio of peak early filling rate over peak atrial filling rate (PeFR to PaFR ratio).
|
SPSS-PC statistical software package was used for analysis (version 11.5; SPSS, Inc., Chicago, IL). P < 0.05 was considered to indicate statistical significance. Results are expressed as mean (SD). Characteristics were compared between patients and controls using the two-sample t test or Mann-Whitney U test. Bivariate correlation coefficients were calculated using Spearmans method (rs).
Analysis of covariance with correction for BSA (calculated using weight and height) or BMI was used to compare biventricular short axis data between patients and controls. Analysis of covariance with correction for heart rate was used to compare biventricular indices of diastolic filling between patients and controls. Ventricular volumes were log10 transformed before analysis of (co)variance was applied.
| Results |
|---|
|
|
|---|
|
Significant positive correlations (Spearmans method) were found for GH dose with adult height (rs= 0.47, P = 0.01) and height SD score (rs = 0.47, P = 0.01). Negative associations of GH dose were found with systolic blood pressure (rs = –0.36, P = 0.04), diastolic blood pressure (rs = –0.38, P = 0.04), and mean blood pressure (rs = –0.40, P = 0.03).
Biventricular size and ejection fractions
The biventricular CMR measures normalized for BSA of patients and controls are shown in Table 2
. Compared with controls, LV and right ventricular (RV) end-diastolic volume (RVEDV, P = 0.02, and LVEDV, P < 0.001), end-systolic volume (RVESV, P = 0.04, and LVESV, P < 0.001), and stroke volume (RVSV, P = 0.03, and LVSV, P = 0.02) were significantly smaller in patients. Biventricular ejection fraction (EF), wall mass, and cardiac index were comparable with controls, although LVEF showed a tendency to be higher in patients (P = 0.05).
|
At analysis of covariance with correction for BMI, comparable differences were found between patients and controls (data not shown). There were no differences between GH dose subgroups, both with adjustment for BSA or BMI.
Atrioventricular valve inflow
Results of the quantitative analysis of atrioventricular inflow are shown in Table 3
. For logistic reasons, valve measurements could not be performed in all patients. Mitral valve inflow in patients showed a smaller early peak filling rate (P = 0.03), a smaller early FF (P = 0.04), a smaller ratio of peak early filling rate over peak atrial filling rate (P = 0.04), and a larger atrial filling fraction (P = 0.06). After correction for heart rate, significant differences were no longer present.
|
No significant differences in tricuspid or mitral valve inflow indices were present between GH dose subgroups.
| Discussion |
|---|
|
|
|---|
Cardiac morphology
Our finding of absence of myocardial hypertrophy is in line with previous findings from studies conducted during GH treatment in children with TS (7, 8), Noonan syndrome (12, 13), GH deficiency (14), and idiopathic short stature (15, 16). However, in contrast with previous papers that reported comparable LV volumes with healthy controls in TS (7, 8), our patients had smaller biventricular volumes and stroke volumes. A detailed look at the results presented by Sas et al. (7) shows they found normal SD scores for LV dimensions at the start of GH treatment that significantly decreased to values less than 0 after 7 yr of treatment. In line with these data, Noordam et al. (12) also found negative SD score changes for LV dimensions during GH treatment in patients with Noonan syndrome.
This may suggest cardiac growth in TS (and other) patients treated with GH during childhood, is disproportionate to somatic growth. However, we do not know of clinical studies in TS that specifically focused on the effects of GH on growth of different organ systems to (dis)prove this hypothesis. Experimental models of GH and IGF-I deficiency reported differential growth effects on different organs, which, however, do not explain our observations (17, 18). A recent echocardiographic study in TS showed no evidence for disproportionate cardiac growth between patients treated with GH and those not treated with GH (11).
Previously described reversibility of GH-induced ventricular remodeling (4, 19) may be an alternative explanation for the apparent differences between our and previous results. After initial myocardial growth in response to short-term GH treatment, Amato et al. (20) found cardiac dimensions and LV mass decreased after discontinuation of GH treatment in GH-deficient adults. Accordingly, ventricular volumes in TS measured during GH treatment may have appeared relatively normal, compared with controls (7, 8). However, after discontinuation of GH treatment, differences may have become more apparent (our patients), assuming discontinuation of treatment led to involution of ventricular size. Unfortunately, we do not have repetitive MRI measurements to support this theory.
Finally, fetal studies suggested that cardiac hypoplasia is present in Turner patients (21). Matura et al. (11) found no difference in cardiac dimensions corrected for BSA between TS patients treated with GH and patients not treated. Their results may indicate GH does not affect ventricular size relative to body size. We, however, compared our patients with healthy controls. The differences we found therefore may be interpreted to reflect the above-mentioned cardiac hypoplasia in TS rather than a GH effect. This is supported by the absence of differences between our GH dose groups.
Cardiac function
Initially GH excess is associated with normal to increased ventricular function (4, 6, 8, 22). Patients with acromegaly, on the other hand, develop LV diastolic dysfunction over time (23, 24). Moreover, in acromegaly diastolic dysfunction relates to disease activity and can be detected before evident systolic dysfunction or hypertrophy (4, 23).
Compared with controls, our TS patients generally showed normal RV diastolic filling, with some shifting toward late diastolic filling for the LV. Previously Radetti et al. (8) reported comparable findings in TS during GH treatment. Diastolic filling is known to be heart rate dependent (25, 26), and correction for heart rate (which was significantly higher in our patients) resulted in comparable biventricular diastolic filling pattern between our patients and controls. As Radetti et al. previously suggested, the changes seen in LV diastolic filling in TS probably should be regarded as an adaptation to the higher heart rates seen in TS, rather than an early representation of LV dysfunction (8). Nevertheless, ventricular function in TS may become compromised over time by factors other than GH treatment. This was suggested by Andersen et al. (27), who found abnormal relaxation and pseudonormal filling during diastole in an older population. They also found signs of systolic dysfunction in close relation to the appearance of diastolic dysfunction. Overall, our results lead us to conclude that young adults with TS formerly treated with GH, even in high doses, show normal biventricular function.
Heart rate
As was briefly mentioned in the previous section, we found TS patients had a higher mean heart rate, compared with our healthy controls. Increased heart rates have repeatedly been found in TS both prenatally and during postnatal life (28, 29, 30). Recent studies described dysregulation of the sympathetic nervous system, leading to tachycardia, high blood pressure, increased resting norepinephrine levels, and a blunted cathecholinergic response to exercise in TS (30, 31). There is also evidence that resting heart rate negatively correlates with LV chamber size, a correlation (although weak) that we also found within our population (32). As such the smaller ventricular volumes in TS described here, although of no immediate clinical relevance, may be part of the etiology of the increased resting heart rate in TS. Evidence is increasing as to the detrimental effect of higher resting heart rates on long-term (cardiovascular) mortality (33, 34).
In Turner syndrome an increased risk of aortic dissection and related mortality is generally recognized. A higher mean heart rate may contribute to faster progression of aortic dissection. Further research on the etiology of the high resting heart rate repeatedly found in TS, its role in cardiac mortality, and possible therapeutic interventions appears warranted.
Limitations of this study
The absence of an untreated group of Turner patients for comparison is a drawback of our study. However, currently TS is an accepted indication for GH treatment in our country, and it was felt unethical to withhold treatment from patients. Nevertheless, the absence of long-term prognostic cardiac data may be considered the driving force behind the present study. To the best of our knowledge, this study is the first to determine cardiac status in an adult population, in which treatment was discontinued in all patients. Our population represents cardiac status up to 9 yr after discontinuation of GH treatment without interference of data from patients still treated with GH.
Longitudinal MRI data would have been valuable and might have given the answers to some of the remaining questions mentioned earlier. However, cardiac MRI has been available in our institution only during the last 5 yr. We chose to apply MRI because it is the current gold standard for biventricular volumetric quantification. We think our current MRI data supplement the echocardiographic data previous published by Sas et al. (7) on LV mass and volume during GH treatment with results pointing in the same direction.
Finally, several factors have been suggested as appropriate for correction of body size when looking at heart dimensions. We choose to apply a correction for BSA. However, a correction for BMI might also have been appropriate considering the difference in BMI between our patients and controls. Afterward we did statistically correct for BMI, which did not change our overall conclusions.
Final considerations
At least 6 months after discontinuation of high-dosed GH treatment, this study shows TS patients have no myocardial hypertrophy or ventricular dysfunction, both well-known cardiac complications of long-term GH excess. Ventricular volumes at adult age are smaller, compared with controls, with comparable cardiac output achieved at a higher resting heart rate. The impact of these last observations on long-term prognosis remains unclear. However, they may be part of the natural development in TS and as such not linked to the use of GH, which at this point we would consider safe.
| Footnotes |
|---|
First Published Online April 22, 2008
Abbreviations: AFF, Atrial filling fraction; BMI, body mass index; BSA, body surface area; CMR, cardiac magnetic resonance imaging; EF, ejection fraction; FF, filling fraction; LV, left ventricular; LVEDV, LV end-diastolic volume; LVESV, LV end-systolic volume; LVSV, LV stroke volume; MRI, magnetic resonance imaging; PaFR, peak atrial filling rate; PeFR, peak early filling rate; rs, Spearmans method; RV, right ventricular; RVEDV, RV end-diastolic volume; RVESV, RV end-systolic volume; RVSV, RV stroke volume; TS, Turner syndrome.
Received October 16, 2007.
Accepted April 10, 2008.
| References |
|---|
|
|
|---|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |