Williams syndrome (WS) is a congenital, developmental disorder affecting 1 in 8,000 live births. The corrected QT (QTc) interval is prolonged in 13% of patients with WS. No data exist characterizing the ambulatory electrocardiographic findings in WS. A retrospective review of all patients with WS evaluated at our institution from January 1, 1980 to December 31, 2007 was performed. Patients with ≥1 ambulatory electrocardiogram (AECG) with sinus rhythm and measurable intervals were included. QTc measurements were made at the minimum and maximum heart rate. Logistic regression analysis was used to evaluate the correlation of ventricular ectopic complexes with QTc measurements. A statistical probability of p <0.05 was considered significant. Of 270 patients identified, 32 had AECGs available for review. Complete data were available for 56 AECGs from 26 patients (15 female; 58%). Their mean age was 15.6 ± 7.2 years at the initial AECG and 20.6 ± 8.6 years for all AECGs. The QTc interval increased with increasing heart rate. Ventricular premature complexes occurred in 40 (73%) of 56 AECGs and 21 (81%) of 26 patients. Ventricular tachycardia occurred in 5 (9%) of 56 AECGs and 4 (15%) of 26 patients. The mean length of ventricular tachycardia was 3.6 ± 0.5 beats at a rate of 171 ± 40 beats/min. The QTc interval at the minimum heart rate correlated directly with age (p <0.001), total ventricular premature complexes (p = 0.007), ventricular couplets (p = 0.002), and ventricular tachycardia (p = 0.011). The QTc interval at the maximum heart rate correlated directly with age (p <0.001), total ventricular premature complexes (p = 0.016), and ventricular couplets (p = 0.006). In conclusion, the QTc interval correlated with ventricular ectopic complexes in patients with WS. The type of ventricular ectopic complexes suggested an alternate etiology of the QTc prolongation seen in WS from that seen in congenital long QT syndrome.
Williams syndrome (WS) is a congenital, multisystem developmental disorder occurring in 1 in 8,000 live births, resulting from deletion of approximately 28 genes on chromosome 7q11.23. Cardiovascular manifestations are the most common organ system abnormality, occurring in approximately 80% of patients with WS. We have reported corrected QT (QTc) interval prolongation in a series of patients with WS who had a normal QRS duration. Patients with prolonged QTc interval have an increased risk of ventricular tachycardia of the torsade de pointes type. In other patient groups, once QTc prolongation has been demonstrated on the electrocardiogram, exercise stress testing is recommended to estimate the change in QTc interval with exercise. The QTc interval usually shortens with an increasing heart rate in apparently normal patients but can be prolonged in pathologic settings, especially in the recovery period after exercise. However, the cognitive impairment that is almost universal in WS (mean full-scale intelligence scores in the 50 to 60 range) severely limits the patients’ ability to exercise to a meaningful intensity. To evaluate QTc variations with heart rate, we analyzed the QTc results at different heart rates from 24-hour ambulatory electrocardiograms (AECGs). To the best of our knowledge, no data are available characterizing the findings from AECGs in a series of patients with WS. The aims of the present study were to (1) characterize the findings from AECGs in patients with WS, and (2) determine whether the QTc interval on the AECG correlates with ventricular ectopic complexes (VE) in these patients.
Methods
Patients with WS were identified from a database of all such WS evaluated at the Children’s Hospital of Philadelphia. The available charts for patients with WS evaluated in either the Multidisciplinary Williams Syndrome Clinic or the Division of Cardiology from January 1, 1980 through December 31, 2007 were reviewed. The diagnosis of WS was confirmed by the clinical phenotype assessed by an experienced medical geneticist and/or by demonstrating elastin hemizygosity using fluorescence in situ hybridization.
In patients with multiple AECGs, all available AECGs with complete data were reviewed. All AECGs were reviewed independently by reader (R.T.C.) in a blinded fashion. Abnormalities in rhythm, intervals, and/or morphologies were recorded. The QTc interval at the minimum and maximum heart rate and at 100 ± 3 beats/min were calculated using Bazett’s method. The QTc interval at 100 ± 3 beats/min was included in analysis as a method of providing a QTc measurement at a similar heart rate for all subjects. Obstructive lesion severity categorization was assigned according to the patient’s most stenotic lesion severity history, including those patients who had undergone surgical relief of the obstruction (i.e., a patient with a history of severe supravalvar aortic stenosis who subsequently had undergone surgery and was without obstruction at the AECG was included in the severe category).
The available cardiovascular data were reviewed, including history, physical characteristics, and ancillary testing. The surgical records were analyzed to determine the interventions performed. In subjects who died, the clinical charts and autopsy reports were reviewed, where applicable, to determine the possible cause of death. The hospital’s institutional review board approved the study.
The data were analyzed using Stata software, version 12 (StataCorp, College Station, Texas). Summary statistics describing the demographic and electrocardiographic parameters were summarized at the first AECG and for all available AECGs. The primary outcomes of interest included the presence or absence of ventricular couplets or ventricular tachycardia. The association between the outcomes and ventricular premature complexes (VPCs) or QTc interval was investigated using a generalized linear mixed model, an extension of logistic regression analysis for repeated measurement data. In these generalized linear mixed models, all available AECGs were analyzed to determine the relations between the outcomes and explanatory variables, accounting for the correlation of measurements within a single study subject. The limited sample size and number of events dictated restriction to only 1 explanatory variable in any given regression model. Because of the wide range of values observed for total VPCs, a logarithmic (base 10) transformation was used to mitigate the leverage of extreme values and to stabilize variance. The association between the QTc measures and total VPCs was examined using a Poisson mixed model regression, modeling the ventricular complex counts and accounting for repeated measurements. The association between QTc interval and obstruction severity, accounting for repeated observations, was investigated using a linear mixed model. The correlation between QTc interval and age was also estimated, accounting for repeated observations. Statistical probability of p <0.05 was considered significant.
We all had full access to the data and take responsibility for its integrity and have read and agreed to the report as written.
Results
Of the 270 patients identified and previously reported, 32 had AECGs available. Complete data were available for 56 AECGs from 26 patients. The baseline characteristics are listed in Table 1 . Cardiovascular interventions were required in 10 (38%) of 26 patients. The interventions were performed for supravalvar aortic stenosis in 4, branch pulmonary artery stenosis in 4, valvar pulmonary stenosis in 1, and superior mesenteric artery stenosis in 1. Of the 16 patients without a history of intervention, 7 (44%) had no known obstructive cardiovascular lesion, 7 (44%) had only a single lesion of mild severity (i.e., isolated mild branch pulmonary artery stenosis), and 1 had moderate supravalvar aortic and pulmonary stenoses. The remaining patient had severe branch pulmonary stenosis and had been deemed early in childhood to be a poor candidate for intervention.
| Characteristic | Value |
|---|---|
| Patients (n) | 26 |
| Female patients (n) | 15 (58%) |
| Total ambulatory electrocardiograms (n) | 56 |
| Ambulatory electrocardiograms per patient | 2.6 ± 2.2 |
| Initial ambulatory electrocardiogram for surveillance | 19 (73%) |
| Age at initial ambulatory electrocardiogram (years) | 15.6 ± 7.2 |
| Age at all ambulatory electrocardiograms (years) | 20.6 ± 8.6 |
| Minimum heart rate (beats/min) | 55 ± 11 |
| Mean heart rate (beats/min) | 85 ± 13 |
| Maximum heart rate (beats/min) | 143 ± 25 |
| Sinus rhythm | 56 (98%) |
| PR interval (ms) | 145 ± 33 |
| QRS interval (ms) | 86 ± 12 |
| Mean QTc interval at minimum heart rate (ms) | 425 ± 34 |
| Mean QTc interval at 100 beats/min (ms) | 447 ± 35 |
| Mean QTc interval at maximum heart rate (ms) | 456 ± 34 |
As shown in Figure 1 , VE, including VPCs, couplets, and ventricular tachycardia, were common in the study cohort ( Figure 1 ). The median VPCs per 24 hours was 5 (range 0 to 19,462). The median age of the patients with ventricular tachycardia was 21.9 years (range 8.1 to 39.7) The mean length of ventricular tachycardia was 3.6 ± 0.5 beats at an average rate of 171 ± 40 beats/min. Ventricular tachycardia morphologies were polymorphic (non–torsade de pointes morphology) in 4 episodes and monomorphic in 1. In patients with ventricular tachycardia, the median QTc interval at the minimum heart rate was 488 ms (range 350 to 523) and the median QTc interval at the maximum heart rate was 489 ms (range 403 to 525). VPCs were not associated with the heart rate immediately before the VE. No correlation was found between VE and a history of cardiovascular intervention or lesion severity (if present).
The total number of VPCs correlated directly with the QTc interval at the minimum heart rate (p = 0.007; Figure 2 ) and the QTc interval at the maximum heart rate (p = 0.016; Figure 3 ). The increased risk of ventricular couplets and/or ventricular tachycardia with increasing QTc interval at the minimum and maximum heart rate is listed in Table 2 . No correlation was found between QTc interval at 100 ± 3 beats/min and VPCs, ventricular couplets, or ventricular tachycardia.
| QTc Interval (ms) | Ventricular Couplets | Ventricular Tachycardia | |||||
|---|---|---|---|---|---|---|---|
| OR | 95% CI | p Value | OR | 95% CI | p Value | ||
| QTc at minimum heart rate | 0.002 ⁎ | 0.011 ⁎ | |||||
| Minimum | 350 | 1.00 | Reference | 1.00 | Reference | ||
| 25th percentile | 405 | 14.53 | 2.64–80.00 | 6.14 | 1.53–24.67 | ||
| 50th percentile | 420 | 30.14 | 3.44–264.30 | 10.07 | 1.71–59.14 | ||
| 75th percentile | 440 | 77.83 | 4.85–1,249.75 | 19.15 | 1.99–184.27 | ||
| Maximum | 523 | 4,524.18 | 21.14–968,363.8 | 300.89 | 3.78–2,935.25 | ||
| QTc at maximum heart rate | 0.006 ⁎ | 0.180 | |||||
| Minimum | 388 | 1.00 | Reference | 1.00 | Reference | ||
| 25th percentile | 434 | 4.15 | 1.51–11.41 | 2.65 | 0.64–11.03 | ||
| 50th percentile | 453 | 7.52 | 1.79–31.52 | 3.99 | 0.53–30.07 | ||
| 75th percentile | 473 | 14.06 | 2.15–91.90 | 6.12 | 0.43–86.38 | ||
| Maximum | 557 | 197.67 | 4.63–8,445.57 | 37.50 | 0.19–7,461.66 | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
