Characterization of Left Ventricular Diastolic Function by Tissue Doppler Imaging and Clinical Status in Children With Hypertrophic Cardiomyopathy

Study End Points

The primary end point of the study was defined as patients who died, experienced a cardiac arrest, or had documented VT, either sustained or nonsustained. Sustained VT was defined as lasting longer than 30 seconds. The secondary end point was defined as patients who developed significant cardiac symptoms, including chest pain on exertion, syncope, dyspnea, or exercise intolerance.

Patient Analysis

Demographic data that included age at diagnosis, gender, presence of a positive family history, and medical therapy were collected. Patients were questioned about whether they had experienced any of the defined symptoms. New York Heart Association (NYHA) classification was recorded for each patient at the start and end of the study. Twenty-four-hour Holter monitors were reviewed to determine the incidence of significant arrhythmia, defined as VT. ECGs were reviewed and abnormalities documented; QRS and QTc duration were measured. Dyssynchronous ventricular activation was defined on the ECG by the presence of intraventricular conduction delay or left bundle-branch block.¹⁸ Exercise capacity was determined with a ramp protocol (Bruce or modified Bruce protocol). Maximum oxygen consumption (V̇o₂) was defined as the mean of the highest values obtained during the last 15 seconds of the protocol. Holter monitors, ECG, and treadmill tests were performed within 1 month of echocardiograms.

Echocardiographic Analysis

A single blinded observer performed the echocardiographic analysis (C.J.M.). Patients were defined as having asymmetric (predominantly septal hypertrophy), concentric (involving the entire LV), or posterior (isolated to the posterior LV wall) LV hypertrophy. 2D measurements included LV end-diastolic and end-systolic dimensions, posterior wall thickness, interventricular septal thickness, and LV ejection fraction.¹⁷ In the parasternal short-axis view, the LV was divided into 4 quadrants (anterior septum, posterior septum, lateral, and posterior wall segments). The greatest LV wall thickness was measured in each of these quadrants in diastole and systole at the level of the mitral valve papillary muscles.⁶ Maximal LV wall thickness was defined as the greatest thickness in any of these segments. The peak LV outflow tract gradient was determined from the modified Bernoulli equation. Pulsed Doppler was used to record mitral and tricuspid inflow patterns at the leaflet tips in the apical 4-chamber view.¹⁷ TD imaging was obtained from an apical 4-chamber view to obtain longitudinal annular velocities at the lateral mitral wall, septum, and lateral tricuspid wall adjacent to the AV valve hinge points.¹⁶ Filters and gains were adjusted to allow a clear tissue signal and minimize background noise.

LA volumes and LV mass were determined by the truncated ellipsoid method and M-mode, respectively, for each patient and were indexed to body surface area.^19,20 Mitral inflow Doppler was measured in standard fashion to determine peak E- and A-wave velocities, deceleration time of the transmitral E wave, and isovolumic contraction and relaxation times.¹³ Pulmonary venous inflow Doppler patterns were analyzed to determine the peak velocities of the systolic, diastolic, and atrial reversal waves.¹³ The Tei index, defined as the sum of isovolumic contraction and relaxation times divided by LV ejection time, was calculated as reported previously.^20a Systolic (Sa), early diastolic (Ea), and late diastolic (Aa) TD velocities were measured at the lateral mitral, septal, and lateral tricuspid walls and subsequently averaged over 3 cardiac cycles in accordance with previous reports (Figure 1).^12–14 Transmitral E/Ea ratios (lateral and septal) were calculated for each patient.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. TD profile demonstrating significantly reduced mitral septal Ea velocities in (A) child with HCM and VT and (B) asymptomatic child with HCM.

Statistical Analysis

Statistical analysis was performed with SSPS Sigma Stat statistical software (SPSS for Windows version 3.0, SPSS Inc). Data are expressed as mean±SD or median (25th to 75th percentile) based on whether they have a normal distribution or not. Forward stepwise regression was performed to determine predictors of the primary and secondary end points. A probability value <0.05 was required for retention within the final stepwise regression model. Statistical significance was taken as P<0.05.

Patient Characteristics

Eighty patients with HCM (age range 1 to 18 years) were compared with 80 age- and gender-matched controls. The median age at diagnosis was 12 years (range 1 to 18 years), and median duration of patient follow-up was 26 months (range 2 to 51 months). The pattern of LV hypertrophy was asymmetric in 54 patients, concentric in 25, and isolated posterior wall hypertrophy in 1 patient. One additional patient had significant LV hypertrophy, which initially normalized and then recurred. Nineteen patients had LV outflow tract obstruction with a peak gradient >20 mm Hg (mean gradient 16 mm Hg, range 10 to 64 mm Hg). Eighteen patients had a positive family history of HCM (23%) ranging from 1 affected family member to 1 family with 20 affected members. In this affected family, 13 members died, 8 of whom died of SCD. Seventy-four patients were treated with β-blockers and 6 with calcium channel blockers.

Fourteen patients reached the primary end point during the study period. Eight patients had sustained VT and underwent placement of an implanted cardiac defibrillator. There were 3 patients who experienced aborted cardiac arrests and 1 patient with a resuscitated near-drowning incident. There was 1 cardiac-related death in a patient with significant LV outflow tract obstruction (peak gradient 37 mm Hg) who had no prior history of VT or cardiac arrest. Twenty-one patients met the secondary end point, of whom 15 patients were NYHA class II, 5 were class III, and 1 was class IV.

Echocardiographic Findings in HCM Patients Versus Controls

2D echocardiographic and TD variables comparing HCM patients with controls are presented in Table 1. As expected, HCM patients had smaller LVs, thicker interventricular septal thickness and LV posterior walls, and increased LV ejection fractions and Tei indices than controls. There were no significant differences in mitral inflow Doppler velocities between groups. There was a significant reduction in all early diastolic TD velocities and septal late diastolic velocity in HCM patients compared with controls.

Predictors of Primary End Point

Echocardiographic variables including TD velocities are shown in Table 2 for patients who met the primary end point. Univariate correlation coefficients for variables entered into the stepwise regression analysis are presented in Table 4. The only significant variable predictive of the primary end point by stepwise regression analysis was the septal E/Ea ratio (r=0.61, R²=0.37, P<0.001). Interestingly, although the lateral transmitral E/Ea ratio was also increased in patients who met the primary end point, this did not reach statistical significance (P=0.31).

Predictors of Secondary End Point

Echocardiographic and TD parameters of patients who developed the secondary end point are outlined in Table 3. By univariate analysis, septal E/Ea and maximum V̇o₂ predicted development of significant cardiac symptoms (Table 4). There was strong inverse correlation between maximum V̇o₂ and septal E/Ea (y=54−2x, r=−0.74, P<0.01). Given the strong correlation between septal E/Ea and maximum V̇o₂, when both were entered into the forward stepwise regression model, only maximum V̇o₂ was predictive of patients experiencing symptoms (r=0.427, R²=0.182, P<0.001). Interestingly, by univariate analysis, lateral transmitral E/Ea ratio did not correlate with patients who experienced symptoms. ECG data demonstrated prolonged QRS and QTc duration among symptomatic patients. In addition, a significant number of symptomatic patients had ECG evidence of dyssynchronous ventricular activation compared with asymptomatic children (Table 5).

Exercise Capacity

Upright exercise testing was performed in 38 patients >8 years of age, 16 (67%) of whom were symptomatic (Table 5). There were significant differences in exercise duration and maximum V̇o₂ between symptomatic and asymptomatic patients. Maximum V̇o₂ correlated inversely with septal E/Ea ratio (r=0.740, P<0.01; Figure 2). By linear regression analysis, there was a significant inverse relationship between NYHA class and V̇o_2max (r=0.669, P<0.001) and between NYHA class and septal E/Ea ratio (r=0.576, P<0.001). Of the symptomatic patients who underwent exercise testing, 81% had a blunted heart rate response compared with 18% of patients who were symptom free; blunted blood pressure response occurred in 50% and 10% of symptomatic and asymptomatic patients, respectively (P<0.01). Correlation coefficients for NYHA class and exercise variables in symptomatic patients are presented in Table 6.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2. Graph correlating maximum V̇o₂ and transmitral E/Ea septal ratio. There is a strong inverse relationship between maximal V̇o₂ and E/septal Ea ratio.

Mitral Inflow Patterns and TD

Among adults with HCM, routine transmitral Doppler inflow velocities have been shown to correlate poorly with symptoms, exercise capacity, mean LA pressure, and LV hypertrophy.^9,10,25 Previous adult studies reported HCM patients have lower mitral E-wave velocities, higher A-wave velocities, prolonged E-wave deceleration time, and a mitral E/A ratio <1; however, there is significant overlap, and many patients have either a normal or pseudonormalized pattern.^26–28 We have shown that in children with HCM, transmitral E- and A-wave velocities failed to distinguish between patients who reached the clinical end points of this study. The mitral E-wave deceleration time was significantly prolonged in patients with HCM compared with normal controls. Significant variation in mitral inflow velocities with altered loading conditions makes this an unreliable echocardiographic predictor of diastolic dysfunction in children with HCM. The failure of conventional mitral inflow Doppler velocities to predict clinical outcomes in children with HCM led us to evaluate other potentially prognostic echocardiographic variables.

This is the first study to demonstrate that children with HCM have lower early diastolic TD velocities than age- and gender-matched controls. Matsumura et al²⁵ reported similar findings in an adult population with HCM. More importantly, by forward stepwise regression, we were able to identify that the septal E/Ea velocity ratio predicted children at risk of death, VT, and cardiac arrest. This finding has not been reported previously in adult or pediatric populations. The majority of patients in the present study had asymmetric septal hypertrophy, and the finding that the septal E/Ea ratio may predict children who will meet the primary end point may not be surprising. Disorganized myocardial fiber arrangement is likely a major determinant of impaired LV relaxation in HCM patients. A significant increase in the septal E/Ea ratio is most likely secondary to both anatomic and hemodynamic abnormalities in these children. Increasing myocardial cell hypertrophy and fiber disarray may also affect electrical conduction in the ventricular septum, predisposing these patients to an increased propensity for ventricular dysrhythmias and SCD. In adult studies, increased LV end-diastolic pressure is consistent with an increased septal E/Ea ratio, and this may play a significant role in the development of symptoms, worsening NYHA class, and decreased exercise performance in these children. Peak V̇o₂ predicted children at risk of developing significant cardiac symptoms. Although by univariate analysis, several factors predicted development of symptoms, only the peak V̇o₂ was predictive by stepwise progression.

2D Echocardiographic LV Measures

The present study demonstrated that children who met the primary and secondary end points were more likely to have increased thickness of the LV posterior wall and septum, increased maximal LV wall thickness, and larger LA volumes. However, none of these parameters, including maximal LV thickness, were predictive of death, VT, or cardiac arrest by stepwise forward regression. This would corroborate data from the previous adult studies that proved that patients with the most ventricular and septal hypertrophy were not necessarily those most at risk of SCD or other significant cardiac events.⁶

Exercise Capacity

Matsumura et al²⁵ demonstrated that an increased lateral E/Ea ratio was significantly associated with maximum V̇o₂ in adults. The present study demonstrated strong correlation between septal E/Ea ratio and maximum V̇o₂ in pediatric HCM patients. We also showed that NYHA class correlated significantly with maximum V̇o₂, irrespective of presence of LV outflow tract obstruction, hence validating the use of NYHA class among patients in the present study. Hemodynamic mechanisms for impaired exercise tolerance in children with HCM remain poorly defined. Postulated mechanisms in adults include elevated LA pressure, impaired augmentation of stroke volume secondary to abnormal diastolic filling, or impairment in LA and LV systolic performance.^29,30 From the present study, a significantly shorter Doppler mitral A-wave to pulmonary reverse A-wave duration difference in symptomatic patients would suggest elevated LA pressure to be a similar hemodynamic mechanism in children. Children who were symptomatic also had significantly increased QRS duration and a significant propensity to dyssynchronous ventricular activation by ECG criteria. Similarly, abnormal ventricular depolarization-repolarization with TD has been reported recently in patients with repaired tetralogy of Fallot as a potential risk factor for VT and SCD, although this was not correlated with symptoms.³¹ Recent studies have proposed such abnormalities in electrical activation of the ventricle may predispose to diastolic dysfunction in adults with HCM. It remains debatable whether ventricular pacing improves diastolic relaxation and hemodynamics in such cases.^32,33

Study Limitations

Correlation of hemodynamic catheterization data and TD velocities may have provided additional data corroborating our hypothesis that symptomatic patients have higher LA and LV end-diastolic pressures. Determination of symptoms in younger children, a large portion of the present cohort, may be less precise, and exercise testing is limited in younger children.