Analysis of Echocardiograms in a Large Heterogeneous Cohort of Patients With Friedreich Ataxia




Although Friedreich ataxia (FA) is associated with cardiomyopathy, the severity and evolution of cardiac disease is poorly understood. To identify factors predicting cardiomyopathy in FA, we assessed echocardiograms from a large heterogenous cohort and their relation to disease traits. The most recent echocardiograms from 173 subjects with FA were analyzed in a core laboratory to determine their relation to disease duration, subject age, age of onset, functional disability score, and GAA repeat length. Mean age of the cohort was 19.7 years, mean age of disease onset was 10.6 years, and mean shorter GAA length was 681 repeats. Echocardiograms collectively illustrated systolic dysfunction, diastolic dysfunction, and hypertrophy. Measurements of hypertrophy correlated moderately with each other (r = 0.39 to 0.79) but not with measurements of diastolic dysfunction (r <0.35). Diastolic measurements correlated poorly with each other, although 26% of the cohort had multiple diastolic abnormalities. The most common diastolic dysfunction classification was pseudonormalization. Classification of diastolic dysfunction was predicted by GAA repeat length but not by age or gender. Ejection fraction was below normal in 20% of the cohort. In linear regression analysis, increasing age predicted decreasing ejection fraction. Functional disability score, a measurement of neurologic ability, did not predict any echocardiographic measurements. In conclusion, hypertrophy and diastolic and systolic dysfunctions occur in FA and are substantially independent; diastolic dysfunction is the most common abnormality with most patients having an assigned diastolic dysfunction class of pseudonormalization.


Friedreich ataxia (FA) is an autosomal recessive neurodegenerative disorder associated with cardiomyopathy and other system features. Most hypotheses on cardiac progression in FA have been based on a model in which hypertrophy occurs early in the disorder with subsequent conversion to fibrosis. The small size of most patient cohorts, variability of assessment, and inconsistent follow-up have limited the understanding of the direct relation between wall hypertrophy and development of systolic dysfunction. Studies have simplified interpretation of data by stratifying cohorts based on subject age or age of onset, thus removing analysis of the temporal course. Such approaches may also remove the understanding of the degree to which cardiomyopathy in FA is a continuum rather than a categorical process. We analyzed echocardiograms from a large diverse cohort at a central core laboratory to identify the diversity of cardiac features of FA and to integrate these findings into a model portraying cardiac severity related to genetic severity and age.


Methods


The study protocol was approved by institutional review boards of all participating institutions. The most recent clinically indicated echocardiogram was acquired on VHS or compact disk for 173 patients with clinically confirmed FA. Patients were recruited through a natural history study in FA, clinic populations of the investigators, and the Friedreich Ataxia Parents Group. A subgroup of studies represented the baseline evaluation (n = 70) from a therapeutic trial of idebenone. Functional disability scores (an index of neurologic dysfunction) were assigned from a pre-existing natural history database (72%) or using a review of medical records when no natural history visit was available. Ninety-six percent had genetic confirmation with GAA repeat length available.


Echocardiograms were quantified without knowledge of patients’ demographics at the echocardiographic core laboratory at the University of Pennsylvania. Digital echocardiographic images and Doppler velocity signals were analyzed using commercially available software (Tomtec Imaging Systems, Munich, Germany). Standard echocardiographic parameters were measured: left ventricular (LV) end-systolic dimension, LV end-diastolic dimension, shortening fraction, ejection fraction, interventricular septal thickness in diastole, posterior wall thickness in diastole, relative wall thickness in diastole (RWTd), peak blood flow velocities during rapid filling (E wave) and atrial systolic contraction (A wave), transmitral E/A ratio (E/A), tissue Doppler index E′, tissue Doppler index A′, deceleration time, and isovolumic, relaxation time, LV mass measured in M-mode, and wall motion score. For most analyses LV mass was indexed by meters 2.7 to decrease the effects of age and size ; wall thicknesses and LV cavity diameters (interventricular septal thickness in diastole, posterior wall thickness in diastole, LV end-diastolic dimension, LV end-systolic dimension) were indexed by body surface area. Normalized diastolic function values for age were available from the American Society of Echocardiography for >15 years of age. For children ≤15 years old, normal ranges for diastolic parameters were defined from previous studies.


Results were also classified by LV geometry as normal (RWTd <0.42, normal indexed LV mass), concentric remodeling (RWTd >0.42, normal indexed LV mass), concentric hypertrophy (RWTd >0.42, increased indexed LV mass), or eccentric hypertrophy (RWTd <0.42, increased indexed LV mass). To assess progression of diastolic function, subjects were assigned to 1 of 4 classes: normal, impaired relaxation (prolonged isovolumic relaxation time; low E/A), pseudonormal (E/A above lower limit of normal; prolonged isovolumic relaxation time), or restrictive filling (shortened isovolumic relaxation time; E/A above lower limit of normal).


Statistical analysis was performed using STATA SE 11 (STATA Corp., College Station, Texas) including creation of linear regression models between echocardiographic parameters and GAA repeat length, gender, and age. Functional disability score was also included in some models. Analysis of variance was used to compare GAA repeat length, gender, and age between abnormal and normal populations for selected echocardiographic parameters and to examine differences in classifications of hypertrophy and diastolic dysfunction. Pairwise correlations were used to assess relations among individual measurements. The Centers for Disease Control body mass index calculator and growth charts were used to convert height and body mass index measurements to percentile ranks based on gender and age.




Results


Mean age of the cohort was 19.7 ± 11.6 years, disease duration was 8.8 ± 7.7 years, and average age of onset was 10.6 ± 7.9 years. Length of the shorter GAA allele, available in 96% of subjects, was 681 ± 189 GAA repeats. Mean functional disability score was 3.3 ± 1.5, equivalent to a patient walking with moderate assistance. The cohort was predominantly women (62%) with 66% of subjects <18 years of age. Seven percent of subjects had diabetes and 76% had scoliosis.


Specific findings were grouped into measurements of systolic function (shortening fraction, ejection fraction, wall motion score), hypertrophy/mass (indexed LV mass, RWTd, interventricular septal thickness in diastole indexed by body surface area, posterior wall thickness in diastole indexed by body surface area), and diastolic function (E/A, E′/A′, E/E′, isovolumic relaxation time). Overall, data for most measurements were normally distributed (skewness <1) such that parametric statistics and linear regression approaches could be used.


Systolic function varied widely in the cohort and measurements of systolic function correlated modestly to moderately with each other ( Table 1 ; Supplemental Table 2 ). Ejection fraction was <50% in 30 patients but <40% in only 6 patients. Shortening fraction showed less variation with only 9.3% decreasing below normal ( Table 2 ). Overall relations to disease-related features were of modest significance but R 2 values were generally low. Age was the strongest predictor of decreased systolic function in linear regression models accounting for gender and GAA repeat length ( Table 3 ). Measurements of systolic function (ejection fraction and shortening fraction) were not predicted by GAA length in linear regression analysis modeled with the entire cohort. Wall motion score was significantly predicted only by age (p <0.01) and correlated poorly with measurements of hypertrophy and diastolic function ( Table 2 ; Supplemental Table 2 ). Hypertrophy measurements correlated moderately with each other and were generally increased ( Table 2 ). Classifications of LV geometry (based on RWTd and indexed LV mass) were abnormal in 82%, with 42% having concentric remodeling, 35% having concentric hypertrophy, and only 5% having eccentric hypertrophy. However, these classifications showed no significant relation to GAA repeat, gender, or age in linear regression analysis or by analysis of variance (data not shown). Similarly, when assessed by linear regression, few relations of measurements of hypertrophy associated with disease features and R 2 values were low. RWTd was increased in 77% of the population but showed no relation to disease features ( Tables 1 and 2 ). Gender predicted higher indexed LV mass in linear regression models.



Table 1

Correlations of examined parameters






































































































Gender Age AOO FDS GAA
Shortening fraction −0.02 −0.29 −0.13 −0.30 0.06
Ejection fraction 0.05 −0.33 −0.17 −0.23 0.05
Wall motion score 0.04 0.30 0.12 0.21 −0.05
Isovolumic relaxation time −0.23 0.00 −0.10 0.08 0.17
Transmitral E/A ratio −0.10 −0.24 −0.29 −0.01 0.15
E′/A′ 0.01 −0.22 −0.22 −0.13 0.17
E/E′ −0.04 −0.08 −0.14 0.13 0.11
Intraventricular septal thickness index −0.13 −0.29 −0.22 −0.11 0.25
Posterior wall thickness index −0.13 −0.48 −0.34 −0.25 0.18
Left ventricular internal diameter index in diastole 0.03 −0.36 −0.36 −0.42 −0.02
Left ventricular internal diameter index in systole 0.03 −0.10 −0.06 −0.16 −0.06
Relative wall thickness in diastole −0.13 −0.19 −0.21 0.08 0.18
Left ventricular mass index −0.27 −0.07 −0.14 0.05 0.16

Pearson correlation values of echocardiographic parameters with demographics and disease features. Left ventricular mass was indexed by height 2.7 and wall thicknesses were indexed by body surface area.

AOO = age of disease onset; FDS = functional disability score.


Table 2

Summary values of echocardiographic parameters






























































































Mean ± SD High (%) Low (%)
Systolic function
Left ventricular internal diameter in diastole (cm) 4.01 ± 0.58 — —
Left ventricular internal diameter in systole (cm) 2.71 ± 0.65 — —
Stroke volume (cm 3 ) 42.9 ± 20.2 — —
Shortening fraction (%) 32.9 ± 8.66 9.30 9.30
Ejection fraction (%) 54.5 ± 8.63 1.4 20.4
Hypertrophy
Posterior wall thickness index in diastole (cm/m 2 ) 0.70 ± 0.18 62.9
Intraventricular septal thickness index in diastole (cm/m 2 ) 0.76 ± 0.24 52.7
Relative wall thickness in diastole 0.52 ± 0.13 77.9
Left ventricular outflow tract diameter in diastole (cm) 1.96 ± 0.23 —
Left ventricular mass index (g/m 2.7 ) 48.0 ± 17.8 40.2
Diastolic function
Transmitral E/A ratio 1.79 ± 0.59 8 4
E′/A′ 2.02 ± 0.71 5 1
E/E′ 6.97 ± 2.42 24 —
Isovolumic relaxation time 91.4 ± 21.1 85 1

Systolic function measurements were decreased in a significant number of subjects (normal ranges, shortening fraction 20% to 44%, ejection fraction 50% to 70%). Hypertrophy measurements were increased with a large percentage of the cohort above the normal range (normal ranges, posterior wall indexed by body surface area <0.62 cm/m 2 , interventricular septal thickness indexed by body surface area <0.69 cm/m 2 , relative wall thickness in diastole 0.42, indexed left ventricular mass <51 g/m 2.7 in men and 48 g/m 2.7 in women. Diastolic function was also abnormal in a large percentage of the cohort by the isovolumic relaxation time criterion, but E/A ranges remained in the normal range (normal range, E/E′ <8; see Supplemental Table 1 for others).


Table 3

Multivariate analysis of echocardiographic parameters























































































Systolic Hypertrophy Diastolic
EF SF WMS IVSTi PWTi RWTd LVMi E/A E′/A′ E/E′ IVRT
GAA 0.12 0.41 0.03 0.07 0.96 0.12 0.08 0.54 0.67 0.48 0.08
Gender 0.17 0.25 0.39 0.38 0.59 0.28 <0.01 0.38 0.96 0.66 0.02
Age <0.01 <0.01 <0.01 0.04 <0.01 0.30 0.55 <0.01 0.03 0.62 0.40
Overall <0.01 0.02 0.02 <0.01 <0.01 0.06 <0.01 0.01 0.06 0.65 0.03
R 2 0.14 0.09 0.07 0.13 0.08 0.05 0.10 0.08 0.08 0.02 0.09

Linear regression analysis of systolic, hypertrophic, and diastolic parameters, with hypertrophic measurements indexed by body surface area or height; p values for individual variables are shown in addition to R 2 and p values for the models. When accounting for multiple comparisons by setting the standard of significance at a p value <0.01, age predicted measurements of systolic function and scattered measurements of hypertrophy and diastolic function. Surprisingly, GAA repeat length did not clearly predict any parameters. R 2 values were relatively low, showing the unexplained variability in echocardiographic parameters.

Only gold members can continue reading. Log In or Register to continue

Related posts:

  1. Association of Frontal QRS-T Angle–Age Risk Score on Admission Electrocardiogram With Mortality in Patients Admitted With an Acute Coronary Syndrome
  2. Vitamin D Deficiency and Supplementation and Relation to Cardiovascular Health
  3. Sudden Death Associated With Danon Disease in Women
  4. Comparison of Patients With Pulmonary Arterial Hypertension With Versus Without Right-Sided Mechanical Alternans

Stay updated, free articles. Join our Telegram channel
Tags:
Dec 15, 2016 | Posted by in CARDIOLOGY | Comments Off on Analysis of Echocardiograms in a Large Heterogeneous Cohort of Patients With Friedreich Ataxia

Full access? Get Clinical Tree
Get Clinical Tree app for offline access