Left Atrial structure and function in hypertrophic cardiomyopathy sarcomere mutation carriers with and without left ventricular hypertrophy

Hypertrophic cardiomyopathy (HCM) is caused by mutations in genes encoding sarcomere proteins [1]. Impaired left ventricular (LV) relaxation, altered myocardial energetics, and increased myocardial extracellular volume (ECV) have been demonstrated in HCM subjects without clinical disease, indicating that sarcomere mutations cause cardiac abnormalities prior to the establishment of LV hypertrophy (LVH) [2‐5]. While the clinical diagnosis of overt HCM is defined by exceeding a threshold of LV wall thickness, typically greater than 1.3 to 1.5 cm, genetic testing allows for the early identification of at-risk family members when LV wall thickness is still within the normal range (LVH−) [6]. This unique scenario allows for the study of early-stage disease in sarcomere mutation or genotype-positive/LVH-negative (G+/LVH−) individuals, herein referred to as “preclinical HCM”.

Historically, LV morphology and function have been extensively studied in HCM; however, evaluation of the left atrium (LA) has been largely underutilized [7, 8]. An increase in LA volume (LAV) has been established as a marker of adverse cardiac events among broad groups of patients with cardiovascular disease [9‐13]. Recently, data have shown that metrics of LA function can provide prognostic information beyond measurement of LA volume. For example, early detection of LA dysfunction may provide additional insight into the pathophysiology and aid in the clinical management of common cardiovascular diseases such as atrial fibrillation, myocardial ischemia, heart failure and cardiomyopathy [9, 10, 12].

LA function can be comprehensively assessed by measuring the three components that contribute to LV filling: reservoir, conduit, and pump phases. The reservoir phase starts at ventricular systole and encompasses LV isovolumetric contraction, ejection, and isovolumetric relaxation. During the reservoir phase, blood is received from the pulmonary veins and the LA fills, growing in volume as the mitral valve remains closed. Next, the LA acts as a conduit as the mitral valve opens and passive emptying occurs. LAV decreases during early ventricular diastole, driven by a suction effect on the LA as the LV relaxes and expands. Lastly, the atrial pump phase occurs during late ventricular diastole as the atrial muscle contracts and executes an active pump function, concluding ventricular filling. Technically, the LA reservoir, conduit, and active pump functions can be calculated from LAV at its maximum (end-systole, just prior to mitral valve opening), minimum (end-diastole, at mitral valve closure), and immediately before atrial contraction (before the electrocardiographic P-wave). From these three LAVs, total, passive, and active emptying functions can be calculated (LATEmF, LAPEmF and LAAEmF).

In this study, we hypothesized that LA function is abnormal in preclinical HCM mutation carriers, declines further in those with overt HCM, and correlates with other metrics of disease severity. To test these hypotheses, we examined the relationship between LA function, clinical parameters and established biomarkers for heart failure in a well-characterized genotyped cohort with preclinical and overt HCM and mutation-negative healthy family members.

Basic demographics of the study population are presented in Table 1. Mean age, gender, blood pressure and BMI did not differ between the control (N = 15) and preclinical groups (N = 24) but overt HCM subjects (N = 34) were 6 to 9 years older (p < 0.001). LV wall thickness did not differ between the control and the preclinical groups. LV wall thickness was greater in the overt HCM group as compared to preclinical and control cohorts (p < 0.001 for both). LVEF and LV mass was similar between controls and preclinical subjects but was higher in subjects with overt HCM (<0.01 for both). LV end diastolic volume and maximal LV outflow tract velocity did not differ between the 3 groups. Serum NT-proBNP levels were significantly higher in the overt HCM group compared with controls and preclinical HCM (p < 0.001 for both). LGE was only observed in the overt HCM group but not in the control and preclinical cohorts.

The analysis of LA size was performed by two reviewers. We found a mean bias of 1.6 ± 8.5 mls for the LAVmin, of 2.9 ± 6.1 mls for the LAVmax and of 1.6 ± 5.8 mls for the LAVbac between reviewers, with a coefficient of variation ranging from 0.11 to 0.19. Representative images are provided in Fig. 1. LAV indexed to BSA were numerically larger in overt HCM subjects, but did not differ significantly between the control, preclinical HCM and overt HCM subjects after adjustment for age, gender, BSA and family relationships (Table 2). In contrast, measures of LA function differed between the 3 groups, as summarized in Table 2. The most consistent relationship was observed for LATEmF, which was lower in both the preclinical HCM group (64 ± 8%) and the overt HCM group (59 ± 10%) as compared to normal controls (70 ± 7%; p = 0.002 and p = 0.005, respectively; Table and Fig. 1a). LATEmF did not differ significantly comparing overt with preclinical HCM (p = 0.2, Table 2). Additionally, LAPEmF was lower in the overt HCM group (35 ± 11%) as compared to controls (47 ± 10%, p = 0.006; Table 2 and Fig. 2b). In analyses further stratified by the extent of LGE, overall, LATEmF and LAPEmF showed an incremental decline in control, preclinical and overt HCM groups; smallest in individuals with more LGE enhancement (p < 0.001 for trend for both; Fig. 3a and b). This pattern was not seen for LAAEmF (p = 0.1, Fig. 3c). Furthermore, while defining cut-offs to differentiate control from overt HCM, and control from preclinical patients is beyond the scope of this study, measurement of the area under the curve (AUC) for LA function yield the following results: the overall AUC for LATEmF is 0.28 for preclinical and 0.23 for HCM. Using a cutoff of 58% for LATEmF, the AUC is 0.60 for preclinical and 0.60 for HCM. A LATEmF cutoff of 28% is 100% sensitive and 0% specific for HCM, correctly classifying 63% of subjects. A LAREmF cutoff of 47.5% is 100% sensitive and 0% specific for preclinical status, correctly classifying 56% of subjects.

Multivariate regression models were constructed to assess the relationship between LA function and measures of cardiac structure, function and serum biomarkers (Table 3). All 3 measures of LAV were positively correlated with LV mass, the extent of LGE, and serum NT-proBNP after adjustment for age, gender, BSA and family relationship (p ≤ 0.01 for all, Table 3). In analyzing LA function, both LATEmF and LAPEmF were inversely correlated with these metrics (p < 0.05 for all), as well as with interventricular septal thickness (p ≤ 0.001 for both). LV end-diastolic volumes and LV end-systolic volumes were correlated with LAVmax but not any measures of LA function. Tissue Doppler early diastolic myocardial relaxation velocity (global E’) was positively associated with LAPEmF (p = 0.05) and negatively associated with LAVmin, LAVbac (p < 0.05 for both). No measure of LAV or function correlated with LVEF or serum levels of high sensitivity troponin I, matrix metalloproteinase-1 (MMP1), or matrix metalloproteinase inhibitor-1 (TIMP1).

In this study, we leveraged a robustly characterized study population to investigate whether CMR-derived measures of LA size and function differed between healthy controls and sarcomere mutation carriers with preclinical and overt HCM. We identified the following major findings: 1. LAV did not significantly differ between mutation carriers with preclinical or overt disease or controls; 2. LA dysfunction, reflected by reduced LATEmF, is detectable in sarcomere mutation carriers with both preclinical and overt HCM compared with healthy controls; 3. Patients with overt HCM, particularly those with the greatest amount of LGE, demonstrated the most marked reduction of LA function, with decreased LATEmF and LAPEmF; 4. Impaired LA function and increased LV volumes were associated with higher serum NT-proBNP levels and a greater extent of LGE.

These data provide additional insights regarding early phenotypic manifestations in preclinical HCM sarcomere mutation carriers. Our work suggests that changes in LA function precede changes in LA size in the pathogenesis of HCM sarcomere mutations. In the preclinical cohort, all 3 measures of LA size were similar to those of control subjects. Yet when LA function was assessed, the LATEmF was lower in preclinical HCM, and reduced in overt HCM. Given our prior work demonstrating increased extracellular volume fraction, a quantitative measure of the extent of extracellular expansion and a marker of interstitial fibrosis, in preclinical HCM subjects [3], we speculate that the impaired LA function seen in this cohort reflects altered myocardial tissue properties. The reduction in total LA function was primarily reflected in the passive emptying component, while the active emptying function was less affected.

Another interesting finding of this study was that the overt HCM group had significantly lower LATEmF and LAPEmF compared to preclinical HCM and controls. Further, when the overt HCM group was stratified by the presence or absence and extent of LGE, most of this difference was due to overt HCM subjects with a greater extent of LGE. We speculate that the development of fibrotic changes in the ventricular myocardium may precipitate a decline in LA function, particularly LAPEmF. Although a primary atrial myopathy may also be present, these findings suggest that alterations in myocardial tissue composition may also drive the development of LA dysfunction in HCM.

While defining cut-offs to differentiate control from overt HCM, and control from preclinical patients is beyond the scope of this study (due to the limitations of sample size as well as the fact that we do not have data as to whether any of our preclinical progress to overt HCM), we provide AUC results. However, this study focuses instead on metrics of LA size and function to better characterize the pathophysiologic impact of sarcomere mutations and therefore help elucidate underlying disease biology.

Fully understanding the clinical implications of impaired LA function in either overt or preclinical HCM will require further longitudinal investigation. However, a reduction in LA function has been shown to have prognostic implications in broad populations [9, 10]. For example, among patients referred for a stress CMR, a reduction in LAPEmF was shown to be a significant predictor of adverse cardiac events [10]. Similarly, among patients with atrial fibrillation, a reduction in LAPEmF was shown to be a robust predictor for recurrence after pulmonary vein isolation [10]. In this study, LA total function was abnormal and passive function trended lower in the preclinical group. It remains to be studied whether reduced LA function will be associated with earlier progression to overt HCM, a greater susceptibility to atrial fibrillation, more prominent symptoms of heart failure, and/or poorer prognosis in these subjects. However, some prior data have shown the association between measures of LA size and function and adverse outcomes in HCM [7, 8]. For example, in a study of 242 HCM patients Debonnaire and colleagues showed than LAV and LA strain were predictive of new onset AF in patients with HCM [8].

The penetrance or clinical expression of a HCM gene mutation is age dependent and commonly occurs during young adulthood. When phenotypic conversion does occur, LV wall thickness can increase rapidly and the pathological manifestations of HCM may arise suddenly. In line with its natural history, subjects with overt HCM in this study were slightly older than preclinical and control groups. Much has yet to be understood of the preclinical stages of HCM. In subjects who are at risk, based on their family history and genetic makeup, it is currently unpredictable when and if the overt HCM will manifest. Developing non-invasive tools to potentially identify those subjects at greater risk for disease onset would be important for clinical management. These subjects could potentially be followed more closely with physician visits and echocardiographic assessment.

Our study has several limitations. We describe a small study group without follow up to evaluate clinical outcomes or longitudinal disease evolution. As an observational study, inferences about causality cannot be determined. Furthermore, we did not measure atrial LGE or measure the ECV to correlate with our finding of LA dysfunction in sarcomere mutation carriers. Lastly, 20% of the cohort did not have both 2 and 4-chamber views needed to quantify LA function. However, when we compared baseline characteristics between those subjects that underwent CMR and had quantifiable LA function versus those that were unable to be quantified, we found no significant differences between groups.

In conclusion, in a well-characterized and genotyped cohort of subjects with preclinical HCM, overt HCM and familial controls, we found that LA dysfunction is detectable in preclinical HCM subjects before the presence of LA enlargement or LVH. Furthermore, LA dysfunction is more pronounced as LVH and ventricular fibrosis are manifest. Impaired LA function appears to have physiologic relevance, being associated with a greater extent of LGE and higher NT-proBNP levels. These data support the hypothesis that HCM is a progressive disorder in which at-risk sarcomere mutation carriers begin to have physiological manifestations before clinical criteria for overt disease are met.