Left Atrial Volume during Stress Is Associated with Increased Risk of Arrhythmias in Patients with Hypertrophic Cardiomyopathy.

INTRODUCTION: In patients affected by hypertrophic cardiomyopathy (HCM), left atrial volume index (LAVi) is associated with an increased risk of tachyarrhythmias and major clinical events. To date, the clinical meaning of LAVi measured during exercise (stress LAVi [sLAVi]) has not yet been investigated in HCM. This study sought to evaluate the correlation between LAVi/sLAVi and clinical outcome (risk of arrhythmias and heart failure [HF]) in patients with HCM. METHODS AND
RESULTS: We enrolled a total of 51 consecutive patients with HCM (39 men; mean age: 39.41 ± 17.9 years) who underwent standard and stress echocardiography, following a common protocol. During follow-up (median follow-up was 1.82 years), the following composite endpoints were collected: ARRHYT endpoint (atrial fibrillation, paroxysmal supraventricular tachycardia, nonsustained ventricular tachycardia (VT), sustained VT, ventricular fibrillation, syncope of likely cardiogenic nature, and sudden cardiac death) and HF endpoint (worsening of functional class and left ventricular ejection fraction, hospitalization, and death for end-stage HF). Eight patients were lost at follow-up. ARRHYT endpoint occurred in 13 (30.2%) patients (8, 18.6%, supraventricular and 10, 23.2%, ventricular arrhythmias), whereas HF endpoint occurred in 5 (11.6%) patients. sLAVi (mean value of 31.16 ± 10.15 mL/m2) performed better than rLAVi as a predictor of ARRHYT endpoint (Akaike Information Criterion: 48.37 vs. 50.37, if dichotomized according to the median values). A sLAVi value of 30 mL/m2 showed a predictive accuracy of 72.1% (C-statistics of 0.7346), with a high negative predictive value (87.5%).
CONCLUSION: These findings encourage future studies on sLAVi, as a potential predictor of arrhythmias and adverse outcome in patients with HCM.

INTRODUCTION

Hypertrophic cardiomyopathy (HCM) is an inherited heart muscle disorder defined by left ventricular hypertrophy in the absence of abnormal loading conditions.[1]

The presentation and clinical course of the disease are extremely variable: patients can be asymptomatic throughout life or develop complications as heart failure (HF), sudden cardiac death (SCD), and arrhythmias.[123] Thus, it is important to identify patients at risk of unfavorable outcomes at an early stage.[456]

Left ventricular outflow tract (LVOT) obstruction is present in about one-third of patients at rest, and another 40% develop LVOT obstruction during exercise.[7] Obstructive HCM (OHCM) patients have major risk of developing clinical symptoms, arrhythmias, and clinical events.[8]

Moreover, patients with OHCM present generally with enlarged left atrium, which is “per se” a predictor of arrhythmias and clinical events.[9] Although most published studies use anteroposterior left atrial (LA) diameter,[10] comparable findings using LA volume indexed to body surface area are reported.[1112]

LA volume, measured by two-dimensional echocardiography (Echo), is the most accurate measure of LA size[1314] because little variations in the linear dimension are often associated with large variation in volume as the result of asymmetric LA remodeling.[15]

Indeed, it is well known that LA volume index (LAVi) is associated with a major risk of atrial fibrillation (AF) and supraventricular arrhythmias, and it is also a marker of diastolic dysfunction, diastolic HF, and major clinical events.[16]

To date, the clinical meaning of LAVi, measured during stress Echo, in HCM patients has not yet been investigated. This study sought to evaluate the correlation between LAVi, measured rest (rLAVi) and stress (sLAVi), and arrhythmias in HCM patients.

MATERIALS AND METHODS

Study population

In this observational, prospective study, we collected data from a cohort of 51 consecutive patients affected by HCM, followed from July 2014 to March 2017 at Cardiomyopathy Clinic, Monaldi Hospital, Naples, Italy.

HCM was diagnosed by two-dimensional Echo, in the presence of unexplained maximal wall thickness ≥15 mm in index cases or ≥13 mm in their family members.[1]

We performed complete clinical evaluation, electrocardiogram (ECG), and standard and stress Echo in all patients, following a common protocol.

Echocardiographic analysis

Echocardiographic measurements at rest and during exercise were carried out using a Vivid E9 ultrasound system, GE Healthcare, Chicago, Illinois, United States. Data were stored digitally for review and analysis using EchoPAC software, GE Healthcare, Chicago, Illinois, United States.

Bicycle exercise Echo was performed during upright posture. The patient pedals against an increasing workload at a constant cadence. The workload is escalated in a step-wise fashion while imaging is performed. Although imaging can be done throughout the exercise protocol, in most cases, interpretation is based on a comparison of resting and peak exercise images, including chamber dimensions (atria and ventricles), systolic and diastolic function, and tissue Doppler parameters. According to international recommendation, LAVi was determined using the area–length method.[14] rLAVi was considered as the mean value of three different measures, calculated at rest, in accordance with the American Society of Echocardiography guidelines.[14]

In the same way, sLAVi was considered as the mean value of three different measures calculated at the peak of bicycle exercise.

Follow-up data

Patients were followed up periodically, every 6–12 months, according to the decision of physicians. Patients who refused follow-up visits were not considered in the analysis of endpoints. Follow-up was programmed until the visit closest to 18 months. At each visit, Holter monitor report or implantable cardioverter defibrillator interrogation, was recorded.

At follow-up, the following composite endpoints were considered: arrhythmias (ARRHYT) endpoint (AF; supraventricular tachycardia [SVT]; nonsustained ventricular tachycardia [ns-VT]; sustained-VT; ventricular fibrillation; SCD) and HF endpoint (significant worsening of NYHA class and left ventricular ejection fraction, hospitalization, and death for end-stage HF).

Thus, we analyzed the correlation between rLAVi and sLAVi and the composite ARRHYT endpoint.

Statistical analysis

Normally distributed and skewed continuous data were presented as mean ± standard deviation and median ± interquartile range, respectively, whereas percentages were used for categorical data. Comparison between groups was based on Student's t (for paired and unpaired data), Mann–Whitney U, Fisher's exact, or Chi-square tests, when appropriate.

Thereafter, given current lack of generally accepted cutoffs for LA volume in these patients, the study population was also stratified, first in quartiles, and then dichotomized in two groups, according to the median value of LAVi (≤ or > of the median value). This stratification was performed both for sLAVi and rLAVi. Intraobserver variability was good for both sLVAI and rLAVi (intraclass correlation coefficient =0.60, confidence interval [CI]: 0.3-0.89; 0.87, CI 0.7-0.93.

In addition, univariate logistic regression analyses were performed to identify association between LAVi (at rest and stress) and the ARRHYT endpoint. We tested four models, where the dependent variable was the ARRHYT endpoint, and the independent variable was sLAVi or rLAVi as continuous variables or as dichotomic variables (in model A: sLAVi as dichotomic variables, in model B: sLAVi as continuous variables, in model C: rLAVi as dichotomic variables, and in model D: rLAVi as continuous variables) We compared these four models using Akaike Information Criterion (AIC) as major criteria.

For the best model, a receiver operating curve estimation was used to calculate C-statistics for composite endpoint. Thus, sensitivity, specificity, and negative and positive predictive values of the model were calculated.

We did not consider multivariable analysis owing to the exiguity of our sample size.

RStudio Team (2015). RStudio: Integrated Development for R. RStudio, Inc., Boston, MA; URL: http://www.rstudio.com/. was used for statistical analysis. P < 0.05 was considered statistically significant, with all tests being two sided.

RESULTS

We studied 51 consecutive patients (mean age: 39.41 ± 17.9 years; male = 39, female = 12). The mean follow-up was 1.71 years, and the median follow-up was 1.82 years. Clinical characteristics of the study population are reported in Table 1.

Table 1

Clinical features

Variable	Value
Age (years), mean	39.41±17.9
Sex male, %	39 (76.5)
BSA, m2	1.9±0.2
FH-SCD, n (%)	12 (23.5)
FH-HCM, n (%)	13 (25.5)
NYHA class, n (%)
I	27 (52.9)
II	22 (43.1)
III	2 (3.9)
IV	0 (0)
NT-proBNP, mean±SD, ng/mL	446.7±385.6
Peak VO2, mean±SD, mL/kg/min	18.75±4.6
Syncope, n (%)	8 (15.7)
History of supraventricular ARRHYT, n (%)	18 (35.3)
History of SVT, n (%)	17 (33.3)
History of AF, n (%)	4 (7.8)
History of ns-VT, n (%)	14 (27.5)
History of s-VT, n (%)	0
History of VF	0
Beta blockers, n (%)	39 (76.5)
Amiodarone	2 (3.9)
ICD, n (%)	7 (13.7)

AF=Atrial fibrillation, BSA=Body surface area, FH=Family history, HCM=Hypertrophic cardiomyopathy, ICD=Implantable cardioverter defibrillator, ns-VT=Nonsustained ventricular tachycardia, SCD=Sudden cardiac death, s-VT=Sustained ventricular tachycardia, SVT=Supraventricular tachycardia, VF=Ventricular fibrillation, NYHA=New York Heart Association, SD=Standard deviation, ARRHYT=Arrhytmic endpoint

Stress echocardiography results

Stress Echo findings are reported in Table 2. At the peak of exercise, we observed a significant increase of LAVi, left ventricular outflow tract gradient, tricuspid S wave, and tricuspid annular plane systolic excursion.

Table 2

Echocardiographic changes from rest to stress

	Rest	Stress	P
LAVi, mL/m2	28.04±9.9	31.16±10.1	<0.001
LVOT gradient max, mmHg	18.4±15.7	41.2±29	<0.001
Mitral regurgitation, n (%)	47 (92.2)	47 (92.2)	0.712
Mild	33	30	0.541
Moderate	14	17	0.518
Severe	0	0	-
Tricuspid regurgitation	47 (92.2)	47 (92.2)	0.713
Mild	46	46	0.713
Moderate	1	1	0.475
Severe	0	0	-
RV-S, cm/s	11.3±2.6	13.5±1.8	<0.001
TAPSE, mm	22.4±3.8	28.2±4.1	<0.001

LAVi=Left atrial volume index, LVOT=Left ventricular outflow tract obstruction, RV=Right ventricle, TAPSE=Tricuspid annular plane systolic excursion, RV-S=Right ventricular peak systolic tissue velocity

sLAVi was significantly increased in 46 patients (90.2%) compared to the rest (mean values: 31.16 ± 10.15 mL/m2 vs. 28.05 ± 9.89 mL/m2, respectively, P < 0.001).

Follow-up

The follow-up was completed in 43 patients (84%) [Table 3]. The composite HF endpoint occurred in five (11.6%) patients. Eight patients (18.6%) showed supraventricular arrhythmias, whereas 10 patients (23.2%) had ventricular arrhythmias and 13 patients (30.2%) presented atrial and ventricular arrhythmias (ARRHYT endpoint).

Table 3

Follow-up data

Variable	Value
NYHA class, n (%)
I	27 (62.7)
II	14 (32.6)
III	2 (4.7)
IV	0
NT-ProBNP (ng/mL)	516.8±278.9
End-stage evolution, n (%)	1 (2.3)
Heart failure hospitalization, n (%)	0
Syncope, n (%)	1 (2.3)
Supraventricular ARRHYT, n (%)	8 (18.6)
AF	1 (2.3)
SVT	7 (16.3)
Ventricular arrhytmias, n (%)	10 (23.2)
ns-VT	10 (23.2)
s-VT	0
VF	0
ICD	9 (20.9)
Appropriate shock	0
Inappropriate shock	1 (2.3)
Death, n (%)	0

AF=Atrial fibrillation, ICD=Implantable cardioverter defibrillator, ns-VT=Nonsustained ventricular tachycardia, s-VT=Sustained ventricular tachycardia, SVT=Supraventricular tachycardia, VF=Ventricular fibrillation, NYHA=New York Heart Association, ARRHYT=Arrhytmic endpoint

The comparison of rLAVi (A) and sLAVi (B) in patients with and without ARRHYT endpoint is reported in Figure 1.

Figure 1

Rest left atrial volume index in patients with (right) and without (left) ARRHYT=Arrhytmic endpoint (a). Stress left atrial volume index in patients with (right) and without (left) ARRHYT endpoint (b)

Among groups in different rLAVi and sLAVi quartiles, patients in the 3rd and 4th ones had a significant higher occurrence of ARRHYT endpoint, particularly for sLAVi [Figure 2].

Figure 2

Rest left atrial volume index in patients with and without ARRHYT endpoint (a). Stress left atrial volume index in patients with and without ARRHYT endpoint (b). Left atrial volume index data are divided in quartiles. Dark gray (0): Absence of ARRHYT endpoint. Light gray (1): Occurrence of ARRHYT=Arrhytmic endpoint

On analysis of the four models (A, B, C, and D), the AIC values were, respectively, 48.37, 52.78, 50.37, and 50.76. Hence, the model A (sLAVi as independent variable, dichotomized according to values > or ≤ of the median value, 30 mL/m2) was proved to be the most accurate.

Patients with sLAVi >30 mL/m2 had a significantly increased incidence of ARRHYT endpoint compared with patients with sLAVi ≤30 mL/m2 (10 out of 19, 52.6% vs. 3 out of 24, 12.5%, P value = 0.00768), being with odds ratio of 7.78 (CI = 1.88, 41.37) and the relative risk of 4.2 (CI = 1.34, 13]) as shown in Figure 3. The distribution of ARRHYT endpoint between patients with sLAVi ≤ and >30 mL/m2 is summarized in Table 4.

Figure 3

ARRHYT endpoint: Comparison between stress left atrial volume index ≤ 30 mL/m2 and > 30 mL/m2. Dark gray (0): Absence of ARRHYT endpoint. ARRHYT=Arrhytmic endpoint. Light gray (1): Occurrence of ARRHYT endpoint

Table 4

Occurrence of endpoints according to sLAVi

sLAVi	≤30 mL/m2 24 (55.8%)	>30 mL/m2 19 (44.2%)	P
ARRHYT endpoint, n (%)	3 (12.5)	10 (52.6)	0.007
AF, n (%)	0	1 (5.3)	0.906
SVT, n (%)	2 (8.3)	4 (21.0)	0.452
ns-VT, n (%)	3 (12.5)	7 (36.8)	0.130
s-VT, n (%)	0	0	-
VF, n (%)	0	0	-
Syncope of likely cardiogenic nature, n (%)	0	1 (5.3)	0.906
SCD, n (%)	0	0	-

AF=Atrial fibrillation, ns-VT=Nonsustained ventricular tachycardia, SCD=Sudden cardiac death, s-VT=Sustained ventricular tachycardia, SVT=Supraventricular tachycardia, VF=Ventricular fibrillation, ARRHYT=Arrhytmic endpoint, sLAVI=Stress left atrial volume index

C-statistics, sensitivity, specificity, positive predictive value, negative predictive value, and predictive accuracy of this cutoff are reported in Table 5.

Table 5

Predictive performance of sLAVi >30 mL/m2

Index	Value
C-statistics	0.7346
Sensitivity (%)	76.9
Specificity (%)	70.0
Positive predictive value (%)	52.6
Negative predictive value (%)	87.5
Predictive accuracy (%)	72.1

sLAVI=Stress left atrial volume index

DISCUSSION

LA size increases in response to the following two main pathophysiologic conditions: pressure overload (e.g., mitral stenosis and increased left ventricular filling pressure due to diastolic and/or systolic dysfunction) and volume overload (e.g., mitral regurgitation [MR] and left-to-right shunts).[17] Thus, increased LA volume may be considered as an adaptive response in order to maintain left ventricular stroke volume and cardiac output. However, continued LA enlargement may result in decreased LA compliance, reduced reservoir and contractile pump functions, and eventually increased incidence of atrial arrhythmias.[18]

In patients with HCM, the cause of LA enlargement is commonly related to systolic anterior movement-related MR and elevated left ventricular filling pressure.[19]

From a clinical point of view, LA anteroposterior diameter is a well-known marker of HCM severity, and a close monitoring, due to the risk of arrhythmias and adverse events, is recommended.[20] In recent years, much more attention has been concentrated around LA volume, which is a more precise marker of hemodynamic load, disease progression, and unfavorable outcome.[1121]

An increased LA volume in HCM is associated with greater hypertrophy, more diastolic dysfunction, and higher filling pressures.[12]

Compared with LA diameter, both LA volume and LA strain represent more powerful markers of new-onset AF.[22] Particularly, in those patients with normal LA diameter, a LA volume ≥37 mL/m2 and LA strain ≤23.4% improve the prediction of new-onset AF in HCM patients (C-statistics 73% and 83%, respectively).

However, the role of sLAVi in patients with HCM is still unknown. There is increasing interest for atrial size/function during exercise in physiology and different disease conditions. The dynamic of LA size in normal individuals and athletes heart has been previously investigated.[2324] During physical exercise, the increase of cardiac output, despite a reduction in filling time (due to higher heart rate), seems to be related to a progressive atrial remodeling, an increase in reservoir function, and a consequent increased ventricular preload.[25] These phenomena can be exacerbated in a peculiar disease model such as HCM, where dynamic MR, LVOT obstruction, and even diastolic dysfunction may play a crucial role during exercise.

To our knowledge, this is the first report of the value of sLAVi, measured at the peak of exercise, as a simple and noninvasive parameter to improve risk stratification of arrhythmias in patients with HCM. In this study, we enrolled 51 consecutive HCM patients who underwent stress Echo following a common protocol. The main finding of this study was the correlation between sLAVi and arrhythmic events (both atrial and ventricular) in patients with HCM, with an incremental risk in the population with sLAVi >30 mL/m2 (predictive accuracy: 72.1%, C-statistics: 0.7346, and negative predictive value: 87.5%). Furthermore, in our study, LAVi at stress performed better than LAVi at rest as a predictor of arrhythmias (AIC: 48.37 vs. 50.37, if dichotomized according to the median values).

In a recent meta-analysis, Liu et al. demonstrated that AF and ns-VT were independent predictors of cardiovascular death.[26] The independent role of sLAVi as a predictor of arrhythmias, but also of unfavorable outcome, should be further reinvestigated in prospective, large-scale, collaborative studies.[27]

Study limitations

Our findings are based on a single-center study with a relatively small sample size. Multivariate models on a significant number of HCM patients, and considering rest/stress parameters (i.e., LAVi, MR, and LVOT obstruction), are warranted to define the independent prognostic role of LAVi.

CONCLUSIONS

Our study is the first study suggesting the feasibility and potential utility of sLAVi in HCM. We showed that patients with LAVi >30 mL/m2 had an increased risk of arrhythmias when compared to patients with HCM. This suggests that sLAVi could be a potential and additional useful marker of arrhythmias in patients with HCM. Further studies are needed to verify the direct relation between sLAVi and arrhythmias and its possible role as an independent predictor of prognosis.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.