Clin Res Cardiol DOI 10.1007/s00392-015-0873-9

ORIGINAL PAPER

Heart rate turbulence and deceleration capacity for risk prediction of serious arrhythmic events in Marfan syndrome Benjamin N. Schaeffer1 • Meike Rybczynski2 • Sara Sheikhzadeh2 • ¨ . Akbulak2 • Julia Moser1 • Mario Jularic1 • Doreen Schreiber1 • Ruken O Anne Daubmann3 • Stephan Willems1 • Yskert von Kodolitsch2 • Boris A. Hoffmann1

Received: 21 February 2015 / Accepted: 20 May 2015 Ó Springer-Verlag Berlin Heidelberg 2015

Abstract Objective Marfan syndrome (MFS) is associated with a substantial risk for ventricular arrhythmia and sudden cardiac death (SCD). We used heart rate turbulence (HRT) and deceleration capacity (DC), to evaluate the risk stratification for these patients. Methods We enrolled 102 patients [45 male (44.1 %), age 40.5 ± 14.6 years] with MFS. Blood samples were obtained to determine N-terminal pro-brain natriuretic peptide (NT-proBNP) levels. Transthoracic echocardiography studies were conducted to evaluate heart function parameters and a 24-h holter ECG was performed. An analysis of two HRT parameters, turbulence onset (TO) and turbulence slope (TS), and DC was performed. Therefore, optimal cut-off values were calculated. Primary endpoint was the combination of SCD, ventricular arrhythmia and arrhythmogenic syncope. Secondary endpoint was total mortality. Results During a follow-up of 1145 ± 491 days, 12 (11.7 %) patients reached the primary and 8 (7.8 %) patients the secondary endpoint. Patients reaching the primary were significantly older, had a higher burden of premature ventricular complexes and NT-proBNP levels and lower values of LVEF, DC and HRT TS. Multivariate

& Benjamin N. Schaeffer [email protected] 1

Department of Cardiology - Electrophysiology, University Heart Center, University Hospital Hamburg, Martinistr. 52, 20246 Hamburg, Germany

2

Department of Cardiology, University Heart Center, University Hospital Hamburg, Hamburg, Germany

3

Department of Medical Biometry and Epidemiology, University Hospital Hamburg, Hamburg, Germany

analysis identified NT-proBNP (HR 1.25, 95 % CI 1.01–1.56, p = .04) and the abnormal HRT (abnormal TS and/or TO (HR 7.04, 95 % CI 1.07–46.27, p = .04) as independent risk predictor of arrhythmogenic events. Conclusion Patients with Marfan syndrome are at risk for severe ventricular arrhythmias and SCD. Abnormal HRT parameters and NT-proBNP values are independent risk factors for arrhythmogenic events and SCD. The assessment of these tools may help predicting SCD patients with MFS. Keywords Marfan syndrome  Heart rate turbulence  Risk stratification  Sudden cardiac death

Introduction Marfan syndrome (MFS) is an autosomal-dominant inherited disorder of the connective tissue involving multiple organ systems. Cardiovascular manifestations include aortic root dilatation, and regurgitation of the aortic and mitral valve [1]. Fatal consequences of aneurysms and dissections of the thoracic aorta are the major cause of death in these patients [2]. Furthermore, there is evidence indicating an increased risk for sudden cardiac death (SCD) and ventricular arrhythmias [3–5]. Other studies have identified ventricular arrhythmias in up to one-fifth of adult patients with MFS and even greater prevalence in children [3, 5, 6]. Arrhythmia related death was seen in 2.5–4 %, indicating the substantial risk [3, 5]. Previous studies have identified premature ventricular complexes (PVCs) and N-terminal pro-brain natriuretic peptide (NT-proBNP) as predictors of arrhythmogenic events in patients with MFS [4, 5]. Other ECG-based characteristics and indicators for potential arrhythmia such as repolarization abnormalities

123

Clin Res Cardiol

failed to predict arrhythmogenic and SCD events in MFS [5, 7]. Heart rate turbulence (HRT) [8, 9] and deceleration capacity (DC) [10, 11] are holter ECG-recording (HECG) based tools which have shown to predict risk of SCD in patients with cardiovascular diseases such as acute and post-myocardial infarct phase [8, 11, 12], congestive heart failure and cardiomyopathy [13, 14], diabetes mellitus [15], Aortic stenosis [16] and Chagas disease [17]. These easyto-use tools are based on the analysis of changes in heart rate after a PVC (heart rate turbulence) and the capacity of the heart to reduce the rate by each beat (deceleration capacity) [8–10]. Heart rate turbulence and deceleration capacity are mediated by the autonomic nervous system and serve as marker of autonomic reflex activity (HRT) and autonomic tonic activity (DC) [8–11]. So far, autonomic dysfunction does not seem to be of great clinical relevance in MFS. But in general, alterations in the autonomic nervous system profoundly affect the risk of malignant ventricular arrhythmia and SCD, and therefore we aimed for new algorithms determining autonomic dysfunction. An improved risk stratification of SCD may be essential to initiate life-saving prevention measures such as ICD-implantation. Thus, we evaluated a new tool for SCD risk stratification in patients with MFS by analysis of HRT and DC.

Materials and methods Population We enrolled 102 consecutive patients with diagnosed MFS in this prospective register study. This is an extended cohort of Marfan patients, which have been previously investigated in our center [5]. No power or sample size calculation was conducted. Inclusion criteria were: (1) diagnosed MFS in accordance to revised Ghent criteria [18, 19]; (2) signed informed consent to the study; (3) age [18 years. Exclusion criteria were: (1) survived SCD or documented sustained ventricular arrhythmia ([30 s); (2) coronary heart disease; (3) persistent atrial fibrillation; (4) pacemaker rhythm. In 88 %, a causative FBN1-gene mutation was verified. Patients were seen between January 2007 and November 2010 in our outpatient clinic either for annual or half-annual follow-ups or at the time of their first visit with clinical presentation of MFS. No patients had to be excluded. The final study group comprised 44 men (43 %) and 58 women with a mean age of 40.5 ± 14.6 years. The study was conducted in accordance with the provisions of the Declaration of Helsinki and amendments.

123

Baseline examination, transthoracic echocardiography and 24 h-holter ECG At study entry, all patients underwent clinical examination. For analysis of NT-proBNP, venous blood samples were obtained and analyzed (Elecsys 2010 analyzer, Roche Diagnostics, Mannheim, Germany). A transthoracic echocardiography (TTE) and a 24 h-holter ECG (24 hHECG) were performed as described previously [5]. TTE was performed in parasternal, apical and sub-xiphoid views using a 4S probe and an echocardiography system (iE33, Philips Medical Systems, Eindhoven, The Netherlands). LVEF measurements were calculated offline with an ultrasound workstation (syngo 3.5, Siemens Medical Solutions, Erlangen, Germany) from apical two and four chamber 2D views using Simpson’s rule. Analysis of HRT and DC All patients underwent 24 h-HECG using a five-channel high-resolution digitized recorder (Medilog AR12, Schiller Medilog Inc., Baar, Switzerland). The recordings were processed and analyzed as described before [5] and transformed into the Schiller-BIN-format by using the holter ECG-analysis software (Medilog Darwin Holter Analysis, Schiller Medilog Inc., Baar, Switzerland) for further analysis. HRT and DC analysis was performed using a custommade software (libRASCH-0.7.4, Technical University Munich). Heart rate turbulence was analyzed as described by Schmidt et al. [8] and Bauer et al. [9, 10] (comprehensive overview). A minimum number of 25 PVCs were required for reliable reconstruction of PVC-tachogram. Quantification of heart rate turbulence was made by two parameters: (1) turbulence onset (TO) and (2) turbulence slope (TS). Turbulence onset represents the percentage change of the heart rate after a PVC is compared to the heart rate before the PVC (early acceleration). It is defined as [8]: TO ¼

ðRR1 þ RR2 Þ  ðRR2 þ RR1 Þ  100 % ðRR2 þ RR1 Þ

where RR1 and RR2 are the two RR-intervals (RRI) following the PVC, and RR-1 and RR-2 are the two RRI preceding the PVC. As published before, a TO \0 % or \25 PVCs/24 h was considered normal, whereas TO C0 % was considered abnormal [9]. TS was generated by analysis of the steepest positive regression slope over 5 consecutive RRI within the following 15 sinus beats after the PVC [9]. To ensure detection free of artifacts or errors, the following criteria were required: sinus rhythm at least 5 beats prior to PVC. Detection and elimination of interpolated PVCs (minimum of [20 % prematurity of the PVCs and a

Clin Res Cardiol

minimum of [120 % compensatory pause afterwards, referred to the mean of last five RRI preceding the PVC). A minimum of 15 sinus beats after the PVC. Using the phase-rectifying signal averaging (PRSA) technique [10] to process RR-intervals obtained from HECG recordings, an analysis of deceleration- and acceleration-related modulation of the heart rate can be made, which is quantified by the deceleration capacity (DC) and acceleration capacity (AC) [10]. This approach has proven to identify post-myocardial infarction patients with a high risk for death [10]. For all relevant variables, optimal cut-off values were calculated using receiver operator characteristics (ROC) analysis. Cut-off values of TS, TO and DC were used to assign patients into groups: For HRT, patients were assigned into one of three groups: Group 0 (normal): TO \0 % and TS C cut-off value or less than 25 PVCs in 24 h. Group 1 (abnormal): TO C0 % or TS \ cut-off value, group 2 (abnormal): TO C0 % and TS \ cut-off value [15]. According to previous description, DC values equal or greater than the determined cut-off were considered normal. DC values lower than the cut-off were considered abnormal [10]. Endpoint and follow-up The primary endpoint was defined as a combination of SCD, as well as survived cardiac arrest, sustained ventricular tachycardia, ventricular fibrillation (VF) and arrhythmogenic syncope. SCD was defined as observed cardiac arrest or death within 1 h after the onset of acute symptoms, or an unexpected death in patients known to have been well within the previous 24 h. Sustained ventricular tachycardia was defined as a regular broad-complex arrhythmia (QRS-complex width [120 ms) of a duration [30 s. Ventricular fibrillation was defined as a rapid ventricular rhythm with variation in waveform and interval length characterized by a disorganized rhythm with the absences of QRS-complexes. Arrhythmogenic syncope was suspected at \-2 Calgary Syncope Symptom Score [20] and further diagnostic (in hospital monitoring, holter ECG) was initiated to identify the underlying cause. Secondary endpoint was total mortality. All patients were seen annually in our outpatient clinic, none were lost to followup.

(HR) and 95 % confidence interval (CI) with the following covariates: age, left ventricular ejection fraction, deceleration capacity, turbulence slope, premature ventricular complexes and NT-proBNP. The assumption of proportional hazards was checked using the Schoenfeld residuals. All covariates satisfied the proportional hazard assumption. Kaplan–Meier survival curves were created for analysis of the freedom from primary endpoint categorized by prior defined groups of HRT and DC and LVEF, PVC, DC, TS and NT-proBNP. Log-rank test was performed for statistical comparison. Receiver operator characteristics (ROC) analysis was performed to determine sensitivity and specificity of, LVEF, PVC, DC, TS and NT-proBNP in predicting the primary endpoint. A p value \.05 was considered statistically significant. All statistical analysis was performed using statistical software package (SAS 9.4, SAS Institute Inc., Cary, NC, USA).

Results All 24 h-HECG recordings were applicable for analysis. Table 1 shows the baseline characteristics and outcome. During a mean follow-up time of 1145 ± 491.5 days, 12 (11.7 %) patients reached the combined primary endpoint and 8 (7.8 %) patients the secondary endpoint (Table 1). Three (2.9 %) patients died due to SCD of which, in two cases, post-mortem examination verified intact aorta and no coronary artery disease. In 9 (8.8 %) cases, sustained ventricular tachycardia were documented, three (2.9 %)

Table 1 Baseline characteristics and outcome Total population (n)

102 patients

Follow-up (days)

1145 ± 491.5

Age (years)

40.5 ± 14.6

Sex (male gender)

44 (43.1 %)

Height (cm)

184.4 ± 9.9

Weight (kg)

76.5 ± 18.2

BMI (kg/m2)

22.4 ± 4.6

Beta-blocker therapy

79 (77.6 %)

Primary endpoint reached Sudden cardiac death

12 (11.7 %) 3 (2.9 %)

Statistical analysis All numerical data was shown as mean ± standard deviation (SD). Univariate analysis was performed using student’s t test. The Cox regression model was performed for multivariate analysis to identify independent risk factors of SCD and total mortality, and shown as the hazard ratio

Sustained ventricular tachycardia

9 (8.8 %)

ICD-implantation

3 (2.9 %)

Arrhythmogenic syncopea

1 (1 %)

Secondary endpoint reached

8 (7.8 %)

Non-SCD-related deathb

5 (4.9 %)

SCD sudden cardiac death a

Patient also received ICD

b

Two patients died by sepsis, both reached combined endpoint, three due to aortic rupture

123

Clin Res Cardiol

patients received an implantable cardioverter defibrillator (ICD). 5 (4.9 %) non-SCD-related deaths occurred during the follow-up period. Three individuals who refused surgery died due to aortic rupture, two patients died due to sepsis (Table 1). ROC analyses were performed to determine optimal cutoff values for LVEF, DC, TS and NT-proBNP at 44 %, 5.38 ms, 3.95 ms/RRI and 464.9 pg/ml with an area under the curve (AUC) of 0.689, 0.712, 0.749, 0.804, respectively (Fig. 1). The Kaplan–Meier curves display freedom of the

combined primary endpoint for these variables and the corresponding p values determined by log-rank test (Fig. 2). Uni- and multivariate analysis Patients who have reached the primary endpoint showed a significantly higher age, occurrence of PVCs and NTproBNP-levels and lower LVEF, DC and HRT TS (Table 2).

Fig. 1 ROC analysis. a LVEF cut-off: 44 %, AUC 0.689; b DC cut-off: 5.38 ms, AUC 0.712; c TS cut-off: 3.95 ms/RRI, AUC 0.749; d NTproBNP cut-off point: 464.9 pg/ml, AUC 0.804

123

Clin Res Cardiol

Fig. 2 Kaplan–Meier curves displaying freedom of combined primary endpoint during follow-up (days). Dotted lines represent patients [ cutoff value. a LVEF, p \ .001; b DC p = .015; c TS p = .003; d NT-proBNP p \ .001

The results of the multivariate analysis using the Cox proportional hazard regression analysis are shown in Fig. 3. NT-proBNP was an individual predictor of the primary composite endpoint (HR 1.25, 95 % CI 1.01–1.56, p = .04, Fig. 3a). In the prediction of mortality, age (HR 1.73, 95 % CI 1.05–2.83, p = .03) was the only significant variable (Fig. 3b). Analyzing the combination of HRT parameters TS and TO in multivariate analysis, abnormal HRT (HRT group 1 and 2 = abnormal TS or/and TO) was an independent predictor of the primary endpoint (HRT group 1 and 2 vs. group 0: HR 7.04, 95 % CI 1.07–46.27, p = .04). Patients with abnormal HRT were statistically significantly older (52.0 ± 13.7 years, p = .002), were more frequent under beta-blocker therapy (84.6 %, p = B .001) showed more PVCs (1730 ± 3228 per 24 h, p = .003), and had a lower LVEF (43.8 ± 14.6 %, p = .02) compared with HRT group 0 (Table 3).

Discussion In this study, we showed that analysis of 24-h-HECG based heart rate turbulence parameters predict the risk of ventricular arrhythmias and SCD in patients with MFS. 11.7 % of the patients reached the combined primary endpoint and 7.8 % the secondary endpoint during a mean follow-up of 3 years. A moderate elevated NT-proBNP cut-off level reconfirms our previous findings showing this parameter to serve as another independent predictor for SCD in patients with MFS [5, 21]. Sudden cardiac death and ventricular dysfunction The substantial risk of SCD in patients with MFS has been described before [3, 4]. Yetman et al. [4] showed that onequarter of a followed cohort of 70 patients with MFS suffered from ventricular arrhythmias during a mean

123

Clin Res Cardiol Table 2 Univariate analysis for comparison between patients with MFS: age, LVEF, PVC, DC, HRT turbulence slope and NT-proBNP differ significantly Total population (n = 102)

Combined endpoint reached (n = 12)

Combined endpoint not reached (n = 90)

p

Age (years)

40.5 ± 14.6

52.9 ± 12.1

38.9 ± 14.1

B.01

Sex (male)

44 (43.1 %)

6 (50.0 %)

38 (42.2 %)

.62

BMI (kg/m2) Beta-blocker (n, %)

22.4 ± 4.6 79 (77.4 %)

22.5 ± 3.1 11 (91.7 %)

22.4 ± 4.7 59 (65.5 %)

.77 .07

LVEF (%)

51.4 ± 12.4

42.5 ± 18.4

52.3 ± 11.1

B.01

PVC (/24 h)

529 ± 1625

2110 ± 3419

317 ± 1081

B.01

History of aortic surgery

46 (45.1 %)

8 (66.7 %)

38 (42.2 %)

.11

SDNN (ms)

146.3 ± 47.1

134.5 ± 14.3

148.1 ± 6.7

.45

DC (ms)

5.5 ± 2.5

3.8 ± 2.7

5.6 ± 2.5

.01

Turbulence onset (%)

-0.0148 ± 0.031

-0.0057 ± 0.091

-0.0162 ± 0.033

Turbulence slope (ms/ RRI)

9.5 ± 7.7

2.7 ± 2.1

11.8 ± 7.9

B.01 B.01

NT-proBNP (ng/l)

623.8 ± 1738

2671 ± 4078

350.8 ± 855.6

MVP (n, %)

54 (52.9 %)

8 (66 %)

46 (51 %)

.35

.31

LVEF left ventricular ejection fraction, PVC premature ventricular complex, SDNN standard deviation of normal-to-normal, NT-proBNP N-terminal pro-brain natriuretic peptide, MVP mitral valve prolapse

follow-up of 6 years assuming an increase of prevalence over time as the this chronic disease proceeds. Yetman et al. [4], Rybcyznski et al. [21] and Kiotsekoglou et al. [22] showed an impaired left ventricular (LV) and right ventricular (RV) function comprising abnormal systolic contractility, abnormal ventricular diastolic relaxation, and dilatation of the ventricles in patients with MFS. These structural alterations are considered to be responsible for electrophysiological changes including an altered repolarization, slowed intra-ventricular conduction and occurrence of PVCs resulting in ventricular arrhythmias and SCD [4, 5]. The underlying factors of impaired LV and RV function in MFS are not fully understood. A diminished structural integrity due to the Fibrillin-1-mutation with a dysfunction of the contractile and relaxation properties, as well as extracellular remodeling and increased wall stiffness may contribute to these findings [5, 23]. Patients reaching the endpoint showed a significantly lower LVEF, although LVEF did not appear as an independent risk factor by multivariate analysis. The level of impaired LV function in patients reaching the combined endpoint is moderate (LVEF 42.5 vs. 52.3 %) and considerably better than the common used cut-off value by which an ICD-Implantation for primary prevention of SCD is considered. In this study, we reconfirmed our findings that NT-proBNP, a parameter known for risk prediction of cardiovascular events in general population and specific diseases [24, 25], is a beneficial tool to improve prediction of arrhythmic events and SCD in MFS. However, further reliable parameters are needed to advance individual risk stratification for these patients.

123

Heart rate turbulence In this study, we accessed a new model for risk stratification in MFS by analysis of HRT. Abnormal HRT is supposed to represent depressed baroreflex sensitivity with an attenuated vagal and sympathetic response of the heart [9]. The results of this study implicate an imbalanced autonomic dysfunction in patients with MFS, reflected by an abnormal HRT, with an elevated risk for ventricular arrhythmias and SCD. HRT has been well investigated in large cohorts of post-myocardial infarct patients with an independent risk prediction for all cause of death, in particular, arrhythmia and SCD [8, 11]. The value of prediction of SCD and ventricular arrhythmia in patients with cardiomyopathies in contrast is diverse. A large study including 651 patients with mixed causes of congestive heart failure (50 % ischemic CMP) by Cygankiewicz et al. [26] showed HRT to be a risk predictor for heart failure and arrhythmic deaths. In contrast, a study by Grimm et al. [13] in 242 patients with idiopathic dilated cardiomyopathy showed a significant prediction of transplant-free survival but not for arrhythmic events. Moore et al. [27] (UK-HEART) described TS as an independent predictor of death due to decompensated heart failure in patients with mild to moderate heart failure but there was no significant association between HRT and SCD. Koyama et al. [14] identified TS as a predictor of death and hospitalization due to chronic heart failure but not for arrhythmic events in a cohort of 50 patients with congestive heart failure. It is likely that differences in the cause and the severity of heart failure, as well as patient’s characteristics contribute

Clin Res Cardiol

Fig. 3 Forrest plot: multivariate analysis of independent risk factors in patients reaching the primary endpoint. NT-proBNP (HR 1.25, 95 % CI 1.01–1.56, p = .04) showed a significant difference. b Forrest plot: multivariate analysis of independent risk factors in

patients reaching the secondary endpoint. Age was the only independent risk predictor for mortality (HR 1.73, 95 % CI 1.05–2.83, p = .03)

Table 3 Patients with normal and abnormal HRT (abnormal TS or/and TO) Normal HRT (HRT group 0, n = 89)

Abnormal HRT (HRT group 1 and 2, n = 13)

p

Endpoint reached

7 (7.8 %)

5 (38.5 %)

Age (years)

38.9 ± 14.0

52.0 ± 13.7

.001 .002

Sex (male)

38 (43 %)

6 (46.2 %)

.81

Beta-blocker (n, %)

30 (33 %)

11 (84.6 %)

B.001

LVEF (EF %)

52.2 ± 11.8

43.8 ± 14.6

.02

PVC (/24 h)

353 ± 1168

1730 ± 3228

.003

History of aortic surgery

37 (41 %)

9 (69.2 %)

.06

NT-proBNP (ng/l)

612 ± 1843

705 ± 717

.86

MVP (n, %)

44 (49 %)

10 (76.9 %)

.6

Follow-up (days)

1160 ± 485

1048 ± 543

.44

123

Clin Res Cardiol

to these inconsistent findings. In our study, patients with MFS and elevated risk for arrhythmia and SCD showed lower LVEF values although the LVEF was not an independent risk factor at multivariate analysis. Alpendurada et al. [28] described an independent Marfan cardiomyopathy entity in a subset of patents with MFS in a cardiac MRT study with rather low-grade reduction of LVEF. The ability of HRT and especially TS to predict SCD and ventricular arrhythmias in patients with MFS better than LVEF may reflect a complex mechanism of heart failure leading to autonomic dysfunction in these patients. Despite no clear validation, the before mentioned structural changes of the myocardium due to Fibrillin-1-mutation leading to a multifaceted functional and structural impairment of the ventricles may give rise to a more severe autonomic dysfunction accompanied by a better compensated LVEF, and represent a possible underlying cause. According to the aforementioned studies, [14, 27] especially TS seems to be affected in the setting of heart failure, as this parameter differed significantly within patients reaching the endpoint in our study. The presence of alterations in the autonomic system in patients with MFS is supported by our previous findings, which showed a significantly reduced normalized low frequency (LF) component and LF/high frequency (HF) ratio in heart rate variability analysis reflecting depressed vagal activity. Heart rate variability (SDNN), a parameter of good predictive value of arrhythmic events following myocardial infarction [29, 30], in our previous and in the present study, was not affected in patients with MFS [5, 7]. It therefore does not seem to be a reliable parameter for SCD risk prediction in MFS.

Deceleration capacity Deceleration capacity and the combination of HRT and DC, described as severe autonomic failure (SAF), is a strong predictor of death and indicator for high-risk postmyocardial infarction patients [10, 11]. To our knowledge, there are no studies concerning the role of DC in risk evaluation in patients with heart failure alone. Further, data regarding possible autonomic dysfunction in MFS is rare. In this study, DC did not predict arrhythmogenic events, SCD or mortality as an independent risk factor in multivariate analysis at the determined cut-off of 5.38 ms. DC is believed to reflect autonomic tonic activity (DC) [10, 11]. Our results may indicate that patients with MFS and a risk for severe arrhythmic events have a larger tendency towards impaired autonomic reflex activity displayed by abnormal HRT than to impairment of autonomic tonic activity displayed by normal DC; although, a larger cohort of patients may alter our findings. HRT and mitral valve prolapse (MVP) Previous studies showed an impaired HRT in patients with merely MVP [32]. This common entity among patients with MFS (53 % in this study) showed no significant impact on the achievement of the combined endpoint in our study (66 vs. 51 %; p = .38). These findings are consistent with early observations of Chen et al. [6] who described ventricular arrhythmias in children with MFS with or without substantial valve disease. In contrast, Yetman et al. [4] reported an association between MVP and ventricular arrhythmias in their study, although arrhythmogenic events occurred even in the absence of MVP.

HRT and beta-blocker Limitations There is only limited data on the effect of beta-blocking agents to HRT in general and specifically in heart failure. Lin et al. [31] showed in a small cohort of 12 patients with congestive heart failure (NYHA III and IV), the restoration of impaired TS values after 3 months treatment with the selective beta-blocker atenolol to normal levels. This was concluded to be a result of improved baroreceptor control. TO was marginally affected in this study. In patients with MFS, beta-blockers are widely used (71 % in this study). Patients reaching the endpoint showed a tendency towards higher beta-blocker usage (p = .07) but there were only insignificant differences in beta-blocker therapy within the different HRT groups; this may indicate that the underlying factors for impaired HRT in MFS are not affected by betablockers. This would be in line with findings from Schmidt et al. [8] showing HRT as mortality predictor in post-myocardial infarction patients with impaired LVEF despite beta-blockade.

123

Marfan syndrome is a rather rare disease. Therefore, the number of patients included in this study is limited. Statistical analysis may be bounded by this fact to some degree. A larger study population possibly would have shown further statistically significant parameters like the deceleration capacity. There was no healthy control group in this trial. Therefore, an increased risk for severe arrhythmogenic events cannot be demonstrated. Although, others as well as our groups have shown in previous studies that MFS is clearly associated with an increased risk of arrhythmogenic events compared to healthy population [4, 5, 7]. The evaluation of HRT and DC is based on single HECG recordings. Repetitive HECG recordings may have shown further information about the time course as this chronic disease proceeds. Sudden cardiac death may be multifactorial and diseases such as coronary artery disease (CAD) may also increase the risk of SCD. Due to the lower

Clin Res Cardiol

age of the patients in this study and the absence of ECG signs of ischemia, we believe that CAD is unlikely.

Conclusion SCD and severe ventricular arrhythmias remain a substantial risk for patients with MFS. We showed that analysis of NT-proBNP and abnormal HRT parameter independently predicts SCD in MFS. This easy to perform screening test may promote the risk stratification of SCD in MFS. Acknowledgments Special thanks to Prof. Dr. Georg Schmidt, Alexander Mu¨ller and Raphael Schneider, Department of Cardiology, Technical University, Munich, Germany for providing the custommade software. Conflict of interest interest.

All Authors declare that there is no conflict of

References 1. von Kodolitsch Y, Rybczynski M (2004) Cardiovascular aspects of the Marfan syndrome—A systematic review. In: Robinson PN, Godfrey M (eds) Marfan syndrome: a primer for clinicians and Scientists. Kluwer Academic Publishers, Dordrecht [ISBN 0-30648238-X:45-69] 2. Devereux RB, Roman MJ (1999) Aortic disease in Marfan’s syndrome. N Engl J Med. 340:1358–1359 3. Savolainen A, Kupari M, Toivonen L, Kaitila I, Viitasalo M (1997) Abnormal ambulatory electrocardiographic findings in patients with the Marfan syndrome. J Intern Med 241(3):221–226 4. Yetman AT, Bornemeier RA, McCrindle BW (2003) Long-term outcome in patients with Marfan syndrome: is aortic dissection the only cause of sudden death? J Am Coll Cardiol 41(2): 329–332 5. Hoffmann BA, Rybczynski M, Rostock T, Servatius H, Drewitz I, Steven D, Aydin A, Sheikhzadeh S, Darko V, von Kodolitsch Y, Willems S (2013) Prospective risk stratification of sudden cardiac death in Marfan’s syndrome. Int J Cardiol 167(6):2539–2545 6. Chen S, Fagan LF, Nouri S, Donahoe JL (1985) Ventricular dysrhythmias in children with Marfan’s syndrome. Am J Dis Child 139(3):273–276 7. Aydin A, Adsay BA, Sheikhzadeh S, Keyser B, Rybczynski M, Sondermann C, Detter C, Steven D, Robinson PN, Berger J, Schmidtke J, Blankenberg S, Willems S, von Kodolitsch Y, Hoffmann BA (2013) Observational cohort study of ventricular arrhythmia in adults with Marfan syndrome caused by FBN1 mutations. PLoS One 8(12):e81281 8. Schmidt G, Malik M, Barthel P, Schneider R, Ulm K, Rolnitzky L, Camm AJ, Bigger JT, Scho¨mig A (1999) Heart-rate turbulence after ventricular premature beats as a predictor of mortality after acute myocardial infarction. Lancet 353(9162):1390–1396 9. Bauer A, Malik M, Schmidt G, Barthel P, Bonnemeier H, Cygankiewicz I, Guzik P, Lombardi F, Mu¨ller A, Oto A, Schneider R, Watanabe M, Wichterle D, Zareba W (2008) Heart rate turbulence: standards of measurement, physiological interpretation, and clinical use: International Society for Holter and Noninvasive Electrophysiology Consensus. J Am Coll Cardiol 52(17):1353–1365

10. Bauer A, Kantelhardt JW, Barthel P, Schneider R, Ma¨kikallio T, Ulm K, Hnatkova K, Scho¨mig A, Huikuri H, Bunde A, Malik M, Schmidt G (2006) Deceleration capacity of heart rate as a predictor of mortality after myocardial infarction: cohort study. Lancet 367(9523):1674–1681 11. Bauer A, Barthel P, Mu¨ller A, Ulm K, Huikuri H, Malik M, Schmidt G (2009) Risk prediction by heart rate turbulence and deceleration capacity in postinfarction patients with preserved left ventricular function retrospective analysis of 4 independent trials. J Electrocardiol 42(6):597–601 12. Sade E, Aytemir K, Oto A, Nazli N, Ozmen F, Ozkutlu H, Tokgo¨zoglu L, Akso¨yek S, Ovu¨nc¸ K, Kabakc¸i G, Ozer N, Kes S (2003) Assessment of heart rate turbulence in the acute phase of myocardial infarction for long-term prognosis. Pacing Clin Electrophysiol 26(2 Pt 1):544–550 13. Grimm W, Schmidt G, Maisch B, Sharkova J, Mu¨ller HH, Christ M (2003) Prognostic significance of heart rate turbulence following ventricular premature beats in patients with idiopathic dilated cardiomyopathy. J Cardiovasc Electrophysiol 14(8): 819–824 14. Koyama J, Watanabe J, Yamada A, Koseki Y, Konno Y, Toda S, Shinozaki T, Miura M, Fukuchi M, Ninomiya M, Kagaya Y, Shirato K (2002) Evaluation of heart-rate turbulence as a new prognostic marker in patients with chronic heart failure. Circ J 66(10):902–907 15. Balciog˘lu S, Arslan U, Tu¨rkog˘lu S, Ozdemir M, Cengel A (2007) Heart rate variability and heart rate turbulence in patients with type 2 diabetes mellitus with versus without cardiac autonomic neuropathy. Am J Cardiol 100(5):890–893 16. Zuern CS, Eick C, Rizas KD, Stoleriu C, Barthel P, Scherer C, Mu¨ller KA, Gawaz M, Bauer A (2012) Severe autonomic failure in moderate to severe aortic stenosis: prevalence and association with hemodynamics and biomarkers. Clin Res Cardiol 101(7): 565–572 17. Ribeiro AL, Schmidt G, Sousa MR, Lombardi F, Gomes ME, Perez AA, Barros MV, Machado FS, Rocha MO (2003) Heart rate turbulence in Chagas disease. Pacing Clin Electrophysiol 26(1 Pt 2):406–410 18. Rybczynski M, Bernhardt AM, Rehder U, Fuisting B, Meiss L, Voss U, Habermann C, Detter C, Robinson PN, Arslan-Kirchner M, Schmidtke J, Mir TS, Berger J, Meinertz T, von Kodolitsch Y (2008) The spectrum of syndromes and manifestations in individuals screened for suspected Marfan syndrome. Am J Med Genet A 146A(24):3157–3166 19. De Paepe A, Devereux RB, Dietz HC, Hennekam RC, Pyeritz RE (1996) Revised diagnostic criteria for the Marfan syndrome. Am J Med Genet 62(4):417–426 20. Sheldon R, Rose S, Connolly S, Ritchie D, Koshman ML, Frenneaux M (2006) Diagnostic criteria for vasovagal syncope based on a quantitative history. Eur Heart J 27(3):344–350 21. Rybczynski M, Koschyk DH, Aydin MA, Robinson PN, Brinken T, Franzen O, Berger J, Hofmann T, Meinertz T, von Kodolitsch Y (2007) Tissue Doppler imaging identifies myocardial dysfunction in adults with Marfan syndrome. Clin Cardiol 30(1): 19–24 22. Kiotsekoglou A, Sutherland GR, Moggridge JC, Kapetanakis V, Bajpai A, Bunce N, Mullen MJ, Louridas G, Nassiri DK, Camm J, Child AH (2009) Impaired right ventricular systolic function demonstrated by reduced atrioventricular plane displacement in adults with Marfan syndrome. Eur J Echocardiogr 10(2):295–302 23. Ramirez F, Dietz HC (2007) Marfan syndrome: from molecular pathogenesis to clinical treatment [Curr Opin Genet Dev 17 (2007) 252–258]. Curr Opin Genet Dev 17(4):367 24. Leistner DM, Klotsche J, Pieper L, Palm S, Stalla GK, Lehnert H, Silber S, Ma¨rz W, Wittchen HU, Zeiher AM (2013) Prognostic value of NT-pro-BNP and hs-CRP for risk stratification in

123

Clin Res Cardiol

25.

26.

27.

28.

primary care: results from the population-based DETECT study. Clin Res Cardiol 102(4):259–268 Franke J, Lindmark A, Hochadel M, Zugck C, Koerner E, Keppler J, Ehlermann P, Winkler R, Zahn R, Katus HA, Senges J, Frankenstein L (2014) Gender aspects in clinical presentation and prognostication of chronic heart failure according to NT-proBNP and the Heart Failure Survival Score. Clin Res Cardiol 104(4):334–341 Cygankiewicz I, Zareba W, Vazquez R, Vallverdu M, GonzalezJuanatey JR, Valdes M, Almendral J, Cinca J, Caminal P, de Luna AB, Muerte Subita en Insuficiencia Cardiaca Investigators (2008) Heart rate turbulence predicts all-cause mortality and sudden death in congestive heart failure patients. Heart Rhythm 5(8):1095–1102 Moore RK, Groves DG, Barlow PE, Fox KA, Shah A, Nolan J, Kearney MT (2006) Heart rate turbulence and death due to cardiac decompensation in patients with chronic heart failure. Eur J Heart Fail 8(6):585–590 Alpendurada F, Wong J, Kiotsekoglou A, Banya W, Child A, Prasad SK, Pennell DJ, Mohiaddin RH (2010) Evidence for Marfan cardiomyopathy. Eur J Heart Fail 12(10):1085–1091

123

29. Kleiger RE, Miller JP, Bigger JT Jr, Moss AJ (1987) Decreased heart rate variability and its association with increased mortality after acute myocardial infarction. Am J Cardiol 59:256–262 30. Kleiger RE, Miller JP, Krone RJ, Bigger JT Jr (1990) The independence of cycle length variability and exercise testing on predicting mortality of patients surviving acute myocardial infarction. The Multicenter Postinfarction Research Group. Am J Cardiol 65:408–411 31. Lin LY, Hwang JJ, Lai LP, Chan HL, Du CC, Tseng YZ, Lin JL (2004) Restoration of heart rate turbulence by titrated betablocker therapy in patients with advanced congestive heart failure: positive correlation with enhanced vagal modulation of heart rate. J Cardiovasc Electrophysiol 15(7):752–756 32. Gunduz H, Arinc H, Kayardi M, Akdemir R, Ozyildirim S, Uyan C (2006) Heart rate turbulence and heart rate variability in patients with mitral valve prolapse. Europace 8(7):515–520