Clinical application of the QRS-T angle for the prediction of ventricular arrhythmias in patients

with the Fontan palliation

Tuong-Vi Tran1 and Daniel Cortez, MD1,2,3

1 University of Colorado School of Medicine, Aurora, USA

2 Electrophysiology, Penn State Milton S. Hershey Medical Center

3 Clinical Sciences, Lund University, Lund, Sweden

Correspondence:

Daniel Cortez, MD

Electrophysiology fellow

9 Tiffany Building,

Hershey, PA 17033

Word count: 4257

1

Abstract

Fontan palliation patients are at risk for ventricular arrhythmias post-operatively. This study aimed to

evaluate whether differences in the spatial QRST angle can reliably predict ventricular arrhythmias in

patients who had undergone Fontan palliation. 

A total of 117 patients who had Fontan palliation and post-Fontan catheterization were included.

Ventricular arrhythmias were identified in 9 patients. Measurements of ECG parameters including QRS

vector magnitude, QRS duration, corrected QT interval, and spatial peaks QRS-T angles were performed,

and compared between those with and without ventricular arrhythmias. 

The only ECG parameter to distinguish those with versus those without VA was the SPQRS-T angle

(p<0.001), which at a cut-off value of 102.9 degrees gave sensitivity, specificity, positive and negative

predictive values of 100.0%, 57.0%, 17.6% and 100.0%, respectively. Only the spatial peaks QRS-T

angle differentiated those with and without ventricular arrhythmia development with a univariate HR

1.237 (95% CI 1.021 to 1.500) and a multivariate HR of 1.032 (1.009 to 1.056) when catheter measured

parameters were taken into account.  

In Fontan patients, the spatial peaks QRS-T angle is a significant independent predictors of ventricular

arrhythmias. Clinical usefulness of this parameter remains to be seen and should be tested prospectively.

Key words: Fontan, vectorcardiography, ventricular arrhythmias

2

Introduction

Since its introduction over forty years ago as surgical repair of tricuspid atresia, the Fontan procedure has

gone through multiple revisions to extend its applicability to other univentricular pathologies including

hypoplastic left heart syndrome, double outlet left/right ventricle, heterotaxia, and pulmonary atresia

intact left ventricular septum (1,2). Although yielding excellent survival, children who have undergone a

Fontan palliation remain at increased risk for medical complications as residua and sequalae from the

natural primary disease course, and multiple surgical procedures. These include rhythm abnormalities

such as brady-arrhythmias, sick sinus syndrome, and/or atrial and ventricular arrhythmias, plastic

bronchitis, protein losing enteropathy, stroke and thrombus formation (3-6). Among the morbidities for

patients with the Fontan palliation, arrhythmias were documented as early as the first decade of Fontan

introduction, and continue to be a challenging management objective for cardiologists. Therefore,

predicting the development of arrhythmias is important, both for management decisions, and patient

consultation. Several tools have been introduced to try to predict arrhythmias including heart rate

variability9, and signal-averaged P wave duration (7-11). In Fontan patients, no single

electrocardiographic predictor has shown prognostic value for identifying those at risk for ventricular

arrhythmias. 

The spatial peaks QRS-T (SPQRS-T) angle, defined as the angle between the directions of ventricular

depolarization and repolarization in 3-dimensional space, has important diagnostic utility for ventricular

arrhythmias in patients with HCM, ischemic heart disease, and congenital heart diseases (12-18).

However, to date, no study has demonstrated its prognostic value in patients with the Fontan palliation. 

Furthermore, in one form of congenital heart disease, tetralogy of Fallot, the QRS vector magnitude

(QRSvm) has demonstrated significant ability to identify those at risk for ventricular arrhythmias peri-

operatively, independent of magnetic resonance imaging measured ventricular volumes or gadolinium

enhancement (19,20).

We predict that the spatial QRS-T angle and the depolarization vector magnitude will have prognostic

value for identification of ventricular arrhythmias in patients with the Fontan palliation. 

3

Patients and Methods

Study population

This study was approved by the Institutional Review Boards at the University of Colorado. 

A blinded retrospective analysis was performed of 117 patients with history of Fontan palliation and post-

Fontan cardiac catheterization from 1997-2015 at University of Colorado Hospital systems (including the

Children’s Hospital of Colorado) were reviewed. All 117 patients met inclusion criteria for cardiac

catheterizations after Fontan palliation as well as electrocardiogram performed within 30 days prior to

cardiac catheterization. Only patients with interpretable ECG’s with adequate baseline measurements

were included. Twenty seven patients either did not have an ECG or an interpretable ECG within 30 days

of the cardiac catheterization and thus were excluded. Since no patients had arrhythmias at baseline, no

patients were on anti-arrhythmics at the time of ECG assessment.  

Patients who had sustained spontaneous ventricular arrhythmias (VA, as determined by ECG, Holter,

exercise stress test, internal cardioverter-defibrillator, pacemaker or by telemetry monitoring) were

identified as well as those with non-sustained VA. Sustained spontaneous VA is defined as 30 seconds or

more or those that were associated with symptoms. Non-sustained VA is defined as 3 beats or more of

VA. For these patients, the arrhythmia had to have occurred after their first post-Fontan cardiac

catheterization.

Comparisons were performed between Fontan palliation patients with spontaneous sustained/non-

sustained VA versus those without VA. 

Electrocardiograms

Sinus rhythm ECGs (Phillips, NV, USA) were performed at 25mm/sec speed with 10mm/mV for limb

and precordial leads. ECGs were analyzed within 30 days prior to the first post-Fontan palliation cardiac

catheterization. Sokolow-Lion left ventricular hypertrophy (LVH) and right ventricular hypertrophy

4

(RVH) were defined as the S-wave in V1 plus the R-wave in V5 sum of 3.5mV or greater and the R-wave

in V1 plus the S-wave in V5 sum of 1.05mV or greater. Measurements of the QRS vector magnitude

(QRSvm, millivolts), QRS duration (QRSd, milliseconds), corrected QT interval (QTc, milliseconds), and

spatial peaks QRS-T angles (SPQRS-T angle, degrees) were performed. The QRSvm was calculated as

the square root of the sum of the squared QRS-wave magnitudes in leads V6, II and one half of the QRS-

wave amplitude in V2 (QRSV6^2+QRSII^2+(0.5*QRSV2)^2,) based on the QRS-wave magnitude as

defined by the visually transformed Kors’ Quasi-orthogonal method, as previously described and as

previously used in congenital heart disease patients (19,20,21,22). The Spatial peaks QRS-T angle is

defined as the angle between the maximum depolarization and maximum repolarization vectors in 3-

dimensional space, as calculated by the Kors’ regression-related method via the visual estimation method

as previously described (21,22). Spatial QRS-T and QRS vector magnitudes are shown in Figure 1.

Calculations of the above parameters are shows in Figure 2. All measures were assessed utilizing digital

calipers.  

Statistics

Data was assessed for normality using Shapiro-Wilk testing. Non-normally distributed continuous data

are presented as median and interquartile ranges (1st and 3rd quartiles), while normally distributed data is

presented as mean ± standard deviation.  Student T-testing, Mann-Whitney U-testing and contingency

table testing were used to identify significant differences between groups. Significant p-values were

<0.05. Relative risks were calculated to estimate risk for parameters identified as significantly different

by comparative analysis. Survival analysis utilizing Kaplan Meijer curves and Cox regression were

performed for significant parameters. Pearson and Spearman correlation coefficients were used as

appropriate for parametric and non-parametric data. Intra-observer and inter-observer variability were

estimated by intra-class correlation coefficients based on a 10% sample of the population. Repeatability

was performed by VT and DC. Data analysis was performed using SPSS (IBM, Chicago, IL).

5

Results

A total of 117 Fontan patients were included in the study. Nine patients eventually developed ventricular

arrhythmias, including 4 (44.4%) with sustained ventricular arrhythmias and 5 (55.6%) with non-

sustained VA (2 assessed by intracardiac device measure, 2 by inpatient telemetry in the emergency

room, and 4 by 24-hour or 30-day event monitoring and 1 by exercise stress testing). The median age at

time of Fontan procedure was 2.0 years (interquartile range 2.0 to 3.5 years). A total of 40 (34.2%) patient

had hypoplastic left heart syndrome (HLHS), 33 (28.2%) patients had tricuspid atresia, 8 (6.8%) patients

had atrioventricular septal defect, 12 (10.3%) patients had double outlet right ventricle, 5 (4.2%) patients

had double inlet left ventricle, 7 (6.0%) patients had pulmonary atresia intact ventricular septum, while 12

(10.3%) patients had single ventricles otherwise not classified. Eight three (70.9%) patients had extra-

cardiac conduits while 34 (29.1%) patients had intra-cardiac conduits. Demographic data is otherwise

presented in Table 1.

Ventricular arrhythmia versus no ventricular arrhythmia

Out of the ECG criteria assessed, only the spatial QRS-T angle significantly differentiated those Fontan

patients with eventual development of ventricular arrhythmias. 

The spatial QRS-T angle had median values of 114.0 (109.9 to 118.3) and 92.0 (67.0 to 132.8),

respectively for those with versus those without eventual ventricular arrhythmias (p-value <0.001).

Sensitivity, specificity, positive predictive values and negative predictive values were 100.0%, 57.0%,

17.6%, and 100.0%, respectively with AU ROC of 0.63 (0.55 to 0.71). Relative risk for eventual

ventricular arrhythmias was 2.4 (1.9 to 2.9). The univariate hazard ratio was 1.237 (95% CI 1.021 to

1.500), while the multivariate HR was 1.033 (1.009 to 1.056) when mean PAP and oxygen extraction

were taken into account.

6

Lower ventricular pulmonary artery pressures and oxygen saturations were present in those with eventual

ventricular arrhythmias, but did not show significant hazard ratios in multivariate analyses (Table 1). 

Sustained ventricular arrhythmias versus no sustained ventricular arrhythmias

Only 4 patients had sustained ventricular arrhythmias and all had HLHS. Only the QRSd and SPQRS-T

angle significantly differentiated those with sustained ventricular arrhythmias versus those without

sustained ventricular arrhythmias. QRSd HR 1.075 (95% CI 1.025 to 1.127) and SPQRS-T angle HR

1.075 (95% CI 1.035 to 1.141).

HLHS only

Age nor sex significantly differentiated those patients with HLHS and sustained VA versus those HLHS

without sustained VA. Median time to sustained VA was 8 months (IQR 8 to 24 months), while median

follow-up for those without sustained VA was 12.0 months (IQR 12.0 to 24.0 months). All 4 patients with

sustained VA had fenestrated Fontans while only 28 of those 36 patients without sustained VA had

fenestrations (77.8%,p-value 0.003). Those patients with HLHS and sustained VA had lower pulmonary

saturations with median value 59.0% (IQR 57.0 to 60.0%) and median value 63.0% (IQR 60.0 to 67.0%,

p-value 0.046), respectively for those HLHS with sustained VA versus those without VA without

significant HR. Systemic arterial saturation was also lower in those HLHS patients with sustained VA

versus those without sustained VA with median value of 83.0% (IQR 81.0 to 84.0%) and (88.0%, IQR

86.0 to 91.0%, p-value 0.001), respectively for those HLHS patients with sustained VA versus those

without sustained VA, without significant HR. Of ECG markers only the QRSd and SPQRS-T angle

differentiated those HLHS patients with sustained VA versus those without sustained VA. The QRSd had

a median value of 138.0ms (IQR 132.0 to 143ms) and 98.0ms (88.0 to 102.0ms, p-value 0.010),

respectively for those with sustained VA versus those without sustained VA and HR of 1.056 (95% CI

1.007 to 1.108). The SPQRS-T angle had a median value of 139 degrees (IQR 135.5 to 140.0 degrees)

and 112.6 degrees (IQR 80.7 to 132.1 degrees, p-value 0.009), respectively for those with versus those

without sustained VA with HR of 1.061 (95% CI 1.010 to 1.083).

7

Correlation Coefficients

The Pearson correlation coefficient between the SPQRS-T angle and PAP was 0.14, otherwise no other

electrocardiographic parameters had higher correlations with catheter measured saturations or pressures.

Subgroup Correlations

HLHS

In the HLHS group age significantly correlated with the QRSd at 0.711. the QTc at

0.364, PAP at -0.461, 0.405, systemic saturation -0.430, PA saturation -0.028. The SPQRS-T

angle significantly correlated with systemic arterial saturation 0.381, systemic systolic blood

pressure -0.385, the QTc at 0.346 and QRSvm at -0.417. The QRSvm also significantly

correlated with QRSd at -0.392.

DORV

In the DORV group, age significantly correlated with the QRSd at 0.775 (p-value 0.002),

the QTc at 0.797 (p-value 0.003), the SPQRS-T angle at 0.760 (p-value 0.004) and the QRSvm

at -0.760 (p-value 0.004). The PAP significantly correlated with SBP at 0.977 (p-value <0.001)

and right ventricular EDP at 0.750 (p-value 0.005). Systemic systolic BP correlated significantly

with right ventricular EDP at 0.794 (p-value 0.002). The QRSd significantly correlated with

systemic and pulmonary artery saturations at 0.627 (p-value 0.029) and 0.726 (p-value 0.041).

The SPQRS-T angle also correlated with PA saturation at 0.726 (p-value 0.041), QRSd at 0.981

(p-value <0.001), and the QTc at 0.677 (p-value 0.016). The QRSvm significantly correlated

with pulmonary and systemic artery saturations at -0.849 (p-value <0.001) and -0.726 (p-value

0.041), as well as QRSd at -0.612 (p-value 0.035), the QTc at -0.677 (p-value 0.016).

AVSD

In AVSD patients, the QRSvm correlates significantly with PAP (-0.752), systemic

saturation (-0.718) and with QRSd (-0.718).

8

Repeatability

Intra-class correlation coefficients for the SPQRS-T angle for inter-observer and intra-observer variability

were 0.91 and 0.93, respectively, based on a 10% sample of the population. Intra-class correlation

coefficients for the QRSvm for inter-observer and intra-observer variability were 0.90 and 0.91,

respectively, based on 10% sample of the population.

Discussion

Main Result 

Our study indicates that the SPQRS-T angle, derived from 12-lead ECG, may be clinically useful to

predict ventricular arrhythmias in pediatric Fontan patients. A widened spatial QRS-T angle

independently indicates increased risk of arrhythmia development regardless of baseline hemodynamics.

Similar to previous studies utilizing electrocardiographic parameters to identify arrhythmia risk, the

negative predictive value of this parameter seems most helpful (7,10,13, 18-20). The main aim of this

study was to introduce the idea and determine if the SPQRS-T angle may have clinical utility. The study

demonstrates that this may be the case, however, prospective prediction needs further validation. If

prospectively determined, this may be a useful screening test to identify which patients may need further

arrhythmia assessment via outpatient monitoring such as Holter or event monitoring. Thus if

independently reproduced, this may provide the first ECG tool which may be useful for risk stratification

in Fontan patients at risk for ventricular arrhythmias. Specifically in HLHS patients, sustained ventricular

arrhythmia risk seems to be conveyed by a longer QRSd and a higher SPQRS-T angle, with the former

being readily available on any ECG, with the latter freely available online (22).

Spatial peaks QRS-T angle

The SPQRS-T angle integrates the heterogeneity in the myocardial action potential duration and myocyte

morphology local and throughout the heart. A widened SPQRS-T angle has been previously observed in

patients with poor myocardial performance (23) and with myocardial infarction (12,24). It was also

9

associated with cardiac mortality (15, 23-28), systemic hypertension (29), left ventricular hypertrophy and

diabetes (13,16-18,23). These pathologies directly and/or indirectly result in inhomogeneity of ventricular

repolarization. Thus, it is expected that some degree of dispersion of repolarization or depolarization-

repolarization vector differential may exist in a patient with a single ventricle over time, contributing to

the increased risk of ventricular arrhythmias from either early after depolarizations or late after

depolarizations. Similar to tetralogy of Fallot, where right ventricular enlargement occurs, many of our

patients had large spatial QRS-T angles. However, dissimilar to tetralogy of Fallot, the spatial peaks

QRS-T angle in our study was able to differentiate those with development of ventricular arrhythmias.

Although not in the aims of our project, we speculate that perhaps some ventricular remodeling such as

abnormal thickening takes place in Fontan palliation patients (at least those with systemic right

ventricles), that places them at higher risk for ventricular arrhythmias. Another population of patients

with hypertrophy, those with hypertrophic cardiomyopathy, were observed to have higher spatial peaks

QRS-T angles with direct association to ventricular arrhythmias. However, Fontan palliation patients

typically do not have such myocardial disorganization as those with hypertrophic cardiomyopathy (13,16-

18).  

QRS vector magnitude

Dissimilar to studies in tetralogy of Fallot, the QRSvm did not predict future ventricular arrhythmia risk

in patients with the Fontan palliation (19,20). Therefore, repolarization heterogeneity or differentiation

from ventricular depolarization likely plays a greater role, given that the spatial and frontal plane QRS-T

angles did differentiate those Fontan patients with eventual ventricular arrhythmias from those who did

not. Unfortunately, this parameter has not been tested in other types of congenital heart disease, thus

normal values in congenital heart disease are not well studied. 

QRS duration and QTC 

10

QRS prolongation was previously shown to be associated with ventricular arrhythmias after Fontan (30),

and its ability to predict arrhythmogenicity in right ventricular dysplasia was also reported (31). We did

not identify any association between QRSd and ventricular arrhythmia in our whole Fontan cohort. As

only a measure of time-lapse in depolarization, QRSd is limited in its ability to measure repolarization

heterogeneity, which appears to play a role in the risk development of ventricular arrhythmia in Fontan

palliation.

Mean pulmonary artery pressure and oxygen extraction

Although mainly used as correlation measures, the mean pulmonary artery pressure, measured during the

first post-Fontan catheterization, was found to be a significant univariate and multivariate parameter for

development of VA. A lower mean PAP was associated with higher spatial QRS-T angle by correlation

coefficient and was significantly associated with development of VA. Perhaps a lower mean PAP allows

more flow into the systemic ventricle and thus would not be protective in the eventual increasing end-

diastolic pressure and overall systemic function. Unfortunately due to the retrospective nature of this

study, that was unable to be fully evaluated. It is of interest, and warrants further evaluation. 

Sustained VA and HLHS

The only 4 patients with sustained VA had HLHS. These patients were identified by the QRSd

and the SPQRS-T angle with those parameters yielding the only significant HR’s. The QRSd, similar to

previous cohort of Fontan patients, did identify those at risk for sustained VA’s, however in the

referenced study, non-sustained VA was included in that cohort (30). The SPQRS-T angle has not been

tested, specifically in HLHS patients, thus no comparison to previous studies in available.

Limitations

The retrospective nature of this study is a limitation and may not have been all inclusive for ventricular

arrhythmias which may have occurred while not monitored. In addition, the small size of patients who

11

had sustained ventricular arrhythmias limits the interpretation of the resultsed the ability to assess for

significant differences between them and the rest of between them and the rest of the cohort. Furthermore,

not all patients with Fontan palliation were evaluated, only those with cardiac catheterizations, such that

measures could be compared between hemodynamics and electrical conduction abnormalities. ECG at

one time point may not be reflective of a dynamic disease process, such as congenital heart diseases.

Furthermore, since electrolytes were not checked routinely on the day of the ECG, changes in

repolarization/depolarization based on electrolyte changes were not able to be assessed. Temporal

changes in vectorcardiographic parameters may impact the association with ventricular arrhythmia risk

over longer periods of follow-up. Other limitations include that transformation methods were used, due to

lack of Frank lead vectorcardiograms, and that baseline vectorcardiograms are also not well quantified in

patients who are status post-Fontan palliation. Furthermore, during follow-up, a number of confounders

might have arisen, which might not have been accounted for, therefore our results should be viewed in

light of this information. 

Conclusion 

The spatial QRS-T angle has promising utility in the prediction of ventricular arrhythmias in our Fontan

cohort along with identification of sustained ventricular arrhythmias in HLHS patients, specifically. Since

the treatment of arrhythmias in patients with Fontan palliation is difficult; the ability to rule out

arrhythmia development therefore becomes crucial in the management of these patients. If independently

reproduced prospectively, the spatial QRS-T angle have significant potential to aid in the early detection

and management of ventricular arrhythmias. 

Compliance with ethical standards

No funding was given for this study.

Conflicts of interest: no conflicts of interest to disclose.

12

Ethical approval: all procedures performed in studies were in accordance with ethical standards of the

institutional research committee and with the 1964 Helsinki declaration and its later amendments.

References

1. Fontan F, Baudet E (1971) Surgical repair of tricuspid atresia. Thorax 26:240-248

2. Gale AW, Danielson GK, McGoon DC, et al (1979) Modified Fontan operation for univentricular

heart and complicated congenital lesions. J Thorac Cardiovasc Surg 78:8310838

3. Tweddell JS, Nersesian M, Mussatto KA, et al (2009) Fontan palliation in the modern era:

Factors impacting mortality and morbidity. Ann Thorac Surg 88:1291-9

4. Peters NS, Sommerville J (1992) Arrhythmias after the Fontan procedure. Br Heart J 68:199-204

5. Deal BJ. Late arrhythmias following Fontan surgery (2012) World J Pediatr Congenit Heart Surg

3:194-200

13

6. Mertens L, Hagler DJ, Somerville J (1998) Protein-losing enteropathy after the Fontan operation:

an international multicenter study. PLE study group. J Thorac Cardiovasc Surg 115:1063-1073

7. Durongpisikul K, Porter CJ, Cetta F, et al (1998) Predictors of early- and late-onset

supraventricular tachyarrhythmias after Fontan operation. Am Heart Assoc 98:1099-1107

8. Gelatt M, Hamilton RM, McCrindle BW, et al (1994) Risk factors for atrial tachyarrhythmias

after the Fontan operation. J Am Coll Cardiol 24:1735-1741

9. Rydberg A, Karlsson M, Hörnsten R. (2008) Can analysis of heart rate variability predict

arrhythmia in children with Fontan circulation? Pediatr Cardiol 29:50-55

10. Tuzcu V, Ozkan B, Sullivan N, et al (2000) P wave signal-averaged electrocardiogram as a new

marker for atrial tachyarrhythmias in postoperative Fontan patients. J Am Coll Cardiol 36:602-

607

11. Yamada T, Fukunami M, Shimonagata T, et al (2001) Identification of sinus node dysfunction by

use of P-wave signal-averaged electrocardiograms in paroxysmal atrial fibrillation: A prospective

study. Am Heart J 142:286-293 

12. Voulgari C, Pagoni S, Tesfaye S, et al (2013) The Spatial QRS-T Angle: Implications in Clinical

Practice. Current Cardiology Review 9:197-210

13. Cortez D, Graw S, Mestroni L (2016) In hypertrophic cardiomyopathy, the spatial peaks QRS-T

angle identifies those with sustained ventricular arrhythmias. Clin Cardiol 39:459-463

14. Rautaharju PM, Prineas RJ, Wood J. et al (2007) Electrocardiographic predictors of new-onset

heart failure in men and in women free of coronary heart disease (from the Atherosclerosis in

Communities [ARIC] study). Am J Cardiol 100:1437-1441

15. Tamazaki T, Froelicher VG, Myers J, et al (2005) Spatial QRS-T angle predicts cardiac death in a

clinical population. Heart Rhythm 2:73-78

16. Poplock Potter SL, Holvqvist F, Platonov PG, et al (2010) Detection of hypertrophic

cardiomyopathy is improved when using advanced rather than strictly conventional 12-lead

electrocardiogram. J Electrocardiol 43:713-718

14

17. Cortez D, Sharma N, Cavanagh J, et al (2015) The spatial QRS-T angle outperforms the Italian

and Seattle ECG-based criteria for detections of hypertrophic cardiomyopathy. J Electrocardiol

48:826-833 

[18.] Cortez D, Sharma N, Cavanaugh J, et al (2016) Lower spatial QRS-T angle rules out sustained

ventricular arrhythmias in children with hypertrophic cardiomyopathy. Cardiol Young [Epub

ahead of print]Graw S, Mestroni L (2016) In hypertrophic cardiomyopathy, the spatial peaks

QRS-T angle identifies those with sustained ventricular arrhythmias. Clin Cardiol [E-pub ahead

of print, May 13, 2016]

18.[19.] Cortez D, Ruckdeschel E, McCanta AC, et al (2015) Vectorcardiographic predictors of

ventricular arrhythmia inducibility in patients with tetralogy of Fallot. Journal of Electrocardiol

48:141-4

19.[20.] Cortez D, Barham W, Ruckdeschel E, et al (2016) Non-invasive predictors of ventricular

arrhythmias in patients with tetralogy of Fallot undergoing pulmonary valve replacements. JACC

Electrophysiol [E-pub, ahead of print].

20.[21.] Kors JA, van Herpen G, Sittig AC, et al (1990) Reconstruction of the Frank vectorcardiogram

from standard electrocardiographic leads: diagnostic comparison of different methods. Eur Heart

J 11:1083-1092.

21.[22.] Cortez D, Sharma N, Devers C, Devers E, Schlegel TT (2014) Visual Transform applications

for estimating the spatial QRS-T angle from the conventional 12-lead ECG: Kors is still most

Frank. J. Electrocardiol 47:12-19

22.[23.] Voulgari C, Tentolouris N, Moyssakis I, et al (2006) The Spatial QRS-T angle: association

with diabetes and left ventricular performance. Eur J Clin Invest 36:608-613

23.[24.] Rautaharju PM, Ge S, Nelson JC, et al (2006) Comparison of mortality risk for

electrocardiographic abnormalities in men and women with and without coronary heart disease

(from the Cardiovascular Health Study). Am J Cardiol 97:309-315

15

24.[25.] Rautaharju PM, Kooperberg C, Larson JC, et al (2006) Electrocardiographic predictors of

incidence congestive heart failure and all-cause mortality is postmenopausal women: the

Women’s Health Initiative. Circulation 113:481-489

25.[26.] Rautaharju PM, Kooperbert C, Larson JC, et al (2006) Electrocardiographic abnormalities

that predict coronary heart disease events and mortality in postmenopausal women: the Women’s

Health Initiaive. Circulation 113:473-480

26.[27.] Rautaharju PM, Prineas RJ, Wood J, et al (2007) Electrocardiographic predictors of new-

onset heart failure in men and in women free of coronary heart disease (from the Atherosclerosis

in Comminites [ARIC] Study). Am J Cardiol 100:1437-1441

27.[28.] Kardys I, Kors JA, Van der Meer I, et al (2003) Spatial QRS-T angle predicts cardiac death in

a general population. Eur Heart J 24:1357–1364

28.[29.] Dilaveris P, Gialafos E, Pantazis A et al (2001) The spatial QRS-T angle as a marker of

ventricular repolarisation in hypertension. J Hum Hypertens 15:63–70

29.[30.] Ghelani SJ Keenan EM, Mah D, et al. Prolonged QRS, mechanical dyssynchrony and

ventricular dilation are associated with ventricular arrhythmias late after the Fontan (2016) J Am

Coll Cardio 67:936.

30.[31.] Peters S, Trummel M. Diagnosis of Arrhythmogenic Right Ventricular Dysplasia-

Cardiomyopathy: Value of Standard ECG Revisited (2003) Ann of Noninvasive

Electrocardiology 8:238-245.

16

Figure 1: Spatial QRS-T angle and QRS wave vector magnitude

17

Figure 2: Calculation of spatial QRS-T angle and QRS vector magnitude.

18

Figure 3 Kaplan-Meijer curves for the spatial QRS-T angle for ventricular arrhythmia identification, p-

value<0.001.

19

Table 1: Fontan patients with eventual development of ventricular arrhythmias versus those who do not

develop ventricular arrhythmias. Variables and associated p-values presented. 

  Ventricular arrhythmias (9) No ventricular arrhythmias (108) p-valueage at catheterization/EKG (years) 9.0 (3.0 to 9.0) 0.8 (0.2 to 3.0) 0.758Age at Fontan (years) 5.0 (1.6 to 6.0) 2.8 (2.0 to 3.5) 0.176Time from ECG/cath (months) to VA 8.0 (8.0 to 108.0) N/A N/Amale gender (%) 5 (55.6%) 71 (65.7%) 0.718syndromic (%) 0 (0.0%) 8 (7.4%) 0.051Syncope (%) 0 (0.0%) 9 (8.3%) 0.065Hypoplastic left ventricle (%) 4 (44.4%) 36 (33.3%) 0.489Tricuspid atresia (%) 4 (44.4%) 29 (26.8%) 0.268Single Ventricle (%) 1 (11.1%) 11 (9.4%) 1.000Atrioventricular septal defect (%) 0 (0.0%) 8 (7.4%) 1.000Double out let right ventricle (%) 0 (0.0%) 12 (11.1%) 0.595Double inlet left ventricle (%) 0 (0.0%) 5 (4.6%) 1.000Pulmonary atresia IVS (%) 0 (0.0%) 7 (6.5%) 1.000Extracardiac conduit (%) 8 (88.9%) 75 (69.4%) 0.282Intracardiac conduit (%) 1 (11.1%) 33 (30.6%) 0.282Fenestration (%) 8 (88.9%) 75 (69.4%) 0.282Protein losing enteropathy (%) 4 (44.4%) 18 (16.7%) 0.063

20

Plastic Bronchitis (%) 0 (0.0%) 4 (3.7%) 1.000Mean PAP (mmHg) 10.9 ± 2.0 14.3 ± 3.9 <0.001ventricular EDP (mmHg) 6.4 ± 2.7 7.9 ± 3.8 0.152Systolic blood pressure (mmHg) 77.6 ± 6.2 79.1 ± 10.9 0.506Pulmonary artery saturation (%) 65.9 ± 6.6 63.6 ± 8.6 0.343Systemic saturation (%) 86.3 ± 3.5 86.4 ± 5.3 0.936oxygen extraction (%) 20.4 ± 3.5 25.4 ± 15.3 0.012Pre-excitation 0 (0.0%) 4 (3.7%) 1.000Sokolow-Lion RVH 4 (44.4%) 62 (57.4%) 0.501Sokolow-Lion LVH 0 (0.0%) 5 (4.6%) 1.000PR duration (ms) 128.3 ± 19.3 126.0 ± 19.7 0.740QRS duration (ms) 107.4 ± 30.5 90.7 ± 18.8 0.142Corrected QT interval (ms) 457.7 ± 63.6 447.4 ± 40.2 0.644Spatial peaks QRS-T angle (deg) 114.0 (109.9 to 118.3) 92.0 (67.0 to 132.8) <0.001QRS vector magnitude (mv) 2.8 ± 1.6 2.1 ± 0.7 0.197

*ECG-electrocardiogram, PAP-pulmonary artery mean pressure, EDP-ventricular end-diastolic pressure,

RVH-right ventricular hypertrophy, LVH-left ventricular hypertrophy.

Table 2: Sensitivity, specificity, positive and negative predictive values (PPV and NPV), area under the

receiver operating characteristic curves (AU ROC), as well as relative risks (RR) for the spatial peaks

QRS-T angle, comparing those with versus those without ventricular arrhythmia development.

  Angle (degree) Sensitivity (%) Specificity (%) PPV (%) NPV (%) AU ROC (95% CI)SPQRS-T angle 102.9 100.0% 57.0% 17.6% 100.0% 0.63 (0.55 to 0.71)

21

Table 3: Hazard ratios for ventricular arrhythmias based on cut-off values for QRS-T angles.

  Univariate Hazard ratio  p-valueSPQRS-T angle 1.237 (1.021 to 1.500) 0.001Mean PAP 0.531 (0.353 to 0.798) 0.002Oxygen extraction 0.964 (0.894 to 1.404) 0.350

Multivariate Hazard ratioSPQRS-T angle 1.033 (1.009 to 1.056) 0.007Mean PAP 0.414 (0.229 to 0.749) 0.004

22

23