- Email: [email protected]
The electrical determinants of increased wall thickness and mass in left ventricular hypertrophy
Journal Pre-proof The electrical determinants of increased wall thickness and mass in left ventricular hypertrophy
Maren Maanja, Todd Schlegel, Rebecca Kozor, Magnus Lundin, Björn Wieslander, Timothy C. Wong, Erik Schelbert, Martin Ugander PII:
S0022-0736(19)30522-9
DOI:
https://doi.org/10.1016/j.jelectrocard.2019.09.024
Reference:
YJELC 52945
To appear in:
Journal of Electrocardiology
Please cite this article as: M. Maanja, T. Schlegel, R. Kozor, et al., The electrical determinants of increased wall thickness and mass in left ventricular hypertrophy, Journal of Electrocardiology(2019), https://doi.org/10.1016/j.jelectrocard.2019.09.024
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2019 Published by Elsevier.
Journal Pre-proof
The electrical determinants of increased wall thickness and mass in left ventricular hypertrophy
Maren Maanja MD1, Todd Schlegel MD1,2, Rebecca Kozor MBBS PhD3,4, Magnus Lundin MD1, Björn Wieslander MD PhD1, Timothy C. Wong MD MS5, Erik Schelbert MD MS5,
1
ro of
Martin Ugander MD PhD1,4
Department of Clinical Physiology, Karolinska Institutet, and Karolinska University
Hospital, Stockholm, Sweden Nicollier-Schlegel SARL, Trélex, Switzerland
3
Department of Cardiology, Royal North Shore Hospital, Sydney, Australia
4
The Kolling Institute, Royal North Shore Hospital, Northern Clinical School, Sydney
re
-p
2
Department of Medicine, University of Pittsburgh Medical Center, Pittsburgh, PA, USA
na
5
lP
Medical School, University of Sydney, Sydney, Australia
Jo
ur
Correspondence Martin Ugander, MD, PhD, Professor University of Sydney Kolling Building, Level 12, Room 612017 Royal North Shore Hospital St Leonards, NSW 2065, Australia E-mail: [email protected] Phone: +61481134220 Manuscript word count: 3138 Data from the current manuscript have been presented in part at: MALT meeting, 04/18/18 – 04/21/18, Uppsala, Sweden Society for Cardiovascular Magnetic Resonance Imaging Congress 2019, Seattle, WA, USA
Journal Pre-proof
Abstract Introduction: Left ventricular hypertrophy (LVH), defined as an increased left ventricular mass (LVM), can manifest as increased wall thickness, ventricular dilatation, or both. Existing LVH criteria from the electrocardiogram (ECG) have poor sensitivity. However, it is unknown whether changes in wall thickness and mass, respectively, can be separately detected by the ECG.
ro of
Methods: Patients undergoing cardiovascular magnetic resonance and resting 12-lead ECG were included. Exclusion criteria were clinical confounders that might influence the ECG,
-p
including myocardial scar. Advanced ECG (A-ECG) analysis included conventional ECG
re
measures and amplitudes, derived vectorcardiographic and polarcardiographic measures, and singular value decomposition of waveform complexity. A-ECG scores for 1) increased LVM
lP
index (LVMI), and 2) increased global wall thickness index (GTI) beyond the upper limit of
na
normal in healthy volunteers, respectively, were derived using multivariable logistic regression. The area under the curve (AUC) and its bootstrapped confidence interval (CI) for
ur
each score were compared to those of Cornell voltage, Cornell product, and Sokolow-Lyon
Jo
voltage conventional ECG-LVH criteria.
Results: Out of 485 patients (median [interquartile range] age 51 [38-61] years, 54% female), 51 (11%) had increased LVMI and 65 (13%) had increased GTI. The A-ECG scores for increased LVMI (AUC [95% CI] 0.84 [0.78-0.90]), and increased GTI (0.80 [0.74-0.85]) differed, and had a higher AUC than the conventional ECG-LVH criteria (p<0.001 for all).
Conclusions: Increased LVMI differed from increased GTI in its electrophysiological manifestation by A-ECG. New A-ECG scores outperform conventional ECG criteria for LVH in determining increased LVMI and GTI, respectively.
2
Journal Pre-proof Keywords: Electrocardiography; Left Ventricular Hypertrophy; Magnetic Resonance Imaging
Highlights
We developed multivariable logistic scores using conventional and advanced ECG measures for diagnosing increased left ventricular mass and increased global wall
ro of
thickness, respectively. Performance of the logistic advanced ECG scores was higher than that of conventional ECG criteria for left ventricular hypertrophy for detecting increased mass and wall
Different ECG measures were selected into the advanced ECG score for detecting
re
-p
thickness.
Increased mass and increased wall thickness differ in their electrophysiological
ur
na
manifestation by advanced ECG.
Jo
lP
increased mass and wall thickness, respectively.
3
Journal Pre-proof Abbreviations Advanced Electrocardiography
CMR
Cardiovascular Magnetic Resonance
ECG
Electrocardiogram
ECV
Extracellular Volume fraction
GLS
Global Longitudinal Strain
GTI
Global wall Thickness Index
LGE
Late Gadolinium Enhancement
LV
Left Ventricle
LVER
Left Ventricular Electrical Remodeling
LVH
Left Ventricular Hypertrophy
LVMI
Left Ventricular Mass Index
VCG
Vectorcardiogram
Jo
ur
na
lP
re
-p
ro of
A-ECG
4
Journal Pre-proof
Introduction Conventional electrocardiographic (ECG) criteria for detecting left ventricular (LV) hypertrophy (LVH) including the Cornell voltage criteria, the Cornell voltage product criteria, and the Sokolow-Lyon criteria, all lack sensitivity (1). LVH is defined as an increased LV mass (LVM), but depending on the relation between mass and LV end diastolic volume (LVEDV), LVH can be classified into three patterns (2): 1) concentric remodeling, defined as
ro of
a normal LV mass with an increased wall thickness; 2) eccentric hypertrophy, defined as an increased mass with an increase in volume, yielding a preserved normal wall thickness; and 3)
-p
concentric hypertrophy, defined as a concomitant increase in both mass and wall thickness. Importantly, the different patterns have an increased risk of coronary heart disease (3). LVM
re
together with global wall thickness (GT) have recently been shown to be able to classify LVH
lP
into these categories, as illustrated in Figure 1, and this classification has been shown to be of
na
prognostic utility (4).
The ECG holds more proven diagnostic information than is usually considered or visible to
ur
the eye. Specifically, valuable diagnostic information can be gleaned from both the three-
Jo
dimensional vectorcardiogram (5) and QRS-wave and T-wave complexity measures such as eigenvectors obtained through singular value decomposition (6). These measures can be derived from a resting 12-lead ECG by Advanced ECG (A-ECG) analysis (7). By selecting a combination of conventional and A-ECG measures using statistical models, new diagnostic scores can be derived (7). Conventional ECG criteria for LVH are typically based on high QRS amplitudes, which have been attributed to the hypertrophied cardiomyocytes (8). However, changes in LVH are not solely attributed to hypertrophied cardiomyocytes; other complex changes such as focal or diffuse myocardial fibrosis and other pathologies may interact with the electrical propagation (8). Recent developments in cardiovascular magnetic
5
Journal Pre-proof resonance (CMR) imaging makes it possible to assess cardiac mass, volume and function, and to quantitatively characterize myocardial tissue characteristics such as myocardial infarction and diffuse myocardial fibrosis with high precision (9). With CMR as the reference, there are compelling reasons to reassess the role of the ECG in detecting the electrical changes in the different types of LVH.
It is unknown how wall thickness and mass, respectively, are manifested electrically in the
ro of
advanced ECG in LVH. Therefore, the aims of this study were to 1) create advanced ECG scores for increased LVM and increased GT, respectively, and 2) to compare these scores to conventional criteria for LVH. We hypothesized that A-ECG can be used to differentially
-p
detect increased LVM indexed to body surface area (LVMI) compared to increased GT
re
indexed to body surface area (GTI), and that the scores can outperform conventional ECG
lP
criteria for LVH.
ur
Study population
na
Methods
Jo
Patients were identified from a prospectively acquired database of patients referred for a clinical cardiovascular magnetic resonance (CMR) imaging at University of Pittsburgh Medical Center (UPMC; Pittsburgh, PA, USA). The study was approved by the UPMC Institutional Review Board, and all participants provided written informed consent. Inclusion criteria were completion of a contrast-enhanced CMR including measures of LVM and EDV, and an ECG with sinus rhythm acquired within 30 days of CMR. To study electrical consequences of only increased mass and wall thickness, respectively, we excluded the following potential CMR-related confounders that could potentially affect the ECG: previous myocardial infarction or non-ischemic myocardial scar by late gadolinium enhancement
6
Journal Pre-proof (LGE), hypertrophic or dilated cardiomyopathy, amyloidosis, severe valvular heart disease, and significant coronary artery stenosis by stress perfusion CMR. We also excluded the following clinical parameters: previous revascularization, and any prior cardiac surgery; and the following ECG-related parameters: treatment with digitalis, heart rate ≥ 100 beats/min, paced rhythm, atrial fibrillation or flutter, abundant premature atrial or ventricular beats (bigeminy or trigeminy), QRS duration ≥120 ms, fascicular blocks, bundle branch blocks, and inadequate ECG records. Baseline data included age at CMR, sex, body mass index (BMI),
ro of
body surface index (BSA), LVM, LVM indexed to BSA (LVMI), left ventricular end diastolic volume (LVEDV), LVEDV indexed to BSA (LVEDVI), global wall thickness (GT), GT
-p
indexed to BSA (GTI), extracellular volume fraction (ECV), left ventricular ejection fraction
re
(LVEF), global longitudinal strain (GLS), hypertension, diabetes mellitus, tobacco smoking status, treatment with beta-blockers, angiotensin-converter enzyme inhibitors (ACEi),
ur
Subgroups
na
reflected in their medical record.
lP
angiotensin receptor blockers (ARB), aspirin or other antiplatelet, and loop diuretics as
From a database of 1828 patients, 485 patients were eligible for the study. To create the A-
Jo
ECG diagnostic scores for increased LVMI and GTI, respectively, patients were divided into two subgroups: 1) normal and increased LVMI, where increased LVMI was defined as >63.7 g/m2 for females, and >81.6 g/m2 for males, based on the mean + 1.96 standard deviations (SD) in healthy volunteers (4), and 2) normal and increased GTI, where increased GTI was defined as >4.2 mm/m2 for females, and >4.4 mm/m2 for males, based on the mean + 1.96 SD in healthy volunteers (4). Thirty-six patients had missing baseline data for comorbidities and medication. CMR acquisition and analysis Quantification of volumes and dimensions, and late gadolinium enhancement imaging 7
Journal Pre-proof The CMR acquisition and analysis was performed as described previously (10). The CMR images were acquired using a 1.5 T scanner (Magnetom Espree, Siemens Healthcare, Erlangen, Germany) using 32 channel phased array surface coils. The exam included standard breath-held segmented cine imaging with a balanced steady-state free precession sequence. A short-axis stack was acquired and used to measure LV volumes, mass, and ejection fraction in end-diastole and end-systole. Late gadolinium enhanced (LGE) imaging was performed using a phase-sensitive inversion recovery (PSIR) pulse sequence, ten minutes after a 0.2 mmol/kg
ro of
intravenous bolus injection of gadoteridol (Prohance, Bracco Diagnostics, Princeton, NJ, USA). In patients with limited breath-hold capability, a single-shot motion corrected PSIR
-p
sequence was used (11, 12). Typical acquisition parameters were field of view 360x270 mm2,
re
256x128 mm2, and slice thickness 6 mm.
lP
Quantification of the extracellular volume fraction
ECV was calculated in quantitative T1 maps that were acquired with an ECG-gated single-
na
shot-modified Look Locker inversion recovery sequence (MOLLI). Native T1 maps were
ur
acquired followed by post-contrast T1 maps after the gadolinium bolus injection. To avoid
calculated as:
Jo
partial volume effects, T1 was measured in the middle third of the myocardium. ECV was , where λ =ΔR1myocardium/ΔR1bloodpool and
ΔR1=1/T1postcontrast - 1/T1precontrast (13).For blood ECV calculations a circular region was traced in the middle of the blood pool to avoid partial volume effects by papillary muscles. The final ECV values were averaged from mid-ventricular and basal short axis slices. Hematocrit measures were acquired on the day of CMR scanning. CMR data were analyzed using a Leonardo work station (Siemens Healthcare, Erlangen, Germany).
Quantification of global longitudinal strain
8
Journal Pre-proof GLS, a measure of impaired left ventricular function, was analyzed using the semi-automated tissue feature tracking software (CVi42, Circle Cardiovascular Imaging Inc., Calgary, Canada). Endocardial and epicardial borders were manually traced in 2-, 3- and 4 chamber end-diastolic images. Strain analysis tracings were inspected visually throughout the cardiac cycle, and manual adjustments were performed if traces deviated from myocardial movement. Calculation of global wall thickness index
ro of
Global wall thickness is the mean wall thickness of the whole left ventricle in end-diastole, measured in millimeters, and it was calculated according to previously validated methods as:
-p
GT =0.05+1.60 • LVM 0.83 • LVEDV -0.49 (4).
re
ECG acquisition and analysis
lP
Resting 12-lead ECG data for each subject were collected from the local ECG storage system (MUSE® Cardiology Information System, Version 8.0 SP2, GE Healthcare, IL, USA) and
na
exported into anonymized xml files with coded subject identification numbers. Advanced ECG semi-automatic software developed in-house was used to analyze the xml files. The
ur
following measures were derived and analyzed from the 12-lead ECG: conventional ECG
Jo
measures of scalar durations, amplitudes, and axes; derived vectorcardiographic (VCG) and polarcardiographic (14, 15) measures of three-dimensional spatial directions and magnitudes of the electrical activation pattern, via Kors’ transform (5); and QRS- and T-wave complexity measures by singular value decomposition (7, 16). The following conventional ECG criteria for LVH were analyzed: Cornell voltage criteria, defined as the S wave in lead V3 (SV3) plus the R wave in lead aVL (RaVL) where LVH is defined as >2.8 mV for males and >2.0 mV for females (17); Cornell product criteria, defined as the Cornell voltage times the QRS duration, where LVH is defined as ≥244mV•ms (18);
9
Journal Pre-proof and Sokolow-Lyon criteria, defined as the sum of SV1 plus the larger of the RV5 or RV6, where LVH was defined as a >3.5 mV (19). Statistical analysis Statistical analysis was performed in R version 3.4.1 (R Foundation for Statistical Computing, Vienna, Austria) using packages dplyr (20) for data transformation, and ggplot2 (21) for data visualizations. The A-ECG scores for increased LVMI and GTI, respectively, were derived
ro of
using stepwise forward logistic regression, and continuous A-ECG measures were analyzed. To determine the diagnostic accuracy of the scores, the area under the receiver operating characteristics curve (AUC) was calculated and bootstrapped 2000 times to obtain 95%
-p
confidence intervals (CI). The A-ECG scores that yielded the highest AUC were chosen, and
re
the Youden index (22) was used to optimize cut-offs for the scores’ sensitivity and specificity
lP
to detect increased LVMI and GTI, respectively. A minimum of ten events per incremental AECG parameter was accepted for the scores (23). The diagnostic scores were constructed to
na
show the likelihood of having increased LVMI or GTI, respectively, ranging from 0% to
•100
Jo
ur
100%, calculated from the logistic regression as:
The AUC for the two A-ECG scores were compared to each other and to the AUC for the conventional ECG criteria, respectively, using the conventional cut-offs for LVH as described and the DeLong test (24), and illustrated using forest plots. Pearson’s correlation coefficient was used to evaluate if the A-ECG parameters in the respective scores were interrelated, and expressed as its square (R2). The Kolmogorov-Smirnov test was used to test if data were normally distributed, and differences between subgroups’ baseline data were tested using the chi-square test or Mann-Whitney U test, as appropriate, and described using percentage and
10
Journal Pre-proof median [interquartile range], respectively. A p-value <0.05 was considered statistically significant.
Results We identified 485 patients (age 51 [38-61] years, 54% female), in which 51 (11%) patients had increased LVMI and 65 (13%) had increased GTI. Sixteen (3%) patients had an isolated
ro of
increased LVMI, and 30 (6%) patients had an isolated increased GTI. Four hundred and four patients (83%) had both a normal LVMI and GTI. Based on the number of patients with
-p
increased LVMI and GTI, respectively, a maximum of five and six A-ECG parameters were
re
accepted for the LVMI and GTI scores, respectively. Increased GTI was associated with older age (p=0.002), and a higher LVEF, compared to increased LVMI (p=0.17, p=0.96,
lP
respectively). Baseline characteristics of the study population are presented in Table 1. A total
na
of 34 (7%) patients had missing data on baseline co-morbidity and indication for CMR. A-ECG score versus conventional criteria for increased LVMI
ur
The measures that contributed to the A-ECG score for increased LVMI are presented in Table
Jo
2. They included two derived vectorcardiographic measures, and two conventional ECG measures, specifically: 1) the area of the QRS vector magnitude (mV•s); 2) the QRS loop area outside the left lower quadrant in the left sagittal plane (%); 3) the QTc interval (ms); and 4) the T wave amplitude in lead aVR (µV). The A-ECG score for detecting increased LVMI had a sensitivity of 75%, and a specificity of 81%, and a higher AUC [95% CI] (0.84 [0.78-0.90]) than the Cornell voltage criteria (0.63 [0.57-0.70]), the Cornell product criteria (0.60 [0.550.67]), and the Sokolow-Lyon criteria (0.57 [0.52-0.62]), as shown in Figure 2 (p<0.001 for all). A-ECG score versus conventional criteria for increased GTI
11
Journal Pre-proof The measures that contributed to the A-ECG score for increased GTI are presented in Table 2. They included one measure of QRS wave complexity, two derived vectorcardiographic measures, one conventional ECG measure, and one derived polarcardiographic measure, specifically: 1) the 6th Eigenvalue (singular value) of the QRS; 2) the bulginess of T vector loop (mV); 3) the azimuth of the maximum QRS magnitude in the left sagittal plane (radians); 4) the T amplitude in lead I (µV); and 5) the elevation (latitude) of the polar vector (degrees).
ro of
The A-ECG score for detecting increased GTI had a sensitivity of 71% and a specificity of 80%, and a higher AUC (0.80 [0.74-0.85]) than the Cornell voltage criteria (0.58 [0.54-0.63]),
0.58]), as shown in Figure 1 (p<0.001 for all).
-p
the Cornell product criteria (0.57 [0.53-0.61]), and the Sokolow-Lyon criteria (0.54 [0.51-
re
Electrocardiographic differences between increased LVMI and increased GTI
lP
The two A-ECG scores were also compared to each other. The GTI A-ECG score had a higher AUC than the LVMI A-ECG score for detecting increased GTI [0.80 [0.74-0.85] vs
na
0.71 [0.64-0.78], p=0.02). However, the GTI A-ECG score could also detect increased LVMI
ur
as well as the dedicated LVMI A-ECG score (0.83 [0.88-0.88] vs 0.84 [0.78-0.90], p=0.88,
Jo
see Figure 1. The largest R2 between any two parameters in the respective scores was 0.063, indicating that the parameters in each score were at most weakly associated.
Discussion
The main findings of this study were that, in a consecutive clinical cohort of patients without myocardial scar or other confounders that may interact with the ECG besides LVH, newly constructed A-ECG scores outperformed conventional LVH ECG criteria (Cornell voltage, Cornell voltage product, and Sokolow-Lyon criteria) for determining increased LVMI and GTI, respectively. Furthermore, increased LVMI and GTI differed in their electrocardiographic manifestation by A-ECG.
12
Journal Pre-proof Although the two scores included different measures, the GTI A-ECG score could diagnose high LVMI on par with the dedicated LVMI A-ECG score, while the LVMI A-ECG score could not diagnose high GTI as well as the dedicated GTI A-ECG score. This suggests that on a group level, increased GTI associates with distinct electrocardiographic abnormalities beyond those associated with increased LVMI, but not vice versa. Depending on the relation between increased mass and wall thickness, LVH can be divided into concentric remodeling, eccentric hypertrophy, and concentric hypertrophy. Increased mass includes eccentric
ro of
hypertrophy and concentric hypertrophy, while increased wall thickness includes concentric remodeling and concentric hypertrophy. Thus, as the subgroups overlap, the common LVH
-p
denominator in both scores are patients with concentric hypertrophy and the electrical
re
differences can be found in patients with either concentric remodeling or eccentric hypertrophy, as shown in Figure 1. The accuracy of conventional ECG LVH criteria remained
lP
poor, as expected (1, 25).
na
Previous studies and the advanced ECG score parameters
ur
Essentially all measures that comprised the constituents of the A-ECG scores derived in this study, or corresponding measures, have been previously studied in ventricular hypertrophy.
Jo
For example, in relation to the area of the QRS vector magnitude, it has been shown that an increased area of the QRS improved the diagnosis of hypertrophic hearts (26). In healthy subjects, the largest percentage of the QRS loop area resides in the lower quadrants of the left sagittal plane (27). Furthermore, studies have found correlations between prolonged QTc and LVH in hemodialysis (28) and hypertensive (29) patients. The bulginess, or the distortion, of the T vector loop, also referred to as Tavplan, is the distance from a preferential twodimensional plane to the three-dimensional T loop (30), and the bulginess of the T vector is sensitive to ischemia followed by coronary occlusion (31). In a study where patients with LVH by Sokolow-Lyon index were excluded, patients with hypertension had more
13
Journal Pre-proof pronounced T loop changes than those without hypertension (32). However, hypertension is associated with LVH (33), and LVH could not be ruled out by ECG (32). Regarding the T wave amplitudes in leads aVR and I, hemodialysis patients with upright Twaves in lead aVR had not only poorer outcomes, but also relatively increased LVMI (34). Moreover, T wave inversion in other leads has been shown to correlate with LVH (35). Others have also found that the azimuth of the QRS at maximal magnitude in the sagittal (36)
ro of
horizontal (37) planes was significantly associated with LVH. The latitude of the polar vector was also a constituent of the GTI A-ECG score. In contrast to the vectorcardiogram from which it can be derived, the polarcardiogram also plots the
-p
direction and magnitude of the heart’s electrical vector over time, and in a coordinate system
re
that is spherical (polar) (38). In a previous study, less than half of the patients with LVH had a
lP
polar vector that deviated from the polar vector in normal controls (39). In summary, essentially all (eight of nine) parameters selected by the stepwise forward logistic regression
na
model for diagnosing increased LVMI and GTI have previously been studied in, and found to
ur
be associated with, LVH. The final parameter, the lower-level QRS Eigenvalue that also contributed to the score for diagnosing increased GTI, has not, to our knowledge, previously
Jo
been studied in LVH. However, all lower-level QRS eigenvalues are known to be relatively increased in heart failure (7), of which hypertrophic remodeling is a typical manifestation. Study limitations This study has several limitations. First, its experimental design was optimized for attempting to ascertain the differential electrocardiographic manifestations of different morphological aspects of LVH. In order to optimally isolate the electrical determinants of increased LVMI and GTI, respectively, we excluded patients with other pathologies that may interact with the ECG. Thus, the findings in this study are not applicable to general clinical cohorts wherein other co-morbidities are often present that overlap with LVH. Second, due to a lack of 14
Journal Pre-proof statistical power, the exact electrical differences between the LVH types of concentric remodeling, eccentric hypertrophy, and concentric hypertrophy could not be determined in the current study. However, such studies are justified. Third, a total of 34 (7%) patients had missing data on baseline co-morbidity and indication for CMR. Finally, besides the conventional ECG LVH criteria evaluated in this study, other sets of criteria have been recommended by the American Heart Association (8), but were not specifically studied.
ro of
Clinical applications The present study gives insight into the different electrical properties in LVH related to increased LVMI or GTI. Based on the improved diagnostic performance of the newly
-p
proposed A-ECG scores over conventional ECG criteria for LVH, we suggest that wherever
re
digital 12-lead ECGs are available, these scores could be additionally employed.
lP
Electrophysiology and anatomy are fundamentally different phenomena, and the changes observed on the ECG should preferentially be referred to as left ventricular electrical
ur
working group (40).
na
remodeling, in accordance with the suggestions of the Journal of Electrocardiology LVH
Jo
Conclusions
Increased LVMI and GTI differed in their electrophysiological manifestation by A-ECG. New A-ECG scores outperform conventional ECG criteria for LVH in determining increased LVMI and GTI, respectively. Sources of funding The study was supported in part by grants from the Swedish Research Council, Swedish Heart and Lung Foundation, Stockholm County Council, and Karolinska Institutet. Disclosures
15
Journal Pre-proof Dr. Schlegel is a principal of Nicollier-Schlegel SARL, a company that performs ECG
Jo
ur
na
lP
re
-p
ro of
research consultancy using the software used in the present study.
16
Journal Pre-proof
References 1.
Pewsner D, Jüni P, Egger M, Battaglia M, Sundström J, Bachmann LM.
Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. BMJ. 2007;335(7622):711. 2.
Ganau A, Devereux RB, Roman MJ, de Simone G, Pickering TG, Saba PS, et
al. Patterns of left ventricular hypertrophy and geometric remodeling in essential
3.
ro of
hypertension. J Am Coll Cardiol. 1992;19(7):1550-8. Gupta S, Berry JD, Ayers CR, Peshock RM, Khera A, de Lemos JA, et al. Left
-p
ventricular hypertrophy, aortic wall thickness, and lifetime predicted risk of cardiovascular
4.
re
disease:the Dallas Heart Study. JACC Cardiovasc Imaging. 2010;3(6):605-13. Lundin M, Heiberg E, Nordlund D, Gyllenhammar T, Steding-Ehrenborg K,
lP
Engblom H, et al. Mean wall thickness improved characterization and prognosis in left
na
ventricular hypertrophy [abstract]. In: Proceedings of the 22nd Annual Meeting of the Society
5.
Kors JA, van Herpen G, Sittig AC, van Bemmel JH. Reconstruction of the
Jo
ID#549719.
ur
for Cardiovascular Magnetic Resonance; 2019, Feb 6-9; Bellevue, WA, US. Abstract
Frank vectorcardiogram from standard electrocardiographic leads: diagnostic comparison of different methods. Eur Heart J. 1990;11(12):1083-92. 6.
Rautaharju PM, Kooperberg C, Larson JC, LaCroix A. Electrocardiographic
predictors of incident congestive heart failure and all-cause mortality in postmenopausal women: the Women's Health Initiative. Circulation. 2006;113(4):481-9. 7.
Schlegel TT, Kulecz WB, Feiveson AH, Greco EC, DePalma JL, Starc V, et al.
Accuracy of advanced versus strictly conventional 12-lead ECG for detection and screening
17
Journal Pre-proof of coronary artery disease, left ventricular hypertrophy and left ventricular systolic dysfunction. BMC Cardiovasc Disord. 2010;10:28. 8.
Bacharova L, Ugander M. Left ventricular hypertrophy: The relationship
between the electrocardiogram and cardiovascular magnetic resonance imaging. Ann Noninvasive Electrocardiol. 2014;19(6):524-33. 9.
Karamitsos TD, Francis JM, Myerson S, Selvanayagam JB, Neubauer S. The
ro of
role of cardiovascular magnetic resonance imaging in heart failure. J Am Coll Cardiol. 2009;54(15):1407-24.
Schelbert EB, Piehler KM, Zareba KM, Moon JC, Ugander M, Messroghli DR,
-p
10.
et al. Myocardial Fibrosis Quantified by Extracellular Volume Is Associated With Subsequent
re
Hospitalization for Heart Failure, Death, or Both Across the Spectrum of Ejection Fraction
Kellman P, Larson AC, Hsu LY, Chung YC, Simonetti OP, McVeigh ER, et al.
na
11.
lP
and Heart Failure Stage. J Am Heart Assoc. 2015;4(12).
Motion-corrected free-breathing delayed enhancement imaging of myocardial infarction.
Piehler KM, Wong TC, Puntil KS, Zareba KM, Lin K, Harris DM, et al. Free-
Jo
12.
ur
Magn Reson Med. 2005;53(1):194-200.
breathing, motion-corrected late gadolinium enhancement is robust and extends risk stratification to vulnerable patients. Circ Cardiovasc Imaging. 2013;6(3):423-32. 13.
Arheden H, Saeed M, Higgins CB, Gao DW, Bremerich J, Wyttenbach R, et al.
Measurement of the distribution volume of gadopentetate dimeglumine at echo-planar MR imaging to quantify myocardial infarction: comparison with 99mTc-DTPA autoradiography in rats. Radiology. 1999;211(3):698-708. 14.
Belletti DA, Lyon AF. The initial 0.04 second polar vector of the QRS loop as
an aid in the detection of a myocardial infarction. Jpn Heart J. 1969;10(3):267-72.
18
Journal Pre-proof 15.
Pipberger H. Use of Computers in Interpretation of Electrocardiograms:
Lippincott Williams & Wilkins, Baltimore, MD; 1962. 16.
Batdorf BH, Feiveson AH, Schlegel TT. The effect of signal averaging on the
reproducibility and reliability of measures of T-wave morphology. J Electrocardiol. 2006;39(3):266-70. 17.
Casale PN, Devereux RB, Kligfield P, Eisenberg RR, Miller DH, Chaudhary
ro of
BS, et al. Electrocardiographic detection of left ventricular hypertrophy: development and prospective validation of improved criteria. J Am Coll Cardiol. 1985;6(3):572-80. Molloy TJ, Okin PM, Devereux RB, Kligfield P. Electrocardiographic detection
-p
18.
of left ventricular hypertrophy by the simple QRS voltage-duration product. J Am Coll
Sokolow M, Lyon TP. The ventricular complex in left ventricular hypertrophy
lP
19.
re
Cardiol. 1992;20(5):1180-6.
Wickham H, Francois R, Henry L, Müller K. Dplyr: A Grammar of Data
Manipulation. 2017.
Wickham H. Ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag
New York. 2009.
Jo
21.
ur
20.
na
as obtained by unipolar precordial and limb leads. Am Heart J. 1949;37(2):161-86.
22.
Youden W. Index for rating diagnostic tests. Cancer. 1950;3(1):32-5.
23.
Peduzzi P, Concato J, Feinstein AR, Holford TR. Importance of events per
independent variable in proportional hazards regression analysis. II. Accuracy and precision of regression estimates. J Clin Epidemiol. 1995;48(12):1503-10.
19
Journal Pre-proof 24.
DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two
or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44(3):837-45. 25.
Sjöberg S, Sundh F, Schlegel T, Maynard C, Rück A, Wagner G, et al. The
relationship between electrocardiographic left ventricular hypertrophy criteria and echocardiographic mass in patients undergoing transcatheter aortic valve replacement. J
26.
ro of
Electrocardiol. 2015;48(4):630-6. Okin PM, Roman MJ, Devereux RB, Pickering TG, Borer JS, Kligfield P. Time-
voltage QRS area of the 12-lead electrocardiogram: detection of left ventricular hypertrophy.
Chou TC, Masangkay MP, Young R, Conway GF, Helm RA. Simple
re
27.
-p
Hypertension. 1998;31(4):937-42.
28.
na
Circulation. 1973;48(6):1262-7.
lP
quantitative vectorcardiographic criteria for the diagnosis of right ventricular hypertrophy.
Alonso MAG, Lima VACC, Carreira MAMQ, Lugon JR. Reproducibility and
ur
Reliability Of QTc and QTcd Measurements and Their Relationships with Left Ventricular
29.
Jo
Hypertrophy in Hemodialysis Patients. Arq Bras Cardiol. 2017;109(3):222-30. Porthan K, Virolainen J, Hiltunen TP, Viitasalo M, Väänänen H, Dabek J, et al.
Relationship of electrocardiographic repolarization measures to echocardiographic left ventricular mass in men with hypertension. J Hypertens. 2007;25(9):1951-7. 30.
Badilini F, Maison Blanche P, Fayn J. Relationship between 12-lead ECG QT
dispersion and 3D-ECG repolarization loop. Computers in Cardiology. 1995; 785-8. 31.
Nowinski K, Jensen S, Lundahl G, Bergfeldt L. Changes in ventricular
repolarization during percutaneous transluminal coronary angioplasty in humans assessed by QT interval, QT dispersion and T vector loop morphology. J Intern Med. 2000;248(2):126-36.
20
Journal Pre-proof 32.
Rubulis A, Jensen J, Lundahl G, Tapanainen J, Wecke L, Bergfeldt L. T vector
and loop characteristics in coronary artery disease and during acute ischemia. Heart Rhythm. 2004;1(3):317-25. 33.
Shenasa M, Shenasa H. Hypertension, left ventricular hypertrophy, and sudden
cardiac death. Int J Cardiol. 2017;237:60-3. 34.
Jaroszyński A, Jaroszyńska A, Siebert J, Dąbrowski W, Niedziałek J, Bednarek-
patients. Clin Exp Nephrol. 2015;19(6):1157-64.
Kannel WB, Gordon T, Castelli WP, Margolis JR. Electrocardiographic left
-p
35.
ro of
Skublewska A, et al. The prognostic value of positive T-wave in lead aVR in hemodialysis
ventricular hypertrophy and risk of coronary heart disease. The Framingham study. Ann
Varriale P, Alfenito JC, Kennedy RJ. The vectorcardiogram of left ventricular
lP
36.
re
Intern Med. 1970;72(6):813-22.
37.
na
hypertrophy. Analysis and criteria (Frank Lead system). Circulation. 1966;33(4):569-76. Romhilt DW, Greenfield JC, Estes EH. Vectorcardiographic diagnosis of left
Macfarlane PW, vanOosterom A, Pahlm O, Kligfield P, Janse M, Camm J.
Jo
38.
ur
ventricular hypertrophy. Circulation. 1968;37(1):15-9.
Comprehensive Electrocardiology. London: Springer; 2011. 39.
Pipberger H, Carter T. Analysis of the normal and abnormal vectorcardiogram
in its own reference frame. Circulation. 1962;25:827-40. 40.
Bacharova L, Estes EH, Bang LE, Hill JA, Macfarlane PW, Rowlandson I, et al.
Second statement of the working group on electrocardiographic diagnosis of left ventricular hypertrophy. J Electrocardiol. 2011;44(5):568-70.
21
Journal Pre-proof
Tables Table 1. Baseline characteristics for the two sub-groups.
Characteristics
All
Number, n Female sex, n (%) Age, years
ECV, %
LVEDV, mL
Normal LVMI 485
434
51
263 (54)
230 (53)
33 (65)
51
51
54
(38–61)
(38–61)
(42–65)
27.2
27.2
(25.0–30.0)
(24.9–29.9)
147.0
144.6
LVEDVI, mL/m
73.9 (64.2–87.0)
GLS, %
LVEF, %
LVM, g LVMI, g/m2
J
l a
rn
u o
(123.4–179.9) 2
Increased LVMI
27.9
(122.1–174.1)
(149.0–237.4)
72.8
98.6
(63.5–84.2)
(149.0–237.4)
-17.9
-13.7
(-19.7– -15.4)
(-19.8– -15.9)
(-16.3– -10.1)
62.0
62.0
58.0
(56.0–67.0)
(57.0–67.0)
(45.3–65.0)
100.6
96.7
163.7
(81.0–129.1)
(79.8–121.2)
(132.0–213.4)
49.5
47.5
79.9
-17.6
Increased GTI
p value
420
65
-
0.11
225 (54)
38 (58)
0.46
0.17
50
58
0.002
(37–60)
(49–65)
27.1
27.9
(25.0–29.7)
(24.9–30.9)
148.3
140.7
(125.1–179.9)
(111.0–196.7)
73.8
76.7
(63.3–86.5)
(64.1–90.8)
-17.9
-15.7
(-19.9– -15.8)
(-17.5– -12.2)
62.0
63.4
(56.4–67.0)
(55.0–67.0)
97.0
137.6
(79.8–123.3)
(111.1–184.7)
47.4
70.3
0.23
(25.7–30.3) 205.6
Normal GTI
-
e
r P
f o
o r p
p value
<0.001
<0.001
<0.001
0.002
<0.001
<0.001
0.11
0.22
0.38
<0.001
0.96
<0.001
<0.001
Journal Pre-proof (42.2–60.2)
(41.7–56.1)
(68.4–93.2)
6.8
6.6
8.7
(5.9–7.9)
(5.8–7.5)
(8.1–10.6)
3.4
3.3
4.6
(3.0–3.9)
(3.0–3.7)
(4.1–5.3)
28.3
28.3
28.2
(24.3–34.0)
(24.2–34.1)
(24.4–32.4)
2.0
2.0
1.9
(1.8–2.2)
(1.8–2.2)
(1.8–2.2)
Hypertension
179 (40)
151 (38)
28 (62)
Diabetes mellitus
54 (12)
40 (10)
14 (31)
Current smoker
61 (14)
50 (12)
Ex-smoker
132 (30)
118 (29)
GT, mm
GTI, mm/m
2
2
BMI, kg/m 2
BSA, m
Aspirin or antiplatelet
121 (27) other
162 (36)
8.7
(5.8–7.4)
(8.2–10.1)
3.3
4.6
(2.9–3.6)
(4.5–5.2)
<0.001
f o
0.60
<0.001
<0.001
25.8
(24.4–34.9)
(23.2–30.1)
2.0
1.9
(1.8–2.2)
(1.7–2.1)
0.001
148 (38)
31 (56)
0.008
<0.001
40 (10)
14 (26)
0.001
11 (24)
0.03
51 (13)
10 (18)
0.30
14 (31)
0.81
118 (30)
14 (26)
0.48
100 (25)
21 (47)
0.002
103 (26)
18 (33)
0.31
141 (35)
21 (47)
0.13
137 (35)
25 (46)
0.13
128 (32)
23 (51)
0.01
130 (33)
21 (38)
0.46
l a
rn
u o
J
6.7
28.6
Smoking status, n (%)
ACEi/ARB
(61.3–87.5)
<0.001
o r p
0.27
e
Co-morbidity, n (%)*
Medication, n (%)
(41.7–56.2)
r P
0.004
<0.001
Beta-blockers
151 (34)
Loop diuretics
60 (13)
51 (13)
9 (20)
0.17
51 (13)
9 (16)
0.50
Insulin
39 (9)
28 (7)
11 (24)
<0.001
29 (7)
10 (18)
0.008
Oral hypoglycemic
19 (4)
17 (4)
2 (4)
0.95
14 (4)
5 (9)
0.06
124 (28)
110 (27)
14 (31)
0.60
106 (27)
18 (33)
0.38
Statin
General indication for CMR: known or suspected disease, n (%)*
23
Journal Pre-proof Arrhythmia syncope
or
164 (34)
152 (35)
12 (24)
0.10
152 (36)
12 (18)
0.005
Cardiomyopathy
194 (40)
172 (40)
22 (43)
0.63
164 (39)
30 (46)
0.28
Coronary disease/
208 (43)
185 (43)
23 45)
0.74
181 (43)
27 (41)
0.81
artery
vasodilator stress test Pericardial disease
22 (5)
19 (4)
3 (6)
0.63
Thoracic assessment
15 (3)
14 (3)
1 (2)
0.63
aorta
p e
ro
f o
Co nti
20 (5)
2 (3)
0.54
12 (3)
3 (5)
0.45
nuo us
data are given as number, n (%), or median (interquartile range), and compared between Normal LVMI and Increased LVMI, and Normal GTI
r P
and Increased GTI, respectively, using Chi-square or Mann-Whitney U test. ACEi indicates angiotensin-converting enzyme inhibitors; ARB,
l a
angiotensin II receptor blocker; BMI, body mass index; BSA, body surface area; CMR; cardiovascular magnetic resonance imaging; ECV,
n r u
extracellular volume fraction; LVEDV, left ventricular end-diastolic volume; LVEDVI, LVEDV index; GLS, global longitudinal strain; LVEF, left ventricular ejection fraction; LVM, left ventricular mass; LVMI, LVM index; GT, global wall thickness; GTI, GT index. *Data from 451 (93%) patients.
o J
24
Journal Pre-proof
Table 2. A-ECG parameters in the increased left ventricular mass index (LVMI) and global wall thickness index (GTI) scores. ECG parameter
Z-value
p value
AUC (95% CI)
Increased LVMI score 1)
The area of the QRS vector magnitude (mV•s)
6.0
<0.001
2)
The QRS loop area outside the left lower quadrant in the left sagittal plane (%)
-3.3
<0.001
T wave amplitude in lead aVR (µV)
-3.2
0.001
4)
QTc (ms)
2.7
0.007
ro of
3)
-p
0.84 (0.78-0.90)
Increased GTI score 4.4
2)
The bulginess of T vector loop (mV)
4.2
3)
The azimuth of the maximum QRS magnitude in the left sagittal plane (radians)
3.5
na
T wave amplitude lead I (µV)
5)
The latitude of the polar vector (degrees)
Jo
ur
4)
<0.001
re
6th Eigenvalue (singular value) of the QRS
lP
1)
<0.001
<0.001 0.80 (0.74-0.86)
-3.5
0.002
-2.8
0.006
Journal Pre-proof Figure 1. A schematic figure of how normal and increased left ventricular (LV) mass index and LV global wall thickness index, respectively, can classify left ventricular hypertrophy into three patterns: concentric remodeling, eccentric hypertrophy, and concentric hypertrophy. The LV global wall thickness index is the mean wall thickness of the entire left ventricle as measured by MRI. Cutoffs defining increase beyond normal (not explicitly shown) are the
Jo
ur
na
lP
re
-p
ro of
sex-specific 95% upper limit of normal in healthy volunteers.
26
Journal Pre-proof Figure 2. Area under the curve (AUC) for the increased left ventricular mass index (LVMI) score, increased global wall thickness index (GTI) score, Sokolow-Lyon, Cornell voltage, and
Jo
ur
na
lP
re
-p
ro of
Cornell product for detecting increased LVMI and GTI, respectively.
27
Figure 1
Figure 2
Maren Maanja, Todd Schlegel, Rebecca Kozor, Magnus Lundin, Björn Wieslander, Timothy C. Wong, Erik Schelbert, Martin Ugander PII:
S0022-0736(19)30522-9
DOI:
https://doi.org/10.1016/j.jelectrocard.2019.09.024
Reference:
YJELC 52945
To appear in:
Journal of Electrocardiology
Please cite this article as: M. Maanja, T. Schlegel, R. Kozor, et al., The electrical determinants of increased wall thickness and mass in left ventricular hypertrophy, Journal of Electrocardiology(2019), https://doi.org/10.1016/j.jelectrocard.2019.09.024
This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
© 2019 Published by Elsevier.
Journal Pre-proof
The electrical determinants of increased wall thickness and mass in left ventricular hypertrophy
Maren Maanja MD1, Todd Schlegel MD1,2, Rebecca Kozor MBBS PhD3,4, Magnus Lundin MD1, Björn Wieslander MD PhD1, Timothy C. Wong MD MS5, Erik Schelbert MD MS5,
1
ro of
Martin Ugander MD PhD1,4
Department of Clinical Physiology, Karolinska Institutet, and Karolinska University
Hospital, Stockholm, Sweden Nicollier-Schlegel SARL, Trélex, Switzerland
3
Department of Cardiology, Royal North Shore Hospital, Sydney, Australia
4
The Kolling Institute, Royal North Shore Hospital, Northern Clinical School, Sydney
re
-p
2
Department of Medicine, University of Pittsburgh Medical Center, Pittsburgh, PA, USA
na
5
lP
Medical School, University of Sydney, Sydney, Australia
Jo
ur
Correspondence Martin Ugander, MD, PhD, Professor University of Sydney Kolling Building, Level 12, Room 612017 Royal North Shore Hospital St Leonards, NSW 2065, Australia E-mail: [email protected] Phone: +61481134220 Manuscript word count: 3138 Data from the current manuscript have been presented in part at: MALT meeting, 04/18/18 – 04/21/18, Uppsala, Sweden Society for Cardiovascular Magnetic Resonance Imaging Congress 2019, Seattle, WA, USA
Journal Pre-proof
Abstract Introduction: Left ventricular hypertrophy (LVH), defined as an increased left ventricular mass (LVM), can manifest as increased wall thickness, ventricular dilatation, or both. Existing LVH criteria from the electrocardiogram (ECG) have poor sensitivity. However, it is unknown whether changes in wall thickness and mass, respectively, can be separately detected by the ECG.
ro of
Methods: Patients undergoing cardiovascular magnetic resonance and resting 12-lead ECG were included. Exclusion criteria were clinical confounders that might influence the ECG,
-p
including myocardial scar. Advanced ECG (A-ECG) analysis included conventional ECG
re
measures and amplitudes, derived vectorcardiographic and polarcardiographic measures, and singular value decomposition of waveform complexity. A-ECG scores for 1) increased LVM
lP
index (LVMI), and 2) increased global wall thickness index (GTI) beyond the upper limit of
na
normal in healthy volunteers, respectively, were derived using multivariable logistic regression. The area under the curve (AUC) and its bootstrapped confidence interval (CI) for
ur
each score were compared to those of Cornell voltage, Cornell product, and Sokolow-Lyon
Jo
voltage conventional ECG-LVH criteria.
Results: Out of 485 patients (median [interquartile range] age 51 [38-61] years, 54% female), 51 (11%) had increased LVMI and 65 (13%) had increased GTI. The A-ECG scores for increased LVMI (AUC [95% CI] 0.84 [0.78-0.90]), and increased GTI (0.80 [0.74-0.85]) differed, and had a higher AUC than the conventional ECG-LVH criteria (p<0.001 for all).
Conclusions: Increased LVMI differed from increased GTI in its electrophysiological manifestation by A-ECG. New A-ECG scores outperform conventional ECG criteria for LVH in determining increased LVMI and GTI, respectively.
2
Journal Pre-proof Keywords: Electrocardiography; Left Ventricular Hypertrophy; Magnetic Resonance Imaging
Highlights
We developed multivariable logistic scores using conventional and advanced ECG measures for diagnosing increased left ventricular mass and increased global wall
ro of
thickness, respectively. Performance of the logistic advanced ECG scores was higher than that of conventional ECG criteria for left ventricular hypertrophy for detecting increased mass and wall
Different ECG measures were selected into the advanced ECG score for detecting
re
-p
thickness.
Increased mass and increased wall thickness differ in their electrophysiological
ur
na
manifestation by advanced ECG.
Jo
lP
increased mass and wall thickness, respectively.
3
Journal Pre-proof Abbreviations Advanced Electrocardiography
CMR
Cardiovascular Magnetic Resonance
ECG
Electrocardiogram
ECV
Extracellular Volume fraction
GLS
Global Longitudinal Strain
GTI
Global wall Thickness Index
LGE
Late Gadolinium Enhancement
LV
Left Ventricle
LVER
Left Ventricular Electrical Remodeling
LVH
Left Ventricular Hypertrophy
LVMI
Left Ventricular Mass Index
VCG
Vectorcardiogram
Jo
ur
na
lP
re
-p
ro of
A-ECG
4
Journal Pre-proof
Introduction Conventional electrocardiographic (ECG) criteria for detecting left ventricular (LV) hypertrophy (LVH) including the Cornell voltage criteria, the Cornell voltage product criteria, and the Sokolow-Lyon criteria, all lack sensitivity (1). LVH is defined as an increased LV mass (LVM), but depending on the relation between mass and LV end diastolic volume (LVEDV), LVH can be classified into three patterns (2): 1) concentric remodeling, defined as
ro of
a normal LV mass with an increased wall thickness; 2) eccentric hypertrophy, defined as an increased mass with an increase in volume, yielding a preserved normal wall thickness; and 3)
-p
concentric hypertrophy, defined as a concomitant increase in both mass and wall thickness. Importantly, the different patterns have an increased risk of coronary heart disease (3). LVM
re
together with global wall thickness (GT) have recently been shown to be able to classify LVH
lP
into these categories, as illustrated in Figure 1, and this classification has been shown to be of
na
prognostic utility (4).
The ECG holds more proven diagnostic information than is usually considered or visible to
ur
the eye. Specifically, valuable diagnostic information can be gleaned from both the three-
Jo
dimensional vectorcardiogram (5) and QRS-wave and T-wave complexity measures such as eigenvectors obtained through singular value decomposition (6). These measures can be derived from a resting 12-lead ECG by Advanced ECG (A-ECG) analysis (7). By selecting a combination of conventional and A-ECG measures using statistical models, new diagnostic scores can be derived (7). Conventional ECG criteria for LVH are typically based on high QRS amplitudes, which have been attributed to the hypertrophied cardiomyocytes (8). However, changes in LVH are not solely attributed to hypertrophied cardiomyocytes; other complex changes such as focal or diffuse myocardial fibrosis and other pathologies may interact with the electrical propagation (8). Recent developments in cardiovascular magnetic
5
Journal Pre-proof resonance (CMR) imaging makes it possible to assess cardiac mass, volume and function, and to quantitatively characterize myocardial tissue characteristics such as myocardial infarction and diffuse myocardial fibrosis with high precision (9). With CMR as the reference, there are compelling reasons to reassess the role of the ECG in detecting the electrical changes in the different types of LVH.
It is unknown how wall thickness and mass, respectively, are manifested electrically in the
ro of
advanced ECG in LVH. Therefore, the aims of this study were to 1) create advanced ECG scores for increased LVM and increased GT, respectively, and 2) to compare these scores to conventional criteria for LVH. We hypothesized that A-ECG can be used to differentially
-p
detect increased LVM indexed to body surface area (LVMI) compared to increased GT
re
indexed to body surface area (GTI), and that the scores can outperform conventional ECG
lP
criteria for LVH.
ur
Study population
na
Methods
Jo
Patients were identified from a prospectively acquired database of patients referred for a clinical cardiovascular magnetic resonance (CMR) imaging at University of Pittsburgh Medical Center (UPMC; Pittsburgh, PA, USA). The study was approved by the UPMC Institutional Review Board, and all participants provided written informed consent. Inclusion criteria were completion of a contrast-enhanced CMR including measures of LVM and EDV, and an ECG with sinus rhythm acquired within 30 days of CMR. To study electrical consequences of only increased mass and wall thickness, respectively, we excluded the following potential CMR-related confounders that could potentially affect the ECG: previous myocardial infarction or non-ischemic myocardial scar by late gadolinium enhancement
6
Journal Pre-proof (LGE), hypertrophic or dilated cardiomyopathy, amyloidosis, severe valvular heart disease, and significant coronary artery stenosis by stress perfusion CMR. We also excluded the following clinical parameters: previous revascularization, and any prior cardiac surgery; and the following ECG-related parameters: treatment with digitalis, heart rate ≥ 100 beats/min, paced rhythm, atrial fibrillation or flutter, abundant premature atrial or ventricular beats (bigeminy or trigeminy), QRS duration ≥120 ms, fascicular blocks, bundle branch blocks, and inadequate ECG records. Baseline data included age at CMR, sex, body mass index (BMI),
ro of
body surface index (BSA), LVM, LVM indexed to BSA (LVMI), left ventricular end diastolic volume (LVEDV), LVEDV indexed to BSA (LVEDVI), global wall thickness (GT), GT
-p
indexed to BSA (GTI), extracellular volume fraction (ECV), left ventricular ejection fraction
re
(LVEF), global longitudinal strain (GLS), hypertension, diabetes mellitus, tobacco smoking status, treatment with beta-blockers, angiotensin-converter enzyme inhibitors (ACEi),
ur
Subgroups
na
reflected in their medical record.
lP
angiotensin receptor blockers (ARB), aspirin or other antiplatelet, and loop diuretics as
From a database of 1828 patients, 485 patients were eligible for the study. To create the A-
Jo
ECG diagnostic scores for increased LVMI and GTI, respectively, patients were divided into two subgroups: 1) normal and increased LVMI, where increased LVMI was defined as >63.7 g/m2 for females, and >81.6 g/m2 for males, based on the mean + 1.96 standard deviations (SD) in healthy volunteers (4), and 2) normal and increased GTI, where increased GTI was defined as >4.2 mm/m2 for females, and >4.4 mm/m2 for males, based on the mean + 1.96 SD in healthy volunteers (4). Thirty-six patients had missing baseline data for comorbidities and medication. CMR acquisition and analysis Quantification of volumes and dimensions, and late gadolinium enhancement imaging 7
Journal Pre-proof The CMR acquisition and analysis was performed as described previously (10). The CMR images were acquired using a 1.5 T scanner (Magnetom Espree, Siemens Healthcare, Erlangen, Germany) using 32 channel phased array surface coils. The exam included standard breath-held segmented cine imaging with a balanced steady-state free precession sequence. A short-axis stack was acquired and used to measure LV volumes, mass, and ejection fraction in end-diastole and end-systole. Late gadolinium enhanced (LGE) imaging was performed using a phase-sensitive inversion recovery (PSIR) pulse sequence, ten minutes after a 0.2 mmol/kg
ro of
intravenous bolus injection of gadoteridol (Prohance, Bracco Diagnostics, Princeton, NJ, USA). In patients with limited breath-hold capability, a single-shot motion corrected PSIR
-p
sequence was used (11, 12). Typical acquisition parameters were field of view 360x270 mm2,
re
256x128 mm2, and slice thickness 6 mm.
lP
Quantification of the extracellular volume fraction
ECV was calculated in quantitative T1 maps that were acquired with an ECG-gated single-
na
shot-modified Look Locker inversion recovery sequence (MOLLI). Native T1 maps were
ur
acquired followed by post-contrast T1 maps after the gadolinium bolus injection. To avoid
calculated as:
Jo
partial volume effects, T1 was measured in the middle third of the myocardium. ECV was , where λ =ΔR1myocardium/ΔR1bloodpool and
ΔR1=1/T1postcontrast - 1/T1precontrast (13).For blood ECV calculations a circular region was traced in the middle of the blood pool to avoid partial volume effects by papillary muscles. The final ECV values were averaged from mid-ventricular and basal short axis slices. Hematocrit measures were acquired on the day of CMR scanning. CMR data were analyzed using a Leonardo work station (Siemens Healthcare, Erlangen, Germany).
Quantification of global longitudinal strain
8
Journal Pre-proof GLS, a measure of impaired left ventricular function, was analyzed using the semi-automated tissue feature tracking software (CVi42, Circle Cardiovascular Imaging Inc., Calgary, Canada). Endocardial and epicardial borders were manually traced in 2-, 3- and 4 chamber end-diastolic images. Strain analysis tracings were inspected visually throughout the cardiac cycle, and manual adjustments were performed if traces deviated from myocardial movement. Calculation of global wall thickness index
ro of
Global wall thickness is the mean wall thickness of the whole left ventricle in end-diastole, measured in millimeters, and it was calculated according to previously validated methods as:
-p
GT =0.05+1.60 • LVM 0.83 • LVEDV -0.49 (4).
re
ECG acquisition and analysis
lP
Resting 12-lead ECG data for each subject were collected from the local ECG storage system (MUSE® Cardiology Information System, Version 8.0 SP2, GE Healthcare, IL, USA) and
na
exported into anonymized xml files with coded subject identification numbers. Advanced ECG semi-automatic software developed in-house was used to analyze the xml files. The
ur
following measures were derived and analyzed from the 12-lead ECG: conventional ECG
Jo
measures of scalar durations, amplitudes, and axes; derived vectorcardiographic (VCG) and polarcardiographic (14, 15) measures of three-dimensional spatial directions and magnitudes of the electrical activation pattern, via Kors’ transform (5); and QRS- and T-wave complexity measures by singular value decomposition (7, 16). The following conventional ECG criteria for LVH were analyzed: Cornell voltage criteria, defined as the S wave in lead V3 (SV3) plus the R wave in lead aVL (RaVL) where LVH is defined as >2.8 mV for males and >2.0 mV for females (17); Cornell product criteria, defined as the Cornell voltage times the QRS duration, where LVH is defined as ≥244mV•ms (18);
9
Journal Pre-proof and Sokolow-Lyon criteria, defined as the sum of SV1 plus the larger of the RV5 or RV6, where LVH was defined as a >3.5 mV (19). Statistical analysis Statistical analysis was performed in R version 3.4.1 (R Foundation for Statistical Computing, Vienna, Austria) using packages dplyr (20) for data transformation, and ggplot2 (21) for data visualizations. The A-ECG scores for increased LVMI and GTI, respectively, were derived
ro of
using stepwise forward logistic regression, and continuous A-ECG measures were analyzed. To determine the diagnostic accuracy of the scores, the area under the receiver operating characteristics curve (AUC) was calculated and bootstrapped 2000 times to obtain 95%
-p
confidence intervals (CI). The A-ECG scores that yielded the highest AUC were chosen, and
re
the Youden index (22) was used to optimize cut-offs for the scores’ sensitivity and specificity
lP
to detect increased LVMI and GTI, respectively. A minimum of ten events per incremental AECG parameter was accepted for the scores (23). The diagnostic scores were constructed to
na
show the likelihood of having increased LVMI or GTI, respectively, ranging from 0% to
•100
Jo
ur
100%, calculated from the logistic regression as:
The AUC for the two A-ECG scores were compared to each other and to the AUC for the conventional ECG criteria, respectively, using the conventional cut-offs for LVH as described and the DeLong test (24), and illustrated using forest plots. Pearson’s correlation coefficient was used to evaluate if the A-ECG parameters in the respective scores were interrelated, and expressed as its square (R2). The Kolmogorov-Smirnov test was used to test if data were normally distributed, and differences between subgroups’ baseline data were tested using the chi-square test or Mann-Whitney U test, as appropriate, and described using percentage and
10
Journal Pre-proof median [interquartile range], respectively. A p-value <0.05 was considered statistically significant.
Results We identified 485 patients (age 51 [38-61] years, 54% female), in which 51 (11%) patients had increased LVMI and 65 (13%) had increased GTI. Sixteen (3%) patients had an isolated
ro of
increased LVMI, and 30 (6%) patients had an isolated increased GTI. Four hundred and four patients (83%) had both a normal LVMI and GTI. Based on the number of patients with
-p
increased LVMI and GTI, respectively, a maximum of five and six A-ECG parameters were
re
accepted for the LVMI and GTI scores, respectively. Increased GTI was associated with older age (p=0.002), and a higher LVEF, compared to increased LVMI (p=0.17, p=0.96,
lP
respectively). Baseline characteristics of the study population are presented in Table 1. A total
na
of 34 (7%) patients had missing data on baseline co-morbidity and indication for CMR. A-ECG score versus conventional criteria for increased LVMI
ur
The measures that contributed to the A-ECG score for increased LVMI are presented in Table
Jo
2. They included two derived vectorcardiographic measures, and two conventional ECG measures, specifically: 1) the area of the QRS vector magnitude (mV•s); 2) the QRS loop area outside the left lower quadrant in the left sagittal plane (%); 3) the QTc interval (ms); and 4) the T wave amplitude in lead aVR (µV). The A-ECG score for detecting increased LVMI had a sensitivity of 75%, and a specificity of 81%, and a higher AUC [95% CI] (0.84 [0.78-0.90]) than the Cornell voltage criteria (0.63 [0.57-0.70]), the Cornell product criteria (0.60 [0.550.67]), and the Sokolow-Lyon criteria (0.57 [0.52-0.62]), as shown in Figure 2 (p<0.001 for all). A-ECG score versus conventional criteria for increased GTI
11
Journal Pre-proof The measures that contributed to the A-ECG score for increased GTI are presented in Table 2. They included one measure of QRS wave complexity, two derived vectorcardiographic measures, one conventional ECG measure, and one derived polarcardiographic measure, specifically: 1) the 6th Eigenvalue (singular value) of the QRS; 2) the bulginess of T vector loop (mV); 3) the azimuth of the maximum QRS magnitude in the left sagittal plane (radians); 4) the T amplitude in lead I (µV); and 5) the elevation (latitude) of the polar vector (degrees).
ro of
The A-ECG score for detecting increased GTI had a sensitivity of 71% and a specificity of 80%, and a higher AUC (0.80 [0.74-0.85]) than the Cornell voltage criteria (0.58 [0.54-0.63]),
0.58]), as shown in Figure 1 (p<0.001 for all).
-p
the Cornell product criteria (0.57 [0.53-0.61]), and the Sokolow-Lyon criteria (0.54 [0.51-
re
Electrocardiographic differences between increased LVMI and increased GTI
lP
The two A-ECG scores were also compared to each other. The GTI A-ECG score had a higher AUC than the LVMI A-ECG score for detecting increased GTI [0.80 [0.74-0.85] vs
na
0.71 [0.64-0.78], p=0.02). However, the GTI A-ECG score could also detect increased LVMI
ur
as well as the dedicated LVMI A-ECG score (0.83 [0.88-0.88] vs 0.84 [0.78-0.90], p=0.88,
Jo
see Figure 1. The largest R2 between any two parameters in the respective scores was 0.063, indicating that the parameters in each score were at most weakly associated.
Discussion
The main findings of this study were that, in a consecutive clinical cohort of patients without myocardial scar or other confounders that may interact with the ECG besides LVH, newly constructed A-ECG scores outperformed conventional LVH ECG criteria (Cornell voltage, Cornell voltage product, and Sokolow-Lyon criteria) for determining increased LVMI and GTI, respectively. Furthermore, increased LVMI and GTI differed in their electrocardiographic manifestation by A-ECG.
12
Journal Pre-proof Although the two scores included different measures, the GTI A-ECG score could diagnose high LVMI on par with the dedicated LVMI A-ECG score, while the LVMI A-ECG score could not diagnose high GTI as well as the dedicated GTI A-ECG score. This suggests that on a group level, increased GTI associates with distinct electrocardiographic abnormalities beyond those associated with increased LVMI, but not vice versa. Depending on the relation between increased mass and wall thickness, LVH can be divided into concentric remodeling, eccentric hypertrophy, and concentric hypertrophy. Increased mass includes eccentric
ro of
hypertrophy and concentric hypertrophy, while increased wall thickness includes concentric remodeling and concentric hypertrophy. Thus, as the subgroups overlap, the common LVH
-p
denominator in both scores are patients with concentric hypertrophy and the electrical
re
differences can be found in patients with either concentric remodeling or eccentric hypertrophy, as shown in Figure 1. The accuracy of conventional ECG LVH criteria remained
lP
poor, as expected (1, 25).
na
Previous studies and the advanced ECG score parameters
ur
Essentially all measures that comprised the constituents of the A-ECG scores derived in this study, or corresponding measures, have been previously studied in ventricular hypertrophy.
Jo
For example, in relation to the area of the QRS vector magnitude, it has been shown that an increased area of the QRS improved the diagnosis of hypertrophic hearts (26). In healthy subjects, the largest percentage of the QRS loop area resides in the lower quadrants of the left sagittal plane (27). Furthermore, studies have found correlations between prolonged QTc and LVH in hemodialysis (28) and hypertensive (29) patients. The bulginess, or the distortion, of the T vector loop, also referred to as Tavplan, is the distance from a preferential twodimensional plane to the three-dimensional T loop (30), and the bulginess of the T vector is sensitive to ischemia followed by coronary occlusion (31). In a study where patients with LVH by Sokolow-Lyon index were excluded, patients with hypertension had more
13
Journal Pre-proof pronounced T loop changes than those without hypertension (32). However, hypertension is associated with LVH (33), and LVH could not be ruled out by ECG (32). Regarding the T wave amplitudes in leads aVR and I, hemodialysis patients with upright Twaves in lead aVR had not only poorer outcomes, but also relatively increased LVMI (34). Moreover, T wave inversion in other leads has been shown to correlate with LVH (35). Others have also found that the azimuth of the QRS at maximal magnitude in the sagittal (36)
ro of
horizontal (37) planes was significantly associated with LVH. The latitude of the polar vector was also a constituent of the GTI A-ECG score. In contrast to the vectorcardiogram from which it can be derived, the polarcardiogram also plots the
-p
direction and magnitude of the heart’s electrical vector over time, and in a coordinate system
re
that is spherical (polar) (38). In a previous study, less than half of the patients with LVH had a
lP
polar vector that deviated from the polar vector in normal controls (39). In summary, essentially all (eight of nine) parameters selected by the stepwise forward logistic regression
na
model for diagnosing increased LVMI and GTI have previously been studied in, and found to
ur
be associated with, LVH. The final parameter, the lower-level QRS Eigenvalue that also contributed to the score for diagnosing increased GTI, has not, to our knowledge, previously
Jo
been studied in LVH. However, all lower-level QRS eigenvalues are known to be relatively increased in heart failure (7), of which hypertrophic remodeling is a typical manifestation. Study limitations This study has several limitations. First, its experimental design was optimized for attempting to ascertain the differential electrocardiographic manifestations of different morphological aspects of LVH. In order to optimally isolate the electrical determinants of increased LVMI and GTI, respectively, we excluded patients with other pathologies that may interact with the ECG. Thus, the findings in this study are not applicable to general clinical cohorts wherein other co-morbidities are often present that overlap with LVH. Second, due to a lack of 14
Journal Pre-proof statistical power, the exact electrical differences between the LVH types of concentric remodeling, eccentric hypertrophy, and concentric hypertrophy could not be determined in the current study. However, such studies are justified. Third, a total of 34 (7%) patients had missing data on baseline co-morbidity and indication for CMR. Finally, besides the conventional ECG LVH criteria evaluated in this study, other sets of criteria have been recommended by the American Heart Association (8), but were not specifically studied.
ro of
Clinical applications The present study gives insight into the different electrical properties in LVH related to increased LVMI or GTI. Based on the improved diagnostic performance of the newly
-p
proposed A-ECG scores over conventional ECG criteria for LVH, we suggest that wherever
re
digital 12-lead ECGs are available, these scores could be additionally employed.
lP
Electrophysiology and anatomy are fundamentally different phenomena, and the changes observed on the ECG should preferentially be referred to as left ventricular electrical
ur
working group (40).
na
remodeling, in accordance with the suggestions of the Journal of Electrocardiology LVH
Jo
Conclusions
Increased LVMI and GTI differed in their electrophysiological manifestation by A-ECG. New A-ECG scores outperform conventional ECG criteria for LVH in determining increased LVMI and GTI, respectively. Sources of funding The study was supported in part by grants from the Swedish Research Council, Swedish Heart and Lung Foundation, Stockholm County Council, and Karolinska Institutet. Disclosures
15
Journal Pre-proof Dr. Schlegel is a principal of Nicollier-Schlegel SARL, a company that performs ECG
Jo
ur
na
lP
re
-p
ro of
research consultancy using the software used in the present study.
16
Journal Pre-proof
References 1.
Pewsner D, Jüni P, Egger M, Battaglia M, Sundström J, Bachmann LM.
Accuracy of electrocardiography in diagnosis of left ventricular hypertrophy in arterial hypertension: systematic review. BMJ. 2007;335(7622):711. 2.
Ganau A, Devereux RB, Roman MJ, de Simone G, Pickering TG, Saba PS, et
al. Patterns of left ventricular hypertrophy and geometric remodeling in essential
3.
ro of
hypertension. J Am Coll Cardiol. 1992;19(7):1550-8. Gupta S, Berry JD, Ayers CR, Peshock RM, Khera A, de Lemos JA, et al. Left
-p
ventricular hypertrophy, aortic wall thickness, and lifetime predicted risk of cardiovascular
4.
re
disease:the Dallas Heart Study. JACC Cardiovasc Imaging. 2010;3(6):605-13. Lundin M, Heiberg E, Nordlund D, Gyllenhammar T, Steding-Ehrenborg K,
lP
Engblom H, et al. Mean wall thickness improved characterization and prognosis in left
na
ventricular hypertrophy [abstract]. In: Proceedings of the 22nd Annual Meeting of the Society
5.
Kors JA, van Herpen G, Sittig AC, van Bemmel JH. Reconstruction of the
Jo
ID#549719.
ur
for Cardiovascular Magnetic Resonance; 2019, Feb 6-9; Bellevue, WA, US. Abstract
Frank vectorcardiogram from standard electrocardiographic leads: diagnostic comparison of different methods. Eur Heart J. 1990;11(12):1083-92. 6.
Rautaharju PM, Kooperberg C, Larson JC, LaCroix A. Electrocardiographic
predictors of incident congestive heart failure and all-cause mortality in postmenopausal women: the Women's Health Initiative. Circulation. 2006;113(4):481-9. 7.
Schlegel TT, Kulecz WB, Feiveson AH, Greco EC, DePalma JL, Starc V, et al.
Accuracy of advanced versus strictly conventional 12-lead ECG for detection and screening
17
Journal Pre-proof of coronary artery disease, left ventricular hypertrophy and left ventricular systolic dysfunction. BMC Cardiovasc Disord. 2010;10:28. 8.
Bacharova L, Ugander M. Left ventricular hypertrophy: The relationship
between the electrocardiogram and cardiovascular magnetic resonance imaging. Ann Noninvasive Electrocardiol. 2014;19(6):524-33. 9.
Karamitsos TD, Francis JM, Myerson S, Selvanayagam JB, Neubauer S. The
ro of
role of cardiovascular magnetic resonance imaging in heart failure. J Am Coll Cardiol. 2009;54(15):1407-24.
Schelbert EB, Piehler KM, Zareba KM, Moon JC, Ugander M, Messroghli DR,
-p
10.
et al. Myocardial Fibrosis Quantified by Extracellular Volume Is Associated With Subsequent
re
Hospitalization for Heart Failure, Death, or Both Across the Spectrum of Ejection Fraction
Kellman P, Larson AC, Hsu LY, Chung YC, Simonetti OP, McVeigh ER, et al.
na
11.
lP
and Heart Failure Stage. J Am Heart Assoc. 2015;4(12).
Motion-corrected free-breathing delayed enhancement imaging of myocardial infarction.
Piehler KM, Wong TC, Puntil KS, Zareba KM, Lin K, Harris DM, et al. Free-
Jo
12.
ur
Magn Reson Med. 2005;53(1):194-200.
breathing, motion-corrected late gadolinium enhancement is robust and extends risk stratification to vulnerable patients. Circ Cardiovasc Imaging. 2013;6(3):423-32. 13.
Arheden H, Saeed M, Higgins CB, Gao DW, Bremerich J, Wyttenbach R, et al.
Measurement of the distribution volume of gadopentetate dimeglumine at echo-planar MR imaging to quantify myocardial infarction: comparison with 99mTc-DTPA autoradiography in rats. Radiology. 1999;211(3):698-708. 14.
Belletti DA, Lyon AF. The initial 0.04 second polar vector of the QRS loop as
an aid in the detection of a myocardial infarction. Jpn Heart J. 1969;10(3):267-72.
18
Journal Pre-proof 15.
Pipberger H. Use of Computers in Interpretation of Electrocardiograms:
Lippincott Williams & Wilkins, Baltimore, MD; 1962. 16.
Batdorf BH, Feiveson AH, Schlegel TT. The effect of signal averaging on the
reproducibility and reliability of measures of T-wave morphology. J Electrocardiol. 2006;39(3):266-70. 17.
Casale PN, Devereux RB, Kligfield P, Eisenberg RR, Miller DH, Chaudhary
ro of
BS, et al. Electrocardiographic detection of left ventricular hypertrophy: development and prospective validation of improved criteria. J Am Coll Cardiol. 1985;6(3):572-80. Molloy TJ, Okin PM, Devereux RB, Kligfield P. Electrocardiographic detection
-p
18.
of left ventricular hypertrophy by the simple QRS voltage-duration product. J Am Coll
Sokolow M, Lyon TP. The ventricular complex in left ventricular hypertrophy
lP
19.
re
Cardiol. 1992;20(5):1180-6.
Wickham H, Francois R, Henry L, Müller K. Dplyr: A Grammar of Data
Manipulation. 2017.
Wickham H. Ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag
New York. 2009.
Jo
21.
ur
20.
na
as obtained by unipolar precordial and limb leads. Am Heart J. 1949;37(2):161-86.
22.
Youden W. Index for rating diagnostic tests. Cancer. 1950;3(1):32-5.
23.
Peduzzi P, Concato J, Feinstein AR, Holford TR. Importance of events per
independent variable in proportional hazards regression analysis. II. Accuracy and precision of regression estimates. J Clin Epidemiol. 1995;48(12):1503-10.
19
Journal Pre-proof 24.
DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two
or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics. 1988;44(3):837-45. 25.
Sjöberg S, Sundh F, Schlegel T, Maynard C, Rück A, Wagner G, et al. The
relationship between electrocardiographic left ventricular hypertrophy criteria and echocardiographic mass in patients undergoing transcatheter aortic valve replacement. J
26.
ro of
Electrocardiol. 2015;48(4):630-6. Okin PM, Roman MJ, Devereux RB, Pickering TG, Borer JS, Kligfield P. Time-
voltage QRS area of the 12-lead electrocardiogram: detection of left ventricular hypertrophy.
Chou TC, Masangkay MP, Young R, Conway GF, Helm RA. Simple
re
27.
-p
Hypertension. 1998;31(4):937-42.
28.
na
Circulation. 1973;48(6):1262-7.
lP
quantitative vectorcardiographic criteria for the diagnosis of right ventricular hypertrophy.
Alonso MAG, Lima VACC, Carreira MAMQ, Lugon JR. Reproducibility and
ur
Reliability Of QTc and QTcd Measurements and Their Relationships with Left Ventricular
29.
Jo
Hypertrophy in Hemodialysis Patients. Arq Bras Cardiol. 2017;109(3):222-30. Porthan K, Virolainen J, Hiltunen TP, Viitasalo M, Väänänen H, Dabek J, et al.
Relationship of electrocardiographic repolarization measures to echocardiographic left ventricular mass in men with hypertension. J Hypertens. 2007;25(9):1951-7. 30.
Badilini F, Maison Blanche P, Fayn J. Relationship between 12-lead ECG QT
dispersion and 3D-ECG repolarization loop. Computers in Cardiology. 1995; 785-8. 31.
Nowinski K, Jensen S, Lundahl G, Bergfeldt L. Changes in ventricular
repolarization during percutaneous transluminal coronary angioplasty in humans assessed by QT interval, QT dispersion and T vector loop morphology. J Intern Med. 2000;248(2):126-36.
20
Journal Pre-proof 32.
Rubulis A, Jensen J, Lundahl G, Tapanainen J, Wecke L, Bergfeldt L. T vector
and loop characteristics in coronary artery disease and during acute ischemia. Heart Rhythm. 2004;1(3):317-25. 33.
Shenasa M, Shenasa H. Hypertension, left ventricular hypertrophy, and sudden
cardiac death. Int J Cardiol. 2017;237:60-3. 34.
Jaroszyński A, Jaroszyńska A, Siebert J, Dąbrowski W, Niedziałek J, Bednarek-
patients. Clin Exp Nephrol. 2015;19(6):1157-64.
Kannel WB, Gordon T, Castelli WP, Margolis JR. Electrocardiographic left
-p
35.
ro of
Skublewska A, et al. The prognostic value of positive T-wave in lead aVR in hemodialysis
ventricular hypertrophy and risk of coronary heart disease. The Framingham study. Ann
Varriale P, Alfenito JC, Kennedy RJ. The vectorcardiogram of left ventricular
lP
36.
re
Intern Med. 1970;72(6):813-22.
37.
na
hypertrophy. Analysis and criteria (Frank Lead system). Circulation. 1966;33(4):569-76. Romhilt DW, Greenfield JC, Estes EH. Vectorcardiographic diagnosis of left
Macfarlane PW, vanOosterom A, Pahlm O, Kligfield P, Janse M, Camm J.
Jo
38.
ur
ventricular hypertrophy. Circulation. 1968;37(1):15-9.
Comprehensive Electrocardiology. London: Springer; 2011. 39.
Pipberger H, Carter T. Analysis of the normal and abnormal vectorcardiogram
in its own reference frame. Circulation. 1962;25:827-40. 40.
Bacharova L, Estes EH, Bang LE, Hill JA, Macfarlane PW, Rowlandson I, et al.
Second statement of the working group on electrocardiographic diagnosis of left ventricular hypertrophy. J Electrocardiol. 2011;44(5):568-70.
21
Journal Pre-proof
Tables Table 1. Baseline characteristics for the two sub-groups.
Characteristics
All
Number, n Female sex, n (%) Age, years
ECV, %
LVEDV, mL
Normal LVMI 485
434
51
263 (54)
230 (53)
33 (65)
51
51
54
(38–61)
(38–61)
(42–65)
27.2
27.2
(25.0–30.0)
(24.9–29.9)
147.0
144.6
LVEDVI, mL/m
73.9 (64.2–87.0)
GLS, %
LVEF, %
LVM, g LVMI, g/m2
J
l a
rn
u o
(123.4–179.9) 2
Increased LVMI
27.9
(122.1–174.1)
(149.0–237.4)
72.8
98.6
(63.5–84.2)
(149.0–237.4)
-17.9
-13.7
(-19.7– -15.4)
(-19.8– -15.9)
(-16.3– -10.1)
62.0
62.0
58.0
(56.0–67.0)
(57.0–67.0)
(45.3–65.0)
100.6
96.7
163.7
(81.0–129.1)
(79.8–121.2)
(132.0–213.4)
49.5
47.5
79.9
-17.6
Increased GTI
p value
420
65
-
0.11
225 (54)
38 (58)
0.46
0.17
50
58
0.002
(37–60)
(49–65)
27.1
27.9
(25.0–29.7)
(24.9–30.9)
148.3
140.7
(125.1–179.9)
(111.0–196.7)
73.8
76.7
(63.3–86.5)
(64.1–90.8)
-17.9
-15.7
(-19.9– -15.8)
(-17.5– -12.2)
62.0
63.4
(56.4–67.0)
(55.0–67.0)
97.0
137.6
(79.8–123.3)
(111.1–184.7)
47.4
70.3
0.23
(25.7–30.3) 205.6
Normal GTI
-
e
r P
f o
o r p
p value
<0.001
<0.001
<0.001
0.002
<0.001
<0.001
0.11
0.22
0.38
<0.001
0.96
<0.001
<0.001
Journal Pre-proof (42.2–60.2)
(41.7–56.1)
(68.4–93.2)
6.8
6.6
8.7
(5.9–7.9)
(5.8–7.5)
(8.1–10.6)
3.4
3.3
4.6
(3.0–3.9)
(3.0–3.7)
(4.1–5.3)
28.3
28.3
28.2
(24.3–34.0)
(24.2–34.1)
(24.4–32.4)
2.0
2.0
1.9
(1.8–2.2)
(1.8–2.2)
(1.8–2.2)
Hypertension
179 (40)
151 (38)
28 (62)
Diabetes mellitus
54 (12)
40 (10)
14 (31)
Current smoker
61 (14)
50 (12)
Ex-smoker
132 (30)
118 (29)
GT, mm
GTI, mm/m
2
2
BMI, kg/m 2
BSA, m
Aspirin or antiplatelet
121 (27) other
162 (36)
8.7
(5.8–7.4)
(8.2–10.1)
3.3
4.6
(2.9–3.6)
(4.5–5.2)
<0.001
f o
0.60
<0.001
<0.001
25.8
(24.4–34.9)
(23.2–30.1)
2.0
1.9
(1.8–2.2)
(1.7–2.1)
0.001
148 (38)
31 (56)
0.008
<0.001
40 (10)
14 (26)
0.001
11 (24)
0.03
51 (13)
10 (18)
0.30
14 (31)
0.81
118 (30)
14 (26)
0.48
100 (25)
21 (47)
0.002
103 (26)
18 (33)
0.31
141 (35)
21 (47)
0.13
137 (35)
25 (46)
0.13
128 (32)
23 (51)
0.01
130 (33)
21 (38)
0.46
l a
rn
u o
J
6.7
28.6
Smoking status, n (%)
ACEi/ARB
(61.3–87.5)
<0.001
o r p
0.27
e
Co-morbidity, n (%)*
Medication, n (%)
(41.7–56.2)
r P
0.004
<0.001
Beta-blockers
151 (34)
Loop diuretics
60 (13)
51 (13)
9 (20)
0.17
51 (13)
9 (16)
0.50
Insulin
39 (9)
28 (7)
11 (24)
<0.001
29 (7)
10 (18)
0.008
Oral hypoglycemic
19 (4)
17 (4)
2 (4)
0.95
14 (4)
5 (9)
0.06
124 (28)
110 (27)
14 (31)
0.60
106 (27)
18 (33)
0.38
Statin
General indication for CMR: known or suspected disease, n (%)*
23
Journal Pre-proof Arrhythmia syncope
or
164 (34)
152 (35)
12 (24)
0.10
152 (36)
12 (18)
0.005
Cardiomyopathy
194 (40)
172 (40)
22 (43)
0.63
164 (39)
30 (46)
0.28
Coronary disease/
208 (43)
185 (43)
23 45)
0.74
181 (43)
27 (41)
0.81
artery
vasodilator stress test Pericardial disease
22 (5)
19 (4)
3 (6)
0.63
Thoracic assessment
15 (3)
14 (3)
1 (2)
0.63
aorta
p e
ro
f o
Co nti
20 (5)
2 (3)
0.54
12 (3)
3 (5)
0.45
nuo us
data are given as number, n (%), or median (interquartile range), and compared between Normal LVMI and Increased LVMI, and Normal GTI
r P
and Increased GTI, respectively, using Chi-square or Mann-Whitney U test. ACEi indicates angiotensin-converting enzyme inhibitors; ARB,
l a
angiotensin II receptor blocker; BMI, body mass index; BSA, body surface area; CMR; cardiovascular magnetic resonance imaging; ECV,
n r u
extracellular volume fraction; LVEDV, left ventricular end-diastolic volume; LVEDVI, LVEDV index; GLS, global longitudinal strain; LVEF, left ventricular ejection fraction; LVM, left ventricular mass; LVMI, LVM index; GT, global wall thickness; GTI, GT index. *Data from 451 (93%) patients.
o J
24
Journal Pre-proof
Table 2. A-ECG parameters in the increased left ventricular mass index (LVMI) and global wall thickness index (GTI) scores. ECG parameter
Z-value
p value
AUC (95% CI)
Increased LVMI score 1)
The area of the QRS vector magnitude (mV•s)
6.0
<0.001
2)
The QRS loop area outside the left lower quadrant in the left sagittal plane (%)
-3.3
<0.001
T wave amplitude in lead aVR (µV)
-3.2
0.001
4)
QTc (ms)
2.7
0.007
ro of
3)
-p
0.84 (0.78-0.90)
Increased GTI score 4.4
2)
The bulginess of T vector loop (mV)
4.2
3)
The azimuth of the maximum QRS magnitude in the left sagittal plane (radians)
3.5
na
T wave amplitude lead I (µV)
5)
The latitude of the polar vector (degrees)
Jo
ur
4)
<0.001
re
6th Eigenvalue (singular value) of the QRS
lP
1)
<0.001
<0.001 0.80 (0.74-0.86)
-3.5
0.002
-2.8
0.006
Journal Pre-proof Figure 1. A schematic figure of how normal and increased left ventricular (LV) mass index and LV global wall thickness index, respectively, can classify left ventricular hypertrophy into three patterns: concentric remodeling, eccentric hypertrophy, and concentric hypertrophy. The LV global wall thickness index is the mean wall thickness of the entire left ventricle as measured by MRI. Cutoffs defining increase beyond normal (not explicitly shown) are the
Jo
ur
na
lP
re
-p
ro of
sex-specific 95% upper limit of normal in healthy volunteers.
26
Journal Pre-proof Figure 2. Area under the curve (AUC) for the increased left ventricular mass index (LVMI) score, increased global wall thickness index (GTI) score, Sokolow-Lyon, Cornell voltage, and
Jo
ur
na
lP
re
-p
ro of
Cornell product for detecting increased LVMI and GTI, respectively.
27
Figure 1
Figure 2