Ventricular Gradient as a Risk Factor in Survivors

of Acute Myocardial Infarction

VELISLAV BATCHVAROV, KATERINA HNATKOVA, AZAD GHURAN,
JAN POLONIECKI, A. JOHN CAMM, and MAREK MALIK
From the Department of Cardiological Sciences, St. George’s Hospital Medical School, London, England

BATCHVAROV, V., ET AL.: Ventricular Gradient as a Risk Factor in Survivors of Acute Myocardial
Infarction. The total cosine between R and T (TCRT) (angular difference between the spatial QRS and
T wave loops) is a technical elaboration of the concept of ventricular gradient (VG), whose power as a risk
stratifier in post-MI patients has already been demonstrated. Recently, it was reported that TCRT differed
significantly between healthy men and women, which suggested that its predictive power might be gen-
der dependent. The aim of the study was to investigate TCRT and its association with cardiac mortality
in male and female survivors of acute MI. TCRT was measured from digital Frank orthogonal XYZ-lead
ECGs recorded before hospital discharge in 681 survivors of acute MI (82% men, age: men 57.0 ± 8.4 years,
women 59.6 ± 8.1 years, P = 0.002). During a follow-up censored at 5 years, cardiac mortality rates were
9.7% and 12.1% in men and women, respectively (P = 0.42). There were no significant difference in TCRT
between men and women (−0.150 ± 0.704 vs −0.070 ± 0.731, P = 0.26). In univariate Cox regression
analysis, TCRT < −0.88 was related to a 5-year cardiac mortality in men (relative risk [RR] 3.67, 95%
confidence interval [CI] 2.13–6.34, P = 1.9 × 10 −6 ), and women (RR 5.16, 95% CI 1.83–14.56, P = 0.0015).
Depressed TCRT was strongly associated with increased long-term cardiac mortality in survivors of acute
MI. Its predictive power did not differ significantly between the sexes. The role of TCRT as a risk-stratifier
in post-MI patients deserves further prospective assessment in multivariate models with established risk
factors. (PACE 2003; 26[Pt. II]:373–376)
ventricular gradient, repolarization, gender differences, risk stratification, myocardial infarction

Introduction which suggested that its value as a risk stratifier

The conflict between the cost of health care
and the prophylactic efficacy of implantable car-
dioverter defibrillators (ICDs),1,2 and the design of
future antiarrhythmic drug trials enhance the need
for accurate risk stratification of postmyocardial
infarction (post-MI) patients. However, the predic-
tive value of presently established risk stratifiers is
modest even with the use of multiple variables.3
Recently, a new electrocardiographic (ECG)
descriptor of ventricular repolarization was intro-
duced, which quantifies the vectorial deviation be-
tween the QRS and T wave loops (total cosine
between R and T, TCRT).4 TCRT is based on and
further develops the classical concept of ventric-
ular gradient (VG).5,6 One study indicated that
TCRT was a strong and independent predictor of
adverse outcome in post-MI patients.7
However, a recently by published extensive
analysis of continuous 24-hour, 12-lead digital
ECGs demonstrated that TCRT differed signifi-
cantly between healthy young men and women,8
Supported in part by The Wellcome Trust, London, England
and the British Heart Foundation, London, England.
Address for reprints: Velislav Batchvarov, M.D., Dept. of Cardi-
ological Sciences, St. George’s Hospital Medical School, Cran-
mer Terrace, London SW17 0RE, England. Fax: +44-20-8725-
0846; e-mail: vbatchva@sghms.ac.uk
may be dependent on the patient’s sex.
This study analyzed a prospectively collected
database of survivors of acute MI with a relatively
long follow-up to compare TCRT and its associa-
tion with long-term cardiac mortality in male and
female patients.
Methods
The study used the database of the St. George’s
Post-Infarction Research Survey which screened
1,338 patients (≥75 years of age) who were ad-
mitted to St. George’s Hospital for acute MI be-
tween 1984 and 1994.9 Patients with valvular heart
disease, permanent pacemakers, atrial fibrillation,
and a QRS complex >120 ms were excluded by the
protocol of the survey. Before hospital discharge, a
signal-averaged ECG (SAECG) was recorded in 681
patients in Frank orthogonal bipolar XYZ leads us-
ing an ART 1200 EPX device with a 40–250 Hz
filter (Arrhythmia Research Technology, Austin,
TX, USA). The study population consisted of 557
men (mean age 57.0 ± 8.4 years) and 124 women
(mean age 59.6 ± 8.1 years, P = 0.002). TCRT
was calculated as a cosine of the angle between
the spatial QRS and T wave loops reconstructed
in a minimum dimensional space from the digi-
tally recorded XYZ leads of the SAECG recordings.
Details of the physiological background and the

PACE, Vol. 26 January 2003, Part II 373


method of calculation of TCRT have been previ-
ously published.4,10
All data were censored at 5 years. The end-
point of the study was overall cardiac death.
Statistical Analysis
Previous studies have suggested that lower
values of TCRT were associated with adverse out-
come in post-MI patients.7,11 Therefore, TCRT
was dichotomized using the 20th percentile (i.e.,
TCRT = −0.88). Kaplan-Meier curves were com-
pared by log-rank test. Univariate Cox regression
analysis with case-wise deletion of missing data
was performed for 5-year cardiac mortality rate
separately in men and in women. Clinical vari-
ables are presented as percentages for categorical
variables, and mean ± SD for continuous variables,
and compared by chi-squared or nonparametric
Mann-Whitney test, as appropriate. The associa-
tion between TCRT and left ventricular ejection
fraction (LVEF) was estimated using the Pearson
correlation coefficient. A P value <0.05 was con-
sidered statistically significant.
Results
During the mean follow-up of 35.2 ± 20.7
months, there were 69 (10.1%) cardiac deaths, 15
(12.1%) of 124 in women and 54 (9.7%) of 557 in
men (P = 0.42).
The characteristics of the patients are pre-
sented in Table I. Survivors had significantly
higher LVEF and TCRT compared to nonsurvivors,
while there were no significant differences in age
or percentage of patients with anterior and non-Q
MI between the two groups.
There were no significant differences in LVEF
(44.8 ± 14.5 vs 47.3 ± 15.3%, P = 0.13), percent-
age of patients with anterior MI (50.4 vs 46.8%,
P = 0.46), non-Q MI 33.9 vs 34.7%, P = 0.87), and
TCRT (−0.150 ± 0.704 vs −0.070 ± 0.731, P = 0.26)
between men and women.
Clinical Characteristics of the Patients
Parameter
(% of all patients)
Age (years)
LVEF (%)
Anterior infarction
Non-Q infarction
Beta-blockers on discharge
TCRT
The table presents basic clinical characteristics of survivors and victims of 5-year cardiac mortality.
LVEF = left ventricular ejection fraction; TCRT = total consine between R and T.
374
BATCHVAROV, ET AL.
Figure 1 presents the cumulative cardiac mor-
tality in men (top panel) and in women (bottom
panel). In both groups, patients with TCRT <
−0.88 had significantly higher 5-year cardiac
mortality rate (>30%), whereas in those with
TCRT ≥ −0.88 the 5-year cardiac mortality rate
was 10%.
In univariate Cox regression analysis, TCRT <
−0.88 was related to a 5-year cardiac mortality
in both men (relative risk [RR] 3.67, 95% confi-
dence intervals [CI] 2.13–6.34, P = 1.9 × 10−6 ), and
women (RR 5.16, 95% CI 1.83–14.56, P = 0.0015).
TCRT was only weekly correlated to LVEF (r =
0.18).
Discussion
In a large and relatively low risk post-MI pop-
ulation, this study showed that decreased TCRT
measured at hospital discharge was associated
with increased long-term cardiac mortality. No sig-
nificant differences in TCRT between men and
women were observed, and, in both groups, pa-
tients with TCRT < −0.88 had 5-year cardiac mor-
tality rate >30% compared to 10% in those with
TCRT ≥ −0.88.
TCRT quantifies the deviation between the
directions of depolarization and repolarization
loops, reviving the concept of VG.5,6,12 It has been
suggested decades ago that VG is a global measure
of the variations of ventricular action potential du-
ration and morphology,13,14 which could serve as
an ECG index of vulnerability to arrhythmia.15,16
While VG and TCRT share the same general phys-
iological background, their methods of calcula-
tion differ substantially.4 Zabel et al.7 analyzed a
database of 280 post-MI patients and found that
patients with decreased TCRT had decreased sur-
vival and a higher incidence of arrhythmic events
during follow-up. In multivariate analysis, TCRT
was independently predictive of primary endpoint
events.
Table I.
Survivors Victims of Cardiac
(n = 612) Death (n = 69) P Value
57.3 ± 8.4 58.8 ± 8.4 0.17
46.6 ± 14.2 33.9 ± 14.4 5.4 × 10−10
49.2 55.1 0.15
36.9 33.3 0.54
44.6 18.8 1.3 × 10−5
−0.088 ± 0.708 −0.552 ± 0.571 2.0 × 10−7
January 2003, Part II PACE, Vol. 26
 VENTRICULAR GRADIENT AFTER MI

Figure 1. Cumulative 5-year cardiac death-free survival. Top panel—men, bottom panel—
women. Note that in both groups, patients with TCRT < −0.88 had 5-year cardiac mortality
rate of >30%, while in those with TCRT ≥ −0.88 the mortality rate was approximately 10%.

In a recently published extensive analysis
of multiple continuous 24-hour, 12-lead digital
ECGs, Smetana et al.8 reported that TCRT was
significantly lower in healthy men than healthy
women during the 24-hour period of recording
(0.35 ± 0.37 vs 0.67 ± 0.20, P = 8 × 10−11 ). In
both men and women, TCRT exhibited a signifi-
cant circadian pattern with lowest values during
morning hours.
In the study population of men and women
survivors of acute MI of slightly, although sta-
tistically significantly different mean age, with
similar LVEF, site of MI, and frequency of non-
Q MI, TCRT and its strong association with car-
diac mortality did not differ significantly between
the sexes. It can be hypothesized that the mor-
phological and electrophysiological changes ac-
companying ischemic heart disease attenuate the
well-known sex differences in ventricular repolar-
ization. However, the healthy population in the
study of Smetana et al.8 was significantly younger
(average age of men and women 27.0 years and

PACE, Vol. 26 January 2003, Part II 375

 BATCHVAROV, ET AL.
26.8 years, respectively) than this study popula-
tion (57.0 years and 59.6 years, respectively). Ex-
perimental17,18 and clinical studies19 have sug-
gested that sex hormones may play a role in
the gender differences observed in ventricular
repolarization.
Conclusions
Depressed TCRT, an ECG descriptor of ven-
tricular repolarization, which revives the classi-
References
1. Moss AJ, Hall J, Cannom DS, et al., for the Multicenter Automatic
Defibrillator Implantation Trial Investigators. Improved survival
with an implanted defibrillator in patients with coronary disease
at high risk for ventricular arrhythmia. N Engl J Med 1996; 335:
1933–1940.
2. Buxton AE, Lee KL, Fisher JD, et al., for the Multicenter Unsus-
tained Tachycardia Trial Investigators. A randomized study of the
prevention of sudden death in patients with coronary artery dis-
ease. N Engl J Med 1999; 341:1882–1890.
3. Zipes DP, Wellens HJJ. Sudden Cardiac Death. Circulation 1998;
98:2334–2351.
4. Acar B, Yi G, Hnatkova K, et al. Spatial, temporal and wavefront
direction characteristics of 12-lead T wave morphology. Med Biol
Eng Comput 1999; 37:574–584.
5. Wilson FN, Macleod AG, Barker PS. The T deflection of the elec-
trocardiogram. Tr A Am Physicians 1931; 46:29–38.
6. Wilson FN, Macleod AG, Barker PS, et al. The determination and
the significance of the areas of the ventricular deflections of the
electrocardiogram. Am Heart J 1934; 10:46–61.
7. Zabel M, Acar B, Klingenheben T, et al. Analysis of 12-lead T-
wave morphology for risk stratification after myocardial infarction.
Circulation 2000; 102:1252–1257.
8. Smetana P, Batchvarov V, Hnatkova K, et al. Sex differences in
repolarisation homogeneity and its circadian pattern. Am J Physiol
Heart Circ Physiol 2002; 282:H1889-H1897.
9. Odemuyiwa O, Malik M, Farrell T, et al. Comparison of the pre-
dictive characteristics of heart rate variability index and left ven-
tricular ejection fraction for all-cause mortality, arrhythmic events
and sudden death after acute myocardial infarction. Am J Cardiol
1991; 68:434–439.
10. Malik M, Acar B, Gang Yi, et al. QT dispersion does not rep-
376 January 2003, Part II
cal concept of VG, was significantly strongly as-
sociated with increased long-term cardiac mor-
tality rate in male and female survivors of acute
MI. The gender difference in TCRT reported in
healthy young men and women were not present
in middle-aged post-MI patients. The role of
TCRT as a risk stratifier in survivors of acute
MI patients deserves further study in multivari-
ate regression models including established risk
factors.
resent electrocardiographic interlead heterogeneity of ventric-
ular repolarization. J Cardiovasc Electrophysiol 2000; 11:835–
843.
11. Hnatkova K, Ryan SJ, Bathen J, et al. T-wave morphology differ-
ences between patients with and without arrhythmic complication
of ischaemic heart disease. J Electrocardiol 2001; 34(Suppl.):113–
117.
12. Acar B, Koumen HL. SCD-based on-line exercise ECG signal or-
thogonalization. IEEE Trans Biomed Eng 1999; 46:311–321.
13. Abildskov JA, Burgess MJ, Millar K, et al. New data and concepts
concerning the ventricular gradient. Chest 1970; 58:244–248.
14. Surawicz B. ST-T abnormalities. In PW Macfarlane, TD Veitch-
Lawrie (eds.): Comprehensive Electrocardiology: Theory and Prac-
tice in Health and Disease. Oxford, Pergamon Press, Inc., 1989,
pp. 521–524.
15. Urie PM, Burgess MJ, Lux RL, et al. The electrocardiographic recog-
nition of cardiac states at high risk of ventricular arrhythmias. An
experimental study in dogs. Circ Res 1978; 42:350–358.
16. Abildskov JA, Evans AK, Lux RL, et al. Ventricular recovery proper-
ties and QRST deflection area in cardiac electrograms. Am J Physiol
1980; 239:H227–H231.
17. Liu XK, Katchman A, Drici MD, et al. Gender difference in the
cycle length-dependent QT and potassium currents in rabbits. J
Pharmacol Exp Ther 1988; 285:672–679.
18. Pham TV, Sosunov EA, Gainullin RZ, et al. Impact of sex and
gonadal steroids on prolongation of ventricular repolarization
and arrhythmias induced by Ik-blocking drugs. Circulation 2001;
103:2207–2212.
19. Rodriguez I, Kilborn MJ, Liu XK, et al. Drug-induced QT prolonga-
tion in women during the menstrual cycle. JAMA 2001; 285:1322–
1326.
PACE, Vol. 26