A New Electrocardiographic Marker of Hypertensive Cardiac Damage

In this issue of the Journal, Ferrucci and coworkers1 report the results of a cross‐sectional study aimed at assessing the value of conventional (ie, Sokolow‐Lyon, Cornell voltage, and product criteria) and new electrocardiographic indexes (ie, ventricular activation time, P‐wave analysis, peak to end time interval of T wave) as markers of subclinical cardiac damage in newly diagnosed untreated essential hypertensive patients with no echocardiographic evidence of left ventricular (LV) hypertrophy (LVH). The authors were able to show that LV mass index (LVMI) was significantly higher (+12 g/m²) in hypertensive patients than in their normotensive counterparts and, more important, that LV mass increase in the former group was paralled by an increase in the time interval between the peak and the end of the T wave (Tp‐Te interval). Notably, no differences were documented between groups in traditional electrocardiographic (ECG) criteria, as well as in ventricular activation time (VAT) and P‐wave parameters. Before addressing these findings in detail, some general considerations on available evidence in this area might be useful.

Target organ damage (TOD) related to hypertension is currently regarded as an intermediate stage in the continuum of vascular disease and a powerful determinant of total cardiovascular risk independent from blood pressure (BP) levels and conventional risk factors.2 Several markers of TOD (ie, QRS voltages, LVMI, carotid intima‐media thickness, pulse wave velocity, ankle‐brachial index, glomerular filtration rate, and urinary albumin excretion) have been shown to be associated with increased incidence of cardiovascular events.3, 4, 5 Because of the role of high BP per se or associated risk factors in determining subtle structural and functional alterations in target organs, signs of organ involvement should be carefully investigated in the initial evaluation of each hypertensive patient in order to quantify total cardiovascular risk. This point is of paramount relevance as the management of hypertension largely depends on global cardiovascular risk, which may range from low to very high levels.

In the past decades, most attention has focused on LVH as a cardinal sign of hypertensive TOD because of the high prevalence of this phenotype and its association with adverse cardiovascular outcomes. Numerous studies, indeed, have consistently shown that LVH, determined either by standard 12‐lead ECG or echocardiography, is a strong predictor of myocardial infarction, congestive heart failure, sudden cardiac death, stroke, and total mortality in the general population and in specific groups, such as patients with systemic hypertension, coronary heart disease, congestive heart failure, chronic renal failure, and atrial fibrillation.6, 7, 8 LVH can be diagnosed by several methods with different sensitivity and specificity.

Standard 12‐lead ECG is the first‐line method, owing to its limited cost, wide availability, and good reproducibility. Unfortunately, ECG has poor sensitivity as documented by numerous studies where LV mass was also estimated by echocardiography or, less frequently, by computerized tomography or magnetic resonance. In a review of 21 studies aimed at assessing the accuracy of six ECG criteria compared with echocardiographc parameters in the hypertensive setting, Pewsner and coworkers9 found that median sensitivity ranged from 10% for Gubner index to 21% for Sokolow‐Lyon index and median specificity ranged from 89% for Sokolow‐Lyon index to 99% for Romhilt‐Estes score.

Since the pioneering contributions by Lewis and Einthoven, dozens of ECG voltage and non‐voltage criteria have been proposed for LVH diagnosis in the attempt to improve sensitivity without affecting specificity.10 These criteria have shown varying performance characteristics in identifying LVH, particularly in relation to body size, age, ethnicity, and sex. Substantial evidence exists regarding the impact of body size on diagnostic accuracy of voltage‐based ECG criteria for LVH: patients with low body mass index (BMI) show increased QRS voltages; conversely, obese patients are less likely to display elevated voltages, despite obesity being a powerful determinant of cardiac hypertrophy. The usefulness of QRS voltage correction by BMI in order to ameliorate ECG detection of LVH in hypertensive patients has been recently tested by Angeli and coworkers11 in a study in 2747 patients with untreated essential hypertension. The authors found that amplification of Cornell voltage by BMI improved ECG performance for LVH diagnosis. The BMI‐corrected Perugia score (ie, Cornell BMI product and/or typical strain pattern) was associated with significantly higher values of area under the curve compared with traditional LVH criteria. Among tested criteria, the new score provided the highest sensitivity and specificity values (36.1% and 90.5%, respectively).

In recent years, many studies have addressed the value of new ECG markers in detecting subclinical cardiac damage and predicting cardiovascular outcome. Fragmented QRS complexes (ie, presence of additional R wave, notching of R or S wave) have been shown to be associated with myocardial fibrosis or LVH. Kadi and coworkers demonstrated that LVMI was significantly higher among 45 hypertensive patients with fragmented QRS complexes as compared with 45 age‐ and sex‐matched patients with normal ECG findings.12 In the logistic regression analysis, fragmented QRS complexes independently correlated with LVMI. Increased P‐wave duration and P‐wave dispersion (ie, difference between maximum and minimum P‐wave duration in simultaneously recorded 12‐lead ECG) have been linked to hypertension, obesity, LVH, left atrial enlargement, and increased risk of incident atrial fibrillation. A close relationship between P‐wave dispersion and BP levels, left atrial area, and LVMI has been reported by Chavez and coworkers in the pediatric setting.13 The authors showed that P‐wave dispersion was increased in hypertensive (40±12 ms) and prehypertensive children (37±9 ms) as compared with their normotensive counterparts (31±9 ms). Both left atrial area and LVMI exhibited a progressive increase from normotensive to prehypertensive and hypertensive children.

Tp‐Te interval is an index of transmural dispersion of repolarization and an emerging new marker for ventricular arrhythmogenesis and repolarization heterogeneity. Prolongation of this interval has been described in pathological settings such as hypertrophic cardiomyopathy, post‐myocardial infarction, long QT syndrome, inducible ventricular tachycardia, end‐stage renal disease, repaired tetralogy of Fallot, Brugada syndrome, mitral valve prolapse, and Chagas disease.14 Importantly, prolonged Tp‐Te interval has been associated with increased risk of mortality in most of the above‐mentioned conditions. Information on this parameter in systemic hypertension is scanty and available findings not univocal. Sauer and coworkers examined 84 consecutive unselected patients referred for exercise echocardiography and found an association between impaired LV diastolic function and Tp‐Te interval above the median; this was not the case for hypertension or LVMI.15 Karagaac and coworkers investigated the relationship between BP circadian variations and Tp‐Te interval and Tp‐Te/QT ratio in 70 patients with newly diagnosed hypertension who fulfilled the metabolic syndrome criteria.16 Tp‐Te (91±12 ms vs 74±10 ms), Tp‐Te/QT (0.24±0.02 vs 0.20±0.02), and Tp‐Te/QTc (0.22±0.02 vs 0.18±0.02) were significantly higher in the nondipper than in the dipper group, respectively. Of note, altered ventricular repolarization was unrelated to LVMI, LV geometry, and diastolic function, as all these variables were similar in dippers and nondippers.

The report by Ferrucci and coworkers1 provides a new piece of evidence on the association between prolonged Tp‐Te interval and systemic hypertension by showing that this index was significantly higher (2.9±0.5 mm) in the hypertensive group compared with the normotensive group (2.2±0.3 mm). Moreover, in multivariable analysis, Tp‐Te interval emerged as the only independent correlate of hypertension, whereas BMI, LVMI, and Em/Am ratio failed to maintain a significant association. The authors included in their analysis 32 never‐treated hypertensive patients (mean age, 44±9 years; BMI 27.0±2.6 kg/m²) without previous cardiovascular diseases, chronic kidney disease, or arrhythmias, and 18 normotensive controls (mean age, 42±10 years; BMI 24.7±5.3 kg/m²); the two groups were carefully selected on the basis of both clinic and 24‐hour ambulatory BP monitoring. LVM indexed to body surface area and to height to the allometric power of 2.7 was approximately 18% and 23% higher in the hypertensive group than in the control group, respectively. Seven hypertensive patients (23%) were found to have LV concentric remodelling (ie, increased relative wall thickness and normal LV mass) and the remaining 67% had normal LV geometry. The early LV structural abnormalities documented in the hypertensive patients were unrelated to prognostically validated measures of LV diastolic dysfunction, as both E/A and Em/Am ratio did not differ between groups. This was also the case for left atrium area, a reliable marker of chronically elevated LV filling pressure and of diastolic dysfunction.

A relevant aspect of the study by Ferrucci and coworkers1 is that no differences in a number of convential and new ECG indexes for LVH were found between normotensive and hypertensive individuals. Indeed, several conventional ECG parameters (such as Sokolow‐Lyon index, Cornell voltage index, Cornell voltage product index, and QT and QTc interval) and P‐wave dispersion, P‐wave area, and VAT did not significantly differ between the groups. Overall, these findings may support the view that Tp‐Te interval is a more sensitive biomarker of pressure overload and of early myocardial changes related to hypertension as compared with all of the above‐mentioned ECG parameters. A few additional points deserve to be mentioned. In the entire population, 11 patients (23%) with normal LVMI fulfilled LVH criteria according to Cornell voltage index, four (8%) according to Sokolow‐Lyon index, and six (12%) according to Cornell product index. This finding may be related to the fact that the study included patients younger than 55 years, potentially leading to false‐positive results, as high voltages may be frequently found in young/middle‐aged patients even in the absence of echocardiographic LVH.17 As Tp‐Te interval has been related to sex and heart rate,18 investigations on the clinical value of this parameter should carefully take into account these variables. Unfortunately, no information was provided by the present study on sex distribution in the two groups.

Conclusions

The study by Ferrucci and coworkers supports the concept that prolonged transmural repolarization may occur in early stages of systemic hypertension and precede frank LVH development. The usefulness of Tp‐Te interval for detecting cardiac TOD and improving cardiovascular risk stratification needs to be re‐examined in larger population samples and in different ethnic groups unbiased by inclusion/exclusion criteria such as age, echocardiographic LVH, and chronic kidney disease.

Disclosure

The authors report no conflicts of interest.