Methods

The study protocol was approved by the Bioethical Committee on Medical Research, School of Medicine, Kanazawa University. This study was carried out in accordance with the Declaration of Helsinki. Written informed consent was obtained from every patient for the administration of gadolinium-based contrast agents. HC was diagnosed by the presence of a non-dilated and hypertrophied LV on 2-dimensional echocardiography or CMR showing LVMWT at end-diastole ≥15 mm in the absence of other diseases that could account for the hypertrophy. Patients with infiltrative cardiomyopathies, such as cardiac amyloidosis or cardiac sarcoidosis, were excluded from the study. End-stage HC was defined as LV ejection fraction (LVEF) <50%. Patients with borderline hypertrophy (13 to 14 mm) were included in this study because progressive wall thinning can occur especially in end-stage HC. The study population comprised 108 consecutive patients with HC who underwent late gadolinium enhancement (LGE)-CMR at our institution from 2008 to 2015. We previously published studies on CMR-determined right ventricular hypertrophy in 106 patients with HC and on the relation between fragmented QRS complexes and CMR-determined fibrosis in 108 patients with HC. The same patients comprised the present study population. Because the presence of right or left bundle branch block can interfere with the ECG diagnosis of LV hypertrophy, 19 patients with HC with bundle branch block were excluded from the study, leaving 89 patients with HC for analysis.

The ECG voltage indexes were defined as follows on the basis of previous studies: total 12-lead QRS amplitude; sum of QRS voltages in all 12 leads, Cornell voltage index: RaVL + SV3 ; and Sokolow-Lyon voltage index: SV1 + RV5 or RV6, whichever is larger. Standard M-mode and 2-dimensional echocardiographic studies were performed to identify and quantify morphologic features of the LV in accordance with guidelines of the American Society of Echocardiography. CMR imaging was performed with a 1.5 T scanner (GE Medical Systems, Milwaukee, Wisconsin). The LV mass was assessed from the LV endocardial and epicardial contours by applying Simpson’s rule and multiplying the muscle volume with the density of myocardium (1.05 g/cm ³ ). Conventional 2-dimensional delayed enhancement images were obtained 10 minutes after intravenous administration of gadolinium–diethylenetriaminepentaacetic acid (0.2 mmol/kg), with the following setting: field of view 350 mm, matrix 224 × 192, slice thickness 8 mm (spatial resolution 1.6 × 1.8 × 8 mm), no gap between each slice, and the number of excitation 1. Areas of LGE at different standard deviations (SDs) (2, 4, 6, 8, and 10 SDs) above the normal myocardial signal were quantified as previously reported. Briefly, the LGE area was measured by manual planimetry in each short-axis image. The percentage area of LGE was calculated by dividing the sum of the LGE areas by that of the total LV area.

Values were expressed as the mean ± SD. Comparisons between continuous variables were made using the Student’s unpaired t test for parametric variables and the Mann-Whitney test for nonparametric variables. Categorical data were compared using the chi-square test. Receiver-operating characteristic analyses were used to estimate the discriminating ability for end-stage HC for each ECG voltage index. A Bonferroni correction was used for multiple comparisons in the area under the curve. Sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were defined as previously reported. Pearson’s correlation was used to assess the relation between the ECG voltage indexes and LV mass, LVMWT, or LVEF. Spearman’s rank correlation was used to assess the relation between the ECG voltage indexes and the extent of LGE. A p value <0.05 was considered as statistically significant, and all analyses were performed using JMP software, version 9.0.2 for Mac (SAS Institute, Cary, North Carolina).

Results

Baseline characteristics of the 89 patients are presented in Table 1 . We first examined which ECG voltage index had the best correlation with the extent of LV hypertrophy in the 89 patients with HC. Of the 3 ECG voltage indexes, only the Sokolow-Lyon index was correlated with the LV mass ( Figure 1 ). Instead, the total QRS voltage index had the best correlation with the LVMWT ( Figure 1 ) followed by the Cornell index ( Figure 1 ). We next examined whether ECG voltage indexes were associated with LV systolic function in HC. The Sokolow-Lyon index was best correlated with LVEF ( Figure 1 ) followed by the total QRS voltage index ( Figure 1 ). We then assessed the diagnostic values of the ECG voltage indexes for discriminating patients with end-stage HC (LVEF <50%) from those with preserved LVEF (LVEF ≥50%). Receiver-operating characteristic analysis revealed an excellent area under the curve for the total QRS voltage index and Sokolow-Lyon index ( Figure 2 , Table 2 ). The greatest sensitivity and specificity were obtained when the cut-off value was set at 16.8 mV for the total QRS voltage index and 2.3 mV for the Sokolow-Lyon index ( Figure 2 , Table 2 ).

Correlation between the different ECG voltage indexes and LV mass (A–C) , LVMWT (D–F) , or LVEF (G–I) in the patients with HC. Of the 3 ECG voltage indexes, only the Sokolow-Lyon index was correlated with the CMR-determined LV mass (r = 0.25, p = 0.017) ( C ). The total QRS voltage index had the best correlation with the LVMWT (r = 0.31, p = 0.0029) ( E ) followed by the Cornell index ( D ); the Sokolow-Lyon index had no correlation with the LVMWT ( F ). The Sokolow-Lyon index was best correlated with LVEF (r = 0.44, p <0.0001) ( I ) followed by the total QRS voltage index ( H ), but the Cornell index was not correlated with LVEF ( G ).

Diagnostic values of the ECG voltage indexes for discriminating patients with end-stage HC (LVEF <50%) from those with preserved LVEF (LVEF ≥50%). Receiver-operating characteristic analysis showed an area under the curve of 0.54 for the Cornell index, 0.87 for the total QRS voltage index, and 0.89 for the Sokolow-Lyon index. Area under the curve values of the total QRS voltage index and Sokolow-Lyon index were significantly greater than that of the Cornell index (total QRS voltage vs Cornell; p <0.0001, Sokolow-Lyon vs Cornell; p = 0.0002).

An increased extent of myocardial fibrosis has been associated with LV systolic dysfunction in HC. We next asked whether the total QRS voltage and Sokolow-Lyon indexes are correlated with the extent of LGE. The total QRS voltage index was not correlated with LGE at any threshold definition ( Table 3 , upper panel), whereas there was a negative correlation between the Sokolow-Lyon index and the extent of LGE measured at 2, 4, and 6 SDs ( Table 3 , upper panel, Figure 3 ). Examples of a low Sokolow-Lyon index in a patient with massive LGE together with a high Sokolow-Lyon index in a patient with focal LGE are shown in Figure 4 . The scatterplots of the ECG voltage indexes showed that voltages varied even at the same extent of LVMWT ( Figure 1 ). We then sought to investigate whether the total QRS voltage or Sokolow-Lyon index relative to LVMWT (mV/1 mm LVMWT) is better correlated with the extent of LGE than unadjusted voltage indexes. The total QRS voltage index/1 mm of LVMWT was negatively correlated with the extent of LGE measured at any threshold definition ( Table 3 , lower panel). Additionally, the Sokolow-Lyon index/1 mm of LVMWT had a better correlation with the extent of LGE measured at any threshold definition ( Table 3 , lower panel) compared with the unadjusted Sokolow-Lyon index ( Table 3 , upper panel). Correlations between the extent of LGE and the adjusted total QRS voltage index by LVMWT or the adjusted Sokolow-Lyon index by LVMWT were observed even after exclusion of 11 patients with end-stage HC (LVEF <50%; Table 4 ).

Table 3

Correlation between late gadolinium enhancement measured at different threshold definitions and total QRS voltage or Sokolow-Lyon index in the study patients with hypertrophic cardiomyopathy

Late Gadolinium Enhancement	Total QRS Voltage Index		Sokolow-Lyon Index
	Spearman’ s Coefficient	p Value	Spearman’ s Coefficient	p Value
Measured at 2 standard deviations	-0.12	0.26	-0.33	0.0015
Measured at 4 standard deviations	-0.099	0.35	-0.29	0.0056
Measured at 6 standard deviations	-0.072	0.51	-0.24	0.024
Measured at 8 standard deviations	-0.056	0.60	-0.20	0.055
Measured at 10 standard deviations	-0.042	0.69	-0.17	0.11

Late Gadolinium Enhancement	Total QRS Voltage Index /1 mm of Left Ventricular Maximal Wall Thickness		Sokolow-Lyon Index /1 mm of Left Ventricular Maximal Wall Thickness
	Spearman’ s Coefficient	p Value	Spearman’ s Coefficient	p Value
Measured at 2 standard deviations	-0.32	0.0026	-0.43	<0.0001
Measured at 4 standard deviations	-0.31	0.0032	-0.43	<0.0001
Measured at 6 standard deviations	-0.27	0.011	-0.39	<0.0001
Measured at 8 standard deviations	-0.24	0.025	-0.34	0.0009
Measured at 10 standard deviations	-0.22	0.037	-0.30	0.0037

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Effects of Polyunsaturated Fatty Acid Treatment on Postdischarge Outcomes After Acute Myocardial Infarction
2. Impact of Statin Use on Development of New-Onset Diabetes Mellitus in Asian Population
3. Usefulness of Von Willebrand Factor Activity Indexes to Predict Therapeutic Response in Hypertrophic Cardiomyopathy
4. The Role of Cardiac Imaging in the Evaluation of Newly Diagnosed Heart Failure

Stay updated, free articles. Join our Telegram channel

Join