Left ventricular hypertrophy is 1 of the most frequent cardiac manifestations associated with an unfavorable prognosis. However, many different causes of left ventricular hypertrophy exist. The aim of the present study was to assess the diagnostic value of common electrocardiographic (ECG) parameters to differentiate Fabry disease (FD), amyloidosis, and nonobstructive hypertrophic cardiomyopathy (HC) from hypertensive heart disease (HHD) and aortic stenosis (AS). In 94 patients with newly diagnosed FD (n = 17), HHD (n = 20), amyloidosis (n = 17), AS (n = 20), and HC (n = 20), common ECG parameters were analyzed and tested for their diagnostic value. A stepwise approach including the Sokolow–Lyon index, corrected QT duration, and PQ interval minus P-wave duration in lead II to overcome P-wave abnormalities was applied. A corrected QT duration <440 ms in combination with a PQ interval minus P-wave duration in lead II <40 ms was 100% sensitive and 99% specific for the diagnosis of FD, whereas a corrected QT duration >440 ms and a Sokolow–Lyon index ≤1.5 mV were found to have a sensitivity and specificity of 85% and 100%, respectively, for the diagnosis of amyloidosis and differentiation from HC, AS, and HHD. Moreover, a novel index ([PQ interval minus P-wave duration in lead II multiplied by corrected QT duration]/Sokolow–Lyon index) proved to be highly diagnostic for the differentiation of amyloidosis (area under the curve 0.92) and FD (area under the curve 0.91) by receiver operator characteristic analysis. In conclusion, a combined analysis of PQ interval minus P-wave duration in lead II, corrected QT duration, and Sokolow–Lyon index proved highly sensitive and specific in the differentiation of FD, amyloidosis, and HC compared to HHD and AS. Analysis of these easy-to-assess ECG parameters may be of substantial help in the diagnostic workup of these 5 conditions.
The aim of the present study was to investigate the value of common electrocardiographic (ECG) parameters in the diagnosis of patients with Fabry disease (FD), amyloidosis, hypertensive heart disease (HHD), aortic stenosis (AS), and nonobstructive hypertrophic cardiomyopathy (HC).
Methods
ECG and echocardiographic findings in 94 patients ≥18 years old with newly diagnosed FD, HHD, amyloidosis, HC, and AS were retrospectively studied. Patients with HHD and those with AS served as matched controls. All patients were selected from the echocardiographic database if electrocardiography was performed at the time of the echocardiography and if patients met the following criteria. In patients with FD the diagnosis was confirmed by low or missing activity of α1-galactosidase levels in leukocyte homogenates in men, mutational analysis of the α1-galactosidase genes, or kidney biopsy. Obligate heterozygous women were identified by pedigree analysis of families with FD. Importantly, none of the patients with FD was under enzyme replacement therapy at time of diagnosis. None of the patients had a known cardiac variant of FD. Clinical diagnosis of HHD was based on a long-standing history of arterial hypertension and confirmation by clinical assessment and ambulatory blood pressure monitoring with a systolic blood pressure >140 mm Hg and a diastolic blood pressure >90 mm Hg and increased left ventricular muscle mass index without any other evident cause. Patients with amyloidosis were included only if amyloidosis was confirmed by biopsy. Left ventricular hypertrophy owing to AS was diagnosed if a significant pressure gradient (mean pressure gradient >20 mm Hg) was found in association with left ventricular hypertrophy. Diagnosis of HC was confirmed by genetic analysis in 15 patients. In the remaining 5 patients diagnosis was based on patient and family histories, typical echocardiographic findings, and clinical exclusion of other differential diagnoses. Groups were matched for age and left ventricular muscle mass index to avoid their influence on any ECG parameters of interest. Accordingly, heart rate-dependent indexes (PQ and QT intervals) were adjusted. Patients with previous pacemaker implantation, atrial fibrillation, regional wall motion abnormalities, and/or history of coronary artery disease were excluded from the study. Furthermore, in patients with FD, amyloidosis, and AS, presence of arterial hypertension was excluded by ambulatory blood pressure monitoring. Twelve-lead surface electrocardiograms at initial diagnosis were independently analyzed by 2 experienced readers (M.N. and J.S.). Observers were blinded to the cause and stage of disease and the ECG reading was performed by consensus reading. Measurements were taken manually at a sweep of 25 mm/s and, if necessary, 50 mm/s. A normal PQ interval was defined as 120 to 200 ms. A normal QRS duration was defined as 70 to 110 ms. Left ventricular hypertrophy was assessed using the Sokolow–Lyon index adding the S wave in lead V 1 or V 2 (whichever was larger) and the R wave in lead V 5 or V 6 (whichever was larger) with a cutoff of 3.5 mV. Based on criteria described by Klein et al, low voltage was defined as the sum of the S wave in lead V 1 plus the R wave in lead V 5 or V 6 <1.5 mV. Because the different groups were not matched for heart rate, we integrated in our analysis a corrected PQ interval (corrected PQ interval = PQ interval/RR interval ½ ). The PQ interval minus P-wave duration in lead II to overcome P-wave abnormalities was used as an expression of left and/or right atrial enlargement in the course of left ventricular hypertrophy. The QT interval (normal 300 to 440 ms) was measured from the beginning of the QRS complex to the end of the T wave, defined as the intersection of the tangent to the downslope of the T wave and the isoelectric line. Corrected QT duration (normal <440 ms) was calculated using the Bazett formula. QT/corrected QT dispersion was defined as the difference between the maximum and minimum QT/corrected QT duration of the 12 leads. Duration from the peak to the end of the T wave was measured in each precordial lead. For negative or biphasic T waves, peak of the T wave was measured from the nadir of the T wave. Single leads with T waves <1.5 mm in amplitude were not included in the analysis. Furthermore, we introduced a posteriori a novel index consisting of these parameters and tested it for differentiation of the cardiopathies: (PQ interval minus P-wave duration in lead II multiplied by corrected QT duration)/Sokolow–Lyon index (milliseconds squared per millivolt). For simple handling and implementation in diagnostic algorithms, the resulting index is divided by 1,000.
All echocardiographic examinations were performed by experienced operators (R.J. and M.N.). Standard 2-dimensional M-mode echocardiogram from a parasternal long-axis view was used for measurements of left ventricular chamber dimensions including ventricular septal and posterior wall thicknesses, diameter of the left atrium, and sinus of Valsalva. Left ventricular muscle mass was calculated with the formula of Devereux et al. Left ventricular hypertrophy was defined as a left ventricular muscle mass index >134 g/m 2 in men and >110 g/m 2 in women. Left ventricular volumes were calculated using the area–length method. The study was approved by the local ethics committee.
Continuous variables were compared by analysis of variance for repeated measurements with post hoc analysis using Fisher’s probable least-significant differences test and/or Student’s t test. Categorical variables were compared using Fisher’s exact test. Receiver operating curve analysis was performed to test for the diagnostic performance of different ECG parameters. Analyses were performed using SPSS 17.0 (SPSS, Inc., Chicago, Illinois). A p value <0.05 was considered statistically significant.
Results
Ninety-four patients with newly diagnosed FD (n = 17), HHD (n = 20), amyloidosis (n = 17), HC owing to significant AS (n = 20), and HC (n = 20) were included in the study. Demographic data and ECG measurements are presented in Table 1 . No patient had signs of preexcitation or atrioventricular or bundle branch block. Significant differences were observed in several ECG parameters with and without correction for heart rate among the patient groups. The most striking feature of ECG FD was a significantly shorter PQ interval (p <0.0001). Furthermore, the interval from the end of the P wave to the beginning of the QRS complex (PQ interval minus P-wave duration in lead II) was significantly shorter in these patients (p <0.0001). However, the longer PQ interval in particular in patients with HHD and HC indicated that this occurred mainly owing to a longer P-wave duration (p <0.01). The amyloidosis group showed a lower Sokolow–Lyon index (p <0.01), a more pronounced dispersion from the peak to the end of the T wave (p <0.05), and a longer mean corrected QT duration (p <0.05). The latter was also observed in HC (p <0.05). Table 2 presents echocardiographic data of all patient groups. None of the parameters demonstrated a significant difference. Of the analyzed ECG parameters, corrected PQ interval (sensitivity 82%, specificity 90%, cut-off value 144 ms, area under the curve 0.90) and PQ interval minus P-wave duration in lead II (sensitivity 82%, specificity 99%, cut-off value 40 ms, area under the curve 0.94) showed the highest diagnostic performance for the diagnosis of FD as assessed by receiver operating characteristics curve analysis. For the diagnosis of amyloidosis, the Sokolow–Lyon index (sensitivity 76%, specificity 82%, cut-off value 1.85 mV, area under the curve 0.83) and mean corrected QT duration (sensitivity 94%, specificity 52%, cut-off value 416 ms, area under the curve 0.80) demonstrated the best diagnostic performance ( Figure 1 , Tables 3 and 4 ). In contrast, cut-off values currently used for “low voltage” (i.e., ≤1.5 mV) and “long corrected QT duration” (i.e., corrected QT duration >440 ms) had poor diagnostic power when used as a single parameter. Indeed, sensitivity and specificity for diagnosis of amyloidosis changed to 47% and 95%, respectively, with a Sokolow–Lyon index ≤1.5 mV, whereas mean corrected QT duration >440 ms was 65% sensitive and 75% specific. However, the combination of these 2 criteria revealed a sensitivity of 85% and a specificity of 100% for the diagnosis of amyloidosis. For FD a combination of a normal corrected QT duration (i.e., <440 ms) and short PQ interval minus P-wave duration in lead II (i.e., <40 ms) demonstrated superior diagnostic performance (sensitivity 100%, specificity 99%). Figure 2 shows the 2-step algorithm for the diagnosis of amyloidosis and FD using these combined ECG parameters. While testing the novel index for its diagnostic performance, the resulting receiver operating characteristics curve analysis showed at different cut-off values for amyloidosis (cut-off value 15.4 ms /mV, area under the curve 0.92) and FD (cut-off value 6.1 ms /mV, area under the curve 0.91) a sensitivity of 88% and specificity of 78% for FD and 94% and 75% for amyloidosis, respectively ( Figure 1 ).
Variable | FD | HHD | Amyloidosis | AS | HC |
---|---|---|---|---|---|
(n = 17) | (n = 20) | (n = 17) | (n = 20) | (n = 20) | |
Age (years) | 50 ± 7 | 52 ± 11 | 56 ± 15 | 58 ± 17 | 51 ± 9 |
Men | 10 (59%) | 12 (60%) | 10 (59%) | 13 (65%) | 12 (60%) |
Resting heart rate (beats/min) | 62 ± 11 ⁎ | 77 ± 20 | 86 ± 19 | 78 ± 15 | 79 ± 11 |
P-wave duration (ms) | 98 ± 18 | 105 ± 15 † | 83 ± 26 | 83 ± 17 | 101 ± 12 † |
P wave corrected for heart rate (ms) | 100 ± 23 | 117 ± 19 ‡ | 99 ± 31 | 95 ± 23 | 115 ± 11 ‡ |
PQ interval (ms) | 130 ± 20 ⁎ | 167 ± 22 | 172 ± 44 | 159 ± 34 | 164 ± 23 |
Corrected PQ interval (ms) | 131 ± 21 ⁎ | 187 ± 33 | 204 ± 52 | 181 ± 47 | 183 ± 41 |
PQ interval <120 ms ⁎ | 4 (24%) | 0 | 0 | 0 | 0 |
PQ interval minus P-wave duration in lead II (ms) | 31 ± 16 ⁎ | 70 ± 31 | 105 ± 36 | 87 ± 36 | 81 ± 26 |
QRS duration (ms) | 98 ± 16 | 91 ± 14 | 91 ± 32 | 96 ± 23 | 94 ± 19 |
Sokolow–Lyon index (mV) | |||||
Left ventricular hypertrophy | 2.9 ± 1.2 | 2.5 ± 0.9 | 1.7 ± 0.9 † | 3.2 ± 1.1 | 2.9 ± 1.1 |
Right ventricular hypertrophy | 1.0 ± 0.4 ‡ | 1.5 ± 0.6 | 1.3 ± 0.7 | 1.5 ± 0.5 | 1.6 ± 0.4 |
Maximum peak to end of T wave (ms) | 34 ± 14 | 31 ± 20 | 49 ± 38 ‡ | 46 ± 17 | 49 ± 27 ‡ |
Corrected QT mean interval (ms) | 411 ± 45 | 416 ± 47 | 455 ± 43 ‡ | 418 ± 34 | 448 ± 14 |
Corrected QT dispersion (ms) | 60 ± 46 | 71 ± 40 | 91 ± 66 | 80 ± 38 | 85 ± 46 |
Variable | FD | HHD | Amyloidosis | AS | HC | p Value ⁎ |
---|---|---|---|---|---|---|
(n = 17) | (n = 20) | (n = 17) | (n = 20) | (n = 20) | ||
Sinus of Valsalva (mm) | 36 ± 5 | 36 ± 5 | 36 ± 5 | 36 ± 6 | 36 ± 4 | 0.9 |
Interventricular septal thickness (mm) | 16 ± 5 | 14 ± 3 | 16 ± 3 | 16 ± 2 | 15 ± 6 | 0.5 |
End-diastolic volume index (ml/m 2 ) | 47 ± 10 | 63 ± 23 | 49 ± 18 | 69 ± 18 | 49 ± 12 | 0.1 |
Left ventricular mean mass index (g/m 2 ) | 178 ± 67 | 153 ± 25 | 168 ± 35 | 174 ± 51 | 171 ± 42 | 0.1 |
Left ventricular ejection fraction (%) | 70 ± 7 | 61 ± 7 | 60 ± 10 | 61 ± 12 | 69 ± 11 | 0.3 |
Left atrium (mm) | 38 ± 8 | 42 ± 8 | 46 ± 8 | 43 ± 8 | 44 ± 7 | 0.2 |
Mean pressure gradient (mm Hg) | — | — | — | 44 ± 23 | — | — |
Aortic valve area (cm 2 ) | — | — | — | 1.1 ± 0.6 | — | — |
⁎ Comparison by analysis of variance for repeated measurements among patient groups.
Parameter | TP | FN | FP | TN | Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | Accuracy (%) |
---|---|---|---|---|---|---|---|---|---|
Fabry disease | |||||||||
Corrected PQ interval (<144 ms) | 14 | 3 | 8 | 69 | 82 | 90 | 64 | 95 | 88 |
PQ interval minus P-wave duration in lead II (<40 ms) | 14 | 3 | 1 | 76 | 82 | 99 | 93 | 96 | 96 |
Amyloidosis | |||||||||
Sokolow–Lyon index (<1.85 mV) | 13 | 4 | 14 | 63 | 76 | 82 | 48 | 94 | 81 |
Mean corrected QT interval (>416 ms) | 16 | 1 | 37 | 40 | 94 | 52 | 30 | 98 | 60 |