Early noninvasive identification of cardiac amyloidosis (CA) is of growing clinical importance. Low voltage on electrocardiogram (ECG), increased left ventricular (LV) septal thickness (ST), and global longitudinal strain (GLS) on echocardiography, and elevated brain natriuretic peptides (BNP) are used as surrogates of CA. Thirty-five patients (50 ± 14 years, 22 women) underwent electrocardiography to analyze low-voltage QRS (<15 mV) pathologic Q waves, poor R-wave progression, ST-T abnormalities, and left bundle branch block. An ECG was considered abnormal if at least one ECG alteration was present. Echocardiography was used to analyze LVST, E/E′, and GLS. All participants also had BNP blood testing. 99m Tc-3,3-diphosphono-1,2 propanodicarboxylic acid scintigraphy assumed as a reference method showed CA in 18 patients (51%, CA group) and no accumulation in 17 patients (no CA group). In descending order of accuracy, LVST >14 mm, E/E’ >6.6, GLS <14.1, BNP >129 pg/ml, and an overall abnormal ECG showed good capability to distinguish patients with and without CA. All these parameters were predictors of CA in univariate analysis, whereas low-voltage QRS showed the worst performance. LVST >14 mm (p = 0.002) was the best independent predictor of CA, achieving sensitivity of 78% and accuracy of 89%. However, an LVST >14 mm (p = 0.005) plus an abnormal ECG (p = 0.03) show together a greater sensitivity, equal to 89%, in identifying CA. An integrated evaluation of ECG and echocardiography is a sensitive and low-cost technical approach to identify CA in patients with transthyretin gene mutation.
Cardiac amyloidosis (CA) is a rare secondary cardiomyopathy with a “hypertrophic” appearance due to myocardial extracellular matrix deposition of fibrillar anomalous proteins originated from a precursor-altered protein called “amyloid.” Amyloidosis is classified into several subtypes, which are differentiated on the basis of the amyloid precursor protein. Of many subtypes of systemic or localized amyloidosis (light chain amyloidosis, systemic AA amyloidosis, and systemic senile amyloidosis), familial amyloidotic polyneuropathy (FAP) is an autosomal dominant disease caused by a mutation of the gene coding for transthyretin (TTR). FAP is very rare (prevalence in the United States is 1:100,000), even though in some regions of Portugal and Sweden, higher prevalence values (reaching 3% to 5% of the population) are reported. This subtype of CA evolves in 2 phases: a first phase characterized by the existence of the genetic mutation without cardiac TTR deposition but a potential significant involvement of extracardiac tissue (often nerve tissue) and a second phase characterized by cardiac deposition of TTR. The age of onset of symptoms due to TTR deposition is variable, ranging from the third to the seventh decade of life. The early identification of myocardial TTR deposition represents a crucial key in the management of and prognosis for patients with FAP. In clinical practice, many markers (such as low voltage on electrocardiogram [ECG], increased left ventricular [LV] thickness and longitudinal systolic dysfunction on echocardiography, and elevated N-terminal pro–brain natriuretic peptides [BNP]) are used as surrogate indicators of cardiac amyloid deposition. 99m Tc-3, 3-diphosphono-1,2 propanodicarboxylic acid (DPD) scintigraphy has demonstrated high accuracy in direct identification of amyloid deposition in the myocardium of patients with TTR-related amyloidosis, allowing an early diagnosis of the disease. The aim of our study was to assess whether an integrated diagnostic approach consisting of low-cost and nonionizing techniques (ECG, echocardiography, and BNP blood sampling) could increase the diagnostic potential of CA, assuming 99m Tc-DPD scintigraphy as the reference technique in a series of patients with FAP with no previous history of cardiac disease.
Methods
We have included 36 consecutive patients with TTR-FAP belonging to 7 unrelated families. The study was approved by our institutional review committee. Informed consent was obtained from all patients.
Diagnosis was based on genetic testing for TTR-FAP: the Glu89Gln mutation was detected in 20 patients, the Phe64Leu mutation was detected in 12 patients, and the Thr49Ala mutation was detected in 4 patients. None of the included patients had evidence of monoclonal protein in the serum or urine, a monoclonal population of plasma cells in the bone marrow, or other diseases that could be responsible for secondary amyloidosis. Of these 36 patients, 1 (with mutation Glu89Gln) was excluded because of a previous myocardial infarction. Therefore, the final evaluated population consisted of 35 patients (13 men and 22 women; mean age 50 ± 14 years; age range 32 to 78 years). Of these, 23 patients (66%) had somatic polyneuropathy. Each patient underwent, in the same day, the following examinations: ECG, BNP dosage, 2-dimensional Doppler, strain echocardiography, and a 99m Tc-DPD scan. At enrollment, all patients were asymptomatic or paucisymptomatic according to New York Heart Association (NYHA) functional class and had no clinical history of cardiac disease.
Twelve-lead ECGs were analyzed offline, according to predefined criteria, by 2 independent cardiologists (G.D.B. and C.Z.) unaware of laboratory, echocardiographic, or scintigraphic data. In case of disagreement on ECG interpretation, the 2 observers reanalyzed the tracing in consensus. The considered parameters were rhythm (sinus, supraventricular, or ventricular); heart rate; PR, QRS, QT, and QTc intervals; presence of abnormal Q waves in leads other than aVR (duration >0.04 seconds; voltage >1/3 of the R wave except in leads V 2 to V 3 , where Q waves have been considered abnormal regardless of their voltage and duration); 2 or more abnormal Q waves was considered as a pseudonecrosis pattern; low QRS voltages in peripheral leads (considered to be low when QRS amplitude was <5 mm in height in all limb leads); and poor R-wave progression in the precordial leads, defined as loss of, or no, R waves in leads V 1 to V 3 . Furthermore, ST-segment appearance (normal, elevated/depressed by 1 mm or more); the T waves’ appearance, classified as normal (positive in all leads apart from III, aVR, V 1 , with voltage >0.1 mV), inverted (negative in any lead except III, aVR, V 1 , with voltage >0.1 mV), or flat (voltage < 0.1 mV); and the presence of left bundle branch block (LBBB) were analyzed in all 12 leads.
Standard echocardiographic examinations were performed in all patients using a commercial ultrasound machine (My Lab ALFA; Esaote, Florence, Italy). Parasternal short-axis views at the basal, mid, and apical levels and 4-chamber, 2-chamber, and LV outflow long axis were acquired. LV dimensions (end-diastolic and end-systolic diameters), diastolic thickness of the LV anterior septal thickness (LVST), posterior wall thickness, LV volumes, and ejection fraction (EF) were measured according to the recommendations of the American Society for Echocardiography. LV hypertrophy, defined as mean value of LVST and posterior wall thickness >12 mm, was used as the conventional echocardiographic marker of CA. LV mass was calculated using the Devereux formula. LV diastolic function was quantified in normal, impaired relaxation, pseudonormal, and restrictive patterns according to the recommendations of the American Society for Echocardiography. A 16-segment model was used to divide the LV. A dedicated software package (XStrain, Esaote) was used for an offline quantification of longitudinal strain on 4- and 2-chamber apical views. Global longitudinal strain (GLS) is obtained as the average of 12-segmental longitudinal strain.
A whole-body scan (anterior and posterior projections) was obtained 3 hours after the intravenous injection of 740 MBq of 99m Tc-DPD using a dual-headed gamma camera (MillenniumVG; GE Healthcare, Milwaukee, Wisconsin). Moreover, a thoracic single-photon emission computed tomography (SPECT) scan was obtained in each patient immediately after the whole-body scan, using the same machine. The whole-body scan and transaxial, sagittal, and coronal 99m Tc-DPD SPECT images were visually evaluated in consensus by 2 blinded nuclear medicine physicians in search of cardiac radiotracer accumulation. Cardiac 99m Tc-DPD accumulation was graded as follows: grade 0, absent cardiac uptake and normal bone uptake; grade 1, mild cardiac uptake, inferior to bone uptake; grade 2, moderate cardiac uptake accompanied by attenuated bone uptake; grade 3, strong cardiac uptake with mild/absent bone uptake.
Plasma samples for troponine I and N -terminal pro-BNP measurements were collected on the same day as ECG, echocardiography, and scintigraphy. Plasma N -terminal pro-BNP levels were measured using a validated and commercially available immunoassay kit (Perkin Elmer, Massachusetts). Quantitative data were expressed as mean ± standard deviation and qualitative data as frequency and percentage. A log10 transformation was performed, and BNP values were then normally distributed, which is in accordance with previous data. Receiver operating characteristic (ROC) curves were calculated, and areas under the curves (AUCs) were derived to determine the best cut-off values of BNP, LVST, GLS, and E/E′ for optimal sensitivity and specificity according to the Youden Index. The performance of each parameter or the combination of 2 of these was evaluated by the AUC value of each ROC curve. Pairwise comparison of ROC curves was then carried out to test the statistical significance of the difference among the AUCs. The performance indexes of each ROC curve analysis included sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and kappa concordance index. The difference between groups was tested by analysis of variance or Welch. Variables were submitted to univariate logistic regression analysis to identify predictors of CA. Variables showing a significant association in univariate analysis were included in a multivariate logistic regression model using the forward selection algorithm on the basis of the Wald statistic. A p value <0.05 was considered to be statistically significant. All tests were 2 tailed. Statistical analyses were carried out using JMP statistical software version 4.0.0 (SAS Institute Inc.) and MedCalc 6.00.014 (MedCalc Software, Mariakerke, Belgium).
Results
Whole-body scan and SPECT revealed 99m Tc-DPD accumulation (CA group) in 18 of 35 patients (51%) and no cardiac radiotracer accumulation (no CA group) in the remaining 17 patients (49%). In the CA group, cardiac 99m Tc-DPD accumulation was graded as follows: grade 1, n = 6; grade 2, n = 8; grade 3, n = 4.
The age of patients in the CA group (52 ± 11 years) and no CA group (49 ± 14 years) was similar (p = 0.51). A higher prevalence of male sex (68% vs 6%, p <0.001) and somatic polyneuropathy (84% vs 44%, p = 0.01) was observed in the CA group compared to the no CA group. Only 3 patients in the CA group complained of mild shortness of breath on effort. The only patient with abnormal troponine I was in the CA group. Log-transformed BNP values were significantly greater in the CA group than in the no CA group (2.62 ± 0.84 vs 1.71 ± 0.33; p <0.0001). BNP showed a good predictive value (AUC 0.85, 95% confidence interval [CI] 0.69 to 0.95, p <0.001) in distinguishing between the 2 groups. A BNP cut-off point of >129 pg/ml (abnormal BNP) yielded the best result in terms of combined sensitivity (67%) and specificity (94%). A BNP level >129 pg/ml was found in 13 of 35 patients (37%); of these, 12 of 13 (92%) were in the CA group and only 1 subject (8%) was in the no CA group.
Findings in the 2 groups are reported in Table 1 . Heart rate and PR segment duration were higher in the CA group than the no CA group (p = 0.004 and 0.01, respectively). Poor R-wave progression, number of Q waves, and pseudonecrosis pattern were all significantly more frequently observed in the CA group with respect to the no CA group. No differences in prevalence of QRS voltage, LBBB, and ST-segment abnormalities were found between the CA and no CA groups (p >0.05). Low-voltage QRS was found in 14 of 35 (40%) of the overall patients; of these, 10 (71%) were in the CA group and 4 (29%) were in the no CA group. The incidence of abnormal ECG (poor R-wave progression, pseudonecrosis, LBBB, QRS voltage <15 mV) was significantly higher (p <0.0001) in the CA group than the no CA group. An abnormal ECG was found in 14 of 35 patients (40%); of these, 13 (93%) were in the CA group and only 1 (7%) in the no CA group.
Variable | Cardiac Amyloidosis | P | |
|---|---|---|---|
| Yes (n = 18) | No (n = 17) | ||
| Heart rate (bpm) | 82 ± 12.3 | 66 ± 51.05 | 0.004 |
| PR interval (msec) | 184 ± 24.27 | 160 ± 26.11 | 0.009 |
| QRS (msec) | 87 ± 14.37 | 81 ± 3.75 | NS |
| Q waves | 44% | 0% | 0.002 |
| Pseudonecrosis pattern | 28% | 0% | 0.03 |
| Poor R wave progression | 44% | 0% | 0.003 |
| QTc (msec) | 445 ± 31.8 | 432 ± 21.41 | NS |
| ST abnormalities | 33% | 27% | NS |
| Left bundle branch block | 11% | 0% | NS |
| QRS voltage (mV) | 18 ± 7 | 18 ± 2 | NS |
| QRS voltage < 15 mV | 53% | 25% | NS |
| Abnormal electrocardiogram pattern (mV < 15, left bundle branch block, pseudonecrosis, poor R wave progression) | 68% | 6% | <0.0001 |
As reported in Table 2 , the CA group had many statistically significant differences in morphologic and functional echocardiographic features in comparison with the no CA group. Namely, the CA group showed greater values of LV mass (p <0.001), anterior septal thickness and posterior wall thickness (p <0.0001), diastolic dysfunction (E/E′ ratio and prevalence of pseudonormal and restrictive pattern), and pericardial effusion (p = 0.01) than the no CA group. Particularly, pseudonormal and restrictive pattern were found in 2 and 3 patients, respectively; all these patients were in the CA group. Global LV longitudinal dysfunction was greater (p <0.001) in the CA group than in the no CA group. Anterior septal thickness showed an excellent predictive value (AUC 0.92, 95% CI 0.78 to 0.96) in distinguishing between the 2 groups. An anterior septal thickness cut-off point of >14 mm yielded the best result in terms of combined sensitivity (78%) and specificity (100%). Anterior septal thickness >14 mm was found in 14 of 35 patients (40%), all belonging to the CA group. Similarly, E/E′ showed a high predictive value (AUC 0.90, 95% CI 0.75 to 0.98) in distinguishing between the 2 groups. An E/E′ cut-off point of 6.6 yielded the best result in terms of combined sensitivity (83%) and specificity (94%). An E/E’ >6.6 was found in 13 of 35 patients (37%); of these, 12 (92%) were in the CA group and only 2 (8%) were in the no CA group. GLS showed a good predictive value (AUC 0.89, 95% CI 0.73 to 0.97) in distinguishing between the 2 groups. A GLS cut-off point of <14.1 yielded the best result in terms of combined sensitivity (59%) and specificity (100%). A GLS <14.1 was found in 10 of 35 (29%) patients, all belonging to the CA group.
| Cardiac Amyloidosis | P | ||
|---|---|---|---|
| Yes (n = 18) | No (n = 17) | ||
| Left ventricular septal thickness (mm) | 15 ± 3.1 | 10 ± 1.408 | <0.0001 |
| Left ventricular inferolateral wall thickness (mm) | 12 ± 2.8 | 9 ± 1 | <0.0001 |
| Left ventricular end-diastolic volume (ml) | 84 ± 26.9 | 78 ± 15.8 | NS |
| Left ventricular end-systolic volume (ml) | 34 ± 15.1 | 28 ± 8 | NS |
| Left ventricular ejection fraction (%) | 59 ± 14.6 | 67 ± 6.2 | NS |
| Left ventricular mass (gr) | 285 ± 136.4 | 143 ± 39.3 | <0.0001 |
| E/E′ ratio | 11 ± 5 | 5 ± 1.4 | <0.0001 |
| Pseudonormal or restrictive Diastolic dysfunction (%) | 38 | 0 | 0.04 |
| Pericardial effusion (%) | 44 | 7 | 0.014 |
| Global longitudinal strain (-%) | 13 ± 4.3 | 20 ± 3.1 | <0.0001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
