Are classic predictors of voltage valid in cardiac amyloidosis? A contemporary analysis of electrocardiographic findings

- The prevalence of ventricular conduction delay is high (30%) in cardiac amyloidosis.
- Low voltage may be more prevalent than previously described.
- Predictors of ECG voltage in the general population are not valid in cardiac amyloidosis.
- LBBB and IVCD do not invalidate the use of ECG voltage in cardiac amyloidosis.
- ECG voltage is associated with all-cause mortality.

BackgroundLow voltage electrocardiography (ECG) coupled with increased ventricular wall thickness is the hallmark of cardiac amyloidosis. However, patient characteristics influencing voltage in the general population, including bundle branch block, have not been evaluated in amyloid heart disease.MethodsA retrospective analysis was performed of patients with newly diagnosed cardiac amyloidosis from 2002 to 2014. ECG voltage was calculated using limb (sum of QRS complex in leads I, II and III) and precordial (Sokolow: S in V1 plus R in V5-V6) criteria. The associations between voltage and clinical variables were tested using multivariable linear regression. A Cox model assessed the association of voltage with mortality.ResultsIn 389 subjects (transthyretin ATTR 186, light chain AL 203), 30% had conduction delay (QRS > 120 ms). In those with narrow QRS, 68% met low limb, 72% low Sokolow and 57% both criteria, with lower voltages found in AL vs ATTR. LV mass index as well as other typical factors that impact voltage (age, sex, race, hypertension, BSA, and smoking) in the general population were not associated with voltage in this cardiac amyloidosis cohort. Patients with LBBB and IVCD had similar voltages when compared to those with narrow QRS. Voltage was significantly associated with mortality (p < 0.001 for both criteria) after multivariable adjustment.ConclusionClassic predictors of ECG voltage in the general population are not valid in cardiac amyloidosis. In this cohort, the prevalence estimates of ventricular conduction delay and low voltage are higher than previously reported. Voltage predicts mortality after multivariable adjustment.

66