Introduction: Sudden cardiac death (SCD) can occur in patients with cardiac amyloidosis (CA). The role of implantable cardioverter-defibrillator (ICD) therapy in patients with CA, however, is not clear. The rate of appropriate ICD therapy has been reported to be between 11 and 28%, but the true impact on morbidity and mortality is still uncertain. The purpose of this study was to assess the impact of ICD therapy on morbidity and mortality among CA patients.
Methods: Patients with CA and an ICD who were evaluated at Northwestern Memorial Hospital between 2000 and 2016 were identified. Electronic medical records were reviewed for clinical and ICD data.
Results: 31 patients (24 males, age 69 +/- 10 years) with CA and ICD were included. 25 patients received an ICD for primary prevention and 6 received an ICD for secondary prevention. CA type was AL in 14 patients, TTR in 16, and unknown in 1. Mean ejection fraction was 43 +/- 14%. Five patients (16%) experienced implant-related complications (4 hematomas and/or wound dehiscence requiring evacuation or extraction and one acute subclavian vein thrombosis). Of the four patients who experienced a hematoma and/or wound dehiscence, two were receiving anticoagulation and one was thrombocytopenic. Two patients had high defibrillation thresholds that persisted despite addition of a subcutaneous array. After 15 +/- 11 months of follow-up, 2 patients (6.5%) received an appropriate shock. One of these patients received an ICD for primary prevention and one received an ICD for secondary prevention. 2 patients (6.5%) experienced inappropriate shocks for atrial fibrillation. A total of 12 patients (39%) died (6 with TTR, 5 with AL, 1 with unknown type), all of non-arrhythmic causes.
Conclusions: The rate of appropriate ICD therapy in patients with CA during an average follow-up of 15 months after ICD implantation was low at 6.5% compared to a high nonarrhythmic mortality rate of 39% and a high implant-related complication rate of 16%. These findings raise further questions about the risk-benefit equation for ICDs in patients with CA, and highlight the need for prospective randomized data.