Electromagnetic interference between continuous-flow left ventricular assist device and cardiac resynchronization therapy defibrillator due to an unrecognized mechanism

Overall efficacy and safety of combined continuous-flow LVAD and ICD were recently reported [ ]. However, several cases of electromagnetic interference between ICD and LVAD have been reported, such as telemetry instability between ICD and its programmer or inappropriate ICD therapies caused by noise arising from LVAD [ , , , , ]. However, these reports were cases which used HeartMate® II (Thoratec Corporation, Pleasanton, CA, USA) or HeartWare® (HeartWare International, Inc., Framingham, MA, USA) with ICD (St. Jude Medical Corporation, Sunnyvale, CA, USA or Sorin Group, Milan, MI, Italy). Here we report novel electromagnetic interference between CRT-D (COGNIS®, Boston Scientific, St. Paul, MN, USA) and LVAD (DuraHeart®, Terumo Corporation, Shibuya, Tokyo, JAPAN). A 35-year-old man who had been diagnosed in a health examination in 2001 with premature ventricular contraction, was then diagnosed with dilated cardiomyopathy, by endomyocardial biopsy, in 2003. Whereas the best medical treatments for severe heart failure had been done, he was repeatedly hospitalized. Left ventricular ejection fraction was reduced to 20% with wide QRS complex, and non-sustained ventricular tachycardia was also observed. CRT-D: Contak RENEWALl® (Boston Scientific, St. Paul, MN, USA) was implanted in 2003. CRT-D generator was replaced by COGNIS 100-D® (Boston Scientific, St. Paul, MN, USA) due to low battery in 2010. However, the frequency of hospitalization was gradually increased and a LVAD (DuraHeart, Terumo Corporation, Shibuya, Tokyo, Japan) was implanted as a bridge to transplantation (Fig. 1A ). During LVAD implantation surgery, ventricular arrhythmia therapeutic capability was turned off to avoid inappropriate therapy caused by electromagnetic interference of the electrocautery scalpel. Just after the LVAD was inserted and started, the programming head of CRT-D programmer (Boston Scientific, St. Paul, MN, USA) was placed on the skin above the CRT-D generator to communicate with the CRT-D, and incessant intermittent telemetry failure was observed (Fig. 1B). In the perioperative period, power delivery was supplied by hospital console, which is the attached power supply and modulator of rotation setting of the LVAD. During telemetry failure, no information was obtained by the programmer and any manipulation from the CRT-D programmer could not be accepted. Abnormal pacing or sensing manner was not suspected by other biomonitors. Even after moving to the cardiac care unit, incessant intermittent telemetry failures were also observed in the same action. We tried to move the cable of the programming head and programmer itself from VAD insertion site to avoid this telemetry failure. However, no position could resolve this intermittent telemetry failure. We tested shielding of these devices by steel shielding [ ] (Fig. 1C). When we covered the skin above the CRT-D generator and that above the LVAD pump, this telemetry failure disappeared. In addition, when we switched the origin of power supply to battery, this telemetry loss also completely disappeared. After discharge, this patient had no problems with CRT-D and LVAD function including wand-based telemetry, because the patient always used a battery-derived power supply. We confirmed this interference in vitro using the same models of LVAD, CRT-D, and programmer (Fig. 2A ). In this case, distance between the CRT-D generator and the LVAD pump was 13 cm (Fig. 1A). We tested whether the CRT-D generator located at 13 cm away from the LVAD pump was affected (Fig. 2B). As expected, the hospital console-powered LVAD repeatedly interfered with the communication of the CRT-D and its programmer (Fig. 2C). This phenomenon is only observed when the CRT-D generator is located in a horizontal direction to the LVAD, but not in a vertical direction (Fig. 2C). Battery-powered LVAD did not provoke this interference as clinically observed (Fig. 2C). In addition, we measured the power of electromagnetic waves generated from currently used LVASs including DuraHeart using originally-developed equipment. Dominant frequency from DuraHeart was 100 kHz and those of HeartMateII and EvaHeart are 7.2 KHz and 30 Hz as manufacture's report, respectively. Whereas electromagnetic wave from HeartMate II was detected to 24 cm from LVAS pump, those of DuraHeart and EvaHeart were not detected over 8 cm (Fig. 2D). These results suggested that electromagnetic wave from DuraHeart could not reach the location of CRT-D generator.