Abstract
Conduction disturbances are among the most frequent complications of transcatheter aortic valve replacement (TAVR), typically occurring within the first 72 h after implantation. However, delayed complete atrioventricular (AV) block is rare and may present late with serious clinical consequences. We describe the case of a 72-year-old man with hypertension who developed complete AV block nearly 1 year after TAVR, underscoring the importance of long-term surveillance in patients with high-risk anatomy. The patient initially presented with progressive exertional dyspnea and was diagnosed with severe low-flow, low-gradient aortic stenosis, with a left ventricular ejection fraction of 33%. Preprocedural computed tomography demonstrated a short membranous septum (4.0 mm) and an annulus-membranous septum distance of 1.2 mm, both of which are recognized predictors of conduction disturbances. He underwent transfemoral TAVR with a self-expanding Qiming L26 valve implanted at a depth of approximately 5 mm. Early recovery was uneventful, aside from a small paravalvular leak, which resolved by the 8-month follow-up, at which time echocardiography showed recovery of systolic function to 60%, and sinus rhythm was observed on electrocardiography. At 11 months following TAVR, a community screening ECG revealed 2:1 AV block, and 3 weeks later, he presented with symptomatic complete AV block and a ventricular escape rhythm at 30 bpm. Repeat CT showed increased frame depth (7.6 and 9.8 mm), suggesting possible valve-septum interaction; however, causal attribution is limited by the absence of immediate/serial CT and electrophysiological mapping, and age-related conduction system disease may have contributed to this finding. A dual-chamber permanent pacemaker was implanted, resulting in complete symptomatic recovery and a stable prosthetic valve function. This case report highlights a rare but clinically important phenomenon of very late conduction block after TAVR and supports a risk-stratified approach to anatomical assessment and long-term rhythm monitoring. However, this inference remains hypothesis-generating, given its single-patient nature.