By Joshua Moss, MD

Associate Professor of Clinical Medicine, Cardiac Electrophysiology, Division of Cardiology, University of California, San Francisco

Dr. Moss reports he is a consultant for Abbott, Biosense Webster, and Boston Scientific.

SYNOPSIS: In the MARVEL 2 prospective study of patients with atrioventricular block treated with a leadless ventricular pacemaker, atrial sensing via an accelerometer-based algorithm was largely successful in establishing atrioventricular synchronous pacing.

SOURCE: Steinwender C, Khelae SK, Garweg C, et al. Atrioventricular synchronous pacing using a leadless ventricular pacemaker: Results from the MARVEL 2 study. JACC Clin Electrophysiol 2020;6:94-106.

Leadless pacemakers, implanted in the right ventricle via femoral venous access, and integrating both pacing circuitry and pacing electrodes in a single small capsule, have proven effective in several clinical scenarios. The initial iteration of these devices, which the Food and Drug Administration (FDA) approved in 2016, provide asynchronous (VOO mode) or demand-based ventricular pacing based on ventricular sensing only (VVI mode), without the benefit of atrioventricular (AV) synchrony.

Steinwender et al assessed an enhanced pacing algorithm for providing AV synchronous pacing with a standard Medtronic Micra leadless pacemaker. The software is designed to detect atrial contraction, using the built-in accelerometer to assess movement of the device within the right ventricle during the cardiac cycle. The authors enrolled 77 patients with leadless ventricular pacemakers from 12 centers, including 40 patients with a predominant rhythm of underlying sinus with complete AV block.

The novel algorithm was downloaded onto the implanted device of 75 patients and optimized individually for each. Pacing was tested in a VVI mode (ventricular pacing without atrial sensing), at a lower rate of 50 beats per minute (bpm), then in a VDD mode (ventricular pacing synchronized to atrial contraction based on the accelerometer data). Holter ECG monitoring data were collected in a series of postures at rest, as well as during walking at two different paces.

The primary safety endpoint was based on freedom from atrial undersensing (leading to pauses lasting more than two cycles of the programmed lower rate) and atrial oversensing (leading to tachycardia faster than 100 bpm at rest for more than three minutes). In the 75 patients with the algorithm downloaded, there were no instances of ventricular pauses or oversensing-induced tachycardia in more than 600,000 cardiac cycles analyzed. There were no adverse events believed to be related to the algorithm.

The primary efficacy endpoint was defined as a paced or sensed ventricular beat within 300 msec of at least 70% of ECG-confirmed P-waves, as assessed in the 40 patients with complete AV block. No patients experienced > 70% AV synchrony during VVI pacing at 50 bpm (median, 26.9%). In contrast, 38 of 40 recorded > 70% AV synchrony during VDD pacing using the accelerometer-based algorithm (median, 94.3% AV synchrony). One patient with only 69% AV synchrony as a result of atypical accelerometer signals presented with a history of repaired tetralogy of Fallot. Results were better when resting in sitting or lying positions vs. standing.

Additionally, echocardiograms were collected during both VVI and VDD pacing, with blinded measurement of left ventricular outflow velocity-time integral (VTI) as a secondary endpoint. The VTI, a proxy for stroke volume, increased by an average of 9% during VDD pacing vs. baseline VVI pacing. Sinus rate slowed from an average of 73 bpm during VVI pacing to 66 bpm during VDD pacing. The device also successfully mode switched from VVI to VDD pacing in patients with intermittent AV block, facilitating minimization of ventricular pacing. The authors concluded atrial sensing via an accelerometer-based algorithm was largely successful in establishing AV synchronous pacing with the available leadless right ventricular apical pacemaker.


Leadless pacemakers, currently limited clinically to the Medtronic Micra device, have become an important tool in our pacing armamentarium. A principal disadvantage of these single-chamber devices has been the lack of ability to provide AV synchronous pacing. Without electrodes in the atrium to sense atrial depolarization, ventricular pacing can be delivered based only on timing from a previous sensed or paced ventricular depolarization. In patients with prior device infections, difficult subclavian venous access, limited functional status, or permanent atrial fibrillation, ventricular-only pacing can be adequate or even advantageous. AV pacing has not been shown to improve survival or clearly reduce the risk of heart failure.

However, it is well established that AV pacing provides superior cardiac output, reduces the risk of atrial fibrillation, and minimizes the incidence of pacemaker syndrome. Additionally, for patients with intact sinus node function, faster ventricular pacing rates in response to native chronotropic response may be superior to that afforded by hardware-based activity sensors.

The MARVEL 2 study showed that despite the lack of implanted electrodes in the atrium, atrial activity can be reliably identified via analysis of the device accelerometer data (normally used to drive activity-based pacing rate changes). The hemodynamic impact of atrial contraction creates a signature movement of the device within the right ventricle that can be leveraged to enable ventricular pacing with AV synchrony. Implementation of such an algorithm represents a significant step toward substantially expanding the patient population in which a leadless device may be practical or even preferred.

The principal limitation of the study is the duration of the observational period — only about 30 minutes in various positions at rest and four minutes while walking for patients without new devices. Additionally, the sample size was small, including only 40 patients with sinus rhythm and complete AV block. Additional longer-term data in a larger and more diverse population of patients will be necessary to fully assess the effect of the VDD pacing algorithm. However, the safety data certainly suggest a favorable risk-benefit ratio, especially considering the ability to use an established physical device design and implant procedure.

Some may consider > 70% AV synchrony to be a relatively low bar for success, considering a rate closer to 100% would be expected for a standard dual-chamber pacemaker with atrial and ventricular leads. However, when combined with the algorithm for minimizing ventricular pacing in patients with intermittently intact AV conduction, that degree of AV synchrony may be more than adequate to achieve the benefits associated with dual-chamber devices. Longer-term studies are needed to determine the effect of upgraded software. Based in part on results of the MARVEL 2 study, the FDA approved the Medtronic “Micra AV” device with the necessary algorithms for clinical use in January.