Hemodynamic consequences, capture mode transitions, exercise implications, and echo-guided optimization strategies for AV interval programming in conduction system pacing
Atrioventricular (AV) delay is among the most consequential programmable parameters in a dual-chamber pacemaker. In conventional right ventricular (RV) pacing, the clinical effects of AV delay misprogramming are well-characterized — pacemaker syndrome from loss of AV synchrony, reduced cardiac output, and symptomatic hemodynamic compromise. In left bundle branch area pacing (LBBAP), however, AV delay adjustments carry additional layers of complexity that are not present in standard DDD systems.
LBBAP achieves physiologic ventricular activation by directly engaging the left bundle branch or its fascicles from a deep septal lead position. The paced wavefront propagates through the native His-Purkinje network, producing narrow QRS complexes and synchronous biventricular contraction. But this mechanism introduces a unique dependency: the AV delay must be programmed not only for optimal hemodynamics but also to preserve LBB capture — a parameter-dependent state that can shift between selective capture, nonselective capture, fusion, and loss of capture depending on timing.
This review examines the clinical side effects of AV delay programming in dual-chamber LBBAP systems, covering hemodynamic consequences at both extremes, LBBAP-specific capture considerations, exercise performance implications, and evidence-based optimization approaches.
When the programmed AV delay is too short, the ventricular pacing stimulus is delivered before atrial contraction has completed its contribution to ventricular filling. The mitral valve begins closing — or has already closed — while the atrium is still contracting, producing a cascade of hemodynamic and symptomatic consequences.
The degree of hemodynamic impairment from a short AV delay is patient-dependent. Patients with stiff, hypertrophied ventricles (e.g., hypertensive heart disease, hypertrophic cardiomyopathy, aortic stenosis) are disproportionately affected because they are more reliant on atrial contraction for diastolic filling. Conversely, young patients with highly compliant ventricles may tolerate moderately short AV delays with minimal symptoms.
At the other extreme, an excessively long AV delay allows the native AV node to conduct the atrial impulse to the ventricles before the paced stimulus fires. In a standard RV pacing system, this simply means less ventricular pacing — which may actually be desirable (as in managed ventricular pacing algorithms). In LBBAP, however, this creates a fundamentally different problem.
Beyond the standard hemodynamic trade-offs shared with all dual-chamber pacemakers, LBBAP introduces capture-dependent phenomena that make AV delay programming uniquely consequential.
LBBAP capture exists on a spectrum. At very short AV delays (well ahead of native conduction), the pacing stimulus activates the left bundle branch directly with a surrounding local myocardial component — producing nonselective LBBAP (ns-LBBAP), characterized by a small initial slur followed by rapid Purkinje-mediated conduction. At slightly shorter coupling intervals or higher outputs, pure selective LBB capture (s-LBBAP) may occur, with a characteristic isoelectric stimulus-to-QRS interval followed by narrow activation.
As AV delay lengthens, native AV conduction begins to contribute, creating fusion — the QRS becomes a hybrid of the paced and native wavefronts. At sufficiently long AV delays, native conduction dominates and LBB capture is functionally lost. The clinical implication is that small changes in AV delay can shift the patient between these capture states, each with different ventricular activation patterns and hemodynamic profiles.
Most modern DDD pacemakers offer rate-adaptive AV delay (RAAVD), which automatically shortens the AV interval as the atrial sensed rate increases. The physiologic rationale is sound: at higher heart rates, the diastolic filling period shortens, and a proportionally shorter AV delay maintains optimal E/A timing.
In LBBAP, however, RAAVD introduces a potential hazard. If the AV delay shortens excessively at exercise heart rates, the ventricular pacing stimulus may be delivered so early that diastolic filling is incomplete — reducing preload and stroke volume at the exact moment when cardiac output demand is highest. This is particularly relevant in athletic patients (e.g., competitive rowers, cyclists) who achieve high sustained heart rates and depend on efficient diastolic filling for exercise performance.
A long programmed AV delay consumes a larger proportion of the total cardiac cycle at any given heart rate. This means the device reaches its programmed upper tracking rate (UTR) at a lower atrial rate than it would with a shorter AV delay. When the UTR is reached, the device initiates Wenckebach behavior (progressively lengthening the AV delay until a ventricular beat is dropped) or 2:1 block (every other atrial beat is tracked).
Both responses impose an abrupt ceiling on the paced ventricular rate, which the patient experiences as sudden exercise limitation — a common and frustrating complaint in active individuals with dual-chamber pacemakers. Optimizing AV delay to the shortest hemodynamically tolerable value maximizes the available tracking range before UTR is reached.
of stroke volume (up to 40% in stiff ventricles)
confirms LBB capture via Purkinje conduction
requires echo-guided individualization
shorter AV delay → higher V-pace % → more drain
Suboptimal AV delay in a dual-chamber LBBAP system may present with symptoms that overlap significantly with other cardiac conditions, making clinical correlation essential. Key symptoms that should prompt AV delay evaluation include:
Empiric AV delay programming (e.g., 150 ms for sensed events, 180 ms for paced events) is a reasonable starting point but leaves hemodynamic improvement on the table. Echo-guided optimization is the gold standard and is especially important in LBBAP, where the optimization must satisfy two simultaneous goals: maximizing hemodynamic output and maintaining conduction system capture.
The Ritter method uses transmitral Doppler to identify the AV delay at which the A wave terminates exactly at mitral valve closure. Two measurements are taken — one at a very short AV delay (truncated A wave) and one at a very long AV delay (A wave well before closure) — and a formula derives the optimal interval. This method is straightforward and widely used but does not directly account for stroke volume; it optimizes diastolic filling without confirming systolic output improvement.
The iterative VTI method directly measures stroke volume (via aortic outflow tract VTI) at multiple AV delays, stepping in 20 ms increments across the hemodynamically plausible range (typically 80–250 ms). The AV delay that produces the highest VTI is selected. This is more time-consuming but directly optimizes the parameter that matters: forward stroke volume.
After identifying the hemodynamically optimal AV delay by either method, the LBBAP-specific verification step is critical: confirm that LBB capture is maintained at the chosen interval. This requires a 12-lead ECG (or at minimum leads V1 and V5/V6) at the optimized AV delay to verify that stim-to-LVAT remains ≤75 ms and V1 morphology retains its terminal R wave. If the optimal hemodynamic AV delay falls in a range where capture is lost or variable, the AV delay should be shortened to the nearest interval that preserves LBB capture — accepting a small hemodynamic trade-off to maintain the physiologic activation advantage of conduction system pacing.
| Parameter | Too Short | Optimal | Too Long |
|---|---|---|---|
| Atrial kick | Truncated/lost | Fully utilized | Completed but temporally dissociated |
| LBB capture (LBBAP) | Maintained (ns-LBBAP or s-LBBAP) | Maintained with hemodynamic benefit | Lost — native conduction dominates |
| QRS morphology | Consistent paced morphology | Consistent, narrow | Variable fusion or native QRS |
| Diastolic MR | Absent | Absent | May occur (late diastolic regurgitation) |
| Cardiac output | Reduced (lost atrial contribution) | Maximized | Reduced (loss of CSP benefit) |
| Exercise capacity | Limited if RAAVD overshoots | Optimized | Limited by early UTR/Wenckebach |
| Pacemaker syndrome | Possible (pseudo-syndrome) | Absent | Possible (loss of synchrony) |
| Battery impact | Higher drain (↑ V-pace %) | Balanced | Lower drain if pacing % drops |