Optimization of Coronary Sinus Pacing Site Based on Cardiac Contractility Detected by a Right Atrial Lead–Embedded Accelerometer

Abstract

A narrow QRS duration does not necessarily correlate with a high SonR value. To identify the optimal pacing site, comprehensive evaluation of all available parameters, including the use of MPP, may be necessary.

Cardiac resynchronization therapy (CRT) is an effective and well‐established treatment for patients with heart failure (HF) refractory to medical therapy, left ventricular systolic dysfunction (ejection fraction ≤ 35%), and a prolonged QRS duration. Non‐optimization of atrioventricular (AV) and ventriculoventricular (VV) timing is a major cause of no response to CRT, and the gold standard for optimization is echocardiographic guidance [1].

The SonR system, which was developed by MicroPort, includes an accelerometer embedded in the right atrial lead that detects endocardial vibrations, primarily corresponding to the first heart sound (S1), which reflects mitral and tricuspid valve closure. The peak amplitude of this signal has been shown to correlate with left ventricular contractility and stroke volume, and is used to guide optimization of atrioventricular delay (AVD) and ventriculoventricular delay (VVD) during CRT. Optimization of AVD and VVD is automatically performed once per week, and the timing parameters associated with the highest estimated cardiac output are then programmed. Optimized device settings using this algorithm reduce HF hospitalizations compared with echocardiographic optimization [2]. However, in the SonR system, the determination of the pacing site for the coronary sinus (CS) lead is left to the physician's discretion because no standardized criteria for the pacing site have been established. In practice, a pacing site associated with a narrower QRS duration on an electrocardiogram is generally preferred.

We report a case in which shortening of the AVD during multipoint pacing (MPP) resulted in a considerable increase in the SonR value despite widening of the QRS.

The patient was a 53‐year‐old man with dilated cardiomyopathy (left ventricular ejection fraction of 31%) and complete left bundle branch block (QRS duration: 169 ms), and was classified as New York Heart Association Class II. Despite receiving optimal guideline‐directed medical therapy with bisoprolol 5 mg, perindopril 2 mg, and eplerenone 100 mg, his cardiac function did not improve. Therefore, the decision was made to proceed with CRT implantation.

The selected CRT‐defibrillator device was the GALI system by MicroPort, which incorporates the SonR optimization algorithm. The atrial lead was positioned in the right atrial appendage, and the shock lead was placed at the apex of the right ventricle. An initial attempt to position the CS lead in the lateral vein was unsuccessful because of vessel tortuosity. Therefore, the lead was placed in the posterolateral vein (Figure 1).

FIGURE 1.

Lead positioning of the cardiac resynchronization therapy‐defibrillator system. The atrial lead was placed in the right atrial appendage, the shock lead was positioned at the right ventricular apex, and the CS lead was inserted into the posterolateral vein. (A) Post‐implantation chest radiograph in the anteroposterior view. (B) Intraoperative fluoroscopic image. The red arrow indicates the lateral vein, and the yellow and green arrows indicate the posterolateral vein. The yellow arrow indicates the location of the CS lead placement. CS, coronary sinus.

The optimal AVD and VVD in the SonR system is automatically evaluated once per week (on Monday night) and programmed accordingly. However, it is also possible to temporarily re‐evaluate the optimal AVD and VVD using the programmer, a process that takes approximately 5–15 min.

On the fourth postoperative day, automatic SonR setting was performed, and an AVD of 150 ms and a VVD of 0 ms were selected. On the following day, a propriety of the settings was evaluated. During threshold testing, all electrodes on the CS lead were available for pacing, allowing SonR values to be recorded for each electrode. Multipoint pacing (MPP) was considered feasible because both CS1 and CS4 electrodes were available for stimulation. As our institution limits MPP use to the 1–4 configuration based on current evidence, only MPP 1–4 was employed in this case.

SonR values were measured five times, and the mean of the measurements was used for subsequent analysis. With the initial automatic settings (CS1 pacing, AVD 150 ms, VVD 0 ms), the mean SonR value was 0.436, and the QRS duration was 150 ms. The highest mean SonR value (0.442) among individual electrodes was observed at the CS4 electrode with an AVD of 150 ms, corresponding to a QRS duration of 153 ms. Subsequent manual re‐optimization at the CS4 site resulted in a newly selected AVD of 80 ms and VVD of 0 ms, which led to a further increase in the SonR value to 0.770 and a reduction in QRS duration to 149 ms.

Additionally, MPP was applied by adding pacing from the CS1 electrode at the same site, which resulted in the QRS duration being 156 ms and the mean SonR value decreased to 0.690 (Figure 2, Table 1). Figure 3 presents enlarged electrocardiograms obtained during multipoint pacing (MPP) at AVD settings of 150, 80, and 70 ms.

FIGURE 2.

Electrocardiogram recorded during optimization of the cardiac resynchronization therapy–defibrillator. AVD, atrioventricular delay; CS, coronary sinus.

TABLE 1.

Parameters recorded during optimization of a cardiac resynchronization therapy‐defibrillator.

CS pacing	Intrinsic rhythm	1	2	3	4	4	1–4	1–4	1–4
AVD (ms)		150	150	150	150	80	150	80	70
QRS (ms)	169	150	150	156	153	149	147	156	158
Mean SonR		0.436	0.318	0.414	0.442	0.770	0.374	0.690	0.832

Note: SonR values were measured five times, and the mean of the measurements was used for subsequent analysis.

Abbreviations: AVD, atrioventricular delay; CS, coronary sinus.

FIGURE 3.

Enlarged electrocardiograms recorded during MPP at each AVD setting. AVD, atrioventricular delay; MPP, multipoint pacing.

Consequently, CRT optimization was reattempted with MPP enabled, and an AVD of 70 ms was selected. Although this approach resulted in a prolonged QRS duration of 158 ms, the mean SonR value increased to its highest level of 0.832. On the basis of this result, the final programming was set to MPP (using CS1 and CS4) with an AVD of 70 ms.

Echocardiography is commonly used to determine the optimal AV and VV timing for CRT optimization. However, it is time‐consuming, lacks standardization, and is unable to dynamically adapt to changes in heart rate associated with physical activity or medication changes.

In recent years, device‐based algorithms that allow automatic adjustment of AV and VV intervals have become available [2, 3]. Among them, the SonR algorithm, which is an approach that estimates cardiac output using an accelerometer embedded in the atrial lead, has gained attention. This algorithm has been shown to reduce the risk of HF hospitalization compared with echocardiography‐guided optimization [3]. Nevertheless, there are no established criteria for selecting the optimal pacing electrode on the CS lead when using the SonR system, leaving the decision to the physician's discretion. Furthermore, there is no standardized methodology for determining the optimal lead position.

A narrower QRS complex following pacing is generally associated with a better response to cardiac resynchronization therapy (CRT). As a result, achieving a narrow QRS is often prioritized when selecting the pacing site, and the use of multipoint pacing (MPP) is frequently considered. However, a narrow QRS complex alone does not necessarily predict a favorable CRT response [4]. While QRS narrowing is a conventional indicator of improved electrical synchrony, it does not always reflect mechanical improvement. A paradoxical dissociation was observed in which the highest SonR value—indicating optimal contractile response—occurred at a setting with a relatively wide QRS duration. This may be due to improved LV filling and reduced diastolic regurgitation from better atrioventricular timing, despite less electrical synchrony. Such discordance, although the frequency of such discrepancies is unknown, underscores the potential value of SonR‐guided programming to individualize CRT optimization beyond QRS duration alone.

In the present case, the SonR system was used to identify the optimal pacing site on the CS lead. The highest mean SonR value was recorded at the CS4 electrode. Additional pacing was applied using a second electrode because MPP was available. Although this resulted in a further reduction in the QRS duration, the SonR value paradoxically decreased. When MPP was reprogrammed with a shorter AV delay, the QRS duration widened, but the SonR value increased to its highest level, leading to the adoption of this setting.

Although a detailed investigation of this phenomenon is challenging, one possible explanation is attributable to inefficient left ventricular filling and increased diastolic regurgitation associated with a prolonged AVD. Liang et al. reported that selecting an AV delay to maximize ventricular resynchronization may result in suboptimal acute hemodynamic performance in patients on CRT [5]. They suggested that using a shorter AV delay—one that prioritizes left ventricular filling rather than minimizing QRS duration—may be preferable when programming AV delay for CRT optimization. In this case, Short AVD delay settings ensured optimal timing of atrial contraction relative to early diastolic left ventricular (LV) filling, thereby reducing diastolic mitral regurgitation and enhancing atrioventricular synchrony. This improvement led to increased LV preload and stroke volume, as evidenced by the elevated SonR signal. We suggest that the enhancement in AV synchrony outweighed the benefit of improved LV contraction efficiency associated with QRS fusion, particularly in the setting of cardiac basal pacing. The use of MPP in combination with a shortened AVD may have led to an increase in the SonR signal by optimizing LV filling and reducing diastolic regurgitation. These hemodynamic improvements may have outweighed the negative effects of suppressed intrinsic conduction and a widened QRS complex.

However, a major limitation of this case report is that mechanical dyssynchrony was not assessed using echocardiography, and cardiac output was not measured invasively during the evaluation of SonR values. Although SonR reflects cardiac contractility and has been associated with stroke volume, it does not directly measure cardiac output. In this case, SonR was used as a surrogate marker to guide AVD optimization. The lack of direct cardiac output assessment is a limitation of this report, and future studies combining SonR with invasive or non‐invasive cardiac output monitoring may provide more comprehensive insight into hemodynamic optimization.

In this case, the left ventricular lead was placed in a posterolateral branch of the coronary sinus based on anatomical accessibility and acceptable pacing thresholds. However, the posterolateral position may not necessarily represent the optimal pacing site for all patients. It is possible that alternate CS pacing locations, such as the mid‐lateral or anterolateral veins, could have yielded different mechanical responses. Therefore, real‐time functional parameters like SonR may provide additional value in evaluating the effectiveness of a given pacing site beyond anatomical criteria alone. Further clinical studies incorporating these assessments are required to validate our findings.

In conclusion, a narrow QRS duration does not necessarily correlate with a high SonR value. To identify the optimal pacing site, comprehensive evaluation of all available parameters, including the use of MPP, may be necessary.

Ethics Statement

The authors have nothing to report.

Consent

Informed consents were obtained from the patients to publish the case report.

Conflicts of Interest

The authors declare no conflicts of interest.

Muto K., Ikeda Y., Mori H., and Kato R., “Optimization of Coronary Sinus Pacing Site Based on Cardiac Contractility Detected by a Right Atrial Lead–Embedded Accelerometer,” Journal of Arrhythmia 41, no. 4 (2025): e70182, 10.1002/joa3.70182.

Data Availability Statement

The authors have nothing to report.