Characterizing cardiac contractile motion for noninvasive radioablation of ventricular tachycardia

Abstract

Background

Respiratory motion management strategies are used to minimize the effects of breathing on the precision of stereotactic ablative radiotherapy for ventricular tachycardia, but the extent of cardiac contractile motion of the human heart has not been systematically explored.

Objective

We aim to assess the magnitude of cardiac contractile motion between different directions and locations in the heart.

Methods

Patients with intracardiac leads or valves who underwent 4-dimensional cardiac computed tomography (CT) prior to a catheter ablation procedure for atrial or ventricular arrhythmias at 2 medical centers were studied retrospectively. The displacement of transvenous right atrial appendage, right ventricular (RV) implantable cardioverter-defibrillator, coronary sinus lead tips, and prosthetic cardiac devices across the cardiac cycle were measured in orthogonal 3-dimensional views on a maximal-intensity projection CT reconstruction.

Results

A total of 31 preablation cardiac 4-dimensional cardiac CT scans were analyzed. The LV lead tip had significantly greater motion compared with the RV lead in the anterior-posterior direction (6.0 ± 2.2 mm vs 3.8 ± 1.7 mm; P = .01) and superior-inferior direction (4.4 ± 2.9 mm vs 3.5 ± 2.0 mm; P = .049). The prosthetic aortic valves had the least movement of all fiducials, specifically compared with the RV lead tip in the left-right direction (3.2 ± 1.2 mm vs 6.1 ± 3.8 mm, P = .04) and the LV lead tip in the anterior-posterior direction (3.8 ± 1.7 mm vs 6.0 ± 2.2 mm, P = .03).

Conclusion

The degree of cardiac contractile motion varies significantly (1 mm to 15.2 mm) across different locations in the heart. The effect of contractile motion on the precision of radiotherapy should be assessed on a patient-specific basis.

Graphical abstract

Key Findings.

• ▪The degree of cardiac contractile motion varies across patients, ranging from 1 to 15 mm.
• ▪Cardiac contractile motion differed significantly across locations in the heart and was greatest in the right atrial appendage (vector mean 11.6 mm) and LV lead (vector mean 8.6 mm), while it was less at the aortic annulus (vector mean 6.1 mm).
• ▪No single clinical characteristic predicted the magnitude of cardiac motion.

Introduction

Stereotactic ablative radiotherapy (SAbR) is an emerging therapy for ventricular tachycardia (VT). Success rates vary in the literature, with recurrences of VT after SAbR ranging widely from 59% to 100% beyond the 6-week blanking period.1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15 Factors such as cardiac and respiratory motion may affect the precision and safety of therapy, but the magnitude of motion in different locations and directions in the heart has not been well characterized. Motion management strategies have often focused on accounting for respiratory motion through the use of respiratory gating⁶ or body immobilization equipment⁴ due to known larger magnitude of respiratory motion, but the need to account for cardiac motion has not been as well addressed.

The aim of this study was to quantify the magnitude of cardiac contractile motion at different locations of the heart and orientations in patients with cardiac arrhythmias. We hypothesized that the magnitude of cardiac contractile motion may vary between different directions and locations in the heart.

Methods

Patient population

A total of 31 consecutive patients undergoing catheter or SAbR ablation procedures for VT, premature VT, atrial fibrillation, and atrial flutter were retrospectively enrolled from 2 centers (University of California San Diego and Veterans Affairs San Diego). These patients all had preprocedural high resolution cardiac 4-dimensional computed tomography (4DCT) scans as part of a standard of care evaluation, and had radiopaque implanted cardiac devices such as transvenous pacemakers, implantable cardioverter-defibrillators (ICDs), or prosthetic valves. There were no exclusion criteria for our study. This study was performed in accordance with an Institutional Review Board–approved protocol and adhered to the Helsinki guidelines; all patients provided written informed consent.

4D cardiac CT protocol

Patients in our study underwent high-resolution, dose-modulated, retrospective cardiac-gated CT scans (Revolution, GE Healthcare, or Siemens Force) during expiratory breath hold. The cardiac CT images were obtained with extended intravenous contrast infusion and reconstructed with 0.5-mm slice thickness in different phases of the cardiac cycle (0%–95%).

Image analysis

The displacement of transvenous right atrial (RA) appendage, right ventricle (RV) ICD, coronary sinus lead tips, and prosthetic cardiac devices across the cardiac cycle were measured in orthogonal 3-dimensional (3D) views on a maximal-intensity projection CT reconstruction using imaging analysis software (Horos Project). A representative 3D reconstruction is shown in Figure 1. The lead tips were used as the fiducial for all the study patients. The inferior aspect of the bioprosthetic aortic valve were used as a fiducial when available. The cardiac motion of individual fiducials was assessed on a maximum-intensity projection CT reconstruction using imaging analysis software (Horos Project). For each fiducial, we measured the displacements in the superior-inferior, left-right (LR), and anterior-posterior (AP) directions. The average displacement was calculated as a vector mean: $\sqrt{{\left(SI\right)}^2+{\left(LR\right)}^2+{\left(AP\right)}^2}$, where SI stands for superior-inferior.

Figure 1.

A representative 3-dimensional reconstruction of a patient’s right ventricular lead, which had a displacement of 1.4cm in the left-right direction.

Statistical analysis

We used GraphPad Prism software (GraphPad Software, Version 9, San Diego, CA) to conduct our statistical analyses. For data samples with normal distributions (determined using the Kolmogorov-Smirnov normality test), paired Student’s t test was used to compare intracardiac lead motion. For data samples without normal distribution, we used the nonparametric Wilcoxon signed-rank test. Unpaired Student’s t test, chi-square test, and linear regression were used to assess for predictors of increased magnitude of cardiac motion. For the data analysis, we reported the mean ± SD and analyzed the maximum motion for each fiducial.

Results

Demographics

A total of 31 patients underwent preprocedural cardiac contrast 4DCT scans prior to arrhythmia ablation and had a permanent pacemaker, ICD, and/or a bioprosthetic valve. In this cohort, 25 patients had an RA lead, 29 had an RV lead, 11 had an LV coronary sinus lead, and 8 had a bioprosthetic aortic valve (AV). Baseline clinical characteristics (age, sex, arrhythmia profile, indication for cardiac CT, and cardiac comorbidities) are listed in Table 1.

Table 1.

Baseline patient demographics

Age, y	68 ± 12
Male	29/31 (94)
Hypertension	12/31 (39)
Ventricular tachycardia	23/31 (74)
Atrial fibrillation	16/31 (52)
Congestive heart failure	27/31 (87)
Obstructive CAD	15/31 (48)
LV ejection fraction, %	38 ± 18
Indication for cardiac CT
Ventricular arrhythmia ablation	22/31 (71)
Atrial arrhythmia ablation	5/31 (16)
Other	4/31 (13)

Values are mean ± SD or n/n (%).

CAD = coronary artery disease; CT = computed tomography; LV = left ventricular.

Differences in the magnitude of motion between RV vs LV lead tips

In patients with a biventricular device, the LV lead had a significantly greater contractile motion compared with the RV lead in the AP direction (6.0 ± 2.2 mm vs 3.8 ± 1.7 mm, P = .01) and SI direction (4.4 ± 2.9 mm vs 3.5 ± 2.0 mm, P = .049). There was a greater motion of the RV lead tip in the LR direction (6.1 ± 3.8 mm vs 3.5 ± 1.2 mm, P = .02). The maximal cardiac contractile ventricular motion of all patients was 15.2 mm, occurring in the LR direction of the RV lead tip. Figure 2 shows the mean magnitude and directions of cardiac contractile motion of the RV and LV lead tips. Table 2 shows the magnitude and directions of cardiac contractile motion of all intracardiac fiducials.

Figure 2.

Comparison of average displacements between right ventricular (RV) and left ventricular (LV) leads in different directions. A-P = anterior-posterior; L-R = left-right; S-I = superior-inferior.

Table 2.

Summary of magnitude and directions of cardiac contractile motion of all intracardiac fiducials.

	RV L-R	LV L-R	AV L-R	RA L-R	RV S-I	LV S-I	AV S-I	RA S-I	RV A-P	LV A-P	AV A-P	RA A-P
Minimum, mm	1.2	2.2	1.6	1.7	1.0	1.0	1.0	0.5	1.4	2.1	1.7	2.0
Maximum, mm	15.2	6.0	5.6	15.4	9.3	9.3	5.2	9.7	8.5	8.6	7.2	23.0
Range, mm	14.0	3.8	4.0	13.7	8.3	8.3	4.2	9.2	7.1	6.4	5.5	21.0
Mean, mm	6.1	3.5	3.2	6.3	3.5	4.4	3.2	3.3	3.8	6.0	3.8	8.7
SD, mm	3.8	1.2	1.2	3.3	2.0	2.9	1.5	2.3	1.8	2.2	1.7	5.4
SEM, mm	0.7	0.4	0.4	0.7	0.4	0.9	0.5	0.4	0.3	0.7	0.6	1.1

A-P = anterior-posterior; AV = aortic valve; L-R = left-right; LV = left ventricle; RA = right atrium; RV = right ventricle; S-I = superior-inferior.

Differences in the magnitude of motion between prosthetic AVs vs ventricular leads

The vector mean contractile motion of the prosthetic AVs (6.1 ± 2.0 mm) was less than both the LV lead tip (8.6 ± 2.6 mm, P = .03) and RV lead tip (8.6 ± 3.5 mm, P = .03).The prosthetic AVs had decreased movement than the RV lead tip in the LR direction (3.2 ± 1.2 mm vs 6.1 ± 3.8 mm, P = .02) (Figure 3) and the LV lead tip in the AP direction (3.8 ± 1.7 mm vs 6.0 ± 2.2 mm, P = .03) (Figure 3). There were no differences between the prosthetic AVs vs ventricular leads in the other orientations. The maximal cardiac contractile motion of the prosthetic AV was 7.2 mm, occurring in the AP direction. Table 2 shows the magnitude and directions of cardiac contractile motion of all intracardiac fiducials.

Figure 3.

Comparison of displacements between the aortic valve and ventricular lead tips. A-P = anterior-posterior; AV = aortic valve; L-R = left-right; LV = left ventricle; RV = right ventricle.

Greater motion of the RA appendage lead tip vs all other fiducials

The mean contractile motion of the RA lead was significantly greater than the RV lead in the AP direction (9.0 ± 5.4 mm vs 3.8 ± 1.8 mm, P < .01). The RA lead had greater motion than the LV lead in the LR direction (6.5 ± 3.4 mm vs 3.7 ± 1.3 mm, P = .01).

Predictors of increased magnitude of cardiac contractile motion

Clinical characteristics, including LV systolic ejection fraction, LV end-diastolic diameter in systole and diastole, and presence of coronary artery disease, were not found to be a predictor of magnitude of cardiac contractile motion (Table 3).

Table 3.

Predictors of increased magnitude of cardiac contractile motion

	RV mean	LV mean	AV mean	RA mean
LVEF	.96	.81	.34	.54
LVIDd	.14	.48	.61	.26
LVIDs	.55	.64	.57	.69
History of CAD	.79	.55	.08	.83

Values are P values.

AV = aortic valve; CAD = coronary artery disease; LV = left ventricle; LVEF = left ventricular ejection fraction; LVIDd = left ventricular internal dimension in diastole; LVIDs = left ventricular internal dimension in systole; RA = right atrium; RV = right ventricle.

Discussion

There were 3 major significant findings from this study: (1) cardiac contractile motion varied greatly across patients, ranging from 1 to 15 mm; (2) cardiac contractile motion differed significantly across locations in the heart and was greatest in the RA appendage (vector mean 11.6 mm) and LV lead (vector mean 8.6 mm), while it was less at the aortic annulus (vector mean 6.1 mm); and (3) no single clinical characteristic predicted the magnitude of cardiac motion. To our knowledge, this is the largest study to assess cardiac contractile motion in a relevant human population with cardiac arrhythmias undergoing cardiac ablation therapy. These findings may provide rationale to consider incorporating patient- and location-tailored cardiac contractile motion management to improve the precision of noninvasive stereotactic radioablative therapy of cardiac arrhythmias.

We noted differences in the magnitude of cardiac motion in different regions of the heart and in different directions. As expected, the RV and LV leads tended to exhibit greater motion than the AV. This can be explained by the torsional movement of cardiac contraction, in which the ventricles twist around the great vessels during systole¹⁶; this can be appreciated in representative 4D cardiac CT reconstruction videos in a patient with a normal ejection fraction (Video 1) and in a patient with severely reduced ejection fraction (Video 2).

This is the largest study to analyze the cardiac motion of intracardiac fiducials in patients with arrhythmias who already have undergone or may become candidates for stereotactic noninvasive radioablative therapy. In one recent study, Prusator and colleagues¹⁷ evaluated the VT target displacement in 11 patients. Our study included significantly more patients (31 vs 11 patients), which is powered to better capture a wider range of cardiac motion. We found a greater displacement of the RV lead tip in the LR direction compared with Prusator and colleagues (6.1 mm vs 3.9 mm), and this could be related to the greater heterogeneity of our larger patient sample. Second, our study was intentionally designed to only assess rigid radiopaque structures affixed to the endocardium to ensure that the motion of the same exact point on the heart wall is tracked through time. This was important to maximize the precision and accuracy of our measurements and minimize any error introduced by estimating the location of the VT target on the myocardial wall, which can be vague without a clear radiopaque landmark and subject to estimation error. Finally, we systematically assessed the cardiac motion of LV lead tip, RA lead tip, and prosthetic AV in addition to the ICD lead tip, which enabled sufficient power to compare the differences in motion between different locations of the heart.

While we chose to only evaluate clearly defined radiopaque landmarks, such as lead tips, to precisely estimate the motion of specific myocardial locations, there are methods to help standardize targeting cardiac segments and track the motion of each segment.¹⁸ However, the ability of this technique to track the precise motion of each segment is unknown, particularly as the ventricles move in a 3D torsion motion, which may be difficult to track using a rigid circular 17-segment American Heart Association model (arranged around a rigid circle). Without a radiopaque fiducial, it is difficult to track the same point of the myocardium through time.

There was another recent study that analyzed manually contoured segmentations of structures on the cardiac CT to characterize cardiac motion in 10 patients undergoing transcatheter AV replacement.¹⁴ In that study, the ventricles and coronary artery ostia were contoured and the centroids of the chamber contours were tracked through the cardiac phases to measure the cardiac motion. In this study, the mean displacements were found to be mostly <5 mm, with maximal displacement of 7 mm by the RV centroid in the LR direction. In comparison, the present study found slightly greater cardiac contractile motion (>5 mm) by all the lead tips and prosthetic valve fiducials, with maximal displacement up to 15 mm of the RV lead in the LR direction. Potential reasons for this difference are (1) the coronary artery ostia that were tracked in the prior study are located at the valve annuli, which would be expected to exhibit less motion compared with the ventricular apex; (2) the radiopaque lead tips imbedded in heart walls that were tracked in this study may reduce the potential error from manual contouring and chamber centroid estimation; and (3) a larger sample size of patients. Nevertheless, the mean magnitudes were still small (<1 cm) overall, and this study extends previous findings that there is a wide variation in the range of motion in different locations and orientations. Another study demonstrated a significant interobserver variability among radiation oncologists in contouring cardiac substructures (eg, valves, chambers, subsegments, coronary arteries).¹⁹ This finding further highlights the preference of utilizing ICD leads as fiducials, as it involves a much more precise target.

Most published series of SAbR protocols have not been able to account for cardiac contractile motion of the VT target directly in a patient- or location-specific manner. A few centers have used an intracardiac fiducial-based radiotherapy delivery system that adjusts the beam based on the motion of an existing or temporarily implanted pacing lead in the RV apex, but usually is not directly located at the VT target (CyberKnife; Accuray, Sunnyvale, CA). To our knowledge, fiducial tracking is currently not done for other radiation delivery methods, such as volumetric modulated arc therapy. However, as our results suggest, the motion of the VT target could potentially be different from the motion of the cardiac fiducial, depending on their location in the heart. Further studies are needed to assess whether this difference will affect the precision of therapy delivery utilizing live fiducial tracking. At centers without this capability, most have empirically employed a general margin of error known as the internal target volume, in the ballpark of a 3- to 5-mm expansion to account for both cardiac contractile motion and respiratory motion. Motion strategies such as body immobilization and respiratory gating have been employed at some centers but do not address cardiac contractile motion. Nevertheless, the extent of cardiac contractile motion has not been quantified systematically in 3 dimensions.

Furthermore, respiratory and cardiac motion may be interconnected, and respiratory motion can potentially have physiologic changes on cardiac preload conditions. However, accounting for both cardiac and respiratory motion with gating methods can certainly lead to longer treatment times and is a limitation of considering cardiac motion gating as a motion management strategy. Further studies are needed to develop other cardiac motion management strategies and also to see which patients would benefit the most from cardiac gating.

The large variation in the magnitude of motion between patients and the differences in movement between locations of the heart may provide a rationale to consider tailoring cardiac contractile motion in order to improve precision of noninvasive therapy. For example, if a target near the RV apex exhibits greater cardiac contractile motion, more aggressive cardiac motion management strategies may need to be employed. These strategies may include increasing the planning target volume margin to account for the cardiac motion, utilizing cardiac electrocardiography (ECG)-gating LINACS (TrueBeam; Varian Medical Systems, Palo Alto, CA), or using fiducial tracking systems (CyberKnife). In the majority of published series, no cardiac contractile motion management is employed. On the other hand, if a target is near the AV annulus and exhibits minimal motion, then a smaller planning target volume margin may be employed.

Our study has several limitations. First, our sample size is small but was still sufficiently powered to detect statistically significant differences in motion. Second, we utilized ICD lead tips as fiducials to track cardiac contractile motion as a surrogate of the cardiac wall. In this study, we intentionally limited our analysis to include a radiopaque marker touching the myocardium to clearly visualize motion of a specific location of the heart and limit any estimation errors. The fiducial motion may not correlate exactly with target motion, especially in circumstances in which the target has regional wall motion abnormalities that would lead to less displacement than the fiducial. In the future, sophisticated myocardial tagging imaging protocols may potentially be used for any part of the heart to better track motion of the exact VT target. Further studies are also needed to assess whether ECG gating strategies (to treat only at the same part of the cardiac cycle) can potentially limit the effect of cardiac motion).

Conclusion

Cardiac contractile motion varies significantly across different locations in the heart and different patients, and is greatest in the LV and least at the AV. Further studies are underway to develop optimal strategies to account for cardiac contractile motion, such as patient- and location-tailored planned target volume expansions and cardiac ECG-gated radiotherapy delivery.