Feasibility of intraprocedural integration of cardiac CT to guide left ventricular lead implantation for CRT upgrades

Abstract Background Optimal positioning of the left ventricular (LV) lead is an important determinant of cardiac resynchronization therapy (CRT) response. Objective Evaluate the feasibility of intraprocedural integration of cardiac computed tomography (CT) to guide LV lead implantation for CRT upgrades. Methods Patients undergoing LV lead upgrade underwent ECG‐gated cardiac CT dyssynchrony and LV scar assessment. Target American Heart Association segment selection was determined using latest non‐scarred mechanically activating segments overlaid onto real‐time fluoroscopy with image co‐registration to guide optimal LV lead implantation. Hemodynamic validation was performed using a pressure wire in the LV cavity (dP/dtmax)). Results 18 patients (male 94%, 55.6% ischemic cardiomyopathy) with RV pacing burden 60.0 ± 43.7% and mean QRS duration 154 ± 30 ms underwent cardiac CT. 10/10 ischemic patients had CT evidence of scar and these segments were excluded as targets. Seventeen out of 18 (94%) patients underwent successful LV lead implantation with delivery to the CT target segment in 15 out of 18 (83%) of patients. Acute hemodynamic response (dP/dtmax ≥ 10%) was superior with LV stimulation in CT target versus nontarget segments (83.3% vs. 25.0%; p = .012). Reverse remodeling at 6 months (LV end‐systolic volume improvement ≥15%) occurred in 60% of subjects (4/8 [50.0%] ischemic cardiomyopathy vs. 5/7 [71.4%] nonischemic cardiomyopathy, p = .608). Conclusion Intraprocedural integration of cardiac CT to guide optimal LV lead placement is feasible with superior hemodynamics when pacing in CT target segments and favorable volumetric response rates, despite a high proportion of patients with ischemic cardiomyopathy. Multicentre, randomized controlled studies are needed to evaluate whether intraprocedural integration of cardiac CT is superior to standard care.

cardiomyopathy. Multicentre, randomized controlled studies are needed to evaluate whether intraprocedural integration of cardiac CT is superior to standard care.

K E Y W O R D S
cardiac CT, CRT, image guidance, improving CRT response

| INTRODUCTION
Patients with heart failure and pre-existing pacing or implantable cardioverter-defibrillator (ICD) systems may benefit from cardiac resynchronization therapy (CRT) upgrade, 1 however suboptimal left ventricular (LV) lead placement outside of late activating regions and in myocardial scar may result in suboptimal response. 2 Cardiac magnetic resonance (CMR) can guide LV lead placement targeting late mechanical activation (LMA) and avoiding LV scar. 3 However, 28% of CRT candidates have pre-existing pacing or ICD systems and may be unsuitable for CMR, 4 which is not without risk if patients are pacing dependent. Furthermore, patients with heart failure often find CMR challenging due to long breath holds or prolonged supine periods (30-45 min) and image degradation from lead artifact impedes its utility.
Cardiac computed tomography (CT) has the potential to guide LV lead placement. 5,6 Rapid acquisition of isotropic 3-dimensional (3D) whole heart data sets with submillimetre spatial resolution allows accurate assessment of coronary venous anatomy, 5 regional or global LV systolic function assessment 5,7 and evaluation of LV dyssynchrony or LMA. 5,8 Additionally, CT may detect regional hypoperfusion and myocardial scar 9 albeit with varying results with no current standardized imaging protocols to reliably identify late iodine enhancement. 10 We have previously shown that "offline" preprocedural cardiac CT dyssynchrony analysis produces functional data sets with sufficient temporal resolution to differentiate LMA segments in a separate cohort of 18 patients and that CT target selection retrospectively correlated well with acute hemodynamic response (AHR). 5 We previously showed the utility of "real-time" X-magneic resonance imaging (MRI) guidance for CRT, 3 however, to date standalone Preprocedural cardiac CT with intraprocedural image integration to guide LV lead placement has not been demonstrated.
We set out to test the feasibility of a purpose-built, integrated software platform to process, analyze and overlay CT data within a cardiac catheter laboratory to prospectively guide LV lead implantation. To achieve this, we performed preprocedural cardiac CT with intraprocedural image integration to target LV lead placement with acute hemodynamic validation.

| Recruitment and follow-up
Consecutive patients ≥18 years of age with heart failure and LV ejection fraction (LVEF) less than 40% undergoing CRT upgrade were eligible if they met all study requirements, were on optimal heart failure pharmacotherapy for ≥3 months Before enrollment and could provide informed consent. Patients were ineligible if eGFR < 30 ml/ min/1.73 m 2 , previous iodine contrast allergy or any contraindication to CRT/transvenous LV lead implantation via the coronary sinus (CS).
All patients underwent the following tests at baseline and 6-month follow-up visits post CRT upgrade: New York Heart Association (NYHA) functional class assessment; physical examination; 12-lead resting ECG; two-dimensional transthoracic echocardiogram including Simpson's biplane LV end-diastolic volume (LVEDV), LV endsystolic volume (LVESV) and LVEF; Minnesota living with heart failure questionnaire score (MLWHFQ); 6-min walk test (6MWT).

| Preprocedural cardiac computed tomography dyssynchrony imaging and analysis
All patients underwent dedicated cardiac CT during RV pacing Before CRT upgrade using a 3rd generation dual source scanner (SOMATOM Force; Siemens Healthineers) with temporal resolution up to 66ms. Intravenous metoprolol helped achieve a heart rate less than 65 beats/min in sinus rhythm and less than 100 beats/min in atrial fibrillation. A topogram was used for localization and automatic exposure control. Following injection of 120 ml iodinated contrast material (Omnipaque 350 mg/ml iodine; GE Healthcare) at 5 ml/s via the antecubital vein, a retrospective ECG-gated cardiac CT angiography (CTA) was performed with reference dose settings 100 kV and 288 mAs/rot. Contrast monitoring triggered the scan with 14 s delay after reaching 100HU (at 100 kV) in the descending aorta. Full cardiac function was acquired for motion analysis throughout the cardiac cycle (0%-100%, 5% increments) and coronary venous anatomy for identifying target veins subtending target segments. Comprehensive cardiac CT dyssynchrony assessment was adapted from Behar et al. 5 and calculated using opensource software ( Figure 1). Image registration warping field was applied to a triangulated LV endocardial mesh from ECG-gated cardiac CT angiograms.
Endocardial wall motion was tracked using single semi-automated LV cavity segmentation at end-diastole with motion characterized by GOULD ET AL.

| Pre-implant cardiac CT scar imaging
The initial eight patients underwent end-systolic prospectively ECGtriggered late enhancement scanning with a dual energy scan 7 min after contrast injection. For patients in sinus rhythm with hazards ratio (HR) less than 65 beats/min, scanning was performed at 90/Sn150 kV with 165 and 127 mAs reference, respectively with a full 250 ms reconstruction. A single energy shuttle mode dynamic scan acquiring for 15 s (4-5 cycles) at 80 kV/300 mAs (reference dose settings) was used for patients with atrial fibrillation and/or HR > 65 beats/min. The last 10 patients at 12.5 min following contrast injection underwent single energy shuttle mode dynamic scanning regardless of heart rate or rhythm with reference doses increased by 30% to reference kV 80 and reference mAs 390. Late iodine enhancement images were reconstructed using 2 mm slice thickness and 1 mm increments with medium smooth kernel (Qr36) and iterative correction of iodine beam hardening. The dynamic scan time points were averaged after nonrigid registration. The resulting average volume was loaded into a DICOM viewer and qualitatively evaluated in apical, mid, and basal short-axis slices in a narrow window. For comparison, systolic phases were loaded in synchronized orientation.
2.4 | Intraprocedural image overlay and LV lead guidance using "guide CRT" platform Image overlay using Guide CRT (Siemens Healthineers) and CMRderived target segments has previously been described. 3

,12 Guide
CRT is a custom-built software prototype on a dedicated workstation, integrated into an Artis-Q biplane fluoroscopic angiography system (Siemens Magnetom Artis Combi Suite; Siemens Healthcare GmbH). Rapid, automatic data processing allows information to be extracted from cardiac CT images with an automated protocol for slice registration and LV segmentation. Cardiac CT acquisition, CT-derived AHA target segments were co-registered and overlaid onto fluoroscopic images to guide optimal LV lead delivery ( Figure 3B).

| LV lead implant and hemodynamic assessment
Following CS cannulation, occlusive balloon venography was performed in three fluoroscopic projections (right-anterior-oblique 30°; postero-anterior; left-anterior-oblique 30°). Fluoroscopic images were fused with the CT-derived 3D mesh and target segments subtended by target coronary veins ( Figure 3B). Bony landmarks, existing pacing wires ± sternotomy wires facilitated image fusion and motion compensation. LV lead placement to multiple veins (target and nontarget segments) was performed and hemodynamic assessment undertaken with invasive dP/dt max measurements using a 0.014-inch high-fidelity "wireless" Pressure Wire X (St. Jude Medical) in the LV cavity via a retrograde arterial approach as previously described. 13 LV-dP/dt max measurements were recorded at each pacing site as previously described, 13 using CoroFlow (Coroventis).

| RESULTS
Eighteen patients underwent CT dyssynchrony and scar assessment.
Baseline characteristics are summarized in Table 1 Table SA1). These correlated with regional late gadolinium enhancement (LGE) from historical CMR imaging. A further six patients had LV scar inferred by myocardial wall thinning and regional hypokinesis/ akinesis on CT without identifiable late iodine enhancement or hypoperfusion. Post-processing CT analysis identified proposed target AHA segments outside of scar subtended by target coronary veins in all 18 patients.   Figure SA2).

| Follow-up
Echocardiographic and clinical measures at baseline and 6-month followup are shown in Table 3. LVESV was improved at 6 months compared to baseline (133.8 ± 67.7 vs. 103.5 ± 53.9 ml; p = .003). Furthermore, LVEDV, NYHA functional class, and paced QRS duration were significantly lower at 6 months compared to baseline (Table 3). MLWHFQ scores, 6MWT distance, and NT-proBNP were similar at 6 months follow-up compared to baseline ( Note: Values are presented as mean ± SD, median (IQR) or as n (%).
Abbreviations: 6MWT, 6-min walk test; ACE, angiotensin converting enzyme; ARB, angiotensin receptor blocker; CT, computed tomography; eGFR, estimated glomerular filtration rate; IQR, interquartile range; MLWHF, Minnesota living with heart failure; NT-proBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association.   could not be avoided and in two patients cannulation of the target vein was not possible due to limited coronary venous anatomy. 16 These results suggest promise in targeting optimal LV lead placement, however, this is potentially a time and resource heavy preprocedural planning exercise requiring two cross-sectional imaging modalities with ECGI integration. Nguyên et al. 16 did not report preprocedural imaging and data processing or planning time which is likely to be reasonably long and may limit its clinical utility in real-world clinical practice. Furthermore, validation of the optimal pacing site was not performed using hemodynamic assessment. 16

| Study limitations
This is a small proof of principle study and larger multicentre, randomized controlled studies would be necessary to evaluate whether in- and therefore cardiac CT may be less sensitive to subtle regional motion changes. Additionally, the predicted target vein derived from cardiac CT was either a lateral or posterolateral vein and a larger study would be required to assess the frequency of predicted alternative CS branches and evaluate the effect on clinical and echocardiographic outcomes.
Furthermore, successful LV lead delivery was achieved to the CT-derived target vein (89%) and target segment (83%). Whilst we believe these numbers are good for an image guidance approach, this remains a limitation of using a transvenous approach via the CS for LV lead delivery.
This may also be seen as an advantage in preprocedural planning; if we are able to reliably predetermine whether there is absence of a suitable caliber vein subtending the target LV segment, then it could also be used to identify patients more suited to first-line endocardial LV lead implantation. LV endocardial pacing may be useful in non-responders to conventional CRT, however, the optimal site of stimulation varies greatly between patients. 18 A CT guidance system may therefore help identify which patients are more suited to CT-guided endocardial LV lead stimulation. 19 CT-guided CRT is resource and time intensive and undoubtedly more costly than using intraprocedural late electrical activation, 20,21 however the additional preprocedural anatomical planning and mechanical dyssynchrony data available with a CT-guided approach offers the potential to avoid late electrically activating segments within scarred myocardium and target the latest mechanically and/or electrically activating segments outside of scar. Integration of these techniques may prove advantageous and further studies are required. In the present study, cardiac CT did not reliably identify late iodine enhancement irrespective of the imaging protocol used, however, regional hypoperfusion (hypoattenuating areas) was observed in regions of known LV scar from historical MRI data of the same patients. This raises the question whether all scar actually hyper-enhances with iodinated contrast which may potentially also be seen as hypoattenuating areas in advanced scar formation. Dual energy scanning was used initially aiming for superior late iodine enhancement detection, however, scar identification proved unreliable. The protocol was modified to a single energy shuttle mode dynamic scan at 12.5 min balancing low image noise and low kVp required for late iodine detection with slight improvement. More research into accurate LV scar detection using cardiac CT is required with standardization of CT scar imaging protocols. Additionally, use of cardiac CT may not be feasible in all patients due to significant renal impairment or contrast allergy and given the small risks of ionizing radiation, this may not be an appropriate imaging modality for all patients.

| CONCLUSION
Intraprocedural cardiac CT image overlay guidance for optimal LV lead placement in CRT upgrades is feasible with superior acute hemodynamics when pacing in CT target segments and favorable volumetric response rates, despite a high proportion of patients with ischemic cardiomyopathy. Multicentre, randomized controlled studies are needed to evaluate whether intraprocedural CT image overlay guidance to avoid scar and target late activating regions is superior to standard care.