第一部分 左束支区域起搏的QRS形态与心力衰竭患者心电特征和超声反应性的相关性

研究背景：左束支区域起搏（Left bundle branch area pacing，LBBAP）是一种新兴的生理性起搏术式，不仅可用于缓慢性心律失常患者，也可作为心力衰竭（Heart failure，HF）患者心脏再同步治疗（Cardiac resynchronization therapy，CRT）的新选择。最初，LBBAP强调左束支主干起搏（Left bundle trunk pacing，LBTP），但刻意追求主干夺获导致手术成功率相对较低且反复尝试也会加重间隔损伤、延长手术及放射时间。近来LBBAP的方法学不断发展，出现了左前分支起搏（Left anterior fascicle pacing，LAFP）及左后分支起搏（Left posterior fascicle pacing，LPFP）等分支起搏模式，并各自表现出不同的QRS形态。本研究拟评估左束支主干/分支起搏的心电参数以及超声反应性的可能差异，以探讨HF患者的最佳LBBAP起搏模式。

研究方法：回顾性纳入本中心成功行LBBAP的HF患者，按照起搏QRS形态判定夺获的左束支成分并分组。收集术中影像比较各组间导线头端位置的差异。收集患者自身心律及LBBAP后的心电图测量心室除极参数QRS时限（QRS duration，QRSd）、心室间激动延迟时间（Interventricular delay，IVD）及复极参数QTc、T波峰末间期（TpeakTend，TpTe）和TpTe/QTc并进行组间比较。收集患者基线和随访左室射血分数（Left ventricular ejection fraction，LVEF）和左室舒张末期内径（Left ventricular end-diastolic diameter，LVEDD），以比较各组间心功能改善情况。此外，还比较了各组间导线参数（R波感知、起搏阈值及阻抗）及并发症的发生情况。

研究结果：研究最终纳入64例HF患者，其中LBTP组16（25.0%）例，LAFP组22（34.4%）例，LPFP组26（40.6%）例，三组间基线特征均衡可比。LBBAP的导线头端主要分布于4区和5区，头端位置与不同的LBBAP起搏模式相关：4区更易出现LBTP而5区更易出现LAFP或LPFP。 LBBAP后患者心室除极参数QRSd、IVD及复极参数QTc、TpTe和TpTe/QTc均显著下降，且与起搏模式无关。心功能改善方面，LBBAP后患者LVEF显著提高（LVEF 47.15±10.75 vs. 32.81±6.79%，P<0.001），LVEDD显著降低（LVEDD 56.44±8.44 vs. 65.00±8.94mm，P<0.001），三组间超声反应率及超反应率无组间差异（有反应率：LAFP vs. LBTP vs. LPFP:89.5% vs. 84.6% vs. 75.0%；超反应率：LAFP vs. LBTP vs. LPFP:42.1% vs. 38.5% vs. 30.0%；P>0.05）。此外，三组患者随访期间导线参数均稳定且无明显组间差异。

研究结论：LBBAP在HF患者中的起搏模式主要以LPFP为主。LBBAP后HF患者心室除极同步性及复极稳定性显著改善，且不受起搏模式的影响。不同LBBAP起搏模式间的超声反应率也无明显差异。

第二部分 左束支区域起搏对心力衰竭患者心室复极参数的影响及与双心室起搏的比较

研究背景：传统的双心室起搏（Biventricular pacing，BiVP）可显著改善心力衰竭（Heart failure，HF）患者的心功能并降低死亡率。但该术式也有一定的局限性，如约5%~10%的HF患者因解剖因素等无法植入左室电极，而成功行BiVP的HF患者也有约30%~50%表现为无反应。目前，左束支区域起搏（Left bundle branch area pacing，LBBAP）已逐渐成为心脏再同步治疗（Cardiac resynchronization therapy，CRT）适应症患者的新选择。本研究旨在探讨LBBAP及其不同除极优化方式对HF患者心室复极离散度（Dispersion of ventricular repolarization，DVR）的影响，并与传统BiVP进行比较，从而探讨更有利于HF患者复极稳定性的CRT术式。

研究方法：回顾性纳入本中心成功行LBBAP或BiVP的HF患者。收集患者术前自身心律和术后24小时的起搏优化心电图。测量心室除极参数QRS时限（QRS duration，QRSd）和复极参数（QTc、T波峰末间期[TpeakTend，TpTe]和TpTe/QTc）。在LBBAP组患者中评估LBBAP及不同优化方式（AV间期优化，增加右心室起搏[Right ventricular pacing，RVP]以及LBBAP优化的CRT[LBBAP optimized CRT，LOT-CRT]）对HF患者心室除极和复极参数的影响。此外，还比较了LBBAP和BiVP起搏前后心室除极与复极相关心电参数的变化。

研究结果：研究共纳入109名HF患者，其中LBBAP组49名，BiVP组60名。LBBAP后，HF患者的QRSd（128.83±10.93 vs.171.06±17.12ms，P<0.001）和DVR（QTc：450.50±34.11 vs. 477.07±46.27ms，P<0.001；TpTe：83.13±12.40 vs.105.42±15.11ms，P<0.001；TpTe/QTc：0.20[0.19，0.22] vs.0.23[0.21，0.24]，P<0.001）均较自身心律下显著改善。 LBBAP不同的除极优化方式对心室复极参数影响不同：AV间期优化后，患者的QRSd进一步缩短但复极参数无显著变化；LOT-CRT可进一步缩短起搏QRSd但显著增加TpTe（95.58±15.86 vs.77.84±17.51ms，P=0.003）和TpTe/QTc（0.23±0.02 vs.0.19±0.02，P=0.001）；在LBBAP的基础上增加RVP对QRSd和QTc无明显影响却显著增加了TpTe（91.57±12.45 vs.86.38±12.29ms，P=0.03）和TpTe/QTc（0.21±0.02 vs.0.20±0.02，P=0.03）。此外，LBBAP和BiVP两组间的基线复极参数相同，但起搏心律下LBBAP组的TpTe（84.39±14.12 vs. 113.20±15.68ms，P<0.001），QTc（446.85±38.56 vs. 480.47±34.69ms，P<0.001)以及TpTe/QTc （0.20±0.02 vs. 0.25±0.03，P<0.001）均短于BiVP组。

研究结论：LBBAP可在缩短QRSd的同时显著改善HF患者的DVR。LBBAP不同的除极优化方式对心室除极和复极参数影响不同：AV间期优化和LOT均可进一步缩短LBBAP患者的QRSd，但前者对DVR没有影响后者却显著增加DVR；而在LBBAP的基础上增加RVP不仅未缩短起搏QRSd还增加患者的DVR。此外，LBBAP相比BiVP可在术后即刻为HF患者提供更好的复极稳定性。

第三部分 左束支区域起搏的心室复极参数对心力衰竭患者超声反应性的预测作用

研究背景：目前，左束支区域起搏（Left bundle branch area pacing，LBBAP）已成为终末期心力衰竭（Heart failure，HF）患者心脏再同步治疗（Cardiac resynchronization therapy，CRT）新的术式选择。然而有研究报道LBBAP的术后有反应率仅为70%，且LBBAP反应性的预测因子尚未阐明。已知复极参数与心肌力学和心脏收缩功能密切相关，但其与LBBAP反应性的相关性尚不明确。本研究旨在探讨LBBAP对心室复极参数的影响并进一步评估基线或起搏状态下的心室复极参数对LBBAP反应性的潜在预测价值。

研究方法：回顾性纳入本中心成功行LBBAP的HF患者，测量基线及起搏后24小时的心室复极离散度（Dispersion of ventricular repolarization，DVR）指标QTc、T波峰末间期（TpeakTend，TpTe）和TpTe/QTc，并按照基线QRS形态分组比较起搏前后的复极参数变化。记录基线和随访的心脏超声指标（如左室射血分数[Left ventricular ejection fraction，LVEF]和左室舒张末期内径）评估患者的反应率。用多因素Logistic回归分析的方法评估复极参数与LBBAP反应性之间的相关性。通过受试者工作特征曲线（Receiver operating characteristic，ROC）分析获得最佳截断值，并以该截断值为界分组后比较两组的心功能改善情况。此外，分别在左束支传导阻滞（Left bundle branch block，LBBB）组和非LBBB组评估该截断值的预测作用。

研究结果：研究最终纳入59例成功行LBBAP的HF患者。LBBAP后DVR的变化趋势受基线QRS形态的影响：束支阻滞（Bundle branch block，BBB）患者和CRT升级患者的DVR显著改善，而窄QRS患者略有增加。LBBAP后总研究人群的LVEF显著改善，超声反应率为74.6%（44/59）。在48例宽QRS患者中，有反应组（n=36）和无反应组（n=12）的基线特征均衡可比，而前者的起搏TpTe显著低于后者（77.2[73.3，81.0] vs. 90.1[82.0，99.3]，P<0.01）。多因素Logistic回归分析进一步表明起搏TpTe是宽QRS患者LBBAP反应性的独立预测因子（P<0.05）。ROC分析显示起搏TpTe预测LBBAP反应性的曲线下面积为0.77（95%置信区间 0.60~0.93），最佳截断为81.2ms（P<0.01）。TpTe<81.2ms的患者超声反应率明显较高（93.3% vs. 44.4%，P<0.01）。进一步亚组分析表明，起搏TpTe对LBBAP反应的预测价值在非LBBB组比LBBB组患者更显著。此外，在QRSd<130ms的患者中未观察到起搏TpTe和LBBAP反应性之间的相关性。

研究结论：基线QRS形态可影响LBBAP后HF患者的DVR：BBB患者和CRT升级患者的DVR显著改善而窄QRS患者略有增加。宽QRS患者的起搏TpTe与LBBAP后的超声反应性显著相关，尤其在非LBBB患者。

Part I The association between QRS morphology after left bundle branch area pacing and electrocardiogram characteristics and echocardiographic response in heart failure patients

Background: Left bundle branch area pacing (LBBAP) is a novel physiological pacing technique applicable to patients with bradycardia and also a new option for cardiac resynchronization therapy (CRT). The initial methodology of LBBAP emphasized left bundle trunk pacing (LBTP), resulting in a relatively lower success rate, serious septal injury, and longer operation time. Recently, many physicians proposed the concept of branch pacing, such as left anterior branch pacing (LAFP) and left posterior branch pacing (LPFP), which substantially facilitates the procedure with various pacing QRS morphology. However, whether the outcomes of various pacing modes differ in heart failure (HF) patients is still unclear. This study aimed to compare the electrophysiological characteristics and echocardiographic response rate among those distinct modes of LBBAP.

Methods: HF patients undergoing successful LBBAP at our center were retrospectively included. Different modes of pacing were determined based on paced QRS morphology. The fluoroscopic images were collected to compare the lead tip position between the groups. The electrocardiograms (ECG) before and after LBBAP were used to measure the depolarization parameters (QRS duration [QRSd] and the interventricular delay [IVD]), and the repolarization parameters (QTc, TpeakTend[TpTe], and TpTe/QTc). The left ventricular ejection fraction (LVEF) and left ventricular end-diastolic diameter (LVEDD) of patients at baseline and follow-up were also recorded from medical records systems. The comparison of those ECG and echocardiographic parameters among groups were then performed. In addition, the lead parameters (R wave perception, pacing threshold, and impedance) and complications were compared.

Results: A total of 64 HF patients were finally included, including 16 (25.0%) patients in the LBTP group, 22 (34.4%) patients in the LAFP group, and 26 (40.6%) patients in the LPFP group. The LBBAP lead tips were mainly distributed in zone 4 and zone 5, and the distribution features were significantly related to different pacing modes: LBTP was more likely to be in zone 4 while LAFP or LPFP was prone to locate in zone 5. After LBBAP, the ventricular depolarization parameters (QRSd and IVD) and repolarization parameters (QTc, TpTe, and TpTe/QTc) were significantly decreased, regardless of paced QRS morphology. Besides, the LVEF of the patients was significantly increased (LVEF 47.15 ± 10.75 vs 32.81 ± 6.79%, P<0.001), LVEDD was significantly decreased (LVEDD 56.44 ± 8.44 vs 65.00 ± 8.94 mm, P<0.001). There was no difference in the response rate and super-response rate among groups (response rate: LAFP vs. LBTP vs. LPFP: 89.5% vs. 84.6% vs. 75.0%; super-response rate: LAFP. vs. LBTP vs. LPFP: 42.1% vs. 38.5% vs. 30.0%; P>0.05). In addition, the lead parameters remained stable and no significant differences were observed among groups.

Conclusion: LPFP was the main pacing mode among HF patients after LBBAP. The paced QRS morphology was significantly related to the position of lead tips. After LBBAP, the ventricular depolarization synchronization and repolarization stability were both significantly improved, regardless of pacing modes. There was no significant difference in the echocardiographic response rate among distinct LBBAP modes.

Part II: The immediate changes of repolarization parameters after left bundle branch area pacing and the comparison with traditional biventricular pacing in heart failure patients

Background: Traditional biventricular pacing (BiVP) is an effective strategy for heart failure (HF) patients to improve cardiac function and reduce mortality, although 5%~10% of patients undergo BiVP failure due to anatomical factors, phrenic nerve stimulation, or unacceptable thresholds, and up to 30%~50% of BiVP patients showed non-responsive. Left bundle branch area pacing (LBBAP) has become a new option for cardiac resynchronization therapy (CRT) patients with remarkable performance in improving ventricular depolarization synchronization and cardiac function. Yet its influence on the dispersion of ventricular repolarization (DVR) remains unclear. The study aimed to assess the immediate effects of LBBAP and its optimization approaches on DVR and further compared those effects with BiVP, thus exploring better CRT delivery concerning repolarization stability in HF patients.

Methods: HF patients undergoing successful BiVP or LBBAP implantation were retrospectively and consecutively enrolled for analysis. Electrocardiograms (ECGs) were collected before and around 24 hours after implantation. QRSd and repolarization parameters (QTc, Tpeak-Tend [TpTe], and TpTe/QTc) were measured. The changes of those ECG parameters after LBBAP and various LBBAP optimization approaches (including AV interval optimization, LBBAP-optimized CRT[LOT-CRT], and the addition of right ventricular pacing[RVP]) were assessed in the LBBAP group. Besides, the comparison of the above parameters between LBBAP and BiVP was also performed.

Results: A total of 109 HF patients were included in the study, with 49 patients in the LBBAP group and 60 in the BiVP group. After LBBAP, the QRSd shortened (128.83±10.93 vs.171.06±17.12ms, P<0.001) and DVR significantly improved: QTc (450.50 ± 34.11 vs 477.07 ± 46.27 ms, P<0.001), TpTe (83.13±12.40 vs 105.42±15.11 ms, P<0.001), and TpTe/QTc (0.20 [0.19, 0.22] vs.0.23 [0.21, 0.24], P<0.001) all decreased. Various optimization approaches of LBBAP showed different influences on DVR. Both AV optimization and LOT-CRT further shortened QRSd, but the former showed no effect on DVR while the latter significantly increased DVR (TpTe [95.58 ± 15.86 vs.77.84 ± 17.51ms, P=0.003)]and TpTe/QTc [0.23 ± 0.02 vs.0.19 ± 0.02, P=0.001]). The addition of RVP increased DVR (TpTe [91.57 ± 12.45 vs.86.38 ± 12.29ms, P=0.03] and TpTe/QTc [0.21 ± 0.02 vs.0.20 ± 0.02, P=0.03]) without shortening QRSd. The post-operative DVR decreased in the LBBAP group while increased in the BiVP group compared to baseline. A significant difference was shown between LBBAP and BiVP groups in post-operative TpTe (84.39±14.12 vs. 113.20 ± 15.68, P<0.001), QTc (446.85 ± 38.56 vs. 480.47 ± 34.69, P<0.001), and TpTe/QTc (0.20 ± 0.02 vs. 0.25 ± 0.03, P<0.001) despite their baseline consistency.

Conclusions: LBBAP significantly improved DVR while shortening QRSd in HF patients. Various optimization approaches of LBBAP showed different influences on DVR: both AV optimization and LOT-CRT further shortened QRSd, but the former showed no effect on DVR while the latter significantly increased DVR; the addition of RVP increased DVR without shortening QRSd. LBBAP may provide better repolarization stability in comparison with BiVP immediately after the implantation in HF patients.

Part III The association between repolarization parameters and echocardiographic response in heart failure patients after left bundle branch area pacing

Background: Left bundle branch area pacing (LBBAP) has become a safe and effective strategy for end-stage heart failure (HF) patients indicated for cardiac resynchronization therapy (CRT) and/or ventricular pacing. Yet the response rate for LBBAP was only 70%. The predictive factors for LBBAP response remain to be further elucidated. Repolarization parameters were demonstrated to be associated with cardiac mechanics and systolic function. This study aimed to investigate the effects of LBBAP on repolarization parameters and then evaluate the potential predictive value of those parameters for the echocardiographic response.

Methods: HF patients undergoing successful LBBAP were consecutively included in the study. The parameters reflecting dispersion of ventricular repolarization (DVR), including QTc, Tpeak-Tend (TpTe), and TpTe/QTc, were measured before and around 24 hours after LBBAP implantation. The baseline and pacing parameters were compared according to baseline QRS morphology. Baseline and follow-up echocardiographic indices such as left ventricular ejection fraction (LVEF) and left ventricular end-diastolic diameter were recorded. The association between repolarization parameters and LBBAP response was then evaluated among patients with wide QRS using multivariate logistic analysis. Receiver operating characteristic analysis was further performed to get the optimal cutoff value. The predictive value was then assessed in the left bundle branch block (LBBB) group and the non-LBBB group, respectively.

Results: A total of 59 HF patients were finally included. Tends of repolarization parameters varied according to different QRS configurations at baseline: DVR decreased among patients with bundle branch block (BBB) or CRT upgrade while slightly increased in patients with narrow QRS. LVEF was significantly improved with a response rate equal to 74.6% (44/59) for the total cohort. Among the 48 patients with QRS≥130ms, baseline characteristics were comparable between responders (n=36) and non-responders(n=12), with only post-implant TpTe significantly longer in non-responders than responders（90.1[82.0，99.3] vs. 77.2[73.3，81.0]， P<0.01）. Multivariate analysis including post-implant TpTe and previously identified predictors demonstrated that post-implant TpTe was the independent predictor of LBBAP response (P<0.05). ROC analysis indicated an area under the curve of 0.77 (95% CI, 0.60-0.93) with a cutoff value of 81.2ms (P<0.01). Patients with post-implant TpTe<81.2ms had a significantly higher rate of echocardiographic response (93.3 vs 44.4%, P<0.01). Further subgroup analysis indicated that the predictive value of post-implant TpTe for LBBAP response was more significant in non-LBBB patients than in LBBB patients. Besides, no association between post-implant TpTe and LBBAP response was identified among patients with narrow QRS.

Conclusions: Trends of repolarization parameters after LBBAP varied according to different QRS configurations at baseline: DVR significantly decreased among patients with BBB or CRT upgrade while slightly increased in patients with narrow QRS. Post-implant TpTe was associated with the echocardiographic response after LBBAP among patients with wide QRS, especially for non-LBBB patients.