Projection selection and rapid atrial pacing improves early outcomes after self-expanding transcatheter aortic valves

INTRODUCTION

Transcatheter aortic valve implantation (TAVI) has become an established option for symptomatic severe aortic stenosis across risk groups.1,2 The implantation technique is moving towards a less invasive procedure with shorter hospitalizations.3,4 Because the length of hospital-stay has been related to major adverse cardiovascular events during follow-up,5 optimizing the procedure remains a concern. Different studies have demonstrated that patients can be safely discharged 72 and even 48 hours after TAVI, and various protocols with a minimalist approach have been described so far, mainly with balloon-expandable valves.6-9 Nevertheless, the potential need for a permanent pacemaker implantation (PPI) frequently prolongs hospital admissions for the purpose of rhythm monitorization after TAVI.10,11 This issue is particularly relevant for self-expanding valves,12-14 since the rate of PPI is highest amongst these models and most early-discharge strategies have been reported for balloon-expandable prostheses.9,15

The rate of PPI can be reduced by guiding TAVI by the cusp-overlap projection (COP) to isolate the non-coronary cusp and better control prothesis implantation depth.16,17 In addition, sequential rapid atrial pacing (RAP) after deployment has been proposed to screen for atrioventricular (AV) conduction disorders.18 However, the effect of combining both preventive and diagnostic techniques on clinical outcomes after self-expanding TAVIs remains unknown. Therefore, the present study was designed to determine whether compared to conventional care, the combined use of COP and RAP can improve the management of conduction disturbances and the early outcomes after self-expandable TAVI procedures.

METHODS

Study population

This is a one-center retrospective analysis of consecutive patients who underwent transfemoral self-expandable TAVI between January 2018 to January 2022. All patients had symptomatic severe aortic stenosis and were deemed eligible for TAVI by a multidisciplinary heart team. We included all transfemoral cases irrespectively of surgical risk. From this cohort, we selected patients in sinus rhythm undergoing a self-expandable prothesis implantation procedure. We excluded patients with permanent atrial fibrillation or permanent pacemakers, and patients who did not undergo follow-up in our center. Procedures performed before 2018 were excluded from our analysis to test our hypothesis on a homogeneously contemporary series.

The institutional Ethics Committee approved the study, and all subjects provided written informed consent.

TAVI procedure

Pre-TAVI workup included baseline electrocardiogram (ECG), transthoracic echocardiography, coronary and peripheral angiography, as well as computed tomography. Implanting technique was standard among operators, following the conventional “minimal approach” previously described3,4,19: all invasive lines were minimized, and procedures were performed under minimal sedation and local anesthesia. Femoral access was obtained using ultrasound guidance and double percutaneous vascular closure (Perclose, Abbott, MN, United States) was placed before non-fractionated heparin administration. During the procedure, all patients received a temporary pacemaker via femoral venous access using a balloon-tipped bipolar catheter (5 French, Arrow, Teleflex, USA) positioned in the right ventricular apex. Pre and post valvular dilation during the intervention were performed according to the operator’s judgement. After implantation, a supravalvular angiography was routinely performed to check for paravalvular leak. At the end of the procedure, an angiography of the main vascular access was performed on all patients to verify arteriotomy closure success as well as ilio-femoral vessel integrity. Until May 2020 (standard-of-care cohort), the temporary pacemaker was kept for 24 hours with monitorization in the cardiac intensive care unit, with later transfer to the cardiology ward. Permanent pacemaker was implanted in the presence of persistent or new-onset complete or high degree AV block.

Since June 2020, the procedure was modified to routinely include: a) COP during valve deployment to assess implantation depth, and b) RAP at the end of the procedure in the absence of complete AV block, at rates of 70 to 120 beats/min (or until pacing-induced Wenckebach heart block was observed) in 10 beats/min increments for a total of 20 beats at each increment.18 In patients with persistent complete AV block, permanent pacemaker was placed in the same day. In the absence of immediate complications, patients were transferred to a standard monitored ward bed minimizing ICCU stay unless clinically needed, and temporary pacemaker surveillance was only maintained for 24 hours in the presence of Wenckebach phenomenon or in patients with AV block until permanent pacemaker was implanted. Final determination of patient discharge was based on patient’s characteristics and clinical status by the referring physician.

Follow-up

After implantation, assessment included ECG, echocardiogram, in-hospital, and 90-day clinical follow-up. All patients’ clinical, procedural, in-hospital information, and follow-up data were collected from the electronic medical records.

Study endpoints

The primary safety and efficacy outcomes were complete AV block after implantation with COP and the diagnostic performance of RAP added to COP in self-expandable valves.

The secondary endpoints included the length of hospitalization following TAVI and the composite major adverse cardiovascular events, defined as cardiovascular mortality, PPI, rehospitalization for heart failure decompensation, stroke, and major bleeding and vascular complications occurring 90 days after hospital discharge. In addition, rehospitalizations for healthcare-associated infections within the same period were analyzed separately.

Statistical analysis

Categorical variables are presented as counts (percentages) and were compared using the chi-square test. Continuous variables were tested for normality of distribution by using the Shapiro-Wilk test. According to their distribution, continuous variables are expressed as mean ± standard deviation or median values with interquartile range [IQR], and were compared using Student t test, ANOVA or Wilcoxon test, as appropriate. For the multivariate analysis, we first selected baseline variables that showed a significant association with the outcome (P < .2) and performed single logistic regression. Variables with a P-value < .05 were entered into a multiple logistic regression analysis. Odds ratios (OR) were calculated using uni and multivariate logistic regression, along with 95% confidence intervals. Statistical analyses were conducted using RStudio 4.1.20 P-values < .05 were considered significant.

RESULTS

Baseline characteristics

A total of 273 patients met the inclusion criteria for the analysis. Among them, 134 underwent standard implantations, and 139 underwent the combined use of COP and RAP (COP+RAP) (figure 1). Mean patient age was 81 ± 7 years old and 51.3% were female, with similar prevalence of cardiovascular risk factors and comorbidities between groups. Patients in the standard-of-care group showed a worse NYHA functional class, and there were no differences in the surgical risk (table 1). Both cohorts presented similar rates of conduction disturbances at baseline. The mean transvalvular pressure gradient by ultrasound was 47 [41-58] mm Hg with no significant differences on baseline findings between groups. The COP+RAP group showed a higher valvular calcium score (2429 [1577-3557] Agatston units) and larger left ventricular outflow tract and annular perimeters (P = .006 and P = .04 respectively) (table 1).

Figure 1. Patient flow chart. AF, atrial fibrillation; COP, cusp-overlap projection; RAP, rapid atrial pacing TAVI, transcatheter aortic valve implantation.

Table 1. Demographics, baseline electrocardiographic and imaging characteristics

Procedure and in-hospital outcomes

Procedural details are listed in table 2. All patients underwent TAVI under conscious anesthesia, with more frequent use of midazolam and fentanyl in the COP+RAP cohort in place of dexmedetomidine (P < .001). Also, in line with current recommendations for self- expandable valves, the rate of predilation was significantly higher in the COP+RAP group (P < .001).

The occurrence of intraprocedural conductance disorders was lower with COP in the COP+RAP group (25.8% vs 14.4%; P = .007; OR, 0.49; 95%CI, 0.27-0.91; P = .026) mainly due to a reduction in temporary complete AV block (table 2), whereas rates of new-onset first degree AV block and atrioventricular block (LBBB) remained similar between groups. Same day permanent pacemaker was implanted in 16 patients of the COP+RAP group with persistent AV block. RAP revealed Wenckebach phenomenon at a median stimulation rate of 115 bpm in 24 patients and allowed to reduce the need for temporary pacemaker surveillance from 91.8 % in the standard-of-care group to 28.1 % of the cases in the COP+RAP group (P < .001). Among the patients in the COP+RAP cohort who did not develop Wenckebach phenomenon (n = 98/122, 80.3%), permanent pacemaker was required in only 3% of patients (OR, 0.16; 95%CI, 0.03-0.76; P = .02; negative predictive value = 97%; 95%CI, 91-99%. Patients with pacing-induced Wenckebach had a higher risk of PPI during admission (OR, 6.33; 95%CI, 1.30-34.29; P = .021; positive predictive value 79% 95%CI, 58-93%).

Table 2. Procedural characteristics, electrocardiographic and imaging outcomes

Incidence of AV block and timing of PPI after TAVI is shown in figure 2. Delayed AV block (developed > 48 hours after TAVI) during admission occurred in 8.2% patients in the standard group and 3.6% of the COP+RAP group (P = .14) mostly in the presence of new-onset LBBB (68% of the cases). Adding both maneuvers did not impact the procedure length or led to adverse events, with a low and similar rate of in-hospital complications among groups (table 3). Additionally, reduction of the implantation depth did not lead to a higher rate of residual significant aortic regurgitation (P = .32).

Figure 2. Immediate complete AVB and on-admission permanent pacemaker implantation trends. Upper: intraprocedure cusp-overlap projection and Wenckebach phenomenon during atrial pacing. Midd: comparison of immediate AVB in standard and protocol groups. Bottom: trend in pacemaker implantation during admission. AVB, atrioventricular block; COP, cusp-overlap projection; RAP, rapid atrial pacing TAVI, transcatheter aortic valve implantation.

Table 3. Immediate, and short-term outcomes

The reduction of stay in the intensive care unit along with the lower rate of 24-hour temporary pacemaker surveillance, significantly decreased the length of hospitalization in the COP+ RAP cohort (−3.0 [−2.5 to −3.5] days, P < .0001), enabling to discharge patients within 48 hours in 56.8% of the cases compared to 7.5% in the standard-of-care group (P < .001).

90-day outcomes

At 90-days, the cumulative rate of PPI in the COP+RAP cohort persisted lower than in the standard-of-care group (29.5% vs 18.7%; P = .049). Of all 65 pacemaker implantations, 58 (89%) occurred during the index admission and 7 (11%) following hospital discharge. Among the 3 patients from the COP+RAP requiring late PPI, 2 developed with new-onset LBBB (QRS width 127 and 136 ms respectively), and both had displayed Wenckebach phenomenon at 120 bpm. They were monitored with temporary pacemaker for 24 hours without events but developed high degree AV block within 48 hours after discharge (day 4 after TAVI). The third patient had new-onset first degree AV block (PR interval 240 ms), Wenckebach phenomenon at 120 bpm and narrow QRS (118 ms) but developed high-degree AV block at day 55. No patients with a negative RAP test required late PPI. The univariate logistic analysis of potential baseline and procedural factors of PPI after TAVI is reported in table 4. Interestingly, differences in sedation, predilation, valve model (P = .98) and bicuspid valves (P = .71) did not influence the risk of PPI after TAVI. In the multivariate analysis, prior 1st degree AV block, baseline right bundle branch block, prosthesis posdilation and new-onset LBBB were associated with PPI. The combination of COP with RAP decreased the need of PPI on a 90-day follow-up (OR, 0.48; 95%CI, 0.24-0.92; P = .031) (figure 3).

Figure 3. Odds plot of multivariate analysis of 90-day permanent pacemaker implantation predictors. 95%CI, 95% confidence interval; AVB, atrioventricular block; COP, cusp-overlap projection; LBBB, left bundle branch block; PPI, permanent pacemaker implantation; RAP, rapid atrial pacing; RBBB, right bundle branch block.

Table 4. Main factors associated with PPI after TAVI: unadjusted logistic regression analysis of PPI predictors at 90 days

The COP+RAP group presented a significantly lower percentage of the secondary composite endpoint than the standard-of-care cohort (20.1% vs 10.1%; P = .031). The rates of HF readmission, and mortality were low and similar across groups during the follow-up period (table 3). None of the 3 deaths during follow-up were due to rhythm disturbances. At 90 days, there was a reduction in the number of readmissions due to admission-related infections in the COP+RAP cohort (OR, 0.35; 95%CI, 0.12-0.89 P = .036) (figure 4), with an 8.6 % reduction in the odds of infection for each day of reduced hospitalization (OR, 0.91; 95%CI, 0.85-0.99; P = .023). The most common were urinary (47.4%) and respiratory (42.1%) infections. None of the patients presented infections related to PPI.

Figure 4. Central illustration. Main immediate and short-term findings of combining cusp-overlap projection and rapid atrial pacing when implanting self- expandable valves. OR, odds ratio; TAVI, transcatheter aortic valve implantation.

DISCUSSION

In the present study we demonstrate that combining the COP with RAP, COP reduces the risk of complete AV block after self- expanding TAVI and negative RAP after implantation adequately dismiss underlying AV conduction disturbances requiring PPI.

This strategy reduces the need of temporary PPI and the appearance of late onset AV block, and safely allows an early discharge in sinus rhythm patients receiving a self-expandable valve. Reducing hospital stays lead to a significant reduction in admission-related infections at short-term follow-up (figure 4).

A minimalistic approach in TAVI has become progressively relevant as a way of simplifying the intervention and promoting early recovery.4,6-8 Multiple studies have shown that early discharge after TAVI is possible, preserving the high standards of effectiveness, patient safety and outcomes.3,9,21 However, most of these studies included highly selected patients, with a predominance of balloon-expandable valves, and might not be representative in the current TAVI scenario. Evidence at the beginning of minimalistic TAVI inception found self-expanding prothesis to be a predictor for rhythm disturbances and delayed discharge.5,6 Since their risk of AV conduction disturbances remain in the range of 5% to 30% even with contemporary transcatheter heart valves10,12-14 and there is no appropriate way to assess the risk of AV block development, self-expandable valves have been underrepresented in early discharge protocols so far.3,7,22,23

Postimplantation conduction disturbances remains one of the current weaknesses in TAVI, and great heterogeneity of recommendations persist regarding their management.24,25 In this regard, the aim of our study was to minimize the appearance and improve the assessment of conduction disturbances without increasing invasiveness in patients receiving self-expandable valves, by combining 2 well-described maneuvers.

Guiding the procedure using the COP has already been established as an effective method to guide implantation depth.16,17 Isolating the non-coronary cusp allows a better identification of the most inferior point of the aortic annulus, leading to a higher (more aortic) valve positioning, which in turn reduces the risk of post-TAVI conduction disturbances. Various observational studies have described the COP technique in distinct types of self-expandable valves,15-17,26 and highlighted its potential benefits in decreasing the PPI rate in the Evolut (Medtronic, United States) and the Portico (Abbott Structural Heart, United States) valves.27 In line with these studies, in our cohort, COP significantly decreased the emergence of intra-procedural transient and late onset ( > 48 hours) high-grade AV block irrespectively of the self-expandable valve type, with no noticeable impact on the rate of new-onset LBBB, that remained around 29% in both cohorts.

After implantation, current consensus documents recommend temporary pacemaker surveillance for at least 24 hours. Further management depends on baseline intra-procedure, and post-procedure conduction disturbances,25,28 that can result in longer duration of temporary pacemaker placement or prolonged inpatient rhythm monitoring. In this context RAP, emerges as an easy and interesting strategy to discriminate which patients may require monitoring. Since its publication in 2020,18 different works have recognized the underuse of this technique, although is an important component of routine electrophysiological studies to “stress” the AV conduction.10 It has the limitation of assessing specifically at the level of the atrio-His interval but seems to be a good indicator of patients with underlying conduction disorders with a described strong negative predictive value for PPI in the absence of Wenckebach phenomenon. In our study, the rate of Wenckebach phenomenon during dynamic atrial stimulation was 19.8% and showed a moderate positive predictive value. However, most importantly, the absence of RAP-induced AV block displayed a very high negative predictive value of PPI during follow-up. These results are in agreement with the previous report of RAP on balloon-expandable valves,18 showing adequate reliability of AV Wenckebach testing over delayed conduction disturbances immediately at the end of the procedure in self-expandable valves. The improvement in valve deployment and risk stratification of AV conduction disturbances allowed us to reduce temporary pacemaker surveillance in 63.1% of cases, with less late-onset high-degree AV block and no impact in procedure length or complications, leading to early recovery and minimizing functional decline. Moreover, although early discharge protocols described so far have focused on demonstrating safety, it should be noted that in our simplified protocol cohort there was a significant reduction in the healthcare-related infection readmission rate at 90 days. The reduction in the incidence of admission-related respiratory and urinary infections was likely attributable to the shortened length of hospitalization, that has been associated with adverse events.5 Observed differences in major adverse cardiovascular events, although significant, should be interpreted cautiously, as part of the findings might be a result of technological advances and cumulative expertise.

The adequacy of these combined maneuvers was patent in the multivariate analysis and in the maintained differences in PPI rate at a 90-day follow-up. Nevertheless, special caution should be recommended in patients developing persistent LBBB after TAVI, as it was a predictor of delayed AV block and PPI. These patients would probably benefit from longer monitoring, and in the presence of new-onset LBBB or Wenckebach-induced phenomenon, a formal electrophysiological study or continuous ECG monitoring might have the potential to risk-stratify patients with unclear indications of PPI following TAVI.29-31

Limitations

This is a single-center study with a modest sample size and an observational design; therefore, there is an inherent potential bias due to patient selection. First, both cohorts are not contemporary, which may introduce bias related to technological advancements or increased operator experience over time. Second, RAP is restricted to patients in sinus rhythm, and the cutoff of 120 beats/min for maximum RAP rate was selected according to previous literature but may be insufficient as low risk patients tend to be young and might request testing at higher heart rates. Also, RAP stresses the AV conduction at the atrio-His level, and therefore our results might not properly assess patients with broadened QRS, as high-grade conduction block occurs at the His-Ventricular segment.18,30,31 Third, our protocol has the inconvenience of requiring temporary pacemaker so it can be moved to the atrium at the end of the procedure, which is not feasible in procedures performing pacing through the left ventricular wire. However, we believe it might be a safe and efficient way to risk-stratify patients immediately after the procedure. Finally, our study was focused on patients in sinus rhythm and self-expanding prothesis and therefore the results and conclusions should be interpreted in this context.

CONCLUSIONS

Combining COP during deployment with RAP immediately after TAVI diminishes the incidence of postprocedural AV block and improves the screening of underlying atrio-His conduction disturbances in patients receiving a self-expandable valve. In turn, this allows to reduce the length of temporary pacemaker surveillance and admission days, with differences in PPI rate maintained at a 90-day follow-up. Optimizing strategies of care after TAVI has a significant impact at short-term, driven by a significant reduction in readmissions due to infections.

FUNDING

M. Tamargo was partially supported by grants from the Fundación para la Investigación Biomédica Gregorio Marañón, Spain, and Rio Hortega CM20/00054 from the Instituto de Salud Carlos III, Spain. J. Bermejo was partially supported by INT22/00025 from the Instituto de Salud Carlos III, Spain.

ETHICAL CONSIDERATIONS

The institutional Ethics Committee approved the study, and all subjects provided written informed consent. This manuscript complies with the guidelines of the SAGER guidelines, and possible gender and/or age differences have been taken into account, without significant differences being observed.

STATEMENT ON THE USE OF ARTIFICIAL INTELLIGENCE

No artificial intelligence has been used in the development of this paper.

AUTHORS’ CONTRIBUTIONS

All authors comply with the international characteristics of authorship of scientific articles. Conception and design: M. Tamargo, J. García Carreño, E. Gutiérrez, J. Bermejo, F. Fernández-Avilés. Data collection: M. Tamargo, M. Huanca, E. Ludeña, J. García Carreño. Analysis: M. Tamargo E. Gutiérrez, J. Bermejo. The manuscript was reviewed and approved by all authors.

CONFLICTS OF INTEREST

Nothing to disclose.

REFERENCES