Evolution and prognosis of tricuspid and mitral regurgitation following cardiac implantable electronic devices: a systematic review and meta-analysis

Abstract Aims Significant changes in tricuspid regurgitation (TR) and mitral regurgitation (MR) post-cardiac implantable electronic devices (CIEDs) are increasingly recognized. However, uncertainty remains as to whether the risk of CIED-associated TR and MR differs with right ventricular pacing (RVP) via CIED with trans-tricuspid RV leads, compared with cardiac resynchronization therapy (CRT), conduction system pacing (CSP), and leadless pacing (LP). The study aims to synthesize extant data on risk and prognosis of significant post-CIED TR and MR across pacing strategies. Methods and results We searched PubMed, EMBASE, and Cochrane Library databases published until 31 October 2023. Significant post-CIED TR and MR were defined as ≥ moderate. Fifty-seven TR studies (n = 13 723 patients) and 90 MR studies (n = 14 387 patients) were included. For all CIED, the risk of post-CIED TR increased [pooled odds ratio (OR) = 2.46 and 95% CI = 1.88–3.22], while the risk of post-CIED MR reduced (OR = 0.74, 95% CI = 0.58–0.94) after 12 and 6 months of median follow-up, respectively. Right ventricular pacing via CIED with trans-tricuspid RV leads was associated with increased risk of post-CIED TR (OR = 4.54, 95% CI = 3.14–6.57) and post-CIED MR (OR = 2.24, 95% CI = 1.18–4.26). Binarily, CSP did not alter TR risk (OR = 0.37, 95% CI = 0.13–1.02), but significantly reduced MR (OR = 0.15, 95% CI = 0.03–0.62). Cardiac resynchronization therapy did not significantly change TR risk (OR = 1.09, 95% CI = 0.55–2.17), but significantly reduced MR with prevalence pre-CRT of 43%, decreasing post-CRT to 22% (OR = 0.49, 95% CI = 0.40–0.61). There was no significant association of LP with post-CIED TR (OR = 1.15, 95% CI = 0.83–1.59) or MR (OR = 1.31, 95% CI = 0.72–2.39). Cardiac implantable electronic device–associated TR was independently predictive of all-cause mortality [pooled hazard ratio (HR) = 1.64, 95% CI = 1.40–1.90] after median of 53 months. Mitral regurgitation persisting post-CRT independently predicted all-cause mortality (HR = 2.00, 95% CI = 1.57–2.55) after 38 months. Conclusion Our findings suggest that, when possible, adoption of pacing strategies that avoid isolated trans-tricuspid RV leads may be beneficial in preventing incident or deteriorating atrioventricular valvular regurgitation and might reduce mortality.


Conclusion
Our findings suggest that, when possible, adoption of pacing strategies that avoid isolated trans-tricuspid RV leads may be beneficial in preventing incident or deteriorating atrioventricular valvular regurgitation and might reduce mortality.
• Right ventricular pacing via cardiac implantable electronic device (CIED) with trans-tricuspid right ventricular (RV) leads was associated with increased risk of both significant post-CIED tricuspid regurgitation (TR) and mitral regurgitation (MR).The risk of TR more than quadrupled, while the risk of MR more than doubled following CIED with trans-tricuspid RV lead pacing.
• Conduction system pacing significantly reduced the risk of CIED-related MR.
• Cardiac resynchronization therapy considerably reduced the risk of secondary MR, but did not significantly impact the risk of TR.The reduction in secondary MR post-cardiac resynchronization therapy (CRT) was comparable across a wide range of quantitative parameters used for reporting mitral regurgitation severity.

Introduction
Tricuspid regurgitation (TR) and mitral regurgitation (MR) whether structural or functional are independently associated with poor survival. 1,24][5][6][7][8] However, uncertainty remains as to whether the risk of significant TR is different following transvenous right ventricular pacing (RVP) via CIED with trans-tricuspid RV leads, compared with conduction system pacing (CSP), cardiac resynchronization therapy (CRT), and leadless pacing (LP).The first objective of this systematic review and meta-analysis was therefore to determine the risk of new significant TR defined as ≥ moderate TR between different pacing strategies or device types and the impact of this post-CIED TR on all-cause mortality.
While some studies have suggested that RVP via CIED with transtricuspid RV leads is associated with new or worsening MR, [9][10][11][12][13] others have not, [14][15][16][17][18] and the impact of other pacing strategies such as CSP and LP on MR remains unknown.While narrative reviews have reported that up to 40% of patients who have conventional indications for CRT have significant secondary MR at baseline 19 and that CRT reduces systolic MR by ∼40%, 20,21 the magnitude of benefit remains controversial, and a quantitative synthesis of the extant data on the effects of CRT on secondary MR is lacking.Hence, the second objective of this study was to determine the risk of new significant MR defined as ≥ moderate MR post-CIED implantation and the impact of significantly persistent MR post-CRT on all-cause mortality.Contemporary pacing guidelines increasingly favour CSP over traditional RVP via CIED with transtricuspid RV leads. 22,23Identifying the pacing strategies that are least likely to be associated with atrioventricular valvular insufficiency might potentially have beneficial clinical implications.

Methods
This study was registered with the international prospective register of systematic reviews, PROSPERO (registration number CRD42021274269).

Search strategy
PubMed/MEDLINE, EMBASE, and Cochrane Library databases were systematically searched to identify all relevant English language studies restricted to human adults published from inception until 31 October 2023 (see Supplementary material for search terms).

Inclusion criteria
The inclusion criteria consisted of longitudinal studies reporting on patients with baseline pre-CIED [permanent pacemaker (PPM) or implantable cardioverter defibrillator (ICD), or CRT devices] assessment of TR or MR and follow-up post-CIED assessment of TR and MR by echocardiography.Included were studies that reported on TR or MR in a binary or categorical fashion as presence or absence of significant TR or MR defined as new ≥ moderate TR or MR and/or continuous quantitative measures of TR or MR post-CIED compared with the pre-CIED.

Exclusion criteria
Excluded were non-English language studies lacking an English-translated version, studies that did not report clear assessment of TR or MR pre-CIED and post-CIED, and studies that included tricuspid or mitral valve repair or replacement, or trans-catheter percutaneous procedures such as MitraClip and other exclusion criteria stated in Supplementary material online, Figure S1.

Outcomes
The main outcomes were as follows: (i) Prevalence of significant TR or MR post-CIED compared with prevalence at baseline pre-CIED.(ii) Risk of significant TR or MR post-CIED estimated using pooled odds ratio (OR) and 95% confidence interval (CI) for categorical data.(iii) For continuous or quantitative measures used to assess TR or MR, we used standardized mean difference (SMD) of change in post-CIED compared with pre-CIED TR and MR values.(iv) Comparison of the risk of TR post-CIED between ICD and PPM.(iv) Finally, the impact of post-CIED significant TR or MR on all-cause mortality during follow-up.

Quality assessment of studies
The Newcastle-Ottawa Scale (NOS) was used for assessing the quality of non-randomized studies in meta-analyses.We categorized the studies according to NOS as follows: 0-3 = poor quality, 4-7 = fair quality, and 8-9 = good quality (see last column in Supplementary material online, Tables S1 and S2).

Data extraction
Two authors (M.F.Y. and S.A.E.) extracted the data independently using standardized forms containing pre-defined demographic and clinical information.Discrepancies were resolved by consensus.

Statistical analysis
The pooled prevalence and OR and 95% CI of significant TR or MR for binary data and SMD (95% CI) for each continuous quantitative measure of TR or MR pre-CIED and post-CIED were calculated using random-effects model meta-analyses (DerSimonian-Laird), as some degree of heterogeneity was anticipated a priori.The pooled SMD were considered clinically significant for quantitative reduction of secondary TR or MR post-CIED compared with pre-CIED if the value was < 0 and the 95% CI did not cross 0. Hazard ratios and 95% CI for the risk of mortality associated with post-CIED TR or MR were also computed.
Between-study variation in ORs or HRs for categorical data and SMD for continuous data across studies that were attributable to heterogeneity was assessed by I 2 statistic.Publication bias was assessed by eye-balling funnel plots and contour-enhanced funnel plots and Egger's test of bias.Where there was suggestion of publication bias, this was further explored using trim and fill analysis.Analyses were performed with STATA software 18 (Stata Corp, Texas).

Duration of follow-up for the assessment of tricuspid regurgitation and mitral regurgitation post-cardiac implantable electronic device
The pooled median duration of follow-up post-CIEDs for TR studies at which point TR was assessed was 12 months [interquartile range (IQR) 8.9-20 months], while for MR it was 6 months (IQR 3-12 months).The median duration of follow-up during at which all-cause mortality associated with persistently significant TR post-CIED was ascertained across TR studies was 53 months (IQR 39-70 months), while for MR it was 38 months (IQR 32-48 months).

Implantable cardioverter defibrillator vs. permanent pacemaker and risk of post-cardiac implantable electronic device tricuspid regurgitation
The OR (95% CI) of post-CIED TR in ICD devices compared with pacemaker devices was not significant at 1.14 (0.74-1.75),P = 0.55, I 2 63.07%, and Egger's test P-value = 0.9772 (see Supplementary material online, Figure S10).

Publication bias and small-study effects
Funnel plots of OR against their standard error for TR are shown in Supplementary material online, Figure S11, and MR in Supplementary material online, Figure S12.Within each of these figures, panels A to E depict various pacing strategies, while panel F shows the funnel plot derived from the HR of all-cause mortality.Contour-enhanced funnel plots are shown in Supplementary material online, Figure S13 (panels A to C for TR and D to F for MR).For any asymmetry in funnel plot or discordant distribution of studies in contour-enhanced funnel plot or significant Egger's test P-value observed in the various analyses, trim and fill analysis showed no significant change in results between the observed and the observed + imputed, suggesting less likelihood of publication bias, though the possibility of small-study effects remained when Egger's test P-value was significant.

Discussion
Amongst the 13 723 patients (TR) and 14 387 patients (MR) assessed for the risk of CIED-associated TR and MR, this systematic review and meta-analysis observed the following: (i) All CIED grouped together were associated with a more than two-fold increased risk of  significant TR, while the risk of MR CIED was reduced likely due the high number of CRT studies in whom post-CIED MR significantly decreased.(ii) RVP via CIED with trans-tricuspid RV leads was associated with increased risk of both significant post-CIED TR and MR.The risk of TR more than quadrupled, while the risk of MR more than doubled following implantation of CIED with trans-tricuspid RV leads.(iii) Conduction system pacing significantly reduces the risk of post-CIED MR. (iv) Cardiac resynchronization therapy considerably reduced the risk of secondary MR, but did not significantly impact the risk of TR.The CRT-associated reduction in MR was comparable across a wide range of quantitative parameters used for reporting MR severity such as MR grade, EROA, regurgitant volume, regurgitant fraction, mitral regurgitant jet area, ratio of mitral regurgitant jet area to left atrial area, and vena contracta.(iv) Leadless pacing had a risk-neutral effect on both incident TR and MR post-CIED.(vi) There was no significant difference in post-CIED TR occurrence between ICD and PPM devices.(vii) Post-CIED TR was independently associated with poor survival, and persistence of significant MR post-CRT was independently associated with a two-fold increased risk of all-cause mortality.To the best of our knowledge, this study is the first comprehensive meta-analysis to assess the risk of atrioventricular valvular regurgitation post-CIED across all pacing strategies and device types.
This meta-analysis revealed that significant TR (≥ moderate) increased considerably from 9% pre-CIED to 25% post-CIED.Of note, given that some studies variably excluded patients with any or significant TR at baseline, the observed pre-CIED prevalence of TR in this study is likely an underestimate of real-world prevalence.Prior meta-analyses concluded that the overall incidence of at least one grade increase in TR post-CIED was approximately 25%, and approximately 10% for at least two grades increase. 3,4,8Our findings are similar to those of a prior study that observed OR of new or worsening CIED-associated TR of 2.44 (95% CI 1.58-3.77)when compared with patients without CIED. 4Nonetheless, despite adopting a different methodology, the findings of this meta-analysis complement those of these prior studies with a much larger sample size and additional sub-group analyses across different pacing strategies.We found no significant difference in the rate of    occurrence of CIED-associated TR between ICD and PPM device recipients, which was consistent the findings of others, 9,25,28,40,42,46,56,67 despite two individual studies reporting higher incidence of TR post-ICD devices compared with pacemakers. 43,52This suggests that the size of the lead might not matter.Pacing factors such as the percentage of RV pacing have not been shown to affect the incidence of post-CIED TR. 25,46,55 Right ventricular pacing site in relation to CIED-associated TR has yielded inconsistent results. 56,69his meta-analysis examined non-RV pacing strategies and risk of TR and MR and showed significant benefits of CSP in reducing the risk of   MR in both binary and continuous data analyses, as well as TR in continuous data analysis with significant reduction in TR grade.Contemporary guidelines recommend offering patients CSP, when possible, given that large observational studies have shown its significant superiority over RVP as well as over biventricular pacing with respect to echocardiographic outcomes, all-cause mortality, and heart failure hospitalization. 22,23The results of this present meta-analysis suggest that in addition to the above, CSP should also be considered over RVP via CIED with trans-tricuspid RV leads to reduce the risk of MR and potentially reduce the risk of TR.The benefit of His bundle pacing (HBP) on post-CIED TR has been previously highlighted with improvement seen in TR grade after HPB for CRT as well as for the atrioventricular block. 143Conduction system pacing involves a trans-tricuspid lead, but it remains uncertain why the risk of TR might be less than with conventional RVP.Whether this possible risk reduction in TR might stem from less lead mechanical pressure exerted on the tricuspid valve apparatus by the lighter small diameter (1.4 mm) Medtronic SureScan 3830 lead that has been widely used so far for CSP, compared with the ticker stylet-driven leads or other mechanisms, remains speculative.This is especially so as the size of the lead has not been shown to alter the risk of TR. 9,25,28,40,42,46,56,67 Patients implanted with biventricular pacing devices did not witness any significant change in TR, and this is probably due to less ventricular dyssynchronous changes seen in CRT.These findings are supported by a few studies that showed that CRT compared with non-CRT devices is not associated with significant TR incidence. 18,49The absence of a significant risk of TR in patients with LP and the strong association with transvenous trans-valvular RVP suggest that the presence of transvalvular leads might contribute to the pathophysiologic mechanism of TR post-RVP, and it is possible that a trans-tricuspid lead position rather than RVP per se leads to TR.It is also conceivable that some of the lack of association of LP with TR could be also driven by the mid-to-high septum placement of the leadless pacemakers.However, prior studies have shown conflicting findings between the RV pacing site and risk of TR. 56,69 There is still palpable ambiguity surrounding mechanisms of pacing-induced TR.Various pathophysiologic phenomena have been suggested to interfere with tricuspid valve apparatus leading to TR.5][146] Non-lead-related factors such as atrioventricular dyssynchrony, right ventricular dyssynchrony, pacing-induced left ventricular dysfunction, increased pulmonary artery systolic pressure, RV dilatation, RV dysfunction, tricuspid valve annular dilatation, leaflet tethering, and mal-co-aptation, which eventually culminate to TR, have also been suggested. 7,31,64The risk factors of CIED-associated TR which are not unanimously found include pulmonary artery systolic pressure, lead duration, atrial fibrillation, pre-existing mild TR, diabetes, heart failure, and peripheral vascular disease. 31,63,144,1476][17][18]32,37,38,140 The findings from this meta-analysis, which pooled these studies together, appear to cement RVP via CIED with trans-tricuspid RV leads as an inducer of significant MR post-CIED.The possible mechanisms of pacing-induced MR were first suggested in experimental animal models, which observed that the development of MR was associated with LV dyssynchrony secondary to RVP, leading progressively to worsening LV contractility and increasing LV geometry and left ventricular dilation, mitral annular dilatation, and MR. 148Right ventricular pacing is known to be associated with pacing-induced cardiomyopathy, 149 and functional MR is a now a well-known consequence of LV systolic dysfunction. 150her mechanisms include AV dyssynchrony and adverse changes in the papillary muscle function. 151Given that CSP reduces electrical dyssynchrony, mechanical dyssynchrony, and the chances of left ventricular size and geometry derangements and improves systolic dysfunction, 152 it is therefore understandable why it reduces the risk of functional MR as observed in this meta-analysis.However, like TR, the reason why LP does not lead to significant MR despite its RV location as observed in this study remains unclear.Unlike TR, the increasing frequency of RV pacing worsens MR. 9,140 Improvement in functional MR post-CRT is likely due to an interplay of multiple complex and varied mechanisms leading to LV reverse remodelling ranging from reduction in left ventricular volumes with a decrease in mitral leaflet tenting angles and augmentation of transmitral pressure gradient attributable to increased and coordinated left ventricular contractility to improved coordination of the papillary muscle closing forces, which may facilitate effective mitral valve closure. 20,82,114,123,133,137,138Cardiac resynchronization therapy reduces regurgitant volume and papillary muscle tethering on mitral leaflets leading to improvement in leaflet co-aptation and resynchronization of atrioventricular and inter-and intra-ventricular contraction. 82,123,133ignificant healthcare utilization costs have associated with CRT-nonresponse, highlighting the important role CRT plays in heart failure, 153 and strategies on optimized implementation of CRT use have been proposed. 154

Limitations
This systematic review and meta-analysis had important limitations.First, the majority of included studies was observational cohort studies, many of which were retrospective with all the inherent biases and confounders associated with such a methodology as well as small-study effects.For any asymmetry in funnel plot or discordant distribution of studies in contour-enhanced funnel plot or significant Egger's test P-value observed in the various analyses, trim and fill analysis showed no significant change in results between the observed and the observed + imputed, suggesting less likelihood of publication bias, which increases confidence in the findings.Second, the pooled baseline patient characteristics were widely different amongst the studies, due to inclusion of patients with structurally normal hearts at baseline (bradycardia indications) and patients with heart failure (in CRT studies) and might not be fully representative of each sub-group.Third, due to the high number of missing or non-uniformly reported co-variates across studies, it was not possible to reliably carry out a meta-regression to explore sources of heterogeneity in CIED-associated TR and MR or to assess possible correlates of CIED-associated TR and MR.Fourth, as MR is dynamic and may fluctuate prospectively due to various positive or negative remodelling drivers, it hard to ascribe all the credit of post-CRT secondary MR improvement to CRT, especially as we did not compare with the group without CRT.

Conclusions
This meta-analysis revealed the deleterious effect of transvenous RVP via CIED with trans-tricuspid RV leads on the risk of significant CIED-associated TR and MR with no significant difference between ICD and PPM for TR risk and the beneficial effect of CSP for MR and maybe TR, as well as CRT for MR, while LP was associated with unchanged risk.It also revealed the association of CIED-associated or persistent post-CIED atrioventricular valvular regurgitation with lower survival.Based on these findings, opting for pacing strategies that avoid isolated trans-tricuspid RV leads, when possible, might offer protection against incident or deteriorating TR and MR and might even reduce preexisting significant atrioventricular valvular insufficiency and potentially reduce the risk of all-cause mortality.

Figure 4
Figure 4Risk of significant mitral regurgitation (≥ moderate MR) post-transvenous right ventricular pacing via CIED with trans-tricuspid RV leads compared with pre-CIED implantation (pooled odds ratio and 95% confidence interval).CIED, cardiac implantable electronic device.

Overall 22 AFigure
FigureTricuspid regurgitation post-conduction system pacing.(A) Odds ratio for studies that provided binary data and (B) standardized mean difference (SMD) for studies that provided continuous data.HBP, His bundle pacing; LBBP, left bundle branch pacing.

OverallFigure 6 Figure 7
Figure 6Mitral regurgitation post-conduction system pacing.(A) Odds ratio for studies that provided binary data and (B) standardized mean difference (SMD) for studies that provided continuous data.HBP, His bundle pacing; LBBP, left bundle branch pacing.

Figure 9
Figure 9Pooled hazard ratio (95% confidence interval) of all-cause mortality associated with presence vs. absence of persistent significant tricuspid regurgitation post-all cardiac implantable electronic devices (CIEDs) (A) and secondary mitral regurgitation post-cardiac resynchronization therapy (CRT) (B).