Arrhythmic risk profile in mitral valve prolapse: A systematic review and metanalysis of 1715 patients

Mitral valve prolapse (MVP) is a common clinical condition in the general population. A subgroup of patients with MVP may experience ventricular arrhythmias and sudden cardiac death (“arrhythmic mitral valve prolapse” [AMVP]) but how to stratify arrhythmic risk is still unclear. Our meta‐analysis aims to identify predictive factors for arrhythmic risk in patients with MVP.


| INTRODUCTION
Devereux et al. 1 provided a definition of mitral valve prolapse (MVP)   as the displacement of one or both mitral leaflets >2 mm into the atrium during systole. 2 MVP is relatively common, with a prevalence of 2%-3% in the general population, and is generally considered a benign condition. 3,4However, the outcomes of patients with MVP are highly heterogeneous and depend on associated conditions and consequences of the prolapse itself. 4A subgroup of MVP patients may experience ventricular arrhythmias (VA), primarily originating from the papillary muscles, and sudden cardiac death (SCD), presenting with what is known as "malignant MVP" or "arrhythmic mitral valve prolapse" (AMVP). 3,5,6Recently, the term AMVP has been proposed to define the coexistence of MVP and VA, which can be complex (sustained or nonsustained ventricular tachycardia, ventricular fibrillation, out-of-hospital cardiac arrest) or frequent. 7][10][11] One critical factor contributing to this may be the limited number of studies with large populations, which hinders the accurate identification of risk features and a deeper understanding of the pathophysiological mechanisms underlying arrhythmogenesis in these patients.Consequently, AMVP remains a poorly understood entity, and little is known about its pathophysiology and the factors that can predict arrhythmic events.Therefore, we conducted a meta-analysis to assess the role of clinical history, ECG, echocardiographic, and CMR parameters in stratifying the arrhythmic risk in MVP patients.

| Data sources and searches
We systematically searched Medline, Cochrane, Journals@Ovid, Scopus electronic databases for studies published from inception to December 28, 2022 and comparing AMVP and nonarrhythmic mitral valve prolapse (NAMVP).Two investigators (L.P. and G. V.) independently performed searches including the following terms: ventricular arrhythmias, mitral valve prolapse and sudden cardiac death.Detailed information of our literature search strategy is available in the section Expanded Methods in Supporting Infomation.
This review was registered with the PROSPERO register of systematic reviews (ID: CRD42023395984).

| Study selection and data extraction
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement for reporting systematic reviews and metaanalyses was used in this study.
All studies had to fulfill the following criteria to be included in the analysis: (1) have performed a direct comparison between AMVP and NAMVP, (2) included more than 30 patients, (3) included patients over 18 years, (4) reported one or more features analyzed.
Arrhythmic MVP was defined when VAs at least Lown Grade II were present.Editorials, case series, case reports, reviews, expert opinion, and non-English studies were excluded.Two investigators (L.P. and G. V.) extracted data from each study using a standardized protocol and reporting forms and independently assessed the quality items.
Disagreements were resolved by consensus.The quality of individual studies was assessed using the Newcastle-Ottawa Quality Assessment Scale for cohort studies.

| Statistical analysis
Descriptive statistics are presented as means and standard deviations (SD) for continuous variables or number of cases (n) and percentages (%) for dichotomous and categorical variables.
The Mantel-Haenszel odds ratio (OR) model and mean difference (MD) were used to summarize the data between AMVP and NAMVP.Summary estimates and 95% confidence intervals (CIs) were reported for continuous variables as standardized MD.Freeman-Tukey double arcsine transformation was used to establish the variance of raw proportions.We used the Hartung-Knapp-Sidik-Jonkman method with the random effect model to combine the transformed proportions.The heterogeneity across studies was evaluated by using the χ 2 , τ 2 , and Higgins-I 2 statistics; random effects models of DerSimonian were used due to clinical heterogeneity across the patients included in our study.
Publication bias was assessed using the funnel plot and Egger's test.Statistical analysis was performed using Review Manager

| Ethical approval
Ethical approval was not required because this study retrieved and synthesized data from already published studies.
T A B L E 1 Study baseline characteristics of patients included in the analysis.

| DISCUSSION
Previous studies have assessed risk factors for arrhythmias in MVP.
However, to the best of our knowledge, our study is the first to weigh each risk factor for arrhythmogenicity in MVP.According to our results, not all features indicative of arrhythmogenicity are equally significant in defining the risk of developing VA.Therefore, each factor should be carefully considered to ensure a balanced assessment of arrhythmic risk and guide the selection of appropriate therapeutic strategies.
We demonstrated that the presence of LGE by CMR, TWI, MAD, and bileaflet prolapse are associated with an increased likelihood of developing VA in MVP patients.[22][23] LGE is suggestive of fibrosis, which may result from myocardial stretch due to prolapsing leaflets and abnormal extracellular matrix deposition.
Previous studies by Miller et al. 24  Although with slightly lower risks, other factors associated with arrhythmic risk in MVP include MAD, bileaflet prolapse, a higher amount of LGE, longer mitral leaflets, thicker AML, and longer MAD.
Importantly, all these factors are also indicative of phenotype severity in Barlow disease. 28These findings suggest a relationship between the severity of "myxomatous" degeneration (including leaflet degeneration, MAD, and fibrosis) and arrhythmogenesis in AMVP, as previously proposed in other studies. 29Moreover, Chivulescu et al. 27 associated MAD length with VA, while levels of circulating transforming growth factor-beta (TGF-β) were found to correlate with circumferential MAD and fibrosis in another study. 20It is well known how TGF-β promotes myxomatous degeneration of the mitral valve through its interaction with Filamins. 22,23This aspect becomes more relevant in light of the recently emerged association between AMVP and mutated filamin C (FNLC), a cardiac-expressing filamin that is mutated in cases of arrhythmogenic dilated cardiomyopathy phenotype, suggesting a possible underlying mutated genotype in these patients. 30,31rthermore, the traction exerted on papillary muscles by longer and thicker leaflets could be stronger and more likely to trigger ventricular extrasystoles, especially in the presence of a favorable substrate created by inflammation and fibrosis. 24 the other hand, left ventricular ejection fraction, corrected QT interval, female sex, history of AF, mitral annulus diameter, and MR severity did not differ significantly between AMVP and NAMVP patients.The high heterogeneity observed in QTc measurements may have affected the results, as there are conflicting findings in the literature regarding the role of QTc length as a stratification factor for arrhythmic risk in MVP patients. 7Interestingly, female sex, previously believed to be a risk factor for AMVP, was equally distributed between AMVP and NAMVP groups. 22The retrospective analysis of MVP patients who experienced SCD by Han et al. 32 and Delling et al. 33 also support our findings, as they showed no higher arrhythmic risk associated with female sex.Changes in mitral annulus diameter, significant MR, AF, and reduced LVEF usually manifest at a later age when the arrhythmic pattern is already established or as a consequence of other comorbidities and diseases.Rather than being markers of arrhythmogenicity, these factors should be considered as results of the progressive degeneration of the mitral valve apparatus due to MVP itself and subsequent left atrial dilation.
In conclusion, our study identifies five features associated with AMVP: LGE by CMR, TWI, MAD, and bileaflet prolapse.Each of these features should be weighed according to its relative risk to make a balanced assessment of arrhythmic risk in these patients (Central

| CONCLUSION
Our meta-analysis demonstrated that the presence of LGE by CMR, TWI, MAD, and bileaflet prolapse, along with longer mitral leaflets, thicker AML, and longer MAD, are associated with VAs in MVP.
However, it was observed that these features do not equally represent the risk of arrhythmias.Therefore, each of these factors should be carefully considered to ensure a balanced evaluation of the patient and, consequently, the selection of appropriate follow-up strategy and therapy.

| Limitations
This study has certain limitations.First, being a meta-analysis, it relies on available data from the literature, which limits the investigation of certain interesting features (such as the PickelHaube sign or extracellular volume by T1 mapping) that have been described in the literature but in different contexts unrelated to the comparison between AMVP and NAMVP patients.Furthermore, our metaanalysis is limited by the observational nature of the studies.Finally, some large cohort studies reporting on the risk of arrhythmia based on certain predictors were not included since they did not report specifically as a comparison between arrhythmic and nonarrhythmic patients.

3. 4 |
Figure SB.The funnel plot for visual inspection of the bias showed no bias (Figure SC), which was confirmed by Egger's test (p = .53).
using hybrid imaging with CMR and positron emission tomography have shown that fibrosis in AMVP is preceded by inflammation.Based on this, it could be hypothesized that chronic stretch stress on papillary muscles leads to inflammation, fibroblast proliferation, and subsequent fibrosis deposition, contributing to the development of a proarrhythmic substrate.This mechanism may also explain our finding of increased mechanical dispersion in AMVP patients compared to NAMVP patients, which is consistent with the results reported by Vairo et al.25 A recent study by Scheirlynck et al.20 provided evidence of the relationship between myocardial stretch and VA in AMVP patients.They demonstrated that MVP patients with arrhythmias have higher levels of soluble suppression of tumorigenicity 2, a marker released from stretched myocardium, compared to MVP patients without arrhythmias.20TWI in inferior and lateral leads may be an electrical manifestation of myocardial distress secondary to stretch stress, rather than the expression of a structural abnormality.This hypothesis is supported by findings from Alqarawi et al.,26 who reported the normalization of TWI after mitral valve repair.Chivulescu et al.27 also found an increased risk of complex VA in MVP patients with a more extensive TWI, showing the correlation between TWI, the percentage of extracellular volume by T1 mapping at CMR, and VAs.28 Consistent with these findings, our results indicate that LGE and TWI are the most indicative features of an increased arrhythmic risk (OR: 16.67 and 2.63, respectively) (Central Figure).F I G U R E 5 Forest plots comparing: female sex distribution between AMVP and NAMVP: AMVP, arrhythmic mitral valve prolapse; CI, confidence interval; NAMVP, nonarrhythmic mitral valve prolapse.CENTRAL FIGURE Arrhythmic risk stratification in mitral valve prolapse patients.Color map: Green: low OR of ventricular arrhythmias; light yellow to orange: medium OR of ventricular arrhythmias; red: high OR of ventricular arrhythmias.The figure was partly generated using Servier Medical Art, provided by Servier, licensed under a Creative Commons Attribution 3.0 unported license.

Figure ) .
Figure).Considering these findings, our study represents a first step towards a multiparametric quantitative risk assessment that can guide risk-targeted follow-up strategies and arrhythmia prevention therapies without overtreatment of lower-risk individuals.