Right atrial structural remodeling predict worse outcomes in transcatheter mitral valve repair

Abstract

Background

In the current study, we assess the predictive role of right and left atrial volume indices (RAVI and LAVI) as well as the ratio of RAVI/LAVI (RLR) on mortality following transcatheter mitral valve repair (TMVr).

Methods

Transthoracic echocardiograms of 158 patients who underwent TMVr at a single academic medical center from 2011 to 2018 were reviewed retrospectively. RAVI and LAVI were calculated using Simpson's method. Patients were stratified based on etiology of mitral regurgitation (MR). Cox proportional‐hazard regression was created utilizing MR type, STS‐score, and RLR to assess the independent association of RLR with survival. Kaplan−Meier analysis was used to analyze the association between RAVI and LAVI with all‐cause mortality. Hemodynamic values from preprocedural right heart catheterization were also compared between RLR groups.

Results

Among 123 patients included (median age 81.3 years; 52.5% female) there were 50 deaths during median follow‐up of 3.0 years. Patients with a high RAVI and low LAVI had significantly higher all‐cause mortality while patients with high LAVI and low RAVI had significantly improved all‐cause mortality compared to other groups (p = 0.0032). RLR was significantly associated with mortality in patients with both functional and degenerative MR (p = 0.0038). Finally, Cox proportion‐hazard modeling demonstrated that an elevated RLR above the median value was an independent predictor of all‐cause mortality [HR = 2.304; 95% CI = 1.26−4.21, p = 0.006] when MR type and STS score were accounted for.

Conclusion

Patients with a high RAVI and low LAVI had significantly increased mortality than other groups following TMVr suggesting RA remodeling may predict worse outcomes following the procedure. Concordantly, RLR was predictive of mortality independent of MR type and preprocedural STS‐score. These indices may provide additional risk stratification in patients undergoing evaluation for TMVr.

1. INTRODUCTION

Mitral regurgitation (MR) is the most common valvular disease in the United States, with increasing prevalence as the population ages. ¹ MR is classified broadly as primary/degenerative (DMR) related to abnormalities of the anatomical structures of the mitral valve while secondary functional (FMR) is based on systolic tethering of the anatomically intact leaflets due to ventricular dysfunction (ischemic or other) or atrial dilatation. ² Transcatheter mitral valve repair (TMVr) has been utilized in DMR patients since the publication of the results of the Endovascular Valve Edge‐to‐Edge Repair Study (EVEREST II) trial, which demonstrated safety and efficacy of TMVr as an alternative to surgical valve repair. ³ , ⁴ Subsequently the results of the Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy for Heart Failure Patients with Functional Mitral Regurgitation (COAPT) trial also demonstrated that heart failure (HF) patients with FMR could benefit from TMVr. This pivotal study demonstrated reduced mortality, symptoms, and hospitalizations due to HF exacerbations. ⁵ As a result of this seminal trial, there has been increasing adoption of TMVr as a viable therapeutic strategy for patients with HF and significant MR. ⁶

There is however paucity of data on the phenotypic characteristics of patients likely to develop adverse outcomes following TMVr, particularly among those with advanced HF. While STS score, EuroScore, and Euroscore II are widely used clinical tools for risk stratification, these scoring systems are based on surgical outcomes data and are known to overestimate 1‐year mortality for TMVr. ⁷ Furthermore, these scores heavily rely on clinical parameters that do not account for the proximate compensatory structural changes resulting from the consequences of MR.

Structural changes in atrial dimension/volumes are often a reflection of pressure and volume changes occurring in the ventricles. It has been previously shown that increased right to left atrial volume indices (RAVI and LAVI) portends worse outcomes in patients with pulmonary hypertension (PH), ⁸  cardiogenic shock, ⁸ and in hospital cardiac arrest. ⁹ LAVI in particular is gaining recognition as an indicator of overall cardiac function, and has been termed “the HbA1c of the heart.” ¹⁰ There is limited evidence regarding the prognostic implications of RAVI and LAVI in patients undergoing TMVr. This study sought to investigate the relationship between RAVI and LAVI with all‐cause mortality among patients undergoing TMVr at a single academic medical center.

2. METHODS

2.1. Study design

We retrospectively identified 158 consecutive patients who underwent TMVr at a single academic center from 2011 to 2019. All patients had moderate to severe MR with both DMR and FMR as an underlying etiology. Baseline patient demographics and clinical data were obtained and analyzed using the electronic medical records. Clinical and demographic characteristics were abstracted from the electronic records including, comorbid conditions, transthoracic echocardiography (TTE), procedural, and longitudinal outcomes. Patients were only included if they had echocardiography in the 3 months preceding or directly following TMVr procedure and had adequate TTE images for calculation of RAVI and LAVI.

The primary outcome was all cause mortality. This study was approved by the institutional review board of the University of Virginia and qualified for a consent waiver.

2.2. Echocardiography

All patients included in the study underwent standard pre and post procedure 2D TTE. Experienced sonographers, using standard echocardiographic views (parasternal, apical, and subcostal), obtained TTE images. Using both the 4‐ and 2‐chamber view, LAVI was measured and calculated with the biplanar Simpson method. RAVI was measured in the 4‐chamber view using single plane disc summation as per current societal guideline recommendations for cardiac chamber measurements. ¹¹ All images were obtained using Phillips IE33, Epiq 7CV, or GE Vivid E9 ultrasound systems. Studies were analyzed and processed using Enterprise imaging (Agfa Healthcare N. V.). Offline echocardiographic analysis was performed by three investigators.

2.3. Statistical analysis

Continuous variables were expressed as mean ± standard deviation. Categorical variables were expressed as frequency and percentages. Analysis of comparisons in continuous variables were performed using t‐tests or Wilcoxon test depending on normality. Analysis of differences in categorical variables between RAVI/LAVI groups were performed using χ ² tests. A Cox proportional‐hazard model was created utilizing MR type, STS score, and RAVI/LAVI ratio (RLR) to assess the association of RLR with survival. Survival analysis and Kaplan−Meier plots were used to show differences between RAVI and LAVI. Log rank test was used to show differences between groups. An α value of <0.1 was used for statistical significance. Statistical analysis was performed using SAS 9.4 (SAS Institute).

3. RESULTS

3.1. Patient cohort and procedural outcome

Out of the 158 consecutive patients, 123 (65.9%) had complete hemodynamic and echocardiographic data with adequate follow‐up and were included in the study. The baseline characteristics of these 123 patients stratified by MR type are shown in Table 1. Mean age was 81.3 years ± 11 months and 52.5% were female. Median STS score was 8.1 with the majority of patients having NYHA class 3 or 4 HF symptoms (67.5%). In this cohort, 65% of the patients had DMR. African Americans represented 11% of cohort and were more likely to have FMR versus DMR (p = 0.02). Patients with DMR were more likely to be older, with an ischemic etiology and were more likely to have a higher baseline left ventricular ejection fraction. There were no deaths in the perioperative period. The STS scores were similar between MR groups stratified based on etiologies. With respect to 30‐day follow‐up after the procedure, the MR severity grade was 1.34 ± 0.75, the EROA was 0.44 ± 0.23, and the regurgitant volume was 63.5 ± 27.15.

Table 1.

Patient characteristics as separated by etiology of MR

	Functional MR (N = 43)	Degenerative MR (N = 80)
Age	75.7 (11.8)	84.9 (9.08)	<0.001
Female sex	18 (41.9%)	42 (52.5%)	0.349
BMI	25.7 (5.48)	24.5 (4.84)	0.208
African American race	8 (18.6%)	4 (5.0%)	0.042
NYHA			0.797
1	0 (0%)	1 (1.2%)
2	11 (25.6%)	17 (21.2%)
3	28 (65.1%)	52 (65.0%)
4	4 (9.3%)	10 (12.5%)
STS score	6.83 (4.36)	8.60 (4.68)	0.053
Hemoglobin	12.4 (1.86)	11.9 (1.65)	0.134
Creatinine	1.45 (1.20)	1.25 (0.492)	0.292
GFR	57.3 (23.8)	57.1 (20.7)	0.959
CKD	7 (16.3%)	9 (11.2%)	0.610
CHF	34 (79.1%)	71 (88.8%)	0.238
CAD	29 (67.4%)	38 (47.5%)	0.054
MI	21 (48.8%)	11 (13.8%)	<0.001
CABG	14 (32.6%)	24 (30.0%)	0.930
PCI	16 (37.2%)	12 (15.0%)	0.010
CVA	4 (9.3%)	7 (8.8%)	1.000
Atrial fibrillation	26 (60.5%)	49 (61.2%)	1.000
COPD	8 (18.6%)	15 (18.8%)	1.000
HTN	31 (72.1%)	65 (81.2%)	0.346
HLD	23 (53.5%)	41 (51.2%)	0.962
DM	8 (18.6%)	11 (13.8%)	0.654
ICD	15 (34.9%)	7 (8.8%)	<0.001
CRT	6 (14.0%)	1 (1.2%)	0.013
LVEF	37.4 (14.5)	53.8 (9.70)	<0.001
MR severity
2	0 (0%)	3 (3.8%)
3	15 (34.9%)	24 (30.0%)
4	28 (65.1%)	53 (66.2%)
LAVI (biplanar)	62.2 (25.4)	59.7 (29.8)	0.622
RAVI	40.7 (19.8)	38.8 (26.4)	0.657
RAVI/LAVI (RLR)	0.747 (0.463)	0.694 (0.406)	0.532
mPAP	33.8 (11.3)	31.6 (8.42)	0.280
CO (thermodilution)	4.30 (1.60)	4.59 (1.60)	0.371

Note: Categorical variables presented as N (%). Continuous variables expressed as N (standard deviation).

Abbreviations: CABG, coronary artery bypass graft; CHF, congestive heart failure (systolic or diastolic); CKD, chronic kidney disease; CO (thermodilution), cardiac output as measured by thermodilution on right heart catheterization; CRT, cardiac resynchronization therapy; CVA, previous stroke; DM, previous diagnosis of diabetes mellitus (either type); HLD, previous diagnosis of hyperlipidemia; HTN, documented diagnosis of hypertension; ICD, intracardiac defibrillator; LAVI (biplanar), left atrial volume index by Simpson's biplane method; LVEF, left ventricular ejection fraction on last confirmed TTE before TMVr; MI, previous myocardial infarction; mPAP, mean pulmonary artery pressure; MR, mitral regurgitation; PCI, previous percutaneous intervention; RAVI, right atrial volume index; RLR, ratio of right atrial volume index to left atrial volume index.

3.2. Association of RAVI LAVI with mortality

Over a median follow‐up period of 3 years, there were a total of 50 deaths in both MR groups. Four groups were generated based on whether patients had RAVI above the median (36.8 ml/m²) and LAVI above the median (55.4 ml/m²). Representative TTE images for these for groups are presented in Figure 1. Survival analysis using Kaplan−Meier plots are displayed in Figure 2. There was a statistically significant increase in all‐cause mortality among patients with a RAVI greater than the median and LAVI less than the median value (“high RAVI/low LAVI” group) compared with those having RAVI less than the median and LAVI greater than the median value (“low RAVI/high LAVI group”). This latter group had the best outcomes. Those patients where both the RAVI and LAVI were concordantly above or below the median values had intermediate survival.

Figure 1.

Representative transthoracic echocardiogram images for RAVI/LAVI groups. All images are presented in apical four chamber view. Measurements were made by Simpson method of disc summation using AGFA Enterprise Imaging software. LAVI, left atrial volume indices; RAVI, right atrial volume indices.

Figure 2.

Kaplan−Meier plot for TMVr patients by RAVI and LAVI above and below the median values. Red represents both RAVI and LAVI below median value; green represents RAVI less than median value, LAVI greater than median value; blue represents RAVI greater than median value, LAVI less than median value; purple represents both RAVI and LAVI above median values. Time on X‐axis is in years. Survival probability on Y‐axis calculated from all‐cause mortality (p = 0.0032). LAVI, left atrial volume indices; RAVI, right atrial volume indices; TMVr, transcatheter mitral valve repair. [Color figure can be viewed at wileyonlinelibrary.com]

3.3. Comparison of hemodynamics between RAVI LAVI groups

Preprocedural hemodynamics were compared between the high RAVI/low LAVI group and the low RAVI/high LAVI group using individual t‐tests for each variable assessed. This comparison demonstrates that high RAVI/low LAVI group had significantly higher mean pulmonary artery pressure (mPAP) compared with the low RAVI high LAVI group (p = 0.04) (Table 2), as well as trend for higher calculated pulmonary vascular resistance (p = 0.08). The high RAVI/low LAVI group had also a trend for higher pulmonary capillary wedge pressures (26.72 ± 8.43) compared with the low RAVI/high LAVI group (22.28 ± 2.29) (p = 0.10).

Table 2.

Hemodynamic and echocardiographic findings in RAVI‐LAVI subgroups

Preprocedure values	Entire cohort mean/SD	Low RAVI/high LAVI mean/SD	High RAVI/low LAVI mean/SD	p Value
RVID	5.06 ± 0.99	5.35 ± 1.36	4.87 ± 0.88	0.33
LVID	3.71 ± 1.01	4.08 ± 0.68	3.73 ± 0.72	0.49
mRAP	11.46 ± 1.01	13.33 ± 6.66	18.33 ± 2.51	0.32
PASP	48.00 ± 15.11	47.33 ± 13.38	55.59 ± 15.80	0.07
PADP	20.80 ± 6.93	20.48 ± 7.16	24.27 ± 6.49	0.08
mPAP	31.90 ± 9.48	31.29 ± 8.28	36.91 ± 8.94	0.04
PVR	6.75 ± 3.68	6.01 ± 3.10	7.90 ± 3.45	0.08
PAWP	21.76 ± 6.12	22.29 ± 2.29	26.73 ± 8.44	0.12
TDCI	2.54 ± 0.85	2.69 ± 0.64	2.52 ± 1.03	0.66

Abbrevitiations: LAVI, left atrial volume indices; LVID, left ventricle internal diameter; mPAP, mean pulmonary artery pressure; mRAP, mean right atrial pressure; PADP, pulmonary artery diastolic pressure; PASP, pulmonary artery systolic pressure; PCWP, pulmonary capillary wedge pressure; PVR, pulmonary vascular resistance; RAVI, right atrial volume indices; RVID, right ventricle internal diameter; TDCI, thermodilution cardiac index.

3.4. RAVI‐LAVI ratio by MR type

Additional analyses were performed based on stratification of patients by the median RLR value of 0.6. MR severity was not significantly different between groups. Kaplan−Meier survival plots are displayed in Figure 3. The results demonstrated that the association between high RLR and mortality held for both FMR and DMR (overall p = 0.0038 for difference among the four groups).

Figure 3.

Kaplan−Meier plot for TMVr patients separated by RLR above or below median value of 0.6 and MR etiology. Red represents FMR patients with RLR less than 0.6. Purple represents FMR patients with RLR greater than 0.6. Yellow represents DMR patients with RLR less than 0.6. Blue represents DMR patients with RLR greater than 0.6. Time on X‐axis in years. Survival probability on Y‐axis calculated from all‐cause mortality (p = 0.0038). DMR, degenerative mitral regurgitation; FMR, functional mitral regurgitation; LAVI, left atrial volume indices; MR, mitral regurgitation; RAVI, right atrial volume indices; RLR, RAVI/LAVI ratio; TMVr, transcatheter mitral valve repair. [Color figure can be viewed at wileyonlinelibrary.com]

3.5. Cox proportional‐hazard regression model

A multivariate Cox proportional‐hazard regression model was constructed to assess the independent association between RLR and mortality accounting for both MR type and STS score. The model accounted for MR type as a binary variable and STS as a continuous variable. The model demonstrated that an elevated RLR above the median value of 0.6 was an independent predictor of mortality [HR = 2.304; 95% CI: 1.26−4.21, p = 0.006] even after adjusting for MR type and the STS score.

4. DISCUSSION

In this single center retrospective analysis of patients undergoing TMVr with over 3 years of follow‐up, we found that an elevated right atrial volume to left atrial volume ratio, was associated with worse outcomes. First, patients with high RAVI and low LAVI compared to the median values had significantly worse outcomes than all other groups, including those patients where both RAVI and LAVI were increased beyond median values. Second, patients with low RAVI and high LAVI had significantly better outcomes compared to all other groups. Analysis of preprocedural hemodynamics between groups showed that patients in the high RAVI low LAVI group had significantly higher mPAP and wedge pressure. These findings support a likely association of an elevated RLR with PH and right HF/right ventricular dysfunction that may influence the increased risk of mortality following the procedure. When stratified between patients with DMR and FMR, RLR was still significantly associated with mortality in both groups. Finally multivariate analysis utilizing Cox proportional‐regression modeling to account for MR type and preprocedural STS score demonstrated that RLR was an independent predictor of mortality with a greater than twofold increase in hazard ratio when RLR was greater than the median.

The patient population reviewed in this single center retrospective study had 65% DMR and 35% FMR, but with increasing numbers of FMR procedures occurring later in the study period, in part related to the publication of the findings of the COAPT trial. DMR patients were significantly older than patients with FMR and FMR patients had significantly lower EF as expected compared to DMR patients. Interestingly, African Americans patients who underwent TMVr procedure were significantly more likely to have FMR as opposed to DMR, though the sample size is small. There was no significant difference in mortality between FMR and DMR in the post procedure period. However, as anticipated, FMR patients had a greater mortality after 3 years likely reflecting the natural progression of HF.

MR is a condition that characteristically leads to left atrial dilatation and remodeling. The degree of LA remodeling may be related to disease severity and prognosis. ¹² Further, LA remodeling may directly or indirectly impose structural changes in the right atrium. ¹³ Increased right atrial pressure and volume leads to RA structural remodeling that manifests as chamber enlargement. Furthermore, RA enlargement may often occur before RV enlargement ¹⁴ due to the consequences of increased volume. Previous studies have demonstrated that changes in cardiac chamber volume may precede pressure changes in both the ventricles and atria. ¹³ Consequently, it is possible that volume changes may be more sensitive than pressure changes in detecting early right HF. ¹⁵ Given these considerations, RAVI and LAVI integrate these closely linked pathophysiological processes that underlie MR, thus cross‐linking the systemic circulatory perturbations with the compensatory reactions of the pulmonary circulation.

As one would expect based on the pathophysiological mechanisms of MR and atrial remodeling, TMVr more consistently produces beneficial remodeling of the LA and LV, where right sided remodeling is observed less consistently. Toprak et al. ¹⁶ demonstrated that at 12 months post‐TMVr there were significant reductions in both minimal and maximum LA volume. These structural changes correlate with functional measurements such as LA Strain and LA operating chamber stiffness. ¹⁷ While little research exists regarding RA volume changes following TMVr, inferences can be made from the impact of TMVr on RV function based on the aforementioned cross talk between the systemic and pulmonary circulations. Ledwoch et al. ¹⁸ recently followed changes in RV function at 3 and 6 months following TMVr and found that RV function declined in 20%, was unchanged in 25% and improved in 55% of the patients. Not surprisingly the patients with the greatest improvement in RV function had increased survival time. The study noted that improvement of RV function was an independent predictor of mortality even after controlling for other variables.

From a hemodynamic standpoint, atrial volume changes may be more sensitive than pressure changes in detecting early right HF. ¹⁹ Right ventricular function has been associated with worse outcomes in many cardiovascular conditions. ²⁰ , ²¹ More recently, pulmonary artery pulsatility index, a hemodynamic marker of RV function, has been shown to predict patient survival following Mitraclip. ²² It is tempting to surmise that TMVr produces a greater and more immediate impact on LA remodeling when regurgitant flow is mitigated. Given these considerations, it is possible that patients that potentially benefit the most from TMVr are those who have not had RA remodeling resulting from the structural constraints of an enlarged LA. Conversely patients where the RA has already become overloaded and remodeled, may signal the presence of a compromised RV function from maladaptive changes, either directly from MR or concomitant HF. In these patients, it is likely that TMVr is less efficacious.

Another interesting finding from our study was the paradoxical relationship between elevated LAVI and improved outcomes following TMVr. While virtually all patients included in the study had at least moderate‐severe left atrial enlargement, the low RAVI high LAVI group had significantly better outcomes than the high RAVI low LAVI group. One potential explanation for this is that left atrial dilation and remodeling may be a protective response to reduce left atrial pressures and by extension mPAP and right sided pressures. Supporting this view, patients with low RAVI high LAVI had lower PAWP (p = 0.1) and mPAP (p = 0.05) than those with high RAVI low LAVI.

Additionally, RAVI and RV function are closely linked with PH. The pulmonary vasculature acts as the bridge between left and right‐sided hemodynamics and clearly impacts both cardiac physiology and remodeling. As changes in atrial volume are indicative of pressure and volume changes within the heart, RAVI may parallel PH severity. PH has been associated with adverse outcomes in many cardiovascular conditions. ²³ On the other hand, RLR is also associated with adverse outcomes in patients with PH, ²⁴ cardiogenic shock, ⁸ and in hospital cardiac arrest. ⁹ Studies have demonstrated the independent association of PH with adverse outcomes following TMVr. ²⁵ As such the potential mechanism underling our findings of increased mortality following TMVr with elevated RLR could be related to the influence of PH on outcomes. To this point when analyzing differences in preprocedural hemodynamics from right heart catheterization between the high RAVI low LAVI group and the low RAVI high LAVI group, there was a significant increase in mPAP (p < 0.05) and PVR (p = 0.08) in the high RAVI low LAVI group.

A limitation of the current study is the utilization of all‐cause mortality rather than cardiac‐specific mortality following the procedure. As evidenced by Table 1, the group undergoing TMVr at our institution tended to be an older population with several comorbidities that likely impact mortality. It was attempted to factor these comorbidities in the analysis by utilizing preprocedural STS score in the Cox‐regression model. Another potential limitation of the study was the utilization of single place disc summation to calculate RAVI. While there has been no direct comparison between the single plane disc summation and more sensitive measurement modalities such as cardiac MRI, single plane disc summation using the 4‐chamber apical view has been previously validated in that it correlates with right atrial pressure and outcomes in chronic systolic HF. ²⁶ Given the significance of even minor changes in RAVI in predicting outcomes in the current study, further research utilizing more specialized modalities for chamber measurements such as cardiac CT or MRI would offer valuable contribution to our hypothesis. This was also a single academic center retrospective analysis. However, the use of objective imaging indices as well as a large cohort of patients with TMVr is a major strength of the study.

RLR is an easily accessible measurement on 2D TTE, which is standard of care before TMVr, and has been shown to be reliable with minimal interobserver variability. ⁹ , ²⁴ , ²⁶ The results of the current study suggest that preprocedural RAVI/LAVI could play a valuable role in patient selection and risk stratification for TMVr and provide added insight beyond risk stratification scores such as STS and EuroScore.

5. CONCLUSIONS

In summary, patients with an elevated LAVI and with decreased RAVI had significantly better outcomes than other groups while patients with elevated RAVI and decreased LAVI had significantly poorer outcomes. Concordantly, RLR was predictive of mortality independent of MR type and preprocedural STS‐score.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Miller MS, Cutts J, Donatelle M, et al. Right atrial structural remodeling predict worse outcomes in transcatheter mitral valve repair. Catheter Cardiovasc Interv. 2023;101:217‐224. 10.1002/ccd.30471

DATA AVAILABILITY STATEMENT

Data are available on request from the authors.