Prognostic impact of transcatheter mitral valve repair in patients with exercise-induced secondary mitral regurgitation

Abstract

Aims

Although exercise-induced secondary mitral regurgitation (MR) is known to have a poor prognosis, the therapeutic strategy towards this condition remains to be investigated. In the present study, we aimed to investigate the prognostic impact of transcatheter mitral valve repair (TMVr) using the MitraClip in patients with exercise-induced secondary MR.

Methods and results

Of the 200 consecutive patients with secondary MR who underwent exercise stress echocardiography, 46 (23%) that presented with exercise-induced secondary MR [i.e. increase in effective regurgitant orifice area (EROA) of ≥ 0.13 cm²] were enrolled in the present investigation. The composite endpoints of all-cause mortality and hospitalization for heart failure were evaluated. Of the 46 patients included in the current cohort, 19 (41%) underwent TMVr and 27 (59%) were medically managed (control group). Although the TMVr group tended to present with a greater EROA at rest (0.26 ± 0.10 vs. 0.20 ± 0.08 cm², P = 0.047), there were no differences in the EROA changes during exercise between the two groups (0.18 ± 0.10 vs. 0.18 ± 0.04 cm², P = 0.940). While the TMVr group reported a higher event-free survival rate after the 13-month follow-up period (log-rank P = 0.017), the Cox proportional-hazard analysis suggested the TMVr to be associated with clinical outcomes (hazard ratio: 0.419, P = 0.044).

Conclusion

As opposed to the medical management, TMVr treatment was associated with a lower risk of composite endpoints in patients with exercise-induced secondary MR. Exercise stress echocardiography is considered to have played an important role in decision-making for secondary MR.

Introduction

Secondary mitral regurgitation (MR) is a dynamic pathology occurring in the majority of the patients presenting heart failure with reduced left ventricular ejection fraction (LVEF).^1–8 Exercise can be considered as a leading cause of the pathology. This condition is strongly associated with decreased quality of life, increased rate of hospitalization for heart failure, and shortened survival.^7,9–11 Medical therapy, revascularization, cardiac resynchronization therapy, and surgery have been proposed as treatments for exercise-induced secondary MR.^12–14 However, it is unclear whether a reduction of dynamic secondary MR has favourable clinical outcomes. Transcatheter interventions have also been reported to improve the haemodynamics in patients with secondary MR, but its effect on prognosis is unknown.¹⁵ A recently published randomized trial suggested the performance of a transcatheter mitral valve repair (TMVr) using the MitraClip to both reduce the rate of hospitalization and improve survival in patients with heart failure and secondary MR compared the maximally tolerated guideline-directed medical therapy alone.¹⁶ Therefore, the purpose of the current study was to investigate the prognostic impact of conducting TMVr using the MitraClip in patients with exercise-induced secondary MR.

Methods

Study population

This study reviewed 200 consecutive patients with secondary MR who underwent exercise stress echocardiography (ESE) using the semi-supine bicycle ergometer between January 2009 and December 2019 at the St. Marianna University School of Medicine Hospital. Exercise-induced secondary MR was defined as an increase in the effective regurgitant orifice area (EROA) by ≥0.13 cm² based on previous studies.^9,17 Of the selected patients, 154 were not recruited in the present study due to either an absence of exercise-induced MR (n = 152) or history of open surgery (n = 2, Figure 1). As a result, a total of 46 patients with exercise-induced secondary MR were enrolled. This study was approved by the institutional review board at the St. Marianna University School of Medicine (No. 1288).

Exercise echocardiography

After conducting the Doppler echocardiography at rest, patients performed a symptom-limited bicycle exercise test in the semi-supine position on a dedicated tilting exercise table. Briefly, they started with an initial workload of 10 W, which was maintained for 3 min and then increased workload of 10 W every 3 min. Blood pressure and heart rate were recorded every 1 min, whereas both a 2D and a Doppler echocardiography data were obtained throughout the exercise test. Because the study included patients with atrial fibrillation, exercise testing was performed with continuous oral administration of drugs that regulate the heart rate, such as beta-blockers.¹⁸

Echocardiographic measurements

Echocardiographic examinations were performed using the commercially available ultrasound system (Vivid E9; GE Vingmed Milwaukee, WI, USA). All the echocardiographic and Doppler data were obtained both at rest and during exercise in a digital format and stored on a workstation for offline analysis (EchoPAC, version 12; GE Vingmed Milwaukee, WI, USA). All the measurements were averaged over three cardiac cycles. In case of atrial fibrillation, we acquired three consecutive beats with a caution to select them if occurring after two serial beats with average RR intervals.¹⁹ MR was quantified according to the proximal isovelocity surface area (PISA) method at rest and during exercise, the EROA, and the regurgitant volume (RV), which indicates the severity of secondary MR.²⁰ In particular, the PISA radius was measured in mid-systole from 3 (more) frames using optimal flow convergence, while the most appropriate negative aliasing velocity to obtain PISA was chosen offline on the workstation. The PISA radius was measured in mid-systole. RV and EROA were calculated with standard formulas. Briefly, the diameter of the left ventricular (LV) outflow tract (inner edge) was measured only at rest in the parasternal long-axis view, whereas the mitral and aortic stroke volumes (SVs) were calculated as the pulsed-wave Doppler time–velocity integral × the area of the annulus of each valve, respectively. Successively, while the RV was calculated as the difference between the mitral and aortic SVs, the EROA was measured as the RV/VTI of the MR jet.^21,22 In addition, the biplane Simpson method was used to measure the following volumes: (i) LV end-diastolic (LVEDV), (ii) LV end-systolic (LVESV), stroke, EF, and left atrial (LA) maximal volumes. Moreover, the mitral E- and A-wave velocities were measured by pulsed wave Doppler, with the E-wave velocity was obtained through tissue Doppler imaging in the septal position of the mitral annulus. These measurements were repeatedly assessed during exercise prior to the eventual fusion of the E- and A-waves. In contrast, the systolic pulmonary artery pressure (SPAP) was derived from the jet of the tricuspid regurgitation using the systolic transtricuspid pressure gradient, which was in turn, calculated through both the modified Bernoulli equation (P4v², where v is maximal tricuspid regurgitant jet velocity in m/s) and the addition of the right atrial pressure. Finally, the mean right atrial pressure was estimated according to the most recent American Society of Echocardiography recommendation whereas the right atrial pressure was assumed to be constant from rest to exercise.²³ We defined exercise-induced pulmonary hypertension as SPAP ≥60 mmHg during exercise.^24,25 Tricuspid annular systolic excursion (TAPSE) was measured as the lateral tricuspid annular excursion from end-diastole to end-systole using M-mode in the apical four-chamber view.²⁶ SV was calculated as the pulsed-wave Doppler time–velocity integral × the area of the annulus of the aortic valve. SPAP/cardiac output (CO) ratio was calculated to determine the pulmonary vascular reserve. Right ventricular–pulmonary arterial coupling was assessed by the TAPSE/SPAP ratio.²⁷

Study design and clinical outcomes

In the present study, a medical vs. a TMVr (using the MitraClip) management of exercise-induced secondary MR was assessed. The decision to perform the TMVr using the MitraClip was determined by the patients’ care providers on the basis of the patients’ clinical and echocardiographic findings. All the patients who were considered by cardiologists to require TMVr underwent the procedure within 3 months of their ESE. All the implantation procedures were performed with proctoring from Abbott Vascular. Successively, the technical success with respect to device implantation was defined according to the consensus document from the Mitral Valve Academic Research Consortium.²⁸ Either the medical or the TMVr thereafter as deemed appropriate by each physician in routine clinical practice. The primary outcome was defined according to both all-cause death and hospitalization for heart failure. Specifically, the New York Heart Association (NYHA) functional class was assessed at the time of the ESE in all patients and in those who completed their 12-month follow-up period.

Statistical analysis

The results are expressed as either means ± standard deviations (SDs) or percentages, unless otherwise specified. Data for both the TMVr and control groups were compared using the Student’s t-test, Mann–Whitney U-test, χ² test, or Fisher exact test, as appropriate. In addition, the probability of event-free survival among the TMVr and control groups was measured using the Kaplan–Meier analysis and compared using the two-sided log-rank test. In contrast, the effect of the clinical and echocardiographic parameters was assessed using the Cox proportional-hazard model in both univariate and multivariate analyses. Specifically, the Society of Thoracic Surgeons (STS) score, an overall clinical risk score, and variables with a uni-variate value of P < 0.05 were incorporated into the multivariate models. Finally, given their collinearity, the LVEF that were recorded during exercise and the TMVr were analysed separately from the multivariate analysis model. All statistical analyses were performed with SPSS 22.0 software (SPSS, Inc., Chicago, IL, USA).

Results

Clinical and echocardiographic characteristics

The baseline characteristics of the patients included in the present study are shown in Table 1. Briefly, compared with the control group, the TMVr group was characterized by older age and higher prevalence of the NYHA functional class II (or greater). While the cause of cardiomyopathy was ischaemic in 32.6% of the patients, differences in the prevalence of ischaemic cardiomyopathy and medications between the two groups were not observed. However, the TMVr group had higher STS scores and increased prevalence of high surgical risk (STS score > 8%) than the control group. A summary of the resting and exercise echocardiographic findings is shown in Table 2. Specifically, both the MR severity and SPAP were significantly increased from rest to exercise in the entire cohort (EROA: 0.22 ± 0.09 vs. 0.40 ± 0.11 cm², RV: 37.0 ± 15.2 vs. 52.2 ± 15.1 mL, SPAP: 33.0 ± 12.6 vs. 55.6 ± 14.8 mmHg, all P < 0.001). Although the TMVr group presented smaller LVEDVi and LVESVi as well as higher LVEF, they had a higher MR severity both at rest and during exercise than the control group. However, the EROA changes from rest to exercise were similar between the two groups. Figure 2 presents the relationship between the LV size and the MR severity at rest and during exercise. Exercise increases LVEDV and EROA in this cohort. Compared with the characteristics of the two large previous studies,^16,29 the relationship between LVEDV and EROA during exercise in this study was similar to that of the Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy for Heart Failure Patients with Functional Mitral Regurgitatio (COAPT) study (Figure 2). There were no significant differences in SV or CO between the two groups at rest and during exercise. However, the TMVr group had higher SPAP during exercise and more frequent exercise-induced pulmonary hypertension than the control group. Furthermore, the TMVr had lower TAPSE and TAPSE/SPAP during exercise compared with the control group.

Procedural and clinical outcomes

None of the patients included in the TMVr group experienced any clinical events while waiting for the procedure and they all achieved a technically successful device implantation. Specifically, 10 subjects (52.6%) had one clip implanted, whereas two clips were provided to all the remaining patients. Furthermore, although 1 (5.3%) intra-procedural vascular complication was observed, neither a 30-day mortality nor in-hospital deaths were seen in the TMVr group. When resting echocardiography was performed in all patients of the TMVr group at the time of discharge, MR grades of 1+ (or lower) and 2+ were found in 17 (89.4%) and 2 (10.6%) patients, respectively. The lack of patients with an MR grade of 3 (or greater) was seen in the present study. A representative case of patients in the TMVr group is shown in Figure 3. During the 13-month follow-up period, 1 (2.2%) patient death and 10 (21.7%) hospitalizations for heart failure were observed. In the entire cohort, event-free survival rates of 82 ± 4% and 56 ± 7% were observed at 12 and 24 months, respectively, with higher rates seen in the TMVr group (log-rank P = 0.017, Figure 4). Although we found no differences in the NYHA functional class between the two groups at baseline, the NYHA II/III functional classes were more frequently experienced at 12 months by the patients in the control group ( Figure 5). Table 3 shows both the uni-variate and multivariate Cox proportional-hazard analyses for the prediction of the clinical outcomes. As a result, both the LVEF during exercise and the TMVr therapy were independently associated with the clinical outcomes [hazard ratio (HR): 0.919, 95% confidence interval (CI): 0.853–0.991, P = 0.028, and HR: 0.419, 95% CI: 0.015–0.941, P = 0.044, respectively].

Discussion

The main findings of the current study will now be highlighted: (i) about 25% of patients with secondary MR presented with a significant increase in MR during exercise, (ii) most of the patients with exercise-induced secondary MR did not undergo open surgery, (iii) patients with exercise-induced secondary MR had a poor prognosis, particularly those hospitalized for heart failure, (iv) the TMVr procedure with the MitraClip was safely and successfully performed in patients with exercise-induced secondary MR, and (v) patients with exercise-induced secondary MR who underwent TMVr reported a better prognosis than that in the control group.

Associations between prognosis and exercise-induced secondary MR

Previous studies reported that significant increase in the EROA during exercise is observed in 30% of patients with heart failure and associated with increased mortality and hospitalization for heart failure.^7,9,11,17,30 In the present study, exercise-induced secondary MR was observed in around 25% of the patients with secondary MR, whereas the prognosis of the patients treated with medication (i.e. the control group) was poor, with a 2-year event-free rate of 60%. Exercise-induced MR has been reported to both increase insufficient forward flow and cause an exaggerated increase in the left atrial and pulmonary pressures during exercise.^24,25 Piérard and Lancellotti ³¹ demonstrated that exercise-induced MR deterioration, following the augmentation of the backward flow and pulmonary vascular congestion, may contribute to the development of acute pulmonary oedema. In the current investigation, the control group presented with increased hospitalization due to heart failure than reported in the TMVr group. Moreover, although the severity of the MR at rest was moderate (mean EROA: 0.22 cm²), the event-free survival rate was comparable with that of patients with severe secondary MR at rest in earlier studies.^32,33 This may partly explain the poor prognosis also seen in patients with moderate MR at rest. In addition, such patients are generally underestimated in clinical practice and are less likely to be candidates for invasive treatment. Indeed, low rates of surgery for secondary MR (4% of patients) were seen in this study.

Effect of the TMVr using the MitraClip on exercise-induced secondary MR

The COAPT and Mitra-FR on the performance of the TMVr using the MitraClip for secondary MR have reported different results and attracted attention.^16,29 Nonetheless, Grayburn et al.³⁴ suggested a difference in proportionality between the LV volume and the EROA. In fact, while the MITRA-FR trial enrolled patients with secondary MR proportional to the degree of LV dilation, the COAPT trial demonstrated a disproportionate secondary MR. Although our study found a proportional relationship between the LV size and secondary MR at rest, a disproportionate relationship similar to that found in the COAPT trial was seen during exercise ( Figure 2). Such characteristics may be important for selecting the optimal treatment for patients with heart failure.

Increased LA pressure and decreased LA compliance due to MR resulted in pulmonary venous congestion and increased pulmonary arterial pressure. Pulmonary and RV haemodynamics during exercise significantly correlated with the symptoms of functional limitation.^17,24,27,35 Exercise intolerance is also associated with reduced quality of life and increased hospitalization and mortality in patients with heart failure.³⁶ Previous studies reported that increased secondary MR during exercise is associated with increased SPAP.^17,24 In fact, although CO did not differ between the two groups, the TMVr group had higher SPAP during exercise, more frequent exercise-induced pulmonary hypertension, and worse clinical symptoms than in the control group. Belgian centres reported that TMVr using the MitraClip improved the haemodynamics, including pulmonary artery pressure and CO, during exercise.¹⁵ In this study, the TMVr group showed improvement in heart failure symptoms within 1 year of intervention, and the TMVr increased the quality of life of patients. Improved haemodynamics and exercise capacity during exercise with TMVr may contribute to improved prognosis in patients with secondary MR. In a study evaluating pulmonary and RV circulation during exercise before and after cardiac resynchronization therapy, it was reported that the reduction of secondary MR during exercise contributed to improvement of SPAP, RV function, and its coupling during exercise.³⁷ In this study, the reduction of MR by TMVr was considered to have improved the pulmonary and RV circulation, and improved the prognosis in patients with heart failure and secondary MR.

Study limitation

This study had some limitations. First, although the number of cases in this study is large compared with previous studies on ESE for secondary MR, this investigation is a single-centred study with a small sample size. Therefore, a multi-centred study with a larger sample size is warranted. Second, given the retrospective nature of this study, we found that the medication regimen could not be standardized for secondary MR. However, β-blockers, ACEi/ARB, and mineralocorticoid receptor antagonists were introduced in 85, 80, and 48% of patients, respectively, and at a comparable rate with that of the COAPT study in which heart failure medication was standardized.¹⁶ Sacubitril/valsartan has not been introduced in Japan, and none of the patients enrolled in this study took it. A study on the effect of sacubitril/valsartan on exercise-induced secondary MR is also expected in the future. Third, selection bias could not be excluded from this single-centred study and propensity score-matching analysis was not performed. Resting LV volume and ejection fraction tended to be lower and higher in the TMVr group than in the control group, but in the cohort of this study, resting LV volume and ejection fraction were not associated with prognosis in Cox hazard analysis and were considered to have little impact on the results of this study. One indication for the TMVr using MitraClip in Japan is a LVEF of more than 30%, which may have affected patient selection for TMVr in this study. However, the TMVr contributed to an improved prognosis in the multivariate analysis, despite the older age and higher STS scores and NYHA functional class of the group compared with the control group. Forth, the Doppler methods used to quantify MR have some pitfalls, including the limited measurement of the PISA radius to only one velocity and time point here presented. Nevertheless, an earlier study validated the use of such a method both at rest and during exercise.²⁰ Fifth, recent studies on LA dynamics using the speckle tracking method for patients with MR, and with normal values of LA volumes and function have been conducted; suggesting that LA function has attracted attention.^38,39 However, the study on LA dynamics in this study is insufficient, and further studies are needed to examine which patients benefit from TMVr in terms of LA function. Sixth, in this study, cardiac function was assessed using ESE, but exercise tolerance using cardiopulmonary exercise testing was not evaluated in detail. In recent years, it has been reported that the combination of cardiopulmonary exercise testing and echocardiography is useful for understanding the pathophysiology.⁴⁰ Therefore, further studies are needed to investigate the effectiveness of catheter treatment for secondary MR using cardiopulmonary exercise testing with echocardiography. Seventh, exercise haemodynamic assessment after the TMVr has not been performed in this study. Dynamic assessment before and after TMVr using the MitraClip is expected to identify potential responders. Finally, Nt-proBNP was measured only at rest and there were no data during exercise. BNP during exercise has been reported to be useful in predicting prognosis in patients with MR ⁴¹ and further studies for evaluating the prognostic impact of BNP during exercise are expected.

Conclusions

While patients with exercise-induced secondary MR reported unfavourable clinical outcomes when following a medical therapy, those who underwent the TMVr presented a lower risk of composite endpoints for heart failure.

Conflict of interest: M.I. is a clinical proctor for Abbott Vascular. The other authors have no conflicts of interest to declare.

References