HIGHLIGHTS
- •
Left ventricular stroke work index (LVSWi) and right ventricular stroke work index (RVSWi) independently predict mortality after transcatheter mitral valve repair (TMVR).
- •
Patients with low LVSWi and low RVSWi show significantly lower survival after TMVR.
- •
Risk-score based on LVSWi and RVSWi discriminates between low- and high-risk patients.
- •
LVSWi and RVSWi may help to identify optimal patients for TMVR.
Optimal patient selection for transcatheter mitral valve repair (TMVR) remains challenging. The aim of the study was to assess the impact of left and right ventricular stroke work index (LVSWi, RVSWi) on mortality in patients with chronic heart failure (CHF) undergoing TMVR. One hundred-forty patients (median age 74 ± 9.9 years, 67.9% male) with CHF who underwent successful TMVR were included. Primary end point was defined as all-cause mortality after 16 ± 9 months of follow-up. LVSWi was calculated as: Stroke volume index (SVi) * (mean arterial pressure – postcapillary wedge pressure) * 0.0136 = g/m −1 /m 2 . RVSWi was calculated as: SVi * (mean pulmonary artery pressure – right atrial pressure) * 0.0136 = g/m −1 /m 2 . Receiver operating characteristic (ROC) analysis determined an optimal threshold of 24.8 g/m −1 /m 2 for LVSWi (sensitivity 80.4%, specificity 40.2%, area under the curve (AUC) 0.71 [0.60 to 0.81]; p = 0.001) and 8.3 g/m −1 /m 2 for RVSWi (sensitivity 67.4%, specificity 57.0%, AUC 0.67 [0.56 to 0.78]; p = 0.006), respectively. Kaplan-Meier analysis showed significantly lower survival in patients with LVSWi ≤24.8 g/m −1 /m 2 (20.0% vs 39.4%; log-rank p = 0.038) and in patients with RVSWi ≤8.3 g/m −1 /m 2 (22.1% vs 43.7%; log-rank p = 0.026), respectively. LVSWi of ≤24.8 g/m −1 /m 2 and RVSWi of ≤8.3 g/m −1 /m 2 were independent predictors for all-cause mortality (hazard ratio (HR) 2.83; 95% confidence interval (CI) 1.1 to 7.6; p = 0.04; HR 2.52; 95% CI 1.04 to 6.1; p = 0.041). A risk-score incorporating LVSWi and RVSWi cut-off values from ROC analysis powerfully predicts long-term survival after successful TMVR (log-rank p = 0.02). In conclusion, LVSWi and RVSWi independently predict mortality in patients with CHF undergoing TMVR and might be useful in risk stratification of TMVR candidates.
Transcatheter mitral valve repair (TMVR) using the MitraClip system (Abbott Vascular, Abbott Park, Illinois) has emerged as an effective treatment option for surgical high-risk patients with severe functional mitral regurgitation (MR) and chronic heart failure (CHF). However, several predictors of worse prognosis in patients treated by TMVR have been identified, for example, high levels of NT-proBNP, New York Heart Association (NYHA) functional class IV prior to TMVR and a severely impaired left and right heart function. , , Two recently published randomized controlled clinical trials investigating clinical outcome of TMVR for patients with severe functional MR and CHF yielded different results. , The COAPT trial revealed a benefit in reduction in heart failure hospitalizations and mortality whereas the MITRA-FR study found no differences between treatment groups. One reason for the presumed inconsistent findings of these 2 trials might be related to key differences in patient selection. Thus, optimal patient selection for TMVR still remains a matter of debate and there is an unmet need for identifying additional risk factors of worse prognosis. It is the purpose of the present study to evaluate if left and right ventricular stroke work index (LVSWi, RVSWi) as hemodynamic parameters of cardiac function could help identifying optimal candidates for TMVR using the MitraClip system.
Methods
From March 2015 to April 2018 all consecutive patients with NYHA functional class III or IV suffering from severe MR due to CHF who underwent successful TMVR (MR ≤2+ at discharge) at the Bremen Heart Center in Germany were included. CHF was defined as heart failure from any cause with reduced left ventricular ejection fraction (LV-EF) ≤50%. All patients received optimal medical and device treatment at least 3 months prior to the MitraClip procedure according to the current heart failure guidelines. Patients undergoing TMVR were enrolled if they were judged inoperable or at unacceptable high surgical risk based on the logistic European System for Cardiac Operative Risk Evaluation (logistic EuroSCORE) and if they had a favorable anatomy suitable for the MitraClip procedure. The suitability was determined by a heart team decision. All patients included in the study were fully informed about the procedure and signed a written consent form. The study complies with the Declaration of Helsinki and the locally appointed ethics committee has approved the research protocol.
Transthoracic and transesophageal echocardiographic evaluations were performed at baseline. Severity grade of MR at baseline was assessed according to the current guidelines. , Transthoracic echocardiographic evaluations were performed at pre-discharge, 30 days after the procedure and, if possible, at follow-up. MR severity grade after TMVR was evaluated according to the technique reported by Foster et al.
A baseline invasive hemodynamic study was conducted in all included patients during the screening phase in a conscious non-sedated state. Right heart catheterization was performed using a 6F single lumen, balloon-tipped, flow-directed Swan-Ganz catheter (Arrow International, Inc, Reading, Pennsylvania) to obtain the following variables: pulmonary capillary wedge pressure (PCWP) including v-wave analysis, pulmonary artery systolic, mean and diastolic pressure (PASP, PAP mean, PAP diast.), pulmonary artery oxygen saturation and right atrial pressure (RAP), systemic arterial systolic, mean and diastolic pressure (RR syst., RR mean, RR diast.). Systemic arterial oxygen saturation was obtained from the LV pigtail catheter. Cardiac output (CO) was calculated by the Fick method. Stroke volume (SV) was calculated as CO / heart rate. Stroke volume index (SVi) was calculated as SV / body surface area (BSA). BSA was estimated from the Du Bois formula. Pulmonary vascular resistance (PVR) was calculated as the ratio between the pressure drop along the vascular bed and the CO and converted in metric units (dyn*s*cm −5 ). All values are reported at end-expiration. Pulmonary artery pulsatility index (PAPi) was calculated as (PASP – PAP diast.) / RAP. LVSWi and RVSWi were calculated as: LVSWi = SVI * (RR mean – PCWP) * 0.0136 and RVSWi = SVI * (PAP mean – RAP) * 0.0136, respectively. Transpulmonary gradient (TPG) was calculated as PAP mean – PCWP. Diastolic pulmonary gradient (DPG) was calculated as PAP diast. – PCWP. The cardiac filling pressures (CFP) as the RAP to PCWP ratio was calculated as RAP / PCWP. All MitraClip procedures were performed under general anesthesia using (3D-) transesophageal echocardiographic and fluoroscopic guidance. TMVR with the MitraClip was performed as previously described. ,
Primary end point was all-cause mortality. Periprocedural and in-hospital major adverse events were reported such as death, myocardial infarction, major stroke, renal failure with the need for renal replacement therapy, pericardial tamponade with the need for pericardiocentesis and urgent or emergent cardiovascular surgery for adverse events. Bleeding complications with the need for blood transfusion and surgical re-operation for recurrent MR were documented. Follow-up was conducted at 30 days and after a mean follow-up period of 16±9 months. All follow-up evaluations were conducted by the Bremer Institut für Herz- und Kreislaufforschung (BIHKF), Germany. If a patient was not able to be present at follow-up, a telephone interview was conducted with either the patient himself, the patient`s relatives or general practitioner. Major adverse cardiac and cerebrovascular events (MACCE) were analyzed including all-cause death, major stroke, non-fatal myocardial infarction and surgical or interventional (TMVR) re-do for recurrent severe MR according to the MVARC criteria. Changes in functional capacity, categorized by the NYHA functional class, were assessed as well as a state of health self-assessment based on a standardized health-related quality of life questionnaire (EQ-5D).
Continuous data are expressed as mean ± standard deviation (SD) or median (interquartile range) where appropriate. Mann-Whitney test was used to compare continuous variables. Categorial variables are presented as numbers and proportions and were compared using chi-square or Fisher`s exact test. Receiver operating characteristic (ROC) curves were used to assess the discriminative capacity of LVSWi and RVSWi and to determine related cut-off scores for primary end point. Parametric/nonparametric distribution of data was assumed by Kolmogorov-Smirnow-testing and distribution analysis. Pearson`s and Spearman Rho correlation function was used to analyze the association between LVSWi and LV-EF and between RVSWi and echocardiographic parameters of RV systolic function, e.g. tricuspid annular plane systolic excursion (TAPSE) and doppler tissue imaging S` (DTI-S`). The Kaplan-Meier method was used for survival analysis. Log-rank testing was used to compare event-free survival from primary end point. A multivariable Cox proportional-hazards regression analysis was performed to assess the association between LVSWi and RVSWi and all-cause mortality, adjusting for covariates reported in the literature to be associated with mortality; these included TAPSE ≤16mm, LV-EF ≤25%, severe tricuspid valve regurgitation and those variables significantly different between groups (male gender, creatinine levels ≥1.5mg/dl and NT-proBNP levels ≥10.000ng/l). A ventricular stroke work index (VSWi) risk score predicting all-cause mortality at long-term follow-up was generated from the LVSWi and RVSWi cut-off values from the aforementioned ROC curve analysis. A 2-sided p value <0.05 was considered statistically significant. All statistical analyses were performed using SPSS version 22 (SPSS, Inc, Chicago, Illinois).
Results
A total of 140 consecutive patients were enrolled ( Table 1 ). Mean age of all patients was 74 ± 9.9 years (68% male). Mean LVSWi of the study population was 22.3 ± 10.7 g/m −1 /m 2 and 8.9 ± 4.1 g/m −1 /m 2 for RVSWi, respectively. Clinical follow-up was obtained in all included patients. The rate of in-hospital mortality and the rate of all-cause mortality at 30 days after the procedure was 1.4% and 2.9%, respectively. No procedure related death or emergent cardiovascular surgery for adverse events occurred in the study population. Bleeding complications requiring transfusion of 2 or more units of blood occurred in 6 patients (4.2%, 3 in each group). In 4 cases, the cause of bleeding was related to vascular access site. In 2 cases, the cause of bleeding was related to gastrointestinal hemorrhage. One patient (0.7%) experienced pericardiac tamponade with the need for pericardiocentesis. All patients underwent a single MitraClip procedure. No patient needed an interventional re-do for recurrent MR after TMVR as a second procedure during the follow-up period. The rate of hospitalization for heart failure at 30 days after the procedure and at long-term follow-up was 9.3% and 35.0%, respectively. The primary end point of all-cause mortality at long-term follow-up occurred in 46 patients (33.1%). Patients who died presented lower LVSWi and lower RVSWi, respectively ( Table 2 ). Nonsurvivors were more likely to have higher levels of NT-proBNP, higher levels of creatinine and a higher logistic EuroSCORE ( Table 1 ). The proportion of male gender was significantly higher in nonsurvivors. Patients who died showed a higher proportion of atrial fibrillation and chronic obstructive lung disease ( Table 1 ). The proportion of patients suffering from NYHA functional class IV prior to TMVR and the numbers of clips implanted were similar in both groups ( Table 1 ).
| Variable | All Patients (n=140) | Survivors (n=94) | Non-Survivors (n=46) | p-value |
|---|---|---|---|---|
| Age (years ± SD) | 74 ± 9.9 | 74 ± 10.7 | 74 ± 8.4 | 0.8 |
| Men | 95 (68%) | 57 (61%) | 37 (80%) | 0.023 |
| NYHA functional class IV | 24 (17%) | 12 (13%) | 12 (26%) | 0.13 |
| Coronary artery disease | 85 (61%) | 55 (59%) | 29 (63%) | 0.43 |
| Chronic atrial fibrillation | 67 (48%) | 39 (42%) | 28 (61%) | 0.06 |
| Hypertension | 101 (72%) | 67 (72%) | 33 (72%) | 0.9 |
| Chronic obstructive pulmonary disease | 21 (15%) | 101 (11%) | 11 (24%) | 0.06 |
| Diabetes mellitus | 30 (21%) | 19 (19%) | 12 (26%) | 0.38 |
| Logistic EuroSCORE (mean % ± SD) | 22.6 ± 15.1 | 20.1 ± 13.8 | 27.9 ± 16.6 | 0.001 |
| NT-proBNP (mean ng/l ± SD) | 8430 ± 10972 | 6745 ± 10820 | 12121 ± 10602 | 0.001 |
| Creatinine (mean mg/dl ± SD) | 1.5 ± 0.8 | 1.4 ± 0.8 | 1.8 ± 0.8 | <0.001 |
| Number of Clips implanted | 0.13 | |||
| 1 | 69 (49%) | 51 (54%) | 18 (39%) | |
| 2 | 69 (49%) | 41 (44%) | 28 (61%) | |
| 3 | 2 (1%) | 2 (2%) | 0 |
| Variable | All Patients (n=140) | Survivors (n=94) | Non Survivors (n=46) | p-value |
|---|---|---|---|---|
| Heart rate (mean beats/min ± SD) | 77 ± 15 | 76 ± 15 | 79 ± 15 | 0.09 |
| Systemic arterial systolic pressure (mean mmHg ± SD) | 127 ± 15 | 131 ± 27 | 120 ± 23 | 0.019 |
| Systemic arterial diastolic pressure (mean mmHg ± SD) | 71 ± 13 | 72 ± 14 | 68 ± 11 | 0.24 |
| Systemic arterial mean pressure (mean mmHg ± SD) | 91± 18 | 94 ± 17 | 84 ± 17 | 0.004 |
| Pulmonary artery systolic pressure (mean mmHg ± SD) | 60 ± 16 | 60± 17 | 58 ± 15 | 0.64 |
| Mean pulmonary artery pressure (mean mmHg ± SD) | 39 ± 11 | 39 ± 11 | 37 ± 9 | 0.48 |
| Postcapillary wedge pressure (mean mmHg ± SD) | 28 ± 10 | 29 ± 11 | 27 ± 9 | 0.57 |
| V-wave (mean mmHg ± SD) | 42 ± 15 | 42 ± 16 | 41 ± 15 | 0.61 |
| Pulmonary artery diastolic pressure (mean mmHg ± SD) | 24 ± 9 | 25 ± 10 | 24 ± 8 | 0.97 |
| Right atrial pressure (mean mmHg ± SD) | 13 ± 7 | 13 ± 7 | 13 ± 6 | 0.65 |
| Cardiac output (mean l/min ± SD) | 3.7 ± 1.3 | 3.7 ± 1.4 | 3.6 ± 1.0 | 0.74 |
| Cardiac index (mean l/min/m 2 ± SD) | 1.9 ± 0.6 | 2.0 ± 0.6 | 1.9 ± 0.5 | 0.91 |
| Pulmonary vascular resistance (mean dyn*s*cm -5 ± SD) | 277 ± 196 | 295 ± 211 | 242 ± 159 | 0.25 |
| Cardiac filling pressures (mean mmHg ± SD) | 0.49 ± 0.24 | 0.47 ± 0.20 | 0.53 ± 0.29 | 0.31 |
| Pulmonary artery pulsatility index (mean ± SD) | 3.51 ± 2.67 | 3.67 ± 2.87 | 3.1 ± 2.21 | 0.07 |
| Left ventricular stroke work index (mean g/m -1 /m 2 ± SD) | 22.3 ± 10.7 | 24.1 ± 11.5 | 18.6 ± 7.9 | 0.005 |
| Right ventricular stroke work index (mean g/m -1 /m 2 ± SD) | 8.9 ± 4.1 | 9.4 ± 4.4 | 7.8 ± 3.2 | 0.041 |
| Transpulmonary gradient (mean mmHg ± SD) | 10.6 ± 7.1 | 10.8 ± 6.9 | 10.2 ± 7.5 | 0.38 |
| Diastolic pulmonary gradient (mean mmHg ± SD) | 3.6 ± 7.5 | 3.9 ± 7.3 | 3.0 ± 7.4 | 0.71 |
| Pulmonary artery compliance (mean ml/mmHg ± SD) | 1.61 ± 0.87 | 1.63 ± 0.92 | 1.55 ± 0.78 | 0.66 |
| Stroke volume (mean ml/beat) | 50 ± 21 | 51 ± 22 | 47 ± 17 | 0.22 |
| Stroke volume index (mean ml/m 2 /beat) | 26.1 ± 9.7 | 26.8 ± 10.4 | 24.6 ± 8.2 | 0.16 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
