Abstract
Background
Low serum albumin levels indicate frailty and are strong predictors of poor prognosis after mitral transcatheter edge‐to‐edge repair (M‐TEER); however, albumin levels are expected to improve in response to treatment for mitral regurgitation. No reports have elucidated the clinical effects of albumin changes after M‐TEER. Thus, we aimed to explore the association between early albumin changes and clinical outcomes after M‐TEER and to identify the associated factors.
Methods
Data from 2695 patients enrolled in a Japanese multicenter registry whose serum albumin levels were measured before and 1 month after undergoing M‐TEER were retrospectively reviewed. Changes in albumin (Δ‐albumin) were calculated and divided into 2 groups, namely improved albumin (Δ‐albumin >0) and worsening albumin (Δ‐albumin ≤0) groups. The incidence, predictors, and clinical outcomes associated with early albumin changes were investigated.
Results
After M‐TEER, albumin levels improved in 56.1% of the patients (n=1512). The independent predictors of worsening Δ‐albumin comprised old age, acute procedure success not achieved, higher Clinical Frailty Scale, higher hemoglobin, impaired renal function, and preprocedural higher albumin levels (all P<0.05). Improved Δ‐albumin was independently associated with reduced all‐cause mortality after M‐TEER (hazard ratio [HR], 0.62 [95% CI, 0.52–0.75], P<0.001). Additionally, improved Δ‐albumin was associated with a lower competing risk of heart failure hospitalization (HR, 0.78 [95% CI, 0.64–0.94], P=0.01).
Conclusions
Early improvement in Δ‐albumin may serve as a measure of procedural benefits and a surrogate marker for predicting clinical outcomes after M‐TEER.
Keywords: albumin, all‐cause mortality, dynamic change, mitral regurgitation, mitral‐transcatheter edge‐to‐edge repair
Nonstandard Abbreviations and Acronyms
- APS
acute procedural success
- MR
mitral regurgitation
- M‐TEER
mitral transcatheter edge‐to‐edge repair
- OCEAN
Optimized Catheter Valvular Intervention
- TTE
transthoracic echocardiography
Clinical Perspective.
What Is New?
After mitral‐transcatheter edge‐to‐edge repair, improvement in albumin levels has been associated with reduced all‐cause mortality and a lower competing risk of heart failure hospitalization.
Old age, acute procedure success not achieved, higher Clinical Frailty Scale, higher hemoglobin, impaired renal function, and preprocedural higher albumin levels independently predict worsening albumin change.
What Are the Clinical Implications?
This novel study demonstrates the efficacy of early albumin changes for predicting patient prognosis after mitral‐transcatheter edge‐to‐edge repair.
Mitral transcatheter edge‐to‐edge repair (M‐TEER) is a minimally invasive treatment for patients with mitral regurgitation (MR). Previously, randomized trials have indicated the effectiveness of M‐TEER using the MitraClip device (Abbott Vascular, Menlo Park, CA, USA) in MR management. 1 , 2 Appropriate patient screening, risk assessment before treatment, and careful follow‐up after discharge are required to optimize therapeutic outcomes. Frailty 3 and malnutrition 4 , 5 , 6 have been associated with poor prognosis after M‐TEER; however, the dynamic change in frailty status after M‐TEER remain unclear.
Serum albumin, a biomarker associated with frailty and sarcopenia, 7 serves as a nutritional indicator. Furthermore, it has been associated with important clinical outcomes, including mortality, in patients with heart failure (HF), 8 , 9 and patients undergoing transcatheter aortic valve replacement. 10 Additionally, a subanalysis of the COAPT (Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy for Heart Failure Patients With Functional Mitral Regurgitation) trial demonstrated a significant association between preprocedural low albumin levels and increased mortality after the M‐TEER group and medical therapy group. 4 Furthermore, improved albumin levels have been associated with favorable prognosis in patients with HF, 11 , 12 , 13 suggesting the need to assess albumin trends rather than single‐point levels. However, the trajectory of serum albumin after M‐TEER has not been investigated. Furthermore, it is unclear whether factors associated with albumin changes and the degree of change affect patient prognosis. Investigating early nutritional changes as an indicator of frailty may reflect patient's responsiveness to M‐TEER.
In this study, we aimed to elucidate the association between early albumin changes and clinical outcomes after M‐TEER, stratified by preprocedural levels and to identify the associated factors.
METHODS
The data, analytic methods, and study materials will not be made available to other researchers for purposes of reproducing the results or replicating the procedure.
Study Population and Definitions of Albumin Change
All data were extracted from the Japanese multicenter OCEAN (Optimized Catheter Valvular Intervention) mitral registry. The OCEAN mitral registry encompasses 21 Japanese institutions and prospectively enrolled patients who underwent M‐TEER for substantial MR. 14 , 15 This study was registered with the University Hospital Medical Information Network Clinical Trials Registry via the International Committee of Medical Journal Editors (UMIN000023653). All patients provided informed consent before undergoing M‐TEER, and the study protocol was approved by the institutional review board of each participating institution. This study was conducted per the latest version of the tenets of the Declaration of Helsinki and the guidelines for epidemiological studies issued by the Ministry of Health, Labor, and Welfare of Japan.
Between April 2018 and June 2023, 3764 consecutive patients with symptomatic MR underwent M‐TEER. The decision on the indication for M‐TEER was made based on consensus within the local heart team, particularly when considering surgical risk in patients with multiple comorbidities. This was especially relevant in patients with degenerative MR, for whom surgery is generally the first‐line treatment but who were considered high‐risk for surgery. Except for 59 patients who died within 1 month after M‐TEER and 1010 patients with missing data of albumin value, 2695 patients (76.1%) had serum albumin levels measured before and 1 month after M‐TEER. The delta (Δ) albumin (Δ‐albumin) level was calculated as follows: ([level at 1 month after M‐TEER – preprocedural level]/preprocedural level)×100. The cohort was divided into 2 groups as follows: (1) improved albumin (Δ‐albumin >0) and (2) worsening albumin (Δ‐albumin ≤0). Additionally, the cohort was divided by the median preprocedural level (3.7 g/dL) because the rate of change depends on the preprocedural albumin level by regression to the mean. The low‐albumin group included patients with preprocedural albumin below the median, whereas the high‐albumin group included patients with levels above the median.
Mitral Transcatheter Edge‐to‐Edge Repair
During the follow‐up, the MitraClip device was the only commercially available mitral transcatheter system in Japan. After introducing the MitraClip Generation 2 (G2) system, the latest version of the G4 system (excluding the G3 system) was launched in September 2020. The G4 system comprises 4 size variations (NT, NTW, XT, and XTW) to accommodate different mitral valve morphologies. M‐TEER was conducted in patients with moderate‐to‐severe or severe MR regardless of cause (degenerative MR and functional MR). Baseline MR severity was defined based on the American Society of Echocardiography guidelines, 16 whereas residual MR was assessed as described previously. 17
MR severity was classified as follows: 0+ (none/trivial), 1+ (mild), 2+ (moderate), 3+ (moderate to severe), and 4+ (severe). Indications for M‐TEER were determined by heart team members based on patient background, including age, frailty, and the perioperative risk of cardiac surgery. M‐TEER has been described previously. 14 , 15 All M‐TEER procedures were conducted under general anesthesia and guided by transesophageal echocardiography. An acceptable MR reduction was defined as MR ≤2+ using peri‐ or postoperative transthoracic echocardiography findings. Acute procedural success (APS) was defined as no life‐threatening complications and an achievement of MR ≤2+ in transthoracic echocardiography at discharge. 14 , 15
Clinical Outcome Measures and End Points
Baseline patient characteristics, laboratory data, transthoracic echocardiography, and transesophageal echocardiography findings were obtained from each participating center. Clinical follow‐ups were conducted at baseline, 1 month, 1 year, and annually thereafter. At each follow‐up visit, hospitalizations for HF (HFH) were recorded. For patients unable to visit hospitals, mortality and HFH data were collected through telephone interviews with the patients, their families, or relatives. Baseline patient data, procedural variables, and clinical outcomes were compared between the improved and worsening groups. The primary end points comprised the incidences of all‐cause mortality and HFH after M‐TEER.
Statistical Analysis
Data are presented as means±SD, median for continuous variables, and frequencies (percentages) for categorical variables. Data comparisons were conducted using χ2 tests for categorical covariates and Student's t test or using the Mann–Whitney U test depending on their distribution. Statistical significance was set at P<0.05, and 95% CIs were reported where applicable. A multivariate linear regression analysis was conducted to explore the factors associated with Δ‐albumin, which served as the dependent variable. The multivariate model included patient characteristics associated with cardiovascular disease and comorbidities, M‐TEER procedural factors, and preprocedural albumin levels.
The Kaplan–Meier method was used to estimate cumulative incidences of mortality and HFH; group differences were assessed using the log‐rank test. We also used the cumulative incidence function to estimate the cumulative incidence of HFH, accounting for competing risk, with death without HFH treated as a competing event. The Gray model was developed to analyze the association with the risk of HFH while appropriately addressing competing risks. Univariate Cox regression analysis was conducted to calculate hazard ratios (HR) for all‐cause mortality and HFH during follow‐up. Thereafter, a multivariate analysis was conducted using the baseline clinical characteristics and other variables with a univariate P<0.1 to examine the independent association between Δ‐albumin and clinical outcomes. Similarly, HFH was considered a competing risk and HRs was calculated using Fine and Gray analysis. To evaluate the association between Δ‐albumin and mortality, a multivariable Cox regression analysis with restricted cubic splines was conducted, with knots selected at the quartiles of the Δ‐albumin distribution. Potential effect modification was explored through subgroup multivariate analyses stratified by age (≥80 years), sex, Clinical Frailty Scale (≥4), New York Heart Association class III or IV, estimated glomerular filtration rate (<30 mL/min per 1.73 m2), and cause (functional MR and degenerative MR).
Statistical analyses were conducted using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), a graphical user interface for R (R Foundation for Statistical Computing, Vienna, Austria). 18
RESULTS
Baseline Characteristics and Procedural Variables
Table 1 summarizes the baseline characteristics associated with albumin changes. Compared with preprocedure levels, albumin levels improved in 56.1% of patients 1 month after M‐TEER. Several clinical characteristics differed in the improved Δ‐albumin group, including younger age, lower body mass index, a history of HFH, and New York Heart Association class III/IV, low preprocedural albumin and hemoglobin levels, high B‐type natriuretic peptide level and estimated glomerular filtration rate. Transthoracic echocardiography parameters, such as left ventricular ejection fraction, MR effective regurgitant orifice area, demonstrated differences (all P<0.05). Table 2 summarizes the procedural variables. The APS achievement rate was higher and incidence of access site complications were lower in the improved Δ‐albumin group (all P<0.05). Additionally, baseline patient characteristics by cause of MR and for patients excluded due to death within 1 month after M‐TEER or missing data of albumin (n=1069) are shown in Tables S1 and S2.
Table 1.
Baseline Characteristics of Study Participants
| Δ‐albumin worsening n=1183 | Δ‐albumin improve n=1512 | P value | |
|---|---|---|---|
| Baseline clinical characteristics | |||
| Age, y | 79.2±9.0 | 78.4±9.8 | 0.023 |
| ≥80 y, n (%) | 672 (56.8%) | 838 (55.4%) | 0.482 |
| Male sex, n (%) | 663 (56.0%) | 845 (55.9%) | 0.938 |
| Body height, cm | 156.7±9.9 | 157.3±10.4 | 0.123 |
| Body weight, kg | 53.2±11.7 | 52.8±11.5 | 0.481 |
| Body mass index, kg/m2 | 21.5±3.7 | 21.2±3.4 | 0.021 |
| Heart rate, bpm | 73.6±15.3 | 73.9±15.2 | 0.577 |
| Functional MR, n (%) | 814 (68.8%) | 1087 (71.9%) | 0.089 |
| Clinical frailty scale, points | 4 (3–4) | 4 (3–4) | 0.127 |
| Society of Thoracic Surgeons Predictive Risk of Mortality score for mitral valve replacement, n=2493, (%) | 10.7±8.3 | 10.9±7.5 | 0.481 |
| Prior heart failure hospitalization within 12 mo before enrollment, n. (%) | 818 (69.3%) | 1110 (73.7%) | 0.014 |
| NYHA functional class | 2.6±0.7 | 2.8±0.7 | <0.001 |
| NYHA functional class, III or IV | 672 (56.8%) | 952 (63.0%) | 0.001 |
| Preprocedural laboratory data | |||
| Albumin, g/dL | 3.9±0.5 | 3.5±0.5 | <0.001 |
| B‐type natriuretic peptide, pg/mL, n=2102 | 302.0 (159.8–613.0) | 377.4 (190.6–718.8) | <0.001 |
| Hemoglobin, g/dL | 12.1±1.8 | 11.5±1.8 | <0.001 |
| Estimated glomerular filtration rate, mL/min per 1.73 m2 | 38.7±18.6 | 40.3±20.0 | 0.033 |
| <30 mL/min per 1.73 m2, n (%) | 406 (34.4%) | 460 (30.5%) | 0.034 |
| Postoperative albumin data | |||
| Albumin, g/dL | 3.6±0.5 | 3.9±0.5 | <0.001 |
| Δ‐albumin, % | −6.5±7.6 | 13.1±10.8 | <0.001 |
| Comorbidities | |||
| Hypertension, n (%) | 799 (67.5%) | 999 (66.1%) | 0.434 |
| Dyslipidemia, n (%) | 572 (48.4%) | 743 (49.1%) | 0.698 |
| Diabetes, n (%) | 298 (25.2%) | 401 (26.5%) | 0.452 |
| Chronic kidney disease, n (%) | 1043 (88.2%) | 1289 (85.3%) | 0.223 |
| Dialysis dependent, n (%) | 65 (5.5%) | 86 (5.7%) | 0.866 |
| Atrial fibrillation/flutter, n (%) | 714 (60.4%) | 918 (60.7%) | 0.874 |
| Liver cirrhosis, n (%) | 19 (1.6%) | 25 (1.7%) | 1.0 |
| Chronic obstructive lung disease, n (%) | 110 (9.3%) | 146 (9.7%) | 0.791 |
| Peripheral artery disease, n (%) | 111 (9.4%) | 152 (10.1%) | 0.601 |
| Echocardiographic data | |||
| LV ejection fraction, % | 46.3±16.8 | 44.7±16.5 | 0.011 |
| <40%, n (%) | 498 (42.1%) | 691 (45.7%) | 0.066 |
| LV end‐diastolic diameter, mm | 56.7±10.4 | 56.7±10.3 | 0.933 |
| LV end‐systolic diameter, mm | 43.4±13.7 | 44.0±13.3 | 0.21 |
| Left atrial volume index, cm3/m2 | 89.3±56.5 | 86.3±50.3 | 0.153 |
| MR characteristics | |||
| Moderate–severe MR, n (%) | 1157 (97.8%) | 1492 (98.7%) | 0.099 |
| MR proximal isovelocity surface area effective regurgitant orifice area, cm2 | 0.37±0.19 | 0.39±0.21 | 0.041 |
| MR regurgitation volume, mL | 56.5±39.7 | 56.3±27.0 | 0.86 |
| Transmitral mean pressure gradient, mm Hg | 1.87±1.07 | 1.88±1.13 | 0.836 |
Values are numbers (%) or mean±SD, median with interquartile range. LV indicates left ventricular; MR, mitral regurgitation; and NYHA, New York Heart Association.
Table 2.
Procedural Results
| Δ‐albumin worsening n=1183 | Δ‐albumin improve n=1512 | P value | |
|---|---|---|---|
| Procedural variables | |||
| Procedural time, min | 92.4±45.1 | 90.1±43.1 | 0.205 |
| Device time, min | 66.3±39.2 | 63.3±36.6 | 0.079 |
| Fluoroscopy time, min | 29.4±20.2 | 28.3±19.3 | 0.179 |
| Length of hospital stay, d | 21.5±31.0 | 21.4±21.3 | 0.928 |
| Length of intensive care unit stay, d | 2.2±6.5 | 1.8±3.7 | 0.04 |
| MitraClip detail information | |||
| Number of implanted clips | 1.29±0.49 | 1.29±0.48 | 0.936 |
| Generation 2 NT, n (%) | 762 (50.1) | 939 (48.3) | |
| Generation 4 NT, n (%) | 173 (11.4) | 233 (12.0) | |
| Generation 4 NTW, n (%) | 294 (19.3) | 406 (20.9) | |
| Generation 4 XT, n (%) | 64 (4.2) | 61 (3.1) | |
| Generation 4 XTW, n (%) | 229 (15.0) | 304 (15.7) | |
| Procedural complications and outcomes | |||
| Acute procedural success, n (%) | 1124 (95.0) | 1467 (97.0) | 0.009 |
| Access site related complications, n (%) | 28 (2.4) | 18 (1.2) | 0.024 |
| Pulmonary complications, n (%) | 7 (0.6) | 4 (0.3) | 0.229 |
| Transesophageal echocardiography associated complications, n (%) | 12 (1.0) | 9 (0.6) | 0.271 |
Values are numbers (%) or mean±SD.
Factors Associated With the Rate of Change in Albumin Levels
Table 3 summarizes the multiple linear regression analysis results examining the factors affecting the rate of albumin change. Higher age, higher Clinical Frailty Scale, higher hemoglobin, preprocedural higher albumin levels, and lower estimated glomerular filtration rate were negatively associated with improved Δ‐albumin (all P<0.05). In addition, achieving APS was positively associated with improved Δ‐albumin (P<0.05). In a subanalysis stratified by baseline albumin levels, male sex and liver cirrhosis were negatively associated with improved Δ‐albumin in the low‐albumin group. In the high‐albumin group, as with overall, achieving APS was positively associated with improved Δ‐albumin after M‐TEER. In all groups, higher preprocedure albumin levels were significantly associated with lower rates of change.
Table 3.
Linear Regression Analysis for Δ‐Albumin in Overall, Low‐Albumin and High‐Albumin Groups
| (A) Overall | ||||
|---|---|---|---|---|
| Partial regression coefficient | Standardized regression coefficient (β) | 95% CI | P value | |
| Age (per 1‐y increase) | −0.19 | −0.13 | −0.246 to −0.133 | <0.001 |
| Male (for female) | −0.94 | −0.03 | −1.901 to 0.019 | 0.055 |
| NYHA functional class (per 1 category increase) | −0.07 | 0.00 | −0.769 to 0.620 | 0.83 |
| Body mass index (per 1 kg/m2 increase) | 0.08 | 0.02 | −0.048 to 0.215 | 0.21 |
| Acute procedural success | 2.91 | 0.04 | 0.571– to 5.240 | 0.015 |
| Clinical frailty scale (per 1 category increase) | −1.23 | −0.11 | −1.625 to −0.826 | <0.001 |
| HFH within 12 mo before enrollment | −0.56 | −0.02 | −1.589 to 0.465 | 0.28 |
| Functional MR (for degenerative MR) | −0.01 | 0.00 | −1.237 to 1.209 | 0.98 |
| Heart rate (per 1 bpm increase) | −0.01 | −0.02 | −0.044 to 0.016 | 0.36 |
| Hemoglobin (per 1 g/dL increase) | −0.48 | −0.06 | −0.760 to −0.200 | <0.001 |
| Atrial fibrillation | 0.20 | 0.01 | −0.743 to 1.137 | 0.68 |
| Liver cirrhosis | −2.59 | −0.02 | −6.127 to 0.937 | 0.15 |
| Peripheral artery disease | −0.09 | 0.00 | −1.624 to 1.443 | 0.91 |
| LVEF (per 1% increase) | 0.00 | 0.00 | −0.034 to 0.039 | 0.9 |
| eGFR (per 1 mL/min per 1.73 m2 increase) | 0.08 | 0.11 | 0.055– to 0.103 | <0.001 |
| Preprocedural albumin (per 1 g/dL increase) | −14.80 | −0.58 | −15.78 to −13.82 | <0.001 |
| (B) Low‐albumin group | ||||
|---|---|---|---|---|
| Partial regression coefficient | Standardized regression coefficient (β) | 95% CI | P value | |
| Age (per 1 y increase) | −0.22 | −0.15 | −0.308 to −0.127 | <0.001 |
| Male (for female) | −1.52 | −0.06 | −3.02 to −0.029 | 0.046 |
| NYHA functional class (per 1 category increase) | −0.34 | −0.02 | −1.419 to 0.738 | 0.54 |
| Body mass index (per 1 kg/m2 increase) | 0.14 | 0.04 | −0.07 to 0.35 | 0.19 |
| Acute procedural success | 0.88 | 0.01 | −2.84 to 4.597 | 0.64 |
| Clinical frailty scale (per 1 category increase) | −1.53 | −0.14 | −2.12 to −0.949 | <0.001 |
| HFH within 12 mo before enrollment | −0.87 | −0.03 | −2.59 to 0.856 | 0.32 |
| Functional MR (for degenerative MR) | 0.03 | 0.00 | −1.856 to 1.918 | 0.97 |
| Heart rate (per 1 bpm increase) | −0.01 | −0.01 | −0.056 to 0.035 | 0.65 |
| Hemoglobin (per 1 g/dL increase) | −0.38 | −0.05 | −0.815 to 0.05 | 0.083 |
| Atrial fibrillation | 0.17 | 0.01 | −0.131 to 1.654 | 0.82 |
| Liver cirrhosis | −6.00 | −0.05 | −11.393 to −0.598 | 0.03 |
| Peripheral artery disease | 0.28 | 0.01 | −1.914 to 2.476 | 0.8 |
| LVEF (per 1% increase) | 0.02 | 0.02 | −0.036 to 0.074 | 0.5 |
| eGFR (per 1 mL/min per 1.73 m2 increase) | 0.10 | 0.14 | 0.06– to 0.132 | <0.001 |
| Preprocedural albumin (per 1 g/dL increase) | −17.11 | −0.67 | −19.122 to −15.103 | <0.001 |
| (C) High‐albumin group | ||||
|---|---|---|---|---|
| Partial regression coefficient | Standardized regression coefficient (β) | 95% CI | P value | |
| Age (per 1 y increase) | −0.15 | −0.10 | −0.208 to −0.083 | <0.001 |
| Male (for female) | −0.20 | −0.01 | −1.273 to 0.875 | 0.72 |
| NYHA functional class (per 1 category increase) | 0.11 | 0.01 | −0.679 to 0.889 | 0.79 |
| Body mass index (per 1 kg/m2 increase) | 0.03 | 0.01 | −0.116 to 0.173 | 0.7 |
| Acute procedural success | 4.67 | 0.06 | 2.117– to 7.223 | <0.001 |
| Clinical frailty scale (per 1 category increase) | −0.63 | −0.06 | −1.132 to −0.137 | 0.012 |
| HFH within 12 mo before enrollment | −0.46 | −0.01 | −1.529 to 0.615 | 0.4 |
| Functional MR (for degenerative MR) | −0.04 | 0.00 | −1.437 to 1.355 | 0.95 |
| Heart rate (per 1 bpm increase) | −0.02 | −0.03 | −0.058 to 0.01 | 0.17 |
| Hemoglobin (per 1 g/dL increase) | −0.55 | −0.07 | −0.87 to −0.233 | <0.001 |
| Atrial fibrillation | 0.14 | 0.01 | −0.906 to 1.189 | 0.79 |
| Liver cirrhosis | 1.53 | 0.01 | −2.527 to 5.583 | 0.46 |
| Peripheral artery disease | −0.44 | −0.01 | −2.421 to 1.538 | 0.66 |
| LVEF (per 1% increase) | −0.02 | −0.02 | −0.061 to 0.023 | 0.38 |
| eGFR (per 1 mL/min per 1.73 m2 increase) | 0.06 | 0.08 | 0.03– to 0.087 | <0.001 |
| Pre‐procedural albumin (per 1 g/dL increase) | −10.81 | −0.42 | −12.79 to −8.84 | <0.001 |
(A) Adjusted R2=0.316. (B) Adjusted R2=0.211. (C) Adjusted R2=0.141. Δ‐albumin (%)=([level at 1 month after mitral transcatheter edge‐to‐edge repair—preprocedural level]/preprocedural level)×100. eGFR indicates estimated glomerular filtration rate HFH, heart failure hospitalization; LVEF, left ventricular ejection fraction; MR, mitral regurgitation; and NYHA, New York Heart Association.
Association Between Δ‐Albumin and All‐Cause Mortality
Figure S1 illustrates the distribution of Δ‐albumin. Among the 2695 patients, the median change rate of albumin was +2.9% (interquartile range, −3.5% to +11.4%; minimum, −84.7%; and maximum, +90.9%). The low‐albumin group had a higher distribution of Δ‐albumin compared with the high‐albumin group. The mean follow‐up period was 631.9±444.2 days for the entire cohort, during which 595 patients (22.1%) died. Kaplan–Meier curves with log‐rank tests suggested a significantly higher all‐cause mortality in the worsening Δ‐albumin group across the entire cohort and within subgroups stratified by the baseline median albumin level (Figure 1A through 1C). Table 4 summarizes results of the Cox regression analysis. Univariate Cox regression analysis indicated that Δ‐albumin level improvement was significantly associated with reduced mortality risk (HR, 0.95 per 5% increment [95% CI, 0.92–0.98], and P=0.004). This association persisted even after adjusting for confounding factors (HR, 0.88 per 5% increment [95% CI, 0.85–0.91], P<0.001). The results were consistent across the low‐ (HR, 0.89 per 5% increment [95% CI, 0.86–0.92], P<0.001) and high‐albumin (HR, 0.87 per 5% increment [95% CI, 0.80–0.94] P<0.001) groups. The proportional hazards assumption for all‐cause mortality outcomes was confirmed to be valid for the overall, low‐albumin, and high‐albumin groups. Restricted cubic spline analysis indicated a nonlinear association between all‐cause mortality and Δ‐albumin as a continuous variable (Figure S2). Improved Δ‐albumin was consistently associated with better survival in the various subgroups, even after adjusting for confounders in the multivariate analysis (Figure 2).
Figure 1. Kaplan–Meier curve for all‐cause mortality based on Δ‐albumin.
Kaplan–Meier curve showing all‐cause mortality in the overall cohort (A), low‐albumin group (B), and high‐albumin group (C).
Table 4.
Univariate and Multivariate Cox Regression Analyses for All‐Cause Mortality
| Overall | Low‐albumin group | High‐albumin group | |||||||
|---|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | HR | 95% CI | P value | |
| Univariate analysis | |||||||||
| Δ‐albumin (per 5% increase) | 0.95 | 0.92–0.98 | 0.004 | 0.88 | 0.85–0.92 | <0.001 | 0.87 | 0.80–0.95 | 0.001 |
| Δ‐albumin improve | 0.85 | 0.72–1.00 | 0.048 | 0.56 | 0.46–0.69 | <0.001 | 0.70 | 0.51–0.95 | 0.022 |
| Multivariate analysis | |||||||||
| Δ‐albumin (per 5% increase) | 0.88 | 0.85–0.91 | <0.001 | 0.89 | 0.86–0.92 | <0.001 | 0.87 | 0.80–0.94 | <0.001 |
| Δ‐albumin improve | 0.62 | 0.52–0.75 | <0.001 | 0.60 | 0.49–0.74 | <0.001 | 0.68 | 0.49–0.94 | 0.019 |
Multivariable models were adjusted for age, sex, prior heart failure hospitalization within 12 months before enrollment, New York Heart Association functional class, body mass index, acute procedural success, heart rate, hemoglobin, atrial fibrillation/flutter, estimated glomerular filtration rate, Clinical Frailty Scale, left ventricular ejection fraction, peripheral artery disease, Functional mitral regurgitation, liver cirrhosis, baseline value<median (overall only). HR indicates hazard ratio.
Figure 2. Δ‐albumin improvement and clinical outcomes among subgroups.
CFS indicates Clinical Frailty Scale; DMR, degenerative mitral regurgitation; eGFR, estimated glomerular filtration rate; FMR, functional mitral regurgitation; HR, hazard ratio; MR, mitral regurgitation; and NYHA, New York Heart Association.
Association Between Δ‐Albumin and HFH
During the follow‐up, HFH occurred in 566 (21.0%) patients. Figure 3 illustrates Kaplan–Meier survival curves of competing risk for HFH compared with Δ‐albumin. Improved Δ‐albumin did not differ in HFH regardless of pre‐M‐TEER albumin levels (log‐rank test, P=0.407). Univariable Fine and Gray regression analysis of the association between Δ‐albumin improvement and competing risk for HFH (HR, 0.98 per 5% increment [95% CI, 0.94–1.01], P=0.14). However, after adjusting for confounding factors, including baseline albumin level in the multivariate analysis, Δ‐albumin improvement was associated with lower competing risk of HFH (HR, 0.94 per 5% increment [95% CI, 0.91–0.97], P<0.001). The results were consistent across the low‐ (HR, 0.95 per 5% increment [95% CI, 0.91–0.99], P=0.011) and high‐albumin (HR, 0.91 per 5% increment [95% CI, 0.84–0.98], P=0.015) groups (Table 5).
Figure 3. Kaplan–Meier curve for competing risk for HFH based on Δ‐albumin.
Kaplan–Meier curve showing competing risk for HFH in the overall cohort (A), low‐albumin group (B), and high‐albumin group (C). HFH indicates heart rate hospitalization.
Table 5.
Fine and Gray Competing Risk Regression Analysis for HFH
| Overall | Low‐albumin group | High‐albumin group | |||||||
|---|---|---|---|---|---|---|---|---|---|
| HR | 95% CI | P value | HR | 95% CI | P value | HR | 95% CI | P value | |
| Univariate analysis | |||||||||
| Δ‐albumin (per 5% increase) | 0.98 | 0.94–1.01 | 0.14 | 0.95 | 0.91–0.98 | 0.005 | 0.93 | 0.86–1.01 | 0.078 |
| Δ‐albumin improve | 0.94 | 0.79–1.1 | 0.43 | 0.82 | 0.65–1.03 | 0.09 | 0.78 | 0.59–1.04 | 0.093 |
| Multivariate analysis | |||||||||
| Δ‐albumin (per 5% increase) | 0.94 | 0.91–0.97 | <0.001 | 0.95 | 0.91–0.99 | 0.011 | 0.91 | 0.84–0.98 | 0.015 |
| Δ‐albumin improve | 0.78 | 0.64–0.94 | 0.01 | 0.83 | 0.65–1.07 | 0.15 | 0.67 | 0.49–0.92 | 0.014 |
Multivariable models were adjusted for age, sex, prior heart failure hospitalization within 12 months before enrollment, New York Heart Association functional class, body mass index, acute procedural success, hemoglobin, atrial fibrillation/flutter, estimated glomerular filtration rate, Clinical Frailty Scale, left ventricular ejection fraction, peripheral artery disease, functional mitral regurgitation, chronic obstructive pulmonary disease, baseline value < median (overall only). HFH indicates hospitalizations for heart failure; and HR, hazard ratio.
DISCUSSION
This study has 3 important findings. First, serum albumin improved in 56.1% of patients after 1 month of M‐TEER. The characteristics of patients who did not improve were old age, higher frailty, higher hemoglobin, lower estimated glomerular filtration rate, and higher preprocedural albumin levels. Particularly, albumin improvement after M‐TEER was significantly greater in patients who achieved APS. Second, regardless of baseline albumin level, Δ‐albumin improvement was significantly associated with a reduced risk of all‐cause mortality after M‐TEER. This trend was also consistently observed across various subgroups, including functional MR and degenerative MR, even after adjusting for confounding factors in multivariate analysis. Third, improved Δ‐albumin was associated with a lower competing risk of HFH after M‐TEER. Preprocedural low albumin status is a poor prognostic and unmodifiable factor. Nonetheless, regardless of baseline albumin level, Δ‐albumin facilitated the identification of a population in which M‐TEER reduced all‐cause mortality and HFH in response to invasive treatment.
In this study, the factors that improved the postoperative albumin levels after M‐TEER were clinically convincing. Old age and low baseline albumin levels have been associated with albumin changes from admission to discharge in patients with acute decompensated HF, consistent with our results. 13 Physicians can share preprocedural information about concerns that older adults, patients with high frailty, or those with impaired renal function may experience less postoperative albumin recovery. A higher preprocedural albumin level has been associated with less improvement in albumin level, indicating that individuals with higher albumin levels have less room for improvement. Conversely, the expected improvement in albumin levels in the low‐albumin group suggested that low albumin levels alone do not preclude M‐TEER as a treatment indication. Notably, APS was an independent predictive factor for albumin improvement after M‐TEER in the high‐albumin group. The likelihood of achieving APS should be carefully considered, particularly when treating individuals with relatively better prognosis and high albumin levels after M‐TEER. Conventionally, serum albumin has been used as a nutritional indicator; however, the American Society for Parenteral and Enteral Nutrition reported that inflammation affects albumin levels. 19 , 20 If albumin indicates inflammatory changes associated with HF before and after M‐TEER, albumin changes may reflect the success or failure of M‐TEER and the degree of disease improvement, supporting our results.
The association between Δ‐albumin and mortality risk after M‐TEER is consistent with previous findings on HF 11 , 12 , 13 ; however, the clinical interpretation of albumin changes may vary among diseases. Evidence regarding the association between improved Δ‐albumin and long‐term prognosis after M‐TEER remains unclear; nonetheless, analysis of the OCEAN‐Mitral registry demonstrated that improved Δ‐albumin 1 month after M‐TEER was associated with a lower risk of death. This trend was similar regardless of the preprocedure albumin level. Albumin serves as an important indicator for determining indications for M‐TEER, with lower levels before M‐TEER being associated with worse prognosis. 4 In this study, increased Δ‐albumin was associated with prognosis independently of other factors, regardless of preprocedure levels. It serves as a single index for determining the efficacy of M‐TEER and its indication, providing evidence for the usefulness of baseline or dynamic changes in albumin levels.
The association between Δ‐albumin and competing risk of HFH after M‐TEER is unclear, compared with its association with all‐cause mortality. Albumin was not significantly associated with HFH in a subanalysis of the COAPT trial, 4 and its usefulness as a predictor of HFH remains debatable. Whereas the current results should be carefully interpreted, improvement in Δ‐albumin was significantly associated with a lower competing risk of HFH. Multivariate analysis demonstrated that improvement in albumin levels significantly reduced HFH in both preoperative low‐ and high‐ albumin group. From the standpoint of HFH, the improvement in albumin levels appears more significant in the preoperative low‐albumin group. Dividing patients into high‐ and low‐albumin groups before M‐TEER may have affected its usefulness as an indicator of HFH risk.
Serum albumin is easy to measure and practical. Furthermore, it reflects the nutritional status and systemic health status, including inflammation, serving as an important marker for therapeutic decision‐making. Additionally, albumin changes with treatment and cardiac and renal function, suggesting its potential as a therapeutic target for patients undergoing M‐TEER.
Study Limitations
This study has several limitations. First, this study was a retrospective, non‐randomized, unblinded observational study conducted in a single‐ethnic population (Japanese). As such, it is inherently subject to common biases associated with this type of study design, and similar limitations have been reported in previous cohort studies. In addition, preprocedural albumin levels may have been affected by mean reversion in the rate of change before and after M‐TEER. Second, procedural and postprocedural echocardiographic parameters were not assessed in a central core laboratory. However, we shared a consensus document based on established guidelines with the participating institutions to minimize knowledge and technical gaps before registry enrollment. Third, the multivariate analysis did not include all unpredictable factors, despite including several important clinical variables. Fourth, only the first HFH was assessed in this study, as data on recurrent events were not reliably available due to the retrospective design. The impact of repeated hospitalizations warrants further investigation in future prospective studies. Finally, preprocedural and follow‐up laboratory data for serum albumin were not available for some patients, and missing data might have affected the results. Approximately 30% of patients had missing postprocedural albumin values, and these patients were found to have poorer baseline characteristics (Table S2). The absence of albumin measurements was attributed not only to early mortality but also to complications or transfer to other hospitals, suggesting that these high‐risk patients may be excluded from this study. Nevertheless, the clinical implication of this study lies in demonstrating albumin changes in predicting the long‐term prognosis of patients undergoing M‐TEER using a larger cohort.
CONCLUSIONS
Our study is the first to demonstrate the efficacy of early albumin changes for predicting the long‐term prognosis of patients undergoing M‐TEER. Serum albumin is easy to measure and serves as an excellent predictor of long‐term mortality risk. Even in high‐risk patients with low preprocedure albumin levels, improved Δ‐albumin is significantly associated with both long‐term mortality and competing risk of HFH. For patients with high preprocedure albumin levels, improved Δ‐albumin remains associated with mortality and competing risk of HFH, suggesting the importance of APS achievement. This is because improvement in Δ‐albumin after M‐TEER indicates procedural success. Improved Δ‐albumin is necessary for predicting prognosis after M‐TEER and may serve as a measure to determine the procedural benefits of M‐TEER.
Sources of Funding
The OCEAN‐Mitral registry, which is part of the OCEAN‐SHD (Optimized Catheter Valvular Intervention–Structural Heart Disease) registry, was supported by Edwards Lifesciences, Medtronic Japan, Boston Scientific Japan, Abbott Medical, and the Daiichi‐Sankyo Company.
Disclosures
Drs Yamamoto, Kagase, Kubo, Saji, Izumi, Asami, Izumo, Watanabe, Ohno, Hachinohe, Enta, Shirai, Mizuno, Boda, Kodama, Amaki, and Hayashida are clinical proctors of transcatheter edge‐to‐edge repair for Abbott Medical and have received lecture/consultant fees from Abbott Medical. Dr J. Yamaguchi is a clinical proctor of transcatheter edge‐to‐edge repair for Abbott Medical and has received a lecture fee and a scholarship donation from Abbott Medical. The remaining authors have no disclosures to report.
Supporting information
Tables S1–S2
Figures S1–S2
Acknowledgments
The authors thank all OCEAN‐Mitral investigators who participated in this registry.
This article was sent to Amgad Mentias, MD, Associate Editor, for review by expert referees, editorial decision, and final disposition.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.125.041961
For Sources of Funding and Disclosures, see page 10 and 11.
Contributor Information
Kenichi Shibata, Email: going_steady0125@hotmail.com.
Masanori Yamamoto, Email: masa-nori@nms.ac.jp.
References
- 1. Stone GW, Lindenfeld J, Abraham WT, Kar S, Lim DS, Mishell JM, Whisenant B, Grayburn PA, Rinaldi M, Kapadia SR, et al. Transcatheter mitral‐valve repair in patients with heart failure. N Engl J Med. 2018;379:2307–2318. doi: 10.1056/NEJMoa1806640 [DOI] [PubMed] [Google Scholar]
- 2. Feldman T, Kar S, Elmariah S, Smart SC, Trento A, Siegel RJ, Apruzzese P, Fail P, Rinaldi MJ, Smalling RW, et al. Randomized comparison of percutaneous repair and surgery for mitral regurgitation: 5‐year results of EVEREST II. J Am Coll Cardiol. 2015;66:2844–2854. doi: 10.1016/j.jacc.2015.10.018 [DOI] [PubMed] [Google Scholar]
- 3. Tokuda T, Yamamoto M, Kagase A, Shimura T, Yamaguchi R, Saji M, Asami M, Enta Y, Nakashima M, Shirai S, et al. Clinical impact of baseline frailty status and residual mitral regurgitation after transcatheter edge‐to‐edge repair: insights from the OCEAN‐mitral registry. J Am Heart Assoc. 2024;13:e035109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Feng KY, Ambrosy AP, Zhou Z, Li D, Kong J, Zaroff JG, Mishell JM, Ku IA, Scotti A, Coisne A, et al. Association between serum albumin and outcomes in heart failure and secondary mitral regurgitation: the COAPT trial. Eur J Heart Fail. 2023;25:553–561. doi: 10.1002/ejhf.2809 [DOI] [PubMed] [Google Scholar]
- 5. Scotti A, Coisne A, Granada JF, Driggin E, Madhavan MV, Zhou Z, Redfors B, Kar S, Lim DS, Cohen DJ, et al. Impact of malnutrition in patients with heart failure and secondary mitral regurgitation: the COAPT trial. J Am Coll Cardiol. 2023;82:128–138. doi: 10.1016/j.jacc.2023.04.047 [DOI] [PubMed] [Google Scholar]
- 6. Pezzola E, Tomasoni D, Caneiro‐Queija B, Raposeiras‐Roubin S, Freixa X, Arzamendi D, Benito‐González T, Montefusco A, Pascual I, Nombela‐Franco L, et al. Prevalence and prognostic significance of malnutrition in patients with secondary mitral regurgitation undergoing transcatheter edge‐to‐edge repair. Catheter Cardiovasc Interv. 2024;104:390–400. doi: 10.1002/ccd.31079 [DOI] [PubMed] [Google Scholar]
- 7. Picca A, Coelho‐Junior HJ, Calvani R, Marzetti E, Vetrano DL. Biomarkers shared by frailty and sarcopenia in older adults: a systematic review and meta‐analysis. Ageing Res Rev. 2022;73:101530. doi: 10.1016/j.arr.2021.101530 [DOI] [PubMed] [Google Scholar]
- 8. Gotsman I, Shauer A, Zwas DR, Tahiroglu I, Lotan C, Keren A. Low serum albumin: a significant predictor of reduced survival in patients with chronic heart failure. Clin Cardiol. 2019;42:365–372. doi: 10.1002/clc.23153 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. El Iskandarani M, El Kurdi B, Murtaza G, Paul TK, Refaat MM. Prognostic role of albumin level in heart failure: a systematic review and meta‐analysis. Medicine (Baltimore). 2021;100:e24785. doi: 10.1097/MD.0000000000024785 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Yamamoto M, Shimura T, Kano S, Kagase A, Kodama A, Sago M, Tsunaki T, Koyama Y, Tada N, Yamanaka F, et al. Prognostic value of hypoalbuminemia after transcatheter aortic valve implantation (from the Japanese multicenter OCEAN‐TAVI registry). Am J Cardiol. 2017;119:770–777. doi: 10.1016/j.amjcard.2016.11.019 [DOI] [PubMed] [Google Scholar]
- 11. Jabbour R, Ling HZ, Norrington K, Amaral N, Zaman N, Aggarwal S, Aung N, Barron A, Manisty C, Baruah R, et al. Serum albumin changes and multivariate dynamic risk modelling in chronic heart failure. Int J Cardiol. 2014;176:437–443. doi: 10.1016/j.ijcard.2014.07.096 [DOI] [PubMed] [Google Scholar]
- 12. Nakayama H, Koyama S, Kuragaichi T, Shiba M, Fujiwara H, Takatsu Y, Sato Y. Prognostic value of rising serum albumin during hospitalization in patients with acute heart failure. Am J Cardiol. 2016;117:1305–1309. doi: 10.1016/j.amjcard.2016.01.030 [DOI] [PubMed] [Google Scholar]
- 13. Kato T, Yaku H, Morimoto T, Inuzuka Y, Tamaki Y, Ozasa N, Yamamoto E, Yoshikawa Y, Kitai T, Taniguchi R, et al. Association of an increase in serum albumin levels with positive 1‐year outcomes in acute decompensated heart failure: a cohort study. PLoS One. 2020;15:e0243818. doi: 10.1371/journal.pone.0243818 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Saji M, Yamamoto M, Kubo S, Asami M, Enta Y, Shirai S, Izumo M, Mizuno S, Watanabe Y, Amaki M, et al. Short‐term outcomes following Transcatheter edge‐to‐edge repair: insights from the OCEAN‐Mitral registry. JACC: Asia. 2023;3:766–773. doi: 10.1016/j.jacasi.2023.06.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Kubo S, Yamamoto M, Saji M, Asami M, Enta Y, Nakashima M, Shirai S, Izumo M, Mizuno S, Watanabe Y, et al. One‐year outcomes and their relationship to residual mitral regurgitation after Transcatheter edge‐to‐edge repair with MitraClip device: insights from the OCEAN‐Mitral registry. J Am Heart Assoc. 2023;12:e030747. doi: 10.1161/JAHA.123.030747 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Zoghbi WA, Enriquez‐Sarano M, Foster E, Grayburn PA, Kraft CD, Levine RA, Nihoyannopoulos P, Otto CM, Quinones MA, Rakowski H, et al. Recommendations for evaluation of the severity of native valvular regurgitation with two‐dimensional and Doppler echocardiography. J Am Soc Echocardiogr. 2003;16:777–802. doi: 10.1016/S0894-7317(03)00335-3 [DOI] [PubMed] [Google Scholar]
- 17. Foster E, Wasserman HS, Gray W, Homma S, Di Tullio MR, Rodriguez L, Stewart WJ, Whitlow P, Block P, Martin R, et al. Quantitative assessment of severity of mitral regurgitation by serial echocardiography in a multicenter clinical trial of percutaneous mitral valve repair. Am J Cardiol. 2007;100:1577–1583. doi: 10.1016/j.amjcard.2007.06.066 [DOI] [PubMed] [Google Scholar]
- 18. Kanda Y. Investigation of the freely available easy‐to‐use software “EZR” for medical statistics. Bone Marrow Transplant. 2013;48:452–458. doi: 10.1038/bmt.2012.244 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Evans DC, Corkins MR, Malone A, Miller S, Mogensen KM, Guenter P, ASPEN Malnutrition Committee , Jensen GL. The use of visceral proteins as nutrition markers: an ASPEN position paper. Nutr Clin Pract. 2021;36:22–28. doi: 10.1002/ncp.10588 [DOI] [PubMed] [Google Scholar]
- 20. Soeters PB, Wolfe RR, Shenkin A. Hypoalbuminemia: pathogenesis and clinical significance. JPEN J Parenter Enteral Nutr. 2019;43:181–193. doi: 10.1002/jpen.1451 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Tables S1–S2
Figures S1–S2