Effect of Preoperative Left Ventricular Mass on Outcomes After Aortic Valve Replacement for Aortic Regurgitation

Abstract

Background: We determined the left ventricular mass index (LVMI) cut-off value for the risk of major adverse cardiac and cerebrovascular events (MACCE) in patients undergoing aortic valve replacement (AVR) for aortic regurgitation (AR) and investigated the effect of preoperative left ventricular remodeling on long-term outcomes postoperatively.

Methods and Results: Of the 1,580 patients who underwent surgical AVR at Shiga University of Medical Science between January 2002 and December 2022, we retrospectively analyzed data for 263 patients who underwent surgery for AR. The receiver operating characteristic curve showed that the cut-off value of preoperative LVMI for the incidence of MACCE was 200 g/m² (area under the curve=0.692). We compared postoperative outcomes between patients with preoperative LVMI >200 g/m² (n=92) and those with preoperative LVMI ≤200 g/m² (n=171) after adjusting for preoperative characteristics using inverse probability of treatment weighting. The mean (±SD) follow-up period was 6.9±5.1 years. The rate of MACCE at 10 years was significantly higher in patients with preoperative LVMI >200 g/m² than in those with preoperative LVMI ≤200 g/m² (25.6% vs. 13.5%; P=0.020). In multivariable Cox models, preoperative LVMI >200 g/m² was significantly associated with a higher risk of MACCE (hazard ratio 2.356, P=0.006).

Conclusions: Preoperative LVMI >200 g/m² was associated with a higher rate of MACCE in patients undergoing AVR for AR.

Chronic aortic regurgitation (AR) is a progressive disease that induces volume overload in the left ventricle (LV), leading to increased wall stress and LV mass (LVM). Aortic valve replacement (AVR) is an effective treatment for AR that can achieve postoperative LV reverse remodeling. However, extensive remodeling due to chronic AR is associated with worse outcomes, even after AVR.¹ In the present study, we determined the preoperative LVM index (LVMI) cut-off value for major adverse cardiac and cerebrovascular events (MACCE) in patients undergoing AVR for AR, and investigated the effect of preoperative LV remodeling on long-term outcomes postoperatively.

Methods

All patients had previously provided permission for their medical records to be used for research purposes. The study was approved the Institutional Review Board of Shiga University of Medical Science (Registration no. 2023-090; approval date: November 2, 2023). The procedures in this study were conducted in accordance with the tenets of the Declaration of Helsinki.

Between January 2002 and December 2022, 1,580 patients underwent surgical AVR at Shiga University of Medical Science. Of these 1,580 patients, 485 underwent surgery for AR. However, we excluded 222 patients who underwent concomitant surgery on other valves or concomitant coronary artery bypass grafting or who had infective endocarditis, leaving 263 patients in the present study. Of these patients, 77 (29.3%) were asymptomatic.

Surgical Treatment

Anesthesia was maintained in the standard manner. One (0.4%) patient underwent minimally invasive cardiac surgery via a right minithoracotomy, and the remaining patients (99.6%) underwent median sternotomy. Myocardial protection was provided using antegrade or retrograde cold blood cardioplegia. Valves were selected according to each surgeon’s preference and were implanted in the supra-annular position or in the intra-annular position, also according to the preference of individual surgeons. After implantation of the AVR, the aortotomy was sutured with a 4–0 monofilament horizontal mattress suture and continuous suture, or with a 4–0 monofilament continuous suture in 2 layers.

Echocardiographic Measurements

All echocardiographic examinations were performed by experienced echocardiographers. Patients underwent annual echocardiographic follow-up at Shiga University of Medical Science. LV dimensions were assessed using 2-dimensional M-mode tracings. The LV ejection fraction was calculated using the Simpson method. Measurements of LV end-diastolic diameter (LVEDD), interventricular septal thickness at end-diastole (IVSD), and posterior wall thickness at end-diastole (PWD), in millimeters, were used to calculate LVM using the formula described by Devereux et al.:²

LVM = 0.8(1.04[{LVEDD + IVSD + PWD}³ − LVEDD³]) + 0.6

The LVMI was calculated by dividing LVM by body surface area.

Outcomes Measures and Definitions

The primary outcome was the incidence of MACCE. The secondary outcome was LVMI regression during the follow-up period. MACCE were defined as a composite of all-cause death, non-fatal myocardial infarction, and non-fatal heart failure, non-fatal stroke, and repeated revascularization. Non-fatal myocardial infarction, non-fatal heart failure, and non-fatal stroke were defined as new admissions with a diagnosis of these diseases during the follow-up period that did not result in death.

Statistical Analysis

Continuous variables are presented as the mean±SD and categorical variables are presented as number and percentages. The Kolmogorov-Smirnov test and Shapiro-Wilk test were used to test the normality of data distribution. Comparisons of clinical characteristics between groups were performed using unpaired t-tests for normally distributed variables and the Mann-Whitney U test for skewed variables. Categorical variables were analyzed using Pearson’s χ² test. Probabilities of survival were estimated using the Kaplan-Meier method, for which the patients’ survival time was measured from the date of surgery until death or the final date of follow-up. Groups were compared using the long-rank test. Univariable and multivariable Cox proportional hazards regression analyses were performed to analyze predictors of MACCE. Variables reaching P<0.100 in the univariable analysis and those that were considered clinically important were entered into the multivariable model. All statistical testing was 2-sided, and results were considered statistically significant at P<0.050.

The area under curve (AUC) was calculated by obtaining the receiver operating characteristic (ROC) curve from the logistic regression model for the incidence of MACCE. The cut-off value with the most favorable sensitivity and specificity was determined from the ROC curve using the Youden index.

To reduce the effect of selection bias and potential confounding factors, we adjusted patients’ baseline characteristics using weighted logistic regression analysis and inverse probability of treatment weighting (IPTW). Weights for patients with preoperative LVMI >200 g/m² were the inverse of the propensity scores, and weights for patients with preoperative LVMI ≤200 g/m² were the inverse of 1−propensity score. The following 20 adjustment variables were used to derive the propensity score: age, sex, body mass index, hypertension, diabetes, dyslipidemia, smoking history, history of percutaneous coronary intervention, previous cerebrovascular accident, estimated glomerular filtration rate <60 mL/min/1.73 m², preoperative hemodialysis, redo surgery, emergency operation, bicuspid aortic valve, annuloaortic ectasia, New York Heart Association functional class ≥III, preoperative statin use, preoperative β-blocker use, preoperative renin-angiotensin system inhibitor use, and preoperative angiotensin receptor-neprilysin inhibitor (ARNI) use. The model was well calibrated (Hosmer-Lemeshow test, P=0.873), with reasonable discrimination (C-statistic=0.709). Absolute standardized mean differences were calculated to compare the balance in baseline characteristics between the 2 groups in the unweighted and weighted cohorts. An absolute standardized mean difference of >0.100 was considered a meaningful imbalance.³

All statistical analyses were performed using SPSS version 29.0 (IBM Corp., Armonk, NY, USA) and SAS version 9.4 (SAS Institute, Cary, NC, USA).

Results

ROC analysis showed that the preoperative LVMI cut-off value for the incidence of MACCE was 200 g/m² (AUC=0.651; Figure 1). In the present study, 92 patients had preoperative LVMI >200 g/m² and 171 patients had preoperative LVMI ≤200 g/m².

Figure 1.

Receiver operating characteristic curve from the logistic regression model for the incidence of major adverse cardiac and cerebrovascular events. AUC, area under the curve; CI, confidence interval.

Preoperative patient characteristics are presented in Table 1. In the unweighted cohort, the proportion of men was higher in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group (76.1% vs. 63.2%, respectively; P=0.027). More patients in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group had dyslipidemia (24.6% vs. 12.0%, respectively; P=0.008) and were taking a statin preoperatively (14.0% vs. 3.3%, respectively; P=0.001). In the weighted cohort, the 2 groups were well balanced in their baseline characteristics.

Table 1.

Preoperative Patient Characteristics

	Unweighted				Weighted
	LVMI >200 g/m2 (n=92)	LVMI ≤200 g/m2 (n=171)	P value	ASMD	LVMI >200 g/m2 (SoW=267.19)	LVMI ≤200 g/m2 (SoW=261.30)	P value	ASMD
Age (year)	63.8±14.0	64.5±13.6	0.691	0.051	64.0±14.9	64.5±13.6	0.681	0.035
Male sex	70 (76.1)	108 (63.2)	0.027	0.283	178.11 (66.7)	176.00 (67.4)	0.865	0.015
BMI (kg/m2)	22.8±3.5	22.5±3.5	0.530	0.083	22.6±3.4	22.6±3.5	0.861	0.014
BSA (m2)	1.64±0.22	1.62±0.22	0.368	0.117	1.61±0.21	1.63±0.22	0.313	0.093
Hypertension	59 (64.1)	102 (59.6)	0.479	0.093	162.06 (60.7)	162.00 (62.0)	0.752	0.027
Diabetes	7 (7.6)	19 (11.1)	0.366	0.120	35.48 (13.3)	26.85 (10.3)	0.285	0.093
Dyslipidemia	11 (12.0)	42 (24.6)	0.008	0.330	61.52 (23.0)	52.44 (20.1)	0.410	0.071
Smoking history	41 (44.6)	79 (46.2)	0.801	0.032	125.46 (47.0)	121.74 (46.6)	0.933	0.008
Previous PCI	3 (3.3)	3 (1.8)	0.437	0.095	4.21 (1.6)	4.65 (1.8)	0.855	0.015
Previous CVD	12 (13.0)	12 (7.0)	0.138	0.201	29.38 (11.0)	25.51 (9.8)	0.643	0.039
eGFR <60 mL/min/1.73 m2	40 (43.5)	68 (39.8)	0.561	0.075	112.56 (42.1)	107.63 (41.2)	0.827	0.018
Hemodialysis	2 (2.2)	3 (1.8)	0.813	0.029	9.66 (3.6)	5.23 (2.0)	0.262	0.097
Redo surgery	5 (5.4)	18 (10.5)	0.129	0.189	26.90 (10.1)	22.51 (8.6)	0.567	0.052
Emergency operation	2 (2.2)	4 (2.3)	0.932	0.007	11.59 (4.3)	7.06 (2.7)	0.308	0.087
Bicuspid aortic valve	19 (20.7)	26 (15.2)	0.283	0.144	41.03 (15.4)	42.33 (16.2)	0.791	0.022
Annuloaortic ectasia	8 (8.7)	8 (4.7)	0.235	0.161	16.24 (6.1)	17.35 (6.6)	0.792	0.021
NYHA class ≥III	28 (30.4)	42 (24.6)	0.316	0.130	65.09 (24.4)	68.08 (26.1)	0.655	0.039
Preoperative drugs
Statin	3 (3.3)	24 (14.0)	0.001	0.388	24.25 (9.1)	27.34 (10.5)	0.592	0.047
β-blocker	29 (31.5)	43 (25.1)	0.281	0.142	80.48 (30.1)	71.79 (27.5)	0.503	0.057
RAS inhibitor	53 (57.6)	83 (48.5)	0.162	0.183	133.47 (50.0)	132.14 (50.6)	0.887	0.012
ARNI	0 (0)	1 (0.6)	0.464	0.110	0 (0)	1.00 (0.4)	0.318	0.090
Echocardiographic data
LVEF (%)	51.8±9.9	57.4±9.5	<0.001	0.577	51.5±10.1	57.2±9.7	<0.001	0.576
LVEDD (mm)	67.7±7.1	57.5±6.6	<0.001	1.488	67.0±6.6	57.8±6.6	<0.001	1.389
LVESD (mm)	49.1±7.4	39.5±6.3	<0.001	1.397	48.7±6.9	39.7±6.4	<0.001	1.352
LVMI (g/m2)	245±38	155±28	<0.001	2.696	246±37	156±28	<0.001	2.743
Mild AS	7 (7.6)	6 (3.5)	0.191	0.180	13.52 (5.1)	8.69 (3.3)	0.320	0.090
Mild MR	39 (42.4)	60 (35.1)	0.251	0.150	134.38 (50.3)	95.05 (36.4)	0.001	0.283

Unless indicated otherwise, data are given as the mean±SD or n (%). ARNI, angiotensin receptor-neprilysin inhibitor; AS, aortic stenosis; ASMD, absolute standardized mean difference; BMI, body mass index; BSA, body surface area; CVD, cerebrovascular disease; eGFR, estimated glomerular filtration rate; LVEDD, left ventricular end-diastolic diameter; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic diameter; LVMI, left ventricular mass index; LVMI, left ventricular mass index; MR, mitral regurgitation; NYHA, New York Heart Association; PCI, percutaneous coronary intervention; RAS, renin-angiotensin system; SoW, sum of weights.

Before and after adjustment using IPTW, LVEDD, LV end-systolic diameter (LVESD), and LVMI were greater in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group. LV ejection fraction was lower in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group.

Early Outcomes

Operative and postoperative outcomes are presented in Table 2. Before and after adjustment using IPTW, the implanted valve size was significantly larger in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group. Postoperative echocardiography revealed 3 (1.8%) patients in the LVMI ≤200 g/m² group with a postoperative effective orifice area index <0.85 cm²/m², and 1 (1.1%) patient in the LVMI >200 g/m² group with a mild transvalvular leak after mechanical valve implantation. No patient in either group had a postoperative paravalvular leak. There were no significant differences in postoperative drugs between the 2 groups.

Table 2.

Operative and Postoperative Data

	Unweighted			Weighted
	LVMI >200 g/m2 (n=92)	LVMI ≤200 g/m2 (n=171)	P value	LVMI >200 g/m2 (SoW=267.19)	LVMI ≤200 g/m2 (SoW=261.30)	P value
Operative data
Operation time (min)	210±61	206±62	0.623	205±57	204±60	0.900
Cardiopulmonary bypass time (min)	106±36	105±38	0.794	103±33	104±36	0.808
Aortic clamp time (min)	69±27	67±28	0.708	66±24	67±27	0.750
Bioprosthetic valve	62 (67.4)	115 (67.3)	0.982	183.84 (68.8)	174.74 (66.9)	0.635
Valve size (mm)	25.5±1.8	24.4±2.0	<0.001	25.3±1.9	24.5±1.9	<0.001
Concomitant procedures
Bentall	12 (13.0)	20 (11.7)	0.751	28.23 (10.6)	33.05 (12.6)	0.456
Hemiarch	11 (12.0)	23 (13.5)	0.732	24.76 (9.3)	33.94 (13.0)	0.175
Total arch replacement	3 (3.3)	13 (7.6)	0.117	9.83 (3.7)	17.15 (6.6)	0.134
Postoperative data
ICU stay >48 h	3 (3.3)	5 (2.9)	0.880	5.29 (2.0)	6.37 (2.4)	0.721
Ventilation >48 h	3 (3.3)	4 (2.3)	0.659	5.29 (2.0)	5.03 (1.9)	0.964
30-day mortality	1 (1.1)	0 (0)	0.320	1.63 (0.6)	0 (0)	0.202
Hospital mortality	2 (2.2)	0 (0)	0.158	2.94 (1.1)	0 (0)	0.086
Postoperative drugs
Statin	19 (20.7)	45 (26.3)	0.298	68.96 (25.8)	59.49 (22.8)	0.416
β-blocker	57 (62.0)	110 (64.3)	0.705	158.50 (59.3)	165.24 (63.2)	0.356
RAS inhibitor	36 (39.1)	53 (31.0)	0.193	103.73 (38.8)	86.38 (33.1)	0.168
ARNI	0 (0)	0 (0)	–	0 (0)	0 (0)	–
Echocardiographic data
EOAI (cm2/m2)	1.30±0.31	1.28±0.27	0.583	1.30±0.29	1.28±0.28	0.514
LVEF (%)	46.6±10.6	52.6±8.9	<0.001	46.1±9.3	52.4±8.8	<0.001
LVEDD (mm)	58.6±8.8	50.6±6.5	<0.001	57.6±7.9	51.0±6.4	<0.001
LVESD (mm)	44.8±9.2	37.0±6.7	<0.001	44.3±8.0	37.4±6.5	<0.001
LVMI (g/m2)	193±44	136±34	<0.001	193±42	137±34	<0.001
Mild MR	13 (14.1)	18 (10.5)	0.389	43.12 (16.1)	28.21 (10.8)	0.072

Unless indicated otherwise, data are given as the mean±SD or n (%). EOAI, effective orifice area index; ICU, intensive care unit; LV, left ventricular. Other abbreviations as in Table 1.

Long-Term Outcomes

The mean follow-up was 6.9±5.1 years (maximum: 20.8 years). There were 31 patients who experienced MACCE 10 years after surgery and 64 patients who could be followed up at that time. In the unweighted cohort, the rates of MACCE at 10 years in the LVMI >200 g/m² and LVMI ≤200 g/m² groups were 22.3% and 14.7%, respectively (Figure 2); the curves showed significant differences (P=0.007). In the weighted cohort, the rates of MACCE at 10 years in the LVMI >200 g/m² and LVMI ≤200 g/m² groups were 25.6% and 13.5%, respectively (Figure 3); the curves showed significant differences (P=0.020).

Figure 2.

Cumulative incidence of major adverse cardiac and cerebrovascular events (MACCE) in the unweighted cohort. LVMI, left ventricular mass index.

Figure 3.

Cumulative incidence of major adverse cardiac and cerebrovascular events (MACCE) in the weighted cohort. LVMI, left ventricular mass index.

The causes of MACCE are presented in Table 3. Significantly more patients in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group experienced MACCE during the follow-up period (27.3% vs. 10.7%; P<0.001). Among MACCE, the incidence of all-cause death (7.7% vs. 2.3%; P=0.005) and non-fatal heart failure (16.1% vs. 5.8%; P<0.001) was significantly higher in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group.

Table 3.

Causes of MACCE

	Unweighted			Weighted
	LVMI >200 g/m2 (n=92)	LVMI ≤200 g/m2 (n=171)	P value	LVMI >200 g/m2 (SoW=267.19)	LVMI ≤200 g/m2 (SoW=261.30)	P value
MACCE	25 (27.2)	19 (11.1)	<0.001	77.82 (27.3)	28.01 (10.7)	<0.001
All-cause death	7 (7.6)	5 (2.9)	0.129	20.51 (7.7)	6.06 (2.3)	0.005
Cardiac death	2 (2.2)	0 (0)	0.158	6.20 (2.3)	0 (0)	0.013
MI	0 (0)	0 (0)	–	0 (0)	0 (0)	–
Heart failure	2 (2.2)	0 (0)	0.158	6.20 (2.3)	0 (0)	0.013
Lethal arrhythmia	0 (0)	0 (0)	–	0 (0)	0 (0)	–
Non-cardiac death	5 (5.4)	5 (2.9)	0.355	14.31 (5.4)	6.06 (2.3)	0.069
Pneumonia	1 (1.1)	1 (0.6)	0.656	2.88 (1.1)	1.16 (0.4)	0.404
Stroke	0 (0)	0 (0)	–	0 (0)	0 (0)	–
Sepsis	3 (3.3)	1 (0.6)	0.173	9.43 (3.5)	1.00 (0.4)	0.009
Cancer	0 (0)	0 (0)	–	0 (0)	0 (0)	–
Others	1 (1.1)	3 (1.8)	0.675	2.00 (0.7)	3.90 (1.5)	0.417
Non-fatal MI	0 (0)	0 (0)	–	0 (0)	0 (0)	–
Non-fatal heart failure	14 (15.2)	9 (5.3)	0.018	42.90 (16.1)	15.24 (5.8)	<0.001
Non-fatal stroke	3 (3.3)	3 (1.8)	0.437	7.95 (3.0)	3.99 (1.5)	0.262
Repeat revascularization	1 (1.1)	2 (1.2)	0.952	1.46 (0.5)	2.72 (1.0)	0.522

Unless indicated otherwise, data are given as n (%). MACCE, major adverse cardiac and cerebrovascular event; MI, myocardial infarction. Other abbreviations as in Table 1.

Multivariable Cox proportional hazards analysis showed that predictors of MACCE were previous cerebrovascular disease (hazard ratio [HR] 3.067; 95% confidence interval [CI] 1.382–6.805; P=0.006), preoperative hemodialysis (HR 4.101; 95% CI 1.055–15.935; P=0.042), redo surgery (HR 3.585; 95% CI 1.331–9.643; P=0.011), and preoperative LVMI >200 g/m² (HR 2.356; 95% CI 1.283–4.327; P=0.006; Supplementary Table).

In the LVMI >200 g/m² group, during the follow-up period, 1 patient underwent surgical AVR for AR associated with structural valve deterioration (SVD), 1 patient underwent surgical AVR for aortic stenosis (AS) associated with SVD, 1 patient underwent transcatheter AVR for AR associated with SVD, and 1 patient underwent the Bentall procedure for prosthetic valve endocarditis. In the LVMI ≤200 g/m² group, 4 patients underwent surgical AVR for AR associated with SVD, 2 patients underwent surgical AVR for AS associated with SVD, 1 patient underwent surgical AVR for prosthetic valve endocarditis, 1 patient underwent transcatheter AVR for AR associated with SVD, and 1 patient underwent transcatheter AVR for AS associated with SVD.

LVM Regression During Follow-up

The time course of the LVMI after surgery in the 2 groups is shown in Figure 4. In the LVMI >200 g/m² group, significant regression of LVMI was observed up to 2 years after surgery, and this LVMI was maintained up to 10 years after surgery. Similarly, in the LVMI ≤200 g/m² group, significant regression of LVMI was observed up to 2 years after surgery, and the LVMI was maintained up to 10 years after surgery. During the follow-up period, a significant difference in the LVMI remained between the 2 groups (Table 4).

Figure 4.

Time course of the left ventricular mass index (LVMI) in the groups with preoperative LVMI >200 g/m² and ≤200 g/m². The boxes show the interquartile range, with the median value indicated by the horizontal line; whiskers show the range. Postop, postoperatively; preop, preoperatively. In this figure, “X” shows the mean value.

Table 4.

Time Courses of Left Ventricular Mass Index (LVMI) in the 2 LVMI Groups

	LVMI (g/m2)		P value
	LVMI >200 g/m2 group (n=92)	LVMI ≤200 g/m2 group (n=171)
Preoperatively	245±38	155±28	<0.001
Postoperatively
1 week	193±44	136±34	<0.001
1 year	150±32	117±25	<0.001
2 years	140±31	112±24	<0.001
3 years	134±29	111±25	<0.001
4 years	128±26	110±22	<0.001
5 years	133±30	112±25	<0.001
6 years	132±33	110±22	<0.001
7 years	131±29	107±23	<0.001
8 years	127±26	110±25	<0.001
9 years	133±30	107±27	<0.001
10 years	131±28	111±26	0.003

Unless indicated otherwise, data are given as the mean±SD.

In the LVMI >200 g/m² group, the mean LVMI regressed from 245 g/m² before surgery to 140 g/m² 2 years after surgery (P<0.001), and the percentage change in LVMI was −42.9%. In contrast, in the LVMI ≤200 g/m² group, the mean LVMI regressed from 155 g/m² before surgery to 112 g/m² 2 years after surgery (P<0.001), with a percentage change in LVMI of −27.7%.

Discussion

In the American College of Cardiology/American Heart Association guidelines, AVR for severe AR is recommended in asymptomatic patients with an LV ejection fraction >55% when the LV is severely enlarged (LVESD >50 mm or indexed LVESD >25 mm/m²; Class IIa indication).⁴ Conversely, in the European Society of Cardiology/European Society for Cardio-Thoracic Surgery guidelines, AVR for severe AR is recommended in asymptomatic patients with an LVESD >50 mm, an indexed LVESD >25 mm/m², or a resting LV ejection fraction ≤50% (Class I indication).⁵ There is no mention of the LVMI in either guideline, and little is known about the effect of preoperative LVM on long-term outcomes after AVR for AR.

In this study, we excluded patients who underwent combined valvular surgery or who underwent coronary artery bypass grafting which may affect LV remodeling. In addition, only 6 (2.3%) patients in our cohort underwent previous percutaneous coronary intervention. Therefore, the primary cause of LV remodeling was AR progression, and ischemic cardiomyopathy seemed to have little effect. This allowed us to investigate the influence of preoperative factors on postoperative outcomes.

One of the major findings of this study was that the rate of MACCE was significantly higher in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group; LVMI >200 g/m² was significantly associated with a higher risk of MACCE. In the present study, significantly more patients in the LVMI >200 g/m² group experienced MACCE than in the LVMI ≤200 g/m² group (P<0.001; Table 3). In addition, the incidence of death from heart failure (P=0.013) and non-fatal heart failure (P<0.001) was significantly higher in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group. Higher LVM is associated with an increased risk of heart failure.⁶ Therefore, a higher LVM in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group was likely associated with the development of heart failure, which led to the development of MACCE.

In the present study, the percentage change in LVMI from before surgery to 2 years after surgery was greater in the LVMI >200 g/m² group (−42.9%) than in the LVMI ≤200 g/m² group (−27.7%). However, during the follow-up period, the significant difference in LVMI between the 2 groups remained (Table 4), which was another finding of the present study. The “point of no return” in chronic AR is unknown. However, if the preoperative LVMI becomes too large, it will be difficult to return the LVMI to a normal value. In our entire cohort, echocardiographic data 5 years after surgery could only be obtained for 53.6% (141/263) of patients, so these results should be interpreted with caution. However, significantly greater LVMI in the LVMI >200 g/m² group than in the LVMI ≤200 g/m² group during the follow-up period may have influenced the rates of MACCE.

AS induces chronic pressure overload of the LV, resulting in increased LVM consisting mainly of c-oncentric hypertrophy.^7,8 AVR for AS can minimize the transaortic pressure gradient and regress LVM.^9,10 However, AR induces volume overload in the LV, resulting in increased LVM consisting of eccentric hypertrophy.¹¹ AVR for AR can correct volume overload and allow the cavity size to decrease,^12,13 resulting in LVM regression. A patient-prosthesis mismatch (PPM) after AVR for AS causes a high residual postoperative transaortic pressure gradient¹⁴ and is associated with impaired LVM regression.¹⁵ However, when AVR is performed for AR, the volume overload usually disappears if there is no paravalvular leak, even if there is a PPM. Consequently, eccentric hypertrophy may improve, regardless of the postoperative aortic valve area. Brown et al. examined the effect of PPM on postoperative LVM regression after AVR for AR in 90 patients, including 13 patients with a postoperative indexed aortic valve area <0.85 cm²/m².¹⁶ Brown et al. concluded that LVM regression after AVR for AR was unrelated to the postoperative aortic valve area.¹⁶ However, the mean follow-up in their study was 3.2 years; therefore, this result should be interpreted with caution. With a residual transaortic pressure gradient, LVM associated with concentric hypertrophy may appear in the long term. If the aortic annuls is small and the occurrence of postoperative PPM is highly anticipated, aortic annuls enlargement techniques or aortic valvuloplasty should be considered. In our entire cohort, only 3 (1.1%) patients had a postoperative aortic valve area <0.85 cm²/m². Therefore, postoperative PPM seems to have had little effect on LVM regression in this study.

There were no significant differences in postoperative drugs, including statins (P=0.416), β-blockers (P=0.356), and renin-angiotensin system inhibitors (P=0.168), between the 2 groups (Table 2). No patient took ARNI after surgery in our entire cohort. Helder et al. investigated the effect of postoperative medical treatment on LVM regression in 444 patients undergoing AVR at 6 institutions.¹⁷ In that study, the use of β-blockers or calcium channel blockers at discharge was significantly associated with LVM regression. Conversely, a previous meta-analysis compared the effects of ARNI with those of angiotensin-converting enzyme inhibitors or angiotensin receptor blockers (ACEi/ARB) on cardiac reverse remodeling and found that ARNI distinctly improved LV hypertrophy compared with ACEi/ARB in heart failure patients with reduced ejection fraction, even after short-term follow-up.¹⁸ In addition, several previous studies reported that statin use was associated with the regression of LVM.^19,20 Therefore, the use of these drugs for heart failure may further reduce LVM in addition to the effects of AVR. These drugs should be actively used in patients with greater LVM preoperatively.

ROC analysis showed that the cut-off value for the incidence of MACCE was 200 g/m² for the preoperative LVMI (Figure 1), but the AUC was relatively low (0.651). We compared postoperative outcomes after dividing patients into 2 groups, namely preoperative LVMI >200 g/m² and preoperative LVMI ≤200 g/m², and found that the incidence of MACCE was significantly higher in former group during the follow-up period. This significant difference remained after adjustment for characteristics using the IPTW method. In addition, a multivariable Cox proportional hazards model showed that preoperative LVMI >200 g/m² was an independent predictor of MACCE (HR 2.356; P=0.006). Therefore, although the AUC was relatively low, a preoperative LVMI of 200 g/m² is considered an important indicator in patients undergoing AVR for AR.

Study Limitations

This study has some limitations. First, the study was neither prospective nor randomized. Despite statistical adjustments with IPTW, unmeasured confounders may have affected the postoperative outcomes. Second, this was a small cohort study at a single institution in Japan, which may limit its generalizability. Third, we do not routinely perform preoperative magnetic resonance imaging in patients undergoing surgery for AR, so we could not calculate LVM more accurately using magnetic resonance imaging. Fourth, we were able to review the medications the patients were taking at the time of discharge, but we were not able to review the medications they were taking during the follow-up period after discharge. Fifth, echocardiographic data 5 years after surgery could only be obtained for 53.6% (141/263) of patients. Therefore, a selection bias for differences in LV remodeling between the 2 groups during the follow-up period should be considered. Finally, in our entire cohort, there were only 64 patients (24.3%) who did not experience MACCE 10 years after surgery.

Conclusions

Preoperative LVMI >200 g/m² was associated with a higher rate of MACCE in patients undergoing AVR for AR. During the follow-up period, a significant difference in LVMI remained between the 2 groups.

Acknowledgment

The authors thank Ellen Knapp, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.

Sources of Funding

None.

Disclosures

The authors have no conflicts of interest to declare.

IRB Information

This study was approved by the Institutional Review Board of Shiga University of Medical Science (Reference no. R2023-090).

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-24-0464

References

• 1.   Koga-Ikuta A, Fukushima S, Kawamoto N, Saito T, Shimahara Y, Yajima S, et al. Reverse remodelling after aortic valve replacement for chronic aortic regurgitation. Interact Cardiovasc Thorac Surg 2021; 33: 10–18.
• 2.   Devereux RB, Alonso DR, Lutas EM, Gottlieb GJ, Campo E, Sachs I, et al. Echocardiographic assessment of left ventricular hypertrophy: Comparison to necropsy findings. Am J Cardiol 1986; 57: 450–458.
• 3.   Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med 2009; 28: 3083–3107.
• 4.   Otto CM, Nishimura RA, Bonow RO, Carabello BA, Erwin JP 3rd, Gentile F, et al. 2020 ACC/AHA guideline for the management of patients with valvular heart disease: A report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2021; 143: e72–e227.
• 5.   Vahanian A, Beyersdorf F, Praz F, Milojevic M, Baldus S, Bauersachs J, et al. 2021 ESC/EACTS guidelines for the management of valvular heart disease. Eur Heart J 2022; 43: 561–632.
• 6.   Velagaleti RS, Gona P, Pencina MJ, Aragam J, Wang TJ, Levy D, et al. Left ventricular hypertrophy patterns and incidence of heart failure with preserved versus reduced ejection fraction. Am J Cardiol 2014; 113: 117–122.
• 7.   Fukui M, Genereux P, Cavalcante JL. Assessment of cardiac damage in aortic stenosis. Cardiol Clin 2020; 38: 23–31.
• 8.   Capoulade R, Clavel MA, Le Ven F, Dahou A, Thébault C, Tastet L, et al. Impact of left ventricular remodelling patterns on outcomes in patients with aortic stenosis. Eur Heart J Cardiovasc Imaging 2017; 18: 1378–1387.
• 9.   Dumesnil JG, Honos GN, Lemieux M, Beauchemin J. Validation and applications of indexed aortic prosthetic valve areas calculated by Doppler echocardiography. J Am Coll Cardiol 1990; 16: 637–643.
• 10.   Treibel TA, Kozor R, Schofield R, Benedetti G, Fontana M, Bhuva AN, et al. Reverse myocardial remodeling following valve replacement in patients with aortic stenosis. J Am Coll Cardiol 2018; 27: 860–871.
• 11.   Wang X, Wei W, Wu J, Kang L, Wu S, Li J, et al. Involvement of endoplasmic reticulum stress-mediated activation of C/EBP homologous protein in aortic regurgitation-induced cardiac remodeling in mice. J Cardiovasc Transl Res 2022; 15: 340–349.
• 12.   Austen WG, Corning HB, Moran JM, Sanders CA, Scannell JG. Cardiac hemodynamics immediately following aortic valve surgery. J Thorac Cardiovasc Surg 1966; 51: 461–467.
• 13.   Taniguchi K, Nakano S, Kawashima Y, Sakai K, Kawamoto T, Sakaki S, et al. Left ventricular ejection performance, wall stress, and contractile state in aortic regurgitation before and after aortic valve replacement. Circulation 1990; 82: 798–807.
• 14.   Pibarot P, Magne J, Leipsic J, Cote N, Blanke P, Thourani VH, et al. Imaging for predicting and assessing prosthesis-patient mismatch after aortic valve replacement. JACC Cardiovasc Imaging 2019; 12: 149–162.
• 15.   Tasca G, Brunelli F, Cirillo M, DallaTomba M, Mhagna Z, Troise G, et al. Impact of valve prosthesis-patient mismatch on left ventricular mass regression following aortic valve replacement. Ann Thorac Surg 2005; 79: 505–510.
• 16.   Brown ML, Schaff HV, Suri RM, Li Z, Sundt TM, Dearani JA, et al. Regression in left ventricular mass after aortic valve replacement for chronic aortic regurgitation is unrelated to prosthetic valve size. J Thorac Cardiovasc Surg 2011; 142: e5–e9.
• 17.   Helder MR, Ugur M, Bavaria JE, Kshettry VR, Groh MA, Petracek MR, et al. The effect of postoperative medical treatment on left ventricular mass regression after aortic valve replacement. J Thorac Cardiovasc Surg 2015; 149: 781–786.
• 18.   Wang Y, Zhou R, Lu C, Chen Q, Xu T, Li D. Effects of the angiotensin-receptor neprilysin inhibitor on cardiac reverse remodeling: Meta-analysis. J Am Heart Assoc 2019; 8: e012272.
• 19.   Su SF, Hsiao CL, Chu CW, Lee BC, Lee TM. Effects of pravastatin on left ventricular mass in patients with hyperlipidemia and essential hypertension. Am J Cardiol 2000; 86: 514–518.
• 20.   Nishikawa H, Miura S, Zhang B, Shimomura H, Arai H, Tsuchiya Y, et al. Statins induce the regression of left ventricular mass in patients with angina. Circ J 2004; 68: 121–125.