Outcomes of Watchful Waiting Strategy and Predictors of Postoperative Prognosis in Asymptomatic or Equivocally Symptomatic Chronic Severe Aortic Regurgitation With Preserved Left Ventricular Function

Abstract

Background

The optimal surgical timing for asymptomatic or equivocally symptomatic chronic severe aortic regurgitation with preserved left ventricular ejection fraction remains controversial.

Methods and Results

Two hundred ten consecutive patients (median age 65 years) with asymptomatic or equivocally symptomatic chronic severe aortic regurgitation and left ventricular ejection fraction ≥50% were registered. First, the treatment plans (aortic valve replacement or watchful waiting) after initial diagnosis were investigated. Then, 2 studies were set: Study A (n=144) investigated the prognosis of patients who were managed under the watchful waiting strategy after initial diagnosis; Study B (n=99) investigated the postoperative prognosis in patients who underwent aortic valve replacement at initial diagnosis or after watchful waiting. The primary outcomes were all‐cause death in Study A and postoperative cardiovascular events in Study B. In Study A, 3 died of noncardiovascular causes during a median follow‐up of 3.2 years. In Kaplan–Meier analysis, the survival curve was similar to that of an age‐sex‐matched general population in Japan. In Study B, 9 experienced the primary outcome during a median follow‐up of 5.0 years. In Cox regression analysis, preoperative left ventricular end‐systolic diameter enlargement (hazard ratio, 1.11; P=0.048) and left ventricular end‐systolic diameter >45 mm (hazard ratio, 12.75; P=0.02) were significantly associated with poor postoperative prognosis. In Kaplan–Meier analysis, left ventricular end‐systolic diameter >45 mm predicted a higher risk of the primary outcome (P <0.01).

Conclusions

Watchful waiting was achieved safely in asymptomatic or equivocally symptomatic chronic severe aortic regurgitation with preserved left ventricular ejection fraction. Preoperative left ventricular end‐systolic diameter >45 mm predicted a poor postoperative outcome and may be an optimal cut‐off value for surgical indication.

Nonstandard Abbreviations and Acronyms

AR

aortic regurgitation

AVR

aortic valve replacement

LVEDD

left ventricular end‐diastolic diameter

LVESD

left ventricular end‐systolic diameter

LVESDi

left ventricular end‐systolic diameter index

LVGLS

left ventricular global longitudinal strain

NYHA

New York Heart Association

Clinical Perspective.

What Is New?

• The optimal timing for surgery in asymptomatic or equivocally symptomatic patients with chronic severe aortic regurgitation and preserved left ventricular ejection fraction remains controversial.

What Are the Clinical Implications?

• In this study, the watchful waiting strategy was achieved safely, showing a prognosis similar to that of an age‐ and sex‐matched general population, and surgery after watchful waiting was not a postoperative cardiovascular event risk; thus, it was a feasible approach for patients with chronic severe aortic regurgitation. Left ventricular end‐systolic diameter >45 mm could be an optimal cut‐off value among a population of small body size for predicting postoperative cardiovascular events.

Chronic aortic regurgitation (AR) is characterized by the diastolic reflux of blood from the aorta into the left ventricle, resulting in left ventricular (LV) volume and pressure overload. Although patients with chronic AR generally remain asymptomatic for a long period, the left ventricle eventually fails to maintain a compensated state. ¹ Because of the relatively young onset and gradual progression of AR, a watchful waiting strategy is feasible in clinical settings until it is optimal to perform surgery for chronic severe AR. Considering this specific natural history, echocardiography cut‐off values for surgery in current guidelines have been based mainly on predictors of postoperative prognosis in patients who underwent aortic valve replacement (AVR). ² , ³ , ⁴ , ⁵ However, few studies have demonstrated the prognosis in patients who were managed under the watchful waiting strategy.

Strong evidence has shown that aortic valve surgery improves the prognosis of patients who have chronic severe AR with marked symptoms (New York Heart Association [NYHA] class ≥III) ⁶ , ⁷ , ⁸ , ⁹ or reduced left ventricular ejection fraction (LVEF; <50%). ⁷ , ⁸ , ¹⁰ , ¹¹ , ¹² However, the optimal timing for surgery in asymptomatic or equivocally symptomatic (NYHA class I or II) patients with chronic severe AR and preserved LVEF remains controversial. In this population, the LV end‐systolic diameter (LVESD) has been reported as a better criterion for surgery than LV end‐diastolic diameter (LVEDD). Several studies have revealed that a higher LVESD or LVESD index (LVESDi) indicates progression of LV remodeling ⁵ and poor prognosis after AVR ⁸ , ¹¹ , ¹² , ¹³ , ¹⁴ despite minimal evidence of increased LVEDD. ¹⁵ On the other hand, recent studies have recommended early surgery in conservatively managed patients with a low LVESDi (20–25 mm/m²) due to increased all‐cause mortality. ¹ , ¹⁶ However, these studies may have included patients who did not undergo aortic valve surgery because of high comorbidity, patient's will, or financial reasons. Moreover, considering the natural history of chronic AR and its low cardiovascular event rate, most patients may have died of noncardiovascular causes. The cut‐off values of echocardiographic parameters that indicate the need for surgery should be determined over cardiovascular events.

In order to elucidate the optimal treatment options for asymptomatic or equivocally symptomatic patients with chronic severe AR and preserved LVEF, we investigated (1) the actual management and treatment plans after initial diagnosis, (2) the prognosis of patients who were managed under the watchful waiting strategy, and (3) the optimal cut‐off values of echocardiographic parameters in terms of postoperative prognosis in patients who underwent AVR.

METHODS

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Study Population

This was a retrospective, single‐center cohort study. We registered 506 consecutive patients with severe AR at the National Cerebral and Cardiovascular Center (Osaka, Japan) between January 2012 and December 2019. Patients were initially identified by searching the echocardiography and operation database. We excluded 77 patients with other severe valvular diseases, 65 who had a history of open heart surgery, 35 with severe coronary artery disease, 24 with acute AR, 13 with marked symptoms (NYHA class ≥III), 90 with LVEF <50%, and 2 who refused surgery. The applicability of the exclusion criteria was determined by the medical records in our hospital. Data of the remaining 210 asymptomatic or equivocally symptomatic (NYHA class I or II) patients with chronic severe AR and preserved LVEF (LVEF ≥50%) were used for the final analysis (Figure 1). Patients with NYHA class II were included in the final analysis because previous studies investigating the relationship between postoperative prognosis and symptoms cited by the guidelines were defined as being symptomatic as NYHA class ≥III. ⁶ , ⁷ , ⁸ , ⁹

Figure 1. Patient flow chart.

AR indicates aortic regurgitation; AVR, aortic valve replacement; LVEF, left ventricular ejection fraction; and NYHA, New York Heart Association.

First, we investigated the management plan after the initial diagnosis (either early AVR [defined as those who underwent surgery within 3 months from the diagnosis of severe AR] or watchful waiting [defined as those who did not undergo surgery within 3 months after the diagnosis of severe AR]). Then, 2 studies were set; Study A included patients who were managed under the watchful waiting strategy after the initial diagnosis of severe AR. We investigated the prognosis of these patients. Patients managed under the watchful waiting strategy were defined as those who received echocardiographic follow‐up at least every 6 to 12 months. Study B included patients who underwent AVR, either just after initial diagnosis or after a period of watchful waiting. We investigated the postoperative prognosis and the predictors of prognosis among these patients. Surgical risks were assessed using the EuroSCORE II, ¹⁷ Society of Thoracic Surgeons score, ¹⁸ , ¹⁹ , ²⁰ and JapanSCORE. ²¹ The attending physician selected either a mechanical valve or a bioprosthetic valve based on the age and clinical background of each patient. Informed consent was obtained by giving them the option to opt‐out on the website of the National Cerebral and Cardiovascular Center (Osaka, Japan). The ethical committee of the National Cerebral and Cardiovascular Center approved the study protocol (approval number: R21009). The investigation conformed to the principles outlined in the Declaration of Helsinki.

Measurements and Data Collection

We collected the following clinical data at initial diagnosis, before AVR, and 1 year after surgery from echocardiography and operation databases and medical records: age, sex, height, weight, NYHA class, laboratory data, medications, and echocardiographic data. Comprehensive transthoracic echocardiography was performed by experienced sonographers using high‐quality, commercially available ultrasound systems. The biplane modified Simpson's method was used to measure LVEF and LV volume. Relative wall thickness and LV mass were calculated using the formula described in the guidelines of the American Society of Echocardiography and the European Association of Cardiovascular Imaging. ²² The LV global longitudinal strain (LVGLS) measurements were obtained from baseline gray‐scale images recorded in the apical 2‐, 3‐, 4‐chamber views. The LVGLS was analyzed offline using TOMTEC 2D Cardiac Performance Analysis software (TomTec Imaging Systems, Unterschleissheim, Germany). Severe AR was carefully diagnosed by an integrated approach considering qualitative, semiquantitative, and quantitative parameters according to the guidelines. ² , ³ , ⁴ , ²³ Finally, only patients with severe AR were registered. The following semiquantitative and quantitative echocardiographic parameters were used: the vena contracta width and the regurgitant volume and effective regurgitant orifice area based on the proximal isovelocity surface area. ²³

Outcomes

In Study A, the primary outcome was all‐cause death. The secondary outcome was cardiovascular events, defined as a composite of cardiovascular death, hospitalization due to heart failure, myocardial infarction, ventricular arrhythmia, and infective endocarditis. Patients who underwent AVR during watchful waiting were censored at the operation date in the primary and secondary outcomes. To investigate the proportion of AVR‐free survival, supplementary analysis was performed under a composite of all‐cause death and AVR. In Study B, the primary outcome was postoperative cardiovascular events defined as a composite of cardiovascular death, hospitalization due to heart failure, myocardial infarction, ventricular arrhythmia, infective endocarditis, and re‐operation. Cardiovascular death was defined as death related to heart failure, death from other cardiac causes, or cardiac sudden death. Survival status was confirmed in September 2021 by reviewing clinical records, sending letters, and making telephone calls to patients.

Statistical Analysis

Continuous variables are summarized as the mean±SD if normally distributed and as the median and interquartile range if non‐normally distributed. Normality was assessed by the Shapiro–Wilk W‐test. Comparisons of baseline characteristics were made with a contingency table and the Pearson χ² test for categorical variables, the t test for normally distributed continuous variables, and the Mann‐Whitney U test for non‐normally distributed continuous variables. Kaplan–Meier survival plots were calculated from baseline on the initial diagnosis with severe AR (Study A) or the day performing surgery (Study B) to the times of the primary and secondary outcomes and compared using the log‐rank test. In Study A, the survival distribution of an age‐ and sex‐matched general Japanese population was estimated. For the investigation regarding general population survival, we downloaded an Excel macro file from the Massachusetts General Hospital Biostatistics Center (http://hedwig.mgh.harvard.edu/biostatistics/node/30) and estimated the survivorship following the instructions. ²⁴ We used the Japanese annual survival data downloaded from the Ministry of Health, Labour and Welfare home page (https://www.mhlw.go.jp/english/). Cox proportional‐hazards analysis was used to evaluate the hazard ratio and 95% CI of the parameters associated with the composite end point. To determine appropriate values for the prognostic predictors, LVEF, LVESD, LVESDi, and LVEDD were dichotomized by the median value or the value presented in the current American Heart Association/American Colleague of Cardiology guidelines, the European Society of Cardiology/European Association for Cardio‐Thoracic Surgery guidelines, and the Japanese Circulation Society/Japanese Society for Cardiovascular Surgery/Japanese Association for Thoracic Surgery/Japanese Society for Vascular Surgery guidelines. ² , ³ , ⁴ In Study B, patients were divided into 2 groups according to the median LVESD: the low LVESD group (LVESD ≤median value) and the high LVESD group (LVESD >median value). For sensitivity analysis, propensity score adjustment was performed to reduce the influence of confounding effects related to differences in the backgrounds of patients in the low and high LVESD groups. To calculate the propensity score, we adopted a logistic regression model in which LVESD enlargement was regressed for the following 13 preoperative baseline characteristics: age, sex, body mass index, body surface area, blood pressure (systolic and diastolic), NYHA class, hypertension, atrial fibrillation, aortic aneurysm, LVEF, B‐type natriuretic peptide, and serum creatinine. The C‐statistic was calculated to examine the accuracy of the propensity score. The Hosmer‐Lemeshow test was used to evaluate the compatibility of the multiple logistic regression. In addition, patients in the low and high LVESD groups were matched based on the propensity score using the greedy matching algorithm with a caliper width of 0.2. ²⁵ Kaplan–Meier analysis and the log‐rank test were repeated in the propensity matched cohort. A P value of <0.05 was considered statistically significant. All statistical analyses were performed using SPSS Statistics software for Windows Version 26 (IBM Corp., Armonk, NY).

RESULTS

Baseline Characteristics

The baseline characteristics of the entire patient population (median age, 65 years; male, 71%) are shown in Table 1. Seventy‐nine patients (38%) were categorized as NYHA class II. Among the 210 enrolled patients, 66 underwent AVR just after the initial diagnosis of severe AR. The remaining 144 patients were managed under the watchful waiting strategy; of these, 33 eventually underwent AVR (median 363 [177–968] days after diagnosis). Compared with those who were managed under the watchful waiting strategy, those who underwent AVR just after the initial diagnosis of severe AR were more likely to be in NYHA class II and had a higher B‐type natriuretic peptide level, a larger LV size, and reduced LVGLS. There were no differences in the quantitative parameters of AR between the 2 groups.

Table 1.

Baseline Characteristics of Patients With Asymptomatic Chronic Severe Aortic Regurgitation and Preserved Left Ventricular Ejection Fraction

Variable	Overall population (n=210)	Watchful waiting strategy after initial diagnosis (n=144): Study A	AVR just after initial diagnosis (n=66)	P value
Age, y	65 [46–73]	65 [44–73]	67 [50–73]	0.65
Male, n (%)	148 (71)	100 (69)	48 (73)	0.63
Height, cm	165 [157–172]	163 [157–171]	167 [155–172]	0.54
Body surface area, m2	1.65±0.21	1.74±0.20	1.69±0.24	0.15
BMI, kg/m2	22.4 [20.3–24.4]	22.0 [20.2–23.9]	22.8 [21.1–25.6]	0.04
NYHA Class (I/II), n	131/79	109/35	22/44	<0.01
Systolic blood pressure, mm Hg	132±17	131±17	132±18	0.68
Diastolic blood pressure, mm Hg	61±12	61±12	58±13	0.09
Heart rate (beats/min)	64 [58–72]	64 [58–72]	67 [60–72]	0.16
Hypertension, n (%)	128 (61)	86 (60)	42 (64)	0.59
Dyslipidemia, n (%)	60 (29)	38 (26)	22 (33)	0.30
Diabetes, n (%)	11 (5)	7 (5)	4 (6)	0.72
Atrial fibrillation, n (%)	12 (6)	5 (4)	7 (11)	0.04
Ischemic heart disease, n (%)	5 (2)	2 (1)	3 (5)	0.16
Cerebral infarction, n (%)	11 (5)	8 (6)	3 (5)	0.76
COPD, n (%)	6 (3)	4 (3)	2 (3)	0.92
Laboratory data
Hemoglobin, g/dL	13.6 [12.5–14.6]	13.6 [12.7–14.6]	13.5 [12.1–14.7]	0.50
Serum creatinine, mg/dL	0.86 [0.74–0.97]	0.86 [0.71–0.98]	0.86 [0.74–0.94]	0.97
BNP, pg/mL	51 [22–114]	43 [20–92]	75 [34–162]	<0.01
Echocardiographic data
LVEF, %	58 [54–62]	58 [55–62]	57 [51–62]	0.05
LVGLS, %	20 [17–22]	21 [18–23]	19 [16–20]	<0.01
LVEDD, mm	61±7	60±6	64±8	<0.01
LVESD, mm	41±6	40±6	43±7	<0.01
LVESDi, mm/m2	24 [23–28]	24 [22–27]	26 [23–28]	0.05
LVESV, mL	78 [57–99]	71 [52–95]	92 [64–116]	<0.01
LVESVi, mL/m2	47 [35–57]	45 [34–56]	52 [38–67]	0.02
LAD, mm	39 [35–43]	38 [34–41]	41 [36–44]	<0.01
LAVi, mL/m2	41 [34–50]	40 [33–50]	43 [37–54]	0.04
LV mass index, g/m2	137 [114–165]	131 [107–153]	159 [129–182]	<0.01
Relative wall thickness	0.30 [0.27–0.34]	0.31 [0.27–0.34]	0.29 [0.27–0.34]	0.45
Mitral E/A ratio (n=200)	0.79 [0.58–1.22]	0.80 [0.58–1.22]	0.78 [0.58–1.25]	0.84
E wave, cm/s	60 [44–73]	61 [44–74]	57 [44–73]	0.70
E/e′ average ratio	8.7 [6.5–10.9]	8.7 [6.4–11.3]	8.8 [6.9–10.8]	0.97
TAPSE, mm	23 [21–27]	22 [20–27]	23 [22–25]	0.67
AR vena contracta width	5.9 [5.2–6.4]	5.9 [5.2–6.3]	5.9 [5.5–6.8]	0.29
AR EROA, cm2	0.26 [0.22–0.31]	0.25 [0.22–0.30]	0.27 [0.23–0.33]	0.12
AR regurgitant volume, mL	57 [52–65]	57 [52–65]	59 [52–69]	0.44

Values are median and interquartile range, or n (%). AR indicates aortic regurgitation; AVR, aortic valve replacement; BMI, body mass index; BNP, B‐type natriuretic peptide; COPD, chronic obstructive pulmonary disease; E, peak early mitral inflow velocity; e′, peak early diastolic velocity of the mitral annulus; EROA, effective regurgitant orifice area; LAD, left atrial diameter; LAVi, left atrial volume index; LV, left ventricular; LVEDD, left ventricular end‐diastolic diameter; LVEF, left ventricular ejection fraction; LVESD, left ventricular end‐systolic diameter; LVESDi, left ventricular end‐systolic diameter index; LVESV, left ventricular end‐systolic volume; LVESVi, left ventricular end‐systolic volume index; LVGLS, left ventricular global longitudinal strain; NYHA, New York Heart Association; PWT, posterior wall thickness; and TAPSE, tricuspid annular plane systolic excursion.

Study A

Patient Characteristics and Clinical Outcomes

Among the 144 patients (median age, 65 years; male, 69%), the causes of AR were as follows: degenerative valve in 60 patients, bicuspid valve in 47, annulo‐aortic ectasia in 25, valve prolapse in 7, and other causes in 6. The baseline patient characteristics are shown in Table 1. Only 5 patients (3.5%) were followed up outside our institution, and their cardiovascular events were confirmed by sending letters or making telephone calls; the other 139 patients (96.5%) were followed up in our institution. Patients followed up in our institution received echocardiography at least once a year. During a median follow‐up of 3.2 (1.0–5.9) years, 3 patients (2.1%) died of lung cancer, renal failure, and old age (unknown details), respectively. No patients died of cardiovascular causes. Thirty‐five patients (24%) were categorized as NYHA class II. In univariate Cox proportional‐hazards analysis, NYHA class I was not associated with better outcomes (hazard ratio, 0.62 [95% CI, 0.06–6.81]; P=0.69). Thirty‐three patients (23%) underwent AVR during watchful waiting. Among them, the indication of surgery was as follows: those who reached the cut‐off values of LV dimension in the guidelines, 28 (85%); appearance of symptoms, 4 (12%); infective endocarditis, 1 (3%). In Kaplan–Meier analysis, the estimated 5‐year survival rate was 96.9%. The survival curve was similar to that of the age‐ and sex‐matched general population in Japan (Figure 2A). Since patients who underwent AVR during watchful waiting were censored at the operation date, the 3‐year and 5‐year follow‐up rate was 64% (75/118) and 42% (47/112), respectively. Five patients (3.5%) experienced the secondary outcome (hospitalization due to heart failure, 4; infective endocarditis, 1). The estimated 5‐year event‐free survival rate was 97.4% (Figure 2B). In supplementary analysis, the 5‐year freedom rate from all‐cause death and AVR was 69.7% (Figure S1).

Figure 2. Kaplan–Meier plots of patients who were managed under the watchful waiting strategy (red line) and an age‐ and sex‐matched general population (green line).

A, Overall survival, (B) cardiovascular events.

Study

Patient Characteristics

Study B consisted of 99 patients (median age, 64 years; male, 76%), including 66 who underwent AVR just after the initial diagnosis of severe AR, and 33 in whom AVR was performed after watchful waiting. The baseline preoperative characteristics are shown in Table 2. The indication for surgery among the 66 patients who underwent AVR just after the initial diagnosis of severe AR was as follows: those who reached the cut‐off values of LV dimension in the guidelines, 55 (83%); ascending aortic aneurysm, 5 (8%); those who underwent AVR due to patient's will and the clinical decision by attending physicians, 6 (9%). A bioprosthesis valve was used in 82 (83%) patients and a mechanical valve was used in 17 (17%). Procedural details of the surgery are shown in Table S1. At 1 year after surgery, postoperative echocardiography was followed up in 92 (93%) patients. There was no significant difference between pre‐ and postoperative LVEF (57±5% versus 57%±6%, P=0.79), although LVEDD (65±7 mm versus 48±4 mm, P <0.01), and LVESD (44±6 mm versus 32±5 mm, P <0.01) significantly decreased 1 year after surgery.

Table 2.

Preoperative Baseline Characteristics in Patients Who Underwent Aortic Valve Replacement

Variable	Patients who underwent AVR (n=99): Study B	AVR just after initial diagnosis (n=66)	AVR after watchful waiting strategy (n=33)	P value
Age, y	64 [44–73]	67 [50–73]	49 [41–68]	0.047
Male, n (%)	75 (76)	48 (73)	27 (82)	0.32
Height, cm	167 [157–173]	167 [155–172]	171 [163–174]	0.05
Body surface area, m2	1.71±0.22	1.69±0.24	1.74±0.16	0.17
BMI, kg/m2	22.7 [21.3–24.8]	22.8 [21.1–25.6]	22.4 [21.5–24.0]	0.79
EuroSCORE II, %	1.1 [0.8–1.7]	1.2 [0.8–1.9]	0.9 [0.8–1.6]	0.25
STS score, %	1.0 [0.5–2.1]	1.2 [0.6–2.4]	0.9 [0.5–1.6]	0.19
JapanSCORE, %	1.5 [0.8–2.6]	1.7 [1.1–3.0]	1.0 [0.7–1.8]	0.04
NYHA Class (I/II), n	39/60	22/44	17/16	0.08
Systolic blood pressure, mm Hg	128 [120–142]	131 [120–145]	125 [121–135]	0.53
Diastolic blood pressure, mm Hg	55 [48–65]	55 [48–68]	55 [48–61]	0.45
Heart rate (beats/min)	67 [60–72]	67 [60–72]	67 [60–73]	0.98
Hypertension, n (%)	64 (65)	42 (64)	22 (67)	0.77
Dyslipidemia, n (%)	30 (30)	22 (33)	8 (24)	0.35
Diabetes, n (%)	5 (5)	4 (6)	1 (3)	0.52
Atrial fibrillation, n (%)	8 (8)	7 (11)	1 (3)	0.19
Ischemic heart disease, n (%)	3 (3)	3 (5)	0 (0)	0.21
Aortic aneurysm, n (%)	19 (19)	12 (18)	7 (21)	0.72
Cerebral infarction, n (%)	4 (4)	3 (5)	1 (3)	0.72
COPD, n (%)	3 (3)	2 (3)	1 (3)	0.99
Cause of aortic regurgitation
Bicuspid valve	36 (36)	20 (30)	16 (49)	0.08
Degenerative valve	33 (33)	26 (40)	7 (21)	0.07
Valve prolapse	18 (18)	11 (17)	7 (21)	0.58
Annulo‐aortic ectasia	8 (8)	7 (11)	1 (3)	0.18
Rheumatic valve	1 (1)	1 (2)	0 (0)	0.67
Other causes	3 (3)	1 (2)	2 (6)	0.26
Medication
RASi, n (%)	52 (53)	38 (58)	14 (42)	0.16
β‐Blockers, n (%)	17 (17)	10 (115)	7 (21)	0.45
MRAs, n (%)	8 (8)	7 (11)	1 (3)	0.19
Antiplatelet drugs, n (%)	10 (10)	8 (12)	2 (6)	0.35
Anticoagulants, n (%)	6 (6)	6 (9)	0 (0)	0.07
Digoxin, n (%)	4 (4)	3 (5)	1 (3)	0.72
Laboratory data
Hemoglobin, g/dL	13.5±1.7	13.3±1.8	13.9±1.6	0.09
Serum creatinine, mg/dL	0.88 [0.75–0.98]	0.86 [0.74–0.94]	0.90 [0.77–1.01]	0.27
BNP, pg/mL	65 [27–123]	75 [34–162]	39 [20–86]	0.03
Echocardiographic data
LVEF, %	56 [53–61]	57 [51–62]	56 [55–59]	0.99
LVGLS, %	19±3	19±3	18±3	0.724
LVEDD, mm	65±7	64±8	67±6	0.14
LVESD, mm	44±6	43±7	46±5	0.01
LVESDi, mm/m2	26±4	26±4	27±3	0.27
LVESV, mL	98±38	94±42	104±29	0.261
LVESVi, mL/m2	56±20	53±22	59±15	0.151
LAD, mm	40 [36–44]	41 [36–44]	37 [34–44]	0.09
LAVi, mL/m2	43 [37–50]	43 [37–54]	43 [37–49]	0.52
LV mass index, g/m2	155±34	158±36	150±30	0.29
Relative wall thickness	0.29±0.06	0.30±0.06	0.27±0.06	0.02
Mitral E/A ratio (n=93)	0.80 [0.60–1.30]	0.78 [0.58–1.25]	0.82 [0.50–1.40]	0.96
E wave, cm/s	57 [44–70]	57 [44–73]	59 [43–68]	0.66
E/e′ average ratio	8.5 [6.3–10.2]	8.8 [6.9–10.8]	7.7 [5.9–9.4]	0.07
TAPSE, cm	23 [22–26]	23 [22–25]	24 [22–27]	0.83

Values are median and interquartile range, or n (%). AVR indicates aortic valve replacement; BMI, body mass index; BNP, B‐type natriuretic peptide; COPD, chronic obstructive pulmonary disease; E, peak early mitral inflow velocity; e′, peak early diastolic velocity of the mitral annulus; LAD, left atrial diameter; LAVi, left atrial volume index; LV, left ventricular; LVEDD, left ventricular end‐diastolic diameter; LVEF, left ventricular ejection fraction; LVESD, left ventricular end‐systolic diameter; LVESDi, left ventricular end‐systolic diameter index; LVESV, left ventricular end‐systolic volume; LVESVi, left ventricular end‐systolic volume index; LVGLS, left ventricular global longitudinal strain; MRAs, mineralocorticoid receptor antagonists; NYHA, New York Heart Association; RASi, renin‐angiotensin system inhibitor; STS score, Society of Thoracic Surgeons score; and TAPSE, tricuspid annular plane systolic excursion.

Clinical Outcomes and Predictors of Postoperative Prognosis

During a median follow‐up of 5.0 (3.0–6.8) years after surgery, 9 patients (9.1%) experienced the primary outcome (cardiovascular death, 2; hospitalization due to heart failure, 2; myocardial infarction, 2; infective endocarditis, 1; re‐operation, 3). The patient with infective endocarditis was one of the 3 cases of re‐operation. Only 1 patient died of noncardiovascular causes (old age). The median values of LVEF, LVESD, LVESDi, and LVEDD just before surgery were 56%, 45 mm, 26 mm/m², and 66 mm, respectively. After univariate Cox regression analysis, preoperative LVESD enlargement (per mm; hazard ratio, 1.11 [95% CI, 1.00–1.24]; P=0.048) and preoperative LVESD >45 mm (median; hazard ratio, 12.75 [95% CI, 1.58–102.60]; P=0.02) were significantly associated with poor postoperative prognosis (Table 3). NYHA class, preoperative LVESDi, LVEDD, and the timing of surgery (either just after the initial diagnosis of severe AR or after a period of watchful waiting) (Figure S2) were not associated with the primary outcome.

Table 3.

Univariate Cox Regression Analysis of the Primary Outcome in Patients Who Underwent Aortic Valve Replacement

Variable	HR (95% CI)	P value
NYHA I	0.65 (0.15–2.76)	0.56
AVR after watchful waiting strategy	1.33 (0.32–5.63)	0.70
Preoperative LVEF (per %)	0.97 (0.86–1.10)	0.67
Preoperative LVEF >55%	0.87 (0.23–3.27)	0.84
Preoperative LVEF >56%; median	1.13 (0.30–4.23)	0.85
Preoperative LVGLS (per %)	0.94 (0.73–1.20)	0.63
Preoperative LVGLS <18%; median	1.12 (0.28–4.50)	0.87
Preoperative LVESD (per mm)	1.11 (1.00–1.24)	0.048
Preoperative LVESD >45 mm; median, JCS	12.75 (1.58–102.60)	0.02
Preoperative LVESD >50 mm; AHA & ESC	2.23 (0.54–9.24)	0.27
Preoperative LVESDi (per mm/m2)	1.13 (0.96–1.32)	0.13
Preoperative LVESDi >25 mm/m2; AHA/ESC/JCS	2.48 (0.51–11.99)	0.26
Preoperative LVESDi >26.2 mm/m2; median	2.10 (0.52–8.51)	0.30
Preoperative LVEDD (per mm)	1.06 (0.97–1.17)	0.21
Preoperative LVEDD >65 mm; median, AHA/JCS	3.80 (0.78–18.49)	0.10
Preoperative LVESV (per mL)	1.01 (0.98–1.03)	0.65
Preoperative LVESV >90 mL; median	1.88 (0.41–8.67)	0.42
Preoperative LVESVi (per mL/m2)	1.00 (0.96–1.05)	0.99
Preoperative LVESVi >51 mL/m2; median	0.91 (0.20–4.18)	0.90

AHA indicates American Heart Association/American Colleague of Cardiology guidelines; AVR, aortic valve replacement; ESC, European Society of Cardiology/European Association for Cardio‐Thoracic Surgery guidelines; JCS, Japanese Circulation Society/Japanese Society for Cardiovascular Surgery/Japanese Association for Thoracic Surgery/Japanese Society for Vascular Surgery guidelines; HR, hazard ratio; LVEDD, left ventricular end‐diastolic diameter; LVEF, left ventricular ejection fraction; LVESD, left ventricular end‐systolic diameter; LVESDi, left ventricular end‐systolic diameter index; LVESV, left ventricular end‐systolic volume; LVESVi, left ventricular end‐systolic volume index; LVGLS, left ventricular global longitudinal strain; and NYHA, New York Heart Association.

Prognostic Impact of LVESD

LVESD >45 mm was related to an increased risk of postoperative cardiovascular events compared with LVESD ≤45 mm (1.9% [1/53] versus 17.4% [8/46], P=0.01). In Kaplan–Meier analysis, LVESD >45 mm predicted a higher risk of the primary outcome (log‐rank test, P <0.01) (Figure 3A). Propensity score matching resulted in the creation of 35 matched pairs of patients with LVESD >45 mm and LVESD ≤45 mm. The multiple logistic regression model had a C‐statistic of 0.786. The Hosmer‐Lemeshow test had a P value of 0.09, and the compatibility of the multiple logistic regression was good. The preoperative baseline characteristics in the propensity score matched cohort are shown in Table S2. After propensity score matching, the incidence of the primary outcome was significantly lower in patients with LVESD ≤45 mm than in those with LVESD >45 mm (0.0% [0/35] versus 17.1% [6/35], P=0.01). As in the unmatched cohort, LVESD >45 mm was associated with a higher risk of the primary outcome in the matched cohort (log‐rank test, P <0.01) (Figure 3B).

Figure 3. Kaplan–Meier plots of the primary outcome according to LVESD level.

Blue line, LVESD ≤45 mm; orange line, LVESD >45 mm. A, Crude cohort. B, Propensity score–matched cohort. LVESD indicates left ventricular end‐systolic diameter.

DISCUSSION

The main findings of this study are as follows: (1) the watchful waiting strategy was selected in ≈70% of asymptomatic or equivocally symptomatic patients with chronic severe AR and preserved LVEF; (2) watchful waiting was achieved safely, because no patients died of cardiovascular causes during the follow‐up period and the 5‐year survival rate was similar to that of the age‐ and sex‐matched general population; (3) preoperative LVESD >45 mm predicted postoperative cardiovascular events in patients with asymptomatic or equivocally symptomatic chronic severe AR and preserved LVEF who underwent AVR; (4) surgery after watchful waiting was not a risk of postoperative events.

Because the natural history of chronic AR is often characterized by few if any symptoms and a low cardiovascular event rate during conservative management, echocardiographic cut‐off values that indicate the need for surgery have been determined mainly based on postoperative prognosis. ⁵ Regarding class I recommendations for surgery in the current guidelines for chronic severe AR ² , ³ , ⁴ (symptomatic or reduced LVEF [<50%]), most background data were based on postoperative survival in those who underwent AVR. ⁶ , ⁷ , ⁸ , ⁹ , ¹⁰ , ¹¹ , ¹² Likewise, background evidence for the class IIa recommendation regarding LVESD or LVESDi was based mainly on postoperative survival. ⁸ , ¹¹ , ¹² , ¹³ , ¹⁴ , ²⁶ , ²⁷ On the other hand, 2 recent studies of chronic severe AR were performed in the United States with large sample sizes, and based on prognostic data with conservative management, they proposed an even lower LVESDi of ≥20 mm/m² as a cut‐off value of surgical indication. ¹ , ¹⁶ However, these studies investigated only all‐cause mortality and may have included patients who could not undergo surgery because of high comorbidity, patient's choice, or financial reasons. The overall mortality rates in these studies were 19% ¹ and 17% ¹⁶ at a median follow‐up duration of 6.6 years and 4.9 years, respectively, and the 5‐year mortality rate derived from spline curves differed between these 2 studies. Moreover, the mortality rates in both studies were much higher than that in a study from Belgium, which reported an estimated 5‐year survival of 97% with conservative management; this study found that early surgery did not improve prognosis in asymptomatic chronic severe AR. ²⁸ Prognosis under conservative management is strongly influenced by clinical backgrounds of the study population, especially in patients with chronic AR. Therefore, echocardiographic cut‐off values investigated from conservative management would not be appropriate for determining surgical indications.

In this study, the estimated 5‐year survival rate of patients who were managed under the watchful waiting strategy was similar to that of the age‐ and sex‐matched general population in Japan. Moreover, no patients died of cardiovascular causes during medical follow‐up, and the postoperative prognosis was similar between patients who underwent AVR after initial diagnosis and those who underwent AVR after watchful waiting. Therefore, the watchful waiting strategy was achieved safely, demonstrating that the cut‐off value of LV size as an indication of the need for surgery should be determined based on data related to postoperative prognosis, as previous studies have done. Moreover, patients who underwent AVR just after the initial diagnosis of severe AR had significantly lower LVGLS than those who were managed under the watchful waiting strategy. However, LVGLS did not predict postoperative prognosis, which is inconsistent with several previous studies that reported the association between impaired LVGLS and disease progression or prognosis. ²⁹ , ³⁰ , ³¹ This may be because previous studies have evaluated not postoperative prognosis but prognosis under the conservative treatment or LV function just after surgery in addition to the selected population of asymptomatic chronic severe AR with preserved LVEF (≥50%) in the present study.

In our study, preoperative LVESD >45 mm was significantly associated with an increased risk of postoperative cardiovascular events, and this result was confirmed in the propensity score matched cohort. This is the first study to investigate the prognostic utility of LVESD over cardiovascular events in a selected cohort with no obvious symptoms (NYHA class ≤II) and preserved LVEF (≥50%), along with data regarding the watchful waiting strategy. Regarding obvious symptoms in patients with chronic severe AR, all previous studies cited in guidelines that provide evidence for surgical indications regarding symptoms distinguished patients with NYHA class ≤II from those with NYHA class ≥III. Therefore, the definition of “symptoms,” as a class I recommendation of surgery in guidelines, should be regarded as NYHA class ≥III. Surgery for patients with NYHA class II should be determined with careful decision based on the values of LVEF and LVESD.

The LVESD cut‐off value of 45 mm in the current Japanese guidelines ⁴ (which is 5 mm smaller than that in Western countries) took into account the small Japanese body size and was defined based on previous reports from Japan. ¹¹ , ²⁶ LVESD ≤45 mm appropriately indicated the need for surgery in Japanese patients from the perspective of postoperative cardiovascular events. On the other hand, the cut‐off value of LVESD 50 mm recommended in Western guidelines ² , ³ could be appropriate for Westerners whose body size is larger than that of Japanese.

Contrary to recent previous studies, preoperative LVESDi was not associated with postoperative cardiovascular events in our study cohort. This finding was, however, consistent with other studies from Japan, where body size is smaller (average body surface area 1.6–1.7 m²) than in Western countries. ¹¹ , ²⁶ , ²⁷ Previous studies from the United States and Europe, where the average body surface area is 1.9–2.0 m², investigated the association between high LVESDi and poor prognosis. Considering these differences, overcorrection with body surface area ⁵ might be one of the reasons why LVESDi was not associated with cardiovascular events in Japanese patients; ie, distance (LVESD: a 1‐dimensional number) is divided by area (body surface area: a 2‐dimensional number). This overcorrection of LVESD by body surface area may also be applicable in cases of patients in the United States and Europe with relatively small body size (<1.68 m²) or very large body size. ² With regard to volume data, preoperative LV end‐systolic volume and LV end‐systolic volume index were not associated with postoperative cardiovascular events in the present study, inconsistent with previous reports. ³² The poor accuracy and reproducibility of LV volumes measured by echocardiography may be the reason for this result. It would be reasonable to correct LV end‐systolic volume by body surface area (3‐dimension divided by 2‐dimension) if the LV volumes could be measured accurately.

Our findings suggest that the watchful waiting strategy is a feasible approach for patients who do not meet the criteria for aortic valve surgery in chronic severe AR. Moreover, LVESD >45 mm could be an optimal cut‐off value indicating the need for surgery in patients with small body size who have asymptomatic or equivocally symptomatic chronic severe AR with preserved LVEF. Individuals with this condition are unlikely to experience cardiovascular events during the watchful waiting period, and the perioperative risks remain low when they undergo AVR after the safety of watchful waiting strategy. This finding supports the current approach of watchful waiting in patients with asymptomatic or equivocally symptomatic severe AR and preserved LVEF. Considering operative risks and the fact that surgical intervention at a young age may lead to the need for future re‐operation, aortic valve surgery for chronic severe AR with NYHA II should be determined with careful decision based on the values of LVEF and LVESD.

This study has several limitations. First, it included only a small number of patients from a single center, which may represent a selected cohort. Second, because this was an observational study, there was a selection bias when patients were evaluated for the watchful waiting strategy or referral for surgery. Third, in Study A, the follow‐up period was short, which may have affected the results. However, the major role of Study A was to evaluate the safety of watchful waiting, and the short follow‐up period is inevitable because patients who underwent AVR during watchful waiting were censored at the operation date. Competing risks were not considered when comparing the survival rate of patients who were managed under the watchful waiting strategy with the age‐ and sex‐matched general population, and definitive conclusions cannot be drawn. Fourth, the LVESD cut‐off value of >45 mm may not be appropriate in patients with a higher body surface area. Further studies are needed to identify a more suitable method to correct for racial differences in LV size. ³³ Fifth, although indications for surgical treatment were determined by a multidisciplinary heart team that included cardiologists and cardiovascular surgeons, the patients were not randomized between early surgical management and watchful waiting. Therefore, we cannot exclude the possibility that unidentified confounding factors affected our results. Finally, there was no uniformity in follow‐up, assessment, and decision‐making processes among the individual patients. This limits both the relevance of the findings and the applicability of the results to larger cohorts of patients.

CONCLUSIONS

In conclusion, watchful waiting was achieved safely in asymptomatic or equivocally symptomatic patients with chronic severe AR and preserved LVEF. No patients died of cardiovascular causes during the follow‐up period. The timing of surgery, whether performed just after initial diagnosis or after watchful waiting, was not associated with a difference in postoperative cardiovascular events. In patients with small body size, preoperative LVESD >45 mm was significantly associated with an increased risk of postoperative cardiovascular events and could be an optimal cut‐off value indicating the need for surgery.

Sources of Funding

None.

Disclosures

None.

Supporting information

Tables S1–S2

Figures S1–S2