1 Introduction

Transcatheter aortic valve replacement (TAVR) has emerged as a viable alternative to surgical aortic valve replacement, even for patients with low-risk severe aortic stenosis (AS) [1, 2]. However, paravalvular regurgitation (PVR) remains a major complication associated with increased mortality after TAVR. Although moderate or severe PVR is well-established as a strong predictor of early and long-term mortality, the prognostic impact of mild PVR remains controversial [3, 4]. Variability in findings across studies may reflect differences in patient characteristics, and the prognostic factors for mild PVR have not been fully clarified.

Advances in device technology have reduced the incidence of moderate or severe PVR to less than 5% with new-generation prostheses [1, 2]. Therefore, identifying patients in whom mild PVR impacts prognosis has become clinically relevant.

Volume overload from PVR is a known contributor to poor outcomes after TAVR. However, the mechanisms through which mild regurgitation and its minor volume overload influence prognosis remain unclear. Severe AS causes chronic pressure overload, leading to concentric left ventricular (LV) hypertrophy and diastolic dysfunction. Moreover, many TAVR candidates are older adults with comorbidities, including overlapping heart failure from various etiologies [5]. Elevated preprocedural LV filling pressures due to AS or concomitant heart failure may render even mild PVR hemodynamically significant, potentially worsening clinical outcomes.

This study aimed to investigate the impact of elevated preprocedural LV filling pressure on clinical outcomes of mild PVR following TAVR.

2 Methods

3 Results

3.1 Study Population

The study flow diagram is shown in Figure 1. Among 1217 consecutive patients who underwent TAVR between October 2009 and December 2022, 939 patients were eligible for this analysis. They were categorized based on their baseline E/A ratio and residual PVR: E/A ≤ 1 with none/trace PVR (n = 466), E/A > 1 with none/trace PVR (n = 92), E/A ≤ 1 with mild PVR (n = 300), and E/A > 1 with mild PVR (n = 46). The median follow-up duration was 3.1 years (IQR: 1.9–5.0 years). The baseline and procedural characteristics of the study population are detailed in Table 1. In the overall cohort, the mean age was 83 years, and 62% of the patients were female. Patients with the E/A > 1 were more likely to be male and receiving hemodialysis, and they had higher STS scores. Preprocedural laboratory data showed higher BNP levels and lower hemoglobin, eGFR, and serum albumin levels in the E/A > 1 groups. The baseline echocardiographic parameters are summarized in Table 2. LV ejection fraction was lower, and LV mass index was higher in the E/A > 1 groups compared to that in the E/A ≤ 1 groups. Other echocardiographic measures of LV filling pressure, including DT, peak TR velocity, LA volume index, and E/e’ ratio, were also higher in the E/A > 1 group.

Table 1. Baseline and procedural characteristics.

	None/Trace PVR		Mild PVR
	E/A ≤ 1 (n = 466)	E/A > 1 (n = 92)	E/A ≤ 1 (n = 300)	E/A > 1 (n = 46)	p value
Age, yrs	83.4 ± (5.3)	82.3 ± 6.9	83.9 ± 6.1	83.1 ± 7.0	0.16
Male	149 (32.0)	37 (40.2)	125 (41.7)	22 (47.8)	0.01
Body mass index, kg/m2	22.4 ± 3.7	22.5 ± 3.7	22.0 ± 3.4	21.6 ± 2.8	0.24
Hypertension	379 (81.3)	72 (78.3)	232 (77.3)	30 (65.2)	0.06
Diabetes mellitus	148 (31.8)	35 (38.0)	82 (27.3)	13 (28.3)	0.23
Dyslipidemia	267 (57.3)	56 (60.9)	159 (53.0)	22 (47.8)	0.32
Hemodialysis	44 (9.4)	18 (19.6)	25 (8.3)	8 (17.4)	0.01
COPD	82 (17.6)	13 (14.1)	47 (15.7)	13 (28.3)	0.16
Peripheral artery disease	97 (20.8)	28 (30.4)	66 (22.0)	12 (26.1)	0.22
Prior MI	38 (8.2)	7 (7.6)	24 (8.0)	6 (13.0)	0.69
Prior PCI	104 (22.3)	28 (30.4)	57 (19.0)	12 (26.1)	0.12
Prior CABG	34 (7.3)	13 (14.1)	17 (5.7)	5 (10.9)	0.05
STS score, %	6.3 [4.4, 9.4]	8.0 [5.4, 11.0]	6.1 [4.4, 8.8]	6.8 [4.8, 13.3]	< 0.01
NYHA class III or IV	173 (37.1)	41 (44.6)	103 (34.3)	19 (41.3)	0.32
Laboratory values
BNP, pg/ml	149.5 [77.5, 348.0]	390.4 [195.9, 883.5]	161.7 [89.7, 364.6]	466.5 [217.9, 1173.8]	< 0.01
Hemoglobin, g/dl	11.5 ± 1.7	11.1 ± 1.5	11.4 ± 1.7	10.9 ± 2.0	0.04
eGFR, ml/min/1.73 m2	49.6 ± 22.6	40.6 ± 23.1	50.8 ± 24.0	39.0 ± 22.8	< 0.01
Albumin, g/dl	3.7 ± 0.5	3.6 ± 0.5	3.8 ± 0.4	3.6 ± 0.5	0.01
Procedural characteristics
Balloon-expandable valve	286 (61.4)	68 (73.9)	150 (50.0)	23 (50.0)	< 0.01
Valve with an outer skirt	218 (46.8)	46 (50.0)	165 (55.0)	31 (67.4)	0.02
Balloon pre-dilatation	311 (67.0)	59 (64.8)	209 (69.7)	31 (67.4)	0.81
Balloon post-dilatation	96 (20.8)	11 (12.0)	90 (30.2)	16 (35.6)	< 0.01
Transfemoral approach	342 (73.4)	65 (70.7)	247 (82.3)	35 (76.1)	0.02

• Note: Values are mean ± SD, median [interquartile range], or n (%)
• Abbreviations: AR = aortic regurgitation; AV = aortic valve; BNP = brain natriuretic peptide; CABG = coronary artery bypass graft; COPD = chronic obstructive pulmonary disease; eGFR = estimated glomerular filtration rate; MI = myocardial infarction; NYHA = New York Heart Association; PCI = percutaneous coronary intervention; PVR = paravalvular regurgitation; STS = Society of Thoracic Surgeons.

Table 2. Echocardiographic characteristics.

	None/Trace PVR		Mild PVR
	E/A ≤ 1 (n = 466)	E/A > 1 (n = 92)	E/A ≤ 1 (n = 300)	E/A > 1 (n = 46)	p value
LV geometry and systolic function
LV end-diastolic diameter, mm	44.8 ± 6.8	48.6 ± 7.6	43.7 ± 6.4	47.5 ± 6.3	<0.01
LV end-systolic diameter, mm	29.1 ± 7.4	33.6 ± 9.1	28.2 ± 6.8	33.3 ± 7.8	<0.01
Intraventricular septum diameter, mm	11.2 ± 2.0	11.3 ± 1.7	11.6 ± 1.7	11.3 ± 1.7	0.04
LV posterior wall diameter, mm	10.9 ± 1.8	11.2 ± 1.8	11.3 ± 1.8	11.0 ± 1.7	0.03
LV ejection fraction, %	64.3 ± 11.9	58.2 ± 14.0	65.0 ± 11.7	56.6 ± 14.5	<0.01
LV mass index, g/m2	123.0 ± 34.1	139.5 ± 33.3	123.4 ± 33.2	131.0 ± 28.2	<0.01
LV diastolic function
E-wave velocity, m/s	0.7 ± 0.2	1.1 ± 0.3	0.7 ± 0.2	1.0 ± 0.3	<0.01
A-wave velocity, m/s	1.2 ± 0.3	0.8 ± 0.3	1.1 ± 0.3	0.7 ± 0.2	<0.01
E/A ratio	0.6 ± 0.2	1.5 ± 0.6	0.6 ± 0.2	1.6 ± 0.6	<0.01
Deceleration time, ms	271.3 ± 96.4	216.8 ± 98.4	275.9 ± 96.4	202.6 ± 78.1	<0.01
Septal e’ velocity, cm/s	3.5 ± 1.2	3.9 ± 1.2	3.4 ± 1.1	3.8 ± 1.4	0.01
Lateral e’ velocity, cm/s	4.7 ± 1.8	4.8 ± 1.5	4.6 ± 1.6	5.1 ± 2.0	0.32
Average E/e’ ratio	18.9 ± 8.9	27.9 ± 13.4	18.7 ± 9.3	25.1 ± 11.7	<0.01
LA volume index, ml/m2	48.4 ± 15.6	61.7 ± 15.9	50.0 ± 27.7	62.1 ± 18.6	<0.01
Hemodynamics
AV area, cm2	0.7 ± 0.2	0.7 ± 0.2	0.7 ± 0.2	0.7 ± 0.2	0.62
AV peak velocity, m/s	4.4 ± 0.7	4.4 ± 0.9	4.6 ± 0.7	4.5 ± 0.7	0.11
AV mean gradient, mmHg	48.9 ± 16.2	49.6 ± 19.6	52.4 ± 18.3	50.7 ± 17.9	0.06
TR peak velocity, m/s	2.5 ± 0.3	2.9 ± 0.5	2.5 ± 0.4	3.0 ± 0.5	<0.01
Moderate or severe AR	22 (4.7)	10 (10.9)	11 (3.7)	3 (6.5)	0.05
Moderate or severe TR	8 (1.7)	1 (1.1)	2 (0.7)	3 (6.5)	0.03

• Note: Values are mean ± SD or n (%).
• Abbreviations: AR = aortic regurgitation, AV = aortic valve, LA = left atrial, LV = left ventricular, PVR = paravalvular regurgitation, TR = tricuspid regurgitation.

3.3 Baseline E/A Ratio and PVR

Figure 2 presents the Kaplan–Meier curves for cardiovascular death. No significant difference in cardiovascular death was observed between the none/trace PVR and mild PVR groups (Figure 2A). However, the baseline E/A > 1 group showed a significantly higher prevalence of cardiovascular death than did the E/A ≤ 1 group (Figure 2B). The cumulative incidence of cardiovascular death was highest in the E/A > 1 with mild PVR group (overall log-rank p < 0.01, Figure 2C). In a multivariable Cox proportional hazard model, the E/A > 1 with mild PVR was identified as an independent predictor of cardiovascular death (adjusted HR: 2.38; 95% CI: 1.28–4.42; p < 0.01); in contrast, E/A > 1 with none/trace PVR and E/A ≤ 1 with mild PVR were not significant predictors, using E/A ≤ 1 with none/trace PVR as the reference (Table 3).

Table 3. Cardiovascular death according to estimated baseline LV filling pressures and PVR following TAVR.

	Univariable analysis		Multivariable analysis
	HR (95% CI)	p value	HR (95% CI)	p value
E/A ratio and PVR
E/A ≤ 1 with none/trace PVR	Reference		Reference
E/A > 1 with none/trace PVR	1.62 (0.95−2.78)	0.08	1.16 (0.66−2.03)	0.61
E/A ≤ 1 with mild PVR	1.21 (0.82−1.80)	0.33	1.33 (0.89−2.00)	0.17
E/A > 1 with mild PVR	3.17 (1.77−5.68)	<0.01	2.38 (1.28−4.42)	<0.01
BNP and PVR
BNP < 200 with none/trace PVR	Reference		Reference
BNP ≥ 200 with none/trace PVR	1.84 (1.16−2.90)	<0.01	1.14 (0.68−1.91)	0.62
BNP < 200 with mild PVR	0.73 (0.38−1.43)	0.36	0.76 (0.39−1.50)	0.43
BNP ≥ 200 with mild PVR	2.94 (1.85−4.67)	<0.01	2.11 (1.26−3.53)	<0.01

• Abbreviations: BNP = brain natriuretic peptide, LV = left ventricular, PVR = paravalvular regurgitation, Ref. = reference, TAVR = transcatheter aortic valve replacement.

4 Discussion

This study demonstrated that mild PVR in patients with elevated LV filling pressure at baseline was associated with an increased risk of cardiovascular death. Specifically, mild PVR in patients with a baseline E/A ratio > 1 or BNP levels ≥ 200 pg/mL was identified as an independent predictor of cardiovascular death.

PVR is a significant complication of TAVR and has been recognized as a predictor of cardiac death related to advanced heart failure [14]. Although previous large-scale studies have consistently shown that moderate or severe PVR is associated with both short- and long-term mortality, the prognostic significance of mild PVR remains controversial [3, 4]. In patients with AS, chronic pressure overload induces LV remodeling, resulting in concentric hypertrophy and increased myocardial stiffness. A stiff, normal-sized LV is often unable to adequately compensate for the volume overload induced by PVR following TAVR. This failure to compensate further increases LV end-diastolic pressure and worsens outcomes [15]. This mechanism suggests that the prognostic impact of PVR is not solely determined by the degree of regurgitation but is also influenced by pre-existing diastolic dysfunction and elevated LV filling pressure.

In addition to the pathophysiology of AS itself, older adult patients eligible for TAVR often present with various comorbidities, such as hypertension, diabetes, or chronic kidney disease, all of which contribute to heart failure with diastolic dysfunction. Concomitant coronary artery disease is also common in these patients, and, in some cases, the etiology of AS is related to cardiac amyloidosis [5]. Given this complex clinical background, even mild PVR may adversely affect the prognosis of TAVR patients with elevated LV filling pressures due to advanced AS and other comorbidities. In the present study, mild PVR alone was not significantly associated with cardiovascular death. However, the combination of mild PVR and elevated preprocedural LV filling pressure emerged as an independent predictor of worse outcomes, even after adjusting for potential confounders. In contrast, elevated LV filling pressure alone was associated with increased mortality, consistent with findings from prior studies [16, 17]. A noteworthy finding was that elevated LV filling pressure did not correlate significantly with mortality when combined with none/trace PVR. These observations suggest that mild PVR substantially increases the risk of future cardiovascular death, specifically in patients with elevated preprocedural LV filling pressure. Conversely, patients with a compliant LV capable of tolerating mild PVR or those with reduced LV compliance but no residual PVR may experience better outcomes. The association between elevated baseline BNP levels and residual PVR further supports this hypothesis. On the other hand, this study did not fully elucidate the mechanisms linking mild PVR to adverse outcomes in patients with elevated LV filling pressure. However, postprocedural echocardiography revealed that the E/A ratio following TAVR tended to remain > 1 in patients with mild PVR. This suggests that mild PVR may hinder the expected decline in filling pressure after AS resolution. Further prospective studies are needed to clarify these mechanisms.

The findings of this study corroborate previous analyses conducted on smaller patient cohorts. In a retrospective study of 146 patients with AS with diastolic dysfunction [18], those with severe diastolic dysfunction who developed mild or greater aortic regurgitation (AR) following TAVR experienced increased mortality compared with other patients. However, the study did not clarify the prognostic impact of baseline diastolic dysfunction on mild AR alone. In another retrospective study of 418 patients undergoing TAVR [19], a preprocedural DT of E-wave < 160 ms, typically seen in cases of restrictive filling, was independently associated with mortality among patients with mild AR after TAVR. Given that the population in these previous studies likely had more severe diastolic impairment compared to our cohort, our findings are novel in suggesting that even mild elevated LV filling pressure can worsen prognosis when PVR occurs. The E/A ratio used in this analysis is known to correlate with LV end-diastolic pressure [20]. Although a normal E/A ratio > 1 is common in younger individuals with normal diastolic function, this study cohort was older, had severe AS, and exhibited other indicators of elevated LV end-diastolic pressures, such as increased E/e’ ratio, LA volume index, or TR velocity. Therefore, the E/A ratio > 1 in our cohort is reasonably considered indicative of elevated LV end-diastolic pressure. Furthermore, future research incorporating other parameters related to both systolic and diastolic function may improve prognostic accuracy.

These findings have clinical significance as they may guide therapeutic strategies for managing PVR. In this study, BEVs were associated with a lower incidence of mild PVR compared to SEVs. In contrast, the newer-generation valves with an outer skirt did not significantly reduce mild PVR compared to the previous-generation valves lacking an outer skirt. However, balloon post-dilatation was performed less frequently in the newer-generation valves, suggesting that the outer skirt may have contributed to some degree of PVR reduction. This also implies that post-dilatation remains a viable option for controlling PVR to less than mild when necessary. These findings suggest that using BEVs or performing balloon post-dilatation may be beneficial for patients with elevated LV filling pressure. In contrast, given these techniques also carry risks, including annulus rupture, coronary artery occlusion, conduction disturbance, and embolism [21, 22], aggressively reducing PVR to less than mild may not be necessary for patients without elevated filling pressure. Identifying patients for whom mild PVR could worsen prognosis is crucial for minimizing complications and improving TAVR outcomes. The E/A ratio and BNP levels are practical tools for preoperative risk stratification. Future research should validate these findings through prospective studies and aim to establish them as standard methodologies for clinical practice.

This study has some limitations. First, it was conducted at a single center with a relatively modest sample size; however, it is larger than similar previous studies. Second, the study cohort excluded patients with atrial fibrillation, prior pacemaker implantation, and significant mitral valve disease. Third, the interpretation of postprocedural echocardiographic findings is limited by missing data and potential survival bias. Finally, cardiovascular events other than mortality, such as heart failure hospitalizations, were not assessed because of limitations in medical records data collection. Further prospective studies considering additional clinical outcomes will be needed.

5 Conclusion

Elevated preprocedural LV filling pressure is independently associated with cardiovascular death in patients with mild PVR following TAVR. These findings could help identify patients whose prognosis may be adversely affected by mild PVR and guide therapeutic strategies in TAVR. Future studies should validate these results through prospective trials and explore additional clinical outcomes, such as heart failure hospitalizations, to further refine patient management strategies.

Conflicts of Interest

The authors declare no conflicts of interest.