Von Willebrand Factor Multimers in Aortic Stenosis Surgery: Dynamics and Disease Correlation

Abstract

Aim: The preoperative to early remote postoperative von Willebrand factor (VWF) large multimer index (LMI) in patients with aortic stenosis (AS) remains unclear. Assessment of LMI changes may be better understood if other valvular diseases are integrated. However, this association requires further investigation.

Methods: A quantitative time-course assessment of the VWF, including LMI, was performed in 23 patients with AS who underwent aortic valve replacement. The proposed total valve score was used to study the correlation between the LMI and valvular diseases, including diseases other than AS.

Results: The postoperative VWF LMI was significantly higher; this increase was sustained across nine measurements up to one year postoperatively. The total valve score showed a moderate negative correlation with the VWF LMI (Spearman correlation coefficient r_s = -0.5483, p = 0.0068), demonstrating a stronger negative correlation and a lower p-value than the peak flow velocity and mean pressure gradient.

Conclusions: Improved VWF LMI was maintained one year after AS surgery and correlated better with echocardiographic severity when considering other valvular diseases than AS severity alone. VWF LMI may indicate overall cardiac shear stress and treatment effectiveness in early remote stages.

Shuji Tachioka and Hideaki Kanda contributed equally to this work.

Introduction

Aortic stenosis (AS) is a valvular disease whose prevalence increases with age in developed countries. It results from the degeneration of aortic valve leaflets and accounts for ＞80% of severe cases requiring surgery¹⁾. AS affects approximately 7% of individuals aged ≥ 80 years^2-4). Bicuspid valves are the most common congenital cause, prevalent in 0.5–2% of the general population⁵⁾. Prognosis worsens as subjective symptoms, such as heart failure, syncope, and chest pain, appear^{6, 7)}. The radical treatment for AS is surgical aortic valve replacement (AVR) or transcatheter aortic valve implantation.

Von Willebrand factor (VWF), a multimeric glycoprotein critical for primary hemostasis that promotes platelet adhesion and aggregation⁸⁾, is synthesized as large multimers in vascular endothelial cells and megakaryocytes. VWF’s platelet aggregation capacity increases with molecular weight. Plasma protease, disintegrin, and metalloproteinase with thrombospondin type 1 motif 13 cleave VWF multimers into smaller molecules in a shear stress-dependent manner.

The use of left ventricular assist devices and extracorporeal membrane oxygenation has highlighted that acquired von Willebrand syndrome (AVWS) may cause hemorrhagic complications⁹⁾. In AS, shear stress from blood passing through the stenotic valve disrupts VWF large multimers, leading to AVWS. Recent studies have suggested a positive correlation between AS and AVWS severity^{10, 11)}.

Although AS pathophysiology is well understood, AVWS can also arise from mitral regurgitation (MR) and other valvular diseases^12-15). Furthermore, paravalvular leakage after prosthetic valve replacement¹⁶⁾ and congenital heart disease^17-19) can contribute to AVWS.

Despite reports of VWF multimer analysis in AS, few studies have assessed long-term quantitative VWF multimer levels pre- and postoperatively²⁰⁾. Additionally, no studies have quantitatively linked VWF to the severity of AS and other valvular diseases.

Aim

Herein, we analyzed VWF levels in patients with AS for up to one-year post-surgery and explored the correlation between the VWF large multimer index (LMI), valvular disease, and the proposed total valve score.

Methods

Study Cohorts

Consecutive patients who underwent AVR for AS at Kagoshima University Hospital and the National Hospital Organization Kagoshima Medical Center were recruited between September 2017 and December 2018. The research ethics committees of the Kagoshima University Hospital (170030, July 27, 2017) and the National Hospital Organization Kagoshima Medical Center (30-34, July 5, 2018) approved this study. Written informed consent was obtained from all participants before surgery. This study adhered to the ethical standards of the Human Experimentation Committees of the institutions involved and the 1975 Declaration of Helsinki. Patients undergoing dialysis and those who declined to participate were excluded, resulting in 23 analyzed cases. For the time-course analysis, 19 patients who were followed up for one year postoperatively were included.

Blood Sample Collection

Preoperative venous blood samples were collected using a 21G needle. Postoperative arterial blood samples were collected from the radial artery during the patient’s stay in the intensive care unit, and venous blood samples were collected using a 21G needle in the general ward. Blood samples were collected immediately postoperatively (0d), on days 1 (1d), 3 (3d), 7 (7d), 14 (14d), 1 month (1m), 3 months (3m), 6 months (6m), and 1 year (1y) postoperatively. Each sample was divided into three tubes: a serum-separating tube, a 5.0-mL tube containing 0.5 mL of 3.2% sodium citrate for plasma, and an EDTA-2K tube (Venoject II; Terumo Corp., Tokyo, Japan). The serum tubes were incubated for at least 30 min at 20℃ to allow clotting. Both serum and plasma tubes were centrifuged twice at 25℃ and 1710 × g for 10 min. Supernatants were carefully aliquoted and stored at −80℃. Complete blood counts were measured from the blood collected in EDTA-2K tubes using an XE-5000 automated hematology analyzer.

Echocardiography

Echocardiographic data obtained preoperatively and seven days postoperatively were used. Valvular disease severity was defined according to the 2020 American College of Cardiology/American Heart Association (AHA) guidelines²¹⁾. Data from additional parameters, such as ejection fraction and right ventricular systolic pressure, were also collected.

Analysis of von Willebrand Factor

VWF multimer analysis was performed using 1.0% agarose gel electrophoresis, followed by western blotting with a polyclonal anti-VWF antibody (DAKO, Denmark). Siemens standard human plasma (Siemens Healthineers AG, Erlangen, Germany) was used as a control.

Quantitative evaluation of VWF large multimer loss was performed using two parameters reported by Tamura et al.¹¹⁾: the VWF large multimer ratio (LMR) and LMI. VWF large multimers were quantified using gel electrophoresis, and each lane was scanned individually using ImageJ software, dividing the VWF multimers into three sizes (large, medium, and small). VWF LMR was calculated as the ratio of the large VWF multimer area to the total VWF area. VWF LMI was determined as the ratio of the large multimeric area of a patient sample to that of the control plasma analyzed in the adjacent lane (Fig.1).

Fig.1.

Determination of the von Willebrand factor large multimer ratio and index using Siemens standard human plasma control (C) and patient plasma (P)

The VWF antigen (VWF: Ag) was measured using ELISA. Plasma was diluted 1:99 in phosphate-buffered saline (PBS) containing 0.05% Tween 20 (PBST) and 1% bovine serum albumin. The diluted samples were then applied to 96-well plates pre-coated with 0.2 µg/well anti-VWF mouse monoclonal antibody (10973-MM05, Sino Biological, China). Before the sample addition, plates were blocked with PBST containing 1% bovine serum albumin. After a 2-hour incubation of the samples at 20℃, the plates were washed four times with PBST. Subsequently, 0.2 µg/well of HRP-conjugated anti-VWF rabbit monoclonal antibody (10973-R111-H; Sino Biological, China) was added to each well. Following a 1-hour incubation and another wash step, the wells were treated with EzELISA TMB chromogenic solution (WSE-7145, ATTO, Japan) according to the manufacturer’s protocol. Finally, the absorbance was measured at 450 nm to quantify VWF: Ag levels. A calibration curve was constructed using Siemens standard human plasma samples.

Total Valve Score

Although VWF LMI may decrease in valvular diseases other than AS, which generates shear stress^12-15), no single variable integrates the overall severity of multiple valvular diseases. To address this gap, we propose a novel “total valve score,” which comprehensively assesses the impact of left heart valvular diseases: AS, aortic regurgitation, mitral stenosis, and mitral regurgitation (MR). Table 1 presents a detailed breakdown of the scoring systems.

Table 1.Total valve score

	＜Trace	Mild	Moderate	Severe	Very severe
Aortic stenosis	0	1	3	5	7
Aortic regurgitation	0	1	3	5
Mitral stenosis	0	1	3	5
Mitral regurgitation	0	1	3	5
					/Total 22

Assuming significant shear stress effects in moderate or high-severity left ventricular disease, scores of 1, 3, 5, and 7 were assigned to mild, moderate, severe, and very severe (AS only), respectively. As severity above moderate in other valvular diseases, such as MR, is associated with VWF reduction in large multimers¹⁵⁾, the scores were weighted more heavily for severity greater than moderate. The maximum score for the four valvular left heart diseases was 22 points.

Statistical Analysis

Statistical analyses were performed using GraphPad Prism version 10 (GraphPad Prism Software Inc., San Diego, CA, USA) and MATLAB version R2022b for bootstrapped confidence intervals. Data are presented as the mean±standard deviation for normally distributed data or as the median with the interquartile range (IQR) for non-normally distributed data. The coefficient expresses the best correlation estimate in the correlation studies, whereas the 95% confidence interval (CI) expresses the estimate’s precision. All correlations with the VWF LMI were indicated using Spearman’s rank correlation coefficient, rs, to provide consistency with a key reference that studies the relationship between LMI and AS²²⁾ across all correlations in this study. The Pearson correlation coefficient was not appropriate for the total valve score, ordinal scale, and nonparametric continuous data. The upper and lower bounds of the 95% CI were calculated by bootstrapping with 10,000 re-samples with replacement. A p-value for correlation expresses the probability of achieving the calculated correlation-coefficient value if the null hypothesis (of no correlation in the population) is true; a p-value ＜0.05 is considered statistically significant and provides strong evidence against the lack of correlation in the population.

For between-group comparisons of continuous variables, Student’s t-test was used for normally distributed data; otherwise, the Mann–Whitney U test was used. The Friedman test, a nonparametric analysis of variance, was used to assess differences across all blood collection points. Following a significant Friedman test result, Dunn’s multiple comparison test was used as a post hoc test to compare specific preoperative and postoperative blood collection points.

Results

Clinical Characteristics and Outcomes

Of the 23 patients with AS, 22 had at least moderate AS. The average age of the patients was 75 years, with a nearly equal male-to-female ratio (48% male). Most patients had preexisting medical conditions, including hypertension (70%), dyslipidemia (61%), and a smoking history (55%). Preoperative blood tests revealed an average hemoglobin level of 12.7±1.5 g/dL, with no cases below 9.0 g/dL (Table 2).

Table 2.Clinical characteristics of the study population

Clinical characteristics (N = 23)
Age (years)	75±8
Male (%)	48
Causal factors
Hypertension (%)	70
Diabetes mellitus (%)	22
Dyslipidemia (%)	61
Smoking (%)	55
Blood tests
Hemoglobin (g/dL)	12.7±1.5
Platelet count (×10^3/μL)	176.9±45.2
Total bilirubin (mg/dL)	0.82±0.26
Aspartate aminotransferase (U/L)	25.2±12.6
Alanine transaminase (U/L)	20.1±10.2
Creatinine (mg/dL)	0.99±0.36

Table 3 summarizes the preoperative echocardiographic data. Aortic valve parameters showed a peak flow velocity (PFV) of 4.6±0.8 m/s, peak pressure gradient (PG) of 86.1±30.5 mm Hg, mean PG of 51.5±19.1 mm Hg, aortic valve area of 0.7±0.2 cm², and aortic valve area index of 0.4±0.1 cm²/m².

Table 3.Preoperative echocardiography data (N = 23)

Left ventricular ejection fraction (%)	68±9
Aortic valve
Peak flow velocity (m/s)	4.6±0.8
Peak pressure gradient (mm Hg)	86.1±30.5
Mean pressure gradient (mm Hg)	51.5±19.1
Aortic valve area (cm2)	0.7±0.2
Aortic valve area index (cm2/m2)	0.4±0.1
Severity of stenosis or regurgitation, trace/mild/moderate/severe/very severe (%)
Aortic stenosis	0/4/9/52/35
Aortic regurgitation	13/65/22/0
Mitral stenosis	100/0/0/0
Mitral regurgitation	9/87/0/4
Tricuspid regurgitation	30/65/4/0
Right ventricular systolic pressure (mm Hg)	34.4±7.9

Table 4 details the surgical procedures and indications for AS. AVR alone was performed in 52% of cases; the remainder involved complex surgeries combined with AVR. Not all patients had severe AS; some surgeries were primarily indicated for conditions like aortic enlargement or ischemic heart disease.

Table 4.Operative data (N = 23)

Procedure	Number (%)
AVR§	12 (52%)
AVR + Ascending aortic replacement	4 (17%)
AVR + Maze or PV§ isolation	3 (13%)
AVR + CABG§	1 (4%)
AVR + MVP§	1 (4%)
AVR + TAP§ + Septal myectomy	1 (4%)
Bentall (includes AVR) + CABG	1 (4%)
Main operative indication	Number (%)
Symptomatic severe AS§	17 (74%)
Asymptomatic severe AS	1 (4%)
Symptomatic severe AS & severe MR§	1 (4%)
Symptomatic moderate AS & moderate AR§	1 (4%)
Thoracic aortic aneurysm	2 (9%)
Angina pectoris	1 (4%)

AVR^§ aortic valve replacement; PV^§ pulmonary vein; CABG^§ coronary artery bypass graft; MVP^§ mitral valvuloplasty; TAP^§ tricuspid annuloplasty; AS^§ aortic stenosis; MR^§ mitral regurgitation; AR^§ aortic regurgitation.

All patients received blood transfusions at median doses of 1120 (IQR, 1120-1400), 1440 (IQR, 960-1920), and 0 (IQR, 0-400) mL of red blood cells, fresh frozen plasma, and platelet concentrate, respectively. No deaths occurred within 30 days of surgery. The most frequent major complications were postoperative paroxysmal atrial fibrillation (39%), pleural effusion requiring thoracentesis (4%), wound complications (4%), respiratory failure requiring prolonged ventilation (＞48 h) (13%), and cerebral infarction (9%).

Dynamics of VWF LMI

Western blotting was used to assess VWF multimers, and VWF LMR and LMI were calculated for each blood collection point, as previously described. Preoperatively, the median LMR was 17.3% (IQR: 15.75–23.65%), and the median LMI was 88.1% (IQR: 75.7–109.2%). A loss of VWF large multimers (VWF LMI ＜80%)²²⁾, were identified preoperatively in 7 of 23 patients (30%).

The median postoperative VWF LMI was higher than the preoperative level (Fig.2A). These median values were the best estimates of LMI at each time point (evaluated in 19 patients with data up to one year postoperatively). The time course suggested that LMI increased postoperatively and then maintained over time. The Friedman test showed a statistically significant (p = 0.0088) difference across the time points. The post hoc Dunn’s multiple comparison tests revealed statistically significant differences between the preoperative LMI and those at 0d (p = 0.0203), 7d (p = 0.0001), 1m (p = 0.003), 3m (p = 0.0117), 6m (p = 0.0222), and 1y (p = 0.0289) (Fig.2A).

Fig.2. Time-course of the von Willebrand factor large multimer index (VWF LMI) in median values and IQRs (A, n = 19)

B and C are grouped by preoperative VWF LMI ＜80% (n = 5) or ≥ 80% (n = 14). D and E show the time-course of VWF LMR and VWF antigen (VWF: Ag). ^＊Statistically significant increase from the preoperative level.

When patients were grouped by preoperative VWF LMI (＜80% or ≥ 80%), the median postoperative LMI increase was more pronounced in those with a preoperative LMI ＜80%, suggesting that LMI increased postoperatively and then maintained over time (Fig.2B). The Friedman test revealed a statistically significant (p = 0.016) time-course change for this subset of five patients, although it was less significant than that for the entire cohort. Dunn’s test showed statistically significant differences between the preoperative LMI and postoperative values at 3d (p = 0.0311), 7d (p = 0.0035), 3m (p = 0.0035), and 1y (p = 0.0311).

For the 14-patient group with preoperative LMI ≥ 80%, the time course showed a smaller increase because of surgery. The Friedman test was not significant (p = 0.0986), and only 7d showed significance (p = 0.0246) in the Dunn test (Fig.2C). Thus, postoperative changes were less significant when considering only LMI ≥ 80%.

The median VWF LMR increased postoperatively and remained high throughout the year. The Friedman test revealed a statistically significant change over time (p = 0.0075). Post hoc Dunn’s tests revealed significant differences between the preoperative LMR at 0d (p = 0.0222), 7d (p = 0.0080), 1m (p = 0.0008), and 3m (p = 0.0107). Compared with LMI, statistical significance was lost at 6m and 1y (Fig.2D).

A time-course dynamics analysis in VWF LMI and LMR to examine the median difference between the preoperative values and each follow-up time point (ΔVWF) (Supplementary Fig.1) showed that for the entire group (n = 19), the median LMI difference from preoperative to one-year postoperative assessment was 18.4% (IQR: 2.5–27.75%). When stratified by preoperative VWF LMI (＜80% vs ≥ 80%), the increase in LMI at 1 year postoperatively was more pronounced in the ＜80% group (n = 5). In this group, the median difference in LMI at 1 year was 28.1% (IQR: 22.6–48.3%), compared with 6.95% (IQR: −0.7–6.3%) in those with preoperative LMI ≥ 80% (n = 14). The median difference in VWF LMR over 1 year was only 3.4% (IQR: −0.7–6.3%, n = 19).

Supplementary Fig.1.

Time-course of the differences in von Willebrand factor (ΔVWF LMI), large multimer index (LMI), and large multimer ratio (LMR)

The median VWF: Ag levels exhibited a time course distinct from those of LMI and LMR. The latter exhibited one-year values that were among the highest over time, whereas the former peaked early postoperatively (days 1 and 3) and diminished by one year (Fig.2E). The Friedman test indicated a significant change (p = 0.0001) over time, with the post hoc Dunn’s test confirming statistical significance for the four highest median values: 1d (p＜0.0001), 3d (p＜0.0001), 7d (p＜0.0001), and 14d (p = 0.0003). Unlike LMI and LMR, there were no immediate postoperative changes in preoperative VWF levels.

Correlation between VWF LMI and Valvular Disease

AS severity was defined by the American Heart Association (AHA) using the mean aortic PG (mPG) and PFV²¹⁾. We analyzed the correlations between these quantities and VWF LMI preoperatively and at 7 days postoperatively. The strongest correlations appeared preoperatively, with surgical remediation concentrated in the data clusters and reduced correlations (Fig.3).

Fig.3. Correlation between the von Willebrand factor large multimer index (VWF LMI) and heart disease, shown preoperatively (A–C) and at 7 days postoperatively (D–F)

Dashed lines denote the American Heart Association severity standards for the mean pressure gradient (A & D) and peak flow velocity (B & E). The total valve score (C & F) shows the strongest preoperative correlation.

The preoperative VWF LMI showed a weak negative correlation with PFV (r_s = -0.2853, 95% CI [-0.6316, 0.1480], p = 0.1870; Fig.3A) and mPG (r_s = -0.3617, 95% CI [-0.6730, 0.0728], p = 0.0906; Fig.3B). Each r_s value was the best correlation estimate despite its relatively large imprecision, as denoted by the CI limits and higher p-values.

Our proposed total valve score exhibited a stronger negative correlation with preoperative VWF LMI than echocardiographic parameters and an improved p-value: r_s = –0.5483, 95% CI (–0.7668, –0.2174), p = 0.0068 (Fig.3C). This provides strong evidence of sampling errors. Supplementary Table 1 shows the total preoperative valve scores of all patients.

Supplementary Table 1. Total preoperative valve scores in 23 patients with aortic stenosis

	AS§ severity	score	AR§ severity	score	MR§ severity	score	Total
1	very severe	7	mild	1	mild	1	9
2	severe	5	Ⅱ	1	mild	1	7
3	very severe	7	none	0	trace	0	7
4	very severe	7	mild	1	mild	1	9
5	severe	5	mild	1	mild	1	7
6	mild	1	mild	1	mild	1	3
7	very severe	7	Ⅱ	1	mild	1	9
8	severe	5	mild	1	mild	1	7
9	severe	5	mild	1	mild	1	7
10	very severe	7	mild	1	Ⅱ	1	9
11	severe	5	mild	1	severe	5	11
12	severe	5	trace	0	mild	1	6
13	severe	5	mild	1	mild	1	7
14	severe	5	moderate	3	mild	1	9
15	severe	5	Ⅱ	1	Ⅱ	1	7
16	severe	5	moderate	3	Ⅱ	1	9
17	severe	5	mild	1	mild	1	7
18	very severe	7	moderate	3	mild	1	11
19	moderate	3	mild	1	mild	1	5
20	severe	5	trace	0	trace	0	5
21	very severe	7	mild	1	mild	1	9
22	moderate	3	moderate	3	mild	1	7
23	very severe	7	moderate	3	mild	1	11

AS^§ aortic stenosis; AR^§ aortic regurgitation; MR^§ mitral regurgitation.

Severity “II” was included in mild AS.

Mitral stenosis was excluded from scoring because none of the patients had stenosis.

The 7-day postoperative correlation data are shown in Fig.3D–F. Surgery improved the echocardiographic parameters, with all mean PG and PFV below AHA’s “severe” thresholds (40 mm Hg and 4.0 m/s, respectively). The resulting scatter plots reveal tighter data clustering at higher LMI values, lower correlation coefficients, and higher p-values.

Preoperative peak PG, aortic valve area index, and LMI showed weak, nonsignificant, and negative correlations. There was no correlation between LMI and the preoperative aortic valve area or hemoglobin levels (Supplementary Fig.2).

Supplementary Fig.2.

The correlations between preoperative von Willebrand factor (VWF), large multimer index (LMI), other echocardiographic data (peak pressure gradient [PG], aortic valve area, and aortic valve area index), and hemoglobin (Hb)

Discussion

This study is the first to examine postoperative VWF dynamics using VWF LMI for up to one year after AS surgery in multiple patients. VWF LMI, which was initially reduced in patients with AS, significantly improved postoperatively and remained statistically higher than preoperative levels after one year, particularly in those with preoperative LMI ＜80%.

While the VWF LMR was also analyzed preoperatively and at one year postoperatively, its statistical significance was lost at 6 months and 1 year, suggesting that the LMI may be a better metric at the one-year follow-up. As previously noted¹¹⁾, LMR exhibits greater variability, and our data confirmed that the LMR has higher variability than the LMI, reinforcing its reliability.

In this study, VWF:Ag was also measured, but as shown in Fig.2, VWF:Ag showed an increase 1 day later than VWF LMI compared to preoperative levels. VWF:Ag is thought to increase postoperatively due to secretion caused by surgical invasion and inflammation, as well as blood transfusion. The following reasons are thought to explain why VWF:Ag rises later than VWF LMI. Although VWF is secreted in very high amounts at 0d, its clearance is further enhanced by the effects of cardiopulmonary bypass, so it is estimated that VWF:Ag, which reflects the ‘total amount’ of VWF, may rise slightly later. The fact that VWF LMI rises before VWF:Ag is likely because the aortic valve stenosis and cardiopulmonary bypass, which causes shear stress, is relieved, and because LMI evaluates the ‘distribution’ or ‘proportion’ of VWF multimers rather than antigen quantity, it likely reflects the characteristics of VWF supplemented by autologous secretion or transfusion more rapidly.

The median difference between preoperative and one-year postoperative VWF LMI was 18.4% (IQR: 2.5–27.75%). This difference was even greater (28.1%, IQR: 22.6–48.3%) in patients with a preoperative VWF LMI of ＜80% (Supplementary Fig.1). The median difference in VWF LMR was only 3.4% (IQR: -0.7–6.3%), indicating that VWF LMI more clearly defined the surgical effect size.

Those with a pre-VWF LMI of ＜80% demonstrated a significant postoperative increase in VWF LMI compared with preoperative values, including at one year. Based on the small sample size, this aligns with the pathophysiology of AS and suggests a surgical improvement of shear stress. The absence of significant differences in the group with a preoperative LMI ≥ 80% supports the idea that VWF LMI reflects the therapeutic efficacy of AVWS caused by AS. Taken together, VWF LMI may be a useful indicator of both AVWS severity and treatment effectiveness in the early remote stages.

It is important to acknowledge that VWF multimer durability can vary considerably among individuals, even under similar mechanical shear stress conditions²³⁾. Therefore, severe AS may not necessarily result in reduced VWF LMI. Nevertheless, the VWF LMI remains a valuable indicator of the shear stress impact of AS and the severity of AVWS.

Traditionally, LMR has been used to assess VWF large multimer loss¹⁰⁾. However, owing to immunostaining variability, LMR values often overlap between healthy individuals and patients with AS. To address this limitation, the VWF LMI was introduced¹¹⁾, which exhibits a strong linear correlation with AS severity, making it a more useful biomarker. Therefore, the VWF LMI was employed in this study.

Our study included patients with preoperative VWF LMI ＜80%, indicating a large VWF multimer reduction²²⁾. As previously reported, 67.7% of patients in this category had severe AS¹¹⁾. Our study showed a lower proportion (30%) of patients with preoperative VWF LMI ＜80%. This difference may be because our study included patients with varying AS severities.

The total valve score exhibited a stronger and statistically significant negative correlation with VWF LMI than with individual echocardiographic parameters, likely due to its consideration of the cumulative impact of valvular disease across the left heart rather than AS alone.

Acknowledging the influence of comorbidities other than AS, such as aortic regurgitation and MR findings, is important. MR severity correlates with VWF multimer loss, particularly in moderate and severe cases^{14, 15)}. Similarly, complex heart diseases are associated with an increased incidence of AVWS¹⁸⁾. Therefore, integrating valvular diseases across the left ventricle likely provides a more accurate reflection of the pathophysiology underlying AVWS, as our findings suggest. Although the proposed scoring system requires further validation, our findings suggest that comprehensive left heart valve assessment may better elucidate the pathophysiology of shear stress than AS evaluation alone.

Limitations

This single-center study included a small participant cohort, which widened the confidence intervals and increased the p-values in correlation studies. Blood samples were collected more than one year prior to the analysis, precluding the measurement of VWF activity and other factors that might influence test reliability.

One of the problems with the total valve score proposed in this study is that it treats stenotic and regurgitant diseases similarly. The relationship between the severity of each valvular disease and the VWF LMI is expected to differ, and the weighting of each valvular disease score should be considered. However, evidence regarding the correlation between the severity of each valvular disease and VWF LMI is still insufficient, and there are currently no reports of the quantitative evaluation of VWF, particularly in cases of aortic regurgitation. Further analysis of the VWF should be performed for each valvular disease, and the total valve score should be improved, as necessary. In addition, there were no hemorrhagic complications before or after surgery and no preoperative anemia. Therefore, it is impossible to link clinical outcomes to the results of this study, and it is necessary to increase the number of cases and verify the results further.

Conclusion

This study investigated VWF dynamics in patients with AS who underwent AVR preoperatively to one year postoperatively. Preoperative VWF LMI levels increased after AVR and remained elevated for up to one year postoperatively, especially in patients with a preoperative LMI ＜80%. Interestingly, VWF: Ag peaked on day 1 and decreased steadily toward the one-year mark.

Preoperatively, VWF LMI was negatively correlated with mPG and PFV, albeit with broad confidence levels and high p-values. The total valve score introduced a stronger correlation and lower p-value, suggesting that the VWF LMI may reflect AS severity and overall shear stress across heart disease. The high performance of the total valve score has motivated its further development and application. These findings suggest that VWF LMI may be a useful indicator of AVWS severity related to AS and treatment effectiveness in the early remote stages.

Acknowledgements

The authors wish to acknowledge Dr. Shuji Nagatomi and Dr. Kazuya Terazono (Department of Cardiovascular Surgery, Graduate School of Medical and Dental Sciences, Kagoshima University) for their help in modifying the scoring system used in this study. We thank Editage (Cactus Communications, Mumbai, India) for the English language editing.

Conflicts of Interest

The authors have no conflicts of interest to declare.

Author Contributions

ST: conceptualization, data curation, formal analysis, funding acquisition, validation, investigation, methodology, visualization, original draft writing, writing, reviewing, and editing.

HK: Blood sample collection, conceptualization, data curation, formal analysis, funding acquisition, validation, investigation, methodology, visualization, original draft writing, writing, reviewing, and editing.

SH: Formal analysis, validation, investigation, and methodology.

MY: conceptualization, data curation, formal analysis, funding acquisition, validation, investigation, methodology, writing, reviewing, and editing.

MS: Validation, investigation, and methodology.

DM: Formal analysis, funding acquisition, validation, writing, reviewing, and editing.

KT, KM, KM: Surgical procedure.

HH: Conceptualization, validation, investigation, methodology.

TH: Conceptualization, data curation, methodology, project administration, supervision, writing, reviewing, and editing.

YS: conceptualization, methodology, project administration, supervision, writing, reviewing, and editing.

Institutional Review Board Number

Kagoshima University Hospital: 170030, July 27, 2017.

National Hospital Organization Kagoshima Medical Center: 30-34, July 5, 2018.

Funding Sources

This work was supported by Health Labour Sciences Research Grants [grant numbers 21FC1008 and 24FC1016], the Japan Society for the Promotion of Science (JSPS) KAKENHI grant numbers JP23K08275 (ST) and JP22H02973 (MY), a Terumo Life Science Foundation Research Grant (HK), and a JSPS International Research Fellowship grant number S23122 (DM).

References

• 1)  Iung B, Baron G, Butchart EG, Delahaye F, Gohlke-Bärwolf C, Levang OW, Tornos P, Vanoverschelde JL, Vermeer F, Boersma E, Ravaud P, Vahanian A: A prospective survey of patients with valvular heart disease in Europe: the euro Heart Survey on Valvular Heart Disease. Eur Heart J, 2003; 24: 1231-1243
• 2)  Nkomo VT, Gardin JM, Skelton TN, Gottdiener JS, Scott CG, and Enriquez-Sarano M: Burden of valvular heart diseases: A population-based study. Lancet, 2006; 368: 1005-1011
• 3)  Eveborn GW, Schirmer H, Heggelund G, Lunde P, and Rasmussen K: The evolving epidemiology of valvular aortic stenosis. The Tromsø study. Heart, 2013; 99: 396-400
• 4)  Danielsen R, Aspelund T, Harris TB, and Gudnason V: The prevalence of aortic stenosis in the elderly in Iceland and predictions for the coming decades: the AGES–Reykjavík study. Int J Cardiol, 2014; 176: 916-922
• 5)  Siu SC and Silversides CK: Bicuspid aortic valve disease. J Am Coll Cardiol, 2010; 55: 2789-2800
• 6)  Horstkotte D and Loogen F: The natural history of aortic valve stenosis. Eur Heart J, 1988; 9(suppl E): 57-64
• 7)  Vahanian A, Beyersdorf F, Praz F, Milojevic M, Baldus S, Bauersachs J, Capodanno D, Conradi L, De Bonis M, De Paulis R, Delgado V, Freemantle N, Gilard M, Haugaa KH, Jeppsson A, Jüni P, Pierard L, Prendergast BD, Sádaba JR, Tribouilloy C, Wojakowski W, ESC/EACTS Scientific Document Group: 2021 ESC/EACTS Guidelines for the management of valvular heart disease. Eur Heart J, 2022; 43: 561-632
• 8)  Hassan MI, Saxena A, and Ahmad F: Structure and function of von Willebrand factor. Blood Coagul Fibrinolysis, 2012; 23: 11-22
• 9)  Loscalzo J: From clinical observation to mechanism—Heyde’s syndrome. N Engl J Med, 2012; 367: 1954-1956
• 10)  Vincentelli A, Susen S, Le Tourneau T, Six I, Fabre O, Juthier F, Bauters A, Decoene C, Goudemand J, Prat A, Jude B: Acquired von Willebrand syndrome in aortic stenosis. N Engl J Med, 2003; 349: 343-349
• 11)  Tamura T, Horiuchi H, Imai M, Tada T, Shiomi H, Kuroda M, Nishimura S, Takahashi Y, Yoshikawa Y, Tsujimura A, Amano M, Hayama Y, Imamura S, Onishi N, Tamaki Y, Enomoto S, Miyake M, Kondo H, Kaitani K, Izumi C, Kimura T, Nakagawa Y: Unexpectedly high prevalence of acquired von Willebrand syndrome in patients with severe aortic stenosis as evaluated with a novel large multimer index. J Atheroscler Thromb, 2015; 22: 1115-1123
• 12)  Weinstein M, Ware JA, Troll J, and Salzman E: Changes in von Willebrand factor during cardiac Surgery: effect of desmopressin acetate. Blood, 1988; 71: 1648-1655
• 13)  Wan SH, Liang JJ, Vaidya R, Blackshear JL, and Chen D: Acquired von Willebrand syndrome secondary to mitral and aortic regurgitation. Can J Cardiol, 2014; 30: 1108.e9-1108.e10
• 14)  Blackshear JL, Wysokinska EM, Safford RE, Thomas CS, Shapiro BP, Ung S, Stark ME, Parikh P, Johns GS, Chen D: Shear stress-associated acquired von Willebrand syndrome in patients with mitral regurgitation. J Thromb Haemost, 2014; 12: 1966-1974
• 15)  Takiguchi H, Miura M, Shirai SI, Soga Y, Hanyu M, Sakaguchi G, Soga Y, Arai Y, Watanabe S, Kimura T, Takahama H, Yasuda S, Nakayoshi T, Fukumoto Y, Yaoita N, Shimokawa H, Sakatsume K, Saiki Y, Kaikita K, Tsujita K, Tamura T, Doman T, Yamashita M, Suzuki M, Eura Y, Kokame K, Hayakawa M, Matsumoto M, Okubo N, Sugawara S, Fujimaki SI, Kawate Y, Ando K, Horiuchi H: Mitral regurgitation is associated with similar loss of von Willebrand factor large multimers but lower frequency of anemia compared with aortic stenosis. Res Pract Thromb Haemost, 2024; 8: 102431
• 16)  Pérez-Rodríguez A, Pinto JC, Lourés E, Rodríguez-Trillo A, Cuenca JJ, Batlle J, López-Fernández MF: Acquired von Willebrand syndrome and mitral valve prosthesis leakage. A pilot study. Eur J Haematol, 2011; 87: 448-456
• 17)  Onimoe G, Grooms L, Perdue K, and Ruymann F: Acquired von Willebrand Syndrome in congenital heart disease: does it promote an increased bleeding risk? Br J Haematol, 2011; 155: 622-624
• 18)  Waldow HC, Westhoff-Bleck M, Widera C, Templin C, and von Depka M: Acquired von Willebrand syndrome in adult patients with congenital heart disease. Int J Cardiol, 2014; 176: 739-745
• 19)  Loeffelbein F, Funk D, Nakamura L, Zieger B, Grohmann J, Siepe M, Kroll J, Stiller B: Shear-stress induced acquired von Willebrand syndrome in children with congenital heart disease. Interact Cardiovasc Thorac Surg, 2014; 19: 926-932
• 20)  Yamashita K, Yagi H, Hayakawa M, Abe T, Hayata Y, Yamaguchi N, Sugimoto M, Fujimura Y, Matsumoto M, Taniguchi S: Rapid restoration of thrombus formation and high-molecular-weight von Willebrand factor multimers in patients with severe aortic stenosis after valve replacement. J Atheroscler Thromb, 2016; 23: 1150-1158
• 21)  Otto CM, Nishimura RA, Bonow RO, Carabello BA, Erwin JP, Gentile F, Jneid H, Krieger EV, Mack M, McLeod C, O’Gara PT, Rigolin VH, Sundt TM, Thompson A, Toly C: 2020 ACC/AHA guideline for the management of patients with valvular heart disease: executive summary: A report of the American College of Cardiology/American Heart Association joint committee on clinical practice guidelines. Circulation. https: //www.ahajournals.org/doi/10.1161/CIR.0000000000000923: e35-e71, 2021; 143: e35-e71
• 22)  Okubo N, Sugawara S, Fujiwara T, Sakatsume K, Doman T, Yamashita M, Goto K, Tateishi M, Suzuki M, Shirakawa R, Eura Y, Kokame K, Hayakawa M, Matsumoto M, Kawate Y, Miura M, Takiguchi H, Soga Y, Shirai S, Ando K, Arai Y, Nakayoshi T, Fukumoto Y, Takahama H, Yasuda S, Tamura T, Watanabe S, Kimura T, Yaoita N, Shimokawa H, Saiki Y, Kaikita K, Tsujita K, Yoshii S, Nakase H, Fujimaki SI, Horiuchi H: Von Willebrand factor Ristocetin co-factor activity to von Willebrand factor antigen level ratio for diagnosis of acquired von Willebrand syndrome caused by aortic stenosis. Res Pract Thromb Haemost, 2024; 8: 102284
• 23)  Sakatsume K, Akiyama M, Sakota D, Hijikata W, Horiuchi H, Maruyama O, Saiki Y: Individual variability in von Willebrand factor fragility in response to shear Stress: A possible clue for predicting bleeding risk. ASAIO J, 2022; 68: 1128-1134