Abstract

Abstract 3377

Background

Bleeding from AVWS may occur in aortic stenosis (AS), hypertrophic cardiomyopathy (HCM), mitral regurgitation (MR), and prosthetic valve dysfunction (PVD) due to loss of von Willebrand factor (VWF) high molecular weight multimers caused by shear stress. Traditional laboratory testing for AVWS, including VWF antigen (VWF:Ag) and ristocetin cofactor activity lack sensitivity in this setting, and VWF multimer analysis, a costly and time-consuming test, must be used for confirmative diagnosis. The goal of this study is to develop a clinical and laboratory approach to predict the likelihood of AVWS, and consequently obviate the need for multimer analysis.

Methods

160 patients with or without a bleeding history were prospectively enrolled with mild, moderate, or severe AS (n=65), HCM (n=32), MR (n=25), or with heart valve replacement (n=38). These patients were investigated at the time of clinically indicated echocardiography with VWF latex immunoturbidic activity (VWF:Ltx), VWF:Ag, (abnormal versus normal cutoff of VWF:Ltx/VWF:Ag is <0.80) platelet function analyzer 100 collagen ADP (PFA-CADP) and VWF multimer analysis. Single variable logistic regression analysis was performed to estimate clinical or laboratory variables that were strongly associated with the probability of abnormal multimers, including results of echocardiographic peak velocity and PFA-CADP. Associations with abnormal multimers were summarized with estimated relative risk (RR) and 95% CI. A scoring system was then created to evaluate the risk of abnormal multimers among cardiac patients.

Results

Overall, 50% of patients had abnormal multimers, indicating likely AVWS. The estimated probability of abnormal multimers showed a significant association between peak velocity, PFA-CADP and VWF:Lx/Ag (Table 1).

Table 1
Variable % Abnormal multimers (95% CI) P-value 
Peak velocity, m/sec   <0.001 
<3.0 53 11 (5–23)  
3.0–3.9 33 45 (30–62)  
4.0–4.9 44 73 (58–84)  
≥5.0 30 90 (73–97)  
PFA-CADP, sec   <0.001 
≤120 76 24 (15–34)  
121–180 36 61 (45–75)  
>180 48 83 (70–91)  
VWF:Ltx/Ag ratio   <0.001 
<0.80 39 82 (67–91)  
≥0.80 121 40 (31–49)  
Variable % Abnormal multimers (95% CI) P-value 
Peak velocity, m/sec   <0.001 
<3.0 53 11 (5–23)  
3.0–3.9 33 45 (30–62)  
4.0–4.9 44 73 (58–84)  
≥5.0 30 90 (73–97)  
PFA-CADP, sec   <0.001 
≤120 76 24 (15–34)  
121–180 36 61 (45–75)  
>180 48 83 (70–91)  
VWF:Ltx/Ag ratio   <0.001 
<0.80 39 82 (67–91)  
≥0.80 121 40 (31–49)  

P-values result from likelihood ratio tests in single variable logistic regression analysis. CI= confidence interval.

Logistic regression analysis using forward selection identified only peak velocity (p<0.001) and PFA-CADP (p=0.033) as the factors showing the strongest evidence of association with abnormal multimers. VWF:Lx/Ag was no longer significant after adjusting for these two variables. Estimated relative risk of abnormal multimers for values of peak velocity and PFA-CADP were then derived (Table 2).

Table 2
Variable Estimated RR (95% CI) P-value 
Peak velocity, m/sec  <0.001 
<3.0 1.00 (reference)  
3.0–3.9 4.02 (1.73–9.31)  
4.0–4.9 6.42 (2.96–13.94)  
≥5.0 7.95 (3.71–17.05)  
PFA-CADP, sec  0.033 
≤120 1.00 (reference)  
121–180 2.58 (1.60–4.17)  
>180 3.52 (2.31–5.37)  
Variable Estimated RR (95% CI) P-value 
Peak velocity, m/sec  <0.001 
<3.0 1.00 (reference)  
3.0–3.9 4.02 (1.73–9.31)  
4.0–4.9 6.42 (2.96–13.94)  
≥5.0 7.95 (3.71–17.05)  
PFA-CADP, sec  0.033 
≤120 1.00 (reference)  
121–180 2.58 (1.60–4.17)  
>180 3.52 (2.31–5.37)  

P-values result from likelihood ratio tests in a logistic regression model containing peak velocity and PFA-CADP. Abbreviations: RR= relative risk, CI=confidence interval. Due to strong association between peak velocity and PFA-CADP, estimated relative risks result from single variable models.

From these relative risks a scoring algorithm was created in order to determine who should be screened for abnormal multimers. A value of 1 was subtracted from each relative risk and rounded to the nearest integer to obtain a score for peak velocity (<3 m/sec=0, 3.0–3.9 m/sec=3, 4.0–4.9 m/sec=5, >5.0 m/sec=7), and for PFA-CADP (≤120 sec=0, 121–180 sec=2, > 180 sec=3). The score from each of the two variables was added to obtain a total score ranging from 0 to 10. The risk score was categorized into four levels, which are 0, 2–3, 5–7, and 8–10, as shown in Table 3.

Table 3
 % Abnormal multimers (95% CI) P-value 
Risk score   <0.001 
44 9 (3–22)  
2–3 28 29 (15–48)  
5–7 42 64 (49–77)  
8–10 46 89 (76–95)  
 % Abnormal multimers (95% CI) P-value 
Risk score   <0.001 
44 9 (3–22)  
2–3 28 29 (15–48)  
5–7 42 64 (49–77)  
8–10 46 89 (76–95)  

P-values result from a likelihood ratio test from a single variable logistic regression model. CI= confidence interval.

Conclusion

Both hemodynamic severity and PFA-CADP testing results could be incorporated into a scoring system to identify potential AVWS due to loss of HMWM in patients with various cardiovascular conditions. Further studies are needed to validate such a scoring system in a separate set of patients.

Disclosures:

No relevant conflicts of interest to declare.

Author notes

*

Asterisk with author names denotes non-ASH members.