Correction of Acquired von Willebrand Syndrome by Transcatheter Aortic Valve Implantation

Abstract: Aims. Acquired von Willebrand syndrome (aVWS) is a common complication of severe aortic valve stenosis and can be corrected by surgical valve replacement. Transcatheter aortic valve implantation (TAVI) is gaining importance, but the influence of this new technique on aVWS has never been examined. The objective of this study was to assess the impact of TAVI on aVWS. Methods. We enrolled 15 patients with severe aortic stenosis and high surgical risk admitted for elective TAVI. All patients were successfully treated by TAVI, using either the transfemoral (n = 6) or transapical approach (n = 9). Patients were screened for aVWS by measuring PFA-100 in vitro closure time, von Willebrand factor (VWF) antigen, VWF function, and VWF multimer analysis. Analyses were then repeated 30 minutes, 24 hours, and 7 days after valve replacement. Results. Fourteen of 15 patients showed pathologic alterations of VWF. An inverse correlation was observed between the transvalvular pressure gradient and VWF high-molecular-weight multimers (VWF:HMWM) (r = -0.621; P=.01), which are essential for the platelet dependent hemostatic function of VWF. Transaortic gradient was significantly reduced in all patients following TAVI. Hemostaseologic findings improved in all patients following TAVI, the percentage of VWF:HMWM increased (19.05 ± 5.19% before TAVI to 24.08 ± 4.75% (P=.04) on day 7 post TAVI), and the multimer pattern normalized. Conclusions. Acquired von Willebrand syndrome due to aortic valve stenosis can successfully be corrected by TAVI.

J INVASIVE CARDIOL 2014;26(12):654-658

Key words: aortic stenosis, aVWS, TAVI, TAVR, von Willebrand factor

_________________________________________

The association between acquired von Willebrand syndrome (aVWS) and severe stenosis of the aortic valve is well known.^1,2 Following the high pressure gradient across a severely stenotic valve, shear forces develop that unfold von Willebrand factor (VWF),^3,4 thereby exposing the ADAMTS-13 cleavage site¹ and leading to a reduction of von Willebrand factor high-molecular-weight multimer (VWF:HMWM).¹ These VWF:HMWMs are essential for platelet-dependent hemostatic function of VWF.^5,6 Previously, it has been shown that bleeding symptoms as well as laboratory findings can be reversed by conventional surgical replacement of the aortic valve.^2,7-12 Aortic valve stenosis occurs predominantly in elderly patients in whom morbidity and perioperative risk are often increased. For patients with unacceptably high risk at surgery, transcatheter aortic valve implantation (TAVI) has emerged as a novel therapeutic option.¹³ With the aging population in view, the question whether TAVI leads to the same improvements in hemostasis as the conventional surgical approach gains importance. 

However, the beneficial effect of TAVI on aVWS due to aortic valve stenosis has thus far not been investigated. We assessed the hypothesis that the impact of TAVI on aVWS would be similar to the impact of the classic surgical approach. 

Methods

Patients. In this pilot study, we enrolled 15 sequential patients (6 males and 9 females; mean age, 80.87 ± 6.65 years) with severe aortic valve stenosis (Valve orifice area <1.0 cm²; mean valvular pressure gradient >40 mm Hg) who underwent elective TAVI at our center. Exclusion criteria were congenital bleeding disorder, hematologic diseases, malignant disease, end-stage renal failure, and severe concomitant dysfunction of the mitral, tricuspid, or pulmonary valve. The local ethics committee approved the design of this study; informed consent was obtained from all patients. Preprocedural work-up included blood samples, transthoracic and transesophageal echocardiography, coronary angiography, and computed tomography of the aorta. Subsequently, the decision for TAVI as well as the choice of transfemoral access (n = 6) or transapical access (n = 9) was made by a dedicated heart team consisting of cardiologists, cardiac surgeons, and anesthesiologists.

The Society of Thoracic Surgeons predicted operative mortality (STS-PROM) was 4.87 ± 2.46% and the average logistic EuroSCORE was 14.54 ± 7.95%. Ten Patients suffered from coronary artery disease; 3 had undergone coronary artery bypass surgery (CABG) with patent left interior mammary artery (LIMA) grafts. Four patients were treated by percutaneous coronary intervention (PCI) and stent implantation prior to TAVI. Currently available risk stratification tools for TAVI have significant limitations. Relevant comorbidities, such as porcelain aorta or frailty, are not represented in the scores. All patients included in this study were deemed high surgical risk candidates by an interdisciplinary heart team after an individual assessment of the patient’s risk factors. 

All but 3 patients were taking aspirin at baseline. Four patients received dual-antiplatelet therapy (aspirin plus clopidogrel) due to coronary stent implantation. Patients who were naive to antiplatelet drugs and with no additional indication for anticoagulation received only aspirin 100 mg daily after TAVI. Two patients with atrial fibrillation received oral anticoagulation with phenprocoumon before TAVI and were transiently switched to a regimen of 2 mg enoxaparin per kg body weight during their stay in our hospital. The remaining 13 patients received 40 mg enoxaparin per day for deep vein thrombosis prophylaxis. 

Echocardiography. Transthoracic echocardiography was performed before and after TAVI. Maximum transaortic velocity and pressure gradients were measured. The aortic orifice area was calculated by the continuity equation. 

Laboratory analyses. At baseline, patients were screened for aVWS by plasma electrophoresis of VWF (multimer analysis) and enzyme-linked immunosorbent assay (ELISA) for VWF antigen concentration (VWF:Ag), as well as for VWF collagen binding activity (VWF:CB). The glycoprotein Ib (GpIb)-binding activity was tested by the new Innovance VWF activity test (VWF:Ac) (Siemens Healthcare). Additionally, we measured the in vitro closure time by platelet function analyzer (PFA-100) with adenosine diphosphate (ADP) and epinephrine. Follow-up samples were taken 30 minutes, 24 hours, and 7 days after successful valve implantation. 

Plasma VWF. Ag and VWF:CB were determined using ELISAs. Microtiter plates were coated with a polyclonal rabbit antihuman VWF antibody (A82; Dako Denmark A/S). After incubation with patient’s plasma, bound VWF:Ag was detected after binding of a 2nd peroxidase labeled polyclonal rabbit anti-human VWF antibody (A228; Dako Denmark A/S) and subsequent staining by ortho-phenyldiamin (S2045; Dako Denmark A/S) as the peroxidase substrate. Measuring VWF:CB to human collagen type III (Biozol Diagnostika) bound to the microtiter plate was again performed by using the peroxidase-labeled polyclonal rabbit antihuman VWF antibody and ortho-phenyldiamin substrate as described above. To assess possible functional deficits of VWF, ratios of VWF:CB and VWF:Ac to VWF:Ag were calculated for each sample.   

Multimer analysis was performed by using medium (1.6%) and low-resolution (1.2%) SDS-agarose gel electrophoresis (LGT agarose type VII; Sigma) followed by luminescent visualization as previously described.^14,15 All plasma samples were diluted with an SDS sample buffer according to the VWF:Ag concentration (eg, 1:20 for a VWF:Ag concentration of 100%). Shortly after electrophoresis, VWF multimers were transferred to nitrocellulose filters by electroblotting. After blocking in milk solution, filters were incubated first with rabbit antihuman VWF antibody (A82). Following extensive washing in milk, they were incubated with a peroxidase-labeled goat antirabbit IgG second antibody (Bio Rad Laboratorien). After further washing in milk, followed by rinsing with water, the VWF multimers were overlayed with Lumi-Light western blotting substrate (Roche Diagnostics GmbH) and placed into a video-detection system with a sensitive camera. This camera generated the multimer images. Densitometric analysis was performed using Biozym Alpha Ease FC software to quantify the percentage of VWF multimers with high, medium, and low molecular weight. The cut-off values were defined as 5 bands for low-molecular-weight multimers, 6-10 bands for medium-molecular weight multimers, and >10 bands for high-molecular weight multimers.¹⁶ 

Furthermore, activated partial thromboplastin time, thrombin time, fibrinogen concentration, international normalized ratio, C-reactive protein, lactate dehydrogenase, and haptoglobin were measured and a differential blood count, including schistocyte identification, was performed. 

All tests were carried out immediately after withdrawal of the blood sample with the exception of VWF analyses, which were performed from citrate plasma that was stored at -80 °C and thawed at 37 °C before processing. 

Medical history. We assessed bleeding symptoms with a questionnaire based on a standardized form.¹⁷ Only bleeding symptoms that occurred during the year before TAVI were taken into account. 

TAVI. Except for the access route, there was no difference between the preprocedure and postprocedure patterns for patients treated by the transapical or transfemoral approach. Both groups of patients received general anesthesia and the same periprocedural anticoagulation protocol. Anesthesia was induced with etomidate, sufentanyl, and cysatracurium; sevoflurane and remifentanyl were applied for the maintenance. The types of implanted valve prosthesis were Edwards Sapien XT (n = 8), Medtronic Engager (n = 5), Symetis Acurate TA (n = 1), and Jenavalve (n = 1). Given the clinical indication, the approach was transfemoral (n = 6) or transapical (n = 9). Mean time of operation was 97 ± 55 minutes; an average of 142 ± 58 mL contrast media were given. Before the induction of rapid pacing and the implantation of the valve, 80-100 IU heparin per kg body weight were administered to obtain a prolonged activated clotting time >250 seconds. In some cases, the heparin effect was antagonized by protamine at the end of the procedure. 

Statistical analyses. Descriptive analysis consisted of absolute frequencies in categorical variables and the means and standard deviations (SD) for continuous variables. All continuous variables in this article are presented as mean ± standard deviation. 

Additionally, a longitudinal mixed model was fitted to the measured parameters. The model class was chosen to account for autocorrelation between repeated measurements. Time points were used as fixed effect in all models. If this effect had a two-sided P-value <.05, the pair-wise comparisons were conducted. IBM SPSS Statistics version 19 was used for the described statistical analysis. 

Results

The assessment of bleeding history revealed 3 patients with repeated epistaxis. One patient who took aspirin and clopidogrel had severe symptoms and needed recurrent treatment. No bleeding occurred in any of the patients during treatment at our center. None of the patients presented with relevant hemolysis before or after TAVI. All interventions were successful and without major complication. Transvalvular gradients were successfully reduced in all patients. The mean gradient was reduced from 56.6 ± 13.6 mm Hg to 10.3 ± 4.1 mm Hg (P<.001). Following TAVI, minor paravalvular regurgitation (grade I) was detected in 5 patients. One patient had a second-degree aortic valve insufficiency. 

At baseline, PFA-100 in vitro closure time was increased in all patients for epinephrine as well as for ADP. Following TAVI, ADP-dependent closure time significantly improved. Mean ADP dependent closure time decreased from 231.69 ± 57.85 seconds before TAVI to 179.25 ± 20.47 seconds (P=.04) on day 7 after TAVI. The results became more clear as the 4 patients under treatment with clopidogrel (who all showed a prolongation of closure time above or close to 300 seconds throughout all time points) were excluded: mean ADP-dependent closure time went from 212.60 ± 51.90 seconds before TAVI to 142.30 ± 60.07 seconds (P=.02) on day 7 after TAVI. All study participants had platelet counts within the normal range at baseline, followed by a significant but transient decline and finally a complete recovery (Table 1).

Before TAVI, aVWS defined by the loss or a considerable reduction of VWF:HMWM was diagnosed in 8 patients. Six patients showed a reduction in VWF:HMWM, but did not fulfill the criteria for aVWS. The proportion of high-, intermediate-, and low-molecular weight multimers was normal in only 1 patient. Patients with the highest transvalvular gradient were most affected by the loss of VWF:HMWM. Overall, there was a significant and inverse relationship between transaortic gradient before TAVI and VWF:HMWM (r = -0.621; P=.01) (Figure 1). A significant increase in the concentration of VWF:HMWM was observed after TAVI: 19.05 ± 5.19% before TAVI and 24.8 ± 4.75% (P=.04) on day 7 after TAVI (Table 1). Multimer pattern normalized in all patients. Correction of the multimer pattern was observed after 24 hours (26.28 ± 4.08%, compared to 19.05 ± 5.19% before TAVI; P<.001) (Table 1) and was maintained through day 7 (Figure 2). The analysis of VWF:Ag, VWF:Ac, and VWF:CB revealed a corresponding improvement (Table 1). 

Discussion

The coincidence of bleeding from gastrointestinal angiodysplasia and aortic stenosis was first described by Heyde.¹⁸ In 1986, VWS in children with congenital cardiac defects¹⁹ and cessation of bleeding in patients with Heyde’s syndrome after replacement of the aortic valve⁸ were observed separately. It was later assumed there was a causal link between aortic stenosis and gastrointestinal bleeding by aVWS.²⁰ Anderson et al reported the first cases of hematologic recovery following conventional surgical valve replacement.¹² Reversal of aVWS by surgery was shown to begin as early as the conclusion of cardiopulmonary bypass²¹ and to persist afterward for at least 6 months^1,2 in the absence of patient prosthesis mismatch.² Cases with longer-lasting correction have been reported.⁹ Aortic valve stenosis is a common disease in elderly patients²² and aVWS is common in severe aortic valve stenosis.² Therefore, the prevalence of aVWS due to aortic valve stenosis could very well be underestimated. This underestimation is reflected by the percentage of aVWS due to cardiovascular diseases, which was reported to be 21% in 2000,²³ 40% in 2002,²⁴ and 46% in 2008.¹⁶ The results of this study confirm the observation of a high prevalence of aVWS in patients with severe stenosis of the aortic valve. 

Similar to Vincentelli et al,² this study reports a strong inverse correlation between the height of the transvalvular gradient and the percentage of the VWF:HMWM (Figure 1). Assays for VWF:Ag and functional assays showed results in or above the normal range; only VWF multimer analysis led to the correct diagnosis. Consistent with findings from previous trials,^16,25 this underscores the importance of multimeric analysis in the diagnosis and follow-up of aVWS. Given the fact that the concentration of VWF:Ag is known to be elevated in elderly patients and patients with cardiovascular disease,^26,27 the diagnostic value of VWF multimer analysis might even be accentuated in aVWS due to degenerative aortic valve stenosis. 

PFA-100²⁸ was previously described as being useful in screening for inherited type 1 von Willebrand disease²⁹ and also proved to be reliable in aVWS due to stenosis of the aortic valve.² With respect to the elevated values of VWF:Ag and the functional assays in the participants in this study, PFA-100 was especially useful in measuring the functional impact of the VWF:HMWM loss on hemostasis. As nearly all of the patients in the present study had been treated with aspirin prior to baseline blood sampling, the thromboxane-dependent PFA-100 closure time for epinephrine was generally prolonged and did not change significantly after TAVI. At baseline, ADP-dependent PFA-100 closure time was prolonged in all patients with or without the presence of aspirin or clopidogrel. Consistent with previous findings, we observed a shortening of the ADP-dependent closure time after TAVI,² with the exception of patients treated with clopidogrel. 

All patients showed a significant increase in VWF:HMWM (P<.001) the day after TAVI, similar to the effect that was previously described for surgical valve replacement.^2,7 When comparing access routes (transapical versus transfemoral), we noticed that the increase of VWF:HMWM was apparent immediately following surgery in the transapical group, whereas the normalization of VWF:HMWM in the transfemoral group was more delayed (Figure 3). Since VWF is released as part of inflammatory responses, the impact of surgical trauma associated with thoracotomy in the transapical group might have led to an earlier onset of this effect in this subgroup. Moderate paravalvular regurgitation did not affect the results in a measurable manner in the small collective of this pilot study. 

Conclusion

In conclusion, our study shows for the first time that TAVI can improve aVWS in a similar fashion as has been shown for surgical aortic valve replacement. In conjunction with the other beneficial effects reported for TAVI, this should lead to a lower incidence of bleeding events in this group of elderly and often frail patients who are at high risk for bleeding complications.

Acknowledgments. We thank Thomas Streichert, Reinhard Schneppenheim, and Florian Langer for reviewing this article.

References

1. Pareti FI, Lattuada A, Bressi C, et al. Proteolysis of von Willebrand factor and shear stress-induced platelet aggregation in patients with aortic valve stenosis. Circulation. 2000;102(11):1290-1295.
2. Vincentelli A, Susen S, Le Tourneau T, et al. Acquired von Willebrand syndrome in aortic stenosis. N Engl J Med. 2003;349(4):343-349.
3. Siedlecki CA, Lestini BJ, Kottke-Marchant KK, Eppell SJ, Wilson DL, Marchant RE. Shear-dependent changes in the three-dimensional structure of human von Willebrand factor. Blood. 1996;88(8):2939-2950.
4. Tsai HM. Physiologic cleavage of von Willebrand factor by a plasma protease is dependent on its conformation and requires calcium ion. Blood. 1996;87(10):4235-4244.
5. Doucet-de Bruine MH, Sixma JJ, Over J, Beeser-Visser NH. Heterogeneity of human factor VIII. II. Characterization of forms of factor VIII binding to platelets in the presence of ristocetin. J Lab Clin Med. 1978;92(1):96-107.
6. Sakariassen KS, Bolhuis PA, Sixma JJ. Human blood platelet adhesion to artery subendothelium is mediated by factor VIII-Von Willebrand factor bound to the subendothelium. Nature. 1979;279(5714):636-638.
7. Weinstein M, Ware JA, Troll J, Salzman E. Changes in von Willebrand factor during cardiac surgery: effect of desmopressin acetate. Blood. 1988;71(6):1648-1655.
8. Cappell MS, Lebwohl O. Cessation of recurrent bleeding from gastrointestinal angiodysplasias after aortic valve replacement. Ann Intern Med. 1986;105(1):54-57.
9. Warkentin TE, Moore JC, Morgan DG. Gastrointestinal angiodysplasia and aortic stenosis. N Engl J Med. 2002;347(11):858-859.
10. Yoshida K, Tobe S, Kawata M, Yamaguchi M. Acquired and reversible von Willebrand disease with high shear stress aortic valve stenosis. Ann Thorac Surg. 2006;81(2):490-494.
11. Panzer S, Badr Eslam R, Schneller A, et al. Loss of high-molecular-weight von Willebrand factor multimers mainly affects platelet aggregation in patients with aortic stenosis. Thromb Haemost. 2010;103(2):408-414.
12. Anderson RP, McGrath K, Street A. Reversal of aortic stenosis, bleeding gastrointestinal angiodysplasia, and von Willebrand syndrome by aortic valve replacement. Lancet. 1996;347(9002):689-690.
13. Leon MB, Smith CR, Mack M, et al. Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery. N Engl J Med. 2010;363(17):1597-1607.
14. Ruggeri ZM, Zimmerman TS. Variant von Willebrand’s disease: characterization of two subtypes by analysis of multimeric composition of factor VIII/von Willebrand factor in plasma and platelets. J Clin Invest. 1980;65(6):1318-1325.
15. Budde U, Schneppenheim R, Eikenboom J, et al. Detailed von Willebrand factor multimer analysis in patients with von Willebrand disease in the European study, molecular and clinical markers for the diagnosis and management of type 1 von Willebrand disease (MCMDM-1VWD). J Thromb Haemost. 2008;6(5):762-771.
16. Tiede A, Priesack J, Werwitzke S, et al. Diagnostic workup of patients with acquired von Willebrand syndrome: a retrospective single-centre cohort study. J Thromb Haemost. 2008;6(4):569-576.
17. Koscielny J, Ziemer S, Radtke H, et al. [Preoperative identification of patients with impaired (primary) hemostasis. A practical concept]. Hamostaseologie. 2007;27(3):177-184.
18. Heide EC. Gastrointestinal bleeding in aortic stenosis. N Engl J Med. 1958;259:196.
19. Gill JC, Wilson AD, Endres-Brooks J, Montgomery RR. Loss of the largest von Willebrand factor multimers from the plasma of patients with congenital cardiac defects. Blood. 1986;67(3):758-761.
20. Warkentin TE, Moore JC, Morgan DG. Aortic stenosis and bleeding gastrointestinal angiodysplasia: is acquired von Willebrand’s disease the link? Lancet. 1992;340(8810):35-37.
21. Solomon C, Budde U, Schneppenheim S, et al. Acquired type 2A von Willebrand syndrome caused by aortic valve disease corrects during valve surgery. Br J Anaesth. 2011;106(4):494-500.
22. Eveborn GW, Schirmer H, Heggelund G, Lunde P, Rasmussen K. The evolving epidemiology of valvular aortic stenosis. The Tromso Study. Heart. 2013;99(6):396-400. Epub 2012 Sep 2.
23. Federici AB, Rand JH, Bucciarelli P, et al. Acquired von Willebrand syndrome: data from an international registry. Thromb Haemost. 2000;84(2):345-349.
24. Budde U, Bergmann F, Michiels JJ. Acquired von Willebrand syndrome: experience from 2 years in a single laboratory compared with data from the literature and an international registry. Semin Thromb Hemost. 2002;28(2):227-238.
25. Sucker C, Feindt P, Zotz RB, Stockschlaeder M, Scharf RE. Functional von Willebrand Factor assays are not predictive for the absence of highest-molecular weight von Willebrand Factor multimers in patients with aortic-valve stenosis. Thromb Haemost. 2005;94(2):465-466.
26. Davies JA, Hathaway LS, Collins PW, Bowen DJ. Von Willebrand factor: demographics of plasma protein level in a large blood donor cohort from South Wales in the United Kingdom. Haemophilia. 2012;18(3):79-81.
27. Paulinska P, Spiel A, Jilma B. Role of von Willebrand factor in vascular disease. Haemostaseologie. 2009;29(1):32-38.
28. Kundu SK, Heilmann EJ, Sio R, Garcia C, Davidson RM, Ostgaard RA. Description of an in vitro platelet function analyzer — PFA-100. Semin Thromb Hemost. 1995;21(Suppl 2):106-112. 
29. Castaman G, Tosetto A, Goodeve A, et al. The impact of bleeding history, von Willebrand factor, and PFA-100((R)) on the diagnosis of type 1 von Willebrand disease: results from the European study MCMDM-1VWD. Br J Haematol. 2010;151(3):245-251.

___________________________________________

From the ¹Department of General and Interventional Cardiology, University Heart Center, Hamburg, Germany; ²Medilys Central Laboratory, Coagulation, Asklepios Clinic Hamburg Altona, Hamburg, Germany; ³Department of Medical Biometry and Epidemiology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; and ⁴Department of Cardiovascular Surgery, University Heart Center, Hamburg, Germany.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript submitted January 6, 2014, provisional acceptance given February 5, 2014, final version accepted June 6, 2014.

Address for correspondence: Dr Olivier Marggraf, Abteilung für Allgemeine und Interventionelle Kardiologie, Universitäres Herzzentrum Hamburg, Martinistr. 52 20246 Hamburg, Germany. Email: o.marggraf@gmx.de