Rationale: Transcatheter aortic valve implantation (TAVI) has become the standard therapy for elderly patients with high surgical risks. Paravalvular leakage after TAVI is relatively common and there is conflicting evidence regarding the clinical impact of mild paravalvular leakage in self-expanding devices. Prospective data for self-expanding devices are required to compare the extent of paravalvular leakage as a result of device design. Grading paravalvular leakage after TAVI is difficult. Echocardiography and angiography systematically underestimate paravalvular leakage (PVL) as compared to cardiac MRI. Hemodynamic measurements are used to aid decision making directly after TAVI implantation. Prospective data comparing hemodynamic measurements with cardiac MRI are needed to design an optimal strategy to grade paravalvular leakage peri-operatively in order to optimize TAVI outcomes. The combination of aortic valve stenosis, angiodysplasia and von Willebrand Disease type 2A (vWD-2A) is known as Heyde syndrome. Previous studies have shown a decrease in angiodysplastic lesions after TAVI. However, since PVL after TAVI is relatively common, angiodysplastic lesions tend to reoccur. Prospective data comparing the severity of PVL to the severity of both vWD-2A and angiodysplasia are lacking.

Objective: To assess procedural hemodynamic measurements in patients with paravalvular regurgitation quantified by means of cardiac MRI (CMR) and to analyse its association with impaired clinical outcome during 5-year follow-up. Secondary objectives are to assess whether the severity of vWD-2A correlates with the severity of PVL measured by cardiac MRI, and to prospectively assess the success percentage of TAVI in the treatment of angiodysplasia.

Study design: This is a prospective, single-center clinical trial. Patients will receive a TAVI. After implantation different hemodynamic indices of PVL will be assessed. Within 4-8 weeks after TAVI a cardiac MRI will be performed to quantify the amount of PVL. Standardized clinical follow-up will take place at discharge, 30 days, 3 months, 6 months and 1 year. Telephone follow-up will take place at 2, 3, 4 and 5 years after TAVI. In patients with known angiodysplasia or iron deficiency anemia e.c.i., a videocapsule endoscopy (VCE) will take place before TAVI and 6 months after TAVI. Of note, for the substudy on Heyde syndrome, patients with a different type of TAVI valve (i.e. no Abbott Portico valve) are also allowed to participate.

Study population: Approximately 80 patients with severe symptomatic aortic valve stenosis with an indication for TAVI will be included. At least 76 patients with a cardiac MRI that is of sufficient quality to quantify the amount of PVL will be included.

Intervention: Patients will undergo cardiac MRI on top of standard clinical care within 4-8 weeks after TAVI. A subgroup of patients will also undergo a VCE.

Main study parameters/endpoints: The primary endpoint is defined as PVL regurgitation fraction as measured by cardiac MRI. One secondary endpoint will comprise a composite of device success, early safety and clinical efficacy as defined by the Valve Academic Research Consortium-2 (VARC-2) (1) and will comprise death, vascular complications, stroke/TIA, life-threatening bleeding requiring transfusion, and acute kidney injury requiring dialysis. Another secondary endpoint will be the reduction of angiodysplastic lesions after TAVI as determined by VCE.

Nature and extent of the burden and risks associated with participation, benefit and group relatedness: The hemodynamic indices can be assessed in a standard fashion using a fluid filled pigtail catheter that is placed in the left ventricle as part of the routine protocol. Following TAVI, enrolled patients will undergo cardiac MRI to assess PVL. The risk of cardiac MRI after TAVI implantation is negligible. Extra blood samples will be taken. After one year, patients will be followed by telephonic follow-up. Risk/benefit: the expected benefit is a structured clinical follow-up at 1, 2, 3, 4 and 5 years, at the cost of an extra visit to undergo cardiac MRI.

1. INTRODUCTION AND RATIONALE Transcatheter aortic valve implantation (TAVI) has become the standard of care in elderly patients with severe symptomatic aortic valve stenosis and increased surgical risks due to advanced age and multiple comorbidity(2-4). However, in contrast to surgical aortic valve replacement (SAVR), paravalvular leakage after TAVI is fairly common. More than moderate paravalvular leakage is associated with intermediate and long-term morbidity and mortality(5).

   Since the introduction of TAVI, implantation- and device technology have improved significantly. Nevertheless, mild paravalvular leakage remains a common finding after implantation of new generation devices. Conflicting evidence exists regarding the impact of mild versus none to trivial paravalvular leakage on mortality and morbidity. Several factors might be of influence on these conflicting results. Patients with left ventricular hypertrophy and no previous aortic regurgitation might experience rapidly rising end-diastolic pressures resulting in heart failure even at small regurgitant volumes as seen with mild paravalvular leakage (6-9). Moreover, paravalvular leakage might have greater impact on survival in patients with impaired left ventricular function as compared to patients with preserved left ventricular ejection fraction(10).

   Furthermore, studies have used different definitions of paravalvular leakage, assessed paravalvular leakage at different time intervals and used different prostheses. Since paravalvular leakage might decrease over time in self- expandable TAVI devices, this potentially influenced outcomes(11). The widely used VARC-2 criteria for paravalvular leakage grading comprise echocardiographic measurements only(1). Especially quantitative assessment of paravalvular leakage in the intermediate strata (mild-moderate) using echocardiography (TTE) remains difficult. TAVI under conscious sedation or local anesthesia more or less preclude the use of peri-operative trans esophageal echocardiography (TEE). TTE analysis and grading of paravalvular leakage is limited by patient factors such as pulmonary disease, patient habitus and prior cardio-thoracic surgery. Moreover, TTE has been shown to underestimate paravalvular leakage as compared to cardiac MRI(12-14). However,the presence of clinical relevant paravalvular leakage necessitates post-dilatation; therefore grading of paravalvular leakage preferably should be done intra-operatively.

   Hemodynamic parameters have been shown to be of great value in grading paravalvular leakage post implantation(15-18). These objective parameters are easily obtained during TAVI and have been shown to have prognostic relevance. All of these parameters are a function of left ventricular end-diastolic pressure (LVEDP). For instance, diastolic delta (DD), defined as the gradient between diastolic aortic pressure and LVEDP, ≤18 millimetres of mercury (mmHg) was associated with adverse prognosis(15). Heart rate adjusted diastolic delta (HR-DD) (HR-DD= [DD/heartrate] *80) <25 mmHg/beats per minute (BPM) was associated with mortality at 1 year in a retrospective study(16). Combined with angiography and (transesophageal) echocardiography, it was associated with impaired clinical outcome in balloon-expandable bioprostheses (16, 17). Furthermore, the aortic regurgitation index (ARI), defined as the ratio of the diastolic delta to systolic aortic pressure (Aortic regurgitation index = (DD/Aortic pressure systolic)*100) >25 combined with angiographic or echocardiographic assessment of paravalvular leakage had a negative predictive value of 95%-100% for the occurrence of more than mild paravalvular leakage. ARI <25 was associated with 1 year mortality, even after adjustment for paravalvular leakage severity(18-20). However, interpretation of ARI is challenging; bradycardia might lead to an ARI <25 without the presence of relevant paravalvular leakage. Determination of ARI should therefore be performed at a heart rate between 60-80 BPM and preferably over several cycles, especially when measured during atrial fibrillation or extrasystoles.

   Both echocardiographic and hemodynamic measures provide an estimation of the amount of paravalvular leak. In contrast to echocardiography, MRI provides accurate quantification of paravalvular leakage volume. In this study the investigators want to assess the contributive value of hemodynamic indices of paravalvular leakage based on paravalvular leakage volume measured by means of cardiac MRI and to assess its association with impaired clinical outcomes during 5 year follow-up.

   Aortic valve stenosis is associated with von Willebrand Disease type 2A (21). vWD-2A is an acquired coagulation disorder that is characterized by a decreased platelet adhesion due to a selective deficiency of high-molecular-weight von Willebrand factor (HMW-VWF) multimers. The pathophysiology of this disorder is that proteolysis of HMW-VWF occurs as it passes through the stenotic valve. High shear forces likely induce structural changes in the shape of the von Willebrand factor molecule. These structural changes make the HMW-VWF susceptible to the action of a specific von Willebrand protease. Since the HMW-VWF molecules are the most effective in platelet-mediated hemostasis under conditions of high shear stress, this makes patients susceptible to bleeding from various tissues (22). This association is best recognized for angiodysplasia. Angiodysplasia are vascular anomalies in the gastrointestinal tract. These lesions often lead to obscure or overt bleeding, which causes anemia (23).

   The pathophysiology of angiodysplasia is mostly unknown, however angiogenesis likely plays an important role, as some studies found an increase of certain angiogenesis markers (24). The combination of aortic valve stenosis, vWD-2A and angiodysplasia is known as Heyde syndrome (21). Angiodysplasia has been reported in 3.1-10% of patients with aortic stenosis (25, 26).

   Angiodysplastic lesions can occur in the entire gastrointestinal tract, but are most often located in the small bowel (57-80%). The majority of patients (40-75%) have multiple lesions (27). Angiodysplasia is diagnosed through endoscopic evaluation of the gastrointestinal tract. If a gastroscopy and colonoscopy do not lead to the diagnosis, a video capsule endoscopy (VCE) is performed. Advantages of VCE are that it visualizes the entire gastrointestinal tract (specifically the small bowel), causes less complications compared to other endoscopic modalities and is preferred by the patient. Disadvantages of a VCE are that therapeutic interventions cannot be performed (this requires an enteroscopy) (28). In order to distinguish angiodysplasia from other vascular malformations, all vascular deviations of the mucosae seen during VCE, by an endoscopic nurse or doctor with experience in VCE assessment, are classified according to the criteria of Leenhardt et al. (29).

   There is currently no adequate therapy for angiodysplasia, as these lesions tend to reoccur after endoscopic treatment (27). However, case reports and three retrospective studies have shown that aortic valve replacement in patients with Heyde syndrome does not only treat the valve stenosis, but also the angiodysplasia (23, 30-33). In these studies approximately 79-92% did not require any blood transfusions after aortic valve replacement. Patients who still required blood transfusions experienced a significant reduction in the number of transfusions (31).

   Furthermore, there have also been five prospective studies that showed 31-92% of patients with severe aortic stenosis have vWD-2A (34-38). Two studies looked at surgical aortic valve replacement (SAVR), one study looked at TAVI and two studies looked at both interventions. All studies showed that vWD-2A was no longer present after aortic valve replacement. However, one study discovered that vWD-2A reappeared in the majority of patients at 6 months follow-up (approximately 66-74%). This reappearance was likely caused by the occurrence of valve regurgitation (34). A second study also found significantly lower HMW-VWF values in patients with valve regurgitation (38). Since paravalvular leakage (PVL) is a common complication of TAVI, this finding is important for long-term outcomes of patients with Heyde syndrome (39).

   A recent retrospective study indeed showed that gastro-intestinal bleeding was more prevalent after TAVI compared to SAVR (40). Furthermore, the study of Selvam et al. also found that certain angiogenesis markers were elevated in patients with AS. This level decreased after aortic valve replacement (38).

   It is necessary to further evaluate the use of TAVI in the treatment of Heyde syndrome as it could possibly become a treatment option for gastrointestinal bleeding in patients with Heyde syndrome. Furthermore, vWD-2A markers could possibly correlate to the existence and level of PVL measured with magnetic resonance imaging (MRI)

2. OBJECTIVES 2.1 Primary Objective

   - To assess per-procedural hemodynamic indices of paravalvular leakage and to relate these to paravalvular leakage as quantified by cardiac MRI.

   2.2 Secondary Objectives

   • To exploratory assess short- and long-term outcomes in patients with different severity of paravalvular leakage
   • To assess whether the severity of vWD-2A correlates with the severity of PVL measured by cardiac MRI.
   • To prospectively assess the success percentage of TAVI in the treatment of angiodysplasia

3. STUDY DESIGN The study is designed as a single center, investigator-initiated, clinical trial.

   In case patients want to participate there is no change of standard clinical care. After TAVI, study participants will receive additional cardiac MRI within 1 month. Subsequent 3-month and 1-year out-patient clinical follow-up will be performed. Telephone follow-up will be performed after 2,3,4 and 5 years. The end of the study is defined as the day when the last patient completes her/his last visit planned in the protocol.

4. STUDY POPULATION 4.1 Population (base) Patients with severe symptomatic aortic valve stenosis allocated to TAVI by a dedicated heart team will be screened for potential participation. The definition of severe symptomatic aortic valve stenosis is defined in the inclusion criteria.

Inclusion Criteria:

1. Patients must be >18 years old.
2. Written informed consent is obtained from all patients.
3. Severe aortic valve stenosis is defined as: jet velocity greater than 4.0m/s or Doppler Velocity index <0.25 or an initial aortic valve area (AVA) of ≤ 1.0cm² (indexed EOA≤06 cm²/m²) as measured by trans thoracic echocardiography <6months prior to inclusion.
4. Patients have symptomatic aortic stenosis, as demonstrated by New York Heart Association (NYHA) functional class 2 or greater or other symptoms of aortic stenosis (e.g. syncope or angina in the absence of coronary artery disease).
5. Surgical risk is deemed high or intermediate by Society of Thoracic Surgery (STS) risk score or by documented Heart-team agreement due to frailty or co-morbidities.
6. The aortic annulus diameter as measured by ECG-triggered CT-scanning < 6months prior to inclusion meets the ranges indicated in the instructions for use.
7. The access artery diameters (femoral or subclavian) as measured by CT-scanning < 6 months prior to inclusion meet the ranges indicated in the instruction for use.
8. There are no contra-indications (e.g.: severe claustrophobia, metal implants, severe renal failure) for and patient is willing to undergo cardiac MRI at discharge to 30 days after TAVI.

Exclusion Criteria:

1. Patient is unwilling or unable to comply with study-required follow-up evaluations.
2. There is evidence of a myocardial infarction within 30 days to index procedure.
3. The presence of severe mitral regurgitation or stenosis.
4. The presence of pre-existing prosthetic cardiac device, valve or prosthetic ring in any position.
5. Left ventricular ejection fraction (LVEF) less than 30%.
6. Untreated significant coronary artery disease requiring revascularization.
7. Echocardiographic evidence of intra-cardiac mass, thrombus or vegetation suggesting active endocarditis.
8. The patient is hemodynamically unstable, requiring inotropic or vasopressive and / or mechanical support.
9. The presence of pulmonary edema or intra venous diuretics to stabilize heart failure at index procedure.
10. Renal insufficiency, defined as a serum creatinin greater than 250umol/l or end-stage renal disease requiring dialysis.
11. Morbid obesity, defined as a BMI ≥40.
12. A life expectancy of less than one year.