Measurements

The prospective CAPITA study evaluates the effect of TAVI treatment on cardiac output, CBF, structural brain lesions and cognitive functioning. Prior to TAVI and at 3‑month follow-up, patients will undergo echocardiography (cardiac output, valve function, LV function), brain MRI (CBF using arterial spin labelling (ASL), white matter hyperintensities (WMH), structural lesions) and extensive neuropsychological testing. To evaluate longer-term haemodynamic and cognitive effects of TAVI, echocardiography and neuropsychological testing will be performed again 1 year after TAVI. Baseline measurements will be taken on the day before scheduled TAVI. Follow-up assessments will be performed during an outpatient clinic visit and are combined with a routine cardiologist consultation. Figure 2 displays the study timeline. The CAPITA study is registered at ClinicalTrials.gov (identifier NCT05481008).

Fig. 2

Timeline of CAPITA study. TAVI transcatheter aortic valve implantation, MRI magnetic resonance imaging

Objectives

The co-primary outcome measures are change in global CBF (in ml/100 g per min) on ASL-MRI and change in global cognitive functioning (Z-score) between baseline and 3‑month follow-up. We will also assess the relation between these co-primary outcome measures and their association with cardiac output [11].

In addition, we will explore determinants of impaired cognitive functioning prior to TAVI, e.g. demographics/risk profile, cardiac functioning and brain injury markers. Moreover, we will explore determinants of change in cognitive functioning after TAVI. As the post-procedural cognitive effects of TAVI may be transient [4, 5, 10], we will assess cognitive functioning again 1 year after TAVI.

We will assess the development and possible regression of WMH on Fluid Attenuated Inversion Recovery (FLAIR) MRI (volume and number at baseline and follow-up). WMH are a manifestation of cerebral small vessel disease, which are associated with increased risk of cognitive impairment and stroke [12]. Moreover, we will assess atrophy (brain volume on structural T1-weighted imaging), infarcts (structural T1-weighted imaging, FLAIR and diffusion-weighted imaging (DWI)) and bleeds and microbleeds (susceptibility-weighted imaging (SWI)).

Patient population

Aortic valve stenosis is defined according to the most recent European Society for Cardiology Guidelines [2]. All consecutive patients scheduled for TAVI at the Amsterdam University Medical Centres (Amsterdam UMC) in Amsterdam, the Netherlands will be screened for study inclusion and approached for study participation at the outpatient clinic. Tab. 1 shows the inclusion and exclusion criteria for the study.

Table 1 Inclusion and exclusion criteria of the CAPITA study

The Amsterdam UMC is a tertiary referral hospital where approximately 350 TAVI procedures per year are performed. A multidisciplinary Heart Team decides whether a patient is eligible to undergo TAVI and selects the access route and valve type. Transfemoral access with local anaesthesia is the default approach. If transfemoral TAVI is not feasible, transaortic access is also performed. Patients will be treated with the balloon-expandable Sapien 3 or Sapien 3 Ultra (Edwards Lifesciences Inc., Irvine, CA, USA) or the self-expandable Navitor (Abbott, Abbott Park, IL, USA) or Evolut R (Medtronic Inc., Minneapolis, MA, USA) transcatheter heart valve devices.

Patients will serve as their own controls: baseline measurements in the presence of severe aortic valve stenosis are compared with post-TAVI follow-up measurement. A control group with comparable aortic valve stenosis seems to be unfeasible, as we believe it is not ethical to delay TAVI in patients with a clinical treatment indication. Time intervals for follow-up assessments are 4 weeks before or after the scheduled visit.

Cardiac output

Cardiac output will be assessed by echocardiographic Doppler measurements of the LV outflow tract velocity time integral, which has been validated against thermodilution, pulmonary artery catheter measurements and cardiac MRI [13,14,15]. Transthoracic echocardiography (GE Medical Systems, Horten, Norway) will be performed by trained echocardiographists from the Amsterdam UMC according to the procedural guidelines of the European Association of Cardiovascular Imaging [16]. All echocardiographs will be evaluated by an independent core laboratory blinded for outcomes using an automatic segmentation method and visually checked by the same physician, who is blinded to clinical data and outcomes.

Brain magnetic resonance imaging

Brain MRI will be performed using the same 3T MRI scanner (Ingenia, Philips, Best, the Netherlands) at baseline and 3‑month follow-up. Tab. 2 shows the sequences included in the MRI protocol (structural T1-weighted imaging, FLAIR, ASL, SWI and DWI).

Table 2 Magnetic resonance imaging study protocol

ASL-MRI assesses global and regional CBF (in ml/100 g per min) using an endogenous tracer [17]. The blood will be magnetically labelled in the carotid artery, creating a labelled image and control image. Both sequences include measurements of the static brain, and the difference between these sequences is the magnetisation of inflowing blood. The ASL-MRI acquisition is shown in Fig. 3. Time of day during scanning, room temperature and visuo-auditory stimuli will be kept similar during baseline and follow-up visits. Patients will be instructed to refrain from caffeine for 3 h and from alcohol and smoking for 12 h prior to the scan [18]. ASL-MRI analyses will be performed by an independent core laboratory using pipelines, which are blinded to clinical data and whether scans were performed at baseline or follow-up. Analyses are corrected for brain volume and partial volume effects.

Fig. 3

Arterial spin labelling magnetic resonance imaging acquisition. a Unlabelled scan, b endogenous tracer labelling in carotid artery, c labelled blood enters cerebral circulation and d labelled scan. Cerebral blood flow = labelled scan—unlabelled scan

WMH are quantified using the Lesion Segmentation Toolbox, an automated segmentation method that uses structural T1-weighted imaging and FLAIR sequences [19]. With these segmentations, we will compute WMH location and volumes (in ml). More details of the MRI protocol can be found in Table S1 in the Electronic Supplementary Material.

Neuropsychological assessment

For cognitive screening, we will use the Mini-Mental State Examination (MMSE) and Montreal Cognitive Assessment (MoCA) [20, 21] MMSE is more sensitive for detection of dementia, whereas MoCA was specifically designed to capture milder forms of vascular cognitive impairment. All patients undergo an extensive and standardised neuropsychological assessment [22, 23]. This test battery covers global cognitive functioning and 4 major cognitive domains: memory, executive functioning, attention/psychomotor speed and language [24]. Tab. 3 presents an overview of the cognitive tests and domains.

Table 3 Neuropsychological test batterya

Neuropsychological tests will be performed by trained clinical neuropsychologists blinded to previous cognitive scores and imaging findings. Time of day, location and test sequence will be kept similar during baseline and follow-up visits. At 3 months and 1 year after TAVI, the neuropsychological assessment will be repeated. If patients are unable or unwilling to visit the Amsterdam UMC during follow-up, a telephone neuropsychological assessment will be offered, including the Modified Telephone Interview for Cognitive Status and the Telephone MoCA [25]. All neuropsychological assessments include the Geriatric Depression Scale (depressive symptoms), Starkstein Apathy Scale (symptoms of apathy) and EuroQol-5D including Visual Analog Scale [22, 23].

Neuropsychological test scores are standardised into Z‑scores, which are constructed as: (test score − mean baseline score) / standard deviation of test score. Z‑scores from available tests of each cognitive domain are averaged to create 4 cognitive domain scores. Patients will serve as their own controls, and Z‑scores are based on baseline cognitive test scores for baseline and follow-up. Detailed information on the neuropsychological assessment and analyses are presented in Tables S1 and S2 in the Electronic Supplementary Material.

Other study variables

Other outcomes and adverse events will be defined and reported according to the third Valve Academic Research Consortium and NeuroARC consensus guidelines [24, 26].

Sample size

Based on our pilot data [11], we hypothesise a 10% increase in CBF after TAVI. With an 80% power and α of 0.05, the required sample size is 126. If a TAVI procedure is complicated by the need for permanent pacemaker implantation, subsequent MRI follow-up will be contra-indicated. In addition, some patients will not be able to take part in all follow-up measurements due to concomitant disease, institutionalisation or death. As we estimate 21% of the patients will not be able to undergo follow-up MRI, a total of 152 patients are needed.

Statistical analysis

Clinical data will be entered at CastorEDC.com to create the final dataset. Continuous variables will be tested for normality distribution with visual inspection and the Shapiro-Wilk test. Accordingly, differences between baseline and 3‑month follow-up in CBF, as well as cognitive global and domain-specific Z‑scores, will be tested with the paired-samples t-test or Wilcoxon signed-rank test.

Change in global CBF (ml/100 g per min increase from baseline) will be assessed as a predictor of change in cognitive functioning (Z-score) using mixed models analysis adjusted for sex, age and education. Additionally, change in cardiac output (in l/min) will be assessed as a predictor of change in CBF with mixed models analysis. Other clinical parameters, including blood pressure, will be assessed as potential confounders of change in CBF and cognitive functioning. Finally, we will examine the association of 1‑year cognitive Z with CBF and cognitive functioning during prior visits.

Ethical consideration

The study is conducted in accordance with the Declaration of Helsinki and the Dutch Medical Research Involving Human Subjects Act (Wet medisch-wetenschappelijk onderzoek met mensen). The study protocol was approved by the Medical Ethics Committee of the Amsterdam UMC. All patients must provide written informed consent before participating.

Current status and timeline

The first patient was included in August 2020, and the last patient’s baseline visit was performed in October 2022. One-year follow-up visits are scheduled and ongoing.