Article Text

Download PDFPDF

Original article
Histology of debris captured by a cerebral protection system during transcatheter valve-in-valve implantation
  1. Tobias Schmidt1,
  2. Michael Schlüter2,
  3. Hannes Alessandrini1,
  4. Ozan Akdag1,
  5. Dimitry Schewel1,
  6. Jury Schewel1,
  7. Thomas Thielsen1,
  8. Felix Kreidel1,
  9. Ralf Bader3,
  10. Maria Romero4,
  11. Elena Ladich4,
  12. Renu Virmani4,
  13. Ulrich Schäfer1,
  14. Karl-Heinz Kuck1,
  15. Christian Frerker1
  1. 1Department of Cardiology, Asklepios Klinik St. Georg, Hamburg, Germany
  2. 2Asklepios Proresearch, Hamburg, Germany
  3. 3Department of Cardiothoracic Surgery, Asklepios Klinik St. Georg, Hamburg, Germany
  4. 4CVPath Institute, Inc., Gaithersburg, Maryland, USA
  1. Correspondence to Dr Tobias Schmidt, Department of Cardiology, Asklepios Klinik St. Georg, Lohmühlenstr. 5, Hamburg 20099, Germany;{at}


Objective Histological analyses of debris captured by a cerebral protection system (CPS) during transcatheter valve-in-valve (VIV) procedures have not been reported.

Methods Fifteen consecutive patients with stenotic aortic (n=13) or mitral (n=2) surgical or transcatheter bioprostheses were treated with implantation of a transcatheter heart valve (THV) in the presence of a dual-filter CPS. Mean patient age was 75 years; mean logistic EuroSCORE was 31%. Filters were collected and histological assessment of debris was performed. Patients were followed clinically until discharge.

Results Debris captured by either or both filters was detected in all patients. Acute thrombus was the most common type of debris, found in all patients, followed in frequency by arterial wall tissue (n=12 patients (80%)), calcification (n=11 (73%)) and valve tissue (n=9 (60%)). Less frequently found were organised thrombus (n=5 (30%)), foreign material (n=4 (27%)) and myocardium (n=2 (13%)). A median of 123 debris particles per patient was detected, with a trend towards a greater median number of particles collected in proximal filters (78 vs 39, p=0.065). The average maximum particle diameter was 88 (range 56–175) µm, with a median of 20 particles ≥150 µm. No stroke or transient ischaemic attack (TIA) had occurred by the time of discharge (mean 8 days).

Conclusions Transcatheter VIV procedures were associated with the release of particulate debris into the cerebral circulation in all patients. The type of debris suggests that debris originates predominantly from arterial and valvular passage of the THV.

  • Stroke
View Full Text

Statistics from


Transcatheter valve-in-valve (VIV) procedures for degenerated surgical or transcatheter aortic and mitral valves are valid treatment options, particularly in elderly patients with comorbidities and an increased risk for repeat surgery.1 Several reports described the feasibility and safety of such procedures.2–5 Periprocedural stroke due to cerebral emboli remains a major complication associated with transcatheter aortic valve implantation (TAVI) in native annuli, with major stroke rates reported at 2%–7% in randomised trials.5–7 Randomised studies reporting stroke rates for VIV procedures are missing. In the Valve-in-Valve International Data (VIVID) registry, a 30-day stroke rate for VIV procedures in aortic or mitral prostheses of 2.2% was reported.4 ,8 The Nordic group for aortic VIV implantations also described a 30-day stroke rate of 2.2%.9

A meta-analysis of more than 10 000 patients undergoing TAVI in native annuli showed a 3.5-fold increased mortality at 30 days in patients who had sustained a major stroke.10

Diffusion-weighted MRI studies have revealed new ischaemic cerebral lesions after TAVI in up to 90% of cases, and transcranial Doppler studies have identified balloon valvuloplasty, valve positioning and valve deployment as causes of cerebral embolisation during TAVI.11–13 Linke et al14 described in their MRI study that the number and volume of cerebral lesions can be reduced with the Claret cerebral protection system (CPS, Claret Medical, Santa Rosa, California, USA). Histological data for TAVI in native annuli, but not for VIV, have been described by Van Mieghem et al.15 Since surgical valve replacement is more frequently done with bioprostheses than with mechanical heart valves, the interest in factors associated with VIV procedures is growing.16 ,17

The objective of the present study was to present qualitative as well as quantitative histological findings in patients undergoing VIV procedures with cerebral protection for degenerated aortic and mitral valve bioprostheses.



Between April 2013 and November 2015, 15 consecutive patients with degenerated aortic (n=13) or mitral (n=2) surgical or transcatheter bioprostheses were treated at our institution with transcatheter VIV using cerebral protection. Patients with isolated aortic or mitral valve stenosis as well as patients with combined degeneration of their prostheses (stenosis grade III plus at least regurgitation grade II or regurgitation grade III plus at least stenosis grade II) were included. VIV procedures due to isolated aortic or mitral regurgitation and valve-in-ring procedures in the mitral position were excluded, since degeneration of the prosthesis is different and captured embolic debris would not be comparable. Isolated aortic stenosis was present in 10 patients and additional three patients had combined aortic valve degeneration. Isolated mitral valve stenosis and combined mitral valve degeneration were present in one case each (figure 1).

Mean patient age was 75 years, 9 patients (60%) were men. High surgical risk was reflected in a mean logistic EuroSCORE of 31%. Detailed baseline characteristics are given in table 1. Clinical follow-up was obtained at 48 and 72 h after the procedure and at discharge. In cases of suspected stroke, a neurologist was consulted for potential further diagnostic.

Table 1

Baseline patient characteristics

Claret CPS

Throughout this study, the Montage and Sentinel CPSs (Claret Medical) were used. Both are CE-marked dual-filter devices designed to capture and remove any debris released during the interventions. The Montage is the second and the Sentinel the third generation of the company's commercially available devices. The Sentinel differs from the Montage only in a more ergonomic handle; there was no change to the portion of the device entering the patient. Both devices consist of two polyurethane filter bags with 140 μm diameter pores that are attached by nitinol loops to a 6 Fr compatible catheter with a deflectable distal tip (figure 2). The device is inserted over a 6 Fr sheath via the right radial or brachial artery. Following insertion, the proximal filter is deployed in the brachiocephalic trunk; subsequently, the second, distal filter is deployed in the left common carotid artery (figure 2). When the Claret devices became available, they were used for all VIV procedures at our institution whenever anatomical access was possible. The Montage was replaced by the Sentinel CPS in January 2014 at our institution.

Figure 2

Claret cerebral protection system with two filters. The proximal filter is positioned in the brachiocephalic trunk and the distal filter in the left common carotid artery.

VIV procedure

In preparation of the procedure, the brachiocephalic trunk and the left common carotid artery were visualised by angiography or computer tomography to exclude severe stenosis and severe tortuosity of the concerned vessels.

Before insertion of the CPS a standard loading dose with heparin of 70 IU/kg was initiated and activated clotting time (ACT) measured. According to the device instructions for use, ACT was aimed to be between 250 and 300 s, at least higher than 250 s before introduction. Control of the ACT was performed at least every 30 min after first measurement.

In brief, standard VIV implantations retrograde across the aortic valve were performed with a straight soft wire through an Amplatz left catheter to cross the aortic prosthesis in the same way as performed for TAVI procedures in native annuli. Then, the straight tip wire was replaced by a standard long J-wire and placed in the left ventricle. The J-wire was replaced over a pigtail catheter by an Amplatz super-stiff wire to guide the transcatheter heart valve (THV). After positioning the THV in the surgical prosthesis, implantation was performed. After implantation, the TAVI catheter was removed and the Amplatz super stiff wire was replaced by a pigtail catheter to measure the transvalvular pressure gradient. After that, the pigtail catheter was pulled back and a final angiography was performed to assess the final result.

The transapical approach for TAVI as VIV in the mitral position and the transaxillary approach for TAVI in the aortic position were previously published in detail by our group.2 ,3 ,18

The diameter of the degenerated bioprostheses for sizing of the THV was measured by transoesophageal echocardiography as well as by the reported internal diameter of the manufacturer. We used the Bapat VIV app, which shows valve-related data including the stent internal diameter and helps to decide which THV size to use.19

At the end of the VIV procedure, the filters and collected debris were withdrawn into the catheter and removed from the patient. The filters were cut and stored in 10% neutral buffered formalin solution. Filters were sent to the CVPath Institute (Gaithersburg, Maryland, USA) for histological assessment of captured debris.

Endpoint definitions

Endpoints were defined according to the Valve Academic Research Consortium for TAVI in native aortic annuli (VARC-2)20 and for transcatheter mitral valve repair and replacement (MVARC).21

Histological analysis

Histological analyses were performed in all cases. A total of 30 filters (2 from each patient) were analysed. The filters were photographed (Canon EOS Rebel XSi), examined grossly for visible debris, cut open and all contents were filtered through a 40 µm nylon cell strainer (BD Falcon; Corning, Durham, North Carolina, USA). The majority of particles detected were adherent to the filter, with only a small number of immersed particles retrieved from the fixative. The cell strainer disc was photographed to document successful debris transfer, and area measurements of retrieved debris were performed. The material collected by the cell strainer was placed in a Shandon Nylon biopsy, dehydrated in a graded series of alcohols and embedded in paraffin. Each paraffin block was serially cut at 4–5 μm, with two consecutive sections affixed per slide. A total of 20 sections were obtained and every other slide was stained with H&E and Movat Pentachrome stain.

The sections were evaluated for the presence of thrombus, valve and arterial wall tissue, vascular structures with or without atherosclerotic changes, myocardial fragments, calcification and foreign material. Thrombus was classified as acute if it showed platelets and fibrin with entrapped red blood cells and acute inflammatory cells or chronic if the thrombus showed the presence of spindle-shaped cells with or without macrophages that lined the thrombus, infiltrated it or had any organisation with matrix deposition interspersed between the fibrin/platelet thrombus. Two pathologists reviewed the slides independently and the final diagnosis was based on unanimous agreement.

Morphometric analyses were performed in addition to the previously described method. Histological sections of each filter sample were scanned using the Zeiss Z1 Axio Scanner. Subsequent CZI files were evaluated using Indica Lab's HALO object co-localisation module. The module was tailored to the tissue and a threshold was set to include particles measuring ≥50 µm2 in total area. Data were expressed as a total object count and average minimum, maximum and median diameters. Additionally, individual particles measuring ≥150 µm were reported.


Written informed consent was obtained from all patients for the TAVI procedure as well as for the Claret implantation. For processing of human tissue outside of routine clinical practice, patient consent was sought. Some patients were treated within the Sentinel H Registry and have additionally signed for the present study.

Statistical analysis

Continuous variables are described as means and standard deviations if normally distributed, or as medians plus interquartile range if not. Categorical variables are described with absolute and relative frequencies.


Patients and procedures

All VIV procedures were performed in a hybrid operation room; mild analgosedation was used in 60% and general anaesthesia in 40% of the cases.

For VIV in the aortic position, both self-expanding prostheses and balloon-expandable prostheses were used (CoreValve, n=3; Evolut R, n=8; Sapien XT, n=1; Sapien 3, n=1; table 2). Vascular access was predominantly transfemoral (n=12 (92%)); 1 patient was treated via transaxillary access.

Table 2

Characteristics of surgical and transcatheter prostheses

For VIV in the mitral position, the balloon-expandable Sapien XT prosthesis was used exclusively (table 2). Transapical access was performed in both cases.

Predilation was performed in four cases and postdilation was performed in seven cases. Mean ACT was 328±39 s, while 12.333±3350 IU of heparin was used. Radial access for the CPS was used in 12 cases and brachial access in 3 cases.

Positioning of the Claret CPS was without any difficulties and both filters were positioned in their intended positions in <5 min. Vascular or bleeding complications due to the Claret device did not occur in any of the patients.

Histological findings

Included in the histological analyses were the contents from 30 (15 proximal, 15 distal) filters. Debris was found in all 15 proximal filters (100%) and 13 distal filters (87%).

Seven different types of debris were found in the filters; they were identified as acute thrombus, organised thrombus, valve tissue, arterial wall tissue, calcification, myocardium and foreign material (figures 3 and 4). Acute thrombus was the most common type of debris, captured in all 15 proximal filters and 12 (80%) distal filters (figure 5). Other frequently found types of debris were arterial wall tissue in 80% (12/15), calcification in 73% (11/15) and valve tissue in 60% (9/15) of patients. Foreign material was found in 27% (4/15) of patients, while organised thrombus was found in 33% (5/15) and myocardium in 13% (2/15) of patients (figure 5). In all patients in whom acute thrombus and biological debris (arterial wall, calcification and valve tissue) were found, platelet aggregation to the biological debris was seen. Also, all samples containing acute thrombus and foreign material showed platelet aggregation.

Figure 3

Macroscopic views of filters with debris. (A) Proximal filter. (B) Distal filter.

Figure 4

Microscopic views of debris captured in the filters. (A) Low-power magnification of proximal filter showing calcified matter originating from degenerated aortic valve prosthesis (arrow heads). (B) Magnification of rectangle in (A) showing calcified debris surrounded by platelet-rich thrombi. (C) Organised thrombus (dark red) proteoglycan (bluish-green). (D) Elastic fibres (black) suggest debris originating from the arterial wall. (E) Fragment of valve tissue rich in proteoglycans and collagen. (F) Second fragment of calcified debris surrounded by thrombus and collagen. (A, B and F) H&E staining; (C–E) modified Movat Pentachrome stains. Scale bars: A=2.0 mm, B, C, E and F=200 µm and D=500 µm.

Figure 5

Type of debris according to patient (Pt), proximal filter (prx) and distal filter (dst). See text for extensive explanations.

Various combinations of different types of debris were found in the individual patient. The combination of all types of debris was most common (n=5 (33%)). Thrombus in combination with arterial wall and calcification was second most common (n=4 (27%)).

Numbers and sizes of particulate debris found in the filters are shown in table 3. Overall, 2862 particles were found and analysed in all filters, with 455 particles >150 µm. A median of 123 particles were found per patient in the filter pairs, with a median of 20 particles per patient being >150 µm. The average median particle diameter per patient was 47 µm, with average minimum and maximum particle diameters of 32 and 88 µm, respectively. There was a trend towards higher total particle counts in proximal than in distal filters (median 78 vs 39, respectively; p=0.065). No differences between proximal and distal filters were observed for average particle diameters and the number of particles ≥150 µm.

Table 3

Morphometric analysis

Stroke/transient ischaemic attack

The stroke rate at 2 and 3 days after the procedure was 0%. Also, stroke rate during hospitalisation (mean 8 days) was 0%. Also, by the time of discharge, no TIA had occurred.


TAVI device success was 80% (12/15). TAVI device failure (n=3) was due to a mean transaortic pressure gradient >20 mm Hg in two patients, and one patient (mitral VIV) had died 6 days after the procedure due to multiple organ dysfunction syndrome. This patient had been treated as an emergency case using a veno-arterial extracorporeal membrane oxygenation because of a decompensated mitral valve stenosis prior to the VIV procedure.

Conversion to open heart surgery was not required in any patient. Myocardial infarction did not occur in any patient. Acute kidney injury of Acute Kidney Injury Network (AKIN) stage 2 did not occur, but AKIN stage 3 occurred in two patients. Both patients recovered and dialysis was not required. Two major vascular complications were observed, and one life-threatening and two major bleedings were documented. New pacemaker implantation had not been necessary by the time of discharge. Thirty-day mortality was 7% (n=1, see patient description above). One-year follow-up was obtained from seven patients. At 1 year, two patients were in New York Heart Association Functional Classification (NYHA) class I, three patients in NYHA class II and one patient in NYHA class III.

No further deaths had occurred, thus all-cause mortality at 1 year was 7%. One stroke occurred 5 months after the procedure (7%, 1/15).


Main findings

The main findings of this study of using a dual-filter CPS during transcatheter VIV implantations are as follows:

  • Debris was captured by either or both filters in all patients.

  • Acute thrombus was the most common type of debris; it was found in all patients.

  • Other prevalent types of debris besides acute thrombus were arterial wall tissue, calcification and valve tissue.

  • There was a trend towards more debris found in proximal than in distal filters.

  • No stroke or TIA occurred.

  • The study demonstrated the feasibility and safety of using a CPS during VIV procedures.

Clinical relevance

This is the first report on debris captured during VIV procedures for dysfunctioning aortic and mitral valve prostheses. Surgical valve replacements are more frequently done with bioprostheses instead of mechanical heart valves, which provides another option if degeneration of these prostheses occurs 10–15 years after the first implantation.16 ,17 VIV procedures in the aortic or mitral position have become a therapeutic option since patients with surgical heart valve prostheses are generally older and have an increased mortality risk for repeat valvular surgery.1

It is important to mention that we have used in about 75% self-expanding prostheses, since the CoreValve or Evolut R has a supra-annular leaflet position, which is an advantage for VIV procedures in aortic position especially for small surgical bioprostheses to lower the risk of a postprocedural gradient. The numbers of patients are too small to compare balloon-expandable prostheses with self-expanding prostheses.

The primary types of structural deterioration of bioprosthetic valves in the aortic or mitral positions are calcification and leaflet tear or their combination.22

Stroke rates for VIV procedures, which have only been collected in registries and not in randomised trials, tend to be lower than for TAVI in native aortic stenosis, but major stroke rates for VIV are still as high as 2.2%.4 ,8 ,9 This incidence of stroke or TIA justifies any attempt to prevent such complications, since the prosthetic material is fragile and manipulation of the prostheses shows high rates (100%) of debris captured in these filters.

Potential origins of emboli during transcatheter valve-in-valve implantations

Due to special stains and an exact morphometric analysis of the debris, it was possible to determine and distinguish the origin of the captured debris. Acute thrombus was the most common type of debris, captured in 27 (90%) of all 30 filters analysed. Acute thrombus may originate at any part of the catheter, including the valve delivery system, guide wire and the THV itself, due to its thrombogenic nature. Maleki et al23 reported up to 9% thrombus formation on regular trans-septal sheaths despite adequate anticoagulation, and Gobeil et al24 even reported 48% of significant thrombus formation on guide wires used in routine percutaneous transluminal coronary angioplasty procedures. Thrombogenicity of guide wires during TAVI procedures has been described in case reports/series25 and laboratory testing.26 Heparin for anticoagulation can minimise thrombus formation.26 In our study, anticoagulation was adequate (as reflected by a mean ACT of 328 s throughout the procedures). The rate of acute thrombus is in line with the data of Van Mieghem et al27 where mostly TAVI procedures in native annuli were implanted. However, in that study the ACT was lower at a mean of 230 s.27 Even in recently implanted TAVI prostheses, thrombogenic material was found at the frame of the device as well as on the leaflets. Since thrombogenicity of the catheters, guide wires and the valve itself cannot be completely avoided, heparin with an ACT level over 250 s should be achieved.

Arterial wall tissue, calcified material and valve tissue were also frequently found types of debris. Since it is known that calcification of the mitral and aortic annuli is related to atherosclerosis in other vascular beds, it is not astonishing that during VIV interventions, which require significant manipulation and instrumentation in the ascending aorta, aortic arch and aortic root, can cause embolisation of arterial wall tissue as well as calcified material.28 Bioprosthetic degeneration is a wear over time, tear of the leaflet material as well as calcification of the leaflets.22 During implantation of a second prosthesis into these fragile materials, embolisation of valve tissue and calcification may be expected and is seen in this study. In a study of 81 patients undergoing TAVI in native annuli, embolised valve tissue was rare at only 15%.27 We further investigated the cases with captured biological debris (valve tissue, arterial wall, calcification) in combination with acute thrombus, which show platelet aggregates to the biological debris, which is an expected finding considering that tissue fragments are known to be prothrombogenic in nature.

Also, we further investigated the possibility of foreign material arising from the actual filter and/or coating and did not find evidence; we also manually scraped the surface of filters with examination under high magnification and no foreign material was observed. Therefore, we assume that the captured foreign material likely arose from hydrophilic polymer coatings commonly used on interventional devices, that is, guide wires, catheters, balloons and so on, which is consistent with previously identified materials in patients undergoing intravascular procedures. In animal experiments, no thrombus was seen on the device. We did not study a control group with implanted filters in patients without any valve intervention; ethical guidelines would not be preserved due to the possible risk of complication (bleeding, perforation and so on) in every invasively implanted device. Frerker et al29 recently reported that foreign material was found in 86% of MitraClip procedures where scraping of the guide and device due to clip placement is higher than with TAVI devices.

Morphometric analysis of captured debris during transcatheter VIV implantations

Separation of debris according to the filter location (proximal vs distal) has not been described for TAVI or VIV procedures yet. There was a numerous difference (n=2036 in proximal filter vs n=826 in the distal filter) and a trend towards a higher number of particles collected in proximal than distal filters. The minimum, median and maximum diameters of particles found in the proximal and distal filter were statistically not different. Until now, there has only been a flow model described by Carr et al30 evaluating the distribution of cardiogenic emboli originating at the aortic root and travelling into the cerebral arteries and descending aorta. The emboli to the cerebral arteries depend on different factors, including the aortic anatomy, blood flow, size of embolic debris and cardiac output.30 The authors also found that smaller size particles, as we have found in our morphometric analysis, tend to travel into the cerebral arteries. Peak particle transport to the cerebral arteries occurred for particles between 1.3 and 1.4 mm (largest diameter), with up to 60% of all particles reaching the cerebrum. Larger diameter particles were more preferentially distributed to the descending aorta. Embolic debris captured with the Claret CPS ranged between 150 and 4000 µm (maximum diameter), which has a high tendency to embolise into the cerebral arteries.15

Study limitations

This is a single-centre study with a small number of patients, but VIV procedures are not nearly performed as frequently as TAVI in native annuli. We combined TAVI and VIV procedures for aortic and mitral bioprostheses, since the degeneration process is similar for both types of surgical valves. We included a heterogeneous group of procedures (different failing surgical valve types, different access routes, different THV types, some with pre/postdilatation), which might have influenced the nature of embolic material caught in the protection device. Larger studies are needed to confirm our findings and differentiate between types of debris and types of valve prostheses. More data on histological findings in VIV procedures are required, and diffusion-weighted MRI studies should be performed to correlate the findings with the clinical outcomes of the patients.


Transcatheter VIV procedures were associated with the release of particulate debris into the cerebral circulation in all patients. The type of debris suggests that debris originates predominantly from the arterial and valvular passage of the THV into the degenerated surgical valve prostheses. The occurrence of debris from THV implantations into surgical bioprostheses seems to be higher than in native annuli. A predilection of the embolic debris into the proximal filter seems to be likely.

Key messages

What is already known on this subject?

  • MRI studies have shown that the number and volume of cerebral lesions can be reduced with the Claret cerebral protection system. Histological data for TAVI in native annuli have been described before, yet without assessment of the origin of the captured debris and not for VIV procedures.

What might this study add?

  • We add data on histological findings for VIV procedures with the assessment of the origin of the captured debris and were able to show a trend towards more debris particles found in the proximal than in the distal filters of the cerebral protection device (total number of particles 2036 vs 826, respectively, p=0.065). We found a debris capture rate of 100% per patient; the most prevalent types of debris were acute thrombus (100% of patients), arterial wall (80%), calcification (73%) and valve tissue (60%).

How might this impact on clinical practice?

  • Since in all patients debris was found in the filters, any attempt to prevent a stroke or TIA seems to justify the possible use of a cerebral protection system, especially in degenerated and fragile bioprostheses. Patient informed consent should be modified due to the 100% capture rate even though the captured debris and MRI diagnosed cerebral lesions are not clearly correlated with clinical strokes. In addition a careful attention to an adequate systemic heparinisation seems to be important since acute thrombus was found in all patients.


View Abstract


  • Correction notice Since this paper was first published online tables 1, 2 and 3 have been inserted into the manuscript.

  • Contributors Conception and design or analysis and interpretation of data was done by all of the authors. Drafting of the manuscript or revising it for critical important intellectual content was done by TS, MS, RV, US, K-HK and CF. Final approval of the submitted manuscript was given by all authors.

  • Competing interests TS has received lecture honoraria from Claret Medical, Inc. and Medtronic, Inc.; CF has received lecture honoraria from Claret Medical, Inc. and Edwards and proctor honoraria from Medtronic. US has received lecture honoraria and consultant fees from Claret Medical, Inc. and proctor honoraria and consultant fees from Edwards and Medtronic, Inc.; K-HK has received lecture honoraria from Claret Medical, Inc. and Edwards, as well as research grants from Medtronic, Inc. RV has been in the advisory board of Medtronic, Inc. and has received research grants from Medtronic and Edwards.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.