- Email: [email protected]
In vitro coronary flow after transcatheter aortic valve-in-valve implantation: A comparison of 2 valves
Accepted Manuscript In vitro coronary flow after transcatheter aortic valve-in-valve implantation: A comparison of two valves Sina Stock, MD, Michael Scharfschwerdt, PhD, Roza Meyer-Saraei, PhD, Doreen Richardt, MD, Efstratios I. Charitos, MD, PhD, Hans-Hinrich Sievers, MD, Thorsten Hanke, MD PII:
S0022-5223(16)31485-4
DOI:
10.1016/j.jtcvs.2016.09.086
Reference:
YMTC 11039
To appear in:
The Journal of Thoracic and Cardiovascular Surgery
Received Date: 12 April 2016 Revised Date:
22 September 2016
Accepted Date: 24 September 2016
Please cite this article as: Stock S, Scharfschwerdt M, Meyer-Saraei R, Richardt D, Charitos EI, Sievers H-H, Hanke T, In vitro coronary flow after transcatheter aortic valve-in-valve implantation: A comparison of two valves, The Journal of Thoracic and Cardiovascular Surgery (2016), doi: 10.1016/ j.jtcvs.2016.09.086. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
In vitro coronary flow after transcatheter aortic valve-in-valve implantation: A comparison of two valves.
RI PT
Sina Stock, MD, Michael Scharfschwerdt, PhD, Roza Meyer-Saraei1, PhD, Doreen Richardt, MD, Efstratios I. Charitos2, MD, PhD, Hans-Hinrich Sievers, MD, Thorsten Hanke, MD
SC
Department of Cardiac and Thoracic Vascular Surgery
Address for correspondence: Hans-Hinrich Sievers, MD
M AN U
University of Luebeck, Germany
Department of Cardiac and Thoracic Vascular Surgery, University of Luebeck
TE D
Ratzeburger Allee 160, 23538 Luebeck, Germany Tel.: +49 451 500 2108, Fax: +49 451 500 2051
EP
E-mail: [email protected]
AC C
Sina Stock, MD and Michael Scharfschwerdt, PhD contributed equally to this work.
Article word count: 3493
Funding: This work was supported by the German Heart Foundation/German Foundation of Heart Research [grant number F/30/12]
1 2
University of Luebeck, Department of Cardiology, Angilogy and Intensive Care Medicine, Luebeck, Germany University of Halle (Saale), Department of Cardiac Surgery, Halle (Saale), Germany
ACCEPTED MANUSCRIPT Conflicts of interest: Sina Stock, MD, Thorsten Hanke, MD, Efstratios I. Charitos, MD, PhD, Doreen Richardt, MD and Hans-H. Sievers, MD received travel grants from St. Jude Medical and Edwards Lifesciences. Efstratios I. Charitos, MD, PhD holds significant stock of Edwards Lifesciences. Thorsten Hanke, MD is a consultant for St. Jude Medical. The senior
RI PT
author created the hypothesis and reviewed the manuscript, supervised by the corresponding author. The manuscript was primarily written by the first author. In addition, the experiments were performed solely by the first author as well as the engineer Michael Scharfschwerdt,
AC C
EP
TE D
M AN U
SC
PhD, and the statistical work by Efstratios Charitos, MD, PhD.
ACCEPTED MANUSCRIPT Glossary of abbreviations Transcatheter Aortic Valve-in-Valve Implantation
SAVB
Surgical Aortic Valve Bioprosthesis
GOA
Geometric Orifice Area
THV
Transcatheter Heart Valve
RCF
Right Coronary Flow
LCF
Left Coronary Flow
dPmean
Mean Pressure Gradient
dPmax
Maximum Pressure Gradient
AC C
EP
TE D
M AN U
SC
RI PT
TAVI-ViV
ACCEPTED MANUSCRIPT Abstract Objective: The Transcatheter Aortic Valve-in-Valve Implantation (TAVI-ViV) is an evolving treatment strategy for degenerated surgical aortic valve bioprostheses (SAVB). However, there is some
RI PT
concern regarding coronary obstruction, especially after TAVI-ViV in calcified SAVB with externally mounted leaflets. We investigated in vitro coronary flow and hydrodynamics after
SC
TAVI-ViV in two modern SAVB with externally and internally mounted leaflets.
M AN U
Methods:
Aortic root models including known risk factors for coronary obstruction served for the implantation of SAVB with either externally (St. Jude Trifecta™, size 25) or internally (Edwards Perimount® Magna Ease, size 25) mounted leaflets. Left and right coronary flow as well as hydrodynamics were measured before and after TAVI-ViV with an Edwards Sapien
TE D
XT™, size 23. After the first experimental run, the SAVB leaflets were artificially “calcified”
Results:
EP
and the measurements were repeated.
AC C
In both models, non-calcified and “calcified”, we found no significant reduction in coronary flow, neither when testing Trifecta nor Perimount Magna Ease. Mean pressure gradient increased after TAVI-ViV in the non-calcified model (Trifecta p=0.0001, Perimount Magna Ease p=0.006) and geometric orifice area decreased (both p<0.001). In the “calcified“ model, mean pressure gradient decreased (both p<0.001) and geometric orifice area increased (both p<0.001).
Conclusions:
ACCEPTED MANUSCRIPT In our specific model, TAVI-ViV is feasible in different SAVB (St. Jude Medical Trifecta and Edwards Perimount Magna Ease), non-calcified as well as “calcified”, without risk of coronary obstruction. Nevertheless, prior to clinical application of these results, preoperative
RI PT
thorough assessment including the different limitations of this model is mandatory.
AC C
EP
TE D
M AN U
SC
Abstract word count: 243
ACCEPTED MANUSCRIPT Central message This limited in vitro study shows no impairment of coronary flow after transcatheter aortic valve-in-valve implantation in surgical bioprostheses with internally and externally mounted
RI PT
leaflets.
AC C
EP
TE D
M AN U
SC
(Word count: 25, Characters: 194)
ACCEPTED MANUSCRIPT Perspective Statement Transcatheter aortic valve-in-valve implantation is an emerging treatment for failing bioprostheses potentially reducing coronary flow, especially in bioprostheses with externally mounted leaflets. This was not found in this limited in vitro study comparing bioprostheses
consideration prior to clinical application.
AC C
EP
TE D
M AN U
SC
(Word count: 51, Characters: 400)
RI PT
with externally and internally mounted leaflets. The limitations of this model need
ACCEPTED MANUSCRIPT
1
In vitro coronary flow after transcatheter aortic valve-in-valve implantation:
2
A comparison of two valves.
RI PT
3 Introduction and background
5
The Transcatheter Aortic Valve-in-Valve Implantation technique (TAVI-ViV) is an appealing
6
treatment option for patients with degenerated surgical aortic valve bioprostheses (SAVB) and
7
where a high surgical risk is expected. However, apart from many advantages compared to re-do
8
cardiac surgery, there is some concern regarding coronary obstruction as a potential life-
9
threatening complication 1. In the global Valve-in-Valve registry, this complication has been
10
validated in 3.5% of the patients undergoing TAVI-ViV 2. Besides predisposing anatomic factors,
11
such as a low coronary ostia height, a narrow aortic root or sinotubular junction, it is feared that
12
the geometry of the primarily implanted SAVB and its extension of calcification may influence
13
the risk of coronary obstruction 1-5.
14
Considering modern stented SAVB, two different geometric types are to be distinguished: Firstly,
15
SAVB with internally mounted leaflets, such as the Edwards Perimount® Magna Ease (Edwards
16
Lifesciences LLC, Irvine, USA) and secondly, SAVB with externally mounted leaflets, such as
17
the St. Jude Trifecta™ (St. Jude Medical Inc., St. Paul, USA). Comparing these two different
18
types, the geometry of SAVB with externally mounted leaflets causes a larger geometric orifice
19
area (GOA) than SAVB with internally mounted leaflets but the wide extension of leaflet tissue
20
beyond the diameter of the valve´s stent may influence coronary flow 6 (Figure 1).
21
Gurvitch et al. describe coronary obstruction after TAVI-ViV with a balloon-expandable
22
transcatheter heart valve (THV, Edwards Sapien XT, size 23, e.g. Medtronic CoreValve, size
AC C
EP
TE D
M AN U
SC
4
1
ACCEPTED MANUSCRIPT
26 [Medtronic, Minnesota, USA]) occurring in two patients with highly degenerated SAVB of
24
the type Sorin Mitroflow, size 21 (Sorin Group Inc, Vancouver, Canada). In this patients, who
25
presented with small aortic roots, the externally mounted leaflets led to a close proximity of valve
26
tissue and coronary ostia. In both patients, the leaflet tissue occluded at least one coronary ostium
27
after TAVI-ViV and caused peri-interventional death 1.
28
The objective of this in vitro investigation is to identify specific differences between TAVI-ViV
29
in geometrically different SAVB with special regard to coronary flow but also considering
30
hydrodynamic performance. Potential coronary obstruction shall be identified leading to
31
strategies in pre-interventional planning to avoid or mitigate this complication.
M AN U
SC
RI PT
23
32
Material and methods
33 Valves
35
Surgical aortic valve bioprostheses: In this in vitro investigation, the SAVB with internally
36
mounted leaflets was represented by the Edwards Perimount® Magna Ease (n=5) and the SAVB
37
with externally mounted leaflets by the St. Jude Trifecta™ (n=5), both size 25. After the first
38
experimental run, leaflet calcification was simulated using glue. The leaflet “calcification” was of
39
alternating thickness with a maximum of 5mm.
40
Transcatheter heart valve: To perform TAVI-ViV, we used the balloon-expandable Edwards
41
Sapien XT™ (n=2), size 23, to achieve optimal hydrodynamics 5,7.
AC C
EP
TE D
34
42 43
Aortic root models
2
ACCEPTED MANUSCRIPT
The design of the aortic root models intended to imitate risk factors that are assumed to favor
45
coronary obstruction. Therefore, it was based on investigations by Ribeiro et al. revealing a
46
distance from coronary ostia to the aortic valve annulus of less than 12mm in 86% of the patients
47
suffering from coronary obstruction 8. The mean distance was 10.6mm. Furthermore, an average
48
diameter of the sinuses of Valsalva of 28.1mm became apparent. We created two aortic root
49
models (Figure 2) with different coronary ostia height (8mm and 10mm) using aortic sinus
50
prostheses of the smallest available size, label size 26mm (Uni-Graft W SINUS, B. Braun,
51
Melsungen, Germany). Sinus prostheses were used to imitate physiological conditions as close as
52
possible. The nomenclature refers to the diameter of the annulus and sinotubular junction,
53
whereas the diameter of the sinuses is 33mm. A circular strengthening with felt, situated below
54
the sinuses tissue, simulated the aortic annulus and ensured an identical implantation height of the
55
SAVB. The coronary arteries were simulated by an 8mm Dacron prosthesis each (B. Braun,
56
Melsungen, Germany).
TE D
M AN U
SC
RI PT
44
57
Physiological mock circulation
59
The physiological circulation was imitated by a pulse duplicator allowing for the evaluation of
60
hydrodynamic parameters 9. It enables the adjustment of different cardiac output volumes,
61
afterload and heart rates. A camera on top of the aortic root models made a visual observation
62
and pictures possible.
AC C
EP
58
63 64
Coronary flow device
65
In vivo, the right and left coronary flow (RCF and LCF) depend on the different myocardial flow
3
ACCEPTED MANUSCRIPT
resistance during a heart cycle. The contraction of the heart muscle in systole leads to a higher
67
flow resistance and subsequently to less coronary perfusion compared to diastole. This effect is
68
stronger in the left ventricle than in the right ventricle10.
69
To imitate this physiological coronary flow, the Dacron prostheses, simulating the coronary
70
arteries, were connected to the coronary flow device (Figure 3), which consists of two sealed
71
power chambers with elastic tubes inside. During systole, the power chambers can be
72
pressurized, leading to compression of the elastic tubes and subsequently to reduction of the
73
coronary flow as mentioned above. The power chambers surrounding the right and left coronary
74
artery can be pressurized independently. During diastole, the power chambers are decompressed.
75
The flow curves for LCF and RCF in our model (Figure 4) are comparable to the physiological
76
coronary flow and prove the validity of this coronary flow device.
77
For the experimental tests, pressurization was based on the Flow-based Intraoperative Coronary
78
Graft Patency Assessment 10.
TE D
M AN U
SC
RI PT
66
79 Experimental procedure
81
Per aortic root model, SAVB type and experimental run, we performed five measurements
82
(eFigure 1) taking the mean value.
83
For the first experimental run, the Trifecta™ respectively the Perimount® Magna Ease was
84
sewed into an aortic root model. Thereafter, we inserted the conduit into the mock circulation 9.
85
We determined the values of the following parameters: diastolic LCF and RCF, pressure
86
gradients dPmean and dPmax. The identification of the GOA resulted from photographs taken by the
87
high-speed camera. Subsequently, we explanted the conduit, performed TAVI-ViV with the
AC C
EP
80
4
ACCEPTED MANUSCRIPT
Sapien XT™, commissure-to-commissure with the SAVB, and repeated all measurements.
89
For the second experimental run, we simulated “calcifications” of the SAVB by using glue
90
(Figure 5) and repeated the tests according to the protocol. To simulate leaflet thickening as well
91
as leaflet stiffening, we used two glues with different viscosities. At first, the leaflets of the
92
SAVB were stiffened by applying liquid glue (SEKUNDEN ALLESKLEBER geruchsfrei
93
EASY, UHU GmbH, Buehl/Baden, Germany) on the entire leaflet surface. This was followed
94
by the punctual application of a viscous gel glue (SEKUNDENKLEBER blitzschnell
95
SUPERGEL, UHU GmbH, Buehl/Baden, Germany) on the entire leaflet surface in an eccentric
96
shape, especially across the central section and free edge, to simulate valve thickening and
97
calcific plaques.
98
During the measurements, diastolic pressure was 80mmHg and systolic 120mmHg, stroke rate 64
99
beats per minute and stroke volume 70ml. The test solution was represented by a physiological
100
saline (0.9%) with a density of 1.0046g/cm3 and a dynamic viscosity of 0.9mPa·s at an ambient
101
temperature of 20°C.
102
Concerning coronary flow, the maximum LCF was determined 120ml/min and RCF 80ml/min
103
10
SC
M AN U
TE D
EP
.
AC C
104
RI PT
88
105
Technique of measurement
106
The left ventricular (4cm below the aortic valve) and aortic pressure (6cm above the aortic valve)
107
were measured with two capacitive pressure transducers Envec Ceracore M (Endress + Hauser,
108
Maulburg, Germany), calibrated to a measuring range of -20 to +160mmHg and a resolution of
109
0.02mmHg. 5
ACCEPTED MANUSCRIPT
The sensor of an ultrasonic flowmeter TS-410 (Transonic System Inc., Ithaca, USA) was
111
mounted directly below the aortic valve to record the volume flow through the valve. The sensor
112
works bi-directionally with a resolution of 2ml/min and records flow rates up to 20l/min.
113
LCF and RCF were measured with the ultrasonic device TS-420 and coronary probes of 6mm
114
(Transonic System Inc., Ithaca, USA).
115
A high-speed camera Motionscope HR-1000 (Redlake Imaging Corp., Morgan Hill, USA) above
116
the conduit recorded the characteristics of motion of the aortic valves with 500 pictures per
117
second. Video recordings and flow measurements were started simultaneously by using trigger
118
signals.
M AN U
SC
RI PT
110
119 Analysis and statistics
121
The values for pressure and flow were registered by an analogue-digital converter recording 500
122
individual values per second. Each measurement included ten successive cardiac cycles in order
123
to calculate the mean value ± standard deviation (SD). The analysis of the measurement results
124
was based on the international norm of testing cardiac valve prosthesis (ISO 5840:
125
Cardiovascular implants – Cardiac valve prostheses).
126
The evaluation of GOA was based on photographs taken by the high-speed camera and calculated
127
with ImageJ (NIH Image), using the inner diameter of SAVB as reference value.
128
Statistically, the T-test or the Mann-Whitney U test were applied depending on the given
129
distribution. The level of significance was defined p<0.05. Data processing programs (R version
130
3.1.1 and IBM SPSS statistics version 22) were used for statistical evaluation.
AC C
EP
TE D
120
131 6
ACCEPTED MANUSCRIPT
Results
132 Coronary flow
134
Significant differences in LCF and RCF between the different SAVB types before and after
135
TAVI-ViV as well as coronary height could not be observed, neither in the non-calcified nor in
136
the “calcified” model (all p-values were non-significant) (Table 1).
137
Right coronary flow: The coronary flow after TAVI-ViV in non-calcified SAVB dropped by 7%
138
(coronary height 8mm) and 9% (coronary height 10mm) in Trifecta™ and by 2% (coronary
139
height 8mm) and 3% (coronary height 10mm) in Perimount® Magna Ease. The coronary flow
140
after TAVI-ViV in “calcified” SAVB dropped by 8% (coronary height 8mm) and 10% (coronary
141
height10mm) in Trifecta™. Considering Perimount® Magna Ease, flow after TAVI-ViV
142
increased by 6% (coronary height 8mm) and remained equal (coronary height 10mm).
143
Left coronary flow: After TAVI-ViV in non-calcified SAVB, LCF decreased by 3% (coronary
144
height 8mm) and 9% (coronary height 10mm) in Trifecta™ and remained equal (coronary height
145
8mm) and dropped by 5% (coronary height 10mm) in Perimount® Magna Ease. TAVI-ViV in
146
“calcified” SAVB caused a flow reduction of 8% (coronary height 8mm) and 4% (coronary
147
height 10mm) in Trifecta™ and of 8% (coronary height 8mm) while flow increased by 6%
148
(coronary height 10mm) in Perimount® Magna Ease.
SC
M AN U
TE D
EP
AC C
149
RI PT
133
150
Geometric orifice area
151
(Table 2)
152
Non-calcified model: TAVI-ViV in Trifecta™ and in Perimount® Magna Ease resulted in a
153
decrease of GOA (both p<0.001). We found a significant difference between the two different
154
SAVB types (p<0.001). 7
ACCEPTED MANUSCRIPT
“Calcified” model: The GOA increased after TAVI-ViV in Trifecta™ and in Perimount® Magna
156
Ease (both p<0.001). A statistically significant difference between the two different types of
157
SAVB could not be observed (p-value non-significant).
RI PT
155
158 Pressure gradients
160
(Table 2)
161
Non-calcified model: TAVI-ViV in Trifecta™ and in Perimount® Magna Ease resulted in an
162
increase of dPmax (both p-value non-significant) and of dPmean (p=0.0001, p=0.006). dPmean showed a
163
significant difference between the SAVB types before TAVI-ViV (p=0.009) and dPmax after
164
TAVI-ViV (p=0.036).
165
“Calcified” model: dPmax and dPmean decreased after TAVI-ViV in Trifecta™ and in Perimount®
166
Magna Ease (dPmax: p=0.0003, p=0.0004; dPmean: both p<0.001). Both parameters showed no
167
significant difference between the SAVB types (p-values non-significant).
TE D
M AN U
SC
159
EP
168
Overlapping of Trifecta leaflets
170
The implantation of the Sapien XT in the Trifecta resulted in an overlapping of the Trifecta
171
leaflets around and above the THV (Figure 5). This overlapping occurred in all TAVI-ViV
172
procedures during our test, in the non-calcified model as well as in the “calcified”.
AC C
169
173 Discussion
174 175
Coronary flow 8
ACCEPTED MANUSCRIPT
This in vitro study provides no evidence of a significant difference in coronary flow after TAVI-
177
ViV in SAVB with externally and internally mounted leaflet tissue (Trifecta and Perimount
178
Magna Ease).
179
To date, there are several risk factors known to increase the risk of coronary obstruction. Besides
180
predisposing anatomic aspects, such as a low coronary ostia height or a narrow aortic root and
181
sinotubular junction, the characteristics of the SAVB need to be considered. Risk factors for
182
coronary obstruction are a supra-annular implantation, a high profile of the SAVB and its
183
locational relation to the coronary ostia 1,11,12. Due to clinical studies, Dvir et al. hypothesized that
184
coronary obstruction occurs preferentially in modern SAVB with externally mounted leaflet
185
tissue, such as the Trifecta™
186
occurred more frequently in patients with calcified SAVB compared to patients with degenerated
187
regurgitant SAVB only
188
displacement of calcified and thickened tissue close to the coronary ostia 14.
189
The present in vitro study aims to investigate further these theses mentioned by Dvir et al. in
190
order to improve the treatment of patients with degenerated SAVB and to foresee an increased
191
risk of coronary obstruction.
192
The results of the present study could not verify these theses. A critical obstruction of coronary
193
flow after TAVI-ViV did not occur in SAVB with internally or externally mounted leaflets in the
194
chosen model. In both experimental runs – the non-calcified as well as the “calcified” model –
195
even with low located coronary ostia, TAVI-ViV did not change the profile of coronary flow
196
significantly. The distance between the coronary ostia and leaflet tissue of the SAVB remained
197
sufficiently in the specific size of the aortic root models used in this study (Figure 5). However,
198
in the calcified model, this space between the coronary ostia and the leaflets was reduced (Figure
. Furthermore, Dvir et al. pointed out, that coronary obstruction
. As the main reason for this fatal complication, they suggested the
AC C
EP
TE D
13
1,2
M AN U
SC
RI PT
176
9
ACCEPTED MANUSCRIPT
5). It is imaginable, that with an increasing amount of calcific deposits, this space could be
200
further reduced causing coronary obstruction. Nevertheless, TAVI-ViV in both modern SAVB
201
types is proved to be feasible and the risk of coronary obstruction is estimated similarly in
202
patients with an anatomy alike the geometry of our aortic root models.
203
To minimize the risk of coronary obstruction in general, an exact pre-interventional planning
204
seems to be a pre-requisite 1. A cardio-CT prior to a TAVI-ViV procedure enables exact images
205
of a patient´s anatomy. Apart from aortic root characteristics, such as diameter and height of the
206
sinuses and the sinotubular junction as well as coronary ostia height, the characteristics and
207
radiological appearance of the degenerated SAVB should be considered
208
echocardiography with aortography is recommended
209
investigations of Gurvitch et al. and Ye et al. 4,5. According to these authors, the major influence
210
on coronary perfusion is not only the primarily used SAVB but also a precise pre-interventional
211
planning to avoid coronary occlusion. This is aligned with Linke et al., who reported a successful
212
TAVI-ViV into a modern SAVB (Trifecta, size 23) without coronary obstruction, using exact
213
pre-interventional imaging 5,16.
214
However, risk factors for coronary obstruction shown by pre-interventional imaging are not
215
considered to be an absolute contraindication for TAVI-ViV. At present, there are already some
216
existing strategies to avoid coronary occlusion. Intraoperative transesophageal echocardiography
217
and angiography enable a fast detection of coronary occlusion. The intra-procedural inflation of a
218
balloon is a useful technique to estimate the risk of coronary obstruction during a TAVI-ViV
219
procedure. A balloon similarly sized to the intended THV results in temporary displacement of
220
the SAVB leaflets and reveals by means of aortography the remaining space between coronary
221
ostia and SAVB leaflets17. The intraoperative placement of a guide wire in the coronary artery at
222
risk seems to avoid a dislocation of leaflet tissue as well as to ensure an emergency access to the
M AN U
SC
RI PT
199
. Additionally, a 3D-
. These results are supported by
AC C
EP
TE D
12
15
10
ACCEPTED MANUSCRIPT
coronary artery 12,14,18,19. Furthermore, Dvir et al. even recommend the implantation of a smaller
224
transcatheter heart valve to reduce the risk of coronary obstruction2. This “downsizing” approach
225
is supported by further in vitro investigations of our group20. In addition, the implantation of low-
226
profile, retrievable and repositionable THV, such as the Lotus (Boston Scientific, Marlborough,
227
Massachusetts, USA) or Direct Flow (Direct Flow Medical Inc., Santa Rosa, California, USA),
228
seems to be a valuable option to avoid or deal with coronary obstruction. Successful TAVI-ViV
229
with these THV have already been described21,22. In cases where coronary obstruction is detected,
230
the Lotus and Direct Flow can be retrieved and implanted in a new position. This is
231
supported by Wolf et al., who reported a successful repositioning of a Direct Flow due to right
232
coronary occlusion23. In case of a transapical approach for a TAVI-ViV procedure, the JenaValve
233
(JenaValve Technology Inc., Delaware, USA) might be a valuable option to avoid coronary
234
occlusion. The unique anchoring mechanism of this THV fixes the leaflets of the SAVB and
235
therefore might avoid leaflet dislocation to the coronary ostia.
236
However, due to small patient numbers, the reliability of these strategies is still uncertain. Thus,
237
in patients incorporating the above mentioned risk factors for coronary obstruction, a classical
238
surgical approach might be taken into consideration.
SC
M AN U
TE D
EP
AC C
239
RI PT
223
240
Geometric orifice area
241
TAVI-ViV in non-calcified SAVB led to a significant reduction in GOA. The cause are the
242
different sizes of SAVB (size 25) and THV (size 23). Furthermore, there was a significant
243
difference between the Trifecta™ and the Perimount® Magna Ease, because the externally
244
mounted leaflets of a Trifecta™ allow a larger GOA than the geometry of a Perimount® Magna
245
Ease of identical size 6. This observation is important with regard to TAVI-ViV procedures in 11
ACCEPTED MANUSCRIPT
degenerated SAVB without stenosis. In this case, regurgitation is treated successfully but to the
247
cost of a considerable reduction of GOA. During pre-interventional planning, this fact ought to be
248
taken into account. Ruel et al. support this thesis by suggesting a correlation between a small
249
GOA and a significantly higher rate of post-procedural heart failure 24.
250
TAVI-ViV in “calcified” SAVB resulted in an increase in GOA. The artificial “calcification” of
251
the SAVB caused a reduction of GOA compared to the non-calcified model and nullified the
252
geometrical discrepancy between Trifecta™ and Perimount® Magna Ease. Subsequently, there
253
was no significant influence by the SAVB type.
M AN U
SC
RI PT
246
254 Pressure gradients
256
TAVI-ViV in non-calcified SAVB increased dPmean significantly. The increase is caused by the
257
reduction of GOA mentioned above. The results of this experimental run showed a significant
258
difference between Trifecta™ and Perimount® Magna Ease because a non-calcified Trifecta™
259
owns lower pressure gradients than a Perimount® Magna Ease of the same size 6. Inevitably,
260
TAVI-ViV in Trifecta™ causes a higher increase in pressure gradients than in Perimount®
261
Magna Ease. According to Ruel et al., this is clinically important because of a linear relation
262
between an increase in pressure gradients and a development of heart failure 24.
263
In the “calcified” model, TAVI-ViV decreased dPmax and dPmean significantly. No influence of the
264
SAVB type could be observed, because the artificial calcification nullified the original
265
geometrical characteristics of the Trifecta™ and the Perimount® Magna Ease. The clinical
266
relevance of this reduction in pressure gradients after TAVI-ViV is supported by Chan et al.
267
According to these authors, a reduction of pressure gradients led to a significantly lower number
268
of heart failure incidents in a selected patient cohort compared to patients whose pressure
AC C
EP
TE D
255
12
ACCEPTED MANUSCRIPT
269
gradients remained at the same level 25.
270 Overlapping of Trifecta leaflets
272
Another finding during our tests was an overlapping of the Trifecta™ leaflets around and above
273
the Sapien XT™ (Figure 5) in the non-calcified as well as the “calcified” model. This finding
274
was similar in all tests and is likely caused by the high profile of the Trifecta. This overlapping
275
of the Trifecta leaflets may represent a barrier for forward flow and therefore may lead to
276
pressure gradients, which are not caused by the implanted THV. Furthermore, turbulence may
277
occur downstream of the overlapping leaflets, potentially creating an additional risk of thrombus
278
formation behind the Trifecta™ with unknown consequences. The overlapping of leaflet tissue
279
might be avoided by the implantation of a THV with a higher stent frame compared to the Sapien
280
XT, for example the EvolutR (Medtronic Inc., Minneapolis, USA).
281
TE D
M AN U
SC
RI PT
271
Limitations
283
Valves: Only two Sapien XT were available for the execution of forty TAVI-ViV procedures.
284
Hence, each valve needed to be crimped and dilated as well as explanted from the SAVB several
285
times, leading to an increasing deformation of the THV.
286
Valve “calcification”: The artificial “calcification” using glue is only an approximation to leaflet
287
calcification in vivo. Though the leaflets were stiffened and thickened, transvalvular gradients
288
remained smaller than one would expect in severe aortic valve stenosis. This is probably related
289
to the specific material properties of the glue, which might be less rigid than real calcification.
290
Further limitations are the fact, that we did not study the effect of the location and the shape of
AC C
EP
282
13
ACCEPTED MANUSCRIPT
leaflet calcification and the variations in leaflet height, which both may be potential risk factors
292
for coronary obstruction.
293
Aortic root models: In contrast to the human aorta, the aortic root models lacked elasticity, thus
294
there was no alteration of its diameter during systole and diastole. Furthermore, the sinuses of the
295
aortic root models were all of the same size but in vivo, the right coronary sinus is larger in size
296
than the left and non-coronary sinuses. This may lead to a more frequent obstruction of the left
297
coronary ostium compared to an obstruction of the right coronary ostium. We also simulated only
298
a single size of the aortic bulge but human aortic roots with the same size of the annulus also
299
show smaller sinuses´ sizes than the aortic root models do 26. This anatomy was seen in a patient
300
suffering from coronary obstruction after TAVI-ViV 1.
M AN U
SC
RI PT
291
301
Conclusions
TE D
302
The supposition of SAVB with externally mounted leaflets (Trifecta) decreasing coronary flow
304
pathologically could not be verified. In principle, TAVI-ViV in SAVB with externally as well as
305
internally mounted leaflets is a feasible treatment option for patients with degenerated SAVB and
306
aortic root diameters identical to our limited model, though a detailed pre-interventional planning
307
concerning the individual anatomy of the aortic root and the characteristics of the SAVB is
308
essential because of the variety of pathologies.
AC C
309
EP
303
14
ACCEPTED MANUSCRIPT
Funding
311
This work was supported by the German Heart Foundation/German Foundation of Heart
312
Research [grant number F/30/12].
313
RI PT
310
Acknowledgements
315
We would like to thank Michael Diwoky and Tobias Frin for their excellent data management
316
and analyses and for their assistance in preparing this manuscript for publication.
SC
314
M AN U
317
Conflicts of interest
319
Sina Stock, MD, Thorsten Hanke, MD, Efstratios I. Charitos, MD, PhD, Doreen Richardt, MD
320
and Hans-H. Sievers, MD received travel grants from St. Jude Medical and Edwards
321
Lifesciences. Efstratios I. Charitos, MD, PhD holds significant stock of Edwards Lifesciences.
322
Thorsten Hanke, MD is a consultant for St. Jude Medical. The senior author created the
323
hypothesis and reviewed the manuscript, supervised by the corresponding author. The manuscript
324
was primarily written by the first author. In addition, the experiments were performed solely by
325
the first author as well as the engineer Michael Scharfschwerdt, PhD, and the statistical work by
326
Efstratios Charitos, MD, PhD.
EP
AC C
327
TE D
318
15
ACCEPTED MANUSCRIPT
References
328 329
1.
Gurvitch R, Cheung A, Bedogni F, Webb JG. Coronary obstruction following transcatheter aortic valve-in-valve implantation for failed surgical bioprostheses. Catheter Cardiovasc
331
Interv. 2011;77(3):439-444. doi:10.1002/ccd.22861.
332
2.
RI PT
330
Dvir D, Webb J, Brecker S, et al. Transcatheter aortic valve replacement for degenerative bioprosthetic surgical valves: results from the global valve-in-valve registry. Circulation.
334
2012;126(19):2335-2344. doi:10.1161/CIRCULATIONAHA.112.104505. 3.
Dvir D, Barbanti M, Tan J, Webb JG. Transcatheter aortic valve-in-valve implantation for
M AN U
335
SC
333
336
patients with degenerative surgical bioprosthetic valves. Curr Probl Cardiol.
337
2014;39(1):7-27. doi:10.1016/j.cpcardiol.2013.10.001.
338
4.
Gurvitch R, Cheung A, Ye J, et al. Transcatheter valve-in-valve implantation for failed surgical bioprosthetic valves. J Am Coll Cardiol. 2011;58(21):2196-2209.
340
doi:10.1016/j.jacc.2011.09.009. 5.
Ye J, Webb JG, Cheung A, et al. Transapical transcatheter aortic valve-in-valve
EP
341
TE D
339
implantation: clinical and hemodynamic outcomes beyond 2 years. J Thorac Cardiovasc
343
Surg. 2013;145(6):1554-1562. doi:10.1016/j.jtcvs.2012.05.016.
344
6.
AC C
342
Wendt D, Thielmann M, Plicht B, et al. The new St Jude Trifecta versus Carpentier-
345
Edwards Perimount Magna and Magna Ease aortic bioprosthesis: is there a hemodynamic
346
superiority? J Thorac Cardiovasc Surg. 2014;147(5):1553-1560.
347
doi:10.1016/j.jtcvs.2013.05.045.
348
7.
Ferrari E. Transapical aortic “valve-in-valve” procedure for degenerated stented 16
ACCEPTED MANUSCRIPT
bioprosthesis. Eur J Cardiothorac Surg. 2012;41(3):485-490. doi:10.1093/ejcts/ezr027.
349
350
8.
Ribeiro HB, Webb JG, Makkar RR, et al. Predictive factors, management, and clinical outcomes of coronary obstruction following transcatheter aortic valve implantation:
352
insights from a large multicenter registry. J Am Coll Cardiol. 2013;62(17):1552-1562.
353
doi:10.1016/j.jacc.2013.07.040. 9.
Scharfschwerdt M, Misfeld M, Sievers H-H. The influence of a nonlinear resistance
SC
354
RI PT
351
element upon in vitro aortic pressure tracings and aortic valve motions. ASAIO J.
356
2004;50(5):498-502. 10.
2002:7-11.
358
359
Transonic Systems Inc. Flow-Based Intraoperative Coronary Graft Patency Assessment.
11.
Russ C, Hopf R, Hirsch S, et al. Simulation of transcatheter aortic valve implantation
TE D
357
M AN U
355
360
under consideration of leaflet calcification. Conf Proc IEEE Eng Med Biol Soc.
361
2013;2013:711-714. doi:10.1109/EMBC.2013.6609599. 12.
Webb JG, Dvir D. Transcatheter aortic valve replacement for bioprosthetic aortic valve
EP
362
failure: the valve-in-valve procedure. Circulation. 2013;127(25):2542-2550.
364
doi:10.1161/CIRCULATIONAHA.113.000631.
365
13.
Dvir D, Webb JG, Bleiziffer S, et al. Transcatheter aortic valve implantation in failed bioprosthetic surgical valves. JAMA. 2014;312(2):162-170. doi:10.1001/jama.2014.7246.
366
367
AC C
363
14.
Kapadia SR, Svensson L, Tuzcu EM. Successful percutaneous management of left main
368
trunk occlusion during percutaneous aortic valve replacement. Catheter Cardiovasc Interv.
369
2009;73(7):966-972. doi:10.1002/ccd.21867. 17
ACCEPTED MANUSCRIPT
370
15.
Bapat V, Mydin I, Chadalavada S, Tehrani H, Attia R, Thomas M. A guide to fluoroscopic identification and design of bioprosthetic valves: a reference for valve-in-valve procedure.
372
Catheter Cardiovasc Interv. 2013;81(5):853-861. doi:10.1002/ccd.24419.
373
16.
RI PT
371
Haussig S, Schuler G, Linke A. Treatment of a failing St. Jude Medical Trifecta by
Medtronic Corevalve Evolut valve-in-valve implantation. JACC Cardiovasc Interv.
375
2014;7(7):e81-e82. doi:10.1016/j.jcin.2013.11.025.
376
17.
SC
374
Dvir D, Leipsic J, Blanke P, et al. Coronary obstruction in transcatheter aortic valve-invalve implantation: preprocedural evaluation, device selection, protection, and treatment.
378
Circ Cardiovasc Interv. 2015;8(1):e002079-e002079.
379
doi:10.1161/CIRCINTERVENTIONS.114.002079.
380
18.
M AN U
377
Chakravarty T, Jilaihawi H, Nakamura M, et al. Pre-emptive positioning of a coronary stent in the left anterior descending artery for left main protection: a prerequisite for
382
transcatheter aortic valve-in-valve implantation for failing stentless bioprostheses?
383
Catheter Cardiovasc Interv. 2013;82(4):E630-E636. doi:10.1002/ccd.25037. 19.
EP
384
TE D
381
Urena M, Nombela-Franco L, Doyle D, et al. Transcatheter aortic valve implantation for the treatment of surgical valve dysfunction (“valve-in-valve”): assessing the risk of
386
coronary obstruction. J Card Surg. 2012;27(6):682-685. doi:10.1111/jocs.12003.
387
20.
AC C
385
Stock S, Scharfschwerdt M, Meyer-Saraei R, et al. Does Undersizing of Transcatheter
388
Aortic Valve Bioprostheses during Valve-in-valve Implantation Avoid Coronary
389
Obstruction? An In Vitro Study. The Thoracic and Cardiovascular Surgeon. 2016;64(S
390
01). doi:10.1055/s-0036-1571726.
18
ACCEPTED MANUSCRIPT
391
21.
Conradi L, Silaschi M, Seiffert M, et al. Transcatheter valve-in-valve therapy using 6 different devices in 4 anatomic positions: Clinical outcomes and technical considerations.
393
J Thorac Cardiovasc Surg. 2015;150(6):1557–65–1567.e1–3–discussion1565–7.
394
doi:10.1016/j.jtcvs.2015.08.065.
395
22.
RI PT
392
Castriota F, Cavazza C, Secco GG, Micari A, Cremonesi A. First Lotus aortic valve-invalve implantation to treat degenerated Mitroflow bioprostheses. EuroIntervention.
397
2016;11(13):1545-1548. doi:10.4244/EIJY15M07_01. 23.
Wolf A, Schmitz T, Latib A, Naber CK. Successful repositioning of a direct flow medical
M AN U
398
SC
396
399
25-mm valve due to acute occlusion of right coronary artery during transcatheter aortic
400
valve replacement procedure. JACC Cardiovasc Interv. 2015;8(2):e33-e34.
401
doi:10.1016/j.jcin.2013.11.030. 24.
Ruel M, Rubens FD, Masters RG, Pipe AL, Bédard P, Mesana TG. Late incidence and
TE D
402
predictors of persistent or recurrent heart failure in patients with mitral prosthetic valves. J
404
Thorac Cardiovasc Surg. 2004;128(2):278-283. doi:10.1016/j.jtcvs.2003.11.048.
405
25.
EP
403
Chan V, Rubens F, Boodhwani M, Mesana T, Ruel M. Determinants of persistent or recurrent congestive heart failure after contemporary surgical aortic valve replacement. J
407
Heart Valve Dis. 2014;23(6):665-670.
408 409
26.
AC C
406
Berdajs D. The anatomy of the aortic root. Cardiovascular Surgery. 2002;10(4):320-327. doi:10.1016/S0967-2109(02)00018-2.
410
19
ACCEPTED MANUSCRIPT
Figure legends
412
Figure 1:
413
Above: photograph of Perimount Magna Ease (left) and Trifecta (right). Below: Schematic
414
drawing. The pericardial leaflets (red) are mounted inside the stent frame in Perimount Magna
415
Ease (A) and outside in Trifecta(B).
RI PT
411
SC
416 Figure 2:
418
Photograph of an aortic root model. Two Dacron prostheses (8mm diameter), simulating the
419
coronary arteries, are anastomosed with an aortic sinus prosthesis (26mm diameter).
M AN U
417
420 Figure 3:
422
Schematic drawing of the coronary flow device. The left and right coronary artery (LC, RC)
423
running out of the aortic root model (4) pass a pressurized power chamber each (1). The pressure
424
is regulated by a pneumatic pump (3). The coronary flow is measured at the volume measuring
425
points (Q).
EP
426
TE D
421
Figure 4:
428
Left (above) and right (below) coronary flow curves, measured during an experimental run. For
429
the statistical analysis in this study, the coronary flow during diastole was used (area between A
430
and B).
AC C
427
431 432
Figure 5:
20
ACCEPTED MANUSCRIPT
Aortic root model with valve-in-valve implantation of a Sapien XT in a non-calcified (left) and
434
“calcified” (right) Trifecta. The red arrows indicate the remaining space between leaflet tissue
435
and coronary ostium. This space is obviously reduced in the “calcified” model. The black arrow
436
indicates the “calcification” of a Trifecta, simulated with glue (see Material and Methods
437
section).
RI PT
433
438 eFigure 1:
440
Flow chart of the test protocol. In each aortic root model, we implanted the different types of
441
bioprostheses. Afterwards, this was always followed by a transcatheter aortic valve-in-valve
442
implantation.
M AN U
SC
439
443 444 Video legend
446
The video shows the experimental procedure of our in vitro study. A Trifecta SAVB, size 25,
447
was implanted into the aortic root model with 8mm coronary ostia height and the conduit was
448
inserted into the mock circulation. Afterwards, the TAVI-ViV with a Sapien XT THV, size 23,
449
was performed and the conduit was inserted into the mock circulation likewise.
AC C
EP
TE D
445
21
ACCEPTED MANUSCRIPT
Table 1. Coronary flow (mean value ± standard deviation) TRI
TRI
PERI
PERI
RCF [ml/stroke]
CH 8mm
0.64 ± 0.06
0.60 ± 0.07
0.62 ± 0.08
0.61 ± 0.07
CH 10mm
0.62 ± 0.06
0.58 ± 0.04
SC
non-calcified
+ TAVI-ViV
RI PT
+ TAVI-ViV
0.60 ± 0.06
M AN U
“calcified”
0.62 ± 0.04
CH 8mm
0.67 ± 0.05
0.62 ± 0.04
0.65 ± 0.06
0.69 ± 0.17
CH 10mm
0.64 ± 0.08
0.58 ± 0.06
0.58 ± 0.11
0.58 ± 0.03
0.90 ± 0.08
0.87 ± 0.06
0.89 ± 0.06
0.89 ± 0.09
LCF [ml/stroke] non-calcified
CH 10mm
CH 8mm CH 10mm
0.89 ± 0.05
0.82 ± 0.07
0.87 ± 0.04
0.83 ± 0.03
0.85 ± 0.06
0.79 ± 0.08
0.83 ± 0.04
0.77 ± 0.12
0.76 ± 0.04
0.76 ± 0.13
0.80 ± 0.04
EP
“calcified“
TE D
CH 8mm
0.79 ± 0.11
AC C
TRI: Trifecta™, TAVI-ViV: Transcatheter Aortic Valve-in-Valve Implantation, PERI: Perimount® Magna Ease, CH: Coronary height, RCF: right coronary flow, LCF: left coronary flow
ACCEPTED MANUSCRIPT Table 2. Hydrodynamics (mean value ± standard deviation) TRI
TRI
PERI
PERI
+ TAVI-ViV
+ TAVI-ViV
Geometric orifice area
“calcified“ [cm2] 0.94 ± 0.32
1.43 ± 0.12
2.08 ± 0.03
1.3 ± 0.14
0.77 ± 0.36
1.56 ± 0.17
RI PT
non-calcified [cm2] 2.48 ± 0.05
Pressure gradients
SC
non-calcified
1.42 ± 0.17
12 ± 2
13 ± 2
13 ± 2
14 ± 2
dPmean [mmHg]
4±1
5±1
5±1
6±1
dPmax [mmHg]
23 ± 7
16 ± 1
34 ± 20
15 ± 2
dPmean [mmHg]
12 ± 5
6±1
20 ± 13
6±1
“calcified“
M AN U
dPmax [mmHg]
TRI: Trifecta™, TAVI-ViV: Transcatheter Aortic Valve-in-Valve Implantation, PERI:
AC C
EP
TE D
Perimount® Magna Ease, dP: pressure gradient
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
S0022-5223(16)31485-4
DOI:
10.1016/j.jtcvs.2016.09.086
Reference:
YMTC 11039
To appear in:
The Journal of Thoracic and Cardiovascular Surgery
Received Date: 12 April 2016 Revised Date:
22 September 2016
Accepted Date: 24 September 2016
Please cite this article as: Stock S, Scharfschwerdt M, Meyer-Saraei R, Richardt D, Charitos EI, Sievers H-H, Hanke T, In vitro coronary flow after transcatheter aortic valve-in-valve implantation: A comparison of two valves, The Journal of Thoracic and Cardiovascular Surgery (2016), doi: 10.1016/ j.jtcvs.2016.09.086. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT
In vitro coronary flow after transcatheter aortic valve-in-valve implantation: A comparison of two valves.
RI PT
Sina Stock, MD, Michael Scharfschwerdt, PhD, Roza Meyer-Saraei1, PhD, Doreen Richardt, MD, Efstratios I. Charitos2, MD, PhD, Hans-Hinrich Sievers, MD, Thorsten Hanke, MD
SC
Department of Cardiac and Thoracic Vascular Surgery
Address for correspondence: Hans-Hinrich Sievers, MD
M AN U
University of Luebeck, Germany
Department of Cardiac and Thoracic Vascular Surgery, University of Luebeck
TE D
Ratzeburger Allee 160, 23538 Luebeck, Germany Tel.: +49 451 500 2108, Fax: +49 451 500 2051
EP
E-mail: [email protected]
AC C
Sina Stock, MD and Michael Scharfschwerdt, PhD contributed equally to this work.
Article word count: 3493
Funding: This work was supported by the German Heart Foundation/German Foundation of Heart Research [grant number F/30/12]
1 2
University of Luebeck, Department of Cardiology, Angilogy and Intensive Care Medicine, Luebeck, Germany University of Halle (Saale), Department of Cardiac Surgery, Halle (Saale), Germany
ACCEPTED MANUSCRIPT Conflicts of interest: Sina Stock, MD, Thorsten Hanke, MD, Efstratios I. Charitos, MD, PhD, Doreen Richardt, MD and Hans-H. Sievers, MD received travel grants from St. Jude Medical and Edwards Lifesciences. Efstratios I. Charitos, MD, PhD holds significant stock of Edwards Lifesciences. Thorsten Hanke, MD is a consultant for St. Jude Medical. The senior
RI PT
author created the hypothesis and reviewed the manuscript, supervised by the corresponding author. The manuscript was primarily written by the first author. In addition, the experiments were performed solely by the first author as well as the engineer Michael Scharfschwerdt,
AC C
EP
TE D
M AN U
SC
PhD, and the statistical work by Efstratios Charitos, MD, PhD.
ACCEPTED MANUSCRIPT Glossary of abbreviations Transcatheter Aortic Valve-in-Valve Implantation
SAVB
Surgical Aortic Valve Bioprosthesis
GOA
Geometric Orifice Area
THV
Transcatheter Heart Valve
RCF
Right Coronary Flow
LCF
Left Coronary Flow
dPmean
Mean Pressure Gradient
dPmax
Maximum Pressure Gradient
AC C
EP
TE D
M AN U
SC
RI PT
TAVI-ViV
ACCEPTED MANUSCRIPT Abstract Objective: The Transcatheter Aortic Valve-in-Valve Implantation (TAVI-ViV) is an evolving treatment strategy for degenerated surgical aortic valve bioprostheses (SAVB). However, there is some
RI PT
concern regarding coronary obstruction, especially after TAVI-ViV in calcified SAVB with externally mounted leaflets. We investigated in vitro coronary flow and hydrodynamics after
SC
TAVI-ViV in two modern SAVB with externally and internally mounted leaflets.
M AN U
Methods:
Aortic root models including known risk factors for coronary obstruction served for the implantation of SAVB with either externally (St. Jude Trifecta™, size 25) or internally (Edwards Perimount® Magna Ease, size 25) mounted leaflets. Left and right coronary flow as well as hydrodynamics were measured before and after TAVI-ViV with an Edwards Sapien
TE D
XT™, size 23. After the first experimental run, the SAVB leaflets were artificially “calcified”
Results:
EP
and the measurements were repeated.
AC C
In both models, non-calcified and “calcified”, we found no significant reduction in coronary flow, neither when testing Trifecta nor Perimount Magna Ease. Mean pressure gradient increased after TAVI-ViV in the non-calcified model (Trifecta p=0.0001, Perimount Magna Ease p=0.006) and geometric orifice area decreased (both p<0.001). In the “calcified“ model, mean pressure gradient decreased (both p<0.001) and geometric orifice area increased (both p<0.001).
Conclusions:
ACCEPTED MANUSCRIPT In our specific model, TAVI-ViV is feasible in different SAVB (St. Jude Medical Trifecta and Edwards Perimount Magna Ease), non-calcified as well as “calcified”, without risk of coronary obstruction. Nevertheless, prior to clinical application of these results, preoperative
RI PT
thorough assessment including the different limitations of this model is mandatory.
AC C
EP
TE D
M AN U
SC
Abstract word count: 243
ACCEPTED MANUSCRIPT Central message This limited in vitro study shows no impairment of coronary flow after transcatheter aortic valve-in-valve implantation in surgical bioprostheses with internally and externally mounted
RI PT
leaflets.
AC C
EP
TE D
M AN U
SC
(Word count: 25, Characters: 194)
ACCEPTED MANUSCRIPT Perspective Statement Transcatheter aortic valve-in-valve implantation is an emerging treatment for failing bioprostheses potentially reducing coronary flow, especially in bioprostheses with externally mounted leaflets. This was not found in this limited in vitro study comparing bioprostheses
consideration prior to clinical application.
AC C
EP
TE D
M AN U
SC
(Word count: 51, Characters: 400)
RI PT
with externally and internally mounted leaflets. The limitations of this model need
ACCEPTED MANUSCRIPT
1
In vitro coronary flow after transcatheter aortic valve-in-valve implantation:
2
A comparison of two valves.
RI PT
3 Introduction and background
5
The Transcatheter Aortic Valve-in-Valve Implantation technique (TAVI-ViV) is an appealing
6
treatment option for patients with degenerated surgical aortic valve bioprostheses (SAVB) and
7
where a high surgical risk is expected. However, apart from many advantages compared to re-do
8
cardiac surgery, there is some concern regarding coronary obstruction as a potential life-
9
threatening complication 1. In the global Valve-in-Valve registry, this complication has been
10
validated in 3.5% of the patients undergoing TAVI-ViV 2. Besides predisposing anatomic factors,
11
such as a low coronary ostia height, a narrow aortic root or sinotubular junction, it is feared that
12
the geometry of the primarily implanted SAVB and its extension of calcification may influence
13
the risk of coronary obstruction 1-5.
14
Considering modern stented SAVB, two different geometric types are to be distinguished: Firstly,
15
SAVB with internally mounted leaflets, such as the Edwards Perimount® Magna Ease (Edwards
16
Lifesciences LLC, Irvine, USA) and secondly, SAVB with externally mounted leaflets, such as
17
the St. Jude Trifecta™ (St. Jude Medical Inc., St. Paul, USA). Comparing these two different
18
types, the geometry of SAVB with externally mounted leaflets causes a larger geometric orifice
19
area (GOA) than SAVB with internally mounted leaflets but the wide extension of leaflet tissue
20
beyond the diameter of the valve´s stent may influence coronary flow 6 (Figure 1).
21
Gurvitch et al. describe coronary obstruction after TAVI-ViV with a balloon-expandable
22
transcatheter heart valve (THV, Edwards Sapien XT, size 23, e.g. Medtronic CoreValve, size
AC C
EP
TE D
M AN U
SC
4
1
ACCEPTED MANUSCRIPT
26 [Medtronic, Minnesota, USA]) occurring in two patients with highly degenerated SAVB of
24
the type Sorin Mitroflow, size 21 (Sorin Group Inc, Vancouver, Canada). In this patients, who
25
presented with small aortic roots, the externally mounted leaflets led to a close proximity of valve
26
tissue and coronary ostia. In both patients, the leaflet tissue occluded at least one coronary ostium
27
after TAVI-ViV and caused peri-interventional death 1.
28
The objective of this in vitro investigation is to identify specific differences between TAVI-ViV
29
in geometrically different SAVB with special regard to coronary flow but also considering
30
hydrodynamic performance. Potential coronary obstruction shall be identified leading to
31
strategies in pre-interventional planning to avoid or mitigate this complication.
M AN U
SC
RI PT
23
32
Material and methods
33 Valves
35
Surgical aortic valve bioprostheses: In this in vitro investigation, the SAVB with internally
36
mounted leaflets was represented by the Edwards Perimount® Magna Ease (n=5) and the SAVB
37
with externally mounted leaflets by the St. Jude Trifecta™ (n=5), both size 25. After the first
38
experimental run, leaflet calcification was simulated using glue. The leaflet “calcification” was of
39
alternating thickness with a maximum of 5mm.
40
Transcatheter heart valve: To perform TAVI-ViV, we used the balloon-expandable Edwards
41
Sapien XT™ (n=2), size 23, to achieve optimal hydrodynamics 5,7.
AC C
EP
TE D
34
42 43
Aortic root models
2
ACCEPTED MANUSCRIPT
The design of the aortic root models intended to imitate risk factors that are assumed to favor
45
coronary obstruction. Therefore, it was based on investigations by Ribeiro et al. revealing a
46
distance from coronary ostia to the aortic valve annulus of less than 12mm in 86% of the patients
47
suffering from coronary obstruction 8. The mean distance was 10.6mm. Furthermore, an average
48
diameter of the sinuses of Valsalva of 28.1mm became apparent. We created two aortic root
49
models (Figure 2) with different coronary ostia height (8mm and 10mm) using aortic sinus
50
prostheses of the smallest available size, label size 26mm (Uni-Graft W SINUS, B. Braun,
51
Melsungen, Germany). Sinus prostheses were used to imitate physiological conditions as close as
52
possible. The nomenclature refers to the diameter of the annulus and sinotubular junction,
53
whereas the diameter of the sinuses is 33mm. A circular strengthening with felt, situated below
54
the sinuses tissue, simulated the aortic annulus and ensured an identical implantation height of the
55
SAVB. The coronary arteries were simulated by an 8mm Dacron prosthesis each (B. Braun,
56
Melsungen, Germany).
TE D
M AN U
SC
RI PT
44
57
Physiological mock circulation
59
The physiological circulation was imitated by a pulse duplicator allowing for the evaluation of
60
hydrodynamic parameters 9. It enables the adjustment of different cardiac output volumes,
61
afterload and heart rates. A camera on top of the aortic root models made a visual observation
62
and pictures possible.
AC C
EP
58
63 64
Coronary flow device
65
In vivo, the right and left coronary flow (RCF and LCF) depend on the different myocardial flow
3
ACCEPTED MANUSCRIPT
resistance during a heart cycle. The contraction of the heart muscle in systole leads to a higher
67
flow resistance and subsequently to less coronary perfusion compared to diastole. This effect is
68
stronger in the left ventricle than in the right ventricle10.
69
To imitate this physiological coronary flow, the Dacron prostheses, simulating the coronary
70
arteries, were connected to the coronary flow device (Figure 3), which consists of two sealed
71
power chambers with elastic tubes inside. During systole, the power chambers can be
72
pressurized, leading to compression of the elastic tubes and subsequently to reduction of the
73
coronary flow as mentioned above. The power chambers surrounding the right and left coronary
74
artery can be pressurized independently. During diastole, the power chambers are decompressed.
75
The flow curves for LCF and RCF in our model (Figure 4) are comparable to the physiological
76
coronary flow and prove the validity of this coronary flow device.
77
For the experimental tests, pressurization was based on the Flow-based Intraoperative Coronary
78
Graft Patency Assessment 10.
TE D
M AN U
SC
RI PT
66
79 Experimental procedure
81
Per aortic root model, SAVB type and experimental run, we performed five measurements
82
(eFigure 1) taking the mean value.
83
For the first experimental run, the Trifecta™ respectively the Perimount® Magna Ease was
84
sewed into an aortic root model. Thereafter, we inserted the conduit into the mock circulation 9.
85
We determined the values of the following parameters: diastolic LCF and RCF, pressure
86
gradients dPmean and dPmax. The identification of the GOA resulted from photographs taken by the
87
high-speed camera. Subsequently, we explanted the conduit, performed TAVI-ViV with the
AC C
EP
80
4
ACCEPTED MANUSCRIPT
Sapien XT™, commissure-to-commissure with the SAVB, and repeated all measurements.
89
For the second experimental run, we simulated “calcifications” of the SAVB by using glue
90
(Figure 5) and repeated the tests according to the protocol. To simulate leaflet thickening as well
91
as leaflet stiffening, we used two glues with different viscosities. At first, the leaflets of the
92
SAVB were stiffened by applying liquid glue (SEKUNDEN ALLESKLEBER geruchsfrei
93
EASY, UHU GmbH, Buehl/Baden, Germany) on the entire leaflet surface. This was followed
94
by the punctual application of a viscous gel glue (SEKUNDENKLEBER blitzschnell
95
SUPERGEL, UHU GmbH, Buehl/Baden, Germany) on the entire leaflet surface in an eccentric
96
shape, especially across the central section and free edge, to simulate valve thickening and
97
calcific plaques.
98
During the measurements, diastolic pressure was 80mmHg and systolic 120mmHg, stroke rate 64
99
beats per minute and stroke volume 70ml. The test solution was represented by a physiological
100
saline (0.9%) with a density of 1.0046g/cm3 and a dynamic viscosity of 0.9mPa·s at an ambient
101
temperature of 20°C.
102
Concerning coronary flow, the maximum LCF was determined 120ml/min and RCF 80ml/min
103
10
SC
M AN U
TE D
EP
.
AC C
104
RI PT
88
105
Technique of measurement
106
The left ventricular (4cm below the aortic valve) and aortic pressure (6cm above the aortic valve)
107
were measured with two capacitive pressure transducers Envec Ceracore M (Endress + Hauser,
108
Maulburg, Germany), calibrated to a measuring range of -20 to +160mmHg and a resolution of
109
0.02mmHg. 5
ACCEPTED MANUSCRIPT
The sensor of an ultrasonic flowmeter TS-410 (Transonic System Inc., Ithaca, USA) was
111
mounted directly below the aortic valve to record the volume flow through the valve. The sensor
112
works bi-directionally with a resolution of 2ml/min and records flow rates up to 20l/min.
113
LCF and RCF were measured with the ultrasonic device TS-420 and coronary probes of 6mm
114
(Transonic System Inc., Ithaca, USA).
115
A high-speed camera Motionscope HR-1000 (Redlake Imaging Corp., Morgan Hill, USA) above
116
the conduit recorded the characteristics of motion of the aortic valves with 500 pictures per
117
second. Video recordings and flow measurements were started simultaneously by using trigger
118
signals.
M AN U
SC
RI PT
110
119 Analysis and statistics
121
The values for pressure and flow were registered by an analogue-digital converter recording 500
122
individual values per second. Each measurement included ten successive cardiac cycles in order
123
to calculate the mean value ± standard deviation (SD). The analysis of the measurement results
124
was based on the international norm of testing cardiac valve prosthesis (ISO 5840:
125
Cardiovascular implants – Cardiac valve prostheses).
126
The evaluation of GOA was based on photographs taken by the high-speed camera and calculated
127
with ImageJ (NIH Image), using the inner diameter of SAVB as reference value.
128
Statistically, the T-test or the Mann-Whitney U test were applied depending on the given
129
distribution. The level of significance was defined p<0.05. Data processing programs (R version
130
3.1.1 and IBM SPSS statistics version 22) were used for statistical evaluation.
AC C
EP
TE D
120
131 6
ACCEPTED MANUSCRIPT
Results
132 Coronary flow
134
Significant differences in LCF and RCF between the different SAVB types before and after
135
TAVI-ViV as well as coronary height could not be observed, neither in the non-calcified nor in
136
the “calcified” model (all p-values were non-significant) (Table 1).
137
Right coronary flow: The coronary flow after TAVI-ViV in non-calcified SAVB dropped by 7%
138
(coronary height 8mm) and 9% (coronary height 10mm) in Trifecta™ and by 2% (coronary
139
height 8mm) and 3% (coronary height 10mm) in Perimount® Magna Ease. The coronary flow
140
after TAVI-ViV in “calcified” SAVB dropped by 8% (coronary height 8mm) and 10% (coronary
141
height10mm) in Trifecta™. Considering Perimount® Magna Ease, flow after TAVI-ViV
142
increased by 6% (coronary height 8mm) and remained equal (coronary height 10mm).
143
Left coronary flow: After TAVI-ViV in non-calcified SAVB, LCF decreased by 3% (coronary
144
height 8mm) and 9% (coronary height 10mm) in Trifecta™ and remained equal (coronary height
145
8mm) and dropped by 5% (coronary height 10mm) in Perimount® Magna Ease. TAVI-ViV in
146
“calcified” SAVB caused a flow reduction of 8% (coronary height 8mm) and 4% (coronary
147
height 10mm) in Trifecta™ and of 8% (coronary height 8mm) while flow increased by 6%
148
(coronary height 10mm) in Perimount® Magna Ease.
SC
M AN U
TE D
EP
AC C
149
RI PT
133
150
Geometric orifice area
151
(Table 2)
152
Non-calcified model: TAVI-ViV in Trifecta™ and in Perimount® Magna Ease resulted in a
153
decrease of GOA (both p<0.001). We found a significant difference between the two different
154
SAVB types (p<0.001). 7
ACCEPTED MANUSCRIPT
“Calcified” model: The GOA increased after TAVI-ViV in Trifecta™ and in Perimount® Magna
156
Ease (both p<0.001). A statistically significant difference between the two different types of
157
SAVB could not be observed (p-value non-significant).
RI PT
155
158 Pressure gradients
160
(Table 2)
161
Non-calcified model: TAVI-ViV in Trifecta™ and in Perimount® Magna Ease resulted in an
162
increase of dPmax (both p-value non-significant) and of dPmean (p=0.0001, p=0.006). dPmean showed a
163
significant difference between the SAVB types before TAVI-ViV (p=0.009) and dPmax after
164
TAVI-ViV (p=0.036).
165
“Calcified” model: dPmax and dPmean decreased after TAVI-ViV in Trifecta™ and in Perimount®
166
Magna Ease (dPmax: p=0.0003, p=0.0004; dPmean: both p<0.001). Both parameters showed no
167
significant difference between the SAVB types (p-values non-significant).
TE D
M AN U
SC
159
EP
168
Overlapping of Trifecta leaflets
170
The implantation of the Sapien XT in the Trifecta resulted in an overlapping of the Trifecta
171
leaflets around and above the THV (Figure 5). This overlapping occurred in all TAVI-ViV
172
procedures during our test, in the non-calcified model as well as in the “calcified”.
AC C
169
173 Discussion
174 175
Coronary flow 8
ACCEPTED MANUSCRIPT
This in vitro study provides no evidence of a significant difference in coronary flow after TAVI-
177
ViV in SAVB with externally and internally mounted leaflet tissue (Trifecta and Perimount
178
Magna Ease).
179
To date, there are several risk factors known to increase the risk of coronary obstruction. Besides
180
predisposing anatomic aspects, such as a low coronary ostia height or a narrow aortic root and
181
sinotubular junction, the characteristics of the SAVB need to be considered. Risk factors for
182
coronary obstruction are a supra-annular implantation, a high profile of the SAVB and its
183
locational relation to the coronary ostia 1,11,12. Due to clinical studies, Dvir et al. hypothesized that
184
coronary obstruction occurs preferentially in modern SAVB with externally mounted leaflet
185
tissue, such as the Trifecta™
186
occurred more frequently in patients with calcified SAVB compared to patients with degenerated
187
regurgitant SAVB only
188
displacement of calcified and thickened tissue close to the coronary ostia 14.
189
The present in vitro study aims to investigate further these theses mentioned by Dvir et al. in
190
order to improve the treatment of patients with degenerated SAVB and to foresee an increased
191
risk of coronary obstruction.
192
The results of the present study could not verify these theses. A critical obstruction of coronary
193
flow after TAVI-ViV did not occur in SAVB with internally or externally mounted leaflets in the
194
chosen model. In both experimental runs – the non-calcified as well as the “calcified” model –
195
even with low located coronary ostia, TAVI-ViV did not change the profile of coronary flow
196
significantly. The distance between the coronary ostia and leaflet tissue of the SAVB remained
197
sufficiently in the specific size of the aortic root models used in this study (Figure 5). However,
198
in the calcified model, this space between the coronary ostia and the leaflets was reduced (Figure
. Furthermore, Dvir et al. pointed out, that coronary obstruction
. As the main reason for this fatal complication, they suggested the
AC C
EP
TE D
13
1,2
M AN U
SC
RI PT
176
9
ACCEPTED MANUSCRIPT
5). It is imaginable, that with an increasing amount of calcific deposits, this space could be
200
further reduced causing coronary obstruction. Nevertheless, TAVI-ViV in both modern SAVB
201
types is proved to be feasible and the risk of coronary obstruction is estimated similarly in
202
patients with an anatomy alike the geometry of our aortic root models.
203
To minimize the risk of coronary obstruction in general, an exact pre-interventional planning
204
seems to be a pre-requisite 1. A cardio-CT prior to a TAVI-ViV procedure enables exact images
205
of a patient´s anatomy. Apart from aortic root characteristics, such as diameter and height of the
206
sinuses and the sinotubular junction as well as coronary ostia height, the characteristics and
207
radiological appearance of the degenerated SAVB should be considered
208
echocardiography with aortography is recommended
209
investigations of Gurvitch et al. and Ye et al. 4,5. According to these authors, the major influence
210
on coronary perfusion is not only the primarily used SAVB but also a precise pre-interventional
211
planning to avoid coronary occlusion. This is aligned with Linke et al., who reported a successful
212
TAVI-ViV into a modern SAVB (Trifecta, size 23) without coronary obstruction, using exact
213
pre-interventional imaging 5,16.
214
However, risk factors for coronary obstruction shown by pre-interventional imaging are not
215
considered to be an absolute contraindication for TAVI-ViV. At present, there are already some
216
existing strategies to avoid coronary occlusion. Intraoperative transesophageal echocardiography
217
and angiography enable a fast detection of coronary occlusion. The intra-procedural inflation of a
218
balloon is a useful technique to estimate the risk of coronary obstruction during a TAVI-ViV
219
procedure. A balloon similarly sized to the intended THV results in temporary displacement of
220
the SAVB leaflets and reveals by means of aortography the remaining space between coronary
221
ostia and SAVB leaflets17. The intraoperative placement of a guide wire in the coronary artery at
222
risk seems to avoid a dislocation of leaflet tissue as well as to ensure an emergency access to the
M AN U
SC
RI PT
199
. Additionally, a 3D-
. These results are supported by
AC C
EP
TE D
12
15
10
ACCEPTED MANUSCRIPT
coronary artery 12,14,18,19. Furthermore, Dvir et al. even recommend the implantation of a smaller
224
transcatheter heart valve to reduce the risk of coronary obstruction2. This “downsizing” approach
225
is supported by further in vitro investigations of our group20. In addition, the implantation of low-
226
profile, retrievable and repositionable THV, such as the Lotus (Boston Scientific, Marlborough,
227
Massachusetts, USA) or Direct Flow (Direct Flow Medical Inc., Santa Rosa, California, USA),
228
seems to be a valuable option to avoid or deal with coronary obstruction. Successful TAVI-ViV
229
with these THV have already been described21,22. In cases where coronary obstruction is detected,
230
the Lotus and Direct Flow can be retrieved and implanted in a new position. This is
231
supported by Wolf et al., who reported a successful repositioning of a Direct Flow due to right
232
coronary occlusion23. In case of a transapical approach for a TAVI-ViV procedure, the JenaValve
233
(JenaValve Technology Inc., Delaware, USA) might be a valuable option to avoid coronary
234
occlusion. The unique anchoring mechanism of this THV fixes the leaflets of the SAVB and
235
therefore might avoid leaflet dislocation to the coronary ostia.
236
However, due to small patient numbers, the reliability of these strategies is still uncertain. Thus,
237
in patients incorporating the above mentioned risk factors for coronary obstruction, a classical
238
surgical approach might be taken into consideration.
SC
M AN U
TE D
EP
AC C
239
RI PT
223
240
Geometric orifice area
241
TAVI-ViV in non-calcified SAVB led to a significant reduction in GOA. The cause are the
242
different sizes of SAVB (size 25) and THV (size 23). Furthermore, there was a significant
243
difference between the Trifecta™ and the Perimount® Magna Ease, because the externally
244
mounted leaflets of a Trifecta™ allow a larger GOA than the geometry of a Perimount® Magna
245
Ease of identical size 6. This observation is important with regard to TAVI-ViV procedures in 11
ACCEPTED MANUSCRIPT
degenerated SAVB without stenosis. In this case, regurgitation is treated successfully but to the
247
cost of a considerable reduction of GOA. During pre-interventional planning, this fact ought to be
248
taken into account. Ruel et al. support this thesis by suggesting a correlation between a small
249
GOA and a significantly higher rate of post-procedural heart failure 24.
250
TAVI-ViV in “calcified” SAVB resulted in an increase in GOA. The artificial “calcification” of
251
the SAVB caused a reduction of GOA compared to the non-calcified model and nullified the
252
geometrical discrepancy between Trifecta™ and Perimount® Magna Ease. Subsequently, there
253
was no significant influence by the SAVB type.
M AN U
SC
RI PT
246
254 Pressure gradients
256
TAVI-ViV in non-calcified SAVB increased dPmean significantly. The increase is caused by the
257
reduction of GOA mentioned above. The results of this experimental run showed a significant
258
difference between Trifecta™ and Perimount® Magna Ease because a non-calcified Trifecta™
259
owns lower pressure gradients than a Perimount® Magna Ease of the same size 6. Inevitably,
260
TAVI-ViV in Trifecta™ causes a higher increase in pressure gradients than in Perimount®
261
Magna Ease. According to Ruel et al., this is clinically important because of a linear relation
262
between an increase in pressure gradients and a development of heart failure 24.
263
In the “calcified” model, TAVI-ViV decreased dPmax and dPmean significantly. No influence of the
264
SAVB type could be observed, because the artificial calcification nullified the original
265
geometrical characteristics of the Trifecta™ and the Perimount® Magna Ease. The clinical
266
relevance of this reduction in pressure gradients after TAVI-ViV is supported by Chan et al.
267
According to these authors, a reduction of pressure gradients led to a significantly lower number
268
of heart failure incidents in a selected patient cohort compared to patients whose pressure
AC C
EP
TE D
255
12
ACCEPTED MANUSCRIPT
269
gradients remained at the same level 25.
270 Overlapping of Trifecta leaflets
272
Another finding during our tests was an overlapping of the Trifecta™ leaflets around and above
273
the Sapien XT™ (Figure 5) in the non-calcified as well as the “calcified” model. This finding
274
was similar in all tests and is likely caused by the high profile of the Trifecta. This overlapping
275
of the Trifecta leaflets may represent a barrier for forward flow and therefore may lead to
276
pressure gradients, which are not caused by the implanted THV. Furthermore, turbulence may
277
occur downstream of the overlapping leaflets, potentially creating an additional risk of thrombus
278
formation behind the Trifecta™ with unknown consequences. The overlapping of leaflet tissue
279
might be avoided by the implantation of a THV with a higher stent frame compared to the Sapien
280
XT, for example the EvolutR (Medtronic Inc., Minneapolis, USA).
281
TE D
M AN U
SC
RI PT
271
Limitations
283
Valves: Only two Sapien XT were available for the execution of forty TAVI-ViV procedures.
284
Hence, each valve needed to be crimped and dilated as well as explanted from the SAVB several
285
times, leading to an increasing deformation of the THV.
286
Valve “calcification”: The artificial “calcification” using glue is only an approximation to leaflet
287
calcification in vivo. Though the leaflets were stiffened and thickened, transvalvular gradients
288
remained smaller than one would expect in severe aortic valve stenosis. This is probably related
289
to the specific material properties of the glue, which might be less rigid than real calcification.
290
Further limitations are the fact, that we did not study the effect of the location and the shape of
AC C
EP
282
13
ACCEPTED MANUSCRIPT
leaflet calcification and the variations in leaflet height, which both may be potential risk factors
292
for coronary obstruction.
293
Aortic root models: In contrast to the human aorta, the aortic root models lacked elasticity, thus
294
there was no alteration of its diameter during systole and diastole. Furthermore, the sinuses of the
295
aortic root models were all of the same size but in vivo, the right coronary sinus is larger in size
296
than the left and non-coronary sinuses. This may lead to a more frequent obstruction of the left
297
coronary ostium compared to an obstruction of the right coronary ostium. We also simulated only
298
a single size of the aortic bulge but human aortic roots with the same size of the annulus also
299
show smaller sinuses´ sizes than the aortic root models do 26. This anatomy was seen in a patient
300
suffering from coronary obstruction after TAVI-ViV 1.
M AN U
SC
RI PT
291
301
Conclusions
TE D
302
The supposition of SAVB with externally mounted leaflets (Trifecta) decreasing coronary flow
304
pathologically could not be verified. In principle, TAVI-ViV in SAVB with externally as well as
305
internally mounted leaflets is a feasible treatment option for patients with degenerated SAVB and
306
aortic root diameters identical to our limited model, though a detailed pre-interventional planning
307
concerning the individual anatomy of the aortic root and the characteristics of the SAVB is
308
essential because of the variety of pathologies.
AC C
309
EP
303
14
ACCEPTED MANUSCRIPT
Funding
311
This work was supported by the German Heart Foundation/German Foundation of Heart
312
Research [grant number F/30/12].
313
RI PT
310
Acknowledgements
315
We would like to thank Michael Diwoky and Tobias Frin for their excellent data management
316
and analyses and for their assistance in preparing this manuscript for publication.
SC
314
M AN U
317
Conflicts of interest
319
Sina Stock, MD, Thorsten Hanke, MD, Efstratios I. Charitos, MD, PhD, Doreen Richardt, MD
320
and Hans-H. Sievers, MD received travel grants from St. Jude Medical and Edwards
321
Lifesciences. Efstratios I. Charitos, MD, PhD holds significant stock of Edwards Lifesciences.
322
Thorsten Hanke, MD is a consultant for St. Jude Medical. The senior author created the
323
hypothesis and reviewed the manuscript, supervised by the corresponding author. The manuscript
324
was primarily written by the first author. In addition, the experiments were performed solely by
325
the first author as well as the engineer Michael Scharfschwerdt, PhD, and the statistical work by
326
Efstratios Charitos, MD, PhD.
EP
AC C
327
TE D
318
15
ACCEPTED MANUSCRIPT
References
328 329
1.
Gurvitch R, Cheung A, Bedogni F, Webb JG. Coronary obstruction following transcatheter aortic valve-in-valve implantation for failed surgical bioprostheses. Catheter Cardiovasc
331
Interv. 2011;77(3):439-444. doi:10.1002/ccd.22861.
332
2.
RI PT
330
Dvir D, Webb J, Brecker S, et al. Transcatheter aortic valve replacement for degenerative bioprosthetic surgical valves: results from the global valve-in-valve registry. Circulation.
334
2012;126(19):2335-2344. doi:10.1161/CIRCULATIONAHA.112.104505. 3.
Dvir D, Barbanti M, Tan J, Webb JG. Transcatheter aortic valve-in-valve implantation for
M AN U
335
SC
333
336
patients with degenerative surgical bioprosthetic valves. Curr Probl Cardiol.
337
2014;39(1):7-27. doi:10.1016/j.cpcardiol.2013.10.001.
338
4.
Gurvitch R, Cheung A, Ye J, et al. Transcatheter valve-in-valve implantation for failed surgical bioprosthetic valves. J Am Coll Cardiol. 2011;58(21):2196-2209.
340
doi:10.1016/j.jacc.2011.09.009. 5.
Ye J, Webb JG, Cheung A, et al. Transapical transcatheter aortic valve-in-valve
EP
341
TE D
339
implantation: clinical and hemodynamic outcomes beyond 2 years. J Thorac Cardiovasc
343
Surg. 2013;145(6):1554-1562. doi:10.1016/j.jtcvs.2012.05.016.
344
6.
AC C
342
Wendt D, Thielmann M, Plicht B, et al. The new St Jude Trifecta versus Carpentier-
345
Edwards Perimount Magna and Magna Ease aortic bioprosthesis: is there a hemodynamic
346
superiority? J Thorac Cardiovasc Surg. 2014;147(5):1553-1560.
347
doi:10.1016/j.jtcvs.2013.05.045.
348
7.
Ferrari E. Transapical aortic “valve-in-valve” procedure for degenerated stented 16
ACCEPTED MANUSCRIPT
bioprosthesis. Eur J Cardiothorac Surg. 2012;41(3):485-490. doi:10.1093/ejcts/ezr027.
349
350
8.
Ribeiro HB, Webb JG, Makkar RR, et al. Predictive factors, management, and clinical outcomes of coronary obstruction following transcatheter aortic valve implantation:
352
insights from a large multicenter registry. J Am Coll Cardiol. 2013;62(17):1552-1562.
353
doi:10.1016/j.jacc.2013.07.040. 9.
Scharfschwerdt M, Misfeld M, Sievers H-H. The influence of a nonlinear resistance
SC
354
RI PT
351
element upon in vitro aortic pressure tracings and aortic valve motions. ASAIO J.
356
2004;50(5):498-502. 10.
2002:7-11.
358
359
Transonic Systems Inc. Flow-Based Intraoperative Coronary Graft Patency Assessment.
11.
Russ C, Hopf R, Hirsch S, et al. Simulation of transcatheter aortic valve implantation
TE D
357
M AN U
355
360
under consideration of leaflet calcification. Conf Proc IEEE Eng Med Biol Soc.
361
2013;2013:711-714. doi:10.1109/EMBC.2013.6609599. 12.
Webb JG, Dvir D. Transcatheter aortic valve replacement for bioprosthetic aortic valve
EP
362
failure: the valve-in-valve procedure. Circulation. 2013;127(25):2542-2550.
364
doi:10.1161/CIRCULATIONAHA.113.000631.
365
13.
Dvir D, Webb JG, Bleiziffer S, et al. Transcatheter aortic valve implantation in failed bioprosthetic surgical valves. JAMA. 2014;312(2):162-170. doi:10.1001/jama.2014.7246.
366
367
AC C
363
14.
Kapadia SR, Svensson L, Tuzcu EM. Successful percutaneous management of left main
368
trunk occlusion during percutaneous aortic valve replacement. Catheter Cardiovasc Interv.
369
2009;73(7):966-972. doi:10.1002/ccd.21867. 17
ACCEPTED MANUSCRIPT
370
15.
Bapat V, Mydin I, Chadalavada S, Tehrani H, Attia R, Thomas M. A guide to fluoroscopic identification and design of bioprosthetic valves: a reference for valve-in-valve procedure.
372
Catheter Cardiovasc Interv. 2013;81(5):853-861. doi:10.1002/ccd.24419.
373
16.
RI PT
371
Haussig S, Schuler G, Linke A. Treatment of a failing St. Jude Medical Trifecta by
Medtronic Corevalve Evolut valve-in-valve implantation. JACC Cardiovasc Interv.
375
2014;7(7):e81-e82. doi:10.1016/j.jcin.2013.11.025.
376
17.
SC
374
Dvir D, Leipsic J, Blanke P, et al. Coronary obstruction in transcatheter aortic valve-invalve implantation: preprocedural evaluation, device selection, protection, and treatment.
378
Circ Cardiovasc Interv. 2015;8(1):e002079-e002079.
379
doi:10.1161/CIRCINTERVENTIONS.114.002079.
380
18.
M AN U
377
Chakravarty T, Jilaihawi H, Nakamura M, et al. Pre-emptive positioning of a coronary stent in the left anterior descending artery for left main protection: a prerequisite for
382
transcatheter aortic valve-in-valve implantation for failing stentless bioprostheses?
383
Catheter Cardiovasc Interv. 2013;82(4):E630-E636. doi:10.1002/ccd.25037. 19.
EP
384
TE D
381
Urena M, Nombela-Franco L, Doyle D, et al. Transcatheter aortic valve implantation for the treatment of surgical valve dysfunction (“valve-in-valve”): assessing the risk of
386
coronary obstruction. J Card Surg. 2012;27(6):682-685. doi:10.1111/jocs.12003.
387
20.
AC C
385
Stock S, Scharfschwerdt M, Meyer-Saraei R, et al. Does Undersizing of Transcatheter
388
Aortic Valve Bioprostheses during Valve-in-valve Implantation Avoid Coronary
389
Obstruction? An In Vitro Study. The Thoracic and Cardiovascular Surgeon. 2016;64(S
390
01). doi:10.1055/s-0036-1571726.
18
ACCEPTED MANUSCRIPT
391
21.
Conradi L, Silaschi M, Seiffert M, et al. Transcatheter valve-in-valve therapy using 6 different devices in 4 anatomic positions: Clinical outcomes and technical considerations.
393
J Thorac Cardiovasc Surg. 2015;150(6):1557–65–1567.e1–3–discussion1565–7.
394
doi:10.1016/j.jtcvs.2015.08.065.
395
22.
RI PT
392
Castriota F, Cavazza C, Secco GG, Micari A, Cremonesi A. First Lotus aortic valve-invalve implantation to treat degenerated Mitroflow bioprostheses. EuroIntervention.
397
2016;11(13):1545-1548. doi:10.4244/EIJY15M07_01. 23.
Wolf A, Schmitz T, Latib A, Naber CK. Successful repositioning of a direct flow medical
M AN U
398
SC
396
399
25-mm valve due to acute occlusion of right coronary artery during transcatheter aortic
400
valve replacement procedure. JACC Cardiovasc Interv. 2015;8(2):e33-e34.
401
doi:10.1016/j.jcin.2013.11.030. 24.
Ruel M, Rubens FD, Masters RG, Pipe AL, Bédard P, Mesana TG. Late incidence and
TE D
402
predictors of persistent or recurrent heart failure in patients with mitral prosthetic valves. J
404
Thorac Cardiovasc Surg. 2004;128(2):278-283. doi:10.1016/j.jtcvs.2003.11.048.
405
25.
EP
403
Chan V, Rubens F, Boodhwani M, Mesana T, Ruel M. Determinants of persistent or recurrent congestive heart failure after contemporary surgical aortic valve replacement. J
407
Heart Valve Dis. 2014;23(6):665-670.
408 409
26.
AC C
406
Berdajs D. The anatomy of the aortic root. Cardiovascular Surgery. 2002;10(4):320-327. doi:10.1016/S0967-2109(02)00018-2.
410
19
ACCEPTED MANUSCRIPT
Figure legends
412
Figure 1:
413
Above: photograph of Perimount Magna Ease (left) and Trifecta (right). Below: Schematic
414
drawing. The pericardial leaflets (red) are mounted inside the stent frame in Perimount Magna
415
Ease (A) and outside in Trifecta(B).
RI PT
411
SC
416 Figure 2:
418
Photograph of an aortic root model. Two Dacron prostheses (8mm diameter), simulating the
419
coronary arteries, are anastomosed with an aortic sinus prosthesis (26mm diameter).
M AN U
417
420 Figure 3:
422
Schematic drawing of the coronary flow device. The left and right coronary artery (LC, RC)
423
running out of the aortic root model (4) pass a pressurized power chamber each (1). The pressure
424
is regulated by a pneumatic pump (3). The coronary flow is measured at the volume measuring
425
points (Q).
EP
426
TE D
421
Figure 4:
428
Left (above) and right (below) coronary flow curves, measured during an experimental run. For
429
the statistical analysis in this study, the coronary flow during diastole was used (area between A
430
and B).
AC C
427
431 432
Figure 5:
20
ACCEPTED MANUSCRIPT
Aortic root model with valve-in-valve implantation of a Sapien XT in a non-calcified (left) and
434
“calcified” (right) Trifecta. The red arrows indicate the remaining space between leaflet tissue
435
and coronary ostium. This space is obviously reduced in the “calcified” model. The black arrow
436
indicates the “calcification” of a Trifecta, simulated with glue (see Material and Methods
437
section).
RI PT
433
438 eFigure 1:
440
Flow chart of the test protocol. In each aortic root model, we implanted the different types of
441
bioprostheses. Afterwards, this was always followed by a transcatheter aortic valve-in-valve
442
implantation.
M AN U
SC
439
443 444 Video legend
446
The video shows the experimental procedure of our in vitro study. A Trifecta SAVB, size 25,
447
was implanted into the aortic root model with 8mm coronary ostia height and the conduit was
448
inserted into the mock circulation. Afterwards, the TAVI-ViV with a Sapien XT THV, size 23,
449
was performed and the conduit was inserted into the mock circulation likewise.
AC C
EP
TE D
445
21
ACCEPTED MANUSCRIPT
Table 1. Coronary flow (mean value ± standard deviation) TRI
TRI
PERI
PERI
RCF [ml/stroke]
CH 8mm
0.64 ± 0.06
0.60 ± 0.07
0.62 ± 0.08
0.61 ± 0.07
CH 10mm
0.62 ± 0.06
0.58 ± 0.04
SC
non-calcified
+ TAVI-ViV
RI PT
+ TAVI-ViV
0.60 ± 0.06
M AN U
“calcified”
0.62 ± 0.04
CH 8mm
0.67 ± 0.05
0.62 ± 0.04
0.65 ± 0.06
0.69 ± 0.17
CH 10mm
0.64 ± 0.08
0.58 ± 0.06
0.58 ± 0.11
0.58 ± 0.03
0.90 ± 0.08
0.87 ± 0.06
0.89 ± 0.06
0.89 ± 0.09
LCF [ml/stroke] non-calcified
CH 10mm
CH 8mm CH 10mm
0.89 ± 0.05
0.82 ± 0.07
0.87 ± 0.04
0.83 ± 0.03
0.85 ± 0.06
0.79 ± 0.08
0.83 ± 0.04
0.77 ± 0.12
0.76 ± 0.04
0.76 ± 0.13
0.80 ± 0.04
EP
“calcified“
TE D
CH 8mm
0.79 ± 0.11
AC C
TRI: Trifecta™, TAVI-ViV: Transcatheter Aortic Valve-in-Valve Implantation, PERI: Perimount® Magna Ease, CH: Coronary height, RCF: right coronary flow, LCF: left coronary flow
ACCEPTED MANUSCRIPT Table 2. Hydrodynamics (mean value ± standard deviation) TRI
TRI
PERI
PERI
+ TAVI-ViV
+ TAVI-ViV
Geometric orifice area
“calcified“ [cm2] 0.94 ± 0.32
1.43 ± 0.12
2.08 ± 0.03
1.3 ± 0.14
0.77 ± 0.36
1.56 ± 0.17
RI PT
non-calcified [cm2] 2.48 ± 0.05
Pressure gradients
SC
non-calcified
1.42 ± 0.17
12 ± 2
13 ± 2
13 ± 2
14 ± 2
dPmean [mmHg]
4±1
5±1
5±1
6±1
dPmax [mmHg]
23 ± 7
16 ± 1
34 ± 20
15 ± 2
dPmean [mmHg]
12 ± 5
6±1
20 ± 13
6±1
“calcified“
M AN U
dPmax [mmHg]
TRI: Trifecta™, TAVI-ViV: Transcatheter Aortic Valve-in-Valve Implantation, PERI:
AC C
EP
TE D
Perimount® Magna Ease, dP: pressure gradient
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT