Author + information
- Received June 18, 2013
- Revision received August 21, 2013
- Accepted September 26, 2013
- Published online March 1, 2014.
- Peter Mortier, PhD∗,†∗ (, )
- Yutaka Hikichi, MD‡,
- Nicolas Foin, PhD§,
- Gianluca De Santis, PhD∗,†,
- Patrick Segers, PhD†,
- Benedict Verhegghe, PhD∗,† and
- Matthieu De Beule, PhD∗,†
- ∗FEops, Ghent, Belgium
- †IBiTech-bioMMeda, Ghent University, Ghent, Belgium
- ‡Department of Cardiovascular Medicine, Saga University, Saga, Japan
- §International Centre for Circulatory Health, Imperial College London, London, United Kingdom
- ↵∗Reprint requests and correspondence:
Dr. Peter Mortier, FEops, IIC UGhent, Technologiepark 3, 9052 Ghent, Belgium.
Objectives This study sought to better understand and optimize provisional main vessel stenting with final kissing balloon dilation (FKBD).
Background Main vessel stenting with FKBD is widely used, but many technical variations are possible that may affect the final result. Furthermore, most contemporary stent designs have a large cell size, making the impact of stent platform selection for this procedure unclear.
Methods Finite element simulations were used to virtually deploy and post-dilate 3 stent platforms in 3 bifurcation models. Two FKBD strategies were evaluated: simultaneous FKBD (n = 27) and modified FKBD (n = 27). In the simultaneous FKDB technique, both balloons were simultaneously inflated and deflated. In the modified FKBD technique, the side branch balloon was inflated first, then partially deflated, followed by main branch balloon inflation.
Results Modified FKBD results in less ostial stenosis compared with simultaneous FKBD (15 ± 9% vs. 20 ± 11%; p < 0.001) and also reduces elliptical stent deformation (ellipticity index, 1.17 ± 0.05 vs. 1.36 ± 0.06; p < 0.001). The number of malapposed stent struts was not influenced by the FKBD technique (modified FKBD, 6.3 ± 3.6%; simultaneous FKBD, 6.4 ± 3.4%; p = 0.212). Stent design had no significant impact on the remaining ostial stenosis (Integrity [Medtronic, Inc., Minneapolis, Minnesota], 16 ± 11%; Omega [Boston Scientific, Natick, Massachusetts], 17 ± 11%; Multi-Link 8 [Abbott Vascular, Santa Clara, California], 19 ± 8%).
Conclusions The modified FKBD procedure reduces elliptical stent deformation and optimizes side branch access.
The provisional side branch stenting strategy is currently the gold standard for treating most coronary bifurcation lesions because the systematic use of more complex 2-stent techniques does not improve clinical outcomes (1,2). This approach involves stenting of the main branch and, if necessary, stenting of the side branch.
Even though final kissing balloon dilation (FKBD) is considered mandatory for which both branches are stented (3), the clinical benefit of routine FKBD after main vessel stenting has not yet been proven (4–6). FKBD after main vessel stenting is, however, widely used and is recommended for side branches with >75% ostial stenosis (7–9). Attempts to better understand and to further optimize the technical aspects of this technique are therefore justified.
In this study, computational modeling was used to compare 2 different FKBD strategies and to evaluate the impact of stent design when stenting the main vessel only.
Virtual bench testing
Finite element computer simulations were used to virtually deploy and post-dilate stents in 3 different stenosed bifurcation models, mimicking a range of coronary bifurcation anatomies (Fig. 1). The outer wall diameters were 1.6 times larger than the nondiseased inner diameters because this results in a realistic wall thickness according to the anatomic data reported by Holzapfel et al. (10). The layered structure of the arterial wall was taken into account, and different isotropic hyperelastic material properties were assigned to the different tissue layers (intima, media, and adventitia) and to the plaques (10,11). All simulations were performed using the Abaqus/Explicit finite element solver (Dassault Systèmes, Velizy, France). Details of this method to study bifurcation stenting have been previously described (12–14).
Three 3.0-mm stents with a length of 18 or 20 mm were tested: the Integrity stent (Medtronic, Minneapolis, Minnesota), the Omega stent (Boston Scientific, Natick, Massachusetts), and the Multi-Link 8 stent (Abbott Vascular, Santa Clara, California). Virtual stent models were generated by reverse-engineering actual stent samples using high-resolution micro-computed tomography (CT) imaging (12,15). Material properties for the different metallic alloys were taken from O'Brien et al. (16). All virtual balloon models were validated against their respective compliance charts.
Each stent was deployed 3 times in each bifurcation model to obtain different stent positions with respect to the side branch ostium, leading to 27 stent deployments. The stents were sized according to the distal branch diameter and deployed following the manufacturer's compliance chart to reach a diameter of 3.2 mm.
The proximal stent segment was post-dilated using a short noncompliant Sprinter balloon (Medtronic) sized according to the proximal branch diameter. Such post-dilation is known as the proximal optimization technique (POT) (9).
Two different FKBD strategies were tested in each of the 27 deformed configurations obtained after the POT: simultaneous FKBD and modified FKBD (Fig. 2). In simultaneous FKBD, both the side and main branch balloons were simultaneously inflated and deflated, with a maximal balloon pressure of 12 atm. In modified FKBD (which is based on a suggestion by Dr. Hikichi), the side branch balloon was inflated to a pressure of 12 atm and then deflated to 4 atm. Subsequently, the main branch balloon was inflated to a pressure of 12 atm. Eventually, both balloons were fully deflated. For both strategies, a noncompliant Sprinter and a semicompliant Sprinter Legend balloon were used for the main and side branch, respectively. Both balloons were sized according to the distal branch diameters. Side branch recrossing was performed through a mid-distal cell.
In vitro bench testing
The accuracy of the finite element simulations was verified by comparing the virtually-predicted stent deformations with in vitro observed stent deformations. Flexible polymer replicas of bifurcation model A (Fig. 1) were manufactured using rapid prototyping techniques. One sample of each stent was deployed and post-dilated using the simultaneous FKDB strategy, as described earlier. The final stent deformations were visualized using micro-CT (HMX-ST micro-CT, X-Tek Systems Ltd., Tring, Hertfordshire, United Kingdom) at a voxel resolution of 23.7 or 28.3 μm/voxel. The scanned volumes were exported in a DICOM format file, and the stents were reconstructed by segmentation using 3DSlicer (Massachusetts Institute of Technology, Cambridge, Massachusetts).
The following parameters were quantified for all simulated cases after both simultaneous and modified FKBD (n = 54): ostial area stenosis, strut malapposition, and ellipticity of the proximal stent segment.
Ostial area stenosis was quantified in a planar projection perpendicular to the side branch axis following the definition proposed by Ormiston et al. (17): (A1 − A2)/A1 × 100%, where A1 is the total area of the side branch ostium and A2 is the largest area free of struts.
The stent strut apposition was evaluated by measuring the distance from the centerline of the outer stent surface to the inner surface of the arterial wall. For the Integrity stent, which has circular struts, the initial outermost point of the struts was used for the distance measurements. Strut malapposition was automatically quantified by calculating the percentage of strut length for which the distance to the luminal surface of the bifurcation model exceeded 100 μm.
The elliptical stent deformation in the middle of the proximal stent segment was analyzed by manually fitting an ellipse on the deformed stent cross section. An ellipticity index was derived, defined as the ratio of the maximal-to-minimal ellipse diameter.
The impact of the stent type (Integrity, Omega, or Multi-Link 8) and FKBD procedure (simultaneous or modified) was first studied by means of analysis of variance using IBM SPSS statistical software version 20 (IBM, Armonk, New York). A general linear model was constructed using stent type as an independent factor. The FKBD procedure was included as a repeated-measures factor to account for the fact that both simulations started from the same initially deployed stent configuration. Statistically tested dependent factors were ostial area stenosis, strut malapposition, and the ellipticity index. Interaction between factors was included in the model. If the general linear model test indicated an overall significant effect (assuming a threshold value of p = 0.05), post-hoc testing was performed for pairwise comparison using the Bonferroni correction.
Virtual versus in vitro bench testing
Bifurcation stenting techniques are traditionally investigated and optimized using in vitro bench tests. Stents are deployed in mock artery models, and the resulting deformations are visualized using, for example, micro-CT. In this study, finite element computer simulations were the primary research tool, and only a few in vitro bench tests were performed to evaluate the accuracy of the virtually-predicted stent deformations. Overall, good qualitative agreement was obtained between simulated and in vitro observed deformations, as shown in Figure 3, and differences were mainly due to the variability of the FKBD procedure. For example, stents may have a different position with respect to the side branch ostium, and the balloons during FKBD may have different relative positions.
Impact of the FKBD strategy
Simultaneous FKBD at 12 atm opens the stent cell at the side branch ostium, but a remaining ostial area stenosis of 20 ± 11% was observed. This FKBD strategy also results in elliptical deformation of the stent in the proximal main vessel (ellipticity index, 1.36 ± 0.06). The modified FKBD technique results in a lower remaining ostial area stenosis (15 ± 9%; p < 0.001) and also reduces elliptical stent deformation (ellipticity index, 1.17 ± 0.05; p < 0.001), as illustrated in Figure 4. The number of malapposed stent struts was similar with both FKBD strategies (simultaneous FKBD, 6.4 ± 3.4%; modified FKBD, 6.3 ± 3.6%; p = 0.212). Malapposed stent struts were observed at different locations: at the side branch ostium; in the main vessel near the side branch ostium; at the shoulders of the plaque; and in the proximal stent segment, where struts are present that are not in contact with any of the 2 overlapping balloons (Fig. 4). All results are summarized in Table 1.
Impact of stent design
The impact of stent design was evaluated by combining the results of the simultaneous and modified FKBD simulations (18 for each stent). The conclusions regarding the impact of the stent design remain unaltered when only considering the simultaneous or modified FKBD group. Ostial area stenosis after FKBD was 16 ± 11%, 17 ± 11%, and 19 ± 8% for the Integrity, Omega, and Multi-Link 8 stents, respectively, with no statistically significant differences between the different platforms. Large variations in the remaining ostial area stenosis were observed for each stent. The range of measured ostial area stenosis was 2% to 39% for the Integrity stent, 4% to 43% for the Omega stent, and 6% to 31% for the Multi-Link 8 stent. Figure 5 shows an example of a good and a suboptimal opening of the stent cell toward the side branch for each stent design.
The number of malapposed stent struts was significantly lower for the Integrity (3.8 ± 1.2%) compared with the other 2 designs (Omega, 8.5 ± 4.4%; Multi-Link 8, 6.7 ± 2.2%; Integrity – Omega, p < 0.001; Integrity – Multi-Link 8, p < 0.001). One representative case for each design is depicted in Figure 6, highlighting the locations of the malapposed stent struts.
FKBD is commonly performed after main vessel stenting to improve side branch access while preventing stent distortion within the main vessel. It also helps to optimize stent expansion in the proximal segment when the POT is not used. There are, however, many different ways to perform FKBD, for example, balloons can be inflated and deflated simultaneously or sequentially or equal or unequal pressures can be used, but the impact of these technical variations is not fully understood. In this study, we used computer simulations to compare 2 different FKBD strategies: simultaneous and modified FKBD. The obtained results show that modified FKBD after the POT leads to lower ostial stenosis and reduces elliptical deformation in the proximal main vessel, but does not alter the number of malapposed struts. It remains unknown whether modified FKBD would still be beneficial if the POT is not used.
Less ostial area stenosis after modified FKBD is the result of first inflating the side branch balloon, which maximizes the opening of the cell toward the side branch. In simultaneous FKBD, opening of this stent cell is hindered by the simultaneously inflated main branch balloon. Ormiston et al. (17) reported a similar observation when studying the crush stenting technique using in vitro bench testing.
Reduced elliptical deformation after using the modified FKBD technique is a direct consequence of lowering the pressure in the side branch balloon before inflating the main branch balloon. Using a lower pressure has no negative impact on the correction of the main vessel stent distortion, as shown by the number of malapposed stent struts, which is nearly identical. The similar amount of strut malapposition also indicates that elliptical stent deformation is not correlated with strut malapposition. It should be noted, however, that the absence of such correlation could be related to other factors, such as a difference in stent area.
Reduced elliptical stent deformations after using the modified FKBD technique might be important, because such deformations have been associated with a greater amount of thrombus in an optical coherence tomography study (18). Alternatively, a final proximal post-dilation (19) has been proposed to reduce elliptical stent deformation after FKBD. Although effective, this approach has the limitation of adding a step to the procedure.
A remaining obstruction of the side branch after FKBD cannot be avoided, and the magnitude of this obstruction is unpredictable. Similar findings have been reported based on in vitro and in vivo optical coherence tomography studies (20). For all investigated stents, we have observed cases with an adequate opening of the cell toward the side branch, but also cases in which the final stent deformation was suboptimal. This is the result of the large variation that occurs when performing FKBD: stents may have different longitudinal and rotational positions with respect to the side branch ostium (21), and the point of recrossing can vary. There were no differences among the individual stent platforms with respect to the side branch obstruction, which is slightly surprising considering the fact that the designs have a different number of connectors between subsequent rings. The Omega platform has, for example, 2 connectors, whereas the Multi-Link 8 design has 3 connectors. Furthermore, the (theoretically) maximal cell diameter of these stents is different (Integrity, 4.7 mm; Omega, 5.3 mm; and Multi-Link 8, 4.2 mm) when evaluated by the method described by Mortier et al. (22).
Interestingly, Foin et al. (20) reported a rate of ostial stenosis after FKBD of 21% on the basis of micro-CT– visualized bench tests, which is very similar to the results obtained by our computer simulations. In addition, they reported a maximum ostial stenosis of 46%, which corresponds well to the maximum observed in this study, which again confirms the accuracy of the virtually-predicted stent deformations.
The 3-dimensional visualization of stent strut malapposition, as shown in Figure 6, helps to better explain this unwanted phenomenon. It clearly shows that malapposed struts may be located not only near the side branch ostium, but also at different locations within the main branch. First, malapposition occurs within the main branch near the shoulders of the plaque due to the local vessel wall curvature. Second, struts between the 2 overlapping balloons, as indicated in Figure 4, are sometimes malapposed because of the lack of direct outward pressure on these struts. Finally, malapposition occurs within the main vessel near the side branch ostium because of complex stent deformations in that region.
The number of malapposed struts depends on stent design, which may be important because some clinical studies have suggested that incomplete stent apposition leads to increased stent thrombosis (23,24). The reasons for the observed differences in strut apposition are numerous. A very important stent design characteristic for bifurcation stenting is the maximal expansion capacity of a stent platform. Deployment of a single stent in a main vessel of a large bifurcation typically requires a considerable overexpansion of the proximal stent segment to obtain adequate strut apposition, certainly when the distal branch diameter is used for stent sizing. This is 1 of the recommendations of the European Bifurcation Club (9) and was followed in this study. Each of the 3 3.0-mm stent designs included in this study has a very different maximal expansion capacity, as shown by a recent bench study (25). The 3.0-mm Integrity stent is the same stent as the 4.0-mm size and can be post-dilated to a minimal diameter of 5.4 mm (with a 6.0-mm balloon), whereas the 2 other designs only have the middle stent size and not the larger stent design available. This means that the Omega and the Multi-Link 8 stents were close to their maximal diameter in our tests, especially in bifurcation model B. Consequently, the rings of these stent designs are almost fully stretched (i.e., limited zigzag shape), particularly for the Multi-Link 8 stent, which may reduce the ability of the rings to conform to the local vessel wall curvature and thus lead to a greater number of malapposed struts in the proximal main vessel. This also implies that different results could be obtained when a different stent size (with a different number of crowns and thus a different expansion capacity) was tested. Another aspect influencing strut malapposition in the proximal main vessel is that overexpansion also increases the bumpiness of the outer stent surface. This phenomenon was observed with all stents, but was most pronounced for the laser-cut Omega stent, probably because of its varying rectangular strut cross section (larger strut width at the peaks compared with the width of the straight struts).
All results are based on numerical models, which inherently contain a number of assumptions and approximations and may therefore not correctly represent in vivo stent behavior. For example, simplified bifurcation models were used, which may not fully reflect the large variations in bifurcation anatomies and lesions (e.g., plaque composition and distribution) that are encountered in clinical practice. In addition, stent models were scanned in their crimped configuration, and the residual stresses in the stents introduced by the crimping process are therefore not taken into account. Although the stent models were generated based on high-resolution micro-CT imaging, small differences may exist between these models and the actual stent designs. Furthermore, simplified mechanical properties are used to model the different alloys and the vessel tissue in the computer simulations.
FKBD introduces an elliptical deformation in the proximal main vessel while a remaining unpredictable amount of side branch obstruction cannot be avoided. A modified FKBD procedure is proposed that reduces this elliptical deformation of the stent and optimizes the side branch access.
Medtronic provided financial support and all stent samples for this study. Dr. De Santis is an employee of FEops Drs. Mortier, De Beule, and Verhegghe are shareholders in FEops, an engineering consultancy spinoff of Ghent University; and have served as consultants for Medtronic, Boston Scientific, and Terumo. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- computed tomography
- final kissing balloon dilation
- proximal optimization technique
- Received June 18, 2013.
- Revision received August 21, 2013.
- Accepted September 26, 2013.
- American College of Cardiology Foundation
- Hildick-Smith D.,
- de Belder A.J.,
- Cooter N.,
- et al.
- Steigen T.K.,
- Maeng M.,
- Wiseth R.,
- et al.
- Ge L.,
- Airoldi F.,
- Iakovou I.,
- et al.
- Niemela M.,
- Kervinen K.,
- Erglis A.,
- et al.
- Gwon H.C.,
- Hahn J.Y.,
- Koo B.K.,
- et al.
- Hildick-Smith D.
- Koo B.K.,
- Park K.W.,
- Kang H.J.,
- et al.
- Holzapfel G.A.,
- Sommer G.,
- Gasser C.T.,
- Regitnig P.
- Mortier P.
- Ormiston J.A.,
- Webster M.W.,
- Webber B.,
- Stewart J.,
- Ruygrok P.,
- Hatrick R.
- Otake H.,
- Shite J.,
- Ako J.,
- et al.
- Foin N.,
- Torii R.,
- Mortier P.,
- et al.
- Cook S.,
- Wenaweser P.,
- Togni M.,
- et al.
- Ozaki Y.,
- Okumura M.,
- Ismail T.F.,
- et al.