Rheolytic effects of left main mid-shaft/distal stenting: a computational flow dynamic analysis

Abstract

Background

The aim of this study was to evaluate the rheolytic effects of stenting a mid-shaft/distal left main coronary artery (LMCA) lesion with and without ostial coverage. Stenting of the LMCA has emerged as a valid alternative in place of traditional coronary bypass graft surgery. However, in case of mid-shaft/distal lesion, there is no consensus regarding the extension of the strut coverage up to the ostium or to stent only the culprit lesion.

Methods

We reconstructed a left main-left descending coronary artery (LM-LCA)-left circumflex (LCX) bifurcation after analysing 100 consecutive patients (mean age 71.4 ± 9.3, 49 males) with LM mid-shaft/distal disease. The mean diameter of proximal LM, left anterior descending (LAD) and LCX, evaluated with quantitative coronary angiography (QCA) was 4.62 ± 0.86 mm, 3.31 ± 0.92 mm, and 2.74 ± 0.93 mm, respectively. For the stent simulation, a third-generation, everolimus-eluting stent was virtually reconstructed.

Results

After virtual stenting, the net area averaged wall shear stress (WSS) of the model and the WSS at the LCA-LCX bifurcation resulted higher when the stent covered the culprit mid-shaft lesion only compared with the extension of the stent covering the ostium (3.68 versus 2.06 Pa, p = 0.01 and 3.97 versus 1.98 Pa, p < 0.001, respectively. Similarly, the static pressure and the Reynolds number were significantly higher after stent implantation covering up the ostium. At the ostium, the flow resulted more laminar when stenting only the mid-shaft lesion than including the ostium.

Conclusions

Although these findings cannot be translated directly into real practice our brief study suggests that stenting lesion 1:1 or extending the stent to cover the LM ostium impacts differently the rheolytic properties of LMCA bifurcation with potential insights for restenosis or thrombosis.

Introduction

According to the current international guidelines,^1,2 stenting of the left main coronary artery (LMCA) has emerged as a valid interventional approach instead of traditional coronary bypass surgery.^3–5 Stenting of the LMCA requires a different strategy depending on the location of the lesion: ostium, mid-shaft, distal and bifurcation. While percutaneous treatment of left main bifurcation disease is still problematic, endovascular stenting of ostial and mid-shaft disease is more widely accepted, despite the presence of some technical dark zones. In particular, when mid-shaft lesions or distal lesions are to be treated, to extend the strut coverage up to the ostium, or to stent the lesion only is a matter of operator choice and no scientific data regarding the optimal technique have yet been published.

Intravascular ultrasound (IVUS) usually should help to elucidate eventually unapparent ostial disease and the need to extend the stenting to the ostium, as recommended by the current standard⁶ when facing with any left main (LM) percutaneous coronary intervention (PCI), but in the real-world practice is still matter of operator view and preference. The impact of the two different approaches on both ostium and LAD-left circumflex (LCX) coronary artery bifurcation physiology has never been investigated. Our study is aimed to evaluate, using computational fluid dynamic (CFD) analysis, the impact of stenting a mid-shaft LMCA on both ostial and bifurcation physiology.

Material and methods

Building up the virtual model

For the computational domain analysis, we considered a LM-LCA-LCX bifurcation reconstructed after retrospective analysis of the angiographic records of 100 consecutive patients (mean age 71.4 ± 9.3, 49 males) with LM mid-shaft or distal disease who underwent PCI between 1 January 2016 and 1 January 2017.

The Ethical and Research committee of the Rovigo General Hospital approved the study and waived the need for informed consent. The diameters of the vessels and the ascending aorta are routinely measured before any PCI per institutional protocol and the obtained numerical values were automatically withdrawn from the Cath lab software in a completely retrospective and anonymous form and used only to build up a unique coronary and ascending aorta virtual model: no other clinical of anagraphic data excepted age and gender were used.

The mean diameter of proximal LM, LAD and LCX, evaluated with quantitative coronary angiography (QCA) was 4.62 ± 0.86 mm, 3.31 ± 0.92 mm, and 2.74 ± 0.93 mm, respectively. The mean LAD-LCX bifurcation angle, measured after the diagnostic angiography using an electronic goniometer was 53.6 ± 12.3° whereas the mean length of LM was 19.4 ± 4.1 mm. Mean diameter stenosis by QCA was 82 ± 7.6%. Diameters of LCA and LCX were modelled according to the Finnet’s law⁷ as follows: LM 4.5 mm, LAD 3.5 mm, LCX 2.5 mm, with bifurcation angle set up at 55°. The LM was divided in three sections: ostium, mid-shaft and distal. A plaque inducing a stenosis of 90% was placed at the mid-shaft position (Figure 1). For evaluating the impact of blood flow on ostial physiology the same model including the ascending aorta was created: height of the aortic take off (distance between the origin of the LMCA and the aortic valve plane) was arbitrarly set at 10 mm, while the diameter of ascending aorta just above the left coronary origin was set at 30 mm. The model was constructed using Rhinoceros version 4.0 Evaluation (McNeel and Associates, Indianapolis, IN, USA). Pressure was assumed to be stable at 120/80mmHg.

Figure 1.

Schematic representation of the model created for the analysis of fluid dynamic. A virtual left main has been created on the basis of a real population and divided in three sectors from distal to proximal including bifurcation, mid-shaft and ostium. Vessel diameters followed the Finnet’s Law.

LAD, left anterior descending; LCX, left circumflex; LM, left main.

Stent geometry reconstruction

For the stent simulation, we reconstructed the strut design and linkage pattern of a third-generation, everolimus-eluting stent (ORSIRO stent, Biotronik IC, Bulack, Switzerland), commonly used in our institution. The strut thickness of this stent is characterized by a very ultrathin strut (60 μm up to 3.0 mm diameter stent and 80 μm up to 4.0 mm stent). Computer Aided Design software was used to reproduce the stented geometry as accurately as possible (SolidWorks 2009, Solidworks Corp, Concord, MA, USA). In the first step, we created the solid model of the coronary artery bifurcation and then the expanded stent geometry. For this purpose, a hollow tube with outer diameter equal with both the nominal expanded diameter and thickness of the stent was created. Then, a two-dimensional sketch with the stent’s strut was propagated and wrapped around the tube. Through a cut out, the obtained ring of the stent was propagated axially to create the full-length expanded model.

Virtual implantation

After placing the stent model in the correct position, the stenting procedure was performed following the real procedural steps.

Lesion only

• - Predilation with noncompliant Euphora (Medtronic Inc, USA) balloon 3.0 × 12 mm at 16 atm;

• - Stent implantation: Orsiro 4.0 × 12 mm at 18 atm;

• - Over-dilation with 4.5 × 12 mm noncompliant Euphora (Medtronic Inc, USA) balloon at 20 atm.

Up to the ostium

• - Predilation with noncompliant Euphora (Medtronic Inc, USA) balloon 3.0 × 12 mm at 16 atm;

• - Stent implantation: Orsiro 4.0 × 15 mm at 18 atm;

• - Over-dilation with 4.5 × 15 mm noncompliant Euphora (Medtronic Inc, USA) balloon at 20 atm;

Using Boolean operation, the modified solid model was subtracted from the bifurcation model to obtain the final geometry. We assumed that after stent deployment and implantation there was no residual stenosis.

Computational fluid dynamic analysis

Blood was modelled as a nonNewtonian, viscous and incompressible fluid. Density was defined as 1060 kg/m3, according to the standard values cited in the literature.^8–11 Blood was represented by the Navier–Stokes equation

$$\rho \mathrm{v}.\nabla \mathrm{v}=-\nabla \le .\mathrm{T}-\nabla P$$ (1)

and continuity equation

$$\nabla \cdot \mathrm{v}=0$$ (2)

where v is the three-dimensional velocity vector, P pressure, ρ density and τ the shear stress term. Instead, the Carreau model was applied for the viscosity of blood. Considering that coronary perfusion is mainly diastolic, we performed a steady flow simulation using a basal diastolic pressure of 80mmHg (10,665 Pa). In the analysis. static pressure (Pa), wall shear stress (WSS) (Pa) and Reynolds number were evaluated at both the lesion site and at the LAD-LCX bifurcation.

Static pressure in the vessel has been evaluated in Pascal. From a pathophysiological point of view, low static pressure is generally related to increased vessel wall thickness.

WSS has been defined as the force which is tangentially acting to the surface due to friction. As well known, low WSS are related to the development of greater plaque, higher neo-endothelization and necrotic core progression with a constrictive remodeling whereas high WSS segment develop greater necrotic core and calcium progression with expansive remodeling.^12,13 The numeric grid was obtained with ANSYS Meshing 14.0 (Ansys, Inc., Canonsburg, PA, USA) while the simulations were conducted using the commercial software ANSYS FLUENT 14.0 (Ansys, Inc).

Statistical analysis

Continuous variables were expressed as mean and standard deviation and compared with Student’s t test while categorical variables were presented as percentages and compared using Chi-square and Fisher’s exact test, as appropriate. All statistical analyses were carried out using SPSS statistical software version 19.0 (SPSS Inc., Chicago, IL, USA). A p value <0.05 was considered statistically significant.

Results

After virtual stenting, all the considered parameters increased significantly compared with the baseline value, reflecting the heavy changes on the blood flow physiology in LMCA critical occlusion. As shown in Table 1, the net area averaged WSS of the model and the WSS at the LCA-LCX bifurcation was higher when the stent covered the culprit lesion only compared with the extension of the stent until the ostium (3.68 versus 2.06 Pa, p = 0.01 and 3.97 versus 1.98 Pa, p < 0.001, respectively (Figure 2). Similarly, the static pressure and the Reynolds number were significantly higher after stent implantation covering up the ostium. Intriguingly, that trend was maintained at the lesion site and globally throughout the model. At ostial level the flow resulted clearly more laminar in stenting the lesion 1:1 than in the ostial coverage (Figure 3).

Table 1.

Computed fluid dynamic measurements of the considered parameters. Baseline refers to the model before stent implantation.

	Baseline	Lesion only	Ostial	p
WSS
Mean WSS at the lesion [Pa]	6.21	3.41	2.10	0.01
Mean WSS at bifurcation [Pa]	6.58	3.97	1.98	<0.01
Mean area avereged WSS of the model [Pa]	6.33	3.68	2.06	0.01
Static pressure
Static pressure at the lesion [Pa]	1058.25	10,679.35	10,737.24	0.01
Static pressure at bifurcation [Pa]	1047.91	10,683.95	10,795.50	0.01
Mean area averaged static pressure of the model [Pa]	1051.20	10,683.93	10,785.52	0.01
Reynolds number
Reynolds number at the lesion	3.26	6.157	2.974	<0.001
Reynold number at the bifurcation	3.38	6.210	10.624	0.001
Mean area averaged Reynolds number of the model	3.32	6.210	10.621	0.01

WSS, wall shear stress.

Figure 2.

Computed fluid dynamic representation of the left main bifurcation after stenting the lesion 1:1 (left, upper cartoon) or extending the stent to the ostium (left, lower cartoon). The wall shear stress forces represented at carena resulted lower (blue-green coloured areas) when stenting is extended to cover the ostium (a) than when the lesion itself is covered 1:1 (red-yellow coloured areas, b).

LAD, left anterior descending; LCX, left circumflex; LM, left main.

Figure 3.

Aortic-left main ostium model reconstruction (a) for evaluation of fluid dynamic at the ostium: stenting the lesion 1:1 (a) and stenting up to the ostium (b). The same model is observed into the two different perspectives to better visualize the rheolytic properties of the blood. The flow is much more turbolent when the stent reaches the ostium than when the stent cover only the lesion at mid-shaft.

LM, left main.

Discussion

Our study suggests that when facing with mid-shaft/distal LMCA lesion, the implantation of a stent covering the lesion 1:1 produce a more favourable fluid dynamic profile at both aortic ostial level and bifurcation levels than to extend the stent up to the ostium. Nowadays there is a consistent trend among interventionalists around the world to promote extension of the stent up to the ostium when treating a LMCA mid-shaft or distal disease, even in the absence of angiographic or IVUS demonstration of disease at the ostium.

The reasons for this approach are different. First of all, to protect the LMCA ostium from intraprocedural injuries with potential long-term disease induced by the guiding catheter itself other materials might justify such approach.¹⁴ Secondly, to extent the stent length to covering up the LMCA ostium might be sometimes easier and faster than precisely tailoring the stent to the lesion especially in urgency/emergency settings.

As explained by Wentzel and colleagues,¹⁵ atherosclerotic plaques predominantly form in regions of low endothelial shear stress (ESS), whereas regions of moderate/physiological and high ESS are generally protected. Low ESS-induced compensatory expansive remodeling, plays an important role in the preservation of lumen dimensions during plaque progression. However, when the expansive remodeling becomes excessive this promotes a continued influx of lipids into the vessel wall, vulnerable plaque formation and potential precipitation of an acute coronary syndrome.

Covering the ostium may cause the WSS forces to lower at the bifurcation level and might predispose plaque formation or plaque increase as also suggested by Krams and colleagues¹⁶ and other authors.^17,18 This paradoxical effect could be explained by the turbulent flow induced by the ostial portion of the stent which may alter the physiological fluid dynamic throughout the whole stent up to the left stem bifurcation carina. Indeed, during diastole, the flow reverses, flows back from above to the aortic root and enters the coronary ostium. If the height of the LM coronary ostium is high (>10 mm) then the diastolic flow will enter from above and hit the inferior wall of the LM (high WSS) and leave the roof of the LM out, causing turbulent vortices and low WSS in that site. As result, the lesion will form in the roof of the LM as evidenced by Figure 3(b). Furthermore, during diastole, the flow reverses, flows back from above to the aortic root and enters the coronary ostium. All these effects can be increased and exacerbated by an increment of the heart rate which naturally increases the flow turbulence potentially disturbing the laminar flow .

If the height of the LM coronary ostium is normal (10 mm) then the diastolic flow will enter from below because the diastolic flow hits the aortic root and enters the coronary ostium and hit the roof of the LM (high WSS) and leave the base of the LM out, causing low WSS in that portion of the vessel. The final result, will be the formation of a plaque in the base of the LM, as evidenced by Figure 3(c). While the goal of LMCA mid-shaft lesion treatment by stenting is to keep away to the bifurcation and its complicated treatment, to extend the stent covering the ostium might equalize the advantage of a mid-shaft position of the plaque predisposing to plaque formation at the bifurcation carina and potentially impacting the restenosis rate.¹⁹ To completely translate CFD in clinical practice represents a challenge: calculation of the WSS in vivo actually is possible by angio-computed tomography (CT) and optical coherence tomography (OCT) but is time consuming and might offer an evaluation in a single patients before and after an indexed procedure without any chance to predict different results in terms of WSS if a different technique is applied. In the very near future, integrated OCT/IVUS or CT with specific simulation software might offer a real chance to improve procedure quality by adding prediction of WSS by CFD.

Limitations

Our study considers a virtual left main model. The artery has been considered noncompliant, straight and with a steady diastolic blood flow in a virtual hemodynamically stable patient. However, coronary artery perfusion is mainly diastolic and previous studies have already demonstrated that myocardial motion has a negligible effect on blood flow distribution on the coronary tree. Our model considered an optimal stent deployment without residual stenosis despite in daily clinical practice, the different angles, the amount and circumferential extent of the calcium, the length of the respective lesion, and many other parameters have an obvious impact on the implantation technique and outcomes. Other limitations of the study are that we did not evaluate the time averaged WSS, oscillatory index, and the relative residence time, which had a recognized role in the development of coronary artery stenosis.

Conclusion

Although these findings cannot be translated directly into real practice, since their clinical value is difficult to assess in the real world, our brief study suggests that stenting lesion 1:1 or extending the stent to cover the LM ostium in case of mid-shaft/distal disease, impacts differently the rheolytic properties of LMCA bifurcation with potential insights for restenosis or thrombosis.