Passive Aortic Counterpulsation to Reduce Pressure Pulse During Aortic Prosthesis Insertion and Reduce Endoleaks Formation: A Preliminary In Silico Investigation

Abstract

Over 10% of patients undergoing aortic endograft implantation experience endoleaks within a few years. In the case of type 1a endoleaks, a crack forms between the aorta and the prosthesis collar, allowing blood to pass. This blood fills the aneurysmal sac and can lead to its rupture. None of the strategies, such as prostheses with barbs and hooks or ad hoc pharmacological therapies, can prevent the phenomenon. An alternative approach is to reduce diameter oscillations due to pulsating pressure to improve the endoprosthesis adhesion to the internal vessel walls during the initial post-implantation phases. To reach this objective, we propose to use a passive intra-aortic balloon pump (PIABP) inserted and then maintained inside the vessel immediately after the surgical procedure. We tested our hypothesis in a mechanical mock of the cardiovascular system. A silicon aorta with physiological behavior was created for this purpose. The PIABP was inflated to increasing pressures between systolic and diastolic values (120/80 mmHg). For each aorta and each condition, the variations in aortic diameter between systole and diastole, and the pressure variations, were measured. For the normal aorta, with a PIABP pressure of 110 mmHg, the variations in diameter were reduced by 38%. Assuming an endoprosthesis with a diameter of 30 mm (oversized by 5% compared to the diastolic diameter), the time the oscillations are higher than 30 mm is also reduced by 36%. The results are positive and suggest the usefulness of a biomechanical approach to the problem of type 1a endoleaks. Further in silico and clinical trials are necessary to validate the method.

1. Introduction

Vascular endoprostheses are artificial implantable devices inserted into vascular chambers to correct situations of possible danger, such as aortic dissection and arterial aneurysm rupture. These are made of inextensible plastic materials like Dacron and Teflon, sometimes supported by metallic stents to maintain an optimal shape even in highly irregular vessels. A key issue with these prostheses is the occurrence of endoleaks resulting from imperfect connections between the vascular lumen and the extra prosthetic chamber, leading to inefficiency in therapeutic interventions in about 10% of cases [1,2].

Type I endoleaks consist of an incorrect adhesion of the proximal (type 1a) or distal (type 1b) neck of the endoprosthesis. The former allows rapid blood flow between the prosthesis and the aneurysm, leading to a significant increase in pressures external to the endoprosthesis, and can lead to aneurysm rupture [3].

Type 1a endoleaks occur due to a progressive loss of adhesion between the proximal collar of the endoprosthesis and the vascular part, leading to cracking between the prosthesis and the aortic collar or migration of the prosthesis forward [4]. This issue has two leading causes: (a) the prosthesis is inextensible, so even a tiny diameter difference results in inefficient adaptation between the prosthesis and the pulsating vascular wall; (b) the pressure pulse increases at the junction between the normal and diseased vascular wall due to impedance change and pulse reflection, leading to increased pulsatility, making it more challenging to maintain a tight interface between the prosthesis tissue and the vascular wall [5]. This problem worsens if the aneurysm neck has an unusual and hostile morphology [3,6,7,8,9].

Therefore, ensuring the optimal adhesion of the endoprosthesis to the vascular wall during the implantation phase is crucial. At this stage, a balloon is inflated to occlude the vessel and dilate the prosthesis to adapt it to the wall. The aortic pressure increases significantly (corresponding to the ratio between the cardiac output of the occluded vessel and the total output), causing dilation and an increase in pulsatility at the junction point (systolic pressure rises throughout the arterial system). Endoprostheses are dilated to larger dimensions than aortic ones to promote adherence under increased pressure conditions, with oversizes ranging from 10% to 30% based on vessel physiology [10]. However, this approach stresses the vessel wall, leading to local stiffening and increased tangential forces that may result in endograft misplacement. Additionally, the prostheses feature hooks and barbs on the joint to enhance blood vessel adherence. While effective in the short term [1,11], this method still presents issues concerning the chosen stent oversize percentage during implantation. Stents that are excessively large compared to the aorta’s diameter reduce stent adhesion and the effectiveness of the barbs and hooks [12]. Beyond a 30% oversize, the risk of stent deformation (stent graft folding) significantly rises [13]. For this reason, since it is impossible to intervene during balloon inflation other than limiting the oversize to 10–20%, it is essential to reduce the pulsatility of the artery immediately after the endoprosthesis insertion. This solution may facilitate the adhesion of the prosthesis to the internal vessel wall.

There are few proposals in the literature to reduce local pressure at the aortic collar neck level. These methods are applied during fixation and aim to reduce or eliminate arterial impulse (mean pressure x aortic cross-sectional area x systole duration) through pharmacologically induced hypotension, adenosine-induced asystole, inflow occlusion, and rapid pacing. These methods are helpful; however, they have significant side effects, such as long times to restore correct blood pressure values, the risk of atrial fibrillation, conduction disturbances, arrhythmias, and difficulty in controlling any asystole. They cannot be used on specific categories of patients [14].

We suggest implementing an innovative strategy immediately after the surgical procedure to achieve the same objective and facilitate prosthesis adhesion. Our proposal involves placing a standard Passive Intra-Aortic Balloon Pump [15] (PIABP) as a passive damper. This device lowers systolic pressure and increases diastolic pressure, thereby reducing the pulse pressure [16]. As a result, the artery’s maximum and minimum diameters decrease, diminishing pulsatility in both amplitude and duration. This approach effectively minimizes leaks between the aortic wall and the prosthesis, enhancing the prosthesis’s attachment to the vascular wall. Additionally, if clinically necessary, the effects of the PIABP are transient and can be promptly reversed by deflating the balloon.

This paper presents the preliminary results of the in silico investigation of the above approach. To test our hypothesis, we built a silicon aorta inserted into a mechanical simulator of the cardiovascular system along with a PIABP. We simulated various filling conditions of the passive damper. We measured the maximum diameter and variations over time during pulsation to evaluate if the passive damper could help improve prosthesis adhesion and reduce endoleaks formation.

2. Materials and Methods

Arteries Preparation

In this preliminary work, two silicon-made aortas (Wacker Elastosil 601 A and 601 B, Wacker Chemie AG, Munich, Germany) were created: the first to simulate nearly physiological behavior in terms of diameter variations (approximately 2.5 mm) due to regular physiological pressure pulse ΔP (120/80 mmHg). The aortas were characterized by measuring a wall sample’s stress–strain (σ-ε) curve. Different samples with varying thicknesses were analyzed.

Square sheets of silicone with sides of about 10 cm were made by laying the material on a smooth surface. By laying down different layers of silicone, sheets of various thicknesses ranging from 1.0 mm to 2.5 mm were made for each sheet. A sample 82 mm long and 10 mm wide from each sheet was cut out. The length was chosen so that it was equal to the circumference of a 26 mm diameter tube.

Each specimen was inserted into a previously described strain–stress curve construction device [17]. Such a device allows the elastic specimen’s pullback force to be measured by a strain gauge, following controlled elongation by a stepping motor.

Each sample was mechanically extended: forces and elongations were continuously acquired and recorded with the ANScovery system (Sparkbio Srl, Bologna, Italy). By referring the forces to the application surface and the lengthening to the initial sample length, it was possible to construct the stress–strain curves for each sample. The mean Young’s modulus was calculated for each condition as the slope of the best interpolation curve obtained with a linear regression.

Then, for each sample, the theoretical diameter variations in the physiological conditions (ΔP) were calculated, considering Young’s and Laplace’s laws for a cylindrical chamber. In fact:

σ = Yε

(1)

σ = (∆P∙R)/d

(2)

where σ represents the wall stress, ε the strain, ΔP the pulse pressure (±1 mmHg), Y the Young’s module, R the internal radius, and d the wall thickness.

By combining the two equations and considering that the strain ε can be calculated as the difference in circumference (ΔC) divided by the initial circumference value (C₀) and the relationship between circumference length and diameter (C = πD), it is possible to calculate the theoretical diameter variation (ΔD) as follows:

∆D = (2∆P∙R²)/(Y∙d)

(3)

By inserting the required pulse pressure values (40 mmHg, 1 mmHg = 133.3 Pa), the measured Young’s module and thickness of each sample and the radius (13 mm), it was possible to calculate the theoretical ΔD for each sample.

The sample exhibiting normal physiological behavior, used as a reference to build the final vessel (length: 200 mm, internal diameter (D): 26.00 ± 0.05 mm), was the one with thickness (d) of 1.70 ± 0.05 mm and Young’s modulus (Y) equal to (4.20 ± 0.03)·10⁵ N/m².

Then, the aorta was inserted into a mechanical mock of the cardiovascular system (Figure 1), operating at a fixed frequency of 60 bpm with a mean aortic flow of 50 mL/s.

The syringe pump (Figure 1, A) simulates the left ventricle. The fluid used is a physiologic solution with glycerin to simulate the blood. The blood passes through the aortic valve (Figure 1, C), enters the silicone aorta (Figure 1, I), and is fed into a rigid tube equipped with a stopper (Figure 1, O) capable of varying the internal lumen and thus modulating peripheral hydraulic resistance. At the end of the systemic circulation, blood enters the left atrium (Figure 1, P) and re-enters the ventricle through the mitral valve (Figure 1, B). Inside the aorta, an IABP flask (40 mL, Maquet Mega 8Fr) (Figure 1, N) connected to an external rigid container (Figure 1, M) that serves as a gas reservoir was inserted. This container is connected via a tap (Figure 1, H) to a syringe to allow gas insertion, extraction, and pressure variation inside the IABP.

Aortic and PIABP inflating pressures were acquired with two transducers (Edwards TruWave, Edwards Lifesciences Corp, Irvine, CA, USA) (Figure 1, E,F). The aortic flow signal was measured with an electromagnetic flowmeter (Biotronex 610 square wave EMF, Biotronex Laboratory Inc., Silver Spring, MD, USA) (Figure 1, D) positioned immediately after the aortic valve (Figure 1, C).

Analog signals were then digitalized and acquired with the ANSCovery System (Sparkbio Srl, Bologna, Italy).

To measure the aortic volume, a digital camera (model acA 1300-75gc, Basler, Ahrensburg, Germany) interfaced with a Matrox Concord Card (Matrox, Dorval, QC, Canada) and connected to the ANSCovery System, was positioned above the aorta (Figure 1, L). Videos were acquired at a rate of 25 fps and a frame dimension of 1280 × 1024 pixels.

Aortic volume waveforms were calculated using a modified Simpson method [18]. Each frame was processed, and the contours of the aorta were extracted. Assuming each section was circular, the total volume was calculated as the sum of the volumes of the single slices extractable pixel by pixel. The system’s calibration was performed along two perpendicular axes by framing a sample of known dimensions.

Then, the maximum diameter (Dmax), the minimum diameter (Dmin), and the difference between them (ΔD) were calculated by averaging the diameters along the aorta in the tele-systolic and tele-diastolic phases.

The balloon pressure was regulated from a value equal to the diastolic pressure (80 mmHg) to a value close to the systolic pressure step-by-step of 10 mmHg. Twenty pressure, flow, and volume waveforms were extracted and averaged for each condition. Mean systolic and diastolic pressures, mean pulse pressures, mean aortic flow, and mean diameter variations were then calculated with their standard deviations.

Since the system is mechanical and stable over time, the standard deviations calculated over twenty beats were lower than the experimental resolutions. So, these last values were used as uncertainty. The errors of the arterial pulse and of the diameter difference were calculated according to the error propagation rules.

3. Results

Figure 2 shows an example of a stress–strain curve.

Table 1. Parameters of interest.

Inflating Pressure (mmHg)	Systolic/Diastolic Pressure (±1) (mmHg)	Arterial Pulse (±2) (mmHg)	Dmax ± 0.05 (mm)	Dmin ± 0.05 (mm)	ΔD ± 0.1 (mm)	% ΔD	Mean Aortic Flow (±1) (mL/s)
0	120/78	42	31.20	28.80	2.4	/	48
80	117/77	40	31.04	28.70	2.3	−4%	49
90	115/80	35	30.92	28.70	2.2	−8%	50
100	115/82	33	30.92	28.90	2.0	−17%	51
110	115/85	30	30.56	29.10	1.5	−38%	53
120	116/87	29	30.65	29.00	1.7	−29%	51

shows the results obtained with the aorta for different PIABP inflating pressure conditions.

Without the passive damper, the maximum diameter variation is 2.4 mm, comparable to the expected theoretical physiological value of 2.5 mm.

Figure 3 and Figure 4 show the pressure and flow waveforms at the different inflating pressures, starting from the basal condition (0 mmHg). It is clear that the pulse pressure reduces and reaches maximal effects for filling pressures of 110 mmHg and 120 mmHg.

The aortic flow changes with the IABP inflated at the different conditions and increases of 4 mL/s with a pressure of 110 mmHg.

As the passive damper pressure increases, the pulse progressively decreases, reaching maximum effect at pressures of 120 mmHg (Pulse = 29 mmHg). As expected, the maximum vessel diameter decreases during systole while the diastolic diameter increases. The maximum effect is achieved at a filling pressure of 110 mmHg, slightly smaller than the systolic pressure, resulting in a diameter reduction of −38%.

The mean pressure in the basal condition (120/78 mmHg) is 92 mmHg and becomes 95 mmHg with the PIABP inflated at 110 mmHg. This result is coherent (Hagen–Poiseuille Law) with the increase in the mean aortic flow from 48 mL/s to 53 mL/s.

Figure 5 shows the diameter variations along a beat obtained without the PIABP and the balloon inflated to the best condition (110 mmHg). The dotted line represents a theoretical fixed prosthesis diameter. The prosthesis diameter was chosen as 5% oversized (30 mm) to the minimum diameter (29 mm) during the diastolic phase without the PIABP.

Time intervals Δt₀ and Δt₁₁₀ represent the duration of the endoleak’s presence during a beat for the two conditions.

The Δt0 is much greater (0.44 s) than the time with the PIABP (0.28 s), with a reduction of approximately 36%.

4. Discussion

Endoleaks represent the most common complication following arterial endoprosthesis implantation surgery. These cracks form at different levels and allow the passage of blood into the aneurysm sac, leading to increased pressure and possible rupture. Depending on the site of this passage, there are different endoleak types. This study focuses on type 1 endoleaks, which occur when a crack forms between the proximal part of the endoprosthesis and the artery neck.

The formation of these cracks depends on several factors. The not-always-perfect adherence of the endoprosthesis to the blood vessel is caused by the particular geometry of the artery at the proximal fixation point. The poor elasticity of endoprostheses makes them unsuitable for dilating and contracting to follow the systolic and diastolic phases of the cardiac cycle. At the junction, over time, the pressures increase; locally, the diameter of the artery increases, and the tangential forces that tend to move the stent also increase [4,6].

To overcome this drawback and promote adhesion between the endoprosthesis and the artery, the endoprostheses are dilated to larger dimensions than those of the artery. They are equipped with hooks and barbs anchored to the blood vessel’s internal wall and should keep the prosthesis anchored to the arterial collar. Over time, a thin layer of endothelial cells forms at the adhesion surface between the artery and the prosthesis, which covers the proximal part of the endograft, favoring the stabilization of the prosthesis [19].

However, this phenomenon only occurs after some time. It requires time for the prosthesis to be fixed to the vascular wall to encourage the formation of endothelial tissues and thus strengthen the anchoring.

Since the inflated balloon obstructs the blood vessel during standard surgical procedures, the inflation and dilation phase of the stent must be rapid. The only possible solution is to administer the patient with a pharmacological therapy that maintains low arterial pressure and pulsation in the post-surgery phases. However, pharmacological therapies sometimes have serious side effects, making them not always applicable.

Furthermore, positioning a stent with dimensions greater than 20% of the aortic ones has proven inadequate. This can lead to a deformation of the prosthesis [12,13] and, therefore, to the formation of endoleaks. To overcome these difficulties, we propose inserting a passive counterpulsation device (PIABP), left inserted in the artery in the days following surgery. During this phase, both local pressures and aortic pulsations decrease. At the same time, the duration of the pulsations also decreases, favoring the anchoring of the prosthesis to the artery. The PIABP is a safe device and can be easily deactivated if necessary.

Our experimental results are comparable with the theoretically expected values obtained by applying Equation (3). The measured diameter variation with a pulse pressure of 40 mmHg (2.4 ± 0.1 mm) is lower but comparable to the theoretical one (2.5 ± 0.2 mm), and this underestimation is present for the diameter variation with a pulse pressure of 30 mmHg (theoretical 1.9 ± 0.2 mm; experimental 1.5 ± 0.1 mm). This difference is low, probably due to the mathematically simplified model used for theoretical evaluation and measurement errors in the Young’s modulus calculation. Moreover, the silicon aorta presents a constant Young’s modulus in the range of our measurements, and this is a limitation since blood vessels present variable elastance [20]. Despite that, considering a normal aortic elastance in its linear range, despite these limitations, a 38% decrease in the diameter variation with the PIABO inflated at 110 mmHg is crucial, demonstrating the feasibility of our proposal.

We compared the timing difference during which the arterial wall remains lower than that of the fixed endograft. For our simulation, we selected an enlarged prosthesis diameter of 30 mm (5% oversized) to mitigate the problems associated with oversizing that can damage the aortic wall and endoleaks. A 36% reduction in timing occurs when the aortic diameter exceeds that of the graft, extending the duration when the vessel and the prosthesis share the same diameter, facilitating adhesion. With oversizing above 5%, diameter fluctuations with the PIABP diminish and vanish. Clearly, with significant oversizing (in our tests with prosthesis diameters exceeding 31.0 mm), the benefits of the PIABP are no longer present within the prosthesis but remain crucial just outside it, where the aortic wall experiences higher stress, making it essential to decrease the pressure pulse to prevent wall damage.

Moreover, with the PIABP, the mean pressure and flow increase (+3 mmHg and +5 mL/s, respectively), improving the patient’s hemodynamic performance. This result is coherent with our previous findings [16].

Our proposal must naturally be extended to aortic walls with different stiffness and to a broader range of pressures since the clinical setting is variable and each patient is different. It should be validated with appropriate clinical studies and deepened with finite elements analysis to simulate different and more complicated cases. Although numerical methods are powerful, finite element analyses were not the aim of the present paper since a numerical approach needs additional and huge experimental work.

Our proposal can mechanically lower aortic pressure in the days after intervention and allow the reduction of antihypertensive drugs. This represents an improvement in therapy, with a consequent reduction in pharmacological-induced adverse reactions.

For our testing, we use a standard PIABP balloon for active counterpulsation, and this can frighten clinicians since it implies the insertion of a new and expensive catheter inside the artery that should be implanted for a few days after the intervention while the patient is in an intensive care unit, with higher management and clinical costs. Then, the catheter must be extracted after a few days, increasing the risks and length of hospitalization. This is undoubtedly true, but we are convinced that global costs will decrease if rehospitalization for an endoleaks occurrence can be avoided. The benefits for patients and the health system will be evident. A proposal that should be deepened to simplify the procedures and reduce the procedure costs is to use as the PIABP the same inflation balloon that allows the dilation of the endograft without the need to introduce an additional catheter.

Passive aortic counterpulsation has been proposed to optimize the mechanical matching between the ventricle and artery for treating pulmonary hypertension [21] and aortic dissection [22]. It was tested on patients [23] with promising results. Despite that, the limited presence of devices for passive counterpulsation on the market represents another major limitation for the clinical use of our approach since there is only one commercial device (Angiopulse, Angiodroid Spa, San Lazzaro di Savena, Bologna, Italy).

5. Conclusions

Passive counterpulsation is a suitable method for the local reduction of endovascular pressures and pulsations by mechanical methods without pharmacological intervention. Despite exploring only one normal physiological case, our preliminary work demonstrates that this approach applies to vascular endoprosthesis positioning interventions and can reduce aortic pulsation, resulting in better prosthesis fixation and reduction of the endoleaks phenomenon.