The Development of a Novel Peristaltic Test Stand “Swallow-Sim” for the Mechanical Evaluation of Esophageal Stents

Around 20 million people worldwide are diagnosed with a tumor every year. Of these, 600,000 cases are esophageal carcinoma, also known colloquially as esophageal cancer. Not only worldwide but also in Germany, the annual number of new cases is rising sharply [1]. This makes it statistically one of the eight most common types of cancer worldwide and the six most common cause of cancer-related deaths [2]. Compared to other types of cancer, there is no routine screening, which means that it is a disease that is often only detected in its final stages. If surgical removal is no longer possible, a stent can be used as a palliative action to enable nutrition [3]. The reason for this is that the carcinoma leads to a reduction in the diameter of the esophagus, making swallowing and food intake significantly more difficult (dysphagia) [4]. Dysphagia is said to occur when the diameter is reduced to approximately 12 mm, which depends also on the location within the esophagus [5]. Shape memory metal stents made from a NiTiNol alloy are currently used to ensure that patients can continue to live and eat as comfortably as possible.

When new stents are developed, their mechanical properties are examined after a development and design phase using various test methods. During in silico testing, a virtual investigation is carried out, for example, in the type of a finite element method simulation, where loads on the stent lead to deformations and stresses or strains [6]. Various publications on the design of cardiovascular devices in geometric and material terms are carried out using finite element methods [7–9]. With the help of computer aided design, geometry and material optimizations can be carried out at an early stage. In the case of metal mesh stents, adjustments to the braiding angle of the individual metal rods are often examined and the respective function checked. This is followed by in vitro testing, which involves tests carried out on a laboratory scale [10]. In the case of NiTiNol stents, for example, radial pressure tests are carried out using standardized test procedures. This involves testing the radial resistance of the sample at body temperature (37 °C) and comparing it with the stents currently in use. The ASTM F3067 standard deals with the radial pressure testing of self-expanding stents [11]. In addition, short-term tests are also carried out to evaluate tensile and flexural strength. However, both test methods have the problem of a lack of biokinetics. In addition, an acidic intraluminal environment in the distal esophagus, due to possible gastroesophageal reflux in some patients may also influence the individual conditions. In order to incorporate these additional biokinetic factors, the first in vivo tests will then take place. This means that the stents are now being used in living organisms (e.g., rabbits in the case of esophageal stents) to test their effectiveness and longevity in their entirety for the first time [12]. The results of the organism tests often differ significantly from the laboratory-scale tests [13]. This is not only an ethical problem for society but also an economic problem for the company developing the product. Costs continue to rise due to regulations and rising costs worldwide, with companies regularly incurring costs in the millions for the approval of medical implants or drugs [14]. The ability to detect problems in a new device at an earlier stage of the development process would provide significant savings in terms of both time and resources.

Phase I to III clinical trials are the final stage of the approval process. Depending on the phase, different numbers of test subjects are examined in double-blind studies. This means that neither the doctor nor the patient knows whether they are receiving an innovative stent design or a state of the art one. In this phase, the stents are tested, for example, for migration in the esophagus (slippage) [15]. If the stents slip downward after a period of use due to insufficient radial forces or to damage, they may no longer fulfill their purpose of keeping the esophagus open and, in the worst case, may slip into the stomach, necessitating surgical removal [16–19]. Slippage can be caused by a change in the mechanical properties due to the mechanical influence of peristalsis on the stent, by the effect of temperature or by a chemical attack on the stent material. In a worst-case scenario, all these external influences may converge, leading to dimensional changes, breaks in the wires or changes in the stent covering, all of which can lead to slippage [20–26].

Since in vivo test results are not fully predicted by previous in vitro testing, this research aims to develop a testing apparatus called Swallow-Sim that comes as close as possible to in vivo tests at an earlier stage. The aim of this test stand is to enable a decision on the probability of success of the innovative stent on a laboratory scale. For the developing companies, this not only has the economic advantage that the test stand may filter out undesirable developments at an early stage and thus avoid unnecessary costs, but it also has the potential to significantly reduce the number of animal experiments by raising the probability of success for stents that reach the in vivo stage. The test stand must simulate the biokinetic conditions of the subsequent environment as realistically as possible. Some of the main requirements for the test stand can be derived from this motivation. Primarily, the test stand should be able to reproduce the peristaltic movement (not a short-term tensile or pressure test, but a repetitive swallowing movement). Since the purely mechanical influence is often amplified by other impact factors, the tests should be carried out at clinically relevant temperatures. In this use case, it is the body temperature of approximately 37 °C. As patients with esophageal cancer often have reflux disease, the esophagus or the inserted stent also comes into contact with gastric juice. Among other things, this contains 0.5% hydrochloric acid [27]. This should also be considered in the test stand. To ensure that the esophagus tubes, innovative stents, etc., can be changed as easily as possible, attention should be paid to good accessibility and simple assembly and disassembly. In addition to the requirements mentioned above, a camera with an evaluation option for the long-term tests will also be implemented in the test stand.

In the area of the esophagus, NiTiNol stents in various designs (coated or uncoated, various cross angles, etc.) have been investigated and used for many years [28]. Since these stents often differ only slightly from one another, they are usually examined with a standardized radial pressure test [29]. However, only the radial force exerted by the stent is measured. The reason for this is that the aim of the stent is to keep the esophagus open, which, depending on the patient, requires a pressure of around 100 mm Hg [30]. This corresponds to about 133 mbar. The existing radial pressures from the muscles to the esophagus and stent are strongly influenced by the anatomy of each patient. In individual cases, pressures of up to 150 mm Hg may be present [31–33]. However, as the static force applied during the standardized radial pressure test does not correspond to the reality of dynamic forces inside the human body, the test results have limited applicability to the real application. For context, the muscles around the esophagus consist of two different layers. Viewed from the inside to the outside, there are circularly arranged and longitudinally arranged muscle fibers on the mucosal layer. The longitudinally aligned muscles ensure that the esophagus is shortened during the swallowing process, thereby bundling the ring muscles to further increase the radial pressure [34]. The ring muscles are responsible for ensuring that a bolus (swallowed object) is transported through the esophagus by contracting [35]. The aim of the Swallow-Sim test stand is to simulate the ring muscles, which have a direct influence on the esophagus and therefore the stent, as realistically as possible. A few other devices with similar goals have been reported. In their publication, Xingzhong et al. from the School of Mechanical Engineering in Nanjing present a device based on a cam drive and located in a temperature-controlled gastric juice bath [36]. This test rig has two disadvantages. The first is the permanent immersion of the stent in the gastric juice bath, which does not correspond to reality (both in terms of orientation and the duration and intensity of the acid exposure). In addition, the implementation with the cam drive results in only one-sided loading of the stent. In reality, the ring muscles exert full pressure on the stent in 360 deg. Another publication by Li et al. deals with a swallowing device that transports objects from A to B using foils inflated with air. The device is primarily designed to examine various components that are transported during the swallowing process (sharp, angular, round, etc.) rather than stents [37]. In a publication of the Bhattacharya research group, a device for examining endoprosthetic stents is described [38]. Pneumatic-mechanical ring systems are connected in series along the length of the stent. These close the stent with the aid of a pneumatically controlled complex mechanical closure function or reduce the diameter in the same way as the esophageal function. The design of the device is very complex, which makes it difficult to replace individual components. In addition, only the mechanics of the stent are examined here, the increased temperature (body temperature) and the contact with gastric juice are left out. Furthermore, the four-chamber system does not achieve a truly uniform radial pressure [38]. The movement robot from the publication Boxerbaum et al. comes closest to accurately stimulating the ring musculature, as a network of longitudinal and transverse muscles is stimulated to perform a peristaltic-like movement. However, body temperature and acid resistance are not considered. Furthermore, the publication shows that the movement sequence is clearly too slow, as the control of the respective positions is described as lengthy [39].

In addition to scientific publications, several patents and utility models are also publicly accessible. Ni's patent from 2014 describes an experimental test stand for fatigue testing of stents. A hydraulic or pneumatic system is used to transport a medium through an annular slot, causing the stent to contract radially. The peristaltic movement can be investigated by arranging several nozzles [40]. In the patent “Testing device for fatigue property of medical self-expanding nonvascular lumen stent” Xing describes a device based on a cam drive movement similar to the publication described above. However, the cam drive is used to load the stent on one side, which does not correspond to reality [41]. Two further patents deal with the swallowing mechanism but are not described as a test stand but merely as a swallowing robot for imitating the swallowing mechanism within the scope of the patent [42,43].

As described above, the Swallow-Sim test stand is intended to reproduce as closely as possible the environment in the esophagus. In addition to the basic function of peristalsis, this includes the body temperature and the acidic environment due to gastric juice. The superimposition of these three stresses can compound to produce stent failure. The implementation of the individual subfunctions is described in the sections Acidic Environment, Body Temperature, and Peristaltic Movement. Figure 1 shows a schematic sketch of the test stand Swallow-Sim with all functions. The electronics (2) of the test stand are fixed to the top of the fluid-filled area on the housing of the test stand. The heating elements including the fan (1) are also located there. In the middle of the test stand, there is a casted silicone esophagus (3) with cords (4) fixed around it and a NiTiNol stent (5) on the inside. Under the test device is a reservoir (6) with gastric juice (7), which is pumped through a tube system (8) with the aid of a peristaltic pump (9) by the rotational movement of the latter, thus creating a closed circuit. In the middle of the stent, the temperature sensor is positioned.

To test the acid resistance of the stents, a peristaltic pump (12 V, volume flow 500 ml/s) is used to achieve a continuous wetting state of the stent. The advantage of a peristaltic pump is that it is a closed system. This means that no mechanical components come into contact with the gastric juice containing hydrochloric acid, because the rotor of the system makes a rotational movement outside the tube and compresses the tube to move the gastric juice through the system. There cannot be any interaction between the gastric juice and the peristaltic pump. This can be seen in Fig. 1. The fluid is supplied at a freely selectable point in the esophagus (in this case from above) and then wets the stent and the esophagus. In order to be able to reuse the supply of gastric juice permanently, there is a collection tray on the underside from which the peristaltic pump is fed again. The circulation process can be seen in Fig. 2. The potentiometer allows both a freely selectable dosage and a simple switching on and off, thus simulating reflux disease as realistically as possible. In the human body, reflux causes the transportation of gastric juice from the stomach into the esophagus. A commercially available concentrate with a pH value of 1.5 is used as the gastric juice concentrate [44]. As this investigation in the Swallow-Sim is a worst-case scenario, the gastric juice is applied constantly during the complete test duration. This is realistic for the lower part of the esophagus and stents which can reach all the way to the stomach.

In order to simulate the esophageal environment as realistically as possible, the test stand is heated to body temperature. In a publication, Carbonaro et al. have shown a four-point bending test rig where the apparatus is in a bath at body temperature [45]. This is not possible in our case, as the properties of the stent to be analyzed could change in interaction with the test stand. Air heating is used in this paper. A powerful 300 W PTC heater (ceramic heater) is used for this. The distribution of the warm air in the installation space is realized by using a radial fan with an additively manufactured air guide component. The temperature is brought to a body-like temperature of 37 °C before the start of the experiment and kept constant at this value. To check the temperature, the installation space is continuously monitored at various points using thermal sensors. The different sensors lead to different results, which can be seen in Fig. 3. The results will be described later in the paper. To keep the temperature in the installation space as constant as possible, a complete enclosure is created using polymethyl methacrylate sheets (PMMA or Plexiglas). These have the advantage of being highly transparent and easy to machine. The positioning of the radial fan, including the heating element, is chosen so that no liquid can enter this area in event of a malfunction. For this reason, the components are attached to the top of the test stand. This is visible in the principal sketch of Fig. 1.

Peristalsis is the most important component of the test bench. The aim is to reproduce the swallowing process as realistically as possible. The publications and patents known to the research team all describe a load case that deviates from the radial constriction of the ring muscles around the esophageal wall. Various concepts were tested as part of the development process, but the most promising was found to be a multipart cassette cord system. The cords, which consist of tear-resistant fishing line, are attached and placed at a defined distance around the esophagus including the stent. The cassettes are activated by an electronically operated carriage and the cords are placed radially around the esophagus at equally spaced positions along its length, resulting in a peristaltic movement. Figure 4 shows the components of the cassette system responsible for peristalsis. The complete cassette system comprises 14 individual cassettes, although the number can be extended or reduced at any time as required using a simple plug-in system. The movement sequence of the cassette system is designed to mimic the swallowing process as closely as possible. An additively manufactured slide (1) is moved continuously between two optical end stops using a threaded spindle (2). The movement of the carriage with the curve geometry (see figure inset on the right) and the deflection of the rollers (5) cause the moving parts of the cassette to move to the left (6). The movement causes the respective cord (3) to contract at the actuated height around the replica of the esophagus (4). The circular movement results in a continuous closing process, which should closely resemble the peristalsis of the esophagus. Due to the continuous up and down movement, the stent to be examined is exposed to pressure changes more frequently in a shorter period of time than in reality. This has the advantage that the test duration can be reduced. Another representation of the peristaltic movement can be seen in Fig. 5, where four different states of the downward movement of the carriage and thus the contraction of the cords can be seen. The examinations were also conducted from inside the esophagus using an endoscope. This revealed a similar pattern of behavior to that observed in the human body.

Figure 6 shows the design of the test device. The mechanical stability is generated by means of commercially available aluminum profiles (1). Black or transparent Plexiglas plates are used as covers. These serve to keep the heat inside the test stand. The peristaltic pump (3) pumps the caustic agent in a circuit through the stent to the collection tray (2) to restore the acidic environment. From there, the liquid is returned to the peristaltic pump. Above and below the esophagus (5) is the esophageal receptacle (4/shown in gray). The esophagus, in this case a molded silicone tube (Shore 50 A) with a length of 20 cm, is attached there. The esophagus length depends on different properties of the patient such as age, size, sex, etc., Qun et al. reviewed 252 esophageal manometric studies and showed, that the mean length of the esophagus is 22.9±0.2 cm [46]. During the test validation, this can be replaced with an animal esophagus (e.g., from a dog or pig). The cords are wrapped around the silicone tube. The motor (7) moves the carriage up and down, causing the cassettes (6) to move axially and pull the cords to reduce the diameter. This creates a continuous peristaltic movement. The heater (8) sures a constant elevated temperature in the Swallow-Sim of about 37 °C.

To validate the test stand and the pressures exerted by the cords on the stent, the esophageal manometry device is used. High resolution manometry (HRM) is the most modern and accurate clinical evaluation method of esophageal motility [47]. A pressure measuring device is inserted into the esophagus via the nose and the swallowing process is triggered by administering a sip of liquid [48]. The pressures then generated by the esophageal muscles are recorded and used to discuss and evaluate treatment options with a panel of experts in order to subsequently provide the patient with appropriate treatment. In the case of this publication, the measuring device was used to validate the pressure generated by the cords. Due to the close-meshed arrangement of the sensors, the pressures could be validated separately for each individual cord. The pressures were then compared with the pressures typically occurring in the human esophagus and adjusted by changing the additively manufactured sled. In a healthy esophagus, radial pressures of approximately 100–150 mm Hg occur, which corresponds to approximately 0.133–0.200 bar [30,49]. The MANOalpha device from Jinshan, China was used for the manometry measurement [50]. A few swallowing processes of the machine can be seen in Fig. 7. The results will be described in results on the following pages. A major advantage of HRM over traditional esophageal manometry is the ability to display and examine a complete picture of esophageal function using many pressure measurement sensors. In Swallow-Sim, the pressure that the cords exert on the esophagus can be adjusted by changing the additively manufactured sled or the interaction of amplitude and return springs. For this reason, various additively manufactured sleds were initially examined during the manometry measurements in order to keep the maximum pressures occurring within a range that can be measured meaningfully and was not destructive for the manometry measuring device.

The final stage of the investigations was a complete run of the test rig with an esophageal stent, followed by mechanical testing of the stent to quantify degradation of mechanical performance caused by the simulated esophageal environment. Mechanical testing was performed on two Hanarostent ECD-20-080-070, one after peristaltic loading in the Swallow-Sim test stand and the other in as-received condition. These stents have a length of 80 mm and a diameter of 20 mm in the middle and 26 mm on the tulips. The maximum diameter of the NiTiNol wires is 0.97 mm. Five types of mechanical tests were performed on both stents, with five repetitions for each type of loading. By testing only to a preset deformation point rather than to failure, each stent could be tested multiple times. All tests were done on a universal testing machine type T1000-700-10 kN. The tests, which are shown in Fig. 8, include a typical three-point bend test (1), a flat compression test (in which two plates are pressed axially onto the stent) (2), a planar radial compression test with three arms from top and bottom (3) (local radial compression test is with only one arm from top and bottom) and a typical tensile test in which the stent is clamped in the area of the two tulips and a force–displacement curve is then recorded (4). The results will be discussed in the next section and can be seen in Fig. 9. The stent which was placed in the test stand was subjected to a fictitious duration of six weeks of exposure to gastric juice, body temperature, and recurring peristaltic movement, as six weeks is the minimum length of stay of a NiTiNol-Stent in the human body. In future tests after an optimization of the test bench (for reasons described below), it is planned to extend the test duration up to six months. This is intended to show whether the test stand places similar loads on the stent as the human body and whether further optimizations to the test stand are necessary based on the endurance tests. Typically, stents are in use for around six to eight weeks, which is why the literature recommends testing the stents for at least six to eight weeks [51]. The physiologic swallowing process takes between 10 and 20 s. The actual physiologic transport of a bolus may take longer depending on the bolus and other factors. As the design of the test stand allows the swallowing processes to be significantly accelerated, the length of the test can be reduced. The test length duration was calculated using a spontaneous swallowing frequency of around 0.26 swallows per minute in older people with an increased risk of dysphagia [52]. This means that around 16,200 swallows will affect the stent in six weeks. By adjusting the speed of the stepper motor, the duration of one swallowing process in the test stand can be reduced to only 2 s and thus the six weeks of testing (16,200 swallows) are completed in approximately 9 h. In this study, the test stand is operated under real conditions (body temperature + acidic environment).

The peristaltic movement describes a person's swallowing process. The esophageal muscles are stimulated in a linear movement in an axial direction from the pharynx to the stomach in sequence and contract or constrict the esophagus. This is also done in the test stand as shown in Fig. 5. The left side shows the unstressed state of the cords and therefore no compression of the esophagus. In the image sequence, the muscles (in this case red cords) are stimulated or contracted from top to bottom and thus exert pressure on the esophagus to move the food through the esophagus or simulate a swallowing process. The contraction exerts pressure on the stent inserted in the esophagus.

The manometry measurements were carried out with different sled amplitudes, which can be easily modified due to additive manufacturing. To avoid damaging of the manometry probe, the cassettes were first operated by hand. During the first manual operation, it could already be determined that the pressures were outside the measurable range of the measuring device (>200 mm Hg). By adjusting the sled geometry and thus changing the amplitude, the radial pressure of the cords on the esophagus can be adjusted. After adjusting the amplitude, the targeted pressures of around 130 mm Hg to 150 mm Hg were achieved. It was found that the pressures decrease sharply after a few passes. A wrinkling of the cast esophagus was detected quickly and optimized by adjusting the connection technique. The pressures could then be kept constant in subsequent examinations. This can be seen in Fig. 7, where all amplitudes reach nearly the same pressure. The data from Fig. 7 can also be used to determine swallowing, which was 10 s, corresponding to the actual swallowing time and to the requirements of the body. Figure 7 shows the results of the HRM measurements of the optimized sled. It can be seen that a relatively even distribution of pressure acts on the esophagus and the stent from the top to the bottom on the test stand. Compared to the images from conventional esophageal manometry, the measured values are in a smaller (lower) image section. The reason for this is that the complete image would record the entire measurements of an esophagus length. As the test stand is only used to measure a stent, which has a standard length of approximately 120–150 mm, it is significantly shorter than a complete esophagus, which has a length of approximately 230 mm [46].

Various temperature sensors are used to monitor the temperature inside the test stand. One sensor is located directly in the ribs of the heater, which also controls the temperature. Three further sensors are located in the installation space at the height of the stent in order to generate a larger number of measured values. The results of the temperature measurement can be seen in Fig. 3. Four different lines can be seen. The solid black line represents the course of the temperature sensor, which is in direct contact with the PTC. There is a clear overshoot of this temperature. For a better examination, a sensor is positioned about 80 mm away from the PTC (dotted line). Here it can be seen that the overshoot is not that high, but relatively strong temperature fluctuations can still be seen. However, the decisive values are provided by the sensors at stent height (longer dashed line). These fluctuate over the recorded course around the target body temperature of 37 °C (solid line). This shows that the test stand can also constantly maintain the temperature over a longer period of time. The temperature measurement did not take place over the full duration of the test with stent. In further investigations, the temperature will also be recorded over the full duration of the test.

The distribution of the liquid in the complete circuit was tested using a caustic agent and a few drops of added food coloring. The caustic agent has a similar composition to the gastric juice. In order to ensure that the simulated gastric juice is distributed in a physiologically realistic way over the full length of the esophageal model, a few drops of food coloring were added to the fluid. Thus, the distribution can be followed by observation of the process in the entire circuit. Figure 2 shows a pictorial sequence of the first filling process of the circuit. On the left-hand side, the gastric juice is only in the storage container (borosilicate glass). In the next two partial images, the tube and circuit are filled up to the top of the esophagus. The image on the right shows the circuit completely flooded with gastric juice. The gastric juice is also evenly distributed inside of the esophagus from top to bottom. The images in Fig. 2 were taken without a stent in place. In subsequent tests with a stent, even more complete wetting of the inside of the esophagus was observed.

The final investigations to validate the test stand were carried out on a NiTiNol stent. A before-and-after comparison of the stent will be presented. Due to the high test speed, the test stand was subjected to enormous mechanical stress and wear. This was also evident during the test, as some minor problems arose. Above all, the friction within the individual cassettes causes heavy abrasion, which turned into a viscous mass with the lubricant of the threaded spindle and significantly impaired the sliding properties. In addition, the mechanical loads of the cords on the esophagus also caused damage, which led to the first leaks after a test period of around 4 h (equivalent to 7200 swallows). Due to the mixture of lubricant and abrasion, the thread of the spindle nut was damaged after a period of 9 h. Damage to the stent could be detected during a visual inspection. Some wires of the NiTiNol mesh were broken and the stent's cover was also damaged in several places. For this reason, the trial was discontinued after a duration of 16,200 swallows, which corresponds to a load of around six weeks. The stent was then subjected to further mechanical tests. A displacement-controlled load case was examined in order to investigate the maximum forces still acting. This means that the generated force is measured after a defined path, the reason for this is that the measured forces before and after loading by the test stand can be compared. The results of the tested stents (80 mm × 20 mm) tested before and after peristaltic loading are shown in Fig. 9. The average of the values measured on one stent from each of the five repetitions is shown with a standard deviation. It can be seen that in test 1 (three-point bend test), the force needed to cause 10 mm of deformation decreased from 2.41 N (before) to 1.28 N (after). This corresponds to a percentage decrease of approximately 47%. In the flat compression tests, the values decrease from 5.72 N to 3.28 N (at a distance of 13 mm), which also corresponds to a percentage decrease of around 43%. In the two radial compression force tests (both the local and the radial compression test in the width), the decreases from 10.42 N to 5.48 N and 22.63 N to 11.27 N are of a similar order of magnitude. In the tensile test, a force of 5.15 N is initially required to stretch the stent 10 mm while a force of only 3.39 N is required after the peristaltic testing. This also corresponds to a reduction of approximately 34%. In Fig. 10, the different changes which occurred during testing of the NiTiNol stent for 16,200 swallows in the Swallow-Sim test stand are shown. On the left side, there is a length change of 4.3% to see. In the middle, there are several cracks and breaks of the NiTiNol wires and on the right side holes and cuts in the covering are visible.

The initial results of tests made with the new Swallow-Sim test stand are very promising. Based on the first investigations of NiTiNol stents, which are currently being inserted into the human esophagus, it was confirmed that the mechanically activated cords produce a realistic simulation of the esophageal muscles. Due to the rhythmic movement and the described circular path, a realistic swallowing movement takes place. This is also confirmed by the images taken with the endoscope. It can be seen that the stent closes to a very small diameter, as also occurs in the esophagus. The results were also confirmed by the manometry measurements. With the help of the test stand, the desired pressures can be applied to the esophagus model and the stent inserted there in the desired duration of around 10 s per swallowing process. For the long-term testing, the speed of the Swallow-Sim can be increased (as it is done in this examination), so that the swallow process takes place in 2 s, which is favored for longer test durations. The temperature control of the installation space and the supply of gastric juice work well and thus the desired superimposed load of temperature, medium and mechanics can be investigated. However, the tests on the NiTiNol stents revealed a few optimization options for the test stand. In addition to a mechanical revision of the components (the additively manufactured plastic components in particular still require revision) and an optimized production of the cast esophagus, an enclosure of linear rail to protect it from dust and abrasion is an optimization option. The shown damage like length changes, breaks in the NiTiNol wires or holes in the covering, which are visible in Fig. 10, lead to significant changes in the mechanical behavior of the stent. The damage that occurs to the stent and the decreasing mechanical strength can also be observed repeatedly in implanted stents in humans [20–26]. The comparison between the mechanical changes after the six weeks simulation in the Swallow-Sim and the referenced research papers of the clinical investigations once again shows relatively clearly that the test stand meets the necessary requirements to simulate the human surrounding and that the radial pressures from the cores are within a realistic range. Damage to the stent could lead to eventual migration in a human esophagus. In the test bench no migration or movement of the stent could be detected. For subsequent optimization, an adapted esophageal model of a patient could be used to investigate migration there as well, so that the same friction coefficients between the stent and esophagus are also present.

As described in this paper, an esophageal test stand was developed that incorporated mechanical peristaltic loading and clinically relevant environment conditions (acidity and body temperature) and the feasibility of its use was demonstrated. All desired functions can be implemented and they also fulfill their respective purpose. Two points need to be noted at the present time in order to optimize the test stand. The first point relates to the construction of the cast silicone esophagus. After some time of continuous up and down movement, the first cracks and incisions can be seen. This can be avoided by using friction-optimized cords or other cast silicones. A thicker-walled cast hose should also be targeted here, as the tendency to buckle was shown in the manometry measurements. The second optimization point is aimed at feeding the gastric juice. Currently, this is fed into the stent from above (orally), whereby the real physiologic esophageal exposure with acidic gastric juice occurs from the distal esophago-gastric junction and refluxes upward into the lower half. In the next iteration loop, a lateral feed must be planned, but this should not be a problem due to the self-manufactured esophagus. A further optimization option is the encapsulation of the linear rail and a mechanical improvement of the components used in the cassette system. In this device, stents can be examined with regard to the mechanical loads in an accelerated examination. The influence of long-term stress caused by storage in gastric juice should also be tested in a separate experiment so that potential influences are not neglected here. The project team has identified further possibilities for the area of application through the initial tests. As the muscles of the esophagus with an inserted stent tire after around two months due to the greater pressure, they are fighting against and no longer have any effect, this can also be investigated in the test stand by excluding the respective cords from the circular path. The same applies to the investigation of the influence of a carcinoma. Here, for example, a permanent contraction of the cord could lead to the simulation, of a very pronounced carcinoma. Other cross-sectional technologies are also conceivable. Peristaltic movements take place not only in the esophagus but also in the colon or ureters, for example. For this, the test stand would have to be adapted in terms of its constriction diameter and media load, but can be operated according to the same principle. Another possibility is the usage of monitoring stent positions before and after the swallowing simulation to detect any migration because of damage to the stents. For this usage, there are changes and further examinations necessary like other esophagus systems (because of the friction coefficient), a better sensor monitoring system, etc. In additional to the optimization of the test bench, there should be more testing of different stents to benchmark the test bench.

As part of the joint research project “3D-Stent” with the funding code LSM-2203-0002, we are collaborating with the UKW—University Hospital Würzburg (Professor Dr. med. Alexander Meining and Professor Dr. med. Karl Hermann Fuchs). We are very grateful for the funding from Bayern Innovativ GmbH and the Bavarian State Ministry of Economic Affairs, Regional Development and Energy. Another big thank you also goes to the company Promedia, which greatly supported the research team by providing the manometry measuring device free of charge. Based on these measurements, further optimizations can be made to the test bench.

• This work was partly supported and funded by Bayern Innovativ GmbH and the Bavarian State Ministry of Economic Affairs, Regional Development and Energy. The entire project consortium would like to thank them for this.

No funds other than those already provided under the “Funding” section have been utilized in the context of this publication. The authors confirm that they do not have any other conflicts of interest.

The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request.