A low-cost bioprosthetic semilunar valve for research, disease modelling and surgical training applications

Abstract

OBJECTIVES

This paper provides detailed instructions for constructing low-cost bioprosthetic semilunar valves for animal research and clinical training. This work fills an important gap between existing simulator training valves and clinical valves by providing fully functioning designs that can be employed in ex vivo and in vivo experiments and can also be modified to model valvular disease.

METHODS

Valves are constructed in 4 steps consisting of creating a metal frame, covering it with fabric and attaching a suture ring and leaflets. Computer-aided design files are provided for making the frame from wire or by metal 3D printing. The covering fabric and suturing ring are made from materials readily available in a surgical lab, while the leaflets are made from pericardium. The entire fabrication process is described in figures and in a video. To demonstrate disease modelling, design modifications are described for producing paravalvular leaks, and these valves were evaluated in porcine ex vivo (n = 3) and in vivo (n = 6) experiments.

RESULTS

Porcine ex vivo and acute in vivo experiments demonstrate that the valves can replicate the performance of clinical valves for research and training purposes. Surgical implantation is similar, and echocardiograms are comparable to clinical valves. Furthermore, valve leaflet function was satisfactory during acute in vivo tests with little central regurgitation, while the paravalvular leak modifications consistently produced leaks in the desired locations.

CONCLUSIONS

The detailed design procedure presented here, which includes a tutorial video and computer-aided design files, should be of substantial benefit to researchers developing valve disease models and to clinicians developing realistic valve training systems.

INTRODUCTION

Valvular heart disease is a serious health problem, with a prevalence of 2.5% in the US population [1]. Extensive investigations are ongoing in the valvular heart disease research community to develop novel techniques and technologies for valve repair and replacement. Validation of new approaches requires testing in large-animal in vivo models. Such evaluations must typically include accurate and repeatable creation of the valvular disease model. Currently, there are no readily available research-grade prosthetic heart valves for valvular disease modelling. While it is sometimes possible for researchers to obtain expired clinical valves from hospitals or device manufacturers, these are rarely available in sufficient quantities and sizes to complete a study. Also, while some non-expired clinical valves are available for purchase by researchers, the cost per device can run into thousands of dollars. Additionally, modelling of a specific disease condition may require either alteration or damage to the valve, which further increases cost and adds to the overall complexity of the studies.

A variety of low-cost training simulators have been proposed that include valve models for practicing surgical repair techniques [2–7]. While effective for valve surgery training, the models in these systems are not intended for use as functioning valves or for use in animal studies. There are also fluid-filled pulsatile simulators and training systems that incorporate functioning valves [8–10]; however, these systems are designed primarily for testing of imaging and navigation systems rather than for valve procedures. While some interventional cardiology training systems include replaceable valve inserts [11], these valves are not designed for in vivo use. The most closely related literature is devoted to designing improved clinical valves through the use of new materials as well as tissue engineering [12–14]. While these efforts offer promise for improving future clinical care, they are impractical for valve repair research since they are still in the research stage and their construction involves substantial expertise, technical infrastructure and cost.

In this paper, we present the design and fabrication method of a prosthetic semilunar (aortic or pulmonary) valve that can be custom-made at low cost for surgical training and disease modelling. It is comparable in design to clinical bioprosthetic valves [15] but adapted to be simply made from readily available materials. Additionally, it can be easily modified to incorporate a particular disease. In particular, we demonstrate how to include a paravalvular leak model (PVL) in the design and fabrication process, which was successfully validated in ex vivo and in vivo acute large-animal experiments.

MATERIALS AND METHODS

Fabrication of valve implant

The valve consists of 4 parts: a metal frame that provides structure to the valve, a fabric to cover the frame, a suturing ring that allows a surgeon to implant it inside a heart and the leaflets that provide the actual valve function. The steps involved in fabricating each component are described below. A video illustrating the complete fabrication process is provided (Video 1). Computer-aided design (CAD) models and templates for each component are provided in the Supplementary Material. Over the course of making approximately 10 valves, fabrication time reduced to 2 h with about 1 h spent cutting and sewing the frame components and 1 h spent sizing and suturing the leaflets.

Video 1:

Download video file ^{(9.8MB, mp4)}

Video showing the complete fabrication process.

The metal frame was made first and was sized based on measurements obtained from the literature [16]. Two fabrication techniques were developed, both of which can be used to produce functioning valves (Fig. 1). The first consists of a wire-bending process using a 2-part mould (Fig. 1A and B), which was designed with CAD software and 3D printed in rigid plastic (Objet Alaris 30 printer, VerowhitePlus material). A soft stainless steel (SS) wire (Bend-and-stay 316 l stainless steel wire, 0.020″ diameter, McMaster-Carr, Elmhurst, IL, USA) was placed in the mould, and the 2 mould components were pressed together causing the wire to take the desired shape under the pressure of the mould (Fig. 1C). Optionally, to increase the annular stiffness and strength of the frame, a tungsten inert gas welder can be used to attach a support ring of SS wire to the bottom of the stamped ring (Fig. 1C). The support ring can be made by wrapping (potentially stiffer) SS wire around a bar of slightly smaller diameter than the frame diameter.

Figure 1:

Fabrication of valve frame. (A and B) CAD models of the 2-part mould for wire bending; (C) 3D printed 2-part mould and resulting wire frame; and (D) CAD model of metal 3D printed frame.

An alternate fabrication approach is to employ metal 3D printing from a CAD model (Fig. 1D). Web-based 3D printing vendors (shapeways.com and i.materialise.com) offer 3D printing in a variety of metals including SS and titanium. It should be noted that some commercially available printing methods incorporate additional alloys, such as bronze, into the desired alloy e.g. SS. Such mixtures may be suitable for acute animal trials but may not provide the long-term biocompatibility and performance necessary for survival experiments e.g. due to corrosion in a blood environment. Components made from titanium and chrome-cobalt 3D printing are more expensive than bronze-infused SS, but provide better biocompatibility and durability—resulting in valves that may be better suited to survival studies. An advantage of 3D metal printing over stamping is the greater geometric design freedom and ease with which the design can be resized or adapted to include disease-modelling modifications.

Once the frame is made, it must be covered in fabric to prevent blood flow around the outside of the valve. Many fabric choices are possible. We used a non-woven 0.25 mm-thick polyester fabric. This material is readily available in a medical lab, e.g. in the form of surgical gowns and sterile drapes. It is strong enough to hold thin surgical suture, sufficiently impermeable to resist blood flow under physiological conditions and resistant to temperatures up to 160°C. The fabric is cut using a template, folded in half and sewn to cover the frame, as shown in Fig. 2A and B.

Figure 2:

Valve assembly. (A) Creating the fabric sleeve for the frame. (B) Fabric sleeve sewn over the valve frame. (C) After adding the sewing ring. (D) Fabricating leaflets from pericardium and attaching to the valve. (E) Close-up view of the leaflet suture line. (F) Completed valve.

Similar to clinically used surgical valves, a suturing ring is added around the exterior of the base of the valve. This ring can be used by surgeons to pass suture through during the attachment of the valve to the aortic or pulmonary annulus. In our design, the suturing ring was made out of polypropylene felt (Cole-Parmer, Vernon Hills, IL, USA), which is cut in 5-mm wide strips and sewn on the side of the valve using 4-0 silk surgical suture (Ethicon, Somerville, NJ, USA) (Fig. 2C).

Finally, the leaflets were made from porcine pericardium obtained from fresh cadaveric pigs and fixed using 0.625% glutaraldehyde for 20 min in a flat, non-stretched configuration. The leaflets were then cut from the fixed pericardium sheets (Fig. 2D) in a semicircular shape with a diameter of 1.15 times the inside diameter of the valve frame, with an additional 1.5–2 mm added to account for the suturing margin, following published leaflet sizing recommendations [17]. Double-armed 6-0 Prolene sutures were used to suture the circular edge of the leaflets to the curved edge of the frame fabric (Fig. 2E). The standard technique for leaflet attachment using double-armed suture was used in which the first stitch is passed through the midpoint on the frame, followed by passing the same needle through the midpoint of the leaflet semicircular edge. This is followed by a continuing running suture line from the centre towards opposing commissures using one needle for each direction [17, 18]. It is important that each needle bite goes around the metal frame/wire of the valve. Leaflet’s free edge apposition at the commissures can be adjusted after suturing each leaflet. Alternatively, all 3 leaflets can be sutured in, leaving the last 1 or 2 stitches at the commissures. The leaflet’s free edges can then be precisely positioned against each other and sutured to the frame at that point. As is the case for clinical bioprosthetic valves, after completion, the valves (Fig. 2F) were stored in a glutaraldehyde solution (we used same 0.625% glutaraldehyde solution that was used for fixation of the porcine pericardium sheets). The leaflets can also be made out of alternate biological and non-biological materials. In particular, if the valves are to be used for training or experiments that do not require fully functioning leaflets, they can be cut from thin sheets of elastic materials such as latex or silicone.

Table 1 summarizes the costs of all non-tissue supplies used to fabricate the valves. Note that this price can be reduced using, e.g. expired suture. Pericardium can often be obtained for no cost at facilities that perform animal experiments. Alternately, fresh tissue can be obtained from a local slaughterhouse at modest cost.

Table 1:

Valve material lists and costs for 3 valve designs (prices in US$)

Material	Supplier	Price	No. of valves	Price/valve	Valve design
					Stamped frame	3D printed frame (SS)	3D printed frame (Ti)
3D printed 2-part stamp	Shapeways	10.20	5	2.04	2.04
SS wire	McMaster-Carr	8.85	690	0.01	0.01
3D printed SS frame	Shapeways	6.80	1	6.80		6.80
3D printed Ti frame	I-Materialise	60.90a	1	60.90			60.90
Surgical drapes	Cardinal health	53.76	1000	0.05	0.05	0.05	0.05
4-0 silk sutures	Ethicon	61.56	6	10.26	10.26	10.26	10.26
6-0 Prolene sutures	Ethicon	186.91	12	15.58	15.58	15.58	15.58
Polypropylene felt	McMaster-Carr	5.31	100	0.05	0.05	0.05	0.05
Pericardium	Research 87	27.00	5	5.40	5.40	5.40	5.40
Total investment					$353.60	$341.30	$395.40
Price per valve					$33.44	$38.10	$92.20

Provider details: Shapeways, New York, NY, USA; McMaster-Carr, Elmhurst, IL, USA; I-Materialise, Leuven, Belgium; Cardinal Health, Dublin, OH, USA; Ethicon (Johnson & Johnson), Somerville, NJ, USA; Research87, Boylston, MA, USA.

SS: stainless steel; Ti: titanium.

^aIncludes 15 US$ of delivery costs to the USA.

Valve modifications for simulating defects

The valve design described above can be modified to simulate different disease states that involve prosthetic valves. Having control over the entire design process makes this step easy, reproducible and less expensive in comparison with attempting modification of a clinical valve. For example, leaflet pathologies can be modelled by changing the shape and size of the leaflets (e.g. adding central regurgitation by shortening the leaflet’s free edge length) or by modifying their attachment (e.g. stenosis, by suturing together the adjacent leaflets).

Modelling other diseases may require structural changes to the valve. To illustrate such modifications, we considered the design modifications necessary to model aortic PVLs—a pathological condition in which a gap develops between the native valve annulus and the circumference of the prosthetic valve. This gap allows a regurgitant blood flow from the aorta back into left ventricle even when the leaflets are functioning normally. Such leaks are more frequent after placement of transcatheter valves, however, they also occur with surgically implanted prosthetic valves [19]. Creation of a reliable PVL model required design modifications to ensure that the annular tissue could not seal tightly around the entire circumference of the valve. In particular, we modified the frame and suturing ring to obtain leaks at specific locations around the annulus. When using the wire pressing process for frame fabrication (Fig. 1A and B), these features were added manually after pressing and were comprised of rectangular, oval or crescent-shaped metal rings tungsten inert gas welded to the outer circumference of the frame (Fig. 3A). Alternatively, when using the 3D printing process, we included such features in the CAD design (Fig. 3B). In this design, each of the 3 PVLs were created using a pair of radial struts, which were partly hidden in the suturing ring (Fig. 3C and D). As shown for both designs, each leak has a corresponding interruption in the suturing ring to allow for regurgitant blood flow. The specific design can be selected based on study goals. For example, if image-based leak localization is being studied, the welded rings may provide unnatural and thus undesirable visualization cues.

Figure 3:

Creating paravalvular leaks. (A) Valve design using wire frame and welded rings to create gaps between the sewing ring and the native valve annulus. (B) 3D printed frame comprised bronze-infused stainless steel incorporating pairs of radial struts to create gaps between the sewing ring and the native valve annulus. (C) 3D printed frame covered with fabric. (D) 3D printed frame with interrupted sewing ring added. PVL: paravalvular leak.

Experimental validation

Ex vivo and invivo experiments were used to validate the 2 valve versions designed to produce PVLs. Exvivo experiments (Fig. 4) were carried out in porcine hearts (n = 3) obtained from approximately 80 kg animals providing a size representative of adult human hearts [20]. The aorta was incised, and the native aortic leaflets were removed. The valve was then positioned and sutured as shown in Fig. 4B, with 2–0 valve sutures preloaded with polytetrafluoroethylene pledgets (Ethibond Excel, Ethicon, Somerville, NJ, USA). After closing the aorta, it was connected to a fluid (saline) column with diastolic pressure of up to 120 cmH₂O (85 mmHg). Pulsatile flow was produced by gentle cyclic manual squeezing of the ventricles to simulate the systolic portion of the cardiac cycle (Fig. 4A). To assess valve function, visualization was provided by a rigid endoscope (Fig. 4C) along with transepicardial echocardiography.

Figure 4:

Ex vivo valve testing using porcine hearts. (A) Schematic; (B) valve implantation; and (C) insertion of endoscope for valve visualization.

Acute invivo experiments were also carried out in porcine models (n = 6, 73–80 kg). After entering the thoracic cavity via sternotomy and initiation of cardiopulmonary bypass, cardioplegia was induced using del Nido cardioplegia solution [21]. The aorta was incised and, after removing the native leaflets, the prosthetic valve was surgically sutured onto the native annulus using 2–0 Ethibond Excel pledgeted valve sutures. To ensure creation of PVLs, one suture was omitted from the aortic annulus at the midpoint of each aortic sinus. After implantation and closure of the aortic incision and removal of the aortic clamp, normal sinus rhythm was re-established with direct current cardioversion. Valve function was then assessed using intraoperative epicardial echocardiography. At the conclusion of the experiments, the animals were euthanized, and the hearts were recovered for postoperative ex vivo assessment following the same protocol as described earlier for the ex vivo experiments. Finally, gross structural assessment was performed and findings documented. Prior to in vivo experiments, the study protocol was reviewed and approved by Boston Children’s Hospital Institutional Animal Care and Use Committee.

RESULTS

Prior to use in experiments, each of the newly manufactured valves was inspected for quality. This process also allowed us to characterize and compare the designs. We found that metal frames made using the wire stamping method were more compliant and more prone to deformation during surgery than the 3D printed frames. Moreover, the 2-part stamp printed in plastic was found to wear at the contact of the metallic wire and had to be changed after making 5 valves. No problems were observed with the polyester fabric covering the frame. Furthermore, the suturing ring held sutures securely and remained tightly in place around the frame. The pericardial leaflets maintained their structural integrity and precision of coaptation through experiment completion.

Ex vivo experimental validation

Ex vivo experiments demonstrated that the valves are functional at physiological pressures. Figure 5A and B depicts aortic views of the valve at peak diastole and systole. Proper sealing of the valve at the end of diastole was confirmed visually as well as using echocardiography. Echocardiographic images (Fig. 5C and D) reveal that the function of our lab-fabricated valves is comparable to clinical prosthetic valves, with spectral and colour Doppler analysis showing no significant valvar stenosis and no more than mild valvar regurgitation detected in any valve implants.

Figure 5:

Bioprosthetic valve (no paravalvular leaks) in ex vivo porcine heart. (A and B) Endoscopic aortic views in water at end-diastole and end-systole; (C and D) corresponding 2D echocardiograms.

In vivo experimental validation

Both valve designs for PVL were successfully demonstrated in animal experiments. The 3D printed frame was preferred, however, since it was stiffer and consequently more effective at maintaining its shape during implantation. Figure 6A depicts a valve during implantation. The suturing ring was able to hold the suture and did not fail during the experiments as confirmed by postoperative inspection. While not evaluated in a training scenario, 2 senior cardiac surgeons (C.W.B. and J.E.M. Jr) indicated that the proposed valve was comparable to a clinical valve in terms of sizing, positioning in the annulus and forces needed for suturing.

Figure 6:

In vivo valve testing. (A) Implantation. Arrows indicate PVL locations. (B) 2D echocardiogram with corresponding colour Doppler showing 2 regurgitant jets from PVLs. PVL: paravalvular leak.

In order to validate the capability of the 2 designs (Fig. 3A and D) to produce PVLs, colour Doppler imaging was used to detect regurgitant jets (Fig. 6B). All intended leaks exhibited jets, and there were no additional PVLs. Furthermore, regurgitant jets between the leaflets were either not present or were very mild.

All animals went through a complete experimental procedure, including cardiopulmonary bypass, aortic valve replacement and valve assessment. No adverse events occurred, other than 2 cases of reversible tachyarrhythmia.

After euthanization, the heart was recovered and examined endoscopically in saline, as described above. Figure 7A–D provides endoscopic views of the entire valve as well as close-up views of PVLs at manually induced diastole and systole. These images show that the valve is functioning as intended with coaptation of the leaflets and channels located at the desired PVL locations. Figure 7E–G shows views of a valve with the 3D-printed design (Fig. 3D) after explant. The 3 leaks created by the gaps between the radial struts on the valve frame are clearly visible on both the aortic (Fig. 7E) and the ventricular sides (Fig. 7F and G). On average, the valves were maintained inside the beating hearts for 3 h before animal killing. During this time, the valves remained free of signs of thrombus formation on either the leaflets or the frame, although animals were fully heparinized due to cardiopulmonary bypass. Additionally, there were no gross signs of reaction to the valve and its materials. Close evaluation of metal frames of the valve showed no gross changes related to their implantation in living organisms, including signs of rusting or corrosion.

Figure 7:

Postoperative assessment of explanted heart filled with saline. (A) Aortic view of closed valve; (B) aortic view of open valve; (C) PVL viewed from aorta; (D) PVL viewed from ventricle; (E) examination of aortic side after explant; and (F and G) examination of ventricular side after explant. Blue arrows indicate the PVLs. PVL: paravalvular leak.

DISCUSSION

This paper presents the design, fabrication and experimental validation of simple, low-cost bioprosthetic semilunar heart valves. By describing the fabrication process in detail and providing tutorial videos and CAD files, other researchers will be able to quickly replicate our designs as well as to adapt them for their particular studies. While valve modifications for PVL were demonstrated here, there are many valve pathologies for which inexpensive reproducible models are needed. We believe that the techniques described here to custom-design valves with embedded defects using simple manufacturing processes will inspire new possibilities and expedite progress in valvar heart disease research and treatment.

In addition to research applications, the proposed valve could be used for surgical training, which is one of the fastest developing fields of surgical education. Current valve surgery training is performed on simulators, large-animal models and, ultimately, on patients under the supervision of an experienced surgeon. Although there are a number of simple and inexpensive training simulators, they are intended to facilitate the development of basic skills and manipulations involving heart valves and are not able to replicate the exact experience that comes with performing an actual valve replacement on a living heart. Our hope is that the valve design and fabrication recipe provided here will enable improved training opportunities for valve replacement or valve repair procedures using simulators, ex vivo models and in vivo large-animal models.

SUPPLEMENTARY MATERIAL

Supplementary material is available at ICVTS online.

Funding

This work was supported by the National Institutes of Health under grant (R01HL124020 to Pierre E. Dupont).

Conflict of interest: none declared.