Original Contribution

Intact and Isolated Human Cadaveric Coronary Artery Perfusion Models to Facilitate Research and Education Regarding Coronary Anatomy and Pathology

John L. Petersen, MD, MHS1;  Dmitry B. Levin, BS2;  Vivian W. Hou, MS2;  Carol V. Hagen, CDT1;  Grant Anderson, BS3;  Joanie Block, BS, MA4;  Mark Reisman, MD2

John L. Petersen, MD, MHS1;  Dmitry B. Levin, BS2;  Vivian W. Hou, MS2;  Carol V. Hagen, CDT1;  Grant Anderson, BS3;  Joanie Block, BS, MA4;  Mark Reisman, MD2

Abstract: Cadaveric tissue-perfusion models are well established in the fields of structural heart and peripheral vascular disease; however, less consideration has been given toward coronary artery disease despite comparable prevalence and morbidity. Two tissue-perfusion models were developed to address this need. The first, an intact heart model, allows simulation of percutaneous coronary interventional procedures. The second focuses upon isolated arteries, allowing quantification of simulated procedures. Both models were applied for clinical training and for investigations into medical device behavior. The manner of preparation facilitates access to clinically relevant disease, thus providing a platform to further research on coronary artery disease.

J INVASIVE CARDIOL 2019;31(9):272-277. Epub 2019 June 15.

Key words: atherosclerosis, cadaveric heart, coronary artery, medical simulation, perfusion model

Despite significant improvement in medical imaging and catheter-based management, coronary artery disease (CAD) remains a leading cause of death, hospitalization, impaired quality of life, and health-care expenditure.1,2 This rift stems in part from the challenges associated with studying critical anatomical structures; deeper examination of disease progression is inherently limited by the degree to which such structures can be accessed. Cadaveric models remain a critical bridge in both clinical translation and medical education.

Although tissue models are well established with regard to other diseases of significant morbidity and mortality, human cadaveric models for CAD have remained under-developed. Focus has been directed toward aortic perfusion3-6 or general vascular perfusion for surgical simulation.7,8 In instances where the coronaries are perfused, the emphasis is upon myocardial supply rather than the disease state of the coronaries themselves.9,10 Perfusion of a coronary segment is not novel in and of itself;11 however, only a few published efforts have developed these setups into reliably reproducible and quantifiable platforms for CAD-related investigations.12,13

We have developed and applied two human cadaveric coronary perfusion models. The first utilizes an intact heart in which the in situ orientation of the coronary arteries are preserved. The second model utilizes isolated segments of coronary arteries and allows for co-registration of simulated imaging and procedural treatments. Herein, we describe the techniques to create the models and present representative images generated from their respective use.


Cadaveric heart tissue. Anonymized cadaveric hearts were procured from accredited vendors. Tissue was used in accordance with the consented purpose of the tissue and under the contracted vendor agreement. Hearts were frozen by the vendor following acquisition and shipped as such. After approved use, the tissue was disposed with an accredited tissue disposal contractor.

Intact Heart Model

Dissection and tissue preparation. Prior to dissection, hearts were thawed at room temperature and rinsed with saline to remove clot and post mortem debris from the internal and external surfaces. Hearts were grossly prepared by incising and removing the pericardium and pericardial fat. Residual pulmonary and mediastinal tissue on the aorta, posterior wall of the left atrium, and the pulmonary veins were also removed. The pulmonary veins were then oversewn with suture.

Silastic tubing sized to the vessel was used to cannulate the descending aorta, followed by the introduction of a laparoscopic trocar into the tubing. The trocar and tubing were secured to the aorta using zip ties. Next, the brachiocephalic and left subclavian arteries were cannulated with vascular sheaths; if under-sized for cannulation, the left carotid was tied off. Finally, the prepared specimen was placed upon a ring stand within a custom-designed, three-dimensional (3D)-printed construct in a supine anatomic orientation (Figure 1).

Tissue model perfusion settings. Inflow perfusion was delivered via the sheathed innominate artery by joining the sideport to a perfusion pump (model 1423 Harvard Apparatus). The cannulated distal aorta and coronary sinus were then joined to a passive drain. Standard 0.9% saline was chosen as perfusate. A stroke volume of 20 mL was delivered at a rate of 30 cycles/min. Hearts were carefully monitored to avoid excessive filling.

Use of intact heart perfusion model. Coronary angiography and angioplasty devices were advanced into the aorta via the trocar. Angiography was performed by guiding the catheter to engage with the coronary ostium under fluoroscopic observation (Ziehm Solo; Ziehm Imaging). Radiographic contrast dye was introduced via catheter and the perfusion pump was allowed to run until dye clearance in the coronary was achieved. Typical angiographic views of the right and left coronaries were obtained through right, left, caudal, and cranial angulations. Interventional coronary procedures were simulated in accordance with standard clinical workflow. Following guidewire advancement, catheters, wires, and other angioplasty devices were exchanged according to the simulated procedure.

Isolated Coronary Artery Model

Dissection and tissue preparation. Initial preparation mirrored the intact heart model: hearts were thawed, rinsed, and incised. The heart was then prepared in the form of blocks approximately 75 mm in length. Each contained the main path of a single coronary artery. Anatomical landmarks served as guides for standardizing segment preparation. First, the right and left ostia were identified along with the proximal lengths to estimate the general arterial path. The bifurcation of the left main into the left anterior descending (LAD) and left circumflex (LCX) was identified. Ligation of the LCX at the bifurcation was performed to maximize flow within the LAD. The left side of the heart was then separated from the right, with care given to avoid cutting through any distal regions of the three main coronary paths. Further resections were performed to remove extraneous tissue, resulting in the 75 mm blocks described above. A block containing only the LAD was created by truncating the LCX distal to the ligation; this point of separation then served as the new access site for the LCX. The right coronary artery was truncated prior to giving rise to the peripheral descending artery.

Luer-lockable cannulas sized to each ostia were tied onto the proximal ends of each coronary to provide access. Saline perfusion was used to identify the distal ends of significant arterial branches for ligation to promote flow down the main coronary path. A hemostatic valve was locked onto the proximal cannula. Finally, a standard 0.014˝ angioplasty wire was advanced out the distal end of the artery and the prepared tissue was placed into a custom-designed cage (Figure 2). The cage contained cutouts that provided anchor points for the cannulas as well as a holder for a radiographic ruler. Simultaneous perfusion and access to the artery were achieved through the hemostatic valve. A two-way stopcock line was joined to the sideport of the hemostatic valve. One port was dedicated for contrast injection and the other attached to a line fed by a perfusion pump.

Tissue model perfusion settings. Inflow of 0.9% saline was delivered via a perfusion pump (FlowTek125; United Biologics) at a stroke volume of 1 mL at 60 cycles/min. Prior to any simulated procedures, vessels were primed with approximately 5 seconds of flow.

Use of isolated coronary artery model. Isolated artery angiograms were obtained by hand-injected images under fluoroscopy. Interventional coronary procedures and imaging were performed by advancing devices over the guidewire through the hemostatic valve. The valve was gently closed down to minimize loss during perfusion, with care given to avoid crimping damage to any devices that may extend through the valve (eg, imaging catheters).


Intact heart model. Representative examples of right and left coronary angiograms are depicted in Figure 3. Figure 4 depicts a simulated angioplasty procedure (guidewire advancement, stent positioning, balloon inflation, and postdeployment angiogram).

Isolated coronary artery model. Figure 5A depicts a representative angiogram. Angioscopy, intravascular ultrasound, and intravascular optical coherence tomography images (Figures 5B-5D) have also been successfully generated using this model. Simulated coronary angioplasty procedures are demonstrated through fluoroscopic observation (Figure 6).


Application of cadaveric perfusion models has advanced disease understanding and led to improvements in device therapies, particularly in the structural heart and peripheral vascular fields.3-8 In comparison, the use of cadaveric models to facilitate understanding of CAD has been more limited.12,13 The work described herein showcases two perfusion models focused toward CAD that have been utilized for medical education and device development.

The strength of the intact heart model is the simulation of catheter-based procedures. Angiograms generated by the model (Figure 3) visualize coronary features and are similar to conventional angiographic views obtained in clinical practice. Importantly, contrast clearance is achieved between angiographic runs, allowing the acquisition of multiple angles and assessment of arteries before and after manipulation. The anatomic challenges of catheter-based intervention and difficult angioplasty procedures can be reproduced; notably, variables that affect engagement of the guiding catheter are reproduced. We have utilized the intact heart perfusion model to test a novel angioscopy device for forward-viewing evaluation of chronic total occlusions and vitally contributed toward commercialization efforts.14 A limitation of the intact model is co-registration of the anatomy with imaging and histology, as further processing requires dissection of the prepared specimen, which may distort the native anatomy.

The isolated artery model is more appropriate for studies in which the variables of guide-catheter engagement and clinical angiography are less important to simulate. Accessibility is particularly advantageous when investigating complex CAD; the manner of preparation imparts rapid and consistent device delivery. Development of a custom cage has greatly improved the ability to quantify simulated procedure outcomes, as anatomic landmarks can be co-registered to a radiographic ruler. Finally, the isolated artery model lends itself to histologic analysis by virtue of co-registration and focal sectioning of tissue. We have used the isolated artery model to assess novel devices, to provide imaging training for interventional cardiologists, and to investigate atherectomy device mechanisms.

Advancements have been made in whole-body perfusion to allow simulation of large vascular procedures including aortic endograft placement and transcatheter aortic valve replacement. Garrett et al described a total-body arterial perfusion model in 2001 for simulation of vascular surgery.6 Subsequent aortic models have been developed for medical education, including a combined arterial and venous perfusion model that allows simulation of complex intravascular procedures.3-5,7,8 Isolated coronary models have been used for device development and regulatory approval.12,13 For example, such cadaveric tissue models were used to validate a near-infrared imaging technique to detect lipid-rich plaque; the resulting data were critical in the approval process.

While animal models of atherosclerosis have provided significant insight into molecular mechanisms, they often rely upon genetic alterations to accelerate and exacerbate disease development. The disease developed in these models is generally atypical of atherosclerosis encountered in clinical practice. In humans, atherosclerosis develops over decades as the result of multiple contributing processes; these conditions are often unethical and expensive to replicate in animal models.14,15 Of particular note, there are currently no reliable animal models that present the heavily calcified arteries typical of advanced CAD, a condition that is treated with percutaneous intervention. As donors inherently possess a breadth of disease severity, in this regard cadaveric models provide an additional valuable perspective for furthering disease understanding.

Future development of the models will aim to further understanding of complex CAD, including chronic total occlusions, bifurcation lesions, and heavy arterial calcification. While most of our applications have used previously frozen tissue, expanded access to fresh tissue allows studies in cellular and molecular aspects of human atherosclerosis. Finally, the models may allow for improvements in bench testing, which may reduce some of the need for animal and human research studies as a part of device development.


We have successfully developed and applied two cadaveric coronary artery perfusion models: (1) an intact heart model that simulates typical angiograms and catheter-based procedures; and (2) an isolated coronary artery model that allows for localization and correlation of angiographic and invasive imaging findings. These models have been successfully used for medical education and provide a platform for advanced CAD research.

Acknowledgment. The authors would like to acknowledge Mr. S.B. Joseph Clark and the Swedish Medical Center Foundation for their generous support of this ongoing work. This work was also supported by an ongoing NIH effort (SBIR R43HL139323).


1. Rodríguez T, Malvezzi M, Chatenoud L, et al. Trends in mortality from coronary heart and cerebrovascular diseases in the Americas: 1970-2000. Heart. 2006;92:453-460.

2. Benjamin EJ, Blaha MJ, Chiuve SE, et al. Heart disease and stroke statistics-2017 update: a report from the American Heart Association. Circulation. 2017;135:e146-e603.

3. Chevallier C, Willaert W, Kawa E, et al. Postmortem circulation: a new model for testing endovascular devices and training clinicians in their use. Clin Anat. 2014;27:556-562.

4. Arbatli H, Cikirikcioglu M, Pektok E, et al. Dynamic human cadaver model for testing the feasibility of new endovascular techniques and tools. Ann Vasc Surg. 2010;24:419-422.

5. Boivie P, Hansson M, Engström KG. Embolic material generated by multiple aortic crossclamping: a perfusion model with human cadaveric aorta. J Thorac Cardiovasc Surg. 2003;125:1451-1460.

6. Garrett HE. A human cadaveric circulation model. J Vasc Surg. 2001;33:1128-1130.

7. Delpech PO, Danion J, Oriot D, Richer JP, Breque C, Faure JP. SimLife a new model of simulation using a pulsated revascularized and reventilated cadaver for surgical education. J Vasc Surg. 2017;154:15-20.

8. Carey JN, Minneti M, Leland HA, Demetriades D, Talving P. Perfused fresh cadavers: method for application to surgical simulation. Am J Surg. 2015;210:179-187.

9. Fedorov VV, Glukhov AV, Chang, R, et al. Optical mapping of the isolated coronary-perfused human sinus node. J Am Coll Cardiol. 2010;56:1386-1394.

10. Fedorov VV, Glukhov AV, Ambrosi CM, et al. Effects of KATP channel openers diazoxide and pinacidil in coronary-perfused atria and ventricles from failing and non-failing human hearts. J Mol Cell Cardiol. 2011;51:215-225.

11. Rosenthal E, Montarello JK, Palmer T, Curry PV. Thermal effects of stationary “hot tip” laser coronary probes: an in vitro assessment. Lasers Surg Med. 1989;9:229-236.

12. Gardner CM, Tan H, Hull EL, et al. Detection of lipid core coronary plaques in autopsy specimens with a novel catheter-based near-infrared spectroscopy system. JACC Cardiovasc Imaging. 2008;1:638-648.

13. Kang SJ, Mintz GS, Pu J, et al. Combined IVUS and NIRS detection of fibroatheromas: histopathological validation in human coronary arteries. JACC Cardiovasc Imaging. 2015;8:184-194.

14. Hou VW, Levin DB, Hagen C, Lombardi WL, Seibel E, Petersen J. High-resolution angioscopy of proximal chronic total occlusion microchannels using the scanning fiber endoscope. J Am Coll Cardiol. 2018;71:A1051.

15. Getz GS, Reardon CA. Animal models of atherosclerosis. Arterioscler Thromb Vasc Biol. 2012;32:1104-1115.

16. Shim J, Al-Mashhadi RH, Sørensen CB, Bentzon JF. Large animal models of atherosclerosis — new tools for persistent problems in cardiovascular medicine. J Pathol. 2016;238:257-266.

From the 1Swedish Medical Center, Swedish Heart and Vascular Institute, Division of Cardiology, Seattle, Washington; 2Division of Cardiology, University of Washington Medical Center, Seattle, Washington; and 3LifeCenter Northwest, Bellevue, Washington.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Petersen reports an education grant from Abbott Vascular; speaker’s bureau income from Cardiovascular Systems, Inc.; and personal fees from Veravanti as a subinvestigator for SBIR. The remaining authors report no conflicts of interest regarding the content herein.

The authors report that patient consent was provided for publication of the images used herein.

Manuscript submitted February 15, 2019, accepted February 26, 2019.

Address for correspondence: Mark Reisman, MD, FACC, Clinical Professor of Medicine, Section Head, Interventional Cardiology, Director, Cardiovascular Emerging Therapies University of Washington Medical Center, Auth Endowed Chair in Cardiovascular Innovation, University of Washington, 1959 NE Pacific Street, Box 356422, Seattle, WA 98195. Email: reismanm@uw.edu