Multifunctional passive-heart platforms for in vitro hemodynamic studies

Background and aim of the work The present work describes the design, development and application of innovative in vitro passive-heart platforms for the hemodynamic and biomechanical analysis of therapeutic approaches to cardiovascular pathologies. This work arises from two main collaborations: the first part of the work was carried out at FoRCardioLab, a laboratory co-founded by Politecnico di Milano, Universitá degli studi di Milano and the division of Cardiac Surgery in Luigi Sacco Hospital, in which surgeons and bioengineers cooperate to study new therapeutic approaches to cardiac pathologies. The work described in the second part of the thesis arises from the collaboration between Politecnico di Milano and the Cardiovascular Biomechanics Laboratory of the TU/e, Eindhoven (The Netherlands), where part of the research was performed under the supervision of Prof. van de Vosse and Prof. Rutten. The use of in vitro mock circulation loops for the study of the cardiovascular system and the test of prosthetic devices dates back to the 70’s [1,2] and finds its rationale in the possibility of performing tests in highly controllable experimental conditions and in a cost-effective fashion, thus reducing the need for animal models. In order to represent realistic models, in vitro platforms must be able to replicate the conditions to which the device will be subjected in vivo. Being historically associated with the development of mechanical heart valves, the classical approach to the design of mock circulatory systems was purely hydrodynamic-based. In recent years, however, the need for realistic in vitro models has become more stringent due to the substantial changes of the clinical approaches towards reparative, minimally-invasive and transcatheter techniques [3–6]. For most of such applications, the interaction between implanted device or repaired structure and in vivo environment is not strictly limited to the hemodynamics, but involves anatomical and functional aspects that are crucial for the outcome of the procedure. Paravalvular leakage in transcatheter aortic valve implantation (TAVI) [7,8], complex aortic-mitral interactions following surgical or transcatheter procedures [9,10] and aortic valve insufficiency after continuous flow left ventricular assist device (cf-LVAD) implantation [11,12] are only few examples of these important interplays. Hence, being able to take into account these aspects is nowadays a challenging, yet fundamental, requirement for any modern in vitro mock circulatory loop. Recently, researchers addressed this issue and tried to fill the morphological and anatomical gap with both ex vivo and animal models by developing mock loops in which excised aortic [13–15] or mitral [16,17] valves could be tested. These systems permit both the imaging of valvular structures with high-speed cameras and/or echographic techniques and the execution of surgical procedures on the biological samples. In this way, surgeons can operate in a very familiar environment, where they can simulate surgical procedures and directly analyse their hemodynamic effects. An interesting approach has been recently proposed by Richards et al. [18], who developed a dynamic passive-heart platform for the analysis of mitral valve repair techniques. The system was able to host an entire explanted heart and to provide the left ventricle with a cyclical flow rate thanks to a pulsatile pump connected to the apex. Given the very specific purpose for which the platform was designed, the hemodynamic conditions were far from being physiological. Nonetheless, the passive-heart approach represents a challenging and fascinating development of the excised valves platforms, towards a better replication of the in vivo anatomy and morphology. Indeed, the use of an entire heart allows a perfect preservation of the cardiac structures, still without involving all the complexities and ethical issues of the ex vivo and animal models. Furthermore, the use of entire hearts greatly widen the potential applications of an experimental apparatus by allowing, for example, the simulation of minimally invasive and multi-valvular surgical procedures. The aim of this study is to develop novel in vitro passive-heart platforms that can represent effective tools for research, device testing and training purposes and reduce the need for ex vivo and animal models. The present work describes the development of two passive-heart platforms, which adopted different methodologies for the generation of the cardiac output, and their use for TAVI and LVAD applications. Also, the design of a 4-chamber mock loop, conceived in order to work with both passive or ex vivo beating hearts, is described. To this purpose, the present thesis is structured as follows: • In the first part of the work, the development of an in vitro platform able to house an entire explanted porcine heart and subject it to pulsatile hemodynamic conditions, is described. The system, similar to that proposed by Richards and colleagues [18], dynamically pressurizes the left ventricle by mean of a piston pump connected to the heart apex, and enables the hemodynamic analysis of simulated surgical procedures and the imaging of the valvular structures. The mock loop’s hydrodynamic design was based on an ad-hoc defined lumped-parameter model. • The evolution of the in vitro passive beating heart platform for TAVI applications was successively addressed. The development included both a feasibility study on a novel methodology to reproduce in vitro aortic valve stenosis, and the redesign of the setup layout and components in order to achieve an easy and cost-effective training platform. The proficiency of the system was assessed by performing transcatheter aortic valve and valve-in-valve implantations under multimodal imaging guidance in a catheterization lab. • The research that was carried out at the Cardiovascular Biomechanics Laboratory of the TU/e led to the development of another in vitro passive-heart mock loop. This setup adopted a complementary functioning principle as compared to the first one that was designed. Indeed, this left heart platform mimics the pulsatile pumping function of the heart through the external pressurization of the ventricles, thus trying to reproduce the in vivo dynamic ventricular behavior. The system was also used for a pilot study aimed at analyzing the aortic valve (AV) opening dynamics during continuous-flow left ventricular assist device (cf-LVAD) support. • The final part of the work describes the redesign of an existing model-controlled four-chamber mock loop, that could be used with both passive and beating heart models. An anatomical study was conducted in order to characterize the orientation of the main heart vessels, and to design a standard manifold for their connection to the hydraulic circuit. Preliminary assessments are being performed using passive hearts. In the last Chapter, a general discussion of the PhD dissertation and the conclusive remarks are presented.   3. In vitro hemodynamics and valve imaging in passive beating hearts In this Chapter, the development of a mock apparatus able to house an entire explanted porcine heart and subject it to pulsatile fluid-dynamic conditions is presented, in order to enable the hemodynamic analysis of simulated surgical procedures and the imaging of the valvular structures. The mock loop’s hydrodynamic design was based on an ad-hoc defined lumped-parameter model. The left ventricle of an entire swine heart was dynamically pressurized by an external computer-controlled pulse duplicator. The ascending aorta was connected to a hydraulic circuit, which simulated the input impedance of the systemic circulation; a reservoir passively filled the left atrium. Accesses for endoscopic imaging were located in the apex of the left ventricle and in the aortic root. The experimental pressure and flow tracings were comparable with the typical in vivo curves; a mean flow of 3.5±0.1 lpm and a mean arterial pressure of 101±2 mmHg was obtained. High-quality echographic and endoscopic video recordings demonstrated the excellent potential of the developed system in the observation of the cardiac structures dynamics. The proposed mock loop represents a suitable in vitro system for the testing of minimally invasive cardiovascular devices and surgical procedures for heart valve repair. 4. Evolution of the passive beating heart platform for TAVI applications In this Chapter, the evolution towards TAVI applications of the passive beating heart platform is presented. Our goal was to develop a platform capable of simulating the typical environment physicians are working with when performing TAVI, in order to provide them with a multi-functional tool that could be used for training, research and educational purposes. The key requirements for our design were the possibility of performing multimodal imaging, the simulation of the anatomo-morphological environment and the achievement of physiologic hemodynamic conditions in a simple, reliable and cost-effective system. In order to achieve these goals, the passive beating heart setup was optimized with improved TAVI-specific design solutions. Moreover, a methodology to simulate in vitro aortic valve stenosis by glueing human aortic calcified leaflets to the healthy porcine aortic cusps, was developed. The experimental assessments proved the reliability of the proposed methodology, which allowed clear visualization of the stenotic valve during fluoroscopy and induced mild to severe degrees of aortic stenosis. The passive beating heart platform, with the model of aortic stenosis, was profitably used to perform several TAVI procedures under fluoroscopic guidance and simultaneous intracardiac visualization in a catheterization lab. 5. A novel passive left heart platform for device testing and research In this Chapter, the development of a novel in vitro left heart platform capable of simulating the pulsatile pumping function of the heart through the external cyclic pressurization of the ventricular walls is described. The system is composed by a fluid-filled chamber, in which the ventricles of the heart are housed and sealed by means of a rapid-prototyped vacuum seal, which excludes the atria from any external load. The fluid-filled chamber is connected to a piston pump, whose cyclic action drives the motion of the ventricular walls. The aorta is connected to a mock circulatory system simulating the human systemic impedance, and the left atrium is fed by an adjustable preload. The platform was capable of reproducing physiologic hemodynamic conditions, i.e. aortic pressures of 120/80 mmHg with 4.5 lpm of cardiac output, and allowed for endoscopic imaging of the cardiac structures. The potential of the system for device-testing and visualization studies was also assessed, with a continuous flow LVAD connected to the heart, so to investigate the effects induced on the aortic valve function by different levels of support. Results were in line with clinical observations and previous studies, showing an increased load on the valve for increasing pump speeds, as well as a reduced valve duty cycle. High-speed video recordings of the aortic valve also allowed the visualization of the transition between a fully opening valve and a permanently closed configuration, which only happened within a very narrow pump speed range. In conclusion, the system showed to be an effective tool for the hemodynamic assessment of devices, the simulation of surgical or transcatheter procedures and for visualization studies. 6. Towards the development of four-chamber passive and beating heart platforms This Chapter describes the redesign of an existing 4-chamber purely hydraulic model-controlled mock circulatory loop, that was developed at the Cardiovascular Biomechanics Laboratory of the TU/e and features a complete 4-chamber circulation, whose actuation is controlled by a heart contraction model and a heart rate control model. Our challenging goal was to redesign the system in order to make it capable of working in a 4-chamber mode with entire heart structures, being either passive hearts actuated by model-controlled pumps, or isolated beating hearts. The first design step consisted in the characterization of the spatial orientation of the main afferent and efferent heart vessels, so to design a standard connection with the afterload circuit. These measures were used as inputs to redesign the afterload circuit accordingly to the heart layout, taking into account also the requirements related to the performance of isolated beating heart experiments, e.g. different working modes, blood oxygenation, venting. The setup was then manufactured and assembled. The first experimental assessments are currently being performed with passive hearts, investigating the feasibility of the approach in which external cyclic pressurization is applied to the ventricular walls in a 4-chamber working mode. With this respect, the main issue that is being addressed is represented by the different compliance of the right and left ventricles, which leads to relevant abnormalities in the generation of the stroke volumes. In order to reduce this gap, we are currently investigating several ways to stiffen the right ventricle using both classical chemical fixation methodologies and external mechanical stiffening of the walls. 7. Main findings and future developments The present work explored the potentiality of in vitro passive-heart mock circulatory systems as multi-functional in vitro platforms for research, device-testing, visualization studies, training and educational purposes in the cardiovascular field. Our findings demonstrated the feasibility of this innovative approach, as the developed passive-heart systems were capable of i) closely reproducing the physiologic hemodynamics, ii) allowing intracardiac endoscopy and multimodal imaging, iii) performing surgical and/or transcatheter procedures, iv) assessing cardiovascular devices and v) being used as training platforms. Two platforms were designed, developed and tested in order to assess two alternative approaches for the generation of the stroke volume. The first system, whose design principle was inspired by the worked of Richards et al. [18], pressurized the left ventricle internally by means of a piston pump connected to the heart apex. This actuation methodology ensured the achievement of physiological hemodynamic conditions, excellent imaging capabilities, and no abnormalities in valve function, though it caused a paradoxical motion of the ventricular walls during the cardiac cycle and an altered fluid dynamic field inside the left ventricle. The platform was optimized for TAVI applications, and used in a catheterization lab to perform implantations under fluoroscopic guidance and simultaneous intracardiac visualization on an in vitro model of aortic valve stenosis, that was developed to mimic the real pathological scenario. The second passive-heart mock loop was designed to mimic the pulsatile pumping function of the left heart through the external dynamic pressurization of the ventricular walls, and ensured a better simulation of the dynamic behaviour of the ventricular walls. Physiological hemodynamics were achieved, although mitral valve prolapse was observed in some samples at high stroke volumes, due to the absence of papillary muscle contraction. The potential of the developed system as a platform for device testing was assessed with a pilot study, in which the acute post-operative scenario after the implantation of a cf-LVAD was simulated, and the AV function for different levels of mechanical support was analysed. Future developments, whose initial steps were described in the last Chapter of this dissertation, will face two relevant, yet unavoidable, challenges. On one side, the development of four chamber mock loops, so to broaden the potential applications of these systems and to allow the study of the interactions between left and right circulation, which are often fundamental. On the other side, achieving a model-controlled actuation of passive-heart platforms may definitely close the existing gap between this approach and the state-of-the-art in vitro hybrid loops. With respect to both these directions, the internal pressurization approach may represent a more sensible choice both to develop a full circulatory loop and to implement a model-controlled actuation.

L’obiettivo del progetto di Dottorato è sviluppare apparati sperimentali per la valutazione biomeccanica in vitro di differenti soluzioni terapeutiche per il trattamento di patologie valvolari cardiache. Negli ultimi anni si è assistito ad un crescente interesse verso tipologie di intervento mini-invasive in grado di preservare la struttura anatomo-funzionale fisiologica. In questo panorama risulta fondamentale, sia per il chirurgo sia per il progettista, disporre di apparati sperimentali che permettano di comprendere, in modo ripetibile e controllabile, gli effetti indotti da una protesi o da una tecnica chirurgica sulla biomeccanica valvolare. In particolare, l’integrazione delle strutture cardiache all’interno dei classici sistemi in vitro risulta essere indispensabile al fine di effettuare un’indagine completa e realistica e di fornire un ambiente familiare al clinico. Solo alcuni lavori sono stati pubblicati in questo ambito di ricerca, che sta acquisendo un notevole interesse a livello mondiale. A questo proposito, alcuni autori hanno proposto sistemi in grado di utilizzare cuori interi espiantati, ripristinandone la contrattilità mediante perfusione del miocardio con soluzioni cristalloidi (de Hart et al., 2011). Questi sistemi, nonostante l’estrema realisticità nella riproduzione delle condizioni fisiologiche, risultano estremamente complessi, costosi e difficilmente controllabili. Un approccio innovativo è stato proposto recentemente da Richards et al. (2009), che ha sviluppato un simulatore in cui la pressurizzazione dinamica del ventricolo sinistro viene ottenuta mediante un sistema pompante esterno, connesso al cuore a livello apicale. Questa soluzione permette una notevole semplificazione dei protocolli sperimentali ed una maggiore ripetibilità delle condizioni fluidodinamiche, pur mantenendo una buona aderenza morfologica ed anatomica. Nell’ambito del lavoro di Tesi di Laurea, è stata sviluppata una versione prototipale di un banco prova in grado di ospitare cuori porcini isolati ed azionarli secondo l’approccio proposto da Richards e colleghi. L’attività di ricerca è stata quindi volta (i) in primo luogo allo sviluppo ed all’ottimizzazione del banco prova prototipale esistente (attività svolte nel primo anno); (ii) quindi all’utilizzo del simulatore per l’analisi fluidodinamica di valvole aortiche trans-catetere, che rappresentano l’attuale applicazione di frontiera nel settore (attività svolta nel secondo anno). (iii) Parallelamente, grazie al periodo svolto presso la Technische Universiteit Eindhoven (TU/e), è stato riprogettando un banco prova già esistente, che dispone di un complesso sistema di retroazione in grado di simulare la legge di Frank-Starling, al fine di integrare nel sistema stesso un cuore intero espiantato. Infine, sempre nell’ambito della collaborazione con la TU/e, (iv) è stata sviluppata una nuova piattaforma in vitro in grado di ospitare un cuore intero passivo ed indurne la contrazione attraverso la pressurizzazione esterna dei ventricoli. Questa piattaforma è stata utilizzata anche per una campagna preliminare volta a studiare la dinamica della valvola aortica in cuori supportati tramite dispositivi di assistenza ventricolare a flusso continuo.