Sjoerd van Tuijl

An Ex Vivo Platform to Simulate Cardiac Physiology: A New Dimension for Therapy Development and Assessment:

Purpose Cardiac research and development of therapies and devices is being done with in silico models, using computer simulations, in vitro models, for example using pulse duplicators or in vivo models using animal models. These platforms, however, still show essential gaps in the study of comprehensive cardiac mechanics, hemodynamics, and device interaction. The PhysioHeart platform was developed to overcome these gaps by the ability to study cardiac hemodynamic functioning and device interaction ex vivo under in vivo conditions. Methods Slaughterhouse pig hearts (420 ± 30 g) were used for their morphological and physiological similarities to human hearts. Hearts were arrested, isolated and transported similar to transplantation protocols. After preparation, the hearts were connected to a special circulatory system that has been engineered using physical and medical principles. Through coronary reperfusion and controlled cardiac loading, physiological cardiac performance was achieved while hemodynamic parameters were continuously monitored. Results Normal cardiac hemodynamic performance was achieved both qualitatively, in terms of pulse waveforms, and quantitatively, in terms of average cardiac output (4 l/min) and pressures (110/75 mmHg). Cardiac performance was controlled and kept at normal levels for up to 4 hours, with only minor deterioration of hemodynamic performance. Conclusions With the PhysioHeart platform we were able to reproduce normal physiological cardiac conditions ex vivo. The platform enables us to study, under different but controlled physiological conditions, form, function, and device interaction through monitoring of performance parameters and intra-cardiac visualization. Although the platform has been used for pig hearts, application of the underlying physical and engineering principles to physiologically comparable hearts from different origin is rather straightforward.

Autoregulation of Coronary Blood Flow in the Isolated Beating Pig Heart

The isolated beating pig heart model is an accessible platform to investigate the coronary circulation in its truly morphological and physiological state, whereas its use is beneficial from a time, cost, and ethical perspective. However, whether the coronary autoregulation is still intact is not known. Here, we study the autoregulation of coronary blood flow in the working isolated pig heart in response to brief occlusions of the coronary artery, to step-wise changes in left ventricular loading conditions and contractile states, and to pharmacologic vasodilating stimuli. Six slaughterhouse pig hearts (473 ± 40 g) were isolated, prepared, and connected to an external circulatory system. Through coronary reperfusion and controlled cardiac loading, physiological cardiac performance was achieved. After release of a coronary occlusion, coronary blood flow rose rapidly to an equal (maximum) level as the flow during control beats, independent of the duration of occlusion. Moreover, a linear relation was found between coronary blood flow and coronary driving pressure for a wide variation of preload, afterload, and contractility. In addition, intracoronary administration of papaverine did not yield a transient increase in blood flow indicating the presence of maximum coronary hyperemia. Together, this indicates that the coronary circulation in the isolated beating pig heart is in a permanent state of maximum hyperemia. This makes the model excellently suitable for testing and validating cardiovascular devices (i.e., heart valves, stent grafts, and ventricular assist devices) under well-controlled circumstances, whereas it decreases the necessity of sacrificing large mammalians for performing classical animal experiments.

Assessment of aortic valve pressure overload and leaflet functions in an ex vivo beating heart loaded with a continuous flow cardiac assist device

OBJECTIVES Aortic valve regurgitation, fusion and thrombosis are commonly reported clinical complications after continuous flow ventricular assist device implantations; however, the complex interaction between reduced pulsatile flow physiology and aortic valve functions has not been studied experimentally. To address this, a continuous flow left ventricular assist device was implanted in four swine ex vivo beating hearts and then operated at baseline (device off, no flow) and at device speeds ranging between 8500 and 11,500 rpm under healthy and experimentally created failing heart conditions. METHODS At baseline and after each speed increase, aortic, left ventricular, left atrial and pulse pressure signals were monitored to assess the haemodynamic status of the ex vivo heart, aortic valve opening time and the transvalvular pressure changes. Aortic root and device flows were recorded with flow probes. Left ventricular pressure-volume loops were measured with a conductance catheter. Changes in aortic leaflet motion and end-diastolic aortic root diameter were recorded with epicardial echocardiography. RESULTS A two-chamber healthy and failing ex vivo beating heart model was successfully created. At increasing device flows, aortic valve open time steadily decreased from 36±7% of the baseline cardiac cycle to 0% at 11,500 rpm in the healthy heart and from 18±16 to 0% in failing heart mode (P<0.05). Aortic transvalvular pressure increased from 25±5 mmHg (baseline) to 67±7 mmHg (11,500 rpm) in the healthy heart and from 10±9 mmHg (baseline) to 73±8 mmHg (11,500 rpm) in failing heart mode (P<0.05). Aortic root diameters were significantly increased at speeds exceeding 10 500 rpm in the healthy heart mode (P<0.05 vs baseline) and approached statistical significance in failing hearts. CONCLUSIONS Increasing assist device flows resulted in pressure overload above the aortic leaflets, impaired leaflet functions, caused aortic root dilatation and altered leaflet coaptation at the central portion of the aortic valve in both modes. We conclude that the deleterious effect of the reduced pulsatile flow on the aortic valve functions and haemodynamics is immediate and such an insult may explain the structural changes of the aortic valve causing leaflet fusion and/or regurgitation in the chronic phase.

Arterial pulsatility improvement in a feedback-controlled continuous flow left ventricular assist device: an ex-vivo experimental study.

Continuous flow left ventricular assist devices (CF-LVADs) reduce arterial pulsatility, which may cause long-term complications in the cardiovascular system. The aim of this study is to improve the pulsatility by driving a CF-LVAD at a varying speed, synchronous with the cardiac cycle in an ex-vivo experiment. A Micromed DeBakey pump was used as CF-LVAD. The heart was paced at 140 bpm to obtain a constant cardiac cycle for each heartbeat. First, the CF-LVAD was operated at a constant speed. At varying-speed CF-LVAD assistance, the pump was driven such that the same mean pump output was generated. For synchronization purposes, an algorithm was developed to trigger the CF-LVAD each heartbeat. The pump flow rate was selected as the control variable and a reference model was used for regulating the CF-LVAD speed. Continuous and varying-speed CF-LVAD assistance provided the same mean arterial pressure and flow rate, while the index of pulsatility doubled in both arterial pressure and pump flow rate signals under pulsatile pump speed support. This study shows the possibility of improving the pulsatility in CF-LVAD support by regulating pump speed over a cardiac cycle without compromising the overall level of support.

PPG motion artifact handling using a self-mixing interferometric sensor

Pulse oximeters measure a patients heart rate and blood oxygenation by illuminating the skin and measuring the intensity of the light that has propagated through it. The measured intensities, called photoplethysmograms (PPGs), are highly susceptible to motion, which can distort the PPG derived data. Part of the motion artifacts are considered to result from sensor deformation, leading to a change in emitter-detector distance. It is hypothesized that these motion artifacts correlate to movement of the emitter with respect to the skin. This has been investigated in a laboratory setup in which motion artifacts can be reproducibly generated by translating the emitter with respect to a flowcell that models skin perfusion. The top of the flowcell is a diffuse scattering Delrin skin phantom under which a cardiac induced blood pulse is modeled by a changing milk volume. By illuminating the flowcell, a PPG can be measured. The emitters translation has been accurately measured using self-mixing interferometry (SMI). The motion artifacts in the PPG as a result of emitter motion are shown to correlate with the emitters displacement. Moreover, it is shown that these artifacts are significantly reduced by a least-mean-square algorithm that uses the emitters displacement measured via SMI as artifact reference.

Ex vivo proof-of-concept of end-to-end scaffold-enhanced laser-assisted vascular anastomosis of porcine arteries.

OBJECTIVE The low welding strength of laser-assisted vascular anastomosis (LAVA) has hampered the clinical application of LAVA as an alternative to suture anastomosis. To improve welding strength, LAVA in combination with solder and polymeric scaffolds (ssLAVA) has been optimized in vitro. Currently, ssLAVA requires proof-of-concept in a physiologically representative ex vivo model before advancing to in vivo studies. This study therefore investigated the feasibility of ex vivo ssLAVA in medium-sized porcine arteries. METHODS Scaffolds composed of poly(ε-caprolactone) (PCL) or poly(lactic-co-glycolic acid) (PLGA) were impregnated with semisolid solder and placed over coapted aortic segments. ssLAVA was performed with a 670-nm diode laser. In the first substudy, the optimum number of laser spots was determined by bursting pressure analysis. The second substudy investigated the resilience of the welds in a Langendorf-type pulsatile pressure setup, monitoring the number of failed vessels. The type of failure (cohesive vs adhesive) was confirmed by electron microscopy, and thermal damage was assessed histologically. The third substudy compared breaking strength of aortic repairs made with PLGA and semisolid genipin solder (ssLAVR) to repairs made with BioGlue. RESULTS ssLAVA with 11 lasing spots and PLGA scaffold yielded the highest bursting pressure (923 ± 56 mm Hg vs 703 ± 96 mm Hg with PCL ssLAVA; P = .0002) and exhibited the fewest failures (20% vs 70% for PCL ssLAVA; P = .0218). The two failed PLGA ssLAVA arteries leaked at 19 and 22 hours, whereas the seven failed PCL ssLAVA arteries burst between 12 and 23 hours. PLGA anastomoses broke adhesively, whereas PCL welds failed cohesively. Both modalities exhibited full-thickness thermal damage. Repairs with PLGA scaffold yielded higher breaking strength than BioGlue repairs (323 ± 28 N/cm(2) vs 25 ± 4 N/cm(2), respectively; P = .0003). CONCLUSIONS PLGA ssLAVA yields greater anastomotic strength and fewer anastomotic failures than PCL ssLAVA. Aortic repairs with BioGlue were inferior to those produced with PLGA ssLAVR. The results demonstrate the feasibility of ssLAVA/R as an alternative method to suture anastomosis or tissue sealant. Further studies should focus on reducing thermal damage.

Development of an Ex Vivo, Beating Heart Model for CT Myocardial Perfusion

Objective. To test the feasibility of a CT-compatible, ex vivo, perfused porcine heart model for myocardial perfusion CT imaging. Methods. One porcine heart was perfused according to Langendorff. Dynamic perfusion scanning was performed with a second-generation dual source CT scanner. Circulatory parameters like blood flow, aortic pressure, and heart rate were monitored throughout the experiment. Stenosis was induced in the circumflex artery, controlled by a fractional flow reserve (FFR) pressure wire. CT-derived myocardial perfusion parameters were analysed at FFR of 1 to 0.10/0.0. Results. CT images did not show major artefacts due to interference of the model setup. The pacemaker-induced heart rhythm was generally stable at 70 beats per minute. During most of the experiment, blood flow was 0.9–1.0 L/min, and arterial pressure varied between 80 and 95 mm/Hg. Blood flow decreased and arterial pressure increased by approximately 10% after inducing a stenosis with FFR ≤ 0.50. Dynamic perfusion scanning was possible across the range of stenosis grades. Perfusion parameters of circumflex-perfused myocardial segments were affected at increasing stenosis grades. Conclusion. An adapted Langendorff porcine heart model is feasible in a CT environment. This model provides control over physiological parameters and may allow in-depth validation of quantitative CT perfusion techniques.

Intra-aortic balloon pump support in the isolated beating porcine heart in nonischemic and ischemic pump failure

The blood pressure changes induced by the intra-aortic balloon pump (IABP) are expected to create clinical improvement in terms of coronary perfusion and myocardial oxygen consumption. However, the measured effects reported in literature are inconsistent. The aim of this study was to investigate the influence of ischemia on IABP efficacy in healthy hearts and in shock. Twelve slaughterhouse porcine hearts (hearts 1-12) were connected to an external circulatory system, while physiologic cardiac performance was restored. Different clinical scenarios, ranging from healthy to cardiogenic shock, were simulated by step-wise administration of negative inotropic drugs. In hearts 7-12, severe global myocardial ischemia superimposed upon the decreased contractile states was created. IABP support was applied in all hearts under all conditions. Without ischemia, the IABP induced a mild increase in coronary blood flow and cardiac output. These effects were strongly augmented in the presence of persisting ischemia, where coronary blood flow increased by 49 ± 24% (P < 0.01) and cardiac output by 17 ± 6% (P < 0.01) in case of severe pump failure. As expected, myocardial oxygen consumption increased in case of ischemia (21 ± 17%; P < 0.01), while it slightly decreased without (-3 ± 6%; P < 0.01). In case of progressive pump failure due to persistent myocardial ischemia, the IABP increased hyperemic coronary blood flow and cardiac output significantly, and reversed the progressive hemodynamic deterioration within minutes. This suggests that IABP therapy in acute myocardial infarction is most effective in patients with viable myocardium, suffering from persistent myocardial ischemia, despite adequate epicardial reperfusion.

Biodegradable polymer scaffold, semi‐solid solder, and single‐spot lasing for increasing solder‐tissue bonding in suture‐free laser‐assisted vascular repair

We recently showed the fortifying effect of poly‐caprolactone (PCL) scaffold in liquid solder‐mediated laser‐assisted vascular repair (ssLAVR) of porcine carotid arteries, yielding a mean ± SD leaking point pressure of 488 ± 111 mmHg. Despite supraphysiological pressures, the frequency of adhesive failures was indicative of weak bonding at the solder‐tissue interface. As a result, this study aimed to improve adhesive bonding by using a semi‐solid solder and single‐spot vs. scanning irradiation. In the first experiment, in vitro ssLAVR (n = 30) was performed on porcine abdominal aorta strips using a PCL scaffold with a liquid or semi‐solid solder and a 670‐nm diode laser for dual‐pass scanning. In the second experiment, the scanning method was compared to single‐spot lasing. The third experiment investigated the stability of the welds following hydration under quasi‐physiological conditions. The welding strength was defined by acute breaking strength (BS). Solder‐tissue bonding was examined by scanning electron microscopy and histological analysis was performed for thermal damage analysis. Altering solder viscosity from liquid to semi‐solid solder increased the BS from 78 ± 22 N/cm2 to 131 ± 38 N/cm2. Compared to scanning ssLAVR, single‐spot lasing improved adhesive bonding to a BS of 257 ± 62 N/cm2 and showed fewer structural defects at the solder‐tissue interface but more pronounced thermal damage. The improvement in adhesive bonding was associated with constantly stronger welds during two weeks of hydration. Semi‐solid solder and single‐spot lasing increased welding strength by reducing solder leakage and improving adhesive bonding, respectively. The improvement in adhesive bonding was associated with enhanced weld stability during hydration. Copyright

Estimation of left ventricular pressure with the pump as “sensor” in patients with a continuous flow LVAD

Introduction In long-term ventricular support of patients with LVADs, left ventricular pressure (plv is relevant for indicating the unloading level of the heart. Monitoring of plv over time might give more insight into the increase or decrease in native ventricular function. In this study, we aim to assess dynamic plv noninvasively, using the LVAD as a pressure sensor. Methods Pressure head (dplvad) was estimated from pump flow with a dynamic pump model (1). Estimated dplvad and measured aortic pressure were used to calculate left ventricular pressure. Moreover, parameters dp/dtmax and mean, minimum, and maximum plv were derived. The method was validated with a porcine ex vivo beating heart model by measurements conducted in 4 hearts supported with a Micromed DeBakey VAD and 3 hearts with a Heartmate II VAD. During each measurement, aortic and left ventricular pressure, pump flow, and pressure head were recorded for 30 s with a sampling frequency of 1 kHz. Results The estimation of left ventricular pressure appeared to be accurate for both pumps. The parameters mean and minimum pressure were estimated with high accuracy. The degree of accuracy of the estimated plv was proportional to the degree of accuracy of the dynamic pump model. Conclusions We proved that the LVAD model described in this paper can be used as a pressure indicator to determine LV pressure at any time based on noninvasive measurements of pump flow, aortic pressure, and the properties of the outlet graft.