New Polycarbonate Polyurethanes Show Promise for Durable Polymer Mitral Valves

Polymeric heart valves have long promised the best of both worlds in prosthetic cardiology: the durability of mechanical valves and the gentle blood-flow characteristics of tissue valves. Yet the field has struggled to link polymer chemistry at the molecular scale to how an actual valve behaves inside a beating heart. A new study from researchers at the University of California, Irvine, published in the Annals of Biomedical Engineering, takes a major step toward closing that gap by evaluating two medical-grade polycarbonate-based polyurethanes through a comprehensive multiscale framework spanning molecular structure, leaflet mechanics, hemodynamic function, and early in vivo performance.

The research team, led by Farhad Sadeghi and Arash Kheradvar, compared two aromatic polycarbonate-based thermoplastic polyurethanes: Carbothane AC-4095A (CB95-AC), a polycarbonate polyurethane produced by Lubrizol, and QuadraSil ARCS 90A (Qsil), a polycarbonate-siloxane copolyurethane from Biomerics. Both materials are segmented polyurethanes composed of alternating hard and soft domains—the hard segments, formed from chain extenders and diisocyanates, act as physical crosslinks that stabilize the polymer network, while the soft segments, derived from polycarbonate polyols, provide extensibility. The critical question was whether subtle differences in composition, particularly the siloxane-containing phase in Qsil, would translate into measurable differences in how a mitral valve scaffold performs under the punishing cyclic loads of the cardiac cycle.

The investigators subjected the two polymers to an unusually thorough characterization program. Rheological testing of 20 weight percent polymer solutions in dimethylacetamide revealed that Qsil exhibited consistently higher viscosity and pronounced shear-thinning, or non-Newtonian, behavior, whereas CB95-AC flowed with lower, nearly frequency-independent viscosity consistent with near-Newtonian response. Energy-dispersive X-ray fluorescence confirmed the presence of silicon in Qsil at 26.7 plus or minus 0.15 weight percent, providing direct compositional evidence of the siloxane phase, while no silicon was detected in CB95-AC. Fourier transform infrared spectroscopy showed remarkably similar carbonyl absorption profiles for both materials, with hydrogen-bonded to free carbonyl peak-area ratios of 0.46 for CB95-AC and 0.48 for Qsil, indicating comparable hard-segment association and microphase organization. Differential scanning calorimetry likewise revealed broadly similar thermal transitions, suggesting that the siloxane incorporation did not substantially alter the bulk microstructure.

Mechanical testing, however, exposed meaningful functional differences. CB95-AC exhibited slightly higher tensile stress at low strain, indicating greater initial stiffness, and consistently outperformed Qsil in tear resistance across room-temperature, wet, and accelerated-aging conditions. Qsil, by contrast, demonstrated superior elastic recovery: under cyclic hysteresis testing it showed lower residual strain, and at large 50 percent deformations it retained the lowest permanent deformation of the two. Dynamic mechanical analysis reinforced this picture, showing that CB95-AC possessed a higher storage modulus across the tested frequency range, while Qsil displayed a higher loss factor, or tan delta, reflecting greater energy dissipation and a more compliant viscoelastic character. Under physiologically relevant conditions—wet testing at 37 degrees Celsius—both polymers softened somewhat but showed similar cyclic behavior with reduced residual strain. Oxidative accelerated aging in cobalt chloride and hydrogen peroxide solutions modestly reduced tear resistance in both materials, consistent with known degradation mechanisms in polycarbonate urethanes, though the relative ranking between the polymers remained unchanged.

With these material fingerprints in hand, the team fabricated trileaflet mitral valve scaffolds by dip molding—27 millimeters in diameter with average leaflet thickness of roughly 190 micrometers—reinforced with Nitinol wireframes for structural support. Identical processing parameters were used for both formulations to permit a controlled comparison. The valves were then evaluated in a pulsatile heart-flow simulator incorporating a compliant silicone left ventricular sac that reproduces physiologic mitral flow geometry and compliance. Particle image velocimetry, using a high-speed 5K camera paired with a pulsed Nd:YLF laser sheet and neutrally buoyant tracer particles, enabled reconstruction of time-resolved velocity fields during the cardiac cycle. Both valves generated physiologic pulsatile inflow with peak rates of approximately 5 liters per minute, a central inflow jet, and clear intraventricular vortex formation during diastole—hallmarks of healthy transmitral flow organization.

Subtle but telling differences emerged in the flow fields. The Qsil valve produced a slightly broader and more spatially distributed high-vorticity region, suggesting more diffuse intraventricular recirculation, while the CB95-AC valve generated a narrower inflow jet with more localized vortical structures. The researchers attribute these differences to leaflet compliance: the more compliant Qsil leaflets open more fully and interact more gently with the passing blood, whereas stiffer CB95-AC leaflets channel flow more sharply. This distinction matters because efficient vortex formation has been linked to ventricular filling efficiency, momentum preservation, and reduced energy dissipation, while disorganized flow is associated with increased thrombogenic potential. Neither valve showed visible reverse flow, leakage, or paravalvular leakage when fully closed in the velocity fields, indicating effective leaflet coaptation under the tested conditions.

Durability was assessed through accelerated wear testing at 800 cycles per minute for up to 50 million cycles—roughly equivalent to more than a year of continuous cardiac cycling. Both valve designs maintained symmetric trileaflet opening, effective coaptation, and stable pressure waveforms throughout, with no visible tearing, delamination, or gross deformation. The projected orifice area was slightly larger for Qsil at 329 plus or minus 2 square millimeters compared with 309 plus or minus 22 square millimeters for CB95-AC, an approximately 6.5 percent difference that did not reach statistical significance. For representative valves tracked before and after cycling, orifice areas changed minimally: CB95-AC moved from 314.5 to 315.5 square millimeters, and Qsil from 327 to 318.5 square millimeters. Pressure waveform analysis showed that CB95-AC exhibited an earlier, sharper positive pressure peak consistent with its stiffer dynamic response, whereas Qsil displayed a delayed, broader, more damped profile echoing its compliance. The authors emphasize that this wear testing served as a comparative durability screen rather than formal ISO 5840 qualification, since peak differential pressures of roughly 25 to 30 millimeters of mercury fell below the standard’s mitral requirements.

The study also included a preliminary acute in vivo evaluation, in which a single CB95-AC mitral valve was implanted in a 43-kilogram sheep under cardiopulmonary bypass, with a PET fabric sewing ring facilitating annular fixation. Doppler echocardiography immediately after implantation revealed a clear biphasic transmitral filling pattern: a peak early filling velocity of 68.2 centimeters per second, an atrial filling velocity of 54.7 centimeters per second, and an E/A ratio of 1.2, all consistent with physiologic diastolic filling. Transvalvular pressure gradients were low—approximately 2 millimeters of mercury for the E-wave and 1 for the A-wave—with a deceleration time of 164 milliseconds, a pressure half-time of 48 milliseconds, and a calculated mitral valve area of 4.58 square centimeters, indicating adequate leaflet opening without obstruction. The animal remained hemodynamically stable throughout, supporting procedural feasibility, though the authors are careful to frame this as an initial assessment rather than evidence of long-term performance.

The broader significance of the work lies in its multiscale framework. Rather than treating material selection and valve design as separate problems, the study demonstrates traceable associations across length scales: solution rheology influences dip-molding behavior and coating uniformity; segment-level viscoelasticity governs leaflet stiffness, damping, and recovery; and these leaflet properties in turn shape pressure waveforms, orifice area, and intraventricular flow organization. The siloxane segments in Qsil appear to increase chain flexibility, lower the storage modulus, and enhance energy dissipation, while the harder CB95-AC network resists tearing and responds faster dynamically. Both property profiles have value—durability demands tear and fatigue resistance, while hemodynamics reward compliance and full opening—suggesting that optimal polymeric valve design may require deliberate tailoring of these competing characteristics rather than maximizing any single property.

The authors are candid about limitations and future directions. Molecular weight effects on rheology are being pursued through developing gel permeation chromatography methods; leaflet-scale factors such as local thickness, bending stiffness, and possible anisotropy warrant dedicated testing; and quantitative leaflet kinematics and fluid–structure interaction modeling are underway to move from correlation toward causation. Chronic studies addressing hemocompatibility, calcification, thrombogenicity, and long-term degradation remain essential before clinical translation. Still, the demonstration that two polycarbonate polyurethanes can survive 50 million wear cycles while producing physiologic ventricular flow patterns—and that one can function with low gradients immediately after implantation in a living heart—marks a substantive advance. For the growing field of polymeric and hybrid tissue-engineered heart valves, this work provides both a methodological template and a encouraging signal that the long-sought synthesis of mechanical durability and bioprosthetic hemodynamics may be chemically within reach.