JetValve: Rapid manufacturing of biohybrid scaffolds for biomimetic heart valve replacement

Abstract

Tissue engineered scaffolds have emerged as a promising solution for heart valve replacement because of their potential for regeneration. However, traditional heart valve tissue engineering has relied on resource-intensive, cell-based manufacturing, which increases cost and hinders clinical translation. To overcome these limitations, in situ tissue engineering approaches aim to develop scaffold materials and manufacturing processes that elicit endogenous tissue remodeling and repair. Yet despite recent advances in synthetic materials manufacturing, there remains a lack of cell-free, automated approaches for rapidly producing biomimetic heart valve scaffolds. Here, we designed a jet spinning process for the rapid and automated fabrication of fibrous heart valve scaffolds. The composition, multiscale architecture, and mechanical properties of the scaffolds were tailored to mimic that of the native leaflet fibrosa and assembled into three dimensional, semilunar valve structures. We demonstrated controlled modulation of these scaffold parameters and show initial biocompatibility and functionality in vitro. Valves were minimally-invasively deployed via transapical access to the pulmonary valve position in an ovine model and shown to be functional for 15 h.

Keywords: Biohybrid; Heart valve; Nanofiber; Rapid manufacture; Rotary Jet Spinning; Tissue engineering.

Fig. 1

Automated Rotary Jet Spinning of JetValves. (a) CAD representation of the automated Rotary Jet Spinning system (aRJS). (b) A two-step mandrel collection system was used consisting of (1) a leaflet mandrel and (2) shielding mandrel. (c) At a 30k RPM fiber extrusion rate, 4% w/v P4HB/Gelatin biohybrid solutions had decreased fiber diameter but increased percent scaffold porosity as a function of decreasing polymer content (N=6 production runs per condition, *p<0.5). (d) Stress vs strain plots of 60/40 P4HB/Gelatin blends compared to native leaflet cusps up to 20% strain (blue circumferential (C), red radial (R) fiber alignment; N=6 production runs per condition, N=5 native leaflets; *p<0.5)

Fig. 2

JetValve mandrel scaling and customization. (a) Digital photograph, (left) of 3D printed shielding (upper row) and leaflet (lower row) JetValve mandrels and scaffolds ranging from 30 mm to 3 mm in diameter. Scanning electron microscope images, (right) of miniaturized JetValve mandrels and scaffold, 750 μm in diameter. (b) Shielding mandrel modification for JetValve scaffolds with sinus bulges. The shielding mandrel was compartmentalized into individual, symmetric sinus component “inserts” which could be fixed to a housing sinus “core,” (left). Mandrels were removed from scaffolds without disrupting the structure, digital photographs (right), by removing the connections of the core and inserts.

Fig. 3

JetValve catheter-based deployment and crimping. (a) Transcatheter delivery involved fixing the scaffold in a self-expanding nitinol stent, transapical placement via entry through the right ventricle (RV), positioning via a guide wire system, deployment of the stented scaffold over the native leaflets, and retraction of the catheter through the ventricle. Radial pressure of the released stent held the valve scaffold in place between the RV and pulmonary artery (PA), over the native valve leaflets. (b) Crimping of anchored JetValves from the 30 mm fully extended conformation to the fully crimped 9mm conformation to accommodate implantation.

Fig. 4

JetValve leaflet/conduit anisotropy and porosity. (a) JetValve leaflet anisotropy was comparable to native anisotropy, as indicated by OOP, and was significantly more anisotropic than the conduit for each collection angle. Colorized SEM images, right, indicate local fiber direction, (R) indicates radial direction and (C) indicates circumferential direction (N=3 production runs per condition and N=7 native leaflets, *p<0.5 comparing leaflet vs. conduit). (b) Representative SEM images of JetValve leaflet and conduit scaffold collected at 45° (scale bar 500μm). Porosity of leaflets and conduits as a function of collection angles (N=3 production runs per condition, *p<0.5 comparing angles and ^#p<0.5 comparing leaflet vs conduit for a given angle). Data presented as mean ± s.e.m. (leaflet: grey, conduit: black).

Fig. 5

JetValve biaxial stiffness as a function of biohybrid composition, collection angle, and location. (a) Increasing protein percentage within the biohybrid ratio of spun scaffolds decreased the low strain (0–10%) and high strain (10–20%) biaxial global stiffness of scaffolds (N=6 production runs per condition, N=5 native leaflets; *p<0.5, data presented as mean ± s.e.m. (b) Conduit samples comprised of 60/40 P4HB/Gelatin blends were stiffer than corresponding collection angle leaflet samples for both low and high strains (N=6 production runs per conditions; *p<0.5 between leaflet and conduit stiffness for the same collection angle, data presented as mean ± s.e.m.).

Fig. 6

Shelf life composition and stiffness of JetValves. (a) Initial protein content and P4HB crystallinity was measured by comparing carbonyl stretch and amide FTIR absorbency peak heights. (b) Neither protein content nor polymer crystallinity were affected by fiber extrusion spin speed; Biohybrid blends of 60/40 P4HB/Gelatin exhibited stable relative crystallinity for all spin speeds compared to compositions with higher synthetic polymer content (N=3 production runs per condition). (c) Ratiometric (XPS) hydrated scaffold content over the course of 1 week. Inset: trace amounts of bound HFIP solvent were detected upon hydration (N=3 production runs). (d) The biaxial mechanical properties of 1 week hydrated scaffolds compared to those of fresh-spun scaffolds (N=3 production runs per condition; blue represents circumferential (C), red represents radial (R) fiber alignment; data presented as mean ± s.e.m.).

Fig. 7

In vitro valvular interstitial cell (VIC) infiltration. (a) VICs infiltrated the JetValve leaflet portion of scaffolds in greater abundance than conduit portions by 1 week; by 2 weeks, infiltration depth evened at ~25–26 μm from the scaffold surface (N=6 production runs/tissues per condition, *p<0.5 between like scaffold areas in time, #p<0.5 between leaflet and conduit at the same time point; data presented as mean ± s.e.m) (b) Representative three dimensional reconstructions of VIC nuclei within the JetValve scaffold (red indicates nuclei of cells on the scaffold surface, while blue indicates the nuclei of cells that have penetrated into the scaffold, all images in isometric 3D view, 40,000 μm² area).

Fig. 8

In vitro and in vivo functionality. (a) Top: digital photographs from arterial view of mounted JetValves during systole and diastole at 48 hr (dotted lines highlight the JetValve leaflet and conduit edges). Bottom: flow through the JetValve reached ~175 ml/s during peak systole with complete valve closure during diastole (~30% regurgitant fraction, ~10 mmHg transvalvular pressure). (b) Top: distal three-dimensional echocardiography revealed complete leaflet opening and closing during systole and diastole respectively at 15 hr. Bottom: Doppler imaging showed unrestricted blood flow through the JetValve leaflets during systole and complete closure with minor regurgitation fraction during diastole (RV: right ventricle, PA: pulmonary artery).