In Vivo Anastomosis and Perfusion of a Three-Dimensionally-Printed Construct Containing Microchannel Networks

Abstract

The field of tissue engineering has advanced the development of increasingly biocompatible materials to mimic the extracellular matrix of vascularized tissue. However, a majority of studies instead rely on a multiday inosculation between engineered vessels and host vasculature rather than the direct connection of engineered microvascular networks with host vasculature. We have previously demonstrated that the rapid casting of three-dimensionally-printed (3D) sacrificial carbohydrate glass is an expeditious and a reliable method of creating scaffolds with 3D microvessel networks. Here, we describe a new surgical technique to directly connect host femoral arteries to patterned microvessel networks. Vessel networks were connected in vivo in a rat femoral artery graft model. We utilized laser Doppler imaging to monitor hind limb ischemia for several hours after implantation and thus measured the vascular patency of implants that were anastomosed to the femoral artery. This study may provide a method to overcome the challenge of rapid oxygen and nutrient delivery to engineered vascularized tissues implanted in vivo.

FIG. 1.

Description of sugar-glass printing and initial flow testing. (A) Extrusion print head in the process of printing a sugar-glass lattice. Print head can move x, y, or z planes. (B) Final sugar lattice before casting. The lattice contains a network of filaments supported by a surrounding well. Red line denotes the outer edge of the well that will be filled with polydimethylsiloxane (PDMS) during casting. (C) Schematic of printed sugar-glass network. Drawing on the left denotes sugar filaments after printing, whereas the figure on the right shows filaments before casting. Unwanted filaments are removed using a fine-tipped soldering iron before casting. (D) Final PDMS gel with channel network. Excess PDMS has been removed using a razor blade. (E) Microcomputed tomography reconstruction of internal channel network from a cast PDMS construct. (F) Computational model of flow rates through cast channel geometry. Flow streamlines are color coded corresponding to flow rate. Flow rate at the inlet is equal to 0.12 mL/min. Computational models demonstrate the continuity of flow and patency of channel networks. Scale bar = 4 mm. Color images available online at www.liebertpub.com/tec

FIG. 2.

(A) Incision and exposure of femoral neurovascular bundle. (B) Arrows showing femoral nerve (n), common femoral artery (a), common femoral vein (v) dissected, and skeletonized in situ. (C) Cannulation of proximal common femoral artery with 26-gauge angiocatheter. (D) The angiocatheter is advanced in the arterial lumen (arrow) with brisk blood flash observed inside the catheter. One hundred units of heparin was injected systemically through this catheter. (E) Arrows showing proximal (p) and distal (d) common femoral artery cannulated and secured with silk suture before transecting the intercatheter arterial segment. (F) The proximal catheter is mounted on the PDMS gel inlet and blood flow tested. (G) The distal catheter is mounted on the distal outlet of the PDMS gel, and all hemostatic clamps were removed. (H) Pulsatile blood flow through the gel microchannels and perfusing the distal hind limb. Color images available online at www.liebertpub.com/tec

FIG. 3.

(A) Arrow pointing to site where the femoral artery was ligated and transected, and p indicates the paw of the animal. (B, C) No blood flow to limb. (D–F) Arrows pointing to the PDMS gel with flow through the microchannels. The outline and individual channels of the gel can be seen under Doppler imaging. (G–I) The whole hind limb of the animal has flow. Color images available online at www.liebertpub.com/tec