High-fidelity tissue engineering of patient-specific auricles for reconstruction of pediatric microtia and other auricular deformities

Abstract

Introduction: Autologous techniques for the reconstruction of pediatric microtia often result in suboptimal aesthetic outcomes and morbidity at the costal cartilage donor site. We therefore sought to combine digital photogrammetry with CAD/CAM techniques to develop collagen type I hydrogel scaffolds and their respective molds that would precisely mimic the normal anatomy of the patient-specific external ear as well as recapitulate the complex biomechanical properties of native auricular elastic cartilage while avoiding the morbidity of traditional autologous reconstructions.

Methods: Three-dimensional structures of normal pediatric ears were digitized and converted to virtual solids for mold design. Image-based synthetic reconstructions of these ears were fabricated from collagen type I hydrogels. Half were seeded with bovine auricular chondrocytes. Cellular and acellular constructs were implanted subcutaneously in the dorsa of nude rats and harvested after 1 and 3 months.

Results: Gross inspection revealed that acellular implants had significantly decreased in size by 1 month. Cellular constructs retained their contour/projection from the animals' dorsa, even after 3 months. Post-harvest weight of cellular constructs was significantly greater than that of acellular constructs after 1 and 3 months. Safranin O-staining revealed that cellular constructs demonstrated evidence of a self-assembled perichondrial layer and copious neocartilage deposition. Verhoeff staining of 1 month cellular constructs revealed de novo elastic cartilage deposition, which was even more extensive and robust after 3 months. The equilibrium modulus and hydraulic permeability of cellular constructs were not significantly different from native bovine auricular cartilage after 3 months.

Conclusions: We have developed high-fidelity, biocompatible, patient-specific tissue-engineered constructs for auricular reconstruction which largely mimic the native auricle both biomechanically and histologically, even after an extended period of implantation. This strategy holds immense potential for durable patient-specific tissue-engineered anatomically proper auricular reconstructions in the future.

Figure 1. Digitization process for human ears.

The anatomy of a 5 year-old female was scanned (A, D), processed to remove noise (B, E), and digitally sculpted to obtain the appropriate curvature for the anterior portion of the ear (C, F). Sagittal (A–C) and worm's-eye (D–F) views.

Figure 2. Mold design based on ear anatomy.

The digital images of ears (A) were used to design 7-part molds (B–H) by embedding the solid images of the ear into virtual blocks.

Figure 3. Schematic representation of length and width measurements.

Construct length was measured along the lobule-helix axis. Construct width was defined as the largest dimension measured along an axis perpendicular to the lobule-helix axis.

Figure 4. Ex vivo gross analysis.

Three months after implantation, acellular implants (A) had decreased in size, whereas cellular constructs (B) retained their original anatomic fidelity.

Figure 5. Ex vivo analysis of specimen length and width.

(A) The length of acellular constructs harvested after 3 months was significantly less that that of constructs harvested after 1 month. In contrast, cellular construct length did not change over time. (B) Cellular construct width was significantly greater than acellular construct width at 3 months. * denotes p<0.05.

Figure 6. Safranin O staining of specimens harvested after 1 month.

Acellular constructs (A) and cellular constructs (C) demonstrated evidence of a thin capsule containing spindle-shaped, fibroblast-appearing cells (star). Although the acellular constructs were invaded by mononuclear cells, there was no evidence of cartilage deposition (B). Cellular constructs demonstrated marked cartilage deposition by lacunar chondrocytes (arrows) throughout the construct (B, D). Scale bar = 100 µm.

Figure 7. Histologic comparison of 1-month and 3-month samples by Safranin O and Verhoeff stains.

Low magnification comparison between 1-month (A) and 3-month (B) Safranin O-stained sections (A–F) demonstrates more intense and uniform staining after 3 months (scale bar = 1 mm). Inspection of the edge of 1-month (C) and 3-month (D) samples shows a transition from the fibrous capsule (FC) to a perichondrial layer (PC) to cartilage (scale bar = 100 µm). High magnification comparison at 1-month (E) and 3-month (F) shows mature cartilage formation at both times (scale bar = 50 µm). Verhoeff's stain reveals the presence of elastin at both 1-month (G) and 3-months (H), with a more continuous network of elastin fibers after 3 months (scale bar = 50 µm).

Figure 8. Equilibrium modulus and hydraulic permeability of tissue-engineered and native bovine auricular cartilage.

Tissue-engineered auricular cartilage showed progressive improvement in mechanical properties with increasing time in vivo. The equilibrium modulus (A) and hydraulic permeability (B) of implants at 3 months were not statistically different from those of native bovine auricular cartilage. Data are displayed as mean+standard deviation for n = 4 for 0- and 1-month tissue-engineered samples, n = 5 for 3-month tissue-engineered samples, and n = 6 samples for native cartilage. * denotes p<0.05.