In vivo efficacy proof of concept of a large-size bioprinted dermo-epidermal substitute for permanent wound coverage

**Background:** Full-thickness skin defects from burns, trauma, or surgical excisions are typically treated with autologous split-thickness skin grafts (STSG), but donor-site shortage and scarring limit this approach. Tissue-engineered skin substitutes offer an alternative, but previous bioprinted constructs have been small and not manufactured under good manufacturing practice (GMP) conditions. This study aimed to evaluate the reliability, homogeneity, and in vivo efficacy of a large-size (40 cm²) GMP-compatible bioprinted dermo-epidermal substitute (DES) named Poieskin® in an immunosuppressed murine model, comparing it to the clinical reference HSTSG.

**Methods:** Human dermal fibroblasts and epidermal keratinocytes were isolated from a 25-year-old female donor (BMI 24.6 kg/m²). Poieskin® was fabricated using a Next-Generation Bioprinter (NGB) combining extrusion bioprinting for bovine collagen (4 mg/mL) and laser-assisted bioprinting (LAB) for cell seeding. The dermis comprised three layers of collagen and fibroblasts (12.5 million cells/mL), cultured for 5 days; then keratinocytes (70 million cells/mL) were printed onto the dermis, followed by 2 days immersed culture and 6 days at air–liquid interface. Two 40 cm² batches were produced. Acceptance criteria included size (40 cm² ± 10%), dermal thickness >200 μm, histological score >50, and sterility. Sixteen immunosuppressed female NMRI-Foxn1nu/nu mice (7 weeks old, 20–23 g) received either Poieskin® (n=8) or HSTSG (n=8) on a 1 × 1.5 cm full-thickness back excision. Primary endpoint was graft area measurement from standardized photographs over 16 days. Secondary endpoints included wound contraction, perfusion (laser Doppler imaging), local inflammation ([18F]-FDG MicroPET/CT), and histological analysis (Masson’s trichrome, immunohistochemistry for cytokeratin 5/10, loricrin, collagen 1, αSMA). Statistical analysis used two-way ANOVA or mixed-effects models.

**Key Results:** Poieskin® batches met all acceptance criteria: total thickness 292 ± 26 μm (batch 1) and 226 ± 31 μm (batch 2); histological scores 75% ± 8% and 73% ± 12%, both above the 50% threshold. Immunohistochemistry confirmed proliferative basal keratinocytes (K5+), differentiated suprabasal layers (K10+), and stratum corneum (loricrin+). One HSTSG mouse died during surgery. Mean graft take at day 16 was 91.8% (SD=0.1152) for Poieskin® vs. 100% (SD=0) for HSTSG (mean difference −0.08225; 95% CI −0.2177 to 0.05315; p=0.2942). Graft retraction was significantly faster for Poieskin® (69% contraction) vs. HSTSG (54%) at day 16 (mean difference −16.0%; 95% CI −29.04 to −29.15; p=0.0095). Laser Doppler perfusion showed no significant difference between groups (p=0.1951 at day 16). [18F]-FDG PET/CT revealed no significant metabolic activity increase in grafted areas (day 5: p=0.9998; day 15: p=0.5109). Histological scoring at day 16: Poieskin® mean 75% (SD=8.31, n=8) vs. HSTSG mean 83% (SD=4.99, n=7); mean difference 7.964 (SD=3.588; 95% CI 0.2121 to 15.72; p=0.2423). Collagen 1 staining confirmed human/bovine collagen persistence; αSMA staining suggested neovascularization. No adverse events, infections, or clinical inflammation occurred.

**Clinical Implications:** This study provides the first evidence that a GMP-compatible, large-size (40 cm²) bioprinted DES can achieve engraftment rates (91.8%) comparable to the clinical reference STSG (100%) in a preclinical model, with similar safety and histological quality. The faster contraction of Poieskin® (69% vs. 54%) may be addressed by optimizing dermal matrix crosslinking. The results support progression to human clinical trials for full-thickness skin defects, including burns, excisions, and chronic wounds. The GMP-compliant manufacturing process addresses a key barrier to clinical translation of bioprinted skin substitutes.