2021 J. Leonard Goldner Award Winner: Vancomycin Topically Applied at the Surgical Site Does Not Impair Diabetic Fracture Healing and Dose-Dependently Inhibits Calcified Tissue Formation by Osteoblast Precursors Cells

Abstract

Category: Diabetes; Basic Sciences/Biologics; Trauma Introduction/Purpose: Prophylactic vancomycin treatment decreases the rates of surgical site and deep infections by >70% for diabetic patients undergoing reconstructive foot and ankle surgery. We aimed to identify whether local vancomycin at a clinically relevant dose impaired fracture healing in diabetic rats. Our hypothesis was that local vancomycin powder to the fracture site would not affect long term healing outcomes, but continuous exposure of vancomycin would inhibit differentiation of osteoblast precursor cells and their osteogenic activity in vitro. The 25 mg/kg vancomycin was a modest increase to routine surgical site vancomycin application of 1-2 grams for a 70 kg. adult (21.42 mg/kg). Determining how vancomycin affects diabetic fracture healing is of clinical interest, particularly considering its prophylactic use in foot and ankle patients. Methods: Following induction of a femur fracture in male BB Wistar type 1 diabetic rats, a longitudinal incision was made in the lateral aspect of the femur. Powdered vancomycin (25mg/kg) was administered to fracture site of treated (n=5), and Sham surgery (n=6) control rats. Femurs harvested at 6 weeks after fracture, were X-ray scored, µCT scanned and tested to failure in torsion. Bone marrow and periosteal cells isolated from diabetic bones and MC3T3 cells were plated for MTT survival and osteogenic assays. After 4 days, beta glycerol phosphate, ascorbic acid and either 0, 50, 500, or 5,000 µg/mL vancomycin were added to the media (n=3 per group). Cells were fixed at 7 and 14 days for alkaline phosphatase (ALP) staining and at 28 days for Alizarin Red S staining. Parametric data were analyzed using student t-tests. Non-parametric data were analyzed using a Kruskal-Wallis ANOVA on RANKs and Tukey post-hoc tests. Results: Radiographic scoring did not show differences between the control (2.28+-1.57) and treatment (3.67+-0.75) groups (p=0.093). BV/TV was similar between the control (61.2+-8.4 %) and treatment (56.5+-5.2 %) groups (p=0.081). Mechanical testing found similar values in normalized torque to failure (69.56+-36.80 vs 50.05+-28.82 %N*mm, p=0.356), and torsional rigidity (105.97+-103.62 vs 48.68+-32.95 %Nmm 2 , p=0.291), for control and treatment groups, respectively. ALP staining was comparable between groups at either 7 (p=0.809) or 14 days (p=0.343) in bone marrow or MC3T3 cells. Alizarin Red S staining found dose- dependent decreases in mineralized nodule formation between the untreated (10.16+-5.01 nodules per 3.8cm 2 ), and 500µg/mL vancomycin (3.16+-2.26 nodules per 3.8cm 2 ) bone marrow groups at 28 days (p=0.026). Periosteal and MC3T3 cell viability were only impaired at the 5000µg/mL vancomycin dose (p=0.001). Conclusion: Our study argues that local application of vancomycin does not affect diabetic fracture healing at clinically relevant doses. We studied a nearly 2-fold higher vancomycin dose than the 'standard' dose of 14.3mg/kg and less than the 143.5mg/kg dose that did not impair rat spine fusion. Although in vitro vancomycin did not affect cell viability and osteogenic staining in periosteal and MC3T3 cells, mineralized nodule formation was dose-dependently inhibited at day 28 in bone marrow cells. Our results bring novel insight into the effects of vancomycin in diabetic fracture healing and the how long-term dosing impacts cell viability and osteoblastogenesis.