Development of the Extended Lane and Sandu Score to Assess Osseous Repair

Lane and Sandhu introduced a radiological score to evaluate bone union in animal and human models after experimental bone grafting. We introduce an extension of the Lane and Sandhu Scoring system to serve as a more granular system for robust bone healing assessment in clinical research. We describe the application of this scoring system in the context of an autologous cell-based osteogenic implant (NVD003) designed to improve the rate of bone union. Images from four participants with congenital pseudoarthrosis of the tibia (CPT) and 9 adults with recurrent osseous non-union after conventional surgery were obtained. All participants underwent NVD003 implantation. An extended version of the Lane and Sandhu scoring system (eLSS; range from 0 to 12) was employed to assess bone formation, union, and remodeling by two independent readers longitudinally, with final adjudication in cases of > 1 point disagreement. Interobserver agreement was assessed using quadratic weighted kappa. 206 exams (170 radiographs and 36 CT scans) were studied across the two pathologies (81 scans for participants with CPT and 125 scans for participants with recurrent osseous non-union). Overall agreement between the expert readers was found to be 0.84 (95% CI: 0.79, 0.89), with comparable agreement found per modality (radiographs [0.85] and CT images [0.76]) and per pathology (CPT [0.83] and recurrent osseous non-union [0.84]). These results support that the eLSS has sufficient reliability to warrant further consideration for clinical trial use. Additional studies with a greater number of subjects, raters, and variety of use cases are required to validate and optimize this method. Level of Evidence: III.