Dynamic musculoskeletal loading can make in-vitro porcine lumbar exhibit biomechanical performance similar to those in-vivo.

Abstract

In-vitro spinal testing often relies on passive and static loading, causing discrepancies in kinematic and mechanical behaviors compared to in-vivo performance. These constrain the reliability of biomechanical evaluations. 1) To establish a spinal musculoskeletal loading platform and evaluate the biomechanical responses under various loading protocols. 2) to propose an optimal loading protocol based on this active dynamic loading platform. In-vitro biomechanical study. A musculoskeletal loading platform was developed to replicate global (active and passive) and local muscle loads. Global muscle forces were simulated via active and passive loading modules, while local muscle stabilization was mimicked utilizing a follower load (FL) that applies a compressive force along the lumbar curvature. Four loading protocols were designed and applied using a novel platform for forward flexion tests on 7 porcine lumbar specimens (L3-L5). The protocols included: 1) DynWoFL: dynamic active and dynamic passive muscle loading without FL as a preload, 2) StaPass: dynamic active and static passive muscle forces with a FL as preload, 3) StaLdg: static active and static passive muscle loading with a FL, and 4) DynLdg: dynamic active and dynamic passive muscle loading combined with a FL. Kinematics assessed range of motion (ROM), vertebral translation, and center of rotation (COR) distribution. Mechanical assessments covered intradiscal pressure (IDP) changes and disc surface strain. Under loading protocols with FL (StaPass, StaLdg, DynLdg), the ROMs of L3/L4 and L4/L5 increased to about 4°. FL reduced the COR distribution range, thereby enhancing spinal movement stability. However, FL had minimal effect on surface strain between DynWoFL and DynLdg groups. Compared to DynLdg, the static passive loading group (StaPass) exhibited additional extension of 3° and posterior translation of 2 mm, with greater medial-lateral and cranial-caudal translations. The COR distribution was also located lower. Although no significant difference in IDP increase was found, StaPass showed significantly higher disc strain. The kinematic performance and disc surface strain of StaLdg were similar to StaPass, and static active and passive muscle loading (StaLdg) underestimated IDP increase by 6.95%-14.70%. Among the musculoskeletal loading procedures, DynLdg produced ROM, vertebral translation, and disc strain values that closely matched published in-vivo data, supporting its substitutes for related studies. For research focusing on COR distribution and IDP changes, they were more suitable for providing qualitative rather than quantitative insights. The dynamic musculoskeletal loading protocol enables porcine lumbar spines to replicate biomechanical responses similar to human in-vivo behaviors, establishing it as a reliable alternative for in-vitro biomechanical research in the future.