Abstract
INTRODUCTION
Finite element analysis (FEA) is widely used to study micromotion between the glenoid baseplate and bone, as a pre-clinical indicator for clinical stability in reverse total shoulder arthroplasty (rTSA). Various key parameters such as the number, length, and angle of screws have been shown to influence micromotion [1]. This study explores the influence of screw preloads, an insufficiently studied parameter. Specifically, two rTSA configurations with 18mm and 48mm peripheral screws (PS) were analyzed without screw preloads, followed by analysis of the 48mm PS configuration with an experimentally measured screw preload.
METHODS
FEA models were created to simulate a fixation experiment inspired by ASTM F2028-14. The rTSA configurations used here have a superior and an inferior PS. The assemblies were virtually implanted into a synthetic bone block as per surgical technique. Sliding contacts were defined to model the interface between screw threads-bone, and between baseplate-bone.
To determine the screw preload experimentally, the 48mm screw (n=5) was inserted through a hole in a metal plate, which rested on top of a Futek washer load cell, placed on top of the foam block with a predrilled pilot hole (Figure 1). The screw was inserted using a torque driver until the average human factors torque for the screw driver handle was reached. The resulting axial compressive load due to screw insertion was measured by the washer load cell.
Two step analyses were performed using Ansys version 17.2 for 18mm and 48mm PS, where 756N axial and shear loads were applied sequentially. The model with the 48mm PS was then analyzed in a four step analysis; preload inferior and superior screws, followed by applying the axial and shear loads (Figure 2). Peak overall micromotion including tangential and normal components at the baseplate-bone interface was compared for all three models.
RESULTS
From the experimental study, the mean screw preload for the 48mm screw was determined to be 141±8 lbs. Peak micromotion was predicted at the inferior edge of the baseplate (Figure 3A). In the two models without screw preloads, the model with the 48mm PS predicted 42% lower micromotion than the model with the 18mm PS. The 48mm PS model predicted 63% further reduction in micromotion by including the preload for the two PS. Figure 3B presents the micromotion comparison between these three models.
DISCUSSION
This study demonstrates the significant influence that screw preload can have on evaluating either absolute values or differential performance of rTSA micromotion within the same design family. It further demonstrated that the inclusion of preload in simulation can have as much (or greater) impact on micromotion as other key parameters such as shorter versus longer screws. These findings indicates that it is important to include appropriate values of screw preloads in simulations when comparing designs with different number of peripheral screws or studying the effects of including a central screw on rTSA micromotion.
