Summary

Use of DBVR revealed significant deviations in both the glenoid and humeral component positions compared to the preoperative plan, indicating that surgical implementation is, as expected, not without errors.

Abstract

Introduction

Accurate implant positioning is crucial for achieving optimal shoulder mechanics, the longevity of the implant and surrounding bones, and patient outcomes short and long-term. Three-dimensional (3D) templating software is increasingly used to optimize component selection and placement. This software uses 3D patient-specific bone models generated from CT scans to determine the optimal size and placement of implant components, aiming to achieve an optimal impingement-free range of motion post-operatively. However, precisely executing the planned surgery and achieving accurate implant positioning during the procedure remains challenging, as deviations from the surgical plan can negatively impact outcomes. Previous research has used post-operative CT imaging to measure accuracy of component placement. This approach poses challenges due to metal artifacts, which impair accurate 3D imaging and subsequent error calculations, and the potential risks associated with repeated radiation exposure. Dynamic biplane video radiography (DBVR) offers a powerful alternative for assessing shoulder implant kinematics. It is minimally affected by metal artifacts, allows for dynamic evaluation of the shoulder, and its radiation exposure is significantly lower—approximately one-quarter to one-third—compared to the postoperative CT scans required by CT-based studies. We hypothesize that DVBR will reflect that the measured implant positioning post-surgery will differ from the preoperative plan.

Methods

Twelve subjects (7 males, 5 females; ages 69 ± 8.8 years; BMI 29.3 ± 7.9 kg/m²) who underwent RSA and completed a standardized post-operative protocol were enrolled in this IRB-approved retrospective study, conducted 71.4 ± 17.5 weeks postoperatively. Preoperative planning utilizing 3D CT templating software determined optimal implant positioning. Postoperative DBVR imaging was acquired with patients seated and their arms in a neutral position, resting at their sides. The manufacturer provided patient-specific 3D implant and bone models in the templated position, with local coordinate systems pre-defined according to manufacturer standards. These models were aligned to the DBVR images using DSX Suite software (HAS-motion). Glenosphere position was characterized by glenoid implant version (GV) and inclination (GI) relative to the coronal and transverse scapular planes, respectively. Humeral implant position was described by humeral implant version (HV) and inclination (HI) relative to the coronal and sagittal humeral planes, respectively. Differences between the planned and DBVR-measured implant positions were calculated for each subject and statistically analyzed using a one-sample t-test with significance level set at 0.05 and Bonferroni correction applied.

Results

Results are displayed in Figure 1. For the glenoid implant, the average difference between planned and implemented GV was 6.5 ± 3.9 deg, and the average GI was 8.1 ± 6.6 deg. For the humeral implant, the average difference between planned and implemented HV was 4.9 ± 3.6 deg, and the average HI was 5.1 ± 4.7 deg. All outcomes were statistically different from zero (GV: p < 0.01, η2 = 0.69, O. power = 0.91; GI: p < 0.01, η2 = 0.47, O. power = 0.98; HV: p < 0.01, η2 = 0.55, O. power = 0.99; HI: p < 0.01 , η2 = 0.55, O. power = 0.81).

Discussion

This study evaluated the accuracy of implant positioning in RSA using 3D templating software and DBVR. Our findings reveal significant deviations in both the glenoid and humeral component positions compared to the preoperative plan, indicating that surgical implementation is, as expected, not without errors. Use of DBVR allowed for accurate analysis of shoulder implant kinematics and position with lower patient radiation burden. More evidence is needed to establish clinically meaningful differences and understand these deviations' long-term repercussions on shoulder mobility and implant longevity.