PURPOSE. To validate the efficacy and accuracy of a novel patient specific guide (PSG) and instrumentation system that enables minimally invasive (MI) short stemmed total shoulder arthroplasty (TSA). MATERIALS AND METHODS. Using Amirthanayagam et al.'s (2017) MI posterior approach reduces incision size and eliminates subscapular transection; however, it precludes glenohumeral dislocation and the use of traditional PSGs and instruments. Therefore, we developed a PSG that guides trans-glenohumeral drilling which simultaneously creates a humeral guide tunnel/working channel and glenoid guide hole by locking the bones together in a pre-operatively planned pose and drilling using a c-shaped drill guide (Figure 1). To implant an Affinis Short TSA system (Mathys GmbH), novel MI instruments were developed (Figure 2) for: humeral head resection, glenoid reaming, glenoid peg hole drilling, impaction of cruciform shaped humeral bone compactors, and impaction of a short humeral stem and ceramic head. The full MI procedure and instrument system was evaluated in six cadaveric shoulders with osteoarthritis. Accuracy was assessed throughout the procedure: 1) PSG physical registration accuracy, 2) guide hole accuracy, 3) implant placement accuracy. These conditions were assessed using an Optotrak Certus tracking camera (NDI, Waterloo, CA) with comparisons made to the pre-operative plan using a registration process (Besl and McKay, 1992). RESULTS. 3D translational accuracy of PSG physical registration was: humeral PSG- 2.2 ± 1.1 mm and scapula PSG- 2.5 ± 0.7 mm. The humeral and scapular guide holes had angular accuracies of 6.4 ± 3.2° and 8.1 ± 5.1°, respectively; while the guide hole positional accuracies on the articular surfaces (which will control bone preparation translational accuracy) were 2.9 ± 1.2 mm and 2.8 ± 1.3 mm. Final implantation accuracy in translation was 2.9 ± 3.0 mm and 5.7–6.8 ± 2.2–4.0° across the implants’ three rotations for the humerus and in translation was 2.8 ± 1.5 mm and 2.3–4.3 ± 2.2–4.4° across the implants’ three rotations for the scapula (Figure 3). DISCUSSION. The overall implantation accuracy was similar to results of previously reported open, unassisted TSA (3.4 mm & 7–12°, Hendel et al., 2012, Nguyen et al., 2009). Analysis of the positional PSG registration accuracy very closely mirrors the final implantation accuracy (humerus:2.2 mm vs 2.9 mm, and scapula:2.2 mm vs 2.8mm), thus, this is likely the primary predictor of implantation accuracy. Furthermore, the greatest component of PSG registration error was mediolateral translation (i.e. along the guiding axis) and thus should not affect guide hole drilling accuracy. The drilled guide hole positional and angular error was low for the humerus (2.9 mm and 6.4°) but somewhat higher in rotation (8.1°) for the glenoid which may indicate a slight shift in the PSG prior to guide hole drilling due to the weight of the arm applied when the PSGs are locked together. In conclusion, this work has detailed the step-by-step surgical errors associated with the developed system and demonstrated that it achieves similar accuracy to open, unassisted TSA, while avoiding complications related to muscular transection and dislocation. Therefore, we believe this technique is worthy of clinical investigation
