Preoperative Planning: Patient-Specific Instrumentation

Preoperative Planning: Patient-Specific Instrumentation

Brandon J. Erickson, MD

Patrick Denard, MD

Anthony A. Romeo, MD


Anatomic total shoulder arthroplasty (ATSA) is an effective

treatment option for the management of glenohumeral

osteoarthritis and aims to provide patients with pain relief and improvement in function.1,2 Similarly, reverse total shoulder arthroplasty (RTSA) has emerged as an excellent treatment option for older patients with cuff tear arthropathy, proximal humerus fractures, patients with glenohumeral osteoarthritis and a poorly functioning or nonfunctional rotator cuff, patients with severe glenoid deformity in the setting of osteoarthritis, and other evolving indications.3,4,5 Given the success of ATSA and RTSA, the number of these procedures performed continues to rise each year, and there is every indication this trend will continue.6

While the outcomes following ATSA and RTSA are encouraging, there are still complications that occur. One of the most common complications and/or reason for a poor outcome following ATSA and RTSA is a problem with the glenoid component.7,8 Obtaining optimal exposure of the glenoid can be challenging in certain cases, and, once the glenoid is exposed, placing the component in the ideal position can also be difficult. As such, glenoid component problems can occur in the early or later postoperative setting and include loosening, migration, polyethylene dissociation, malposition leading to shoulder instability, and notching.9,10,11 Studies have cited the rate of glenoid loosening at 1.2% per year following ATSA with a recorded revision rate of 0.8% per year.8 Similarly, the rate of glenoid loosening following RTSA has been cited at 1.7% to 13% and is often attributed to malposition of the glenoid baseplate (most commonly superiorly inclined).12 Therefore, if glenoid position could be improved such that the glenoid was placed in the “optimal” position reproducibly and reliably from case to case, complications following ATSA and RTSA may decrease.

One potential method for improving glenoid component position in every case is the use of preoperative three-dimensional (3D) planning and patient-specific instrumentation (PSI).13 3D planning allows the surgeon to understand patient-specific glenoid inclination and version. PSI can then be created for each specific patient based upon preoperative advanced imaging that, in turn, will allow the surgeon to reproducibly place the glenoid in the ideal position in each case by positioning the glenoid pin in an optimal position. This chapter will discuss the evolution of PSI, its role in both ATSA and RTSA, and the results following use of 3D planning and PSI for glenoid placement in shoulder arthroplasty.


The idea of using technology to improve glenoid component position in shoulder arthroplasty began with intraoperative computer-assisted navigation.14,15 This technology was first attempted in cadaveric shoulders where pre- and postoperative computed tomography (CT) scans were obtained to help plan the glenoid placement and then to check the version and inclination of the implanted glenoid component.15 Nguyen et al compared traditional with computer-assisted glenoid placement in 16 paired cadaveric shoulders.15 The authors found the computer-assisted technique was more accurate in achieving the correct version of the glenoid component as measured on postoperative CT scans. The authors also noted that the largest errors with traditional glenoid implantation occurred during drilling and reaming and that the common error with the traditional method was to overly retrovert the glenoid. This technique was evaluated in patients by Kircher et al who performed a prospective randomized controlled trial, where 26 patients with shoulder osteoarthritis were randomized to traditional glenoid placement or intraoperative navigation glenoid placement.14 The authors measured glenoid version on preoperative CT scans for baseline values and then assessed the glenoid component position on postoperative CT scans at 6 weeks. The authors found that average change of retroversion was significantly better in the navigation group where the glenoid version of the navigation group improved from 15.4° preoperatively to 3.7° postoperatively compared to 14.4° preoperatively to 10.9° postoperatively in the traditional, nonnavigation group. However, the navigation process was aborted for six patients because of technical issues, and the authors noted operating time was significantly longer in the navigation group (170 vs
138 minutes). Unfortunately, the purpose of this study was simply to gauge the accuracy of the navigation and not to determine if placement of the components in an improved position led to improved patient outcomes. While the navigation seemed to improve placement of the components, it was not a feasible long-term solution given the technical complications and added time in the operating room (OR). A new technique that would similarly improve glenoid position but not at a great monetary or time consumption expense had to be developed.

This improvement became known as 3D planning and PSI and was first described by Hendel et al in 2012.13 Hendel et al performed a randomized prospective clinical trial in which 31 patients were randomized for glenoid component placement to either the novel 3D CT scan planning software combined with PSI or a conventional two-dimensional (2D) CT scan, with standard instrumentation. The study was appropriately powered to detect differences in glenoid position based on CT scans between groups. The authors found that PSI significantly decreased the average deviation of the glenoid component position in inclination and medial/lateral offset compared to the non-PSI group and also decreased the frequency of malpositioned implants. This study demonstrated that PSI was not only reproducible but it also improved the accuracy of glenoid component position, both of which were critical factors when evaluating the effectiveness of new technology.


Proper placement of the glenoid in ATSA and RTSA is one of the most controversial topics in shoulder arthroplasty. While PSI can allow precise and accurate glenoid placement based on a preoperative plan, this is irrelevant if the preoperative plan has the glenoid placed in a poor position. Unfortunately, there is significant debate as to the ideal placement of the glenoid in ATSA and RTSA. Some advocate for restoring normal anatomy, while others will use the patient’s current anatomy to place the glenoid component. In the author’s opinion, the ideal position of the glenoid component would be to recreate of the patient’s native version, inclination, and center of rotation. Native position can either be based on of estimates of the “normal” population or by using computer modeling of the premorbid state such as the glenoid vault model.16 Churchill et al used 172 matched pairs of cadaver glenoids in an attempt to determine average native glenoid version, inclination, and size in the general population as a baseline.17 They used 50 black male, 50 white male, 50 black female, and 22 white female cadavers aged between 20 and 30 years. The authors reported that glenoid version for all cadavers averaged 1.23° of retroversion. Interestingly, there was a statistically significant difference in average glenoid version between black and white patients (0.20° vs 2.65° of retroversion, respectively), between black and white males (0.11° vs 2.87° of retroversion, respectively), and between black and white females (0.30° vs 2.16° of retroversion, respectively). However, there was no statistical difference in glenoid version between men and women of the same race and no statistical difference in glenoid inclination based on race or sex. The obvious issues with using such estimates are that these means do not take into consideration individual anatomy and, perhaps more importantly, were based on nonpathologic specimens. An alternative to using mean values is to estimate the premorbid anatomy of the individual based on the glenoid vault method described by Iannotti and Scalise.16 Iannotti, Codsi, Scalise, and others described and validated a 3D glenoid vault model as a template to predict a normal glenoid version in patients undergoing shoulder arthroplasty.16,18,19 This method allows for determination of what the patient’s native glenoid version and inclination was prior to onset of their arthritis, thereby allowing the surgeon to recreate the patients normal anatomy at the time of surgery.

To obtain an “ideal” position, the glenoid component must be placed in proper version and inclination without removing too much bone. Chen et al performed a computational study on 25 CT-reconstructed B2 glenoids and demonstrated that version correction as low as 10° reduced glenoid bone density for glenoid fixation.20 They noted that as version correction increased, there was a gradual depletion of high-quality bone from the anterior portion of B2 glenoids. Hence, while correction of deformity may be necessary to achieve success in ATSA, removing too much subchondral bone during glenoid preparation can compromise fixation of the glenoid component. Decreasing the amount of reaming and accepting some deformity or using an augmented glenoid design are options to maintain glenoid bone stock.

Some authors have found that placing the glenoid component without correcting any deformity has led to excellent short-term outcomes. Service et al reported the 2-year results of 71 ATSAs and compared the results of those patients (n = 21) in whom the glenoid component was implanted in 15° or greater retroversion (mean ± standard deviation [SD], 20.7° ± 5.3°) with those patients (n = 50) in whom the glenoid was implanted in less than 15° retroversion (mean ± SD, 5.7° ± 6.9°).21 The authors found no difference in clinical outcomes scores between the retroverted and nonretroverted group. Furthermore, no patient in either group complained of posterior shoulder instability, and there was no difference in revision rates between the groups. While these results are interesting and worthy of discussion, the results are short term, and there is concern that posterior instability can occur over time leading to glenoid loosening.

There are differences in implant design and biomechanics between ATSA and RTSA and, in the authors’ opinion, ideal placement of the glenoid is different for each procedure. Furthermore, there are patient-specific considerations that should be incorporated into glenoid placement as no two glenoids are exactly the same. Placement of a glenoid in a patient with neutral version and inclination is very different than placement of a glenoid in patients with 15° of superior inclination and 30° of retroversion with 85% posterior humeral head subluxation.

Anatomic Total Shoulder Arthroplasty

The goal of glenoid placement in ATSA is to maximize contact between the glenoid component and the native bone while avoiding violation of the subchondral bone, to avoid overstuffing the joint, and to minimize the potential for postoperative glenoid failure from loosening through version/inclination malpositioning. Several studies have evaluated glenoid placement, and unfortunately, no consensus on placement has been reached. Karelse et al used CT scans of 152 patients undergoing ATSA to place virtual glenoid components and determine forces on these components, which could lead to the rocking horse phenomenon.22 The authors found that by using a best-fit circle based on the inferior glenoid rim rather than a best-fit circle based on the superior glenoid rim, there was a significant reduction in shear forces on the glenoid component. Hence, when preoperatively planning a case for the use of PSI, ensuring the glenoid component fits the inferior portion of the native glenoid may reduce shear forces on the implanted glenoid component.

Some studies surrounding PSI in total shoulder arthroplasty (TSA) have templated the cases to 0° of version and 0° of inclination, while many studies do not mention the templated inclination and version but rather report the difference between the preoperative plan and the postoperative outcome.13,23,24 In the authors’ opinion, if the glenoid vault method is not available to determine the patient’s premorbid anatomy, ideal positioning of the glenoid component should include retroversion of 10° or less, correcting any asymmetry in inclination (the native glenoid averages approximately 10° of superior tilt) and centering where the glenoid is positioned on the scapula.

Jun 23, 2022 | Posted by in ORTHOPEDIC | Comments Off on Preoperative Planning: Patient-Specific Instrumentation

Full access? Get Clinical Tree

Get Clinical Tree app for offline access