ABSTRACT
Patient‐specific instrumentation (PSI) in shoulder arthroplasty has been used to translate preoperative surgical planning into precise implant positioning. However, screws for baseplate fixation using PSI have not been preoperatively planned or verified for proper location and length. This study aims to assess the reproducibility of the 3D‐printed PSI system for baseplate and screw positioning in reverse total shoulder arthroplasty (rTSA) and the role of preoperative screw planning. Postoperative CT data from 30 patients who underwent primary rTSA using PSI were collected. After ideal position planning of the baseplate and screws, a PSI guide was 3D‐printed. Postoperative CT evaluated baseplate version, inclination, and translation. Screw length, insertion angle, and potential penetration of the spinoglenoid and suprascapular notch were investigated. The mean differences between planned and actual implantation were 2.7° ± 5.8° for version, 0.9° ± 3.5° for inclination, and 1.0° ± 5.4° for rotation. The mean translation difference was 1.7 ± 1.0 mm. The mean screw angulation differences were −0.5° ± 6.4° anteroposteriorly and −1.4° ± 7.1° superior‐inferiorly. There was no risk of nerve injury from suprascapular notch involvement because it was considered that the screw was positioned away from the nerve path. The posterior screw was abandoned in 93.3% of patients due to proximity to the suprascapular nerve or insufficient length for bone purchase (mean length: 9.3 ± 2.0 mm). Using PSI, the reproducibility of baseplate and screw placement in rTSA was confirmed. The posterior screw has a limited role due to its length and direction constraints.
Clinical significance
Preoperative planning and PSI enable precise surgery, including proper screw insertion and baseplate positioning.
Keywords: 3D printing, baseplate and screw position, patient‐specific instrumentation, preoperative planning, reverse total shoulder arthroplasty
1. Introduction
1.1. Background
The indications for reverse total shoulder arthroplasty (rTSA) are gradually expanding from end‐stage cuff tear arthropathy to irreparable rotator cuff tear, acute proximal humerus fracture, glenohumeral arthritis with severe glenoid bone loss, and failed shoulder arthroplasty [1, 2, 3, 4, 5]. Proper location of the baseplate and firm fixation with screws are important factors for the long term survival of rTSA [6, 7]. However, there is no proper landmark except glenoid surface for the proper location of baseplate and insertion of screws during the surgery. Thus, accuracy cannot be guaranteed even if proper planning is done, since the surgeon can only see the glenoid surface, and there is variability in the anatomy of the patient [8, 9]. In this case, there is a possibility of malposition of the base plate, and the screw may unintentionally involve the spinoglenoid notch and suprascapular notch causing possible suprascapular nerve damage [10]. In addition, occasionally the screw length is not enough to obtain secure fixation [8]. Recently, several studies have been conducted to properly position the glenoid base plate using 3D printed patient‐specific instrument (PSI) and computer‐assisted surgery [11, 12, 13, 14, 15]. However, computer‐assisted navigation surgery requires a tracker mounted on the coracoid process for the registration of various landmarks on the glenoid and coracoid, and a coracoid process fracture may occur in this process [16, 17]. Several PSI systems are currently in use and reported with high accuracy [10, 11, 14, 15]. However, some studies provoked suspicion about the accuracy of preoperative planning software and patient‐specific instrumentation [10, 13]. While those studies evaluated the accuracy of the baseplate position, screws for baseplate fixation using PSI have not been planned preoperatively or verified for the proper location and length [11, 14, 15]. Also, the previously mentioned studies primarily compared conventional methods with the PSI group, not focusing on examining the reproducibility of surgical planning within the PSI group itself.
1.2. Rationale
The aim of this study was to evaluate the reproducibility and efficacy of the 3D printed patient‐specific instrument especially for the screws as well as the positioning of the baseplate and in reverse total shoulder arthroplasty. Specific attention was provided to the screws inserted by preoperative planning and 3D printed screw guide based on the planning. Our hypothesis is that the screws inserted by use of PSI would show high reproducibility in addition to the baseplate.
1.3. Study Questions
-
1.
How reproducible is the 3D printed PSI system for the positioning of the baseplate and screws in reverse total shoulder arthroplasty through planning‐post operative state distance and angle measurements?
-
2.
What is the role of preoperative planning of the screw through quantative methods?
2. Methods
2.1. Study Design and Setting
This study is a retrospective analysis (Level IV retrospective case series) of prospectively collected data.
2.2. Participants/Study Subjects
We reviewed 45 consecutive patients who underwent primary rTSA at a single University hospital using PSI with the Equinoxe Reverse System (Exactech Inc., Florida, USA) from June 2019 to July 2021. Finally, with the exclusion of 15 patients who did not undergo postoperative CT scans due to the non‐use of PSI caused by technical issues, 30 patients were enrolled in the study.
2.3. Description of Experiment, Treatment, or Surgery
2.3.1. Scapula Segmentation
Each patient had preoperative computed tomography (CT) scans. The CT DICOM file of the shoulder joint had a slice thickness of less than 1 mm with a high resolution of 512*512 pixels. Additionally, the scapula part was segmented using Materialise Mimics 22.0 software (Materialise, Leuven, Belgium). The segmented image was extracted to the 3D reconstruction STL (Stereolithography) file format. This process, which is conducted by engineer with orthopedic surgeons supervisions, was performed in the same manner for both pre‐ and post‐operative conditions.
2.3.2. Implant Planning
The baseplate was positioned on the glenoid using 3‐matic 14.0 software (Materialise, Leuven, Belgium), done by orthopedic surgeons. The baseplate was designed by reverse engineering to position on the glenoid surface. The lower margin of the baseplate was matched to be flush with the inferior margin of the glenoid in accordance with the manufacturer's instructions. After the determination of the coronal plane of the scapula based on the glenoid center point, inferior scapular angle point, and trigonum spinae point, axial and sagittal planes were defined perpendicular to coronal plane [18]. (Figure 1) The glenoid version and inclination were measured on these reference planes that intersect the center of the glenoid surface. The optimal baseplate location was planned to minimize the risk of scapula notching and avoid superior tilt [19]. Baseplate version was set to be less than 10° from the neutral version [20]. Baseplate inclination was determined to be tilted inferiorly 0°–10° [1, 21, 22]. The final location of implant was modified considering the anatomy of individual patients [17]. After the planning of the baseplate location, screws for baseplate fixation were planned for a proper angle and length. Screws were planned to obtain maximum intraosseous length without impairing the neurovascular structures [2, 23]. The superior screw is basically planned to be inserted under the base of coracoid process and penetrate the anterior far cortex of the scapula. It was planned to purchase the maximal intraosseous bone while not to involve the suprascapular notch. The anterior screw is directed posteroinferiorly for the maximal bone purchase. The inferior screw was tilted inferiorly toward the ant/sup aspect of the inferior pillar of the scapula. The posterior screw was directed toward the junction of the glenoid neck and scapular spine with maximal bone purchase. Injury to the suprascapular nerve was avoided by preventing the penetration of the glenoid posterior cortex [2].
Figure 1.
After the determination of the scapular coronal plane based on the three points of the glenoid center point (yellow dot), inferior scapular angle point (Green dot), and trigonum spinae point (red dot), the sagittal and axial planes were defined based on the plane perpendicular to the line connecting the Glenoid Center Point and the Trigonum Spinae Point.
2.3.3. Guide Manufacturing With 3D Printing
The center pin guide and screw guide were designed separately to translate the planning into the actual surgery in 3‐matic 14.0 software (Materialise, Leuven, Belgium). The center pin guide was designed to be accurately fixed to the bone of the anterior edge of the glenoid with 0.15 mm gap with the glenoid bone bed, and a hole for center pin insertion, and a slit to indicate the rotation of the baseplate were made. The screw guide was designed to guide drill for screws and placed on the real baseplate. The center pin guide and screw guide were printed with photopolymer resin SG‐100 (Graphy Inc., Republic of Korea) by use of Sindoh A1+ (Sindoh, Republic of Korea) 3D printer (Figure 2). The SG‐100 photopolymer resin is a biocompatible material that has received FDA and the Ministry of Food and Drug Safety (MFDS, Republic of Korea) approval. It has a heat distortion temperature of at least 130, making it suitable for sterilization in high‐temperature environments such as an autoclave. After the 3D printing process (3D printing—washing—curing—post processing), the final ultrasonic cleaning, packaging, and sterilization were conducted in a clean room environment, ensuring its safety and usability.
Figure 2.
(A) Baseplate and screw planning using 3‐matic 14.0 software (Materialise, Leuven, Belgium). (B) Applied 3D printed center pin guide on the anterior edge of 3D printed glenoid surface. The red bar is a slit to indicate rotation. (C) 3D printed drill guide for screws.
2.3.4. Surgical Technique
All the surgeries were performed using a deltopectoral approach in a 30° conventional beach chair position. As a routine procedure, after the resection of the humeral head and posterior displacement, the joint capsule was released, the labrum and capsule were excised to sufficiently expose the glenoid, and the cartilage was meticulously removed while retaining the osteophyte for precise fitting of the guide pin guide. The 3D printed center pin guide was located on the anterior edge of the glenoid surface, and the secure sitting was confirmed by compression with the surgeon's thumb. After inserting the guide pin, the rotation of the implant was marked by electrocautery in the rotation slit. After the Real Glenoid baseplate was inserted in consideration of the indicated rotation, the second drill guide for screws was set on the baseplate and the screw holes were drilled. The depth of the screw hole was directly measured using a depth gauge; additionally, the screw length was finally determined by comparing it with the planned value. Subsequently, the glenosphere, humeral stem, and polyethylene tray were implanted. After all the implants were inserted, the subscapularis tendon was repaired with the transosseous suture method if possible, and soft tissue was sutured layer by layer (Figure 3).
Figure 3.
(A) Glenoid exposure with the deltopectoral approach; (B) bone bed preparation with sufficiently removal of the labrum and cartilage; (C) the 3D printed guide pin guide was sitting on the anterior edge of glenoid; (D) after insertion of the guide pin, rotation was marked through the slit; (E) the real implant was seated after making the peg hole; (F) after the seating of the screw hole guide, screw drilling was performed as planned.
2.4. Variables, Outcome Measures, Data Sources, and Bias
Postoperative CT was taken on the third day after surgery. 3D reconstruction was performed after removing the metal artifacts using the Materialize Mimics 22.0 software (Materialise, Leuven, Belgium) by an engineer under the supervision of an orthopedic surgeon. After 3D reconstruction of the scapular body and baseplate, both pre‐ and postoperative scapula exactly overlapped with use of the landmarks of the coracoid process, acromion, and scapular body border. Comparisons of the planned implant and the inserted implant were conducted (Figure 4A) by an engineer and a surgeon.
Figure 4.
Evaluation of the actual inserted implant. (A) After overlapping of the pre‐ and postoperative scapula based on the landmarks of the coracoid process, acromion, and body border, the planned implant and the actually inserted post OP implant were compared. (B) Based on the AP and SI axis of the baseplate, the errors of the guide pin placement of the actual inserted baseplate compared with the plan were evaluated. (C) The rotation of the baseplate were measured using coronal plane of the scapula and the bisector line of the baseplate. (D) For screw angulation measurement, after setting the AP and SI plane perpendicular to the baseplate, AP angulation and SI angulation were measured.
2.4.1. Baseplate Evaluation
The version and inclination of the real baseplate was measured based on the same manner with those of the preoperative planning. The anteroposterior (AP) and superoinferior (SI) translation of the base plate from the planning was compared. The rotation of the base plate was determined by comparing the coronal plane of the scapula and the bisector line of the baseplate. Based on the AP and SI axis of the baseplate, the errors of the center pin and baseplate placement were evaluated (Figure 4B,C). If the baseplate position (version, inclination, and rotation) deviated less than 10° in both AP, SI plane from the planned position, the component was considered to be in the correct position. If the center pin placement deviated less than 2 mm in both AP and SI plane, the baseplate was considered as the correct translation [9, 10, 24, 25].
2.4.2. Screw Evaluation
The length of the inserted screws was measured. After setting the AP and SI plane perpendicular to the baseplate, each AP angulation and SI angulation were measured (Figure 4D). The deviation of each screw from the preoperative planning in AP and SI dimension was evaluated. If the screw placement deviated less than 10° in both AP, SI plane, considered correct position [25]. Penetration of the spinoglenoid and suprascapular notch in screw insertion were determined on the postoperative CT scan. Correlation of baseplate rotation deviation and screw length deviation from the preoperative planning of screws was evaluated.
2.5. Demographics, Description of Study Population
The mean age was 73.5 ± 8.5 years. Out of a total of 30 patients, 5 were men and 25 were women. There were 13 irreparable cuff tears, 7 cuff tear arthropathies, 4 osteoarthritis, 3 proximal humerus fractures, 2 avascular necrosis, and 1 neglected shoulder dislocation. In all cases, a small baseplate was used.
2.6. Statistical Analysis, Study Size
All the data are expressed as means ± standard deviations unless otherwise indicated. All statistical analyses were performed with SPSS Statistics 22 (IBM, Armonk, NY, USA). The absolute difference between the measurements of the planned and the implanted baseplate and screws were compared using a paired t‐test. For correlation analysis, Spearman's analysis for continuous data and Kruskal–Wallis test for categorical data were used. The data were checked for normality and homogeneity of variances. The results were considered to be significant at p value < 0.05.
3. Results
3.1. Reproducibility of 3D Printed PSI System for the Positioning of the Baseplate and Screws in rTSA
As shown in Figure 5, normal positioning of the baseplate and screw according to planning and malposition were compared.
Figure 5.
Schematic figure for both normal and malpositioned baseplate and screws.
3.1.1. Baseplate Evaluation
The mean error in version of baseplate from the preoperative plan was 2.7° ± 5.8°, in inclination 0.9° ± 3.5°, and in rotation 1.0° ± 5.4°. Among the patients, 25 patients (83.3%) were positioned correctly within a deviation of 10° in version. All the patients were positioned accurately in inclination, and 27 patients (90%) were positioned accurately in rotation. The position of the guide pin was located at mean 0.7 ± 1.1 mm anterior from the AP plane, and 0.7 ± 1.4 mm superior from the SI plane. The mean error in the guide pin distance were 1.65 ± 1.03 mm (Table 1). Among the patients, 21 patients (70%) were positioned correctly within a 2 mm deviation in both the AP and SI plane (Figure 6).
Table 1.
Results for baseplate version, inclination, and entry point translation. For the version, positive value is anteversion and for the inclination, the positive value is superior tilting.
| Base plate (n = 30) | Data | p |
|---|---|---|
| Version | ||
| Plan (°) | 0.4 ± 4.0 | |
| Real implantation (°) | −2.4 ± 6.0 | |
| Version difference from the planning to real implantation (°) | 2.7 ± 5.8 | 0.014 |
| Inclination | ||
| Plan (°) | 0.1 ± 5.9 | |
| Real implantation (°) | 0.0 ± 6.3 | |
| Inclination difference from the planning to real implantation (°) | 0.9 ± 3.5 | 0.156 |
| Rotation | ||
| Rotation difference from the planning to real implantation (°) | 1.0 ± 5.4 | 0.333 |
| Baseplate angle errora | ||
| Version error (< 10°/10° ≤) | 25 (83.3%)/5 (16.6%) | |
| Inclination error (< 10°/10° ≤) | 30 (100%)/0 (0%) | |
| Rotation error (< 10°/10° ≤) | 27 (90%)/3 (10%) | |
| Guide pin translation | ||
| Anterior offset (mm) | 0.7 ± 1.1 | |
| Superior offset (mm) | 0.7 ± 1.4 | |
| Distance (mm) | 1.7 ± 1.0 | |
The unit is the number of cases, and the number in parentheses is expressed as a percentage.
Figure 6.
A view of the guide pin deviation on AP, SI plane relative to the size of the small baseplate (Exactech Inc., Florida, USA).
3.1.2. Screw Evaluation
Overall, for the anterior‐posterior (AP) angle, 14 of 90 screws (15%) were more than 10° off plan. Additionally, for the superoinferior (SI) angle, 11 of 90 screws (12%) were more than 10° off plan. Among the 90 screws, 71 screws (78.9%) were positioned accurately within 10° deviation in both the AP and SI planes (Figure 7). The screw length error was 1.2 ± 6.2 mm. The details of the superior screw, anterior screw, and inferior screw are in (Table 2). In preoperative planning, all the screws were planned not to involve the suprascapular notch and spinoglenoid notch. However, there was one case of notch involvement (3.33%, suprascapular notch involvement).
Figure 7.
A view of screw angle deviation. In 71 screws (78.9%), the screw was inserted within 10° deviation of both the AP and SI planes.
Table 2.
Details in the measurement of the superior, anterior, and inferior screws. For the AP angle, a positive value is anterior angulation, and for the SI angle, the positive value is superior angulation.
| Screws | Data | p |
|---|---|---|
| Superior screw (n = 30) | ||
| Mean AP angle error (°) | −0.8 ± 6.5 | 0.428 |
| Mean SI angle error (°) | −1.6 ± 3.9 | 0.037 |
| Screw length difference (mm) | −0.9 ± 6.2 | 0.435 |
| AP angle error (< 10°/10° ≤) | 28 (93%)/2 (7%) | |
| SI angle error (< 10°/10° ≤) | 29 (97%)/1 (3%) | |
| Notch involvement | 1/30 (3.3%) | |
| Anterior screw (30) | ||
| Mean AP angle error (°) | −1.0 ± 6.7 | 0.403 |
| Mean SI angle error (°) | 1.0 ± 7.8 | 0.500 |
| Screw length difference (mm) | 4.4 ± 5.1 | < 0.001 |
| AP angle error (< 10°/10° ≤) | 25 (83%)/5 (17%) | |
| SI angle error (< 10°/10° ≤) | 23 (77%)/7 (23%) | |
| Notch involvement | None | |
| Inferior screw (30) | ||
| Mean AP angle error (°) | −2.4 ± 8.9 | 0.145 |
| Mean SI angle error (°) | −1.0 ± 6.6 | 0.416 |
| Screw length difference (mm) | 0.1 ± 6.2 | 0.931 |
| SI angle error (< 10°/10° ≤) | 27 (90%)/3 (10%) | |
| AP angle error (< 10°/10° ≤) | 23 (77%)/7 (23%) | |
| Notch involvement | None | |
Abbreviations: AP, anterior posterior; SI, superior inferior.
3.2. Role of Preoperative Planning of the Screw—Posterior Screw Planning
When we planned the screws for the baseplate fixation, the maximal length of the posterior screw was less than 18 mm (shortest length of screw among the available screw options) in 29 cases (96.7%). The mean length of the posterior screw was 9.3 mm (±2.0 mm). In 19 cases (66.3%), the maximum screw length was less than 10 mm (Table 3, Figure 8). Although we planned and printed the drill guide for all four screws, the posterior screw was abandoned in 28 cases (93.3%). Thus, we evaluated the accuracy of the screw guide only for the anterior, superior, and inferior screws.
Table 3.
Commercially available screws are present in 4 mm increments; additionally, the shortest screw that can be used is 18 mm. In 29 cases (96.7%), the longest screw was measured to be less than 18 mm.
| Posterior screw evaluation | Data |
|---|---|
| Mean maximal planned length (mm) | 9.3 ± 2.0 |
| Maximal planned length | |
| < 10 mm | 19 (63.3%) |
| 10 ~ less than 14 mm | 9 (30%) |
| 14 ~ less than 18 mm | 1 (3.3%) |
| ≥ 18 mm | 1 (3.3%) |
Figure 8.
After simulating the optimal baseplate and screw (sup, ant, inf) position, the length of the longest posterior screw were simulated. (A) The relationship between the usable range of the screw and the posterior glenoid of the scapula. Yellow cone: Area where the bone can be purchased based on the posterior screw hole. (B) The maximum length of the posterior screw that can be used without piercing the glenoid posterior cortex was measured to be 10 mm considering the relationship with the adjacent screw and center peg. The screw length was measured excluding the head as the manufacturer's instructions.
3.3. Other Relevant Findings
3.3.1. Correlation Analysis of Errors in Baseplate Rotation and Screw Length Deviation
There was no statistical significant correlation between the baseplate rotation deviation and screws (superior screw: p = 0.243, r = −0.220, anterior screw: p = 0.287, r = −0.201, and inferior screw: p = 0.431; r = −0.149) (Table 4).
Table 4.
Rotation error is the difference in the actual inserted baseplate rotation compared to the planned value. The screw length error is the difference between the planned value and the actual inserted screw length. All the errors were analyzed by converting them into absolute values regardless of the direction.
| Correlation analysis | p | r |
|---|---|---|
| Rotation error−screw length error | ||
| Superior screw | 0.243 | −0.220 |
| Anterior screw | 0.287 | −0.201 |
| Inferior screw | 0.431 | −0.149 |
3.3.2. Complication and Clinical Result
There were no complications reported related to the PSI procedure. Clinical results were excluded from the analysis since the purpose of this study was to analyze the reproduction of pre‐op planning by PSI on postoperative CT.
4. Discussion
4.1. Background and Rationale
The most important finding of this study was that PSI helps to perform surgery as planned for proper glenoid and screw fixation with high accuracy. Additionally, our research revealed that the posterior screw has a limited role in terms of length and direction in preoperative 3D planning. These results underscore the crucial role of PSI in achieving precise surgical outcomes, particularly in optimizing baseplate and screw placement. Currently, there are many efforts to position the baseplate as planned using PSI, but there are not many reports on 3D planning of screws and implantation using PSI. In the present study, we evaluated the reproducibility of 3D printed guide by comparing preoperative planning and postoperative of the screw using 3D aligning method. On the other hand, this study selected the stereolithography (SLA) 3D printing method, which cures photopolymerization materials using a laser light source. Several prior studies have used selective laser sintering (SLS) as a 3D printing method [10, 14, 15]. Msallem et al. reported that both SLA and SLS methods showed equally high dimensional accuracy, and that the SLA 3D printing method was a reliable choice when considering economic efficiency [26]. Therefore, the SLA 3D printing method used in this study can be said to be a reasonable method.
4.2. Reproducibility of 3D Printed PSI System for the Positioning of the Baseplate and Screws in rTSA
In this study, PSI assisted rTSA was performed including a guide pin guide and screw guide, and the accuracy was confirmed by precise measurement through 3D reconstruction. Several results have been reported for shoulder arthroplasty using PSI [3, 11, 16, 25]. Gauci et al. [27] performed PSI‐assisted total shoulder arthroplasty (TSA) in 17 cases and reported accurate baseplate position with patient‐specific guide pin guide. The guide pin translation in the horizontal plane reported −0.1 ± 1.4 mm and in the vertical plane 0.8 ± 1.3 mm, and showed a precise accuracy of mean error in version 3.4° ± 5.1° and inclination 1.8° ± 5.3°. However, the results of applying PSI to the screw were not included. In our study, the PSI for screws was included, and as a result, all the screws were able to be inserted into safe positions without involving a notch except for one case and firm fixation with sufficient length was obtained with high accuracy. It remains challenging to directly attribute improvements in clinical outcomes to the technical high accuracy achieved through PSI reproducibility demonstrated in this study. However, it is well established that complications such as scapular notching or implant loosening can result from baseplate malposition [28, 29]. In this regard, improving the reproducibility of surgical planning through PSI, which aims to optimize baseplate stability, may help achieve better clinical outcomes.
4.3. Role of Preoperative Planning of the Screw—Posterior Screw Planning
The most notable finding is that the posterior screw for baseplate fixation was too short to purchase firm fixation or at risk to injure the suprascapular nerve since the tip of screw was too close to the path of the nerve. Thereby, the posterior screws were abandoned in 28 out of 30 cases in our series. When only three screws are used for baseplate fixation, there may be concerns that the secure fixation of the baseplate is unobtainable. In another study, a biomechanics study conducted with the Cadaver Model James et al. [30] reported no difference in the displacement at initial fixation between baseplates with two and four screws and concluded that only two screws were necessary. Also, in the previous clinical studies using baseplates with a similar center peg configuration, originally designed for fixation with only two screws, have not reported issues such as glenoid component loosening [31, 32, 33]. DiStefano et al. [2] reported that, when inserting the post screw of the baseplate, the usable angle is very limited and is impossible to access the high bone stock without penetrating the posterior cortex and involving the spinoglenoid notch. In addition, the screw length plays an important role in fixation stability [23]. In our study, when the posterior screw was simulated by 3D planning, the usable length was shorter than the commercially available minimum screw length (18 mm), except for one case. Considering this, we concluded that the posterior screw is not always necessary if at least 3 screws can be inserted, especially when there is a possibility of nerve injury or very short screw length than that provided as a ready‐made product.
4.4. Other Relevant Findings
When we verified the expectation that the rotational error of the baseplate would lead to the deviation of the actual inserted screw length from the planned length, the rotation error is smaller than expected and there was no statistically significant difference between the planned baseplate rotation and the actually inserted baseplate rotation (1.0° ± 5.4°, p = 0.333). Additionally, unlike the authors' predictions, there was no correlation between the two deviation although there was some screw length mismatch (4.4° ± 5.1°, p < 0.001) between the planning and real screw insertion.
As another factor affecting the accuracy of PSI, glenoid deformity could affect the accuracy of PSI. Lau and Keith [13] performed 11 consecutive TSAs (7 TSAs and 4 reverse TSAs) using PSI guides. Among them, more than 10° anteversion or retroversion was reported in five cases (45%), suggesting that the accuracy of PSI may not be as high as previously reported. In Lau and Keith's study, the average native glenoid version of patients was 22° ± 9° retroversion and 17° ± 9° inclination, which was different from other studies, as patients with relatively high deformity. In the case of patients with high glenoid deformity, a bony spur may be unintentionally removed during the process of removing the cartilage and labrum for precise sitting of PSI guide and could affect the accuracy of PSI. In addition, error may occur in the process of correcting deformation with eccentric reaming and this is considered a factor that increases the possibility of error for reasons other than PSI itself with high deformity patients. Clavert et al. [34] reported that glenoid retroversion of 15° or more cannot be satisfactorily corrected simply by reaming through the cadaver study. Iannotti et al. [35] reported that with severe glenoid retroversion it is difficult to place the glenoid component perpendicular to the plane of the scapula by asymmetric reaming.
Among the patients included in this study, there was a patient with a relatively severe glenoid deformity due to neglected shoulder dislocation. The pre‐OP version of this patient was 15°. Additionally, the error of the planned version and the post‐OP version was 11°, and the error in the guide pin translation was also 4.42 mm. It might influence the mean version difference between the planned and real baseplate implantation (2.7° ± 5.8°, p = 0.014). However, these results do not mean that PSI is helpless in severe glenoid deformity. Without PSI, severe glenoid deformity would result in greater difficulty in the intraoperative assessment of glenoid, resulting in an inability to correct the deformity because the anatomical landmarks are deformed [35]. Additionally, the results also reported that the PSI system improved the component position in patients with more severe deformity [3]. Even considering the possibility of errors in severe glenoid deformity, it is thought that the PSI system is of great help in pre OP planning and intra OP guiding. However, factors affecting the accuracy should be recognized in the process.
4.5. Limitations
There are several limitations to the study. First, to know the position of the post OP of the screw and baseplate, the removal of metal artifacts and overlapping of the scapula are essential, and errors may appear in this process. However, the actual inserted screw length and the screw length measured by 3D reconstruction showed an error of less than 1 mm in 74 (82%) of 90 screws, confirming the precise measurement reliability. In this study, the 3D reconstruction of 1 mm slice data was measured very precisely for all patients, making it a strength of this study. Second, 3D reconstruction using CT has a limitation in that soft tissues, such as cartilage and labrum are not included. Considering this, the bony spur was retained in the planning in this study. During the operation, the soft tissue that had not been included in the CT scan was removed as much as possible so that the center pin guide was fit on to the glenoid. Additionally, the center pin guide was designed to have a 0.15 mm gap to be accurately fixed to the anterior edge of the glenoid for some residual soft tissue, which was not completely removed. Third, most of the patients in this study did not have severe bone deformity of native glenoid. Patient‐specific instrumentation with preoperative planning and a 3D printed guide would be more useful in patients with severe glenoid deformity. However, it should be noted that glenoid deformities could be the cause of errors in PSI. Fourth, a priori power analysis was not conducted in this study; therefore, it is difficult to clearly demonstrate whether the sample size ensured sufficient statistical power. However, this study was designed as an exploratory investigation focusing on validating the reproducibility of PSI. To address this limitation, future research should include an appropriate a priori power analysis to determine an adequate sample size, followed by large‐scale, multi‐institutional clinical studies to confirm the observational findings presented in this study. Lastly, the various designs resulting from different anatomical structures of patients, tolerances that can occur during the printing process, and material properties can affect accuracy. To minimize issues, the design and manufacturing processes were standardized by considering potential tolerances to set clearance in the design step, and accuracy was measured after all processes were completed. Despite this, any errors that might occur were considered inherent limitations of the PSI system, and these effects were also included in evaluating the reproducibility of the 3D printed PSI system.
5. Conclusion
PSI helps to perform surgery as planned for proper glenoid and screw fixation with high accuracy. The posterior screw has a limited role in terms of length and direction.
Author Contributions
Wonhee Lee and Woojin Yu were involved in the writing of the manuscript and in the interpretation of the results. Wonhee Lee, HwaYong Lee and Guk Bae Kim contributed to the design and implementation of the application. In‐Ho Jeon and Kyoung Hwan Koh supervised this study. All authors read and approved the final manuscript.
Ethics Statement
Asan medical center institutional review board, receipt number: S2021‐2056‐0002.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgments
This study was supported by a grant from the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health and Welfare, Republic of Korea (Grant number: HI21C0196).
Wonhee Lee and Woojin Yu are Co‐first authors.
References
- 1. Bigorre N., Lancigu R., Bizot P., and Hubert L., “Predictive Factors of Scapular Notching in Patients With Reverse Shoulder Arthroplasty,” Orthopaedics & Traumatology, Surgery & Research: OTSR 100, no. 7 (2014): 711–714. [DOI] [PubMed] [Google Scholar]
- 2. DiStefano J. G., Park A. Y., Nguyen T. Q. D., Diederichs G., Buckley J. M., and Montgomery W. H., “Optimal Screw Placement for Base Plate Fixation in Reverse Total Shoulder Arthroplasty,” Journal of Shoulder and Elbow Surgery 20, no. 3 (2011): 467–476. [DOI] [PubMed] [Google Scholar]
- 3. Hendel M. D., Bryan J. A., Barsoum W. K., et al., “Comparison of Patient‐Specific Instruments With Standard Surgical Instruments in Determining Glenoid Component Position: A Randomized Prospective Clinical Trial,” Journal of Bone and Joint Surgery 94, no. 23 (2012): 2167–2175. [DOI] [PubMed] [Google Scholar]
- 4. Smith C. D., Guyver P., and Bunker T. D., “Indications for Reverse Shoulder Replacement: A Systematic Review,” Journal of Bone and Joint Surgery. British Volume 94, no. 5 (2012): 577–583. [DOI] [PubMed] [Google Scholar]
- 5. Vance D. D., O'Donnell J. A., Baldwin E. L., et al., “Risk of Suprascapular Nerve Injury During Glenoid Baseplate Fixation for Reverse Total Shoulder Arthroplasty: A Cadaveric Study,” Journal of Shoulder and Elbow Surgery 30, no. 3 (2021): 532–537. [DOI] [PubMed] [Google Scholar]
- 6. Eraly K., Stoffelen D., Vander Sloten J., Jonkers I., and Debeer P., “A Patient‐Specific Guide for Optimizing Custom‐Made Glenoid Implantation in Cases of Severe Glenoid Defects: An In Vitro Study,” Journal of Shoulder and Elbow Surgery 25, no. 5 (2016): 837–845. [DOI] [PubMed] [Google Scholar]
- 7. Nguyen D., Ferreira L. M., Brownhill J. R., et al., “Improved Accuracy of Computer Assisted Glenoid Implantation in Total Shoulder Arthroplasty: An In‐Vitro Randomized Controlled Trial,” Journal of Shoulder and Elbow Surgery 18, no. 6 (2009): 907–914. [DOI] [PubMed] [Google Scholar]
- 8. Tarallo L., Porcellini G., and Giorgini A., et al., “The Use of Navigation in Reverse Shoulder Arthroplasty: Preliminary Report OF 11 Cases,” Orthopaedic Proceedings 102–B, no. S1 (2020): 40, 10.1302/1358-992X.2020.1.040. [DOI] [Google Scholar]
- 9. Bokor D. J., O'Sullivan M. D., and Hazan G. J., “Variability of Measurement of Glenoid Version on Computed Tomography Scan,” Journal of Shoulder and Elbow Surgery 8, no. 6 (1999): 595–598, 10.1016/S1058-2746(99)90096-4. [DOI] [PubMed] [Google Scholar]
- 10. Cabarcas B. C., Cvetanovich G. L., Gowd A. K., Liu J. N., Manderle B. J., and Verma N. N., “Accuracy of Patient‐Specific Instrumentation in Shoulder Arthroplasty: A Systematic Review and Meta‐Analysis,” JSES Open Access 3, no. 3 (2019): 117–129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Dallalana R., McMahon R., East B., and Geraghty L., “Accuracy of Patient‐Specific Instrumentation in Anatomic and Reverse Total Shoulder Arthroplasty,” International Journal of Shoulder Surgery 10, no. 2 (2016): 59–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12. Gutiérrez S., Greiwe R. M., Frankle M. A., Siegal S., and Lee W. E., “Biomechanical Comparison of Component Position and Hardware Failure in the Reverse Shoulder Prosthesis,” supplement, Journal of Shoulder and Elbow Surgery 16, no. 3 (2007): S9–S12. [DOI] [PubMed] [Google Scholar]
- 13. Lau S. C. and Keith P. P. A., “Patient‐Specific Instrumentation for Total Shoulder Arthroplasty: Not as Accurate as It Would Seem,” Journal of Shoulder and Elbow Surgery 27, no. 1 (2018): 90–95. [DOI] [PubMed] [Google Scholar]
- 14. Kwak J. M., Jeon I. H., Kim H., Choi S., Lee H., and Koh K. H., “Patient‐Specific Instrumentation Improves the Reproducibility of Preoperative Planning for the Positioning of Baseplate Components With Reverse Total Shoulder Arthroplasty: A Comparative Clinical Study in 39 Patients,” Journal of Shoulder and Elbow Surgery 31, no. 7 (2022): 1488–1498. [DOI] [PubMed] [Google Scholar]
- 15. Yung C. S. Y., Fang C., Fang E., et al., “Surgeon‐Designed Patient‐Specific Instrumentation Improves Glenoid Component Screw Placement for Reverse Total Shoulder Arthroplasty in a Population With Small Glenoid Dimensions,” International Orthopaedics 47, no. 5 (2023): 1267–1275. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Heylen S., Van Haver A., Vuylsteke K., Declercq G., and Verborgt O., “Patient‐Specific Instrument Guidance of Glenoid Component Implantation Reduces Inclination Variability in Total and Reverse Shoulder Arthroplasty,” Journal of Shoulder and Elbow Surgery 25, no. 2 (2016): 186–192. [DOI] [PubMed] [Google Scholar]
- 17. Mourad W., Wiater J. M., Wiater B. P., and Martusiewicz A., “Baseplate Options for Reverse Total Shoulder Arthroplasty,” Current Reviews in Musculoskeletal Medicine 13, no. 6 (2020): 769–775. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Kwon Y. W., Powell K. A., Yum J. K., Brems J. J., and Iannotti J. P., “Use of Three‐Dimensional Computed Tomography for the Analysis of the Glenoid Anatomy,” Journal of Shoulder and Elbow Surgery 14, no. 1 (2005): 85–90. [DOI] [PubMed] [Google Scholar]
- 19. Duethman N. C., Aibinder W. R., Nguyen N. T. V., and Sanchez‐Sotelo J., “The Influence of Glenoid Component Position on Scapular Notching: A Detailed Radiographic Analysis at Midterm Follow‐Up,” JSES International 4, no. 1 (2020): 144–150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Favre P., Sussmann P. S., and Gerber C., “The Effect of Component Positioning on Intrinsic Stability of the Reverse Shoulder Arthroplasty,” Journal of Shoulder and Elbow Surgery 19, no. 4 (2010): 550–556. [DOI] [PubMed] [Google Scholar]
- 21. De Wilde L. F., Poncet D., Middernacht B., and Ekelund A., “Prosthetic Overhang Is the Most Effective Way to Prevent Scapular Conflict in a Reverse Total Shoulder Prosthesis,” Acta Orthopaedica 81, no. 6 (2010): 719–726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Lévigne C., Garret J., Boileau P., et al., “Scapular Notching in Reverse Shoulder Arthroplasty: Is it Important to Avoid it and How?” Clinical Orthopaedics and Related Research (Springer New York LLC, 2011), 2512–2520. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Roche C., DiGeorgio C., Yegres J., et al., “Impact of Screw Length and Screw Quantity on Reverse Total Shoulder Arthroplasty Glenoid Fixation for 2 Different Sizes of Glenoid Baseplates,” JSES Open Access 3, no. 4 (2019): 296–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Denard P. J., Provencher M. T., Lädermann A., Romeo A. A., Parsons B. O., and Dines J. S., “Version and Inclination Obtained With 3‐Dimensional Planning in Total Shoulder Arthroplasty: Do Different Programs Produce the Same Results?,” JSES Open Access 2, no. 4 (2018): 200–204. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Verborgt O., Hachem A. I., Eid K., Vuylsteke K., Ferrand M., and Hardy P., “Accuracy of Patient‐Specific Guided Implantation of the Glenoid Component in Reversed Shoulder Arthroplasty,” Orthopaedics & Traumatology, Surgery & Research: OTSR 104, no. 6 (2018): 767–772. [DOI] [PubMed] [Google Scholar]
- 26. Msallem B., Vavrina J. J., Beyer M., et al., “Dimensional Accuracy in 3D Printed Medical Models: A Follow‐Up Study on SLA and SLS Technology,” Journal of Clinical Medicine 13, no. 19 (2024): 5848. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Gauci M. O., Boileau P., Baba M., Chaoui J., and Walch G., “Patient‐Specific Glenoid Guides Provide Accuracy and Reproducibility in Total Shoulder Arthroplasty,” Bone & Joint Journal 98–B (2016): 1080–1085. [DOI] [PubMed] [Google Scholar]
- 28. Walch G. and Keith P. Boileau, eds. Shoulder Arthroplasty, (Springer, 1999), 10.1007/978-3-642-58365-0. [DOI] [Google Scholar]
- 29. Friedman R. J., Barcel D. A., and Eichinger J. K., “Scapular Notching in Reverse Total Shoulder Arthroplasty,” Journal of the American Academy of Orthopaedic Surgeons 27, no. 6 (2019): 200–209, 10.5435/jaaos-d-17-00026. [DOI] [PubMed] [Google Scholar]
- 30. James J., Allison M. A., Werner F. W., et al., “Reverse Shoulder Arthroplasty Glenoid Fixation: Is There a Benefit in Using Four Instead of Two Screws?,” Journal of Shoulder and Elbow Surgery 22, no. 8 (2013): 1030–1036. [DOI] [PubMed] [Google Scholar]
- 31. Castagna A., Delcogliano M., de Caro F., et al., “Conversion of Shoulder Arthroplasty to Reverse Implants: Clinical and Radiological Results Using a Modular System,” International Orthopaedics 37, no. 7 (2013): 1297–1305, 10.1007/s00264-013-1907-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Young S. W., Everts N. M., Ball C. M., Astley T. M., and Poon P. C., “The SMR Reverse Shoulder Prosthesis in the Treatment of Cuff-Deficient Shoulder Conditions,” Journal of Shoulder and Elbow Surgery 18, no. 4 (2009): 622–626, 10.1016/j.jse.2009.01.017. [DOI] [PubMed] [Google Scholar]
- 33. Martinez A. A., Calvo A., Bejarano C., Carbonel I., and Herrera A., “The Use of the Lima Reverse Shoulder Arthroplasty for the Treatment of Fracture Sequelae of the Proximal Humerus,” Journal of Orthopaedic Science 17, no. 2 (2012): 141–147, 10.1007/s00776-011-0185-5. [DOI] [PubMed] [Google Scholar]
- 34. Clavert P., Millett P. J., and Warner J. J. P., “Glenoid Resurfacing: What Are the Limits to Asymmetric Reaming for Posterior Erosion?,” Journal of Shoulder and Elbow Surgery 16, no. 6 (2007): 843–848, https://www.sciencedirect.com/science/article/pii/S1058274607003886. [DOI] [PubMed] [Google Scholar]
- 35. Iannotti J. P., Greeson C., Downing D., Sabesan V., and Bryan J. A., “Effect of Glenoid Deformity on Glenoid Component Placement in Primary Shoulder Arthroplasty,” Journal of Shoulder and Elbow Surgery 21, no. 1 (2012): 48–55. [DOI] [PubMed] [Google Scholar]