INTRODUCTION
The incidence of total shoulder arthroplasty and reverse total shoulder arthroplasty (RTSA) continues to increase rapidly in the United States, with a five-fold increase over the past decade [1]. As the incidence of arthroplasty has increased, the understanding of implant biomechanics and fixation techniques has improved, and the indications for the procedure have expanded. However, the burden of associated complications and revision operation has also increased [2-4]. Additionally, the design of humeral component has undergone rapid changes with the recent advent of short and stemless humeral components (SHCs).
Several researchers have described several complications associated with stemmed humeral components, including intraoperative complications (malpositioning of components, metaphyseal-diaphyseal humerus fracture, loss of bone stock) and postoperative complications (stress shielding of proximal humerus, periprosthetic fracture, loosening, migration) [5-12]. These complications, as well as the focus on bone preservation and the anticipation of revision surgery, have prompted the development of SHCs.
The advantages of stemless designs include the preservation of humeral bone stock, reduced periprosthetic fracture with the elimination of broaching, more flexibility in reconstruction in cases of altered anatomy such as post-traumatic malunion, less complex revision surgery and the avoidance of stress shielding [13-16]. However, SHCs also have several disadvantages. If there is any doubt about the bone quality of the resected humeral head, a stemmed humeral component must be used to provide stability to the diaphyseal engagement; therefore, careful attention must be paid to patient selection [17]. Additionally, because the level of the humeral neck cutting is higher than when using a stemmed humeral component, glenoid exposure is challenging, and therefore precise implant positioning, soft-tissue release, and retractor placement are important [18]. Other possible disadvantages include the theoretical risk of component loosening before osseous ingrowth has occurred.
Because of the disadvantages surrounding SHCs, several surgeons are hesitant to use SHCs. However, since the introduction of the stemless design in Europe in 2004, a growing number of studies have reported good clinical and radiographic outcomes at mid-term to long-term follow-ups (5–9 years) [14,19-23]. As the SHCs were introduced much later in Korea than Europe, reporting mid- to long-term follow-up clinical outcomes is not yet possible; however, the short-term follow-up results are meaningful.
The purpose of this study was to describe the radiographic and clinical outcomes of stemless RTSA after a minimum of 2-years follow-up. The hypothesis of this study was that stemless RTSA yields good radiographic and clinical outcomes.
METHODS
This study was approved by the Institutional Review Board in School of Medicine, Daegu Catholic University (No. CR-24-035). Informed consent was obtained from all patients in this study. The patients agreed to the study publication, including the use of radiographic images.
Between July 2018 and March 2023, 50 patients underwent 50 stemless RTSAs with Lima SMR glenoid and humeral component (Lima Corporate). Twenty-eight patients (10 males, 18 females) with a follow-up of more than 2 years were retrospectively reviewed. The average age was 71.9 years (range, 61–85 years), and the average follow-up period was 2.2 years (range, 2.0–5.1 years). The average follow-up period of the 22 excluded patients was 1.2 years (range, 0.02–1.9 years). The indications for stemless RTSA were cuff tear arthropathy (n=15), rotator cuff retear (n=3), massive rotator cuff tear (n=9) and rheumatoid arthritis (n=1) (Table 1). The exclusion criteria were as follows: (1) revision arthroplasty and (2) intraoperative component loosening due to insufficient resistance of the trabeculae of proximal humerus.
The bone marrow density (BMD) of the proximal humerus was measured before the stemless RTSA. Patients wore a 30º abduction brace for four weeks after the surgery. Isometric grip exercise and active motion of the elbow and wrist joints while wearing an abduction brace were started 3 days to 1 week after the surgery. Passive forward flexion exercise of the shoulder joint was conducted at the 1-month follow-up. Active forward flexion exercise of the shoulder joint and activities of daily living were performed at the follow-up of 2 to 3 months, depending on the patient’s compliance and clinical evaluation measured in the outpatient clinic.
Preoperative Evaluation of Bone Quality
To evaluate the bone quality of the proximal humerus preoperatively, BMD was measured before performing stemless RTSA. Preoperative BMD was measured in 23 of 28 cases. We arbitrarily divided the measured region into region 1 (R1), region 2 (R2), and region 3 (R3) (Fig. 1). R1 includes the lateral region of the scapula, acromion, humeral head and proximal humerus. R2 includes the humeral head and glenoid surface, and R3 includes the lateral 1/2 of the humeral head. Before performing the surgery, we referenced the BMD values in all regions, particularly the value for R2, which measures the entire humeral head. While no absolute standard is available, we considered that if the value was 0.5 g/cm2 or higher, the bone quality would be sufficient to perform stemless RTSA.
Surgical Technique
The surgical approach mainly involved the deltopectoral approach. After palpating the coracoid process, an 8 to 10 cm longitudinal skin incision was made in the downward direction from a point approximately 1 cm lateral to the coracoid process. The cephalic vein was retracted laterally, and the region between the deltoid and pectoralis major muscle was approached; the humeral attachment of the pectoralis major was then partially detached. The biceps long head tendon was fixed to the surrounding soft tissue, the proximal portion of the tendon was removed, and the subscapularis tendon was peeled off. After dislocation of the humeral head, humeral head cutting was conducted based on 20º of retroversion. Next, glenoid retractors were placed on the posteroinferior, posterosuperior, and anterior sides of the glenoid, and the entire labrum was removed (Fig. 2). After glenoid reaming was performed until the cancellous bone of the glenoid was exposed, we inserted a baseplate into the glenoid. A 40-mm polyethylene glenoid component was then placed on the baseplate to increase the range of motion (ROM).
After positioning the glenoid component, we dislocated the resected humeral head and prepared for insertion of the SHC. A thumb test was first performed to check sufficient resistance of the trabeculae (Fig. 3) [24]. We classified the state of bone preservation into the following three categories: wood-like bone, soft bone, and sponge-like bone. We defined a bone that is compressed with high force as wood-like bone, one that is compressed with moderate force as soft bone, and one that is easily compressed with minimal force as sponge-like bone. In the case of wood-like bone, a SHC was inserted without concern for instability; in cases of sponge-like bone, a stemmed humeral component was inserted. If the bone was judged to be soft after the thumb test, a cancellous bone graft was first performed on the proximal humerus. After repeated checking of the humeral bone stock, the SHC was inserted while preserving the cortical ring of the humeral cross section (Fig. 4).
The stability and motion of the SHC were evaluated using fluoroscopy in the operating room. During fluoroscopy evaluation, we checked the ROM in various directions (forward flexion, internal rotation, external rotation, abduction, adduction). If the SHC was loosened or pulled out, it was replaced with a stemmed humeral component (Fig. 5). The subscapularis tendon was repaired to ensure balanced stability in all patients (Fig. 6).
Clinical and Radiographic Evaluation
Clinical results were documented based on ROM, visual analog scale (VAS) score, Korean Shoulder Scoring system (KSS), and American Shoulder and Elbow Surgeons (ASES) score. Radiologic evaluation was performed for stem alignment, radiolucency, subsidence, and component stability. The assessment of the radiolucent line (RLL) around the SHC was performed in the true anteroposterior (AP) and axillary views by dividing the implant–bone interface in four different zones. The interval between each zone was set at 45º. The AP view was divided into zones 1 to 4, and the axial view was divided into zones 5 to 8. The grade of radiolucency was evaluated as follows: no sclerosis (grade 0), sclerosis and less than 2-mm space (grade 1), sclerosis and more than 2-mm space (grade 2), and radiologic loosening of the component (grade 3) [25].
Postoperative complications including revision, scapular notching, subluxation, dislocation, infection, periprosthetic fracture (intraoperative and postoperative), osteolysis and loosening were investigated at the follow-up. The radiological classification was used to quantify scapular notching, identifying four grades according to the amount of osteolysis [26]. Stem loosening was as defined as RLL of 2 mm or more in 3 zones [27]. Clinical and radiographic assessments were performed at the follow-up of 3 months, 6 months, and 1 year and annually thereafter. A computed tomography scan was performed at the follow-up of 1 year for component evaluation. All data for the study were collected on the basis of a normal standardized clinical investigation.
RESULTS
Clinical Outcomes
When comparing the ROM before surgery and at the final follow-up, significant increases were observed compared with preoperatively in forward flexion (from 112.0º to 162.5º, P<0.01) and internal rotation (from L3 to T12 level, P<0.05), respectively. No significant changes were observed in external rotation (from 43.0º to 45.2º, P=0.762). There was a statistically significant decrease in VAS score at the final follow-up compared with preoperative values at resting, motion and night: from 1.5 to 0.2, from 3.6 to 0.5, and from 2.8 to 0.4, respectively (P<0.01). KSS and ASES scores also significantly increased at the final follow-up, from an average of 64 to 93 (P<0.01) and 17.5 to 27.3 (P<0.01), respectively (Table 2).
Radiographic Outcomes and Complications
RLLs of less than 2 mm were observed in all cases. Most RLLs were found in zone 1 and 2. In zone 1, 26 cases of grade 1 and 2 cases of grade 2 were found at the final follow-up. In zone 2, 28 cases of grade 1 were found. In an excluded case, the SHC was pulled out at the follow-up of 1 week, and a revision was implemented to change it to a stemmed humeral component. The reason was chronic anterior dislocation with huge anterior glenoid defect. Appropriate stability was achieved by performing a glenoid bone graft during revision (Fig. 7). Complications including revision, subluxation, dislocation, infection, periprosthetic fracture (intraoperative and postoperative), osteolysis and loosening were not observed. However, scapular notching was observed in 18 cases. Scapular notching was classified as grade 1 in 15 cases and grade 2 in three cases (Fig. 8).
DISCUSSION
Our study purposed to evaluate the SHC based on radiographic and clinical results at short-term follow-up (a minimum of 2 years). We obtained encouraging radiographic and clinical outcomes. However, this study showed different results from other studies regarding the incidence of postoperative complications. In our study, we observed 18 cases of scapular notching (64.3%). Ballas and Béguin [22] conducted a mid-term study for stemless RTSA. The authors [22] reported one case of metaphyseal-diaphyseal humeral bone crack (1.7%), one case of revision surgery due to early instability (1.7%), and five cases of scapular notching (8.9%). Similar to our study, Nabergoj et al. [28] conducted a short-term study with a minimum 2 years’ follow-up. Scapular notching was observed in 28 cases (24.3%), humeral loosening in five cases (4.3%), and glenoid loosening in four cases (3.5%). The total complication rate was 17.4%, and revision was performed in eight cases. The rate of other postoperative complications in our study was similar or lower compared with other studies, but the rate of scapular notching was higher. According to the Sirveaux classification [26], grade 1 is defined as scapular notching is limited to the glenoid pillar, and grade 2 is defined as when it reaches the inferior screw; grade 3 is defined as when the scapular notching occurs over the inferior screw, and grade 4 is defined as when it extends under the baseplate. When analyzing scapula notching in our study, grade 1 showed the highest rate (15 cases, 53.6%), followed by grade 2 (3 cases, 10.7%). Grades 3 and 4 were not observed.
Scapular notching is associated with RTSA with a variable rate of 4.6%–50.8% and up to 96% [29,30]. It can be classified into two different types, mechanical and biological notching. Mechanical notching occurs when the humeral liner contacts the scapular pillar, while biological notching is a chronic foreign-body reaction resulting from the formation of polyethylene debris [31-33]. Some authors have described grades 1 and 2 as primarily due to mechanical notching, whereas grades 3 and 4 are more likely to be due to a biological reaction [32]. Standard SHCs allow for improved glenoid exposure compared with designs that require more limited head resection [32]. However, designs that do not require a standard humeral neck cut can also make glenoid exposure more technically challenging with the increased importance of precise positioning, soft-tissue release, and retractor placement [18]. In this study, the humeral neck cutting was performed at a higher position compared with the standard humeral neck cutting in all cases (high neck cutting). High neck cutting has the disadvantage of limiting glenoid exposure and making glenoid component fixation more difficult. Considering that all scapular notching that occurred in this study was grade 1 or 2, which is a result of mechanical notching, we speculate that the scapular notching was a result of improper glenoid component positioning due to limited glenoid exposure caused by high neck cutting. However, the high neck cutting may result in a lateralization effect of the humeral component by performing more bone preservation. Therefore, when performing stemless RTSA, the surgeon should be careful to obtain appropriate glenoid component positioning.
In previous studies, RLL was not found or was observed in very small proportions [14,24,34,35]. Beck et al. [14], Churchill et al. [24], and Huguet et al. [34] reported no RLL in their research. Krukenberg et al. [35] described one incomplete RLL. In contrast, RLL of less than 2 mm was observed in all cases in this study. In a cadaveric study, Hudek et al. [36] described that RLL around SHC may be a radiation artifact due to radiation scatter rather than true bone loss or stress shielding, and the imaging artifacts need to be considered in the follow-up evaluation of the SHC. Moursy et al. [37] reported that RLL less than 2 mm does not influence mid-term clinical results. In addition, Habermeyer et al. [19] stated that although the density of cancellous bone in greater tuberosity was reduced in the AP view in 34.9% of 78 patients, there was no clinical influence compared to the group of patients who did not show such changes. However, it is not easy to overlook our results showing that the RLL is concentrated in zone 1 and zone 2, because the clinical significance of metaphyseal stress shielding and relative tuberosity resorption after stemless arthroplasty has not yet been clearly elucidated, and the results of long-term follow-up studies have not yet been published. Considering these problems, we harvest cancellous bone from the resected humeral head and place it into the greater tuberosity before inserting the SHC.
To evaluate the bone quality of proximal humerus preoperatively, we measured BMD before stemless RTSA. We considered that the presence or absence of osteoporosis would be proportional to the values obtained from the proximal humerus. However, when using the independent samples t-test according to the presence or absence of osteoporosis, no correlation was found in all areas: R1 (P=0.158), R2 (P=0.302), and R3 (P=0.431). Thus, the BMD values in R1, R2, and R3 are independent of the presence or absence of osteoporosis. Considering the above results, even if osteoporosis is not observed in the spine or hip, the evaluation of bone quality of the proximal humerus is necessary before performing stemless RTSA. Additionally, we found that males had larger values than females in R1 and R2, but no difference was observed in R3 (R1 [P<0.01], R2 [P<0.05], and R3 [P=0.112]). Based on the above outcomes, we propose that gender should be taken into consideration when considering the surgical indication for performing stemless RTSA. In addition, as previously mentioned, R1 and R2 include the entire humeral head, but R3 includes only the lateral 1/2 of the humeral head, which may be the reason for the above results. While there is no absolute standard for BMD values in the proximal humerus and no correlation with osteoporosis, we speculate that preoperative BMD can be used as a reference for performing stemless RTSA. Verification through future research is necessary.
To the best of our knowledge, this study is the first report of stemless RTSA using an inlay design with a minimum 2-year follow-up in Korea. However, this study has several limitations. First, due to the short follow-up period after RTSA, there was a lack of sufficient research on radiographic and clinical evaluation. Second, the number of cases was small, so sufficient research results were not obtained. Finally, because this was a retrospective study, there is a possibility of errors in patient selection and low reliability of the outcomes.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Full text via PMC
Download Citation
Print
