If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Evidence is building that a functional subscapularis improves function—specifically internal rotation tasks—following reverse total shoulder arthroplasty (rTSA). However, the optimal method for subscapularis repair during rTSA remains unknown with variable healing rates reported. This study aims to investigate the rate of and predictors for healing a lesser tuberosity osteotomy (LTO) following rTSA.
Following local IRB approval, patients with at least one-year follow-up for rTSA managed with an LTO and subsequent repair between March, 2017 and March, 2020 were retrospectively identified. Shoulders were selected for LTO repair based upon preoperative imaging and intraoperative assessment of subscapularis quality. All patients were implanted with a system consisting of a 150° or 155° (constrained) humeral neck-shaft angle and 2.5 to 4.5 millimeters (mm) of glenoid lateralization (Trabecular Metal Reverse Shoulder System; Zimmer Biomet, Warsaw, IN, USA). At a minimum of six months, radiographs were reviewed for an assessment of LTO healing by three independent reviewers. Healing was classified as displaced, fibrous union, or ossified union. For assessing predictors, the repair was considered intact if the LTO fragment was not displaced (fibrous union or ossified union).
Sixty-five rTSA with LTO repair were performed in 64 patients. These patients had an average age of 67.2 years (range: 31-81) and 36 (55.4%; 36/65) were female. At an average follow-up of 15.2 months (range: 8-38), 50 cases (76.9%; 50/65) were classified as having an ossified union. The radiographic healing could not be assessed in a single case. Of the 14 cases without ossific union, 8 (12.3%; 8/65) were displaced and 6 (9.2%; 6/65) were classified as a fibrous union. In logistic regression, only combined humeral liner height predicted LTO displacement (OR=1.4 [95% CI=1.1-1.8]; p=0.01). Humeral loosening was not found in any cases following LTO.
This analysis demonstrates that radiographic healing of LTO repair is more favorable than published rates of healing after subscapularis tenotomy or peel in the setting of rTSA. Subscapularis management with LTO provides the ability to monitor repair integrity with plain radiographs and a predictable radiographic healing rate. The integrity of subscapularis repair may be influenced by use of thicker humeral liners. Further investigation is needed to determine the functional impact of a healed subscapularis following rTSA.
. Although debatable, some authors have advocated for subscapularis repair when performing rTSA, in attempts to improve behind-the-back function, internal rotation strength and possibly decrease the rate of prosthetic instability
Choice of subscapularis management (subscapularis tenotomy, subscapularis peel, and lesser tuberosity osteotomy [LTO]) in anatomic total shoulder arthroplasty (aTSA) has not been shown to consistently affect clinical outcomes
. Healing of the subscapularis in the setting of rTSA, however, remains poorly understood. Given the altered joint position and glenohumeral kinematics of rTSA compared to aTSA, there are theoretical concerns regarding subscapularis healing potential.
Importantly, these authors reported improved functional internal rotation in patients with a healed subscapularis repair. Given the potential influence of subscapularis integrity on functional outcomes following rTSA, further studies are needed to determine the method of repair that maximizes healing. LTO for subscapularis management and repair provides theoretical advantages of subscapularis integrity following rTSA. This study aims to (a) quantify the healing rate and (b) determine predictors of healing for this unique approach. It is hypothesized that in the setting of rTSA, LTO repair will provide a superior healing rate compared to other subscapularis management techniques, and that meaningful predictors of LTO healing will be identified.
Following local institutional review board (IRB) approval, a retrospective analysis was undertaken. All patients undergoing primary rTSA with a repaired LTO by one of two shoulder and elbow fellowship-trained surgeons at a single academic institution between 2017 and 2020 with an identical rTSA implant were identified. Patients undergoing reverse total shoulder arthroplasty for fracture or revision arthroplasty were excluded. Minimum follow-up for inclusion was 6 months with plain radiographs that included an anteroposterior (AP), Grashey, Scapular-Y, and axillary lateral.
The decision to perform and repair an LTO was based upon preoperative imaging and intra-operative assessment of subscapularis integrity. Specifically, repair was only considered if > 2/3 of the subscapularis was still intact on visual inspection and the LTO could be repaired with the arm in 30° of external rotation at the side after rTSA implantation. After release of the rotator interval tissue, the LTO was created with an oscillating saw which began at the deep medial aspect of the bicipital groove moving from lateral to medial. In cases where a portion of the humeral head remained attached to the LTO fragment, this was removed prior to LTO repair to facilitate anatomic reduction to the tuberosity bed. All patients received a rTSA system consisting of 150° or 155° (constrained) humeral neck-shaft angle and 2.5 to 4.5 millimeters (mm) of glenoid lateralization (Trabecular Metal Reverse Shoulder System; Zimmer Biomet, Warsaw, IN, USA). The length of the humeral spacer/liner construct was individualized based on intraoperative stability trialing parameters. Prior to stem insertion, four drill tunnels were placed lateral to the bicipital groove. Four #5 non-absorbable sutures (FiberWire; Arthrex, Naples, FL, USA) were passed through these drill tunnels with the three most inferior sutures passed around the stem (Figure 1). After final reduction, these sutures were then passed just medial to the subscapularis tendon-bone junction and tied, securing the LTO fragment to the proximal humerus. Bone grafting was not routinely performed. Postoperatively, patients were immobilized in a sling for two weeks prior to initiating passive and active-assisted range-of-motion in forward elevation and external rotation. Patients were instructed to avoid motion behind the back (active internal rotation in adduction) for the first six postoperative weeks.
During clinical follow-up, standard radiographs were obtained at two weeks, six-eight weeks, three months, one-year, and two-years. To classify radiographic LTO healing, the categories described by Small et al were utilized
(Figure 2). Films at all time-points were reviewed independently by three reviewers (JDK, BMZ, JRH) with the most recent films used to determine the final healing classification. In cases of disagreement among raters, the consensus of 2 out of 3 observers was used to establish the LTO healing status. For those patients with displaced LTO, time to displacement was also determined.
In addition, patient and surgical characteristics were gathered to assess variables that may predict LTO healing in the setting of rTSA. This included age, gender, humeral spacer/polyethylene liner thickness, and glenoid lateralization. Within this specific rTSA system, metal spacers can be utilized in conjunction with a polyethylene liner to increase the overall humeral height. Therefore, if a metal spacer was utilized, liner size was an aggregate of the spacer size and the polyethylene thickness. For this analysis, cases were considered to have a healed LTO if there was osseous union or a fibrous union (non-displaced).
Descriptive statistics were used to describe demographic factors of the cohort. Intra-observer and inter-observer reliability were calculated using Fliess’ kappa to grade the reliability of measurement among the three observers
. To assess predictors of radiographic LTO healing, appropriate univariate statistical tests were used for continuous and categorical variables. Statistical significance was defined as p-value <0.05. To determine independent predictors of LTO repair failure, logistic regression was performed in which all tested variables were included with a forced entry. For predictive continuous variables, a receiver operating characteristic (ROC) curve was analyzed to determine a threshold that predicted LTO repair failure. Youden’s J-statistic was utilized to calculate this diagnostic threshold.
The final cohort consisted of 65 shoulders in 64 patients. The mean age of patients at the time of surgery was 67.5 years (range 31-81; Table 1) and 36 (55.4%; 36/65) were women. Mean follow-up time was 15.2 (range: 8-39) months. Only 8 cases had a displaced LTO nonunion at the time of latest follow-up (8/65; 12.3%). In a single case, the LTO could not be visualized as displaced or healed by a majority of the reviewers. Of the 56 with non-displaced LTO, 50 (50/65; 77% of entire cohort) had bony union and 6 (6/65; 9%) were fibrous union. In this early follow-up period, there was no evidence of stem subsidence or humeral loosening in any cases. Inter-observer reliability for determination of LTO healing was substantial with kappa value of 0.74 (substantial agreement), both for full classification and for dichotomous healed versus not healed (displaced). In 55 cases (84.6%; 55/65) all three reviewers independently agreed. In the remaining cases, two of the 3 reviewers agreed during independent review.
Table 1Cohort characteristics and breakdown by LTO healing.
When comparing healed versus non-healed (displaced) LTO, there was no difference in the age (67.1 vs. 67.6 years; p=0.89). Male patients were more likely to develop an LTO nonunion (24.1%; 7/29) compared to females (2.8%; 1/35; p=0.02; Figure 3). Females were more likely to develop a non-displaced fibrous union (14.3%; 5/35) than males (3.4%; 1/29; p=0.02). Patients with a LTO nonunion had a thicker humeral liner (mean 6.75 mm; STD=4.7) compared to those with an intact LTO repair (mean 2.0 mm; STD=2.4; p<0.001). No statistically significant difference was noted between males (3.1 mm; STD=4.0) and females (2.0 mm; STD=2.2; p=0.07) for humeral liner thickness. On multivariate logistic regression, overall liner height was the only significant predictor of LTO displacement with a 1 mm increase in liner height associated with a 40% increase in the odds for LTO displacement (OR=1.4 [95% CI=1.1-1.9]; p=0.01). ROC analysis (area under the curve [AUC]=0.74) demonstrated a sensitivity and specificity of radiographic failure for LTO healing of 85.7% and 55.6%, respectively, for a liner thickness of 4.5 millimeters or greater.
Subscapularis repair during rTSA remains a controversial topic with theoretical but poorly-defined advantages. Some authors suggest improved internal rotation and decreased risk for instability with subscapularis repair
. This study aimed to understand the radiographic healing rate of traditional LTO repair using a single rTSA system with a valgus humeral stem and slight lateral glenoid offset. Only 12.3% of repaired LTO in this rTSA cohort were displaced at latest follow-up. Furthermore, over three-quarters of the cohort achieved an ossified union of the LTO.
The healing rates reported in this study are substantially higher than healing following rTSA with other methods of subscapularis management. Using a Grammont-style implant, de Boer et al found that only 40% of selectively repaired subscapularis tenotomies were intact on ultrasound at a minimum of 12 months
. They were unable to detect a difference in functional scores based upon subscapularis repair or postrepair integrity. Similarly, Collin et al found a marginal healing rate (52.6%) for patients undergoing rTSA with a Grammont style implant with lateralized placement of the glenoid component via bone-graft and selective subscapularis tenotomy repair
. While they also noted no difference in patient reported outcome measures, they did find that patients with an intact subscapularis demonstrated significantly greater internal rotation with no loss of external rotation. In contrast, Erickson et al reported an 83.3% healing rate on ultrasound of a double row subscapularis peel repair following rTSA with a 135° humeral neck-shaft angle
. They also did not note a difference in patient reported outcomes of range-of-motion based upon postoperative subscapularis integrity. Lastly, Dedy et al reported on the subscapularis integrity of a cohort of patients treated with mixed subscapularis management techniques (48% LTO) and a mix of Grammont style implants
. They found that 77% of cases had a subscapularis in continuity. However, a majority of the cohort had a subscapularis that was mildly (33%) or severely (31%) attenuated. In their analysis, subscapularis integrity did not appear to impact patient-reported outcomes but did influence achievable internal rotation.
In addition to finding a high incidence of healing, this study identified that at-risk repairs may be predictable. Multivariate analysis demonstrated that polyethylene liner thickness is independently associated with LTO displacement. The majority of displaced LTO non-unions in this cohort occurred in males. However, men trended towards having an increased liner thickness. The humeral implant used in this study is characterized by a valgus neck-shaft angle, leading to relatively greater distalization with increasing liner thickness compared to a system with a varus humeral implant. Increasing the humeral liner thickness may place greater strain on the repair and ultimately lead to early failure. Conceptually, this may explain the variability in previous reports of subscapularis healing,
. Decreasing tension through smaller humeral liners in a Grammont-style system may improve LTO union rates without compromising prosthetic stability, as prior literature suggests improved stability with subscapularis repair.
The findings of this analysis must be interpreted in light of its limitations. First, this is a retrospective analysis with many confounding variables that cannot easily be controlled for, including surgeon bias toward which patients were considered for subscapularis repair, and variable indications for surgery. This complicates the assessment of predictors for displacement following LTO repair. Given the selective use of subscapularis repair, findings of this study would have limited applicability to shoulders with advanced subscapularis damage and/or contracture.
All repairs in this study were performed in a similar technique using the same humeral stem and a consistent range of glenoid offset. In the presence of reasonable subscapularis tissue quality, this method of repair produces consistent healing results. The goal of this study was to report radiographic rates of LTO union. Ossific or fibrous union of a LTO repair does not necessarily dictate a healthy and functional subscapularis tendon, and we cannot make conclusions about association of healing with clinical function. In this analysis, there was no utilization of ultrasound or computed tomography to further confirm subscapularis integrity. Other analyses have demonstrated some clinical improvement with subscapularis repair and subsequent integrity. While we note a suspected improvement in subscapularis integrity with a LTO repair in this setting, we make no effort to correlate this with clinical outcomes. We report that humeral liner thickness has a significant impact on LTO displacement. However, we did not account for the humeral stem placement or any further radiographic parameters that may dictate humeral liner thickness. Although this was a single implant system and single repair technique, findings of radiographic union may not be reproducible in systems with greater lateralization and/or varus humeral neck-shaft angles. Finally, given the retrospective nature of this study, we did not precisely control for the size or dimension of the LTO fragment. Bony surface area for healing could have a potential influence of healing rates following LTO repair. Despite this potential flaw, we were able to discern the position of the LTO in all but one case and were able to demonstrate a consistent rate of LTO healing across a nonuniform range of LTO sizes in this cohort.
In conclusion, subscapularis repair through a lesser tuberosity osteotomy has a high rate of radiographic healing compared to other repair constructs described. Furthermore, repair of the subscapularis with an LTO in the setting of rTSA allows for reproducible radiographic monitoring of the repair integrity with plain radiographs. Consideration should be given to minimizing humeral liner thickness when attempting to repair the subscapularis with a more valgus humeral stem. More research is needed to truly understand the impact of a repaired and functional subscapularis on functional outcomes following rTSA.
de Boer F.A.
van Kampen P.M.
The influence of subscapularis tendon reattachment on range of motion in reversed shoulder arthroplasty: a clinical study.
Conflicts of interest: Jay D. Keener, MD Research support – National Institute of Health; Consultant – Shoulder Innovations and Wright Medical; Royalties – Shoulder Innovations and Wright Medical; Assistant Editor - Journal of Shoulder and Elbow Surgery, Associate Editor – Seminars in Shoulder Arthroplasty: JSES
Alexander W. Aleem, MD is a paid consultant for Wright Medical.
Benjamin Zmistowski is a paid consultant for Zimmer-Biomet.
The other authors, their immediate families, and any research foundation with which they are affiliated have not received any financial payments or other benefits from any commercial entity related to the subject of this article.
Institutional Review Board (IRB) approval was obtained from the Washington University in St. Louis Institutional Review Board (Approval Date: 4/5/21; IRB#: 202104094).
Level of Evidence: Level III; Retrospective Cohort Comparison; Prognosis Study