Do subscapularis integrity and posterosuperior cuff tear severity affect scapular impingement and joint stability during external rotation in lateralized reverse total shoulder arthroplasty?

Abstract

Aims

The biomechanical effects of varying rotator cuff tear severities on medialized versus lateralized reverse total shoulder arthroplasty (RTSA) remain unclear. This study aimed to compare medialized and lateralized RTSA designs based on subscapularis integrity and the severity of posterosuperior cuff tears.

Methods

A total of 12 human in vivo experimental datasets were collected during external rotation (ER) from a neutral position to 45°, with the elbow fixed at 90° and the palm facing inward. These datasets were used as inputs for musculoskeletal shoulder models of both medialized and lateralized RTSA. Inverse dynamic simulations were conducted under two subscapularis conditions—repaired (all bundles intact) and torn (all bundles torn)—across three stages of posterosuperior cuff tear severity. The scapular notching-related impingement stress, joint subluxation, and muscle-tendon forces were compared between the two RTSA configurations.

Results

Subscapularis repair in lateralized RTSA led to greater reductions in impingement stress (16.7% to 26.2%; p < 0.05) and joint subluxation (27.5% to 58.9%; p < 0.001) compared to medialized RTSA (14.1% to 18.6%; p < 0.05 and 14.7% to 22.4%; p < 0.001, respectively) across all posterosuperior cuff tear conditions. Additionally, with subscapularis repair, posterior deltoid force increased more markedly with tear severity in lateralized RTSA (from 2.6% (p = 0.007) to 918.5% (p < 0.001)) than in medialized RTSA (from 4.7% to 784.5%, both p < 0.001).

Conclusion

Lateralizing RTSA and repairing the subscapularis can significantly reduce scapular notching-related impingement, improve joint stability, and support ER torque generation by increasing posterior deltoid force in patients with posterosuperior cuff tears. These findings provide critical insights for surgical planning, supporting the use of a lateralized implant and subscapularis repair to reduce the risk of scapular notching and joint subluxation while enhancing ER torque generation.

Cite this article: Bone Joint Res 2026;15(2):135–147.

Article focus

• Comparing the biomechanical effects of medialized and lateralized reverse total shoulder arthroplasty (RTSA) on scapular notching-related impingement stress, joint stability, and muscle-tendon forces across varying levels of subscapularis integrity and posterosuperior cuff tear severity.

• Quantifying impingement stress and joint subluxation to determine which RTSA configuration, combined with subscapularis status, more effectively minimizes scapular notching-related impingement and improves joint stability.

• Exploring how varying rotator cuff tear conditions influence scapular notching risk, joint stability, and muscle-tendon forces in medialized versus lateralized RTSA.

Key messages

• Lateralizing RTSA and repairing the subscapularis can reduce impingement stress and joint subluxation, even in the presence of posterosuperior cuff tears.

• Subscapularis repair in lateralized RTSA improves external rotation (ER) torque generation by increasing posterior deltoid force.

Strengths and limitations

• This study provides a comprehensive overview of the interaction between RTSA design, subscapularis status, and cuff tear severity. It highlights the potential benefits of lateralized RTSA with subscapularis repair in mitigating impingement, improving joint stability, and enhancing ER torque generation.

• Evaluating joint behaviour under a single motion task may not fully capture the complex subluxation mechanisms encountered in daily activities. Further research should incorporate a broader range of dislocation-prone movements (e.g., extension, adduction, and internal rotation) to provide a more comprehensive understanding and evaluation of joint stability in RTSA.

Introduction

Recent strategies to address the limitations of Grammont’s reverse total shoulder arthroplasty (RTSA) design increasingly favour lateralization of both the glenoid and humerus.¹ Glenoid lateralization reduces scapular notching and improves joint stability,^2,3 while also restoring native rotator cuff tension to enhance external rotation (ER) function.^4,5 Humeral lateralization similarly improves rotator cuff tension, increases deltoid wrapping, and enhances joint compression, further stabilizing the glenohumeral joint.^6,7 However, lateralization introduces specific biomechanical challenges. Shifting the glenoid centre of rotation (COR) laterally increases joint shear forces during abduction, potentially elevating the risk of baseplate failure.⁸ In addition, isolated humeral lateralization does not adequately prevent impingement or scapular notching.⁹ Despite these concerns, lateralized RTSA is now indicated for a broader range of shoulder pathologies, regardless of rotator cuff status.^10-13 This expanded use has reignited debate over the necessity of subscapularis repair in lateralized RTSA.

Biomechanical studies report varied outcomes following subscapularis repair in lateralized RTSA. Subscapularis deficiency has been shown to increase scapular notching-related impingement (i.e., mechanical contact between the posterosuperior scapular neck and humeral liner) by 82.3% and impingement-related subluxation by 21% to 25% during ER, potentially leading to scapular notching and joint instability.^14,15 Conversely, other studies found no significant impact on ER function with or without subscapularis repair.^16,17 While previous studies have explored functional outcomes, the influence of subscapularis repair on scapular notching-related impingement stress, joint subluxation, and ER function in the context of varying posterosuperior cuff (supraspinatus and infraspinatus) tear severities remains unclear in both medialized and lateralized RTSA. Therefore, this study aimed to compare the effects of medialized and lateralized RTSA on impingement stress, joint subluxation, and muscle-tendon forces, based on subscapularis integrity and the severity of posterosuperior cuff tears. We hypothesized that lateralized RTSA would exhibit greater reductions in the impingement stress and joint subluxation during ER due to increased subscapularis force, compared to medialized RTSA.

Methods

The overall simulation process for the musculoskeletal model is illustrated in the workflow (Figure 1). First, each six-degrees-of-freedom (6-DOF) musculoskeletal shoulder model for both medialized and lateralized RTSA was developed by incorporating each prosthetic bone geometry into a previously validated anatomical shoulder model.¹⁸ Second, 12 human in vivo experimental datasets were used as consistent inputs for both RTSA models. Third, segment scaling was performed using parameter optimization to minimize discrepancies between model and experimentally recorded marker trajectories.¹⁸ Fourth, the lateralized RTSA model was validated by comparing impingement stress and joint subluxation against results from a validated finite element model.¹⁹ Finally, impingement stress, joint subluxation, and muscle-tendon forces were compared between the medialized and lateralized RTSA models under both subscapularis-repaired (all bundles intact) and subscapularis-torn (all bundles torn) conditions, across three stages of posterosuperior cuff tear severity.

Fig. 1.

Workflow of the overall simulation process. FDK, force dependent kinematics; F^(FDK), FDK residual force; α^(FDK), FDK translations at the glenohumeral joint; RTSA, reverse total shoulder arthroplasty; q, joint angle; q′, angular velocity; q″, angular acceleration.

Experimental protocol

A total of 12 male participants (mean age 24.7 years (SD 2.3); mean weight 76.8 kg (SD 7.9); mean height 174.7 cm (SD 6.1)), with no history of upper limb injuries, participated in the experiments after signing an informed consent document approved by the Institutional Review Board of Sogang University. The average range of motion in ER for lateralized RTSA patients was between 40° and 50°.^20-23 Thus, an experimental task involving 45° of ER was conducted, as this movement is one of the primary clinical assessments used to evaluate functional outcomes in patients who have undergone RTSA, and is commonly associated with osseous impingement and joint subluxation.^17,24,25 All participants confirmed right-hand dominance and performed 45° of ER from a neutral position, with the elbow fixed at 90° and the palm facing inward (Figure 2). ER movement was captured using a 10-camera, 3D motion analysis system (1 Raptor and 9 Eagle; Motion Analysis Corp, USA) at 400 Hz. Retro-reflective markers were attached to the pelvis, thorax, clavicle, head, scapula, humerus, and right-hand middle knuckle.¹⁸ Marker trajectory data were processed using a zero-lag, fourth-order low-pass Butterworth filter with a 10 Hz cutoff frequency.¹⁸

Fig. 2.

Illustration of 45° external rotation in the musculoskeletal shoulder model of reverse total shoulder arthroplasty.

Surgical technique

Virtual implantation of prostheses was performed in SolidWorks 2024 (SolidWorks Corp, USA), following standard surgical guidelines for medialized (Lima SMR System; Lima Corp, Italy) and lateralized (Coralis Reverse Shoulder System; Corentec, South Korea) RTSA. Glenospheres for both designs were positioned at the centre of each subject’s scaled-generic glenoid.²⁶ Each humeral head was uniformly resected at a 20° retroversion angle for both stem designs, with neck-shaft angles (NSAs) of 135° for the onlay and 150° for the inlay configurations, respectively.²⁷ The medialized RTSA configuration included a 36 mm humeral liner with an inlay design and 150° NSA (Figure 3a), and a 36 mm glenosphere and baseplate without lateralization (Figure 3a). The lateralized RTSA featured a 36 mm humeral liner with an onlay design and 135° NSA (Figure 3b), and a 36 mm glenosphere and baseplate with a 3 mm lateral offset from the glenoid surface (Figure 3b).

Fig. 3.

Illustrations of reverse total shoulder arthroplasty (RTSA) designs. a) Medialized RTSA: 36 mm humeral liner and inlay design with a 150° neck-shaft angle of the humeral stem, along with a 36 mm glenosphere and baseplate without lateralization. b) Lateralized RTSA: 36 mm humeral liner and onlay design with a 135° neck-shaft angle of the humeral stem, along with a 36 mm glenosphere and baseplate with a 3 mm lateral offset from the glenoid surface.

Musculoskeletal modelling

The 6-DOF musculoskeletal shoulder models for both RTSA configurations were developed by incorporating prosthetic geometries into a validated anatomical shoulder model (Figure 4).¹⁸ Both models applied an average scapulohumeral rhythm of 1:1.7 derived from RTSA patients specifically for scapular upward rotation to reflect postoperative scapular behaviour.²⁸ For the other two rotational components (internal rotation and posterior tilt), regression equations from a 3D scapulohumeral rhythm model based on healthy participants’ experimental datasets were adopted.²⁹ This modelling approach was based on the finding that the scapulohumeral rhythm differed from the asymptomatic contralateral shoulder only in upward rotation, with no significant differences observed in the other two rotational axes.³⁰ Muscle actuation was achieved using 97 Hill-type muscle-tendon units representing major scapulohumeral, axiohumeral, and axioscapular groups. Muscle-tendon forces were estimated using a quadratic polynomial muscle recruitment criterion.¹⁸ The resulting force estimations were representative of older adults, as muscle parameters were derived from elderly cadaver data.^31-33

Fig. 4.

Illustrations of reverse total shoulder arthroplasty (RTSA) models. a) Medialized RTSA model. b) Lateralized RTSA model. Scapular reference frames in both models were positioned at the rotational centre of the glenosphere. X-axis, Y-axis, and Z-axis indicate the lateral-medial, superior-inferior, and anterior-posterior directions, respectively.

The scapular notching-related impingement stress, calculated using the force-dependent kinematics (FDK) method,¹⁸ was obtained by dividing the joint contact force by its corresponding contact area (Supplementary Material). This approach assumes quasi-static equilibrium between the FDK residual force and joint translation along the FDK axis, such that translation is resolved when the residual force reaches zero at each time step. Joint subluxation was subsequently quantified as the sum of the square roots of the 3D joint translations between the CORs of the glenosphere and humeral liner (i.e., the gap distance between these two CORs; Figure 5).¹⁹ To validate this musculoskeletal modelling approach, the impingement stress and joint subluxation in the current lateralized RTSA model were compared with those from a previously validated finite element model of lateralized RTSA (36 mm Aequalis Ascend Reverse Flex Implant System; Wright Tornier, USA) during 45° ER.¹⁹

Fig. 5.

Illustration of joint subluxation during external rotation in the musculoskeletal shoulder model of reverse total shoulder arthroplasty. COR, centre of rotation.

Simulation

Simulations were conducted for both medialized and lateralized RTSA models under subscapularis-repaired and subscapularis-torn conditions (Figure 6), across three stages of posterosuperior cuff tear severity defined by the senior surgeon (JHO) (Figure 7): Stage I, isolated supraspinatus bundle tears; Stage II, complete supraspinatus tears and superior bundle tears of the infraspinatus (half of them torn); and Stage III, complete tears of both the supraspinatus and infraspinatus.

Fig. 6.

Illustrations of subscapularis conditions in the musculoskeletal shoulder model of reverse total shoulder arthroplasty. a) Subscapularis-repaired condition (all bundles intact). b) Subscapularis torn condition (all bundles torn).

Fig. 7.

Illustrations of posterosuperior cuff tear severity in the musculoskeletal shoulder model of reverse total shoulder arthroplasty. a) Stage I: isolated supraspinatus bundle tears; b) Stage II: complete supraspinatus tears and superior bundle tears of the infraspinatus (half of them torn); c) Stage III: complete tears of both the supraspinatus and infraspinatus.

Statistical analysis

A three-way analysis of variance (ANOVA) was performed to assess the effects of RTSA design (medialized vs lateralized), subscapularis condition (repaired vs torn), and cuff tear severity (Stages I to III) on the impingement stress, joint subluxation, and muscle-tendon forces in the anterior, middle, and posterior deltoids, subscapularis, infraspinatus, and teres minor. Post hoc paired t-tests with false discovery rate (FDR) correction were used to compare medialized and lateralized designs within each subscapularis condition and cuff tear severity stage.³⁴ The FDR was controlled at 5% using the Benjamini-Hochberg procedure (adjusted p < 0.05). All statistical analyses were conducted in MATLAB R2020b (The MathWorks, USA).

Results

Model validation

The lateralized RTSA model showed good agreement with previously reported data during 45° ER.¹⁹ Peak impingement stress ranged from 21.4 MPa (SD 6.1) to 45.7 MPa (SD 9.1) (Supplementary Figure a), and peak joint subluxation ranged from 1.1 mm (SD 0.1) to 5.0 mm (SD 1.7) (Supplementary Figure b).

RTSA design (medialized vs lateralized)

The peak impingement stress and joint subluxation were consistently higher in medialized RTSA than in lateralized RTSA across all conditions (Table I).

Table I.

Comparison of peak impingement stress and joint subluxation during external rotation between the medialized and lateralized reverse total shoulder arthroplasty models under subscapularis-repaired and -torn conditions, across three stages of posterosuperior cuff tear severity. Data are presented as mean (SD).

Variable	Subscapularis condition	Stage	Medialized RTSA	Lateralized RTSA	p-value*
Impingement stress (MPa)	Repaired	I	42.0 (7.9)	21.4 (6.1)	< 0.001
		II	50.3 (9.8)†	24.1 (6.9)†	< 0.001
		III	65.3 (12.0)‡§	33.7 (6.8)‡§	< 0.001
	Torn	I	50.2 (9.7)	25.7 (7.3)¶	< 0.001
		II	58.6 (11.3)†¶	30.9 (8.8)†¶	< 0.001
		III	80.2 (14.6)द	45.7 (9.3)द	< 0.001
Subluxation (mm)	Repaired	I	6.0 (0.4)	1.1 (0.1)	< 0.001
		II	6.7 (0.5)†	2.1 (0.1)†	< 0.001
		III	9.7 (1.2)‡§	3.6 (1.1)‡§	< 0.001
	Torn	I	7.2 (0.5)¶	2.6 (0.2)¶	< 0.001
		II	8.6 (0.6)†¶	3.3 (0.2)†¶	< 0.001
		III	11.3 (0.8)द	5.0 (1.7)द	< 0.001

Stage I: isolated supraspinatus bundle tears; Stage II: complete supraspinatus tears and superior bundle tears of the infraspinatus (half of them torn); Stage III: complete tears of both the supraspinatus and infraspinatus.

^*The p-values were determined by post hoc paired t-tests with Benjamini–Hochberg correction for multiple comparisons, with a significance threshold set at an adjusted p < 0.05.

^†Significant differences between Stages I and II (p < 0.05).

^‡Significant differences between Stages II and III (p < 0.05).

^§Significant differences between Stages I and III (p < 0.05).

^¶Significant differences between subscapularis-repaired and subscapularis-torn conditions (p < 0.05).

RTSA, reverse total shoulder arthroplasty.

The peak anterior deltoid force was consistently lower in medialized RTSA than in lateralized RTSA across all conditions (Table II). In Stages I and II, peak middle deltoid force was significantly reduced in medialized RTSA compared with lateralized RTSA across subscapularis conditions (Table II). In Stage III, peak posterior deltoid force was consistently lower in medialized RTSA than in lateralized RTSA across subscapularis conditions (Table II).

Table II.

Comparison of peak muscle-tendon forces during external rotation between the medialized and lateralized reverse total shoulder arthroplasty models under subscapularis-repaired and -torn conditions, across three stages of posterosuperior cuff tear severity. Data are presented as mean (SD).

Variable	Subscapularis condition	Stage	Medialized RTSA	Lateralized RTSA	p-value*
Anterior deltoid (N)	Repaired	I	34.6 (4.3)	40.0 (4.5)	< 0.001
		II	31.4 (3.9)†	39.5 (4.5)†	< 0.001
		III	30.1 (4.8)‡‡	39.1 (9.6)	0.038
	Torn	I	34.5 (4.2)	39.6 (4.4)	< 0.001
		II	31.4 (3.9)†	39.1 (4.4)†	< 0.001
		III	29.3 (6.6)‡‡	38.3 (9.7)§	< 0.001
Middle deltoid (N)	Repaired	I	119.5 (16.1)	132.1 (17.8)	< 0.001
		II	118.1 (15.4)	133.2 (16.1)	< 0.001
		III	129.5 (18.4)‡¶	131.1 (13.1)	0.831
	Torn	I	120.9 (18.5)	134.4 (15.5)§	< 0.001
		II	119.4 (17.4)	134.5 (15.5)	< 0.001
		III	124.3 (17.6)¶	133.0 (13.0)§	0.259
Posterior deltoid (N)	Repaired	I	7.6 (1.1)	7.8 (1.5)	0.298
		II	8.0 (1.2)†	8.0 (1.4)†	0.936
		III	67.4 (7.1)‡¶‡	79.1 (11.0)‡¶‡	< 0.001
	Torn	I	7.3 (1.1)	7.6 (1.4)	0.033
		II	7.6 (1.2)†	7.9 (1.4)†	0.001
		III	54.1 (5.5)‡§¶	66.1 (9.6)§¶	< 0.001
Subscapularis (N)	Repaired	I	11.0 (1.9)	29.1 (8.6)	< 0.001
		II	20.0 (3.4)†	34.6 (9.5)†	< 0.001
		III	26.9 (3.2)‡¶	40.7 (10.5)‡¶	< 0.001
	Torn	I	0	0	N/A
		II	0	0	N/A
		III	0	0	N/A
Infraspinatus (N)	Repaired	I	58.2 (6.3)	67.6 (6.2)	0.005
		II	52.2 (6.6)†	57.9 (5.0)†	0.036
		III	0	0	-
	Torn	I	54.5 (5.6)§	63.9 (4.5)§	< 0.001
		II	48.8 (5.8)†§§	54.6 (3.9)†§§	0.009
		III	0	0	-
Teres minor (N)	Repaired	I	0.2 (0.2)	0.7 (0.3)	< 0.001
		II	0.4 (0.3)†	1.7 (0.5)†	0.001
		III	36.5 (4.2)‡¶	44.6 (6.6)‡¶	< 0.001
	Torn	I	0.2 (0.1)	0.7 (0.3)	< 0.001
		II	0.2 (0.2)†§§	1.0 (0.3)†§§	0.001
		III	32.3 (3.5)द	40.1 (5.8)द	< 0.001

Stage I: isolated supraspinatus bundle tears; Stage II: complete supraspinatus tears and superior bundle tears of the infraspinatus (half of them torn); Stage III: complete tears of both the supraspinatus and infraspinatus.

Significant differences between medialized and lateralized RTSA (adjusted p < 0.05).

^*The p-values were determined by post hoc paired t-tests with Benjamini–Hochberg correction for multiple comparisons, with a significance threshold set at an adjusted p < 0.05.

^†Significant differences between Stages I and II (p < 0.05).

^‡Significant differences between Stages I and III (p < 0.05).

^§Significant differences between supscapularis-repaired and subscapularis-torn conditions (p < 0.05).

^¶Significant differences between Stages II and III (p < 0.05).

N/A, not applicable; RTSA, reverse total shoulder arthroplasty.

The peak subscapularis and teres minor forces were consistently lower in medialized RTSA than in lateralized RTSA across all conditions (Table II). In Stages I and II, peak infraspinatus force was significantly reduced in medialized RTSA compared with lateralized RTSA across subscapularis conditions (Table II).

Subscapularis condition (repaired vs torn)

Subscapularis repair in Stages I to III significantly reduced peak impingement stress by 14.1% to 18.6% (p < 0.001) in medialized RTSA and by 16.7% to 26.2% (p < 0.001) in lateralized RTSA (Table I). The same repair stages also significantly decreased peak joint subluxation by 14.7 to 22.4% (p < 0.001) in medialized RTSA and by 27.5 to 58.9% (p < 0.001) in lateralized RTSA (Table I).

In Stage III, subscapularis repair significantly increased peak middle and posterior deltoid forces by 4.1% and 24.6% (p < 0.001 for both), respectively, in medialized RTSA (Table II). In lateralized RTSA, it significantly increased peak anterior and posterior deltoid forces by 2.2% and 19.7% (p < 0.001 for both), respectively, while peak middle deltoid force significantly decreased by 1.4% (p < 0.001) (Table II).

Subscapularis repair in Stages I and II significantly increased peak infraspinatus force by 6.8% (p < 0.001) in medialized RTSA, and by 5.8% (p = 0.002) and 6.2% (p < 0.001), respectively, in lateralized RTSA (Table II).

In Stages II and III, peak teres minor force increased significantly by 71.7% and 13.1% (p < 0.001) in medialized RTSA, and by 68.4% (p = 0.005) and 11.2% (p < 0.001), respectively, in lateralized RTSA (Table II).

Posterosuperior cuff tear severity

Relative to Stage I, peak impingement stress in medialized RTSA significantly increased by 19.7% and 16.7% in Stage II, and by 55.6% and 59.9% in Stage III, in the subscapularis-repaired and -torn conditions, respectively (Table I). In lateralized RTSA, corresponding increases were 12.5% and 19.9% in Stage II, and 57.4% and 77.8% in Stage III (Table I).

Compared to Stage I, peak joint subluxation in medialized RTSA significantly increased by 11.7% and 18.4% in Stage II, and by 62.0% and 56.0% in Stage III, under the repaired and torn conditions, respectively (Table I). In lateralized RTSA, increases were 101.0% and 25.5% in Stage II, and 239.5% and 92.6% in Stage III (Table I).

In medialized RTSA, peak anterior deltoid force significantly decreased by 9.0% (p < 0.001) in Stage II, and by 13.0% (p = 0.026) and 15.0% (p = 0.006) in Stage III, under the repaired and torn conditions, respectively (Table II). In lateralized RTSA, anterior deltoid force decreased by 1.3% and 1.2% (p < 0.001 for both) in Stage II (Table II). In medialized RTSA, peak middle deltoid force significantly increased by 8.4% (p = 0.006) and 2.8% (p = 0.011) in Stage III under the repaired and torn conditions, respectively (Table II). The peak posterior deltoid force significantly increased by 4.7% and 3.5% (p < 0.001 for both) in Stage II, and by 784.5% and 637.0% (p < 0.001 for both) in Stage III, respectively (Table II). In lateralized RTSA, posterior deltoid force increased by 2.6% (p = 0.007) and 3.6% (p < 0.001) in Stage II, and by 918.5% and 768.0% (p < 0.001 for both) in Stage III, in the repaired and torn conditions, respectively (Table II).

Compared to Stage I, peak subscapularis force in medialized RTSA increased significantly by 81.8% (p < 0.001) in Stage II and by 144.5% (p < 0.001) in Stage III (Table II). In lateralized RTSA, subscapularis force also increased significantly, by 18.9% (p = 0.003) in Stage II and 40.0% (p < 0.001) in Stage III (Table II).

For the infraspinatus, peak force in medialized RTSA decreased significantly from Stage I by 10.3% and 10.4% (p < 0.001 for both) in Stage II under subscapularis-repaired and -torn conditions, respectively (Table II). In lateralized RTSA, peak infraspinatus force significantly decreased by 14.3% and 14.6% (p < 0.001 for both) in Stage II, in the subscapularis-repaired and -torn conditions, respectively (Table II).

In medialized RTSA, peak teres minor force significantly increased from Stage I to Stage II by 78.1% and 6.8% (p < 0.001 for both), and from Stage I to Stage III by 16,811.6% and 14,320.3% (p < 0.001 for both), in the subscapularis-repaired and -torn conditions, respectively (Table II). Similarly, in lateralized RTSA, teres minor force increased by 124.5% and 34.7% (p < 0.001 for both) in Stage II, and by 5,959.0% and 5,406.9% (p < 0.001 for both) in Stage III, for the subscapularis-repaired and -torn conditions, respectively (Table II).

Interaction effects

Significant interactions were observed between RTSA design and posterosuperior cuff tear severity for peak impingement stress (p = 0.027), joint subluxation (p < 0.001), and posterior deltoid force (p < 0.001), indicating a greater effect of cuff tear severity on these outcomes in lateralized RTSA compared to medialized RTSA (Table I and Table II).

A significant interaction between RTSA design and posterosuperior cuff tear severity was also found for teres minor force (p < 0.001), with a greater effect observed in medialized RTSA than in lateralized RTSA (Table II).

Additionally, RTSA design significantly interacted with subscapularis condition for posterior deltoid force (p < 0.001), indicating a stronger influence of subscapularis integrity in medialized RTSA than in lateralized RTSA (Table II).

Lastly, a significant interaction between subscapularis condition and posterosuperior cuff tear severity was found for teres minor force (p < 0.001), suggesting that both factors significantly influenced teres minor loading in both RTSA designs (Table II).

Discussion

This study developed 6-DOF musculoskeletal shoulder models of medialized and lateralized RTSA to predict impingement stress and joint subluxation, based on a previously validated anatomical shoulder model.¹⁸ The predicted range of peak impingement stress (21.4 to 45.7 MPa) encompassed a previously reported value of 25 MPa.¹⁹ When considering the mean and SD, the predicted peak joint subluxation (5.0 mm (SD 1.7)) was within 1 mm of the reported value of 7.6 mm,¹⁹ supporting the model’s validity. Overall, the proposed model reliably predicts scapular notching-related impingement stress and joint subluxation during ER.

The primary finding of this study is that lateralized RTSA more effectively reduces the impingement stress and joint subluxation during ER compared to medialized RTSA. Subscapularis repair further enhances these benefits in lateralized RTSA, yielding greater reductions in impingement stress (16.7% to 26.2%; p < 0.001) and subluxation (27.5% to 58.9%; p < 0.001) than in medialized RTSA (14.1% to 18.6%; p < 0.001 and 14.7% to 22.4%; p < 0.001), across all posterosuperior cuff tear conditions. These results suggest that subscapularis repair in lateralized RTSA significantly mitigates both impingement stress and subluxation, even in the presence of posterosuperior cuff deficiencies. The observed improvements are likely due to increased joint compression from subscapularis tension, which enhances joint stability.^14,15,35 Joint instability is commonly associated with impingement and arises from unbalanced force coupling between the anterior (subscapularis) and posterior (infraspinatus and teres minor) rotator cuffs due to anterior or posterior cuff deficiency.^36-38 Therefore, these findings support lateralized RTSA with subscapularis repair as an effective strategy for mitigating scapular notching-related impingement and improving joint stability in patients with posterosuperior cuff tears.

The current results demonstrate that peak muscle-tendon forces in the anterior deltoid, subscapularis, infraspinatus, and teres minor were consistently higher in lateralized RTSA than in medialized RTSA under all conditions (Table II). Notably, when the subscapularis was repaired, posterior deltoid force exhibited a substantially greater increase with posterosuperior cuff tear severity in lateralized RTSA (from 2.6% to 918.5%; p < 0.001) than in medialized RTSA (from 4.7% to 784.5%; p < 0.001). These increases are attributed to the counterbalancing role of the repaired subscapularis, which elevates posterior deltoid loading in both configurations. This is consistent with prior work indicating that subscapularis repair in RTSA increases posterior deltoid force to counteract the added subscapularis tension.³⁹ While the value of subscapularis repair in lateralized RTSA remains debated—some studies report no significant impact on ER function—our findings suggest that it can improve ER torque generation by increasing posterior deltoid force more effectively than in medialized RTSA.^16,17 Since the efficiency of the posterior deltoid and teres minor is critical for restoring ER function in RTSA patients with cuff deficiency,^10,40-42 these results highlight the potential of subscapularis repair in lateralized RTSA to enhance ER function by increasing posterior deltoid force.

This study has several limitations. First, marker trajectory data were collected from young individuals, which may not accurately represent the movement patterns of older adults. However, the scapulohumeral rhythm was modelled at 1:1.7 based on experimental datasets from RTSA patients,²⁸ and muscle parameters were derived from elderly cadaver studies.^31-33 Therefore, despite using young participant motion data, the scapulohumeral motion and estimated muscle-tendon forces in the RTSA model may still approximate those of postoperative RTSA patients. Future studies using subject-specific kinematics from RTSA patients are required to improve the validity and clinical relevance of simulation results. Second, lateralization was analyzed using a standard configuration, which did not encompass other possible glenoid or humeral component configurations.⁵ While the findings are adaptable to varying degrees of lateralization, further research is needed to assess how different lateralization levels affect joint contact mechanics, particularly in the context of subscapularis integrity and posterosuperior cuff tear severity. Third, it could have been beneficial to use a single manufacturer’s system to vary the NSAs, inlay, or onlay stem designs. However, this approach was not feasible for the selected lateralized RTSA (Coralis Reverse Shoulder System), because this system is inherently designed for lateralization. Therefore, to compare the biomechanical differences between medialized and lateralized RTSA configurations, this study used a well-established implant that represents the conventional medialized RTSA (Lima SMR System) and the selected lateralized RTSA designs. Fourth, this study quantified the overall magnitude of joint subluxation using a scalar parameter; however, it did not evaluate the directional components of subluxation. Since dislocation mechanisms may vary between medialized and lateralized RTSA configurations depending on the direction of instability, analyzing directional subluxation vectors could provide additional insights into implant-specific biomechanical behaviors. Future studies incorporating such directional analyses are needed to further elucidate the mechanisms underlying joint instability in RTSA. Fifth, although the selected motion does not fully replicate the kinematics of dislocation-prone activities, it serves as a clinically relevant and standardized task for assessing joint stability, particularly when comparing different implant configurations. Nevertheless, evaluating joint behaviour under a single motion task may not fully capture the complex subluxation mechanisms encountered in daily activities. Further studies incorporating a broader range of dislocation-prone movements—such as extension, adduction, and internal rotation—are warranted to provide a more comprehensive understanding and evaluation of joint stability in RTSA.

This study demonstrated that lateralized RTSA combined with subscapularis repair reduces scapular notching-related impingement stress and joint subluxation, even in the presence of posterosuperior cuff tears. Additionally, subscapularis repair improves ER torque generation by increasing posterior deltoid force. In conclusion, our results suggest that lateralization and subscapularis repair in RTSA can mitigate the impingement, improve joint stability, and support ER torque generation in patients with posterosuperior cuff deficiencies. These findings can inform surgical decisions regarding implant design and subscapularis repair to reduce the risk of scapular notching and joint subluxation while enhancing ER torque generation.

Data Availability

The datasets generated and analyzed in the current study are not publicly available due to data protection regulations. Access to data is limited to the researchers who have obtained permission for data processing. Further inquiries can be made to the corresponding author.