Evaluating coracoid graft geometry for glenoid bone loss: a comparative study of stress distribution and contact mechanics

Abstract

Background

This study aimed to investigate the biomechanical differences between natural coracoid (NC) and modified coracoid (MC) grafts (with a flattened surface to optimize contact with the glenoid cavity), focusing on their impact on contact area, contact volume, and stress distribution in scenarios of glenoid bone loss. The objective was to determine how graft configuration influences biomechanical stability and force distribution under varying conditions of bone loss.

Methods

Three-dimensional models of the glenoid cavity and coracoid process were developed using population-based anatomical averages. Two graft configurations were analyzed: NC and MC. Simulations incorporated progressive glenoid bone loss (0%-20%). Biomechanical analyses evaluated contact area, contact volume, stress distribution, and deformation under a compressive force of 700 N using finite element analysis.

Results

NC grafts demonstrated superior adaptation and a larger contact area in anatomically intact glenoid (0% bone loss). However, as bone loss increased beyond 2%, MC grafts provided more consistent contact areas, better stress distribution, and reduced stress concentration. The flattened surface of MC grafts optimized the biomechanical interaction, particularly under conditions of advanced bone loss, ensuring enhanced stability and reduced risk of localized deformation.

Conclusion

The findings highlight the importance of graft configuration in addressing glenoid bone loss. While NC grafts are preferable for intact or minimally compromised glenoids, MC grafts are more effective in scenarios with bone loss, providing improved biomechanical stability and optimized force distribution. This study underscores the need for tailored surgical strategies to achieve optimal outcomes based on individual anatomical and clinical conditions.

Understanding and analyzing the irregularities associated with glenohumeral bone loss and the morphology of the coracoid process are crucial for the effective management of glenohumeral instability, particularly when using the Latarjet technique. Traumatic anterior shoulder dislocations frequently result in fractures of the anterior glenoid rim, compromising the static stabilizing mechanisms of the glenohumeral joint. Such disruptions significantly impair the glenoid's osseous articular conformity, reducing its capacity to resist shear stress.¹

Improper positioning of the coracoid in the Latarjet procedure has been identified as a potential factor contributing to suboptimal clinical outcomes. This includes misalignment of the coracoid graft, inadequate screw placement, and improper angulation, which may lead to reduced bone contact, graft instability, and compromised biomechanical performance.^2,8,12 Although various autologous bone-grafting techniques have been proposed, no single method has been definitively established as superior for addressing recurrent instability and glenoid bone loss. Unlike other bone block techniques, the coracoid transfer approach offers additional stabilizing effects through capsular and muscular reinforcement, alongside its bony effect. This multifaceted stabilization, commonly referred to as the “triple effect,” was first described by Patte.⁶

The original Latarjet technique involves fixing the coracoid process to the glenoid with its lateral edge flush against the anterior glenoid surface, aiming to restore articular congruence and provide additional stability. However, this conventional approach was later redefined by De Beer and Burkhart, who introduced the congruent-arc technique. In this adaptation, the coracoid is rotated during fixation so that its inferior surface lies contiguous with the glenoid. Proponents of this modification argue that it improves articular surface congruence by better conforming to the glenoid's native concavity, similar to the congruence achieved with the inner table of the iliac crest bone graft or the articular surface of the distal tibia.⁵

The congruent-arc technique, while promising, entails technical challenges, and its advantages may come at the cost of reduced biomechanical strength of graft fixation.⁷ Furthermore, its application can be limited in cases where the coracoid size is insufficient, compromising the technical feasibility of the approach. Conversely, the traditional coracoid positioning, which fixes the graft with its lateral edge flush against the glenoid, provides a greater bone-to-bone contact area. This configuration has the potential to enhance both consolidation rates and biomechanical stability. However, there is a lack of studies investigating how different anatomical configurations influence graft contact with the glenoid and the distribution of pressure across the interface between these structures during fixation. Such analyses are particularly relevant in understanding the effect of angulation and how it impacts surgical outcomes.³

This study aims to evaluate the anatomical compatibility between the glenoid and the coracoid graft under varying conditions, focusing on the extent of contact between these 2 structures. Understanding these interactions is critical for optimizing surgical outcomes, as effective contact improves graft stability, enhances bone consolidation, and reduces the risk of postoperative complications. Specifically, the analysis will assess contact surfaces in both intact glenoids and those with bone loss, comparing outcomes when using a natural coracoid (NC) graft versus a modified coracoid (MC) graft. By investigating the impact of angulation and graft positioning, this research seeks to identify the optimal anatomical configuration that ensures enhanced bone-to-bone contact and improved biomechanical performance.

Materials and methods

Subject-specific geometries

The morphological and morphometric data used in this study were extracted from previously published anatomical studies and validated using three-dimensional modeling techniques. Specifically, dimensions of the glenoid cavity were derived from the work of Nobeschi et al, which details the diameters and flat surface area of the glenoid cavity. Measurements of the coracoid process were based on the study by Dolan et al,^4,9 which provides critical parameters related to bone transfer procedures.

Glenoid cavity dimensions

The morphometric dimensions of the glenoid cavity were obtained from the literature and confirmed using three-dimensional models. The following measurements were considered:

• •The maximum transverse diameter averaged 25.47 mm.
• •The minimum transverse diameter averaged 16.94 mm.
• •The longitudinal diameter was 37.25 mm on average.
• •The flat-interface area (antero-inferior bony footprint used for coracoid fixation after decortication) of the glenoid cavity was 256.68 mm² on average.

Coracoid process dimensions

The coracoid process was also assessed using previously published data and three-dimensional validation. The principal measurements included.

• •Tip width: 18.3 mm (mean).
• •Tip height: 11.5 mm (mean).
• •Distance from the tip to the midpoint: 22.8 mm (mean).

Cortical thickness

The cortical thickness was determined to be an average of 1.875 cm, calculated from measurements in axial, sagittal, and coronal planes. This parameter is critical for biomechanical simulations as it directly influences the resistance to applied forces and the potential for deformation, thereby affecting the stability and integration of the graft during surgical fixation. This parameter was critical for assessing the mechanical resistance of the graft during the biomechanical simulations.¹⁴

Three-dimensional model development

All remaining dimensions (depth, curvature radii, coracoid length, etc.) were taken from peer reviewed computed tomography based anatomic studies.^4,9 Three-dimensional models of the glenoid cavity and coracoid process were developed using Fusion 360. Fusion 360 was selected due to its robust capabilities in parametric design and its ability to accurately simulate complex anatomical structures, making it particularly suitable for biomechanical analyses. The models were designed to reflect anatomical variations observed in the literature and accurately represent the bone interfaces for biomechanical analysis.

• •Glenoid cavity: Modeled as a concave surface to represent its flat area and transverse and longitudinal diameters. Progressive bone loss was simulated by removing the anteroinferior portion of the glenoid cavity in increments of 2%, up to a maximum of 20%.
• •Coracoid process: Modeled to include the length, width, and height of the tip and midpoint. The cortical thickness was uniformly applied to represent the structural resistance. The graft was modeled in 2 configurations (Figs. 1 and 2).

Figure 1.

Natural coracoid (NC): Maintaining the natural curvature of the tip.

Figure 2.

Modified coracoid (MC): Flattened to optimize contact with the glenoid cavity.

While this study focused on symmetric defects for controlled analysis, we recognize that clinical practice often involves oblique defects. These primarily affect graft positioning rather than the fundamental contact mechanism, as the MC graft's flat surface maintains optimal contact with the prepared glenoid bed regardless of defect orientation. For defects exceeding 20%, the contact area plateaus as the defect size surpasses the graft's coverage capacity, though the MC graft retains its biomechanical advantage.

Fixation parameters

The simulations incorporated variations in osteotomy angles and distances to replicate real-life insertion conditions commonly encountered in the Latarjet technique. The osteotomy angles, ranging from 4.5° to 13°, were selected to reflect the anatomical variability observed in clinical practice and to evaluate their influence on optimizing graft-to-glenoid contact. These angles were measured relative to the horizontal plane, simulating different orientations of the coracoid graft after osteotomy preparation. The aim was to identify the configuration that maximized the contact area between the graft and the glenoid cavity during compressive force application. In this study, only the osteotomy/graft angle (4.5°–13°) was varied, while the screw axis was kept perpendicular (0°) to the surface to avoid fixation-angle bias (Figs. 3 and 4).

Figure 3.

The fixation angles ranged from 4.5° to 13°, with the goal of maximizing the contact area between the graft and the glenoid cavity during compressive force application.

Figure 4.

The distances from the Bankart lesion ranged from 0 mm (control) to 5 mm (representing maximum bone loss/20%).

Constitutive and material laws

The mechanical behavior of cortical bone was modeled using an orthotropic material law, accounting for the material's directional properties. This choice reflects the fact that cortical bone represents nearly 80% of the skeletal mass and has a preferred "grain" direction, resulting in a varying Young's modulus depending on orientation.

• •Material properties: All property values used in this study were obtained from the literature. The yield strength was chosen based on corticalized bones in young patients.¹³
• •Orthotropic modeling: This material law accurately describes the stress-strain relationship in the study, with Young's modulus and shear properties reflecting the anisotropic nature of cortical bone.

After applying maximum compressive force to the interface between the glenoid cavity and the coracoid graft, the biomechanical limits of cortical bone were evaluated. Plastic deformation was analyzed based on known biomechanical properties of cortical bone:

• •Elastic modulus (E): 17–20 GPa.
• •Yield strength: 100–130 MPa.
• •Plastic strain limit: 1–2% linear deformation before fracture.¹³

Mesh convergence was verified by iterative refinement, with element size reduced until changes in stress, contact area, and contact volume were below 5%, confirming mesh independence. Based on a cortical thickness of 1.875 cm, the maximum plastic deformation under uniform compression was calculated as approximately 0.185 mm. This threshold was critical for determining whether the bone interface could sustain the applied forces without structural compromise, thereby underscoring the biomechanical compatibility of the graft across different fixation techniques.¹⁴

Contact interface assessment

The reshaped and nonreshaped graft configurations were compared to evaluate contact distribution. Metrics such as the percentage of bone-to-bone contact area, the magnitude of pressure distribution across the interface, and the locations of peak stress were analyzed to determine the efficiency of each configuration in optimizing biomechanical stability.

• 1.Natural coracoid: Distributed compressive forces more uniformly across the bone interface, reducing stress peaks and minimizing the risk of plastic deformation.
• 2.Modified coracoid: Concentrated forces in localized regions due to its curvature, leading to higher stress peaks and increased risk of localized plastic deformation.

Graft placement was standardized to be flush with the glenoid rim in all configurations. Variations positions were not analyzed.

Force application and analysis conditions

A compressive force of 700 N, representative of the torque applied during screw fixation, was uniformly applied to the interface between the glenoid and the graft. This force was selected based on typical values reported in the literature for biomechanical shoulder testing. Throughout the analysis, the cortical integrity of the coracoid was maintained, ensuring that the structural properties of the graft remained intact.¹⁴

The primary parameters evaluated included the following:

• •Stress distribution: Stresses at the bone interface were calculated and visualized using heat maps, focusing on areas of pressure concentration.
• •Maximum stress (MPa): Assessed to identify load peaks in the different graft configurations.
• •Graft adaptation to the glenoid: Measured based on the contact area and contact volume between the bones.
• •Contact area was quantified as the total node-to-node surface area where the distance between the coracoid graft and the glenoid was ≤0.05 mm (the default contact-pair tolerance of the software). This metric reflects the effective surface available for biological healing.
• •Contact volume was defined as the enclosed three-dimensional solid volume when this tolerance was expanded to ≤0.1 mm. This broader threshold captures microinterdigitation and surface roughness, which are important contributors to initial mechanical interlock and stability.

The simulations were performed in a finite element analysis environment, enabling a detailed evaluation of the interaction between the glenoid and the graft under varying anatomical conditions. The objective was to determine how graft shape and cortical preservation influence stress distribution and biomechanical behavior at the bone interface.

Results

The interaction between the coracoid graft and the glenoid was analyzed under varying anatomical and biomechanical conditions, focusing on contact volume, contact area, and the configurations of MC and NC grafts. The results obtained highlight critical aspects related to stability and the biomechanical behavior of the bone interface (Table I and II).

Table I.

Biomechanical parameters of the natural coracoid (NC) graft across progressive glenoid bone loss (0% to 20%).

Glenoid bone loss	Graft rectification	Contact volume (mm³)	Contact area (mm²)	d (Bankart)	Angle relative to the ground
0%	No	9.46	81.03	0	13
2%	No	20.9	107.1	0.5	10
4%	No	9,867	74.13	1	8
6%	No	6,817	61.71	1.5	6
8%	No	5.53	43.35	2	5.5
10%	No	6,096	55.1	2.5	5.5
12%	No	7.04	52,127	3	5.5
14%	No	7.04	52,127	3.5	5.5
16%	No	7.04	52,127	4	5.5
18%	No	7.04	52,127	4.5	5.5
20%	No	7.04	52,127	5	5.5

Table II.

Biomechanical parameters of the modified coracoid (MC) graft with flattened surface, under identical bone loss conditions.

Glenoid bone loss	Graft rectification	Contact volume (mm³)	Contact area (mm²)	d (Bankart)	Angle relative to the ground
0%	Yes	7,559	59.54	0	10
2%	Yes	20,911	106.93	0.5	10
4%	Yes	27.95	157.59	1	10
6%	Yes	31.95	177.04	1.5	10
8%	Yes	34,127	196.60	2	10
10%	Yes	38,245	201.64	2.5	10
12%	Yes	36,326	206,735	3	10
14%	Yes	40,641	211.14	3.5	10
16%	Yes	39,724	214,841	4	10
18%	Yes	22,492	218,161	4.5	10
20%	Yes	37,057	228.00	5	10

Results analysis

The contact area demonstrated significant differences between MC and NC grafts, depending on the presence or absence of glenoid bone loss. When the glenoid showed no bone loss (0%), the NC graft provided a larger contact area of 81.03 mm² compared to the MC graft, which displayed 59.54 mm². This result underscores that, in the absence of bone loss, the natural shape of the NC graft may facilitate better initial adaptation to the glenoid. However, beyond 2% of bone loss, the MC graft exhibited superior outcomes in terms of contact area. The contact area of the MC graft progressively increased, reaching 228.00 mm² at 20% bone loss, while the NC graft stabilized around 52.127 mm² after 10% bone loss. This behavior suggests that modifying the graft is more effective for optimizing contact in scenarios of reduced bone support, enabling better distribution of compressive forces (Fig. 5).

Figure 5.

Graph illustrating contact volume versus glenoid bone loss for modified coracoid (MC) (solid yellow) and natural coracoid (NC) (dashed orange).

Contact volume followed trends similar to those observed for contact area. For NC grafts, the initial volume was 9.46 mm³ (0% bone loss), but it continuously decreased with increasing bone loss, reaching 7.04 mm³ at 20%. Conversely, the MC graft exhibited a different pattern, with a progressive increase in contact volume from 7.559 mm³ (0% bone loss) to a peak of 39.724 mm³ at 16%, before stabilizing. This increase reflects the MC graft's ability to enhance bone interlocking in more advanced bone loss conditions (Fig. 6).

Figure 6.

Graph illustrating contact area versus glenoid bone loss for modified coracoid (MC) (solid yellow) and natural coracoid (NC) (dashed orange).

Stress analysis

The fixation angle played a critical role in graft stability, particularly for NC grafts. In intact glenoids (0% bone loss), the best performance was observed with a fixation angle of 13° relative to the ground. However, as the angle was reduced to 4.5°, both contact volume and contact area decreased significantly. These findings indicate that NC grafts are more sensitive to variations in fixation angle. In contrast, MC grafts demonstrated greater consistency in contact volume and area values, regardless of the fixation angle. This biomechanical versatility makes MC grafts a more reliable option across various surgical scenarios (Fig. 7).

Figure 7.

Biomechanical simulation illustrating stress distribution for Modified (Left) and Natural (Right) coracoid grafts. Preserving the coracoid's cortical integrity reveals distinct force concentration patterns and underscores differences in graft adaptability.

Image analysis

The presented images illustrate the biomechanical analysis of cortical bone subjected to a compressive force of 700 N applied to the glenoid with bone loss, considering 2 configurations of the coracoid graft: modified (left) and natural (right). It is important to emphasize that in both the configurations the cortical integrity of the coracoid was preserved, ensuring the structural stability of the graft.

In the NC graft configuration, the analysis revealed stress distribution with localized areas of concentration, reaching a maximum stress of 29.754 MPa. The irregular surface of the graft resulted in pressure points, particularly at the bone interface contact zones, indicating lower efficiency in stress distribution. This pattern increases the risk of plastic deformation in specific regions. Furthermore, the irregular contact of the graft with the glenoid surface limited its adaptability, leading to reduced contact area and less uniform load distribution, despite the preservation of the coracoid's cortical structure.

Conversely, the MC graft configuration demonstrated a more uniform stress distribution, with a significantly lower maximum stress of 17.472 MPa. The flat and regular surface of the graft allowed better adaptation to the glenoid, optimizing the contact area and minimizing stress peaks. The preservation of the coracoid's cortical structure ensured the graft's resistance, enabling it to withstand compressive forces without plastic deformation, even in scenarios of significant bone loss (Fig. 8).

Figure 8.

Four 3D-printed scapula prototypes (a-d) showing distinct coracoid graft placements (in gray). Variation in orientation highlights how graft positioning influences bone coverage and potential contact area. 1a/b demonstrate good graft-glenoid contact (1a with bone loss, 1b without), whereas 1c/d show suboptimal contact (1c with bone loss, 1d without). 3-D, three-dimensional.

Discussion

The results of this study highlight the biomechanical differences observed between MC and NC grafts, particularly regarding contact area, contact volume, and stress distribution at the interface between the coracoid graft and the glenoid. While the primary objective was not to demonstrate complete restoration of glenohumeral loading mechanics, the data provide valuable insights into how graft configurations influence force distribution and bone contact under varying degrees of glenoid bone loss. When the glenoid exhibited no bone loss (0%), the NC graft demonstrated a larger initial contact area, suggesting better adaptation in anatomically preserved conditions. However, as bone loss exceeded 2%, MC grafts proved more effective in maintaining consistent contact area and ensuring a more uniform force distribution. These findings emphasize the importance of considering graft configuration in cases of significant bone loss, underscoring the need for tailored surgical strategies for optimal biomechanical outcomes.

In addition to the differences observed in the contact area, this study demonstrated that NC grafts, by maintaining the natural anatomy of the coracoid, exhibit more concentrated pressure points on flatter surfaces. This uneven force distribution limits the effective contact between the graft and the glenoid in scenarios requiring significant structural stability. Conversely, MC grafts, with their rectified and uniform surfaces, provided superior contact area as glenoid bone loss increased. This configuration not only enhanced the distribution of compressive forces but also reduced the likelihood of localized stress peaks, positioning MC grafts as the preferable option in reconstruction scenarios with advanced bone loss. These results are particularly relevant in guiding surgical decision-making, where graft modification can significantly impact biomechanical stability and clinical outcomes.

Another fundamental aspect is the capacity of the grafts to tolerate plastic deformation. Studies by Stirma and colleagues have demonstrated that cortical grafts, with an average thickness of 1.875 cm, can undergo deformation of up to 1% before reaching their plastic limit, equivalent to approximately 0.1875 mm.^10,11,14 This parameter was essential for calculating the contact area and volume between the 2 bones, allowing for precise analysis of the bone interface under varying conditions of bone loss. Deformation, by subtly altering the conformation of the graft, proved influential in optimizing the contact between surfaces, particularly in more complex biomechanical scenarios. These findings suggest that in cases of advanced bone loss, the use of MC grafts not only optimizes the contact area but also reduces the likelihood of stress concentration in specific regions, thereby enhancing biomechanical stability. However, in intact glenoids or those with minimal bone loss, the preserved anatomy of NC grafts may provide more appropriate initial adaptation, emphasizing the need for personalized strategies tailored to each patient.

The comparative analysis between the 2 configurations underscores the biomechanical advantages of MC grafts, particularly when glenoid bone loss exceeds 2%. In such cases, MC grafts allow for a more uniform distribution of compressive forces and improved adaptation between the graft and the glenoid, optimizing contact area and reducing stress concentrations in critical regions. This configuration not only reduces stress peaks but also increases the contact area, suggesting greater biomechanical stability and resistance. In contrast, NC grafts demonstrated higher force concentrations in specific areas, which may compromise their efficiency in clinical scenarios. The preservation of the cortical structure in both configurations was fundamental in ensuring the structural integrity of the grafts during biomechanical analyses, highlighting its essential role in maintaining stability and functional integrity. These findings suggest that in cases of advanced bone loss, the use of MC grafts not only optimizes the contact area but also reduces the likelihood of stress concentration in specific regions, thereby enhancing biomechanical stability. However, in intact glenoids or those with minimal bone loss, the preserved anatomy of NC grafts may provide more appropriate initial adaptation. This underscores the importance of considering individualized strategies tailored to each patient's anatomical and clinical conditions to achieve optimal outcomes (Table III).

Table III.

Comparative assessment of contact area, volume, and stress distribution under different glenoid conditions, graft types, and inclinations.

Glenoid condition	Graft type	Inclination (°)	Contact area (mm²)	Contact volume (mm³)	Stress	Visual outcome
Intact (0% loss)	NC	13°	81.03	9.46	High (localized)	✔
Intermediate (10%)	NC	10°	55.1	6.09	Moderate (localized)	✔
Intermediate (10%)	MC	10°	171.4	30.12	Low (uniform)	✔
Advanced (20%)	NC	4.5°	52.12	7.04	High (localized)	✔
Advanced (20%)	MC	4.5°	221.0	39.72	Very low (uniform)	✔

NC, natural coracoid; MC, modified coracoid.

Checkmarks indicate favorable outcomes, whereas crosses denote suboptimal performance.

This study has several limitations that should be acknowledged. Population-based anatomical averages for the glenoid and coracoid were employed, which do not capture the full spectrum of anatomical variability. Accordingly, the findings should be interpreted as representative of typical anatomy rather than as a population-wide statistical analysis. Although this simplification allowed for a focused assessment of compressive stresses at the graft–glenoid interface, it did not account for shear and tensile forces that play a critical role in real-world loading scenarios and may substantially influence stress distribution and failure risk. Future investigations should incorporate these additional forces to provide a more comprehensive understanding of biomechanical behavior under diverse clinical conditions. The load was applied directly to the graft-glenoid interface to isolate mechanical performance under compressive force. Although this approach does not replicate full joint articulation, including the humeral head and cartilage, it enables focused analysis of bone-to-bone contact.

Another limitation is the use of mathematical biomechanical models as an alternative to cadaveric specimens. While the modeling approach ensured uniformity and allowed precise control of variables, it excluded the effects of individual variations in bone quality, density, and anisotropy. In real-world scenarios,⁷ these factors can impact the mechanical properties and behavior of both the glenoid and the coracoid graft. The plastic deformation analysis relied on average biomechanical properties derived from the literature. However, these properties can vary between individuals, particularly in cases involving pathologies or age-related changes in bone composition. As such, the applicability of these results to a broad patient population may be limited.

Micromotion at the graft site was not simulated due to the assumption of rigid fixation. The primary aim of this study was to evaluate bone-to-bone contact at the graft–glenoid interface. In clinical scenarios, micromotion may influence load transfer and healing and should be addressed in future dynamic models.

Conclusion

This study highlights the significant biomechanical differences between coracoid grafts, emphasizing the importance of graft configuration in optimizing contact area and force distribution under conditions of glenoid bone loss. While NC demonstrated superior adaptation in anatomically intact glenoids, MC proved more effective in scenarios of advanced bone loss, providing improved biomechanical stability and reducing stress concentration.

Disclaimers:

Funding: No funding was disclosed by the authors.

Conflicts of interest: The 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.