Remplissage versus latarjet for engaging Hill-Sachs defects without substantial glenoid bone loss: a biomechanical comparison

Abstract

Background: Recurrent shoulder instability is commonly associated with Hill-Sachs defects. These defects may engage the glenoid rim, contributing to glenohumeral dislocation. Two treatment options to manage engaging Hill-Sachs defects are the remplissage procedure, which fills the defect with soft tissue, and the Latarjet procedure, which increases glenoid arc length. Little evidence exists to support one over the other.

Questions/purposes: We performed a biomechanical comparison of the remplissage procedure to the traditional Latarjet coracoid transfer for management of engaging Hill-Sachs defects in terms of joint stiffness (resistance to anterior translation), ROM, and frequency of dislocation.

Methods: Eight cadaveric specimens were tested on a shoulder instability simulator. Testing was performed with a 25% Hill-Sachs defect with an intact glenoid and after remplissage and Latarjet procedures. Joint stiffness, internal-external rotation ROM, and frequency of dislocation were assessed. Additionally, horizontal extension ROM was measured in composite glenohumeral abduction.

Results: After remplissage, stiffness increased in adduction with neutral rotation (12.7 ± 3.7 N/mm) relative to the Hill-Sachs defect state (8.7 ± 3.3 N/mm; p = 0.016). The Latarjet procedure did not affect joint stiffness (p = 1.0). Internal-external rotation ROM was reduced in abduction after the Latarjet procedure (49° ± 14°) compared with the Hill-Sachs defect state (69° ± 17°) (p = 0.009). Horizontal extension was reduced after remplissage (16° ± 12°) relative to the Hill-Sachs defect state (34° ± 8°) (p = 0.038). With the numbers available, there was no difference between the procedures in terms of the frequency of dislocation after reconstruction: 84% of specimens (27 of 32 testing scenarios) stabilized after remplissage, while 94% of specimens (30 of 32 testing scenarios) stabilized after the Latarjet procedure.

Conclusions: Both procedures proved effective in reducing the frequency of dislocation in a 25% Hill-Sachs defect model, while neither procedure consistently altered joint stiffness.

Clinical relevance: In the treatment of shoulder instability with a humeral head bone defect and an intact glenoid rim, this study supports the use of both the remplissage and Latarjet procedures. Clinical studies and larger cadaveric studies powered to detect differences in instability rates are needed to evaluate these procedures in terms of their comparative efficacy at preventing dislocation, as any differences between them seem likely to be small.

Fig. 1

A rendering of the in vitro shoulder simulator is shown, including a mounted specimen with soft tissues removed for clarity. The overlaid arrows indicate the loading vectors for each of the muscle groups: F_DELTS = three deltoid heads; F_SUP = supraspinatus; F_INF = infraspinatus and teres minor; F_SSC = subscapularis; F_LHB = long head of the biceps; F_SHB = conjoint tendon of the short head of the biceps. The simulator is capable of physiologically orienting the scapula and glenohumeral joint in four degrees of freedom and is shown with a potted scapula specimen (with soft tissues omitted for clarity) (A), a humerus (with soft tissues omitted for clarity) (B), a computer-controlled scapular elevation mechanism that achieves repeatable positioning (C), a glenohumeral abduction guide arc and slider (D), a glenohumeral plane of elevation adjustment plate (E), a low-friction deltoid and rotator cuff guide system that routes cables to pneumatic actuators (F), six-degrees-of-freedom tracking markers (G), a cemented humeral rod with an interposed six-degrees-of-freedom load cell (H), and pneumatic actuators used to separately load the rotator cuff, deltoid, and biceps tendons (I).

Fig. 2

A 25% Hill-Sachs defect is shown.

Fig. 3

A remplissage procedure with placement of suture anchors in the valley of the Hill-Sachs defect is shown.

Fig. 4

A Latarjet reconstruction with the coracoid fixed to the anteroinferior glenoid with two cortical screws is shown.

Fig. 5

A flow diagram demonstrates the testing protocol followed for each specimen in each condition (intact, Hill-Sachs defect, remplissage, and Latarjet coracoid transfer). ER = external rotation; IR = internal rotation; N = neutral.

Fig. 6A–D

Graphs show joint stiffness in (A) abduction and external rotation, (B) abduction and neutral rotation, (C) adduction and neutral rotation, and (D) adduction and external rotation among the four specimen states (intact, Hill-Sachs defect, remplissage, and Latarjet coracoid transfer). (A) Joint stiffness in abduction and external rotation is lower (more unstable) in the Hill-Sachs defect state than in the intact state (p = 0.029). Both remplissage and Latarjet restore stiffness to intact levels, with no differences between groups (p > 0.08). (B) No significant differences are noted between the remplissage and Latarjet groups in joint stiffness in abduction and neutral rotation (p = 0.907). (C) Remplissage has significantly greater joint stiffness in adduction and neutral rotation compared to the Latarjet procedure (p = 0.003). Neither procedure is significantly different from intact (p = 1.0). (D) No significant differences are noted between the remplissage and Latarjet groups in joint stiffness in adduction and external rotation (p = 0.137). Values are shown as means with SDs (error bars). HSD = Hill-Sachs defect.

Fig. 7

A graph shows horizontal extension ROM with the arm in abduction and external rotation (60°) among the four specimen states (intact, Hill-Sachs defect, remplissage, and Latarjet coracoid transfer). Remplissage significantly reduces this motion (16.1° ± 12.1°) relative to the Latarjet procedure (34.4° ± 7.8°) (p = 0.043). The Latarjet procedure does not affect this motion relative to the Hill-Sachs defect state (34.3° ± 7.6°) (p = 1.0). Neither procedure significantly affects this motion compared to the intact specimen (29.7° ± 10.5°) (p > 0.19). Values are shown as means with SDs (error bars). HSD = Hill-Sachs defect.

Fig. 8A–B

Graphs show internal-external rotation ROM in (A) adduction and (B) abduction among the four specimen states (intact, Hill-Sachs defect, remplissage, and Latarjet coracoid transfer). (A) No significant effect in internal-external rotation ROM in adduction is noted (p > 0.24). (B) The Latarjet procedure limits internal-external rotation ROM in abduction relative to the Hill-Sachs defect state (p = 0.009), while remplissage does not (p = 1.0). Neither procedure significantly alters this motion compared to the intact specimen (p > 0.13). Values are shown as means with SDs (error bars). IR/ER = internal-external rotation; HSD = Hill-Sachs defect.