Individualized coracoid osteotomy and 3D congruent arc reconstruction of glenoid for the treatment of recurrent anterior shoulder dislocation

Abstract

Background: The present study investigated individualized coracoid osteotomy for 3D congruent arc glenoid reconstruction and evaluated the clinical outcomes in recurrent anterior shoulder dislocation.

Methods: From January 2005 to July 2015, 78 patients with glenoid defect underwent coracoid and conjoint tendon transposition. The patients were divided into the individualized group (n = 34) and the non-individualized group (n = 44). All patients had CT data to reconstruct the shoulder model using Mimics software. In the individualized group, the individual coracoid osteotomy and bone fixation position parameters were measured from preoperative planification through simulating a 3D congruent arc glenoid reconstruction model. The non-individualized group underwent classic Bristow-Latarjet (B-L) procedure. The postoperative evaluation parameters included 3D congruent arc index, coracoid bone position, shoulder osteoarthritis index (Samilson-Prieto) and shoulder function score (Rowe, Constant-Murley score).

Results: The mean follow-up time was 51.0 months (ranging from 24 to 146). The individualized group got 3D congruent arc reconstruction of the glenoid by postoperative CT scanning. Bone position was more precise in the individual group than that in the B-L group. There was a lower incidence of shoulder osteoarthritis (Samilson-Prieto) in the individual group compared with that in the B-L group: 0 vs 13.6% (mild 6/44, P = 0.027), respectively. No significant difference was observed between the individual and B-L groups in rate of re-dislocation: 0 vs 4.5% (2/44, P = 0.315), respectively. The postoperative Rowe and Constant score was significantly improved but was not significantly different between the two groups.

Conclusion: The individual procedure achieved 3D congruent arc glenoid reconstruction. The clinical effects in patients with medium glenoid defect were good, especially the low incidence of shoulder osteoarthritis in middle-term follow-up.

Keywords: Bone defect; Congruent arc; Coracoid transposition; Preoperative planning; Shoulder dislocation; Three-dimensional reconstruction.

Fig. 1

Simulating coracoid osteotomy to repair the bone defect. A1–A2 The red circle was the best-fitting ball from the healthy side glenoid mirrored to the affected side glenoid as the template. The grey plane was the coracoid cutting plane and glenoid bone fixation bed. B1–B2 Simulation drilling of Φ4.3 mm “screws” to determine its length and direction

Fig. 2

Simulation of 3D congruent arc reconstruction of the glenoid by coracoid transposition. a The 3D glenoid reconstruction model. b The axial plane. c The coronal plane. d The sagittal plane. The white sphere was mirrored from the healthy side as the co-radian template. The red line was a virtual coracoid graft contour. The subcoracoid surface in the 3D plane was parallel to the best-fitting sphere

Fig. 3

Preoperative design coracoid osteotomy parameters (A1–A2 one length and one angle) and the glenoid bed fixation data (B1–B2, four lengths)

Fig. 4

The intraoperative sketch. The calliper was for measuring length, and the X-ray film was for measuring angle. A1–A2 The coracoid osteotomy parameters. B1–B2 The glenoid bed fixation data

Fig. 5

a–d Postoperative 3D congruent arc reconstruction of the glenoid. The white ball was the glenoid radian best-fitting ball. The coracoid graft was parallel to the postoperative glenoid radian ball

Fig. 6

The coracoid graft clock position. a The individualized group: all cases were under the equator level. b The B-L group: five cases were above the equator level. The middle point of the graft was defined by the bone clock position

Fig. 7

Postoperative reconstruction of the glenoid. A1–A4 The individualized group and the bone 3D congruent arc with the glenoid. B1–B4 The B-L group and the bone flush with the glenoid surface

Fig. 8

The bone contact surface of the B-L and individual groups. The contact surface of the individual inclined osteotomy was higher than that of vertical osteotomy in the B-L group by 22% (1/cosα-1)

Fig. 9

Postoperative osteoarthritis, avulsion of bone block and re-dislocation of glenohumeral joint. A Individual group, OA degree, normal. Classic Bristow-Latarjet group. B1 Mild OA. B2 Bone avulsion and mild OA. B3 Shoulder re-dislocation after basketball collision. Red arrow: the length of the osteophyte, the black star: the avulsion bone

Fig. 10

Different contact patterns between the humeral head and coracoid grafts. The single plane congruent arc Latarjet: line, the classic Bristow-Latarjet: point, the 3D congruent arc surface. The red broken line represented the contact patterns, and the white ball simulated the humeral head