Introduction: The goal of this study is to report the clinical and radiographic results of 2 types of implants used to treat 3 and 4 parts fractures of the proximal humerus.

Patients: Sixty-three patients (64 shoulders) were reviewed in this retrospective series. Forty women and 23 men were included, the mean age was 64 ± 12 (39–86). A group of 31 patients was managed with a ‘standard’ implant, a second group of 32 patients (33 shoulders) was managed with a ‘fracture’ implant. The delay between initial trauma and the surgical procedure was less than 4 weeks (1–30 days) for all patients.

Methods: All the procedures were carried out by a senior surgeon. The patients were reviewed by an independent observer with a mean follow-up of 59 ± 38 months (12–138) for a clinical and radiographic evaluation.

Results: In the ‘standard implant’ group; 84% of the patients were satisfied or very satisfied regarding the outcome of surgery. The subjective evaluation (SSV score) was 69% (30–100%). The active anterior elevation (AAE) was 117° ± 43° (30–180°), the active external rotation (AER) was 24° ± 20° (0–60°), the active internal rotation (AIR) was up to the T12 vertebra (buttocks-T8). The mean Constant score was 60 ± 20 points (24–95). The radiographic analysis revealed a greater tuberosity that was considered migrated, not healed or lysed in 65% of cases. The acromion – implant height was ≤ 7mm in 52% of the patients. In the ‘fracture implant’ group; all the patients were satisfied or very satisfied regarding the outcome of the surgery. The SSV score was 70% (20–100%). The AAE was 132° ± 36° (45–180°), the AER was 34° ± 16° (0–60°), the AIR was up to the L3 vertebra (buttocks-T8). The mean Constant score was 66 ± 16 points (33–95). The radiographic analysis revealed a greater tuberosity that was considered migrated, not healed or lysed in 33% of cases. The acromion – implant height was ≤ 7mm in 30% of the patients. The patients with a healed greater tuberosity in an adequate position had better Constant scores: 71 points versus 54 points for those with a greater tuberosity not healed/lysed or in a bad position (p=0.03). A healed greater tuberosity in an adequate position was obtained more constantly for the patients in the ‘fracture implant’ group (p=0.02).

Conclusion: A healed greater tuberosity in an adequate position is a significant parameter influencing the outcome of hemiarthroplasty for proximal humerus fractures. A fracture designed implant allows better greater tuberosity positioning and healing.

Introduction: Locking plates and screws have been developed to increase stability of internal fixation in osteoporotic bone. The anatomic design should also facilitate the fracture’s reduction in complex cases.

The aim of this study was analyse the results of locking plates used for fractures of the proximal humerus and to look for specific complications.

Method: Forty four patients (mean age 60; 28 males and 16 females) were treated with a locking plate (Philos, Synthes-Sratec Medical, Switzerland) for trauma of the proximal humerus. There were fourteen 2-part (32%), ten 3-part (23%), ten valgus impacted 4-part (23%), 3 classical 4-part (7%) fractures and 7 non-unions (15%). Five patients presented an associated gleno-humeral dislocation. Patients were reviewed clinically and radiologically at a mean follow-up of 21 months (6–42). A particular attention was paid to the occurrence of specific complications.

Results: A deep infection occurred in two patients (4,5%); two others lost the reduction (4,5%), and one broke his plate (2,3%). We observed 9 cases (20,5%) of avascular necrosis. Six patients (13,6%) had an impaction of the fracture with secondary intraarticular protrusion of the locked screws, which induced a secondary glenoid wear. The impaction sometimes occurred even without any evidence of AVN. Protrusion of screws were more frequent in elderly patients or in cases of non-union. Revision surgery (18 operations) was performed in 16 patients (36%) : 9 isolated material removal; 3 revisions for loss of reduction and malposition of the plate; 2 debridement including implantation of a cement spacer with antibiotics; 4 arthroplasties (2 hemi and 2 total shoulder prostheses).

Conclusions: Proximal humerus locking plates and screws, designed to improve stability in osteoporotic bone, may have specific complications. They do not prevent fracture’s impaction, resulting in an intraarticular protrusion of the locked screws, which may induce a severe secondary wear of the glenoid. This phenomenon could be due to the direction of the forces and stresses applied on the humeral head, which is mainly parallel to the screws and not perpendicular to them. Accordingly, patients operated on with proximal humerus locking plates should be regularly controlled. In case of progressive humeral head impaction, the material should be removed before it damages the glenoid.