Percutaneous plate fixation of displaced proximal humerus fractures: Do minimally invasive techniques improve outcomes and reduce complications?

OBJECTIVE: The aim of this study was to analyze the outcomes and complications of minimally invasive plate osteosynthesis (MIPO) for displaced proximal humerus fractures and elucidate if the percutaneous technique reduces the rate of avascular necrosis (AVN) without jeopardizing fracture reduction. MATERIAL AND METHODS: 118 patients with a displaced proximal humerus fracture were treated with a polyaxial locking-plate through a standardized percutaneous approach. 73 % were women and mean age was 63 years (18-89) with an average follow-up of 51 months (12-256). There were 32 two-part fractures (27 %), 57 three-part fractures (48 %) and 24 four-part fractures (25 %). Shoulder function at the last follow-up, including ROM and ability to perform daily living activities, was objectively evaluated with an adjusted Constant Score (CS). Subjective patient satisfaction was rated with an Visual Analogic Scale from 0 to 10 and the degree of residual pain with a Verbal Rating Scale (VRS). Radiographic analysis at the most recent follow-up evaluated the presence of AVN, degree of residual medial calcar displacement, and the cervicodiaphiseal angle. RESULTS: Forty patients (34 %) had a complication, 25 of them (21 %) requiring further surgery. ROM at the last follow-up was 131° of elevation (40°-180°), 38° of external rotation (SD: 17.7) and internal rotation to L3. Average adjusted Constant Score was 68 (SD: 17.76). Twenty-one patients (18 %) complained of shoulder pain (14 moderate and 7 severe) and seven were not satisfied. Radiographically, 29 patients (25 %) had varus malunion and 17 patients (14 %) showed some degree of AVN. Patients with varus malunion (CS 64.3 versus 69.8, p = 0.16) and AVN (CS 56.9 versus 70.4, p = 0.005) had lower CS. The presence of a varus malunion was directly related to the degree of initial medial calcar displacement (p = 0.001) and deficient calcar reduction at surgery (p = 0.004). AVN was statistically more prevalent when the medial calcar was inadequately reduced (p = 0.01). CONCLUSIONS: MIPO surgery for proximal humerus fractures through an anterolateral approach does not reduce the rate of mechanical complications or AVN compared with standard ORIF techniques. Moreover, percutaneous plating may preclude adequate medial calcar reduction, leading to humeral head malunion and a worse clinical outcome.