Internal fixation of displaced fractures of the proximal humerus with a locked intramedullary blade nail

Abstract

The optimum management of displaced proximal humerus fractures remains a contro-versial topic. The purpose of this study was to report 1 year functional and radiological outcomes after intramedullary nailing for displaced proximal humerus fractures using the Marquardt Locking Blade Nail (LBN) system.

Cases of displaced proximal humeral fractures treated with locking blade fixation under the care of a single surgeon from July 2011 to June 2015 were identified. The fracture pattern was classified according to Neer. The cases were predominantly 2 and 3 part fractures with only four 4 part fractures. The primary outcome measures were the Constant score and the Oxford Shoulder Score (OSS) at 1 year. Secondary measures assessed were the progression to bony union, evidence of avascular necrosis (AVN), loss of fracture alignment, humeral neck-shaft angle after fracture healing and implant related problems. Complication data and re-operations were also reviewed.

62 patients were identified, with 11 patients excluded from the analysis (8 cases lost to follow up, 3 died before 1 year). The mean Constant score at 1 year was 76.0±15.3 (range 44 - 100) and a mean OSS at 1 year of 41.6±5.1 (range 27 - 48). The mean post-operative humeral neck-shaft angle post-operatively was 130.6°±9.6° (range 101° - 145°). All patients achieved bony union. There were 4 cases (7.8%) of AVN. 7 cases (13.7%) underwent a secondary operative procedure during the 1 year follow-up period. Two cases (3.9%) demonstrated a varus loss of alignment post-operatively, and there was 1 case (2.0%) of screw penetration of the glenohumeral joint due to fracture impaction. There was no post-operative infection, axillary nerve injury or screw back out.

Fixation using the LBN achieved robust fixation and satisfactory clinical outcomes at one year follow up in 2 and 3 part fractures. Younger patients had better functional outcome compared to the elderly. The number of 4 part fractures is too small to draw any useful conclusion. AVN remains a difficult problem with these fractures.