Isolated greater tuberosity fractures make up 17–21% of proximal humeral fractures, 30% are associated with shoulder dislocation. Conservative management of minimally displaced fractures (<5 mm) is recommended. There are few guides to which and how many fractures displace over time. A retrospective analysis of isolated greater tuberosity fractures presenting to a shoulder fracture clinic over 1 year was performed. Patients were identified from shoulder fracture clinic lists and a bluespier database. Radiological fracture displacement was measured from the edge of the defect in the humeral head to the closest edge of the greater tuberosity. All measurements were performed by three oberservers on two occasions. Data was analysed to study the relationship between initial displacement and fracture stability and between concurrent dislocation and fracture stability. Inter-observer analysis was performed.Aims
Methods
Reverse total shoulder arthroplasty (RTSA) provides a surgical alternative to conventional shoulder arthroplasty in the rotator cuff deficient shoulder. Short term data has shown consistent improvements in pain and function but higher complication and failure rates have also been reported. The aims of this study were to identify the early and late complications of RTSA, to establish the frequency of glenoid notching, and to assess the post-operative functional outcomes. 21 patients (22 joints) treated with RTSA at Glasgow Royal Infirmary (GRI) between April 2006 and October 2010 were retrospectively reviewed. Indications for surgery included cuff tear arthropathy, revision hemiarthroplasty and fracture malunion. Complication rates were obtained by analysis of follow up data from Bluespier and case notes. Glenoid notching was graded from x-rays by multiple observers using the Sirveaux classification. Outcome was assessed using the Oxford Shoulder Score (OSS) and range of motion (ROM). The complication rate associated with RTSA was 14.3%, effecting 3 patients. One dislocation and 1 ulnar nerve palsy occurred within 30 days post-op. A late complication was represented by 1 dislocation, which required revision. Glenoid notching occurred in 71.4% (15 of 21patients), though the majority had a low Sirveaux classification (grade 1 or 2). OSS increased post-operatively and showed a linear improvement with time (R2 = 0.81) and ROM increased significantly post-op compared with pre-op (p<0.001). The complication rate associated with RTSA at GRI was lower than that reported in literature and the outcome was good as defined by ROM and OSS. The rate of glenoid notching was higher than literature reports but the significance of this is unclear as notching may not be associated with loosening.
There is continued concern over complication rates (20–30% of cases) in locked proximal humeral plating. The most common sequelae of this is screw penetration of the humeral head. This is associated with natural settling of the fracture, malreducition in varus, insufficent medial support of the fracture. The proximity of the screws to the articular surface can also be influential on outcome if collapse occurs. Our operative technique is to establish the rotation of the humeral head where the drill appears closest to the articular margin (by sequential xray screening) and subtract from this to avoid intra-articular penetration of the humeral head. 55 Consecutive patients of average age 56.4 years (14.7–86.1), 17 male and 38 females, who underwent PHILOS plating were identified using Bluespier database. Xrays were analysed for fracture pattern, restoration of neck-shaft angle, plate positioning, number and configuration of screws and presence of screw penetration both intra-operatively and at postoperative follow-up. There were 6.07 screws used per head (total 330). There was one intraoperative screw penetration and 3 patients had evidence of screw penetration at follow-up, which required implant removal (total screw penetration rate of 7%). There was one case of AVN. The mean neck shaft angle was 137 degrees (anatomical 135 degrees). Accurate reduction of fractures and placement of screws in the humeral head using image intensifier can act to minimise risk of screw penetration and make some of the complications of locked proximal humeral plating avoidable.
Locking plate fixation in proximal humeral fractures has demonstrated good results tempered by a significant rate of loss of fixation. Reported rates of failure are typically around 10% of cases but can be 20% or higher. In addition large series are often made up of a diverse patient population, so we have chosen to focus solely on patients confirmed to have significantly reduced BMD who can be considered a subset at high risk of fixation failure. Twenty-three patients (5 male, 18 female) with a proximal humeral fracture treated by locking plate fixation were confirmed on DEXA scanning to be osteopaenic (17), osteoporotic (4) or severely osteoporotic (2). Patients early in the series were reviewed retrospectively and recalled for an updated assessment where appropriate, and the remainder were followed prospectively. The average age was 66 years (range 49 to 82). Follow up was for an average of nine months following surgery (range 2 and a half to 28 months). 17 patients underwent surgery for acute injuries and 6 for established surgical neck non-unions. Seven injuries were 2-part fractures, 12 3-part, 3 were 4-part and one a 2-part surgical neck non-union.1 plate failed due to complete loss of fixation within 2 months in a patient with severe osteoporosis and was treated with removal of metalwork. This was the only injury that failed to unite. Avascular necrosis occurred in three patients with two revised to a hemiarthroplasty. 1 patient had ongoing pain and underwent removal of the plate. Our series demonstrated that locking plate fixation of proximal humeral fractures is associated with a low rate of fixation failure and satisfactory outcomes in patients with significantly reduced bone mineral density.
