Session VI - Upper Extremity

*Fixation of Fractures of the Proximal Humerus: Experience with a New Locking Plate

Marc Saudan, MD (e-Synthes-Stratec); Richard E. Stern, MD (e-Synthes-Stratec); Anne Lubbeke, MD; Robin E. Peter, MD; Pierre Hoffmeyer, MD; University Hospital of Geneva, Geneva, Switzerland

Purpose: Although the majority of fractures of the proximal humerus can be managed nonoperatively, displaced fractures generally require surgical repair to achieve a satisfactory reduction and allow for early mobilization of the shoulder and upper extremity. In the younger patient, the better bone quality in the humeral head allows for secure screw fixation. However, the osteoporosis found in the elderly patient makes internal fixation problematic and is frequently the cause of failure of fixation and poor results. We report on a prospective review of the first 31 patients treated with a new locking plate, Philos (Synthes-Stratec, Oberdorf, Switzerland), for the treatment of displaced fractures of the proximal humerus.

Methods: The titanium plate (also in stainless steel) is available in two lengths, with either three or five holes in the distal portion that is fixed to the proximal diaphysis. Proximally, there are nine holes for 3.5-mm locking screws to be inserted into the head fragment. Ten small holes can be used for anchoring sutures for fragments of the greater or lesser tuberosity.

Between December 2000 and December 2001, 31 patients with proximal humeral fractures were operated upon with use of the new proximal humerus plate. There were 15 men and 16 women with a mean age of 64 years (range, 16 to 101). Twelve patients were under the age of 65 (mean age, 44.6). Nineteen patients (5 men and 14 women) were older than 65 (mean age, 76.3). The mechanism of injury in the older group was always a fall from the patient's own height, with injuries in the younger group secondary to a high-energy mechanism such as a motor vehicle accident. Associated fractures were only noted in the younger group. According to the AO/OTA classification there were 8 11-A3, 16 11-B2, and 7 11-C2 fractures. The operations were performed by residents in various levels of training, but all were personally supervised by an attending surgeon familiar with the technique. The operation was performed with a deltopectoral approach. After reduction and provisional stabilization with Kirschner wires, sutures were placed around any fragments of the greater and lesser tuberosity for passage through holes in the proximal portion of the plate. A minimum of four to six locking screws were inserted into the proximal humeral head fragment. Distally the plate was fixed either with locking or traditional screws. A standard-length plate was used in 21 patients, and a longer plate in 10 patients because of distal extension of the fracture. Postoperatively, a simple sling was used for patient comfort, and physiotherapy was begun with passive and active assisted exercises within the first few days after the operation. No patient required any period of more formal immobilization. We used the score of Constant and Murley as our outcome measurement tool. In addition, we used the age-adjusted Constant score to better evaluate the results between the younger and older group of patients.

Results: The mean interval between the time of accident and the surgical procedure was 9 days (range, 2 to 24). Surgery was performed under general anesthesia in all patients and the mean time for the surgical procedure was 135 minutes (range, 90 to 195). The duration of hospitalization was longer for those patients over the age of 65 years, reflecting the need for correction of associated medical and social problems. The mean follow-up was 12.5 months (range, 6 to 18). The mean raw Constant score in patients less than 65 years of age was 72.6, with an age-adjusted score of 82.0. In the older group of patients, the mean raw score was 52.3, with an age-adjusted score of 81.2. Complications included one case of plate breakage, one nonunion associated with avascular necrosis, and two patients whose fractures had united with partial head necrosis. All cases of avascular necrosis occurred in C2 (four-part) fractures in the older age group. Technical errors were noted in two patients in whom a locking screw in each was placed just through the subchondral bone but was not so recognized on the intraoperative radiographs. The offending screws were left in place, and at the most recent follow-up the patients had no complaints.

Conclusions: Although the treatment of displaced proximal humerus fractures remains controversial, the majority opinion is that such fractures usually demand operative intervention to correct the malposition of the fracture fragments and to maintain enough stability of the fracture to allow for early mobilization. However, the ideal operative procedure is not well defined. Although plate-and-screw fixation is successful in the younger patient, such a technique is dependent upon the hold of the screws in the osteoporotic bone of the elderly. Our results with the locking plate were equally good in the younger and older groups of patients. Avascular necrosis only occurred in the C2 (four-part) fractures. On the basis of this experience, we favor prosthetic replacement for these fractures in the elderly patient with use of the locking plate for all other displaced fractures in any age group.