Arthroplasty for proximal humerus fractures, nonunions, and malunions

Treatment of complex fractures and fracture-dislocations of the proximal humerus in both the acute and late settings represent some of the most difficult injuries to assess and treat in the shoulder girdle. While the vast majority of proximal humerus fractures (more than 85%) are nondisplaced and amenable to nonoperative treatment, the remaining minority may involve multiple fragments or parts, comminution, and articular surface damage and represent a significant therapeutic challenge. Based on Codman's anatomical description of fractures about the proximal humerus [1], Neer [2] published a comprehensive four-part classification scheme of these complex injuries in 1970. Wide use of this system has allowed for more uniformity in evaluation and has directed treatment in this area for the last three decades. In conjunction with adequate evaluation of the proximal humeral anatomy, whether it is an acute fracture or the sequelae of an old injury, the therapeutic course should be guided by the surgical goal of anatomic restoration of the proximal humerus and patient-determined factors, such as age, function, and quality of remaining bone. With respect to acute fractures, all two-part, most three-part, and some fourpart fractures are well treated by closed reduction or open reduction and internal fixation techniques. When soft bone, impaired head vascularity, or articular damage make fixation difficult or ill-advised, prosthetic arthroplasty is a reasonable option. Use of a humeral head prosthesis for proximal humerus fractures was first reported in the 1950s. Multiple designs were emerging [3'7], but Neer's metal prosthesis became the most commonly used. In 1953, Neer [7] reported the first use of his prosthetic humeral head in a complex fracture-dislocation of the proximal humerus. In 1955 and 1970, he reported his series of 27 and 43 patients, respectively, with fracture-dislocations treated with prosthetic replacement [7, 8]. Advances in design since Neer's first-generation monoblock prosthesis and his subsequent redesign in 1973 have included the use of modular head and stem components. Restoration of soft tissue tensioning may be more reliable with a wider variety of head sizes and offset. In addition, should conversion to a total shoulder become necessary, removable heads allow for better visualization of the glenoid while permitting preservation of a well-fixed stem [9, 10].