[Pseudoarthrosis of the scaphoid bone associated with carpal collapse: factors in choice of surgical approach]

The most common fracture involving the wrist is a fracture of the scaphoid bone [1], and only 5% to 10% of these fractures proceed to nonunion. Although not symptomatic initially, most (if not all) nonunions later produce a painful wrist with impaired function, clinically significant loss of motion, increased weakness and degenerative arthritis. Nonunion of the scaphoid bone should be treated by open reduction and internal fixation. Many surgical procedures have been advocated to achieve union. Most widely used technique for the treatment of scaphoid nonunion was described by Russe [2], but this method may overcome the flexion deformity of the scaphoid and carpal deformities. The ununited scaphoid usually undergoes resorption of the fractures surfaces, principally over the anterolateral aspect of the fracture, so that the scaphoid becomes misshapen. The restoration of the exact length and form is enabled by insertion of a tight-fitting trapezoidal corticocancellous graft, a technique described by Fernandez [11].

We analysed the results of treatment of scaphoid nonunion utilized by two bone-grafting techniques and pointed out the need of choice of the best operative method.

From 1977 to 1993, at the Institute of Orthopaedic Surgery and Traumatology in Belgrade, 40 patients were surgically treated for symptomatic nonunion of the scaphoid bone. The mean duration of follow-up was 10.2 years (range, from 6 to 22 years). Eighteen (45%) patients were operated by Fernandez technique and 22 (55%) patients were operated using Russe's technique. Volar approach and Kirschner's wire fixation were performed in both operative methods. We used two rating scales proposed by Cooney [13] to evaluate the results. Objective scale (Table 1a) included the radiographic appearance of the wrist, the range of motion and grip strength. Subjective scale (Table 1b) comprised function, pain perception of a decrease in performance because of limited motion or strength, and satisfaction. These scales were used to compare the objective and subjective results in patients who had postoperatively carpal collapse with the results in patients who had not such deformity.

The union rate was 92.5% in both methods. Russe's technique resulted in union in 20 (91%) of 22 cases with two ununited. Fernandez technique achieved union in 17 (94%) of 18 cases. Fracture union was determined by both clinical and roentgenographic examinations. Correction of the lateral interscaphoid angle was obtained in 14 (82%) patients operated by Fernandez technique and 9 (45%) patients operated by Russe's technique. Correction of dorsal tilt of the lunate were achieved in 6 (30%) patients operated by Russe's technique, and 13 (76.6%) patients operated by Fernandez technique. There was a highly significant correlation (p < 0.01) between increased deformity of the scaphoid and extent of carpal collapse (Graph 1). Also, there was significant difference between two operative techniques regarding correction of lateral interscaphoid angle (p < 0.05). Arthrosis of the wrist was present in all patients. We could not demonstrate a significant difference (p > 0.05) between intensity of degenerative changes and increase of lateral interscaphoid angle, but obviously, the large flexion deformity of the scaphoid the worse intensity of degenerative changes (Graph 2). The grip strength significantly increased after Fernandez technique (p > 0.05) (Graph 3), but wrist motion changed a little. The average objective score was 71 points for the patients in whom the lateral interscaphoid angle was 45 degrees or less, and 63 points for those in whom the angle was more than 45 degrees. This difference was significant (p < 0.05), but we could not demonstrate a significant difference between the two groups in terms of the average subjective score.

In our series, both procedures provided a high union rate [2]. In cases with severe scaphoid shortening and flexion deformity, Russe's procedure has proved to be insufficient to restore anatomic length and correction of carpal alignment [6, 11, 17]. Previous authors have reported that the progression in degenerative changes was slower in patients who had a lateral interscaphoid angle less than 45 degrees [13]. Also, grip strength and range of motion increased in patients in whom flexion deformity of the scaphoid had been corrected [2, 4, 6, 16, 17]. Our study supports these findings, except results regarding the movement. We believe that this was due to postoperative scarring. Discrepancy between the subjective and objective results may have been due to postoperative relief of pain obtained by increased carpal stability or decreased range of motion of the carpal joints due to postoperative scarring. If pain is relieved, patients readily adapt to the functional deficit of decreased range of motion. We concluded that angulatory collapse of the scaphoid resulted in nonunion as well as malunion with secondary functional loss. Recognition and avoidance in acute fractures were important. When recognised late, volar wedge grafting appeared to be a satisfactory method of treatment.