[Long-term outcome of non-vascularized grafts for carpal scaphoid nonunion: 58 cases with 8.8 year follow-up]

We conducted this retrospective study to analyze outcome obtained in 58 cases of carpal scaphoid nonunion treated with a non-vascularized corticocancellous graft. Our objective was to search for factors influencing outcome and identify appropriate indications.

Between 1984 and 1999, 103 patients were treated with a non-vascularized corticocancellous graft. Fifty-seven patients (58 wrists) were retained for analysis (4 deaths, 30 not attending follow-up consultation far from their residence, and 12 lost to follow-up). Mean follow-up was 106 months (range 19-212). Mean age was 36 years (20-62). Mean time to diagnosis of fracture was 35 months (6-252); the fracture was not recognized initially in 45 patients. According to the Schernberg classification, fractures were localized in zone II (n = 11), zone III (n = 40), and zone IV (n = 7). Using the Alnot classification, nonunion was grade I (n = 13), grade IIA (n = 20), grade IIB (n = 22), grade IIIA (n = 2), and grade IIIB (n = 1). Mean immobilization after grafting was 2.7 months.

At three months, bone healing was achieved in 42 wrists. Thirty-six patients were very satisfied, 18 satisfied and four not satisfied. Twenty-seven patients presented significant pain at the graft harvesting site which regressed in all. There were no harvesting sequelae. Compared with the healthy wrist, we noted reduced flexion (8.7 degrees on average), extension (12.8 degrees on average), and radial inclination (10.5 degrees on average). Other wrist motions were not affected, particularly pronation and supination. The index of carpal height was 0.54 on average (range 0.47-0.57); carpal height was diminished in nine wrists. The radiolunate angle was 4.8 degrees on average (-17 degrees to +27 degrees). Dorsal intercalated segmental instability (DISI) was noted in twenty wrists. Bone healing was not achieved in eleven wrists, seven of which presented initial necrosis of the proximal pole. Pain improved in all patients. Persistence of DISI favored osteoarthritic degeneration.

With this technique, we achieved bone healing in 81% of the cases. Non-vascularized grafting cannot provide cure in the event of advanced osteoarthrtic degeneration. Correction of DISI appears to be an indispensable element for the prevention of osteoarthritis. This technique should not be proposed for patients with a necrotic proximal pole, a vascularized graft would be preferable.