Failed Scaphoid Surgery With Nonunion, Bone Defect, and Loose Headless Screw Treated by Nonunion Resection, Bone Grafting, and Scaphoid Plating

Objectives: Failed scaphoid osteosynthesis or failed scaphoid nonunion treatment with headless screw lead to a persistent scaphoid nonunion with loose screw, bone defect that is not just between proximal and distal fragment but also around loose screw within these fragments. Stable intramedullary screw fixation is seldom possible. In these cases, stable fixation maybe performed by using scaphoid plate and angular stable screws. Patient and Methods: Ten consecutive patients with persistent scaphoid nonunion after primary osteosynthesis for fracture treatment or after secondary treatment for scaphoid nonunion were within the group. Size of the fragments and bone defect within proximal and distal fragments did not allow operating surgeon to use headless screw for stabilization. All patients were treated by screw removal, nonunion resection, bone grafting by autograft from iliac crest, and stabilization by 1.5 mm scaphoid plate with angular stable screws through volar approach. Surgery was followed by 6 weeks of casting and 6 weeks of protected mobilization. All patients were followed clinically and radiologically at 2, 6, 12, and 24 weeks after surgery until clear signs of healing or nonhealing of the scaphoid. In case of doubts about the healing process, computed tomography (CT) scan was used for confirmation and patients were followed up to 18 months after surgery. In 2 patients with screw inserted from proximal pole, second approach was used with stabilization through volar approach. In 2 patients, removal of the screw was impossible so it was decided to leave the screw within the bone with grafting around it and applying adjacent plate stabilization on top. Results: Eight patients healed the scaphoid. Wrist synovitis disappeared with scaphoid stabilization. Pain subsided to 0 in 6 with return of full wrist motion and function. In 2 cases, residual motion limitation of 20° and 30° with residual pain at the end of motion arc in heavy load persisted. Two patients that failed to heal have persistent wrist synovitis, motion limitation, and pain with function restriction. One of the persistent nonunions was 9 years old nonunion and second had small proximal fragment with difficulty to fix proximal fragment with screws. In 4 patients, plate was removed in second setting. Discussion: Persistent scaphoid nonunion after previous attempts to heal it with loose headless screw within the bone mean structural problem for operating surgeon due to necessity to reconstruc