Karen S Myung1, David L Skaggs2, Charles E Johnston3, Behrooz A Akbarnia4. 1. Children's Orthopaedic Center, Children's Hospital Los Angeles, Orthopaedic Surgery #69, Los Angeles, CA 90027, USA. 2. Children's Orthopaedic Center, Children's Hospital Los Angeles, Orthopaedic Surgery #69, Los Angeles, CA 90027, USA. Electronic address: DSkaggs@chla.usc.edu. 3. Department of Orthopaedic Surgery, Texas Scottish Rite Hospital for Children, 2222 Welborn Street, Dallas, TX 75219, USA. 4. San Diego Center for Spinal Disorders, 4130 La Jolla Village Drive, Suite 300, La Jolla, CA 92037, USA.
Abstract
STUDY DESIGN: Retrospective review of a multicenter database. OBJECTIVES: To evaluate the incidence of implant-related complications of pedicle screws versus hooks in children with early-onset scoliosis treated with growing rods. SUMMARY OF BACKGROUND DATA: Because growing rods have a high rate of implant complications compared with spinal fusion, this challenging, fusionless group of patients is a provocative environment in which to evaluate implant-related complications of pedicle screws. METHODS: A total of 159 patients with growing rods treated at 18 institutions were included. Inclusion criteria were children aged less than 10 years who had growing rod surgery between 1998 and 2008 with minimum 2-year follow-up from index surgery. Charts and radiographs were evaluated only for complications directly related to a pedicle screw or hook. The researchers evaluated 464 pedicle screws and 643 hooks. RESULTS: Of the 464 pedicle screws, there were 17 complications (3.7%) directly related to screws: acute loss of fixation (2), migration (14), and breakage (1). Of the 643 hooks, there were 47 complications (7.3%): acute loss of fixation (28), migration (16), and unspecified loss of fixation (3). When loss of fixation occurred, the mean time to loss of fixation was similar for both implants: 33 months for hooks and 30 months for screws (p = .95). There were no complications involving neurologic or vascular injury directly related to a hook or screw. CONCLUSIONS: Pedicle screws in growing rod constructs had fewer implant-related complications than hooks in patients with early-onset scoliosis in a minimum 2-year follow-up period (p = .02). It is encouraging that there were no neurological or vascular injuries associated with either implant in 159 children with over 4.5 years of mean follow-up, in a provocative, fusionless environment.
STUDY DESIGN: Retrospective review of a multicenter database. OBJECTIVES: To evaluate the incidence of implant-related complications of pedicle screws versus hooks in children with early-onset scoliosis treated with growing rods. SUMMARY OF BACKGROUND DATA: Because growing rods have a high rate of implant complications compared with spinal fusion, this challenging, fusionless group of patients is a provocative environment in which to evaluate implant-related complications of pedicle screws. METHODS: A total of 159 patients with growing rods treated at 18 institutions were included. Inclusion criteria were children aged less than 10 years who had growing rod surgery between 1998 and 2008 with minimum 2-year follow-up from index surgery. Charts and radiographs were evaluated only for complications directly related to a pedicle screw or hook. The researchers evaluated 464 pedicle screws and 643 hooks. RESULTS: Of the 464 pedicle screws, there were 17 complications (3.7%) directly related to screws: acute loss of fixation (2), migration (14), and breakage (1). Of the 643 hooks, there were 47 complications (7.3%): acute loss of fixation (28), migration (16), and unspecified loss of fixation (3). When loss of fixation occurred, the mean time to loss of fixation was similar for both implants: 33 months for hooks and 30 months for screws (p = .95). There were no complications involving neurologic or vascular injury directly related to a hook or screw. CONCLUSIONS: Pedicle screws in growing rod constructs had fewer implant-related complications than hooks in patients with early-onset scoliosis in a minimum 2-year follow-up period (p = .02). It is encouraging that there were no neurological or vascular injuries associated with either implant in 159 children with over 4.5 years of mean follow-up, in a provocative, fusionless environment.