MANAGEMENT OF PSEUDOARTHROSIS AFTER ARTHRODESIS OF THE SPINE FOR IDIOPATHIC SCOLIOSIS

Sixty-three first, second, or third-time repairs of one or more pseudarthroses were done in fifty-one patients who had had an arthrodesis for idiopathic scoliosis. Forty-five of the patients were female and six were male. The average age was 30.2 years. The indications for the sixty-three repairs were pain (twenty-five repairs), progression of the curve (sixteen), both pain and progression of the curve (twelve), and radiographic changes only (ten). Failure of the implant was identified before 27 per cent of the sixty-three procedures. The pseudarthroses were diagnosed an average of 2.8 years after the initial arthrodeses. Sixty-eight per cent of the defects were visible on plain radiographs preoperatively and 32 per cent were identified at operation. During the time between the original arthrodeses and the repairs of the pseudarthroses, the scolioses increased by a mean of 7 degrees and the kyphoses, by a mean of 10 degrees. Harrington distraction was the most commonly used instrumentation (twenty-six [41 per cent] of the sixty-three procedures), and autogenous iliac bone was the most commonly used material for the graft (thirty-three [52 per cent] of the procedures). Thirty-five (56 per cent) of the sixty-three procedures were performed in patients who had two, three, or four pseudarthroses and twenty-eight (44 per cent), in patients who had one pseudarthrosis. Forty per cent of the pseudarthroses were in the lumbar spine; 33 per cent, in the thoracolumbar spine; and 27 per cent, in the thoracic spine. After an average duration of follow-up of ten years, 60 per cent of all sixty-three first, second, or third-time repairs had resulted in solid fusion of the spine and 62 per cent of the fifty first-time repairs had been successful. The mean loss of correction of the curve was 3 degrees in the frontal plane and 0 degrees in the sagittal plane. The rate of complications was 27 per cent. Of the fifty-one patients, thirteen had seventeen subsequent procedures because of complications. At the time of the most recent follow-up, fifteen patients (29 per cent) had no pain; seventeen (33 per cent), mild pain; and nineteen (37 per cent), moderate or severe pain. A poor outcome was significantly associated with adulthood, pseudarthrosis in the lumbar spine, persistence of a severe curve after the repair, and use of only a distraction implant during the repair. Use of a compression implant significantly improved the rate of healing. Of the six patients who had been treated with anterior arthrodesis, three had recurrent pseudarthroses. The time from the arthrodesis to the repair of the pseudarthroses did not influence the outcome. Since multiple pseudarthroses were commonly found at the time of repair, the entire fusion mass should always be explored. In addition, a compression implant should be used. Finally, it appears to be safe to observe an asymptomatic pseudarthrosis, since the results of delayed repair did not differ from those of earlier repair.