OBJECTIVE

Rod fracture (RF) after adult spinal deformity (ASD) surgery is reported in approximately 6.8%-33% of patients and is associated with loss of deformity correction and higher reoperation rates. The authors' objective was to determine the effect of accessory supplemental rod (ASR) placement on postoperative occurrence of primary RF after ASD surgery.

METHODS

This retrospective analysis examined patients who underwent ASD surgery between 2014 and 2017 by the senior authors. Inclusion criteria were age > 18 years, ≥ 5 instrumented levels including sacropelvic fixation, and diagnosis of ASD, which was defined as the presence of pelvic tilt ≥ 25°, sagittal vertical axis ≥ 5 cm, thoracic kyphosis ≥ 60°, coronal Cobb angle ≥ 20°, or pelvic incidence to lumbar lordosis mismatch ≥ 10°. The primary focus was patients with a minimum 2-year follow-up.

RESULTS

Of 148 patients who otherwise met the inclusion criteria, 114 (77.0%) achieved minimum 2-year follow-up and were included (68.4% were women, mean age 67.9 years, average body mass index 30.4 kg/m2). Sixty-two (54.4%) patients were treated with traditional dual-rod construct (DRC), and 52 (45.6%) were treated with ASR. Overall, the mean number of levels fused was 11.7, 79.8% of patients underwent Smith-Petersen osteotomy (SPO), 19.3% underwent pedicle subtraction osteotomy (PSO), and 66.7% underwent transforaminal lumbar interbody fusion (TLIF). Significantly more patients in the DRC cohort underwent SPO (88.7% of the DRC cohort vs 69.2% of the ASR cohort, p = 0.010) and TLIF (77.4% of the DRC cohort vs 53.8% of the ASR cohort, p = 0.0001). Patients treated with ASR had greater baseline sagittal malalignment (12.0 vs 8.6 cm, p = 0.014) than patients treated with DRC, and more patients in the ASR cohort underwent PSO (40.3% vs 1.6%, p < 0.0001). Among the 114 patients who completed follow-up, postoperative occurrence of RF was reported in 16 (14.0%) patients, with mean ± SD time to RF of 27.5 ± 11.8 months. There was significantly greater occurrence of RF among patients who underwent DRC compared with those who underwent ASR (21.0% vs 5.8%, p = 0.012) at comparable mean follow-up (38.4 vs 34.9 months, p = 0.072). Multivariate analysis demonstrated that ASR had a significant protective effect against RF (OR 0.231, 95% CI 0.051-0.770, p = 0.029).

CONCLUSIONS

This study demonstrated a statistically significant decrease in the occurrence of RF among ASD patients treated with ASR, despite greater baseline deformity and higher rate of PSO. These findings suggest that ASR placement may provide benefit to patients who undergo ASD surgery.