Preservation of Distal Mobile Segments Reduces Risk of Rod Breakage in Adult Spinal Deformity Surgery

Retrospective multicenter observational study. To analyze whether avoiding fusion to the pelvis and preserving distal mobile segments reduces the risk of rod breakage (RB) after adult spinal deformity (ASD) surgery. RB is a major complication in ASD surgery, mostly associated with age, malalignment, and pelvic fixation. However, the impact of different lower instrumented vertebrae (LIV) possibilities as an independent risk factor remains unclear. We included ASD patients (aged 50-75) who underwent fusion of more than six levels, with LIV at or below L4, and a minimum two-year follow-up. Patients were categorized into those who did or did not experience rod breakage. Univariate and multivariate regression analyses assessed biological, surgical, and radiographic risk factors, with particular focus on LIV as an independent variable. Among 642 patients, 84% were female, mean age 64 years, BMI of 26.3, and ASD Frailty Index (ASD-FI) of 0.44. LIV distribution was L4 (52), L5 (44), S1 (48), and Iliac (498). RB occurred in 17.3% (113 patients). Univariate analysis linked RB to various alignment (preoperative global alignment and pelvic retroversion) and surgical factors (blood loss, osteotomies, levels fused, the use of cages, higher one-year coronal Cobb angle, and lower implant density). However, multivariate analysis identified two independent risk factors: implant density (OR: 0.98; 95% CI: 0.97-0.99) and LIV (OR: 0.74; 95% CI: 0.02-0.312), with LIV being the strongest predictor. Notably, RB appearance was significantly lower in patients with LIV at L4 or L5 (2.1%, 2/96) compared with those with LIV at S1 or Iliac (20.3%, 111/546). Selecting the lowest instrumented vertebra at L4 or L5 in ASD surgery significantly reduces rod breakage risk compared with pelvic fixation, independent of other biological, surgical, and radiographic variables. Preserving distal mobile segments, when feasible, may help prevent rod failure.