Primary drivers of adult cervical deformity: Prevalence, variations in presentation, and effect of surgical treatment strategies on early postoperative alignment

BACKGROUND: Primary drivers (PDs) of adult cervical deformity (ACD) have not been described in relation to pre- and early postoperative alignment or degree of correction. OBJECTIVE: To define the PDs of ACD to understand the impact of driver region on global postoperative compensatory mechanisms. METHODS: Primary cervical deformity driver/vertebral apex level were determined: CS = cervical; CTJ = cervicothoracic junction; TH = thoracic; SP = spinopelvic. Patients were evaluated if surgery included PD apex, based on the lowest instrumented vertebra (LIV): CS: LIV ≤ C7, CTJ: LIV ≤ T3, TH: LIV ≤ T12. Cervical and thoracolumbar alignment was measured preoperatively and 3 mo (3M) postoperatively. PD groups were compared with analysis of variance/Pearson χ², paired t-tests. RESULTS: Eighty-four ACD patients met inclusion criteria. Thoracic drivers (n = 26) showed greatest preoperative cervical and global malalignment against other PD: higher thoracic kyphosis, pelvic incidence-lumbar lordosis (PI-LL), T1 slope C2-T3 sagittal vertical axis (SVA), and C0-2 angle (P < .05). Differences in baseline-3M alignment changes were observed between surgical PD groups, in PI-LL, LL, T1 slope minus cervical lordosis (TS-CL), cervical SVA, C2-T3 SVA (P < .05). Main changes were between TH and CS driver groups: TH patients had greater PI-LL (4.47^◦ vs −0.87^◦, P = .049), TS-CL (−19.12^◦ vs −4.30, P = .050), C2-C7 SVA (−18.12 vs −4.30 mm, P = .007), and C2-T3 SVA (−24.76 vs 8.50 mm, P = .002) baseline-3M correction. CTJ drivers trended toward greater LL correction compared to CS drivers (−6.00^◦ vs 0.88^◦, P = .050). Patients operated at CS driver level had a difference in the prevalence of 3M TS-CL modifier grades (0 = 35.7%, 1 = 0.0%, 2 = 13.3%, P = .030). There was a significant difference in 3M chin-brow vertical angle modifier grade distribution in TH drivers (0 = 0.0%, 1 = 35.9%, 2 = 14.3%, P = .049). CONCLUSION: Characterizing ACD patients by PD type reveals differences in pre- and postoperative alignment. Evaluating surgical alignment outcomes based on PD inclusion is important in understanding alignment goals for ACD correction.