Does stopping at C2 matter? A comparative analysis of alignment and craniocervical compensation in C2–T1 versus C3–T1 posterior cervical fusion

Objective: Selection of the proximal upper instrumented vertebra (UIV) in multilevel posterior cervical fusion (PCF) remains controversial. This study compared sagittal alignment, craniocervical compensatory behavior, and perioperative outcomes between C2–T1 and C3–T1 PCF constructs. Methods: A retrospective cohort study was performed, including 79 adult patients who underwent PCF from either C2–T1 ( n = 41) or C3–T1 ( n = 38) with a minimum of 18 months of follow-up. Demographic, perioperative, and complication data were collected. Preoperative and postoperative radiographs were analyzed for global cervical alignment (C2–7 Cobb angle, cervical sagittal vertical axis [cSVA]) and upper cervical parameters, including C2 slope, C0–C2 angle, jaw slope, and T1 slope. Between-group comparisons and linear regression analysis were performed. Results: Baseline demographics and preoperative radiographic parameters were similar between groups. The C3–T1 group demonstrated significantly shorter operative time (198 ± 41 vs. 273 ± 45 min, P < 0.001) and a lower incidence of C5 palsy (5.3% vs. 19.5%, P = 0.048). Radiographically, C2–T1 constructs achieved superior global sagittal correction, with greater improvement in cervical lordosis (+12.5° vs. −4.5°, P = 0.01) and reduction in cSVA (−16.0 mm vs. +4.4 mm, P = 0.04). In contrast, C3–T1 constructs demonstrated significantly greater increases in C2 slope (+9.8° vs. +2.7°, P = 0.045), jaw slope (+4.1° vs. +1.2°, P = 0.038), and T1 slope (+9.4° vs. +1.7°, P = 0.04). Compensation in the C2–T1 group occurred predominantly through increased C0–C2 angulation (+4.8° vs. +1.4°, P = 0.02). Preoperative T1 slope was a strong predictor of postoperative C2 slope increase in the C3–T1 cohort (B = 1.45, P < 0.001). Conclusion: C2–T1 PCF provides superior global sagittal alignment and shifts compensatory demands to the occipitocervical junction, whereas C3–T1 constructs rely more heavily on compensation at the c2-3 level, particularly in patients with higher T1 slope. While C3–T1 fixation reduces operative morbidity, UIV selection should be individualized, incorporating preoperative sagittal parameters to optimize long-term biomechanical outcomes. Clinical Relevance: The findings highlight the importance of individualized surgical planning in multilevel PCF, demonstrating that preoperative sagittal parameters should be incorporated to optimize the selection of the proximal UIV at C2 versus C3, balancing reduced operative morbidity with superior long-term sagittal alignment restoration. Level of Evidence: III