INTRODUCTION: There are over 150,000 anterior cervical discectomy and fusion (ACDF) surgeries performed per year in the US[1]. Approximately 25% of ACDF patients develop symptomatic adjacent segment disease (ASD) within 10 years of the initial surgery[2]. Biomechanical testing of cadaver specimens indicates that single-level arthrodesis increases adjacent segment stress[3-5], and double-level arthrodesis exacerbates these effects[6.7]. Direct measurement of tissue loading requires complex invasive procedures[8,9], therefore adjacent segment range of motion (ROM) is typically used to infer the effects of arthrodesis on adjacent segment loading. The purpose of this study was to determine in vivo if the number of fused motion segments affects fused and adjacent segment ROM. Because cervical spine mechanics are level-dependent[10], intervertebral ROM was compared at corresponding adjacent motion segments: the C4/C5 ROM was evaluated when comparing superior adjacent effects after C5/C6 or C5/C6/C7 arthrodesis, and the C6/C7 ROM was evaluated when comparing inferior adjacent segment effects after C5/C6 or C4/C5/C6 arthrodesis. It was hypothesized that (1) adjacent segment ROM would increase from pre-surgery (PRE) to 1-year post-surgery (1YR-POST) after single and double-level arthrodesis, (2) that the increase in adjacent segment ROM would be greater after double-level arthrodesis, and (3) that motion at the operated C5/C6 motion segment would be less after single-level than after double-level arthrodesis.
METHODS: Thus far, 48 anterior arthrodesis patients have provided informed written consent to participate in this ongoing IRB-approved study. Twenty-three patients (Table 1) with either C5/C6, C4/C5/C6, or C5/C6/C7 level fusions have returned for post-surgical testing (12 M, 11 F; average age: 50±5 yrs.; age range: 41-60 yrs.). All participants sat upright with their head in the neutral position for a single static neutral trial, then performed dynamic full ROM flexion/extension (3 trials) and axial rotation (3 trials) while biplane radiographs were collected at 30 images/s for 3 seconds. Three-dimensional vertebral motion (C1 to C7 for flexion/extension and C2 to C7 for rotation) was determined with sub-millimeter accuracy using a previously validated volumetric model-based tracking process that matched subject-specific bone models obtained from CT to the biplane radiographs[11]. Intervertebral rotations were calculated using ordered rotations[12] of anatomic coordinate systems created within each vertebra[13]. Differences from PRE to 1YR-POST were assessed using a Wilcoxon signed-rank test and differences between groups in the change in adjacent and operated C5/C6 segment ROM were analyzed with Wilcoxon rank-sum test with significance set at p < 0.05 for all tests. 
RESULTS: Although adjacent segment ROM increased from PRE to 1YR-POST for both single- and double-level arthrodesis groups during rotation and flexion/extension (Figure 1), this increase in ROM failed to reach statistical significance (all p > 0.091). The change in adjacent segment ROM from PRE to 1YR-POST was not significantly different between single- and double-level arthrodesis groups (all p > 0.67) (Figure 1). The amount of motion remaining at the C5/C6 arthrodesis segment was not different between single and double-level arthrodesis groups (all p > 0.10) (Figure 2).
DISCUSSION: The most important finding of this study was that we failed to find a significant change in adjacent segment ROM from pre-surgery to 1 year after surgery after single or double-level arthrodesis and that we were unable to detect a difference in the change in adjacent segment ROM between single versus double-level arthrodesis groups. These results contradict the widely reported increase in adjacent segment motion after arthrodesis based upon cadaveric testing[3-7] and supports the notion that current in vitro biomechanical test protocols fail to replicate the in vivo condition[14]. The current results should be interpreted with the understanding that sample size is still relatively small in this ongoing study and that age-matched controls have not yet been tested to determine if the total adjacent segment ROM is a return to normal kinematics or an increase in ROM beyond what is commonly found in similarly-aged asymptomatic individuals. It is also important to note that total ROM may not be the best predictor of eventual adjacent segment pathology. 
SIGNIFICANCE: These in vivo findings contradict years of cadaveric spine biomechanical testing, possibly because in vitro protocols fail to replicate the in vivo condition. Double-level arthrodesis does not appear to increase adjacent segment ROM more than single-level arthrodesis 1 year after ACDF.

0788 - The Effects of Single Versus Double-Level Fusion on Adjacent Segment Motion 1 Year Post ACDF Surgery