To assess the rate of and specific indications for early reoperation after ACDF in a cohort of patients receiving the ACDF as part of their customary care. These results are contrasted with similar patients receiving ACDF as the control arm of five FDA investigational device exemption (IDE) studies.
Multisurgeon retrospective clinical series from a single institution.
One hundred seventy-six patients with spondylotic radiculopathy or myelopathy underwent ACDF by three surgeons between 2001 and 2005 as part of their clinical practices. All patients had at least 2 years of follow-up with final follow-up within 6 months of completion of this study.
Cervical reoperation rates at 2-year follow-up and at 3.5-year follow-up.
Review of medical records and telephone conversations were completed to determine the number of patients who had undergone a revision cervical procedure.
At final follow-up, complete data were available for 159 ACDF patients. Of the 48聽patients who underwent single-level ACDF and met criteria for inclusion in the IDE studies, one patient (2.1%) required additional surgery (adjacent-segment degeneration) within 2 years, the duration of follow-up of the five published IDE studies. Of the 159 patients who received single or multilevel ACDF at a mean follow-up of 3.5 years, 12 patients (7.6%) had undergone revision cervical surgery, with three patients (1.9%) undergoing same-level revisions (posterior fusion) and nine patients (5.7%) undergoing adjacent anterior level fusions. Patients who underwent revision same-level surgery typically had the intervention within the first year (mean, 11 months), whereas those requiring adjacent-level fusions typically had surgery later (mean, 29 months).
The present study identifies a 2.1%rate of repeat surgery within 2 years of a single-level ACDF performed during routine clinical practice, which is lower than that reported in the control arm of the Prestige, ProDisc-C, Bryan, Kineflex|C, and Mobi-C FDA trials (mean=9%). Even with longer follow-up including multilevel cases, our reoperation rate (7.6%) compared favorably with the IDE rates. This discrepancy may reflect different thresholds for reoperation in the control arm of a device IDE study compared with routine clinical practice. Additionally, patients enrolled in the single-level-only IDE trial may have received multilevel procedures outside of the study. This factor could result in a higher rate of subsequent surgeries at adjacent levels not addressed at the index procedure. These data suggest that we need to better understand factors driving treatment and, in particular, decisions to reoperate both in and outside of a device trial.