1487. Artificial Discs in Cervical Disc Arthroplasty: A Network Meta-Analysis for Comparison of Long-term Outcomes

Authors: Mohamad Bydon, MD, FAANS ; Waseem Wahood, MS; Yagiz Yolcu, MD; Sandy Goncalves, MS; Panagiotis Kerezoudis, MD, MS (Rochester, MN)

Introduction: Cervical disc replacement(CDR) has emerged as a safe and effective alternative to ACDF for the surgical treatment of degenerative cervical disc disease. Although comparison of two techniques has been extensively studied in the literature, a thorough comparison of all artificial discs within each other hasn't been performed yet. The objective of the present study was to examine the long-term outcomes of five artificial discs. Methods: A comprehensive electronic literature search in PubMed/Medline, Scopus, Cochrane and EMBASE was conducted for studies on CDR devices for all years available. Heterotopic ossification(HO), adjacent segment disease(ASD), and reoperation comprised the primary outcomes of interest. Pooled descriptive statistics with effect size(ES) and 95% confidence interval were used to synthesize the outcomes for each device. Results: Eighty-four studies(n=7,066) were included in the analysis. Thirty-eight studies with 3,408 patients belonged to Bryan group; thirteen studies with 1,463 patients belonged to Mobi-C group; eight studies with 1,001 patients belonged to Prestige-LP group; twelve studies with 453 patients belonged to ProDisc-C group and thirteen studies with 741 patients belonged to Discover group. Comparison of the incidence of HO revealed a significant variability between the five devices(p<0.001) with ProDisc-C(ES:0.38,95%CI:24%-54%) having the highest incidence rate. Overall rate of ASD was 17%(95%CI:11%-24%) with significant variability(p=0.005),with Bryan having the highest incidence rate of 22%(95%CI:11%-33%) and Discover having the lowest incidence rate of 5%(95%CI:1%-8%). With regard to 4-year reoperation risk, the overall incidence rate was 2%(95%CI:1%-3%) with non-significant variability between devices(p=0.224). The highest rate was observed in the Discover group(ES 3%,95%CI:0%-10%), whereas Mobi-C was associated with the lowest rate(ES 1%,95%CI:0%-2%). Conclusion: Our results show that surgical and clinical outcomes may differ among different CDR devices. These results are important for surgeons so that in addition to patient selection, providers are also aware of device specific nuances while making a decision to perform CDR.