Reoperation for Symptomatic Nonunions in Atlantoaxial (C1-C2) Fusions with and without Bone Morphogenetic Protein: A Cohort of 108 Patients with >2 Years Follow-Up.

Abstract

To determine if there is a difference in reoperation rates for symptomatic nonunions in atlantoaxial (C1-C2) fusions with or without bone morphogenetic protein (BMP) using data from a national spine registry and to analyze the different types of bone grafts used in the non-BMP group. Data from the Kaiser Permanente spine registry were used to identify patients with C1-C2 fusions with >2 years follow-up. Patient characteristics, diagnosis, operative times, length of stay, and reoperations were extracted from the registry. The data set was divided into patients with and without BMP. Further analysis was made of the different types of non-BMP grafts as well as the instrumentation used. In our cohort, we found 58 patients (53.7%) with BMP and 50 patients (46.3%) without with an average follow-up time of 5 years (interquartile range, 2.04-8.49). The BMP versus non-BMP groups differed in admitting diagnosis, operative times, length of stay, and follow-up times. There were no reoperations for symptomatic nonunions in both groups. The non-BMP group included iliac crest graft (with or without allograft [+/-] allograft); lamina (+/- allograft); and allograft alone. Using one of the largest retrospective studies on C1-C2 fusions with and without BMP, we found no difference in reoperation rates for symptomatic nonunions. For the non-BMP group, we found that lamina (+/- allograft) or allograft alone may also be just as effective as iliac crest graft (+/- allograft) in having no reoperations for symptomatic nonunions.