Evaluation of Occipitocervical Arthrodesis Rates with Screw-based Fixation and Osteoinductive Fusion Adjuncts

Occipitocervical (OC) instability may be associated with neurologic impairment and even death. There is a paucity of research on the rate of arthrodesis utilizing modern screw-based constructs coupled with adjuvant osteoinductive agents. We reviewed our experience with OC constructs and compared the fusion rate, functional outcome, and rate of adverse events between recombinant human bone morphogenetic protein (BMP)-2, autologous iliac crest bone graft (ICBG), a combination of BMP and ICBG, and local bone autograft alone. We performed a retrospective cohort analysis of all adult admissions for operative treatment of OC instability utilizing segmental screw-based constructs for OC arthrodesis between January 2003 and September 2012. Data concerning demographic characteristics, diagnostic and procedural details, radiographic pathology, and clinical course were abstracted from medical records. The primary end point was evidence of stable fixation and osseous union on either dynamic lateral radiographs or computed tomography (CT) imaging at most recent follow-up. Secondary end points included functional outcome as determined by Nurick scale and Neck disability index (NDI) at1year postoperation, as well as perioperative morbidity and mortality at 30 days and 3 months. During the study period, 94 patients (mean age: 62 +/- 18 years) underwent OC fixation with segmental screw-based constructs. The four fusion adjunct cohorts analyzed included local autograft alone (32%), ICBG (41%), BMP (14%), or a combination of ICBG and BMP (14%). Notably, demineralized bone matrix was also used in 61% of cases overall, but its use did not differ significantly among the four cohorts (p=0.28). Median radiographic follow-up was 6 months postoperatively (range: 1.5-54 months). Clinical outcomes were assessed at a median postoperative follow-up of 45 months (range: 12-87 months). Overall, radiographic evidence of arthrodesis was present in 83% of patients assessed and was not significantly different between adjunct cohorts (local autograft 92%, ICBG 77%, BMP 88%, and combination of ICBG and BMP 83%; p=0.79). This finding persisted despite adjustment for age, pathology, number of levels instrumented, and attendant procedures. Importantly, neither the presence of arthrodesis nor fusion adjunct was significantly associated with functional outcome in both univariate and adjusted regression models. Additionally, perioperative adverse events occurred in 23% of cases and did not vary significantly in incidence or severity between fusion adjunct cohorts. We present a large series of patients treated for OC instability with rigid fixation utilizing modern segmental screw-based constructs. The use of adjuvant osteoinductive agents (BMP, ICBG, or a combination) produced equivalent rates of arthrodesis, functional outcome, and adverse events compared with use of local autograft alone.