Failure in cervical total disc arthroplasty: single institution experience, systematic review of the literature, and proposal of the RUSH TDA failure classification system

BACKGROUND CONTEXT Cervical total disc arthroplasty (TDA) is an alternative procedure to anterior cervical discectomy and fusion that facilitates neural decompression while both preserving motion of the spinal unit and decreasing the risk for degenerative changes at adjacent segments. However, due to its more recent introduction in clinical practice and low complication rates, the modes by which TDA may fail remain to be described. PURPOSE This study sought to identify the modes and frequencies of cervical TDA failure in order to propose a novel classification system. STUDY DESIGN Retrospective Cohort and Systematic Review. PATIENT SAMPLE Patients who underwent single or two-level TDA for cervical radiculopathy or myelopathy at a single institution and in the literature of medium and large prospective studies. OUTCOME MEASURES Cervical TDA failure, defined as subsequent surgical intervention at the index segment. METHODS This study retrospectively reviewed patients who underwent single or two-level TDA for cervical radiculopathy or myelopathy at a single institution to identify the potential implant failure modes. A systematic review and meta-analysis of prospective data in the literature was then performed to further supplement failure mode identification and to describe the rates at which the various failure types occurred. Statistical analysis included between-group comparisons of Non-Failed and Failed patients and frequencies of each failure type among Failed patients. RESULTS A retrospective review at our institution of 169 patients (201 levels) identified eight failures, for a failure rate of 4.7%. Additionally, seven patients were revised who had the primary surgery at an outside institution. The systematic review of 3976 patients (4525 levels) identified 165 (4.1%) additional failures. Using this data, six primary failure types were classified, with several subtypes. These include recurrent or persistent index-level stenosis (Type I); migration (Type II) presenting as gross extrusion (A) or endplate failure with subsidence/acute fracture (B); instability (Type III) due to mechanical loosening (A), septic loosening (B), or device fracture (C); device motion loss (Type IV) such as “locking” of the device in kyphosis; implantation error (Type V) due to malposition (A) or improper sizing (B); and wear (Type VI) either without osteolysis (A) or with wear-particle-induced osteolysis (B). Stenosis (Type I) was the most common mode of failure found both through retrospective review and in the literature. CONCLUSIONS Cervical TDA fails through six primary mechanisms. While rates of certain failures requiring subsequent surgical intervention are low, it is possible that these complications may become more prevalent upon further longitudinal observation. Thus, future application and validation of this classification system is warranted to evaluate how failure frequencies change over time and with larger patient samples.