Cervical myelopathy secondary to spinal cord compression at C2 is an uncommon disorder that remains incompletely characterized. The differential diagnosis may include causes as diverse as rheumatoid disease, calcium pyrophosphate deposition (CPPD) disorder, congenital anomalies, neoplastic processes, infectious etiology (including tuberculosis), or posttraumatic disorders (1, 2, 3, 4, 5, 6 and 7). Common features include ventral spinal cord compression and, in most cases, hypertrophic tissue adjacent to the dorsal border of C2. Ligamentous laxity or overt destruction leads to occipitocervical hypermobility (1). Vertical migration of the odontoid into the foramen magnum, with corresponding compression of the brainstem, may occur as well (1,4). This constellation of findings at the craniocervical junction clinically manifests as cervical myelopathy and brainstem compression (1,4).

The initial treatment of these disorders was direct decompression through a transoral approach, resecting at least a portion of the C1 ring, a majority of the odontoid process that includes the insertion of the alar ligaments and the attachment of the transverse ligament (8, 9 and 10). This direct decompression necessarily further destabilized the already hypermobile occipitocervical complex. Dorsal augmentation, immediately or at further follow-up, has been advocated either prophylactically or for persistent pain (5,7, 8, 9, 10 and 11). Controversy arises from clinical experience with modern rigid dorsal fusion, with or without decompression. This may lead to spontaneous regression of the hypertrophic tissue and excellent clinical results (12,13), though no high-quality clinical series of dorsal treatment exists. To this date, no clinical trials exist to compare the results of these two divergent paradigms.

Periodontoid pseudotumor, retroodontoid mass, and vertical migration of the odontoid are terms used to describe the observed development of nonneoplastic, noninfectious abnormal tissue accumulating at the occipitocervical junction, typically in rheumatoid arthritis (RA), which can then lead to disruption of normal relationships between the spine and skull base (1,11,14). The prevalence is difficult to ascertain, as modern treatment of RA has led to drastic decreases in prevalence, but RA is thought to affect around 1% of the adult population, and atlantoaxial subluxation may affect up to 5% of those, increasing with time (7). The pathophysiology is incompletely understood but is thought to be related to pannus formation typical in other sites of rheumatoid involvement (7,9,15). Hypermobility is thought to be a necessary component, as is seen with vertical migration of the odontoid, or basilar invagination, severe manifestations of the same pathologic process.

The severe bony destruction seen in advanced disease is typical of end-stage hypermobile appendicular joints, such as the wrist. Histopathologic examination of transorally resected tissue has been consistent with rheumatoid pannus (14,16), either an inflammatory predominance in active disease, seen in younger patients, or a hypercellular, hypovascular fibrous tissue with minimal synovial tissue. The former, type I, is typical of active disease, while the latter, type II, is more consistent with chronic disease and hypermobility. Based partly on these findings, the authors conclude, “Although rheumatoid disease at the craniocervical junction in its early stages is undoubtedly an erosive process similar to that observed in the hands and feet, by the time the patient reaches the ‘chronic stage’ the pathologic mechanisms are no longer metabolic but are rather mechanical” (14).