An analysis of heterotopic ossification in cervical disc arthroplasty: a novel morphologic classification of an ossified mass

Background context: Although the precise cause of heterotopic ossification (HO) remains unclear, it is certain that it increases with time. The reason why the reported occurrence rate has been given as a wider range is that there were no clinical reports that have addressed the occurrence rate based on the morphology and position.

Purpose: The aim of this retrospective study was to determine whether radiological parameters had an influence on the formation of HO and to compare the results after cervical arthroplasty using Bryan (Medtronic Sofamor Danek, Memphis, TN, USA), PCM (Cervitech, Rockaway, NJ, USA), and Prestige LP (Medtronic Sofamor Danek) implants.

Study design/setting: A retrospective study.

Patient sample: Eighty-one patients were included.

Outcome measurement: The occurrence of HO was determined at the four corners of the disc space according to the McAfee classification system. Heterotopic ossifications were classified into Type 1, Type 2, and Type 3 HOs (end plate, traction spur, and teardrop types) based on their morphologic features. The presence of preoperative ossifications, sex, type of device, operated level, hybrid implantation, depth behind the prosthesis, cervical lordotic angle, and segmental angle between the footplates of the prosthesis were assessed as predictors in terms of location and morphologic features.

Methods: Eighty-one patients after 95 cervical arthroplasties using the Bryan (35 segments), PCM (30 segments), and Prestige LP implants (30 segments) underwent postoperative radiographs or three-dimensional computed tomography with a mean follow-up period of 46, 39, and 30 months, respectively, after the operation.

Results: The overall incidence of all cases and one-level subgroup were 64.2% and 60.3%, respectively. According to the types of device, the incidence of HO was 49% (Bryan), 80% (PCM), and 60% (Prestige). Type 1 HO (62.1%) was found only in the posterosuperior disc space. Type 2 HO (13.7%) was primarily detected in the anterosuperior disc space rather than the posterosuperior disc space (3.2%). Type 3 HO (4.2%) developed only in the anterior disc space. In the anterior disc space, the incidence of Type 2 and Type 3 HOs was highest in the PCM group. In the posterior disc space, the Bryan group showed a lower proportion in the high McAfee class than the other device groups. The occurrence of Type 1 HO in the posterosuperior disc space was significantly related with the presence of preoperative ossification (p=.030), ossification in the ligamentum nuchae (p=.027), male sex (p=.042), and PCM device (p=.012). A well-fitting (p<.002) and less lordotic segmental angle (p<.015) were correlated with Type 1 HO. Implantation in the upper cervical level (p=.016) and hybrid implantation with cage (p=.033) or artificial disc (p=.048) on the upper adjacent level were significant risk factors for anterior Type 2 HO. Cervical lordotic angle at 1 month after surgery had a significant connection with the occurrence of anterior Type 2 HO in both groups of all cases (p=.032) and one-level subgroup (p=.000).

Conclusions: Type 1 HO developed mostly in the posterior disc space. Type 2 HO was the dominant type in the anterior disc space. Type 3 HO developed only in the anterior disc space. It is certain that both Type 1 and Type 2 HOs are related to biomechanical stresses (compressive force for Type 1 HO and traction force for Type 2 HO). It is suggested that a cervical arthroplasty should be selected in terms of the implant level, hybrid conditions on the upper adjacent segment, disc design vulnerable to the pseudotranslation, the presence of preoperative ossification, and fitting implants to end plates to reduce the development of HO.