Heterotopic Ossification in the Cervical Total Disc Replacement with Discover Prostheses: A Retrospective 5-Year Follow-Up Study

Study background: Heterotopic ossification (HO) is a well-known complication of cervical total disc replacement (CTDR), a successful alternative to cervical decompression and fusion for degenerative disc disease. Objective: The aim of our study was to evaluate the incidence of HO in single and double-level Discover disc replacement, a relatively new type of artificial implants. Moreover, the related clinical and radiological outcomes and the possible influencing factors were analysed. Methods: Twenty-six women and 39 men were retrospectively followed-up for 5 years. They consecutively underwent one or two-level cervical Discover arthroplasty. Clinical outcome was assessed using the visual analogue scale (VAS). The angular range of motion (ROM) was calculated with periodic dynamic X-rays. A Student’s t test was used to analyse the association between the occurrence of HO and various risk factors. Results: Among the 65 patients enrolled, the overall incidence of HO was 7.69% and the mean follow-up was 63.29 ± 11.90 months. All the cases were diagnosed as Grade III or Grade IV. The mean VAS showed immediate postoperative improvements and the mean ROM was partially maintained. No displacement of the prostheses was detected. Conclusions: During the 5-year follow-up period, the Discover cervical discs have provided optimal clinical outcomes showing a lower and tardive incidence of HO than other reported implants. Although a longer follow-up with a larger group of patients should be investigated, cervical arthroplasty can be still considered a safe and effective procedure in highly selected patients. *Corresponding author: Giuseppe Maida, Division of Spine Surgery, Santa Maria Maddalena Clinic 1, Via Gorizia, 45030 Occhiobello, Rovigo, Italy, Tel: 00393479697287; Fax: 0039-0532247274; E-mail: giuma@vodafone.it Received October 13, 2014; Accepted November 17, 2014; Published November 19, 2014 Citation: Maida G, Gennaro P, Marcati E (2014) Heterotopic Ossification in the Cervical Total Disc Replacement with Discover Prostheses: A Retrospective 5-Year Follow-Up Study. J Spine 3: 194. doi:10.4172/2165-7939.1000194 Copyright: © 2014 Maida G, et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.


Introduction
Cervical Total Disc Replacement (CTDR) is a relatively new procedure whose main goal is to overcome the limitations of the anterior cervical discectomy and interbody fusion (ACDF), a successful treatment for degenerative cervical disc disease when not responding to medical therapy [1,2]. The main shortcomings of ACDF are the postoperative reduced range of motion (ROM) and the increased mechanical stress on adjacent vertebrae. In fact, the subsequent adjacent segment degeneration (ASD) has been reported to require a secondary surgery in 7-15% of the cases [3][4][5]. Bryan, ProDisc-C, and Prestige are some of the previous most widely used disc replacement devices and their clinical and biomechanical outcomes have already been well-documented [6][7][8][9][10][11][12][13].
Heterotopic ossification (HO) after CTDR is more and more often described and it seems not to be related to the presence of muscular tissue, as in hip and knee arthroplasties [14]. Although its pathogenesis is still unclear, many authors associated it to both systemic and local factors [15]. Reidel et al. [16] first described this phenomenon in 1883 and subsequently it was noted in spinal cord injuries. Ever since, many authors reported HO as one of the main complications in dynamic cervical and lumbar implants [16][17][18][19][20][21][22][23][24][25]. In addition, our group published a rare case of HO secondary to an interspinous/interlaminar device [26].
It has been recently shown a low incidence of HO with a relatively short mean follow-up (31.6 months) on 79 patients treated with Discover artificial disc (DePuy Spine, Raynham, MA, USA) [27]. Our study aims to retrospectively analyse the outcome of Discover CTDR, particularly referring to heterotopic ossification with a longer follow-up.

Patient population
Sixty-five consecutive patients who underwent one or two-level CTDR with Discover prosthesis were enrolled in the study from January 2007 to October 2012. Inclusion criteria were subaxial (C3-C7) disc herniation, degenerative disc disease (DDD) or spondylosis with radiculopathy and/or myelopathy, not responding to 6-month conservative treatments. Exclusion criteria included infection, tumour, osteoporosis, metabolic disease, ossification of the posterior longitudinal ligament (OPLL), disc height <4 mm, ligament injury or segmental instability, and kyphotic deformity. Preoperatively, all patients underwent dynamic cervical radiography, magnetic resonance imaging (MRI), and clinical evaluation using visual analogue scale (VAS). Preserved ROM was documented preoperatively and it was calculated as the difference between the angles obtained in the flexion and extension X-rays, as reported by Lee et al. [24]. The study was performed in accordance with the Helsinki Declaration and a prior informed consent was obtained from all the patients. postoperative evaluation, the clinical outcome was improved with a mean decrease of the VAS Score from 8.20 ± 0.84 to 2.20 ± 1.30 ( Table  2). Heterotopic ossification was noted in 5 patients (7.69%): two patients (3.08%) at the 3-year follow-up and 3 patients (4.61%) at the 5-year follow-up. Among them, only 1 patient (1.54%) was symptomatic at the end of the follow-up and his major complaint was neck pain. Four of those cases (6.15%) were Grade III of the McAfee's Classification with a residual maintenance of motion while 1 patient (1.54%) was classified as Grade IV with a complete list of motion. The involved levels were all C5-C6 (40%) and C6-C7 (60%) ( Table 3) where the operated levels were mostly C5-C6 and C6-C7 (49.23 and 38.09%, respectively). The occurrence of HO was not significantly related to the concerned risk factors.

Discussion
The efficacy of CTDR has been widely reported over the past 10 years with a better protection of adjacent segments than ACDF. This feature is still debated, however a recent prospective study concluded that maintaining motion rather than fusion would prevent ASD [30,31]. Indeed, the preservation of motion constitutes the main goal of CTDR along with the maintenance of cervical lordosis, which some authors considered even more relevant to prevent the ASD after CTDR surgery [32,33]. Although most of the commonly used prostheses do not obtain the restoration of the cervical lordosis, the relatively new Discover discs have been described as MRI-compatible ball-and-socket devices, allowing all the directions and restoring the alignment in the cervical region [27,34,35]. Moreover, in patients who underwent CTDR it has been analysed the correlation between preoperative and postoperative ROM considering a poor preoperative ROM as a relative contraindication [36]. Instead, a preoperative disc height over 4 mm and the resection of posterior longitudinal ligament have been reported as favouring factors for a better ROM after CTDR [37]. Indeed, in the present study all the enrolled patients showed a preserved preoperative ROM and a disc height higher than 4 mm.
Several studies analysed the incidence of HO and the related risk factors as one of the most important causes of limited motion after arthroplasties [38]. This phenomenon has been described as an ectopic bone formation around the implant, subsequent to the transformation of primitive cells into osteogenic cells or bone morphogenic protein [39,40]. It was first reported in 2005 when Parkinson published a case of fusion after Bryan disc implant at the C5-C6 level [41]. Nowadays, HO after CTDR is commonly described in the literature and its incidence

Surgical technique
In all cases, a right anterior cervical approach was performed as described by Smith and Robinson and a Discover artificial disc was implanted after a proper discectomy and removal of posterior longitudinal ligament and anterior osteophytes [28]. A lateral and anterior-posterior fluoroscopy was used intraoperatively to confirm the proper device positioning. An abundant irrigation with saline solution was performed during the entire procedure until the end of the wound closure. If not contraindicated, non-steroid anti-inflammatory drugs (NSAIDs) were routinely prescribed in the first two postoperative weeks. Each patient was precociously mobilised without neck collar and dismissed the first day after surgery. After a first evaluation at 30 days, all the patients were followed with annual clinical assessments and dynamic radiographies (Figure 1). Where HO was described, the following associations were analysed: age, sex, treated pathology, and number of levels. Moreover, HO was described using the McAfee's Classification (Table 1) and the flexion and extension ROM was measured after its appearance [29].

Statistical analysis
All statistical analyses were performed using SPSS software 17.0 (SPSS, Inc., Chicago, IL, USA). Radiological and clinical measurements were expressed as mean ± standard deviation (SD). A Student's t test was used to assess the statistical significance of postoperative VAS and overall ROM changes and to analyse the associations between the occurrence of HO and the influencing factors, including age, sex, number of levels, and treated pathology. P<0.05 was considered as statistically significant.

Results
A total of 65 patients were enrolled in our study (26 men and 39 women). The mean patient age was 43.74 ± 10.94 years (range: 28-67 years) and the mean follow-up was 63.29 ± 11.90 months (range: 14-83 months). Among the 65 implanted patients, 61 (93.84%) were monosegments and 4 (6.15%) were bi-segments. No prosthesis displacement and no revision surgery were observed in all the population. At the first    varies widely. Significantly higher rates were noted in male sex, older age, multi-level surgery, presence of bone dust, stress between endplate and prosthesis [19,24,39]. Besides, the postoperative NSAIDs use has been considered to reduce HO formation and fusion rate. For instance, a study published by Heller et al. showed no spontaneous fusion at 2 years with the prescription of NSAIDs for 2 weeks postoperatively [42]. Consequently, a 2-week postoperative NSAIDs treatment was routinely used as prophylaxis.
Moreover, as the presence of bone dust has been widely considered as favouring factors of HO, the removal of anterior osteophytes and the irrigation of the surgical field with abundant saline solution before the insertion of the device were considered mandatory in our series [39]. Furthermore, as a postoperative limitation of motion is considered to affect the occurrence of HO, we decided to early mobilise the patients without neck collar [19]. Thus, we added to our surgical technique all those practical details, which were considered as factors lessening the risk of HO appearance.
In addition, it has been reported the important role of prosthesis type on the occurrence of HO with a greater incidence in PCM discs (80%) and a lower rate in Bryan implants (49%), with also a lower McAfee's Grade in this latter type [19,43]. A recent paper on ProDisc-C found a 37% rate of HO, easily comparable to previously published studies on the same prosthesis [11,44].
Most of the previous studies documented a progressive increase of the incidence and grade of HO as the time passes. The possibility of a plateau after the fourth year of follow-up has been hypothesised because the increase of the bone formation would decrease the ROM as far as the complete fusion is obtained [11,44,45]. In our study, all the patients showed the appearance of HO after the third year of follow-up and all of them were high McAfee's grade. Therefore, our rate of HO development was significantly lower than in previous reports, but more severe, although most of the enrolled patients had excellent clinical outcomes. Only 1 patient showed a high-grade HO (Grade IV) with a complete list of motion and in this case the functional result was the same as an ACDF. Similarly, Malham et al. [11] in his long-term followup study observed a late presentation of HO, with no case identified at 2 years. Indeed, as described by Tu et al. [46] a CT scan could be crucial for the diagnosis of lower grade of HO and to better clarify the role of HO in the natural history of the CTDR.
In conclusion, our paper showed a lower rate of HO with a higher follow-up in respect with previous studies confirming that the use of a well-selected CTDR in the right patient, with a good positioning of the prostheses and a precise surgical technique, is still a safe, satisfactory, and effective procedure.
However, a prospective cohort study with at least a 10-year follow-up using multi-planar CT scans could overcome some of the shortcomings of this study. Furthermore, more examiners with a calculated interobserver agreement could be useful to obtain more reliable results.