- What complications can occur, and which are most frequent?
- How do they present?
- How are they resolved?
- What factors contribute to these complications?
- When should reoperation happen?
Had spine surgery and the pain returned? Send your studies for a second opinion.
Send my studiesCervical disc arthroplasty (CDA) has an overall safety profile comparable to ACDF, but because it uses a mobile, articulating implant instead of a static fusion, it introduces its own distinct set of possible complications:
| Complication | Approx. incidence | Severity |
|---|---|---|
| Heterotopic ossification (any grade) | Up to 40–60% at long-term follow-up | Usually asymptomatic; gradual motion loss |
| Heterotopic ossification causing significant motion loss | ~10–20% | Functional, rarely symptomatic |
| Device subsidence or migration | Uncommon | Variable |
| Core herniation/extrusion | Rare, case-report level | Can cause acute neurological symptoms |
| Catastrophic mechanical device failure | Rare, case-report level | Can cause progressive myelopathy |
| Late periprosthetic infection | Rare | Serious, requires urgent evaluation |
| Adjacent segment disease (10-yr symptomatic) | Lower than ACDF in most comparative trials | May require additional treatment |
By far the most common finding is heterotopic ossification (HO) — new bone forming around the implant, gradually restricting the motion the device was implanted to preserve. Most HO is a radiographic finding without a matching symptom; the rarer but more consequential complications are mechanical device problems, which can range from silent to genuinely dangerous. (Hui et al., 2020; Jin et al., 2013)
A systematic review and meta-analysis pooling over 3,160 patients across 14 randomized controlled trials found comparable overall complication rates between arthroplasty and ACDF, with arthroplasty showing a statistically lower rate of secondary surgical procedures at the index and adjacent levels combined. (Findlay et al., 2018) A separate meta-analysis specifically addressing heterotopic ossification found progression to high-grade HO (Grade III–IV, substantially restricting motion) in roughly 20–30% of segments by 5–10 years, though clinical impact at that grade remains limited in most patients. (Hui et al., 2020)
A large administrative database comparison (Michigan Spine Surgery Improvement Collaborative) found reoperation rates within 2 years of approximately 3–4% after CDA compared to roughly 6–8% after ACDF, consistent with the theoretical advantage of motion preservation in reducing adjacent-level stress — though patient selection differences between groups make direct comparison imperfect. (Miller et al., 2024)
What is heterotopic ossification in a disc arthroplasty?
Almost always identified on imaging rather than by symptoms — a gradual, painless loss of motion at the operated level over months to years. In advanced grades, the segment effectively auto-fuses. Patients rarely notice this directly; it is typically found on routine follow-up X-rays. (Jin et al., 2013)
Can the core of a disc prosthesis fail?
Presents acutely, often with new or recurrent radicular pain or, if the extruded material compresses the cord, myelopathic symptoms. This is a mechanical event, not a gradual degenerative process, so the symptom onset is typically sudden. (Brenke et al., 2015)
How common is mechanical failure of a disc prosthesis?
This is the complication that deserves the most respect. Published cases describe device failure presenting as slowly progressive cervical myelopathy that worsens over months, sometimes without dramatic new pain — the patient simply notices gradually declining hand function, gait changes, or balance problems. In one illustrative case, an M6-C prosthesis failure caused spinal cord compression requiring urgent intervention. (Carrera & Ricks, 2022; Parish et al., 2025) Late implant failure has also been reported presenting with a mix of radiculopathy and myelopathy years after the original surgery. (Pingel et al., 2022)
Can an infection appear years after arthroplasty?
Neck pain, sometimes fever, and elevated inflammatory markers — occasionally presenting long after the original surgery rather than in the immediate postoperative period. (Xia & Winder, 2019)
- Asymptomatic heterotopic ossification: no treatment needed. Even advanced-grade HO that eliminates motion is usually left alone if the patient is asymptomatic — the segment simply behaves like a fusion at that point.
- Core herniation with new neurological symptoms: surgical revision — removal of the extruded material and, in most cases, conversion of that level to fusion.
- Mechanical device failure with progressive myelopathy: urgent surgical revision. Depending on how the device has failed and how it has integrated with the vertebral bodies, this can range from a straightforward implant exchange to a full corpectomy with reconstruction. (Parish et al., 2025)
- Late infection: targeted antibiotics based on culture, with implant removal considered if the infection does not clear or if the device is loose.
- Pre-existing facet joint arthropathy: associated with higher rates of heterotopic ossification, since the same biological process that drives facet degeneration appears to promote bone formation around the implant. (Hui et al., 2020)
- Sagittal alignment and implant positioning: device design and implant placement influence how motion and load are distributed across the segment, which can affect both wear patterns and HO risk. (Patwardhan & Havey, 2020)
- Younger, more active patients: a longer expected lifespan of use means more accumulated motion cycles on the device, which is a consideration for long-term wear, though modern designs have substantially improved durability compared to early-generation implants. (Anderson et al., 2003)
- Off-label or expanded indications: multi-level arthroplasty and use outside FDA-labeled indications carry a different, less well-characterized risk profile than single-level, on-label use. (Turel et al., 2017; FDA IFU, M6-C)
- Improper patient selection: significant pre-existing instability, kyphotic deformity, or advanced facet disease are relative contraindications; using arthroplasty in these settings increases mechanical complication risk.
- Any new or progressive myelopathy suspected to relate to device malposition, migration, or mechanical failure — this is treated as urgent, not elective.
- Core herniation or extrusion causing new radicular or myelopathic symptoms.
- Symptomatic device subsidence with axial pain or new neurological findings.
- Infection not controlled with antibiotics alone, or with a loose/failing implant.
- Notably, heterotopic ossification — even advanced-grade — is not by itself an indication for reoperation if the patient is asymptomatic; the loss of motion alone does not justify the risks of revision surgery.
A device that has "auto-fused" due to heterotopic ossification is not a failed surgery — the patient still generally retains the pain relief and decompression benefit of the original operation. It is a mechanical failure of the specific motion-preservation goal, not a clinical failure, and by itself rarely warrants revision.
Everything that makes revision ACDF complex applies here too — scarred tissue planes, a recurrent laryngeal nerve embedded in fibrosis, and distorted anatomy. But arthroplasty revision adds a layer that static cage revision does not: the implant itself may be osseointegrated into the vertebral endplates. Depending on the device design, bone can grow directly onto or into the prosthesis over time, meaning that extracting it is not simply unscrewing hardware — it can require controlled removal of adjacent bone as well.
This is precisely why the most severe reported cases of arthroplasty device failure have required conversion not just to a simple disc-level fusion, but to a full corpectomy — removing the entire vertebral body adjacent to the failed device to safely access and reconstruct the segment. (Parish et al., 2025) The surgeon has to weigh the risk of aggressive device extraction against leaving a failing implant partially in place, a calculation that does not exist in cage revision.
In addition to the same recurrent laryngeal nerve and esophageal risks that apply to any revision anterior cervical surgery, arthroplasty revision carries a specific additional risk: vertebral body bone loss during device extraction. If the implant has integrated into the endplates, removing it can compromise enough bone stock that a straightforward disc-level revision is no longer possible, forcing the more extensive corpectomy-and-reconstruction approach described above.
There is also a real, if uncommon, risk of acute neurological worsening during device extraction itself, if the cord or nerve roots are in close proximity to fragments of a broken or migrated component. This is part of why device-failure revisions are approached as urgent, carefully planned procedures rather than routine outpatient surgery.
For core herniation and straightforward mechanical failures converted to a disc-level fusion, prognosis is generally good — comparable to a well-indicated primary ACDF, once the compressive lesion has been addressed. Reported single- and multi-level device failure cases converted to fusion have generally shown clinical improvement postoperatively.
For the more severe cases requiring corpectomy — where progressive myelopathy from device failure has already caused measurable neurological deficit before surgery — the reported outcome is that decompression and reconstruction do reverse or halt the progression of myelopathy, but the degree of neurological recovery depends heavily on how long the compression was present before it was addressed. (Parish et al., 2025) This is the central argument for taking any new or worsening neurological symptom after cervical arthroplasty seriously and promptly, rather than waiting to see if it resolves on its own.
Most cervical arthroplasty complications — heterotopic ossification chief among them — require no intervention at all. The minority that do require reoperation, particularly mechanical device failure with neurological compromise, respond best to prompt recognition: the earlier a failing device is identified, the more straightforward — and the better-prognosis — the revision tends to be.
It is worth stating plainly: interbody cages used in ACDF remain the most extensively studied cervical implant, with decades of data across thousands of patients. Disc arthroplasty is a well-validated, effective technology for the right candidate — but for multilevel disease, significant instability, advanced facet arthropathy, or kyphotic alignment, fusion with a cage remains the gold-standard, better-characterized option. The choice between the two is not "newer versus older" — it is a question of matching the right technology to the right anatomy.
On reoperation rates: large-database comparisons have shown somewhat lower 2-year reoperation rates after arthroplasty than after ACDF in well-selected, single-level candidates (roughly 3–4% versus 6–8%, per Miller et al., 2024), and several long-term randomized trials support lower rates of adjacent segment surgery with arthroplasty. But this comparison only holds within the population that was actually eligible for arthroplasty in the first place — it is not evidence that arthroplasty is broadly "safer" than ACDF across all patients, since the two are not typically offered to the same patients.
If imaging or symptoms suggest the device has failed, subsided, or migrated, the critical question is whether it is compressing the spinal cord — the complication most associated with progressive myelopathy and the most time-sensitive risk in arthroplasty failure. This requires prompt, specific imaging (CT to assess bone and implant position, MRI when cord compression is a concern) to characterize exactly what has happened before any decision about revision is made. Unlike an anteriorly migrating cage, a failing arthroplasty device is more likely to threaten the cord than the esophagus, given its position and the mechanics of how these devices typically fail — but both directions of concern should be assessed explicitly, not assumed.
Reintervention for a failing arthroplasty device has to be approached with at least as much judiciousness as ACDF revision — arguably more, given the possibility of an osseointegrated implant discussed earlier. The surgeon performing the revision should have specific, demonstrable experience in anterior cervical revision surgery, ideally including experience with arthroplasty device removal specifically, since this is a narrower and less common skill set than ACDF revision alone. Vital structures — the recurrent laryngeal nerve, esophagus, carotid sheath, and spinal cord itself — are all in immediate proximity, and injury risk to any of them is meaningfully higher than in a first-time operation.
- Does the surgeon have specific experience in anterior cervical revision surgery — and, ideally, arthroplasty device removal specifically?
- Is there a clear, specific surgical plan — including how much vertebral bone may be lost during device extraction, and what the reconstruction will look like?
- Is there a rescue plan if the device cannot be safely removed — including the possibility of conversion to corpectomy, considered in advance rather than improvised mid-surgery?
- Has the surgeon evaluated specifically whether the failure is threatening the spinal cord, and how urgently that needs to be addressed?
