Intracranial erosion of an occipitocervical rod through the foramen magnum following occipitocervical fusion for retro-odontoid pannus: illustrative case

Abstract

BACKGROUND

Retro-odontoid pannus can cause craniocervical instability and cervicomedullary compression. Occipitocervical fusion (OCF) is an established treatment, but hardware-related complications remain significant. Intracranial rod migration through the foramen magnum is exceedingly rare and poses substantial neurological risk.

OBSERVATIONS

A 76-year-old man with retro-odontoid pannus and progressive myelopathy underwent OCF. Months later, he developed wound complications and mechanical failure. Imaging revealed superior rod migration with erosion through the foramen magnum and intracranial extension. A staged surgical strategy was utilized: initial hardware removal and soft tissue reconstruction, followed by targeted left suboccipital craniectomy, durotomy, and ultrasound-guided extraction of the migrated rod. Revision C1 laminectomy and allograft C1–2 arthrodesis restored construct stability. The patient recovered without new neurological deficits and experienced improvement in gait imbalance.

LESSONS

Intracranial rod erosion after OCF is rare but should be suspected in the presence of wound breakdown, new posterior fossa symptoms, or construct instability. Prevention requires meticulous rod contouring to avoid occipital contact, secure cranial fixation, and early postoperative surveillance. Management benefits from multidisciplinary collaboration and intraoperative ultrasound for safe extraction in anatomically constrained regions.

https://thejns.org/doi/10.3171/CASE25837

Craniocervical instability from retro-odontoid pannus (periodontoid pseudotumor) can cause cervicomedullary compression, myelopathy, and disabling occipital pain. Historically associated with rheumatoid arthritis, pannus formation is also seen in noninflammatory/degenerative settings, for example, in cases of chronic atlantoaxial instability, where micromotion between the joint structures is thought to lead to the pannus formation from continuous inflammation over time. On review of the literature, only 7 prior cases have been reported, with the present case being the eighth (Table 1). In these cases, posterior fusion alone may eliminate that micromotion and, alongside posterior decompression, may obviate the need for any ventral operation. Various cases document radiographic pannus regression after posterior stabilization, thus supporting fusion as a primary strategy in appropriately selected patients.^1,–4

TABLE 1.

Review of the literature of posterior fossa rod migration

Authors & Year	Age (yrs)	Initial Procedure	Time to Migration	Possible Cause	Condition After Migration	Condition After Surgical Correction
Kiran et al., 20168	55	Laminectomy & lat mass fixation	1.5 yrs	—	Loss of consciousness & quadriplegia	Death
Plant & Ruff, 201012	13	C1–2 fusion	3 yrs	Trauma	Neck pain	Improved
Yablon et al., 199313	18	C4–5 Harrington rod hook fusion	4 yrs	—	Neck pain & dizziness	Improved
Chun et al., 201014	23	C1–2 fusion	20 mos	—	Occipital headache & dizziness	Refused op
Mahtabfar et al., 202115	70	C1–2 fusion	15 yrs	Rheumatoid	Headache, dizziness, nausea, & vomiting	Stationary course
Ezzat & Salah, 20229	67	C3–6 fusion	2 yrs	Trauma	Headache, facial palsy, & ophthalmoplegia	Improvement in headache only
Basankin et al., 202310	25	C1–4 fusion	4 yrs	—	Neck pain & staggering gait	Improved
Present case	76	Occiput–C6 fusion	11 mos	Infection	Imbalance & gait difficulties	Improved

Occipitocervical fusion (OCF) is effective for craniocervical instability but carries nontrivial complication rates. Contemporary series and reviews describe overall complication rates with a predominance of implant-related problems (loosening, failure, and migration) and wound complications. Reported implant-related complications may approach 32% in some cohorts, with wound problems as high as 23%, and reoperations including hardware removal. Pediatric and mixed series similarly cite early hardware failure modes (occipital screw backout and cervical screw failure) before arthrodesis.^5,6,7,8,9

Intracranial/foramen magnum rod migration is rare but repeatedly documented. Published cases describe cephalad rod displacement breaching the foramen magnum into the posterior fossa/cerebellum months to years after posterior cervical or occipitocervical constructs, sometimes presenting with acute giddiness or loss of consciousness and requiring rod removal via suboccipital exposure. Recent case reports and literature reviews (2016–2024) underscore its rarity, proposed mechanisms (construct mechanics, bone erosion, rod contour/position, host bone quality, and destructive spondyloarthropathy), and preventive strategies (rod contouring, avoidance of contact points, and robust cranial fixation).^5,8,–10

Illustrative Case

Demographics and History

A 76-year-old male with extensive prior spine surgery presented with roughly 1 year of occipital headaches, gait imbalance with falls, dizziness with extension, and Lhermitte-like symptoms. The patient had previously undergone a C5–7 anterior cervical discectomy and fusion (ACDF) and presented 5 years ago with adjacent segment severe cervical spinal canal stenosis at C4–5 and C7–T1 (Fig. 1) He subsequently underwent a C2–T2 posterior cervical decompression and fusion (PCDF) (Fig. 2). After an uneventful postoperative course, the patient presented again to the clinic many years later with examination remarkable for hyperreflexia and a positive Hoffmann sign. MRI of the brain and cervical spine demonstrated a large retro-odontoid pannus (Fig. 3A) and craniocervical instability with severe stenosis in the supine position. He underwent OCF with an occipital plate and bilateral C1 lateral mass screws with extension to his previous construct down to C6 arthrodesis (Fig. 3B and C). As the compressive pannus was primarily at the level of C1, a wide laminectomy and superior and inferior ligamentous disruption were required for adequate decompression, thus necessitating fusion to the occiput instead of terminating at C1. The early postoperative period included wound healing issues that improved with conservative care; later imaging showed persistent pannus but a decompressed cord. Ultimately, the patient developed superior hardware failure with occipitocervical rod erosion through the occipital bone into the posterior fossa (Fig. 4A–C).

FIG. 1.

Images obtained prior to PCDF. A:Lateral radiograph of the cervical spine with the previous C5–7 ACDF construct visualized. B: Sagittal T2-weighted MR image of the cervical spine, demonstrating multilevel degenerative changes and significant spinal canal stenosis at C4–5 and C7–T1. C:Sagittal CT scan of the cervical spine, demonstrating the previous C5–7 ACDF construct and multilevel osseous degenerative changes.

FIG. 2.

Plain lateral (A) and anteroposterior (AP) (B) radiographs of the C2–T2 construct without evidence of hardware complications.

FIG. 3.

A:Sagittal T2-weighted MR image of the cervical spine, demonstrating a large C1–2 pannus with severe central canal stenosis and resultant cord signal change. B and C:Plain lateral (B) and AP (C) radiographs of the occiput–T2 construct without evidence of hardware complication.

FIG. 4.

Preoperative sagittal (A), axial (B), and coronal (C) CT scans demonstrating posterior fossa migration of the rod. Postoperative sagittal (D), axial (E), and coronal (F) CT scans demonstrating sequelae of posterior fossa rod removal.

During his admission, he first underwent removal of occipitocervical-thoracic hardware and cultures with plastic surgery wound closure. Intraoperatively, a small bony opening at the left skull base consistent with rod erosion was seen, but the intracranially migrated rod was not safely retrievable at that setting. He then returned during the same admission for a left suboccipital craniectomy and limited durotomy. With the aid of microscope magnification, intraoperative ultrasound guidance, and careful dissection, the intracranially eroded rod was extracted uneventfully. A C1 laminectomy and C1–2 posterior lateral allograft arthrodesis with cadaveric graft were completed (Fig. 4D–F). The patient was discharged home on postoperative day 8 and later improved in gait and imbalance. His headaches persisted but were managed medically.

Anesthetic Management

Patient and Preoperative Assessment

A 76-year-old male (weight 107.8 kg, height 180.3 cm, BMI 33 kg/m²) presented with a past medical history that included hypertension, hyperlipidemia, atrial fibrillation, and dyspnea on exertion thought to be secondary to tachycardic episodes managed with diltiazem. He also had mixed anemia and well-controlled gastroesophageal reflux disease. No medication allergies were reported. Baseline outpatient vital signs included a blood pressure of 138/73 mm Hg, heart rate of 69 beats per minute, and oxygen saturation of 97% on room air, with an otherwise normal review of systems.

The airway assessment revealed limited neck range of motion but adequate mouth opening and thyromental distance. The Mallampati classification was class IV. As described above, concern existed for bony erosion of the hardware into the skull, raising the risk that even small movements could worsen the pathology.

Airway Strategy and Induction

Given the mixed airway predictors and risk of hardware destabilization, an awake fiber-optic intubation under topical anesthesia and mild sedation was planned. This approach was selected given the patient’s limited neck mobility and Mallampati class IV airway, despite multiple prior anesthetic records demonstrating atraumatic first-attempt endotracheal intubation with video laryngoscopy and adequate mask ventilation. Video laryngoscopy was therefore designated as the backup plan should awake intubation be poorly tolerated.

Standard American Society of Anesthesiology (ASA) monitors (electrocardiogram, noninvasive blood pressure, pulse oximetry, capnography, and temperature monitoring) were applied prior to preoxygenation. The airway was anesthetized with nebulized 4% lidocaine and lidocaine jelly; no additional nerve blocks were performed. Sedation was achieved with 3 mg of midazolam and 50 μg of dexmedetomidine. The airway was secured with a flexible fiber-optic bronchoscope via the orotracheal route on the first attempt while maintaining in-line cervical stabilization.

After confirmation of tube placement, general anesthesia was induced with 100 mg of propofol and 50 mg of rocuronium. A postinduction arterial line was placed for invasive blood pressure monitoring. Large-bore intravenous access was obtained, and blood products were cross-matched in anticipation of possible blood loss.

Intraoperative Management

The patient was positioned prone with his head secured in a Mayfield head clamp, maintaining neutral craniocervical alignment throughout positioning and during transition from supine to prone, verified jointly by the anesthesia and surgical teams. Anesthesia was maintained with sevoflurane (end-tidal concentration 1.0%–2.1%). Phenylephrine (20–80 μg/min) was infused intermittently to maintain a mean arterial pressure at 15% above baseline, thereby optimizing spinal cord perfusion. Arterial blood pressure ranged from 85/54 mm Hg to 161/96 mm Hg, with the lowest recorded blood pressure immediately postinduction, as expected. Heart rate ranged from 73 to 123 beats per minute, and peripheral oxygen saturation ranged from 94% to 100%. Ventilation was adjusted to maintain normocapnia and normoxia. Perioperative antibiotic prophylaxis was administered per institutional protocol. Analgesia consisted of 100 μg of fentanyl and 2 mg of hydromorphone. Postoperative nausea and vomiting prophylaxis included 4 mg of ondansetron and 10 mg of dexamethasone. Estimated blood loss during the first operation was approximately 150 mL, replaced with 2000 mL crystalloid; no transfusion was required. At the conclusion of surgery, the patient was returned to the supine position and extubated awake with intact airway reflexes after confirmation from the neurosurgical team. He was transported to the postanesthesia care unit (PACU) and then to the neurosurgical ward. No anesthetic complications occurred.

During a subsequent suboccipital craniectomy and rod extraction, the same airway plan was utilized—awake fiber-optic intubation followed by general anesthesia. Additionally, the patient’s secretion burden was noted to be worse than with the first operation; 0.2 mg of glycopyrrolate was administered with sedatives to minimize secretions. Alternately, for this operation anesthesia was maintained with isoflurane (end-tidal concentration 0.5%–1.0%). Hemodynamic parameters were supported similarly. His arterial pressure ranged from 61/48 mm Hg (noted secondary to a transducer malposition during prone positioning) to 130/81 mm Hg, with a heart rate ranging between 61 and 143 beats per minute. He remained adequately oxygenated with 99%–100% pulse oximetry. Blood loss was minimal (50 mL), and no hemodynamic perturbations were noted at the time of rod removal. The patient was again extubated in the operating room and transported to the PACU in stable condition.

Surgical Management

In the first part of this staged operation, after commencing a routine posterior cervical approach, we ultimately encountered the prior posterior cervicothoracic hardware at which point all locking caps, bilateral rods, hinged rods to the occipital plate, and occipital plate screws were removed. Bilateral C1 screw heads were fractured and removed, leaving remnants within the bone. Granulation and inflammatory tissue around the craniovertebral junction were cultured and sent for aerobic, anaerobic, and fungal sampling. With the microscope in the field, we were able to visualize a small healed bony channel at the left foramen magnum consistent with chronic rod erosion. Given the uncertainty of the intracranial trajectory, definitive rod extraction was deferred pending imaging; the plastic surgery team performed complex closure with drains placed at the end.

In the second part of this staged operation, a targeted left suboccipital craniectomy of about 1 cm in size was fashioned over the erosive foramen; a dural opening was made. Intraoperative ultrasound was used to localize the rod, and the segment was carefully delivered using dental instruments and Kerrison rongeurs. Hemostasis was achieved, and a small piece of nonsuturable dural substitute was placed intradurally/extradurally over the opening. A C1 laminectomy was then performed, and decortication/arthrodesis at C1–2 with a cadaveric graft was completed. Layered closure in typical fashion followed to complete the case. Of note, intraoperative ultrasound was used to confirm adequate decompression; meticulous hemostasis and copious irrigation were used throughout both procedures.

The chosen approach mirrors prior reports advocating suboccipital exposure for rods that breach the foramen magnum and the posterior cranial fossa, sometimes with durotomy when intradural. Ultrasound guidance facilitates localization, while attention to small working corridors at the craniovertebral junction and avoidance of neural/vascular injury are critical.^9,11

The patient recovered without new neurological deficits. Wound healing improved under combined neurosurgery–plastic surgery management. The intraoperative swabs obtained grew Cutibacterium acnes requiring intravenous penicillin G. By early follow-up, the patient’s imbalance improved substantially; collar wean was initiated per stability and symptom trajectory.

Informed Consent

The necessary informed consent was obtained in this study.

Discussion

Observations

This case highlights a rare but serious complication of OCF: intracranial erosion and migration of a fixation rod through the foramen magnum months after the initial surgery. While OCF is an established strategy for craniocervical instability and is effective in managing retro-odontoid pannus through indirect decompression and mechanical stabilization, hardware-related complications remain clinically significant. The literature reports overall OCF complication rates up to 30%, with implant failures representing a major subset. However, intracranial rod migration specifically is exceedingly rare, with only a handful of documented cases worldwide.

Prior reports describe similar delayed presentations involving cephalad rod displacement into the posterior cranial fossa or cerebellum, sometimes precipitating acute neurological decline, syncope, or hemorrhage. This case mirrors these reports in demonstrating delayed progressive bony erosion at the foramen magnum, ultimately permitting rod violation of the skull base. Proposed mechanisms include inappropriate rod contouring causing occipital contact, thin occipital bone stock, infection-mediated osteolysis, and loss of distal construct constraints. The chronic nature of the bony channel in our case suggests long-standing mechanical erosion rather than acute displacement.

This case further reinforces previously reported associations between wound complications, occult low-grade infection (e.g., C. acnes), and eventual hardware failure. The presence of persistent wound issues and construct loosening should raise suspicion for deeper pathology, as these signs may precede catastrophic migration events. We suspect the mechanism of rod migration in our case to be a combination of progressive mechanical imbalance combined with set screw failure. Namely, we believe the continued stress from the mechanical imbalance likely contributed to loosening of an already faulty or improperly tightened set screw. Chronic wound healing issues could have led to a low-grade latent infection, which may have compounded to hardware failure.

Lessons

The most important lesson from this case is that intracranial rod migration is preventable with meticulous construct planning and vigilant postoperative surveillance. Several technical and clinical principles emerge.

Prevention begins at implantation. Rods should be contoured to avoid direct contact with the occipital squama or foramen magnum. Cranial fixation should be placed into regions of adequate bone thickness, and locking mechanisms must be reinforced to prevent gradual rod advancement. Soft tissue integrity matters. Compromised soft tissue coverage over the occiput increases construct vulnerability. Early collaboration with the plastic surgery team may prevent exposure and infection. Red flags demand action. Persistent wound drainage, posterior fossa symptoms (headache, vertigo, and cranial neuropathies), or new mechanical pain should prompt urgent imaging to detect rod migration before intracranial breach. Multistage surgical strategy may be required. In anatomically constrained revision cases, staged management allows controlled removal of infected or unstable hardware while minimizing neurological risk. Intraoperative ultrasound is particularly valuable for navigating around the brainstem and venous sinuses during rod extraction.

The limitations of this report include its single-patient design and inherent lack of generalizability. Nonetheless, given the rarity of this complication, each case adds material insight into mechanisms, early detection, and operative management strategies. The delayed nature of this complication emphasizes the need for long-term follow-up after OCF, especially in patients with early wound issues or serological evidence of inflammation.^5,8,9

In summary, this case demonstrates that intracranial rod erosion after OCF is rare but clinically significant. Prevention hinges on thoughtful biomechanical planning and early identification of red flags. When migration occurs, multidisciplinary revision, including neurosurgery, plastic surgery, and intraoperative imaging techniques, can achieve safe extraction and stabilization with favorable outcomes.^7,–9,11

Intracranial erosion/migration of an occipitocervical rod through the foramen magnum is rare but clinically significant. In a 76-year-old man treated for retro-odontoid pannus and craniocervical instability, delayed rod erosion required staged hardware removal and suboccipital rod extraction with favorable outcome. Vigilance for wound and mechanical red flags, attention to cranial fixation and rod contouring, and early imaging when symptoms evolve may mitigate risk and enable timely intervention.^7,8

Disclosures

The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper.

Author Contributions

Conception and design: Gunasekaran, Saway, Sirianni, Morgan, Faulkner. Acquisition of data: Gunasekaran, Horne, Sheedy, Saway, Morgan, Faulkner. Analysis and interpretation of data: Gunasekaran, Saway, Morgan. Drafting the article: Gunasekaran, Saway, Morgan, Faulkner. Critically revising the article: Gunasekaran, Saway, Morgan, Faulkner. Reviewed submitted version of manuscript: Gunasekaran, Horne, Saway, Sirianni, Faulkner. Approved the final version of the manuscript on behalf of all authors: Gunasekaran. Statistical analysis: Gunasekaran. Administrative/technical/material support: Gunasekaran, Sheedy, Sirianni. Study supervision: Gunasekaran, Morgan. Participated in patient treatment: Sirianni.

Correspondence

Arun Gunasekaran: Medical University of South Carolina, Charleston, SC. arg214@musc.edu.