Iliac end-to-end structural autograft for complex failure after occipital cervical fusion: illustrative case

Abstract

BACKGROUND

Occipital cervical fusion (OCF) is an effective method for immobilizing the cranial cervical junction (CCJ). OCF failure is uncommon but can be associated with kyphosis after laminectomy. Revision OCF after long-segment cervical laminectomy is challenging, as grafting options are more limited.

OBSERVATIONS

The patient was a 53-year-old man with pseudarthrosis, posterior displacement of the occipital plate, a chronic type II odontoid fracture with anterior subluxation of the fractured dens, a fracture of the C1 vertebrae with anterior displacement, and a previous 6- to 7-cm laminectomy defect from C1 to C5. Two iliac autografts, which were harvested from the left and right posterior superior iliac spine, were used. One end of the graft was placed into the occipital notch, and the other was placed over the C5 and C6 lamina. Two cross connectors were then placed over the combined graft, securing it in place. Three loops of fiber wire were tied to each main rod. The loops were then placed below the combined graft to prevent the rod from fracturing into the cervical cord in the event of repeated instrumentation failure.

LESSONS

Bilateral autologous iliac grafts can be combined to overcome a large cervical laminectomy defect during OCF revision surgery.

https://thejns.org/doi/10.3171/CASE25649

Rigid fixation techniques are commonly used for the implementation of occipital cervical fusion (OCF), which provides immediate immobilization of the cranial cervical junction (CCJ), with a high overall fusion rate.^1–6 The rate of pseudarthrosis after OCF is less than 16% and may be treated with external immobilization in most cases.^1,2 When external immobilization is insufficient, OCF revision surgery may be required. These cases usually involve more complex failures, such as pullout of the occipital plate, subluxation of the occiput, and possible upper cervical cord compression. Although such catastrophic OCF failures are rare, they tend to occur in the setting of postlaminectomy kyphosis, owing to the increased strain on the CCJ. Therefore, achieving fusion during OCF revision surgery can be especially challenging for a long-segment cervical laminectomy because traditional grafting options are limited.

Regarding graft type, structural autografts are associated with superior rates of fusion.⁷ Although more recent techniques and allograft materials have demonstrated that allografts are noninferior to structural autografts, these techniques rely on the placement of structural allografts under compressive forces, which can be difficult in the setting of a previous extensive laminectomy.⁸ Therefore, structural autografts are generally a more favorable option. However, there are no published reports that describe methods to span a multilevel laminectomy defect with a structural iliac autograft. In this case report, we describe a novel method of using bilateral iliac structural autografts to span a four-level laminectomy defect to achieve fusion during OCF revision surgery.

Illustrative Case

A 53-year-old male with an extensive history of multiple anterior and posterior cervical fusions presented to an outside hospital in September 2023 after a fall. He was found to have a type II odontoid fracture and C1 Jefferson fracture. The patient was homeless at the time and discharged against medical advice without receiving surgery. Later, he presented again to an outside hospital in February 2024 with worsening pain, myelopathy, and cervical kyphosis; subsequently, he underwent an occiput-to-C4 fusion with a C1–2 laminectomy. He presented 1 month postoperatively with pseudarthrosis of the occipital plate, worsening anterior subluxation of the dens, and wound dehiscence, necessitating a revision occiput-to-C4 fusion with removal and reinsertion of the occipital plate. Two months (May 2024) after this revision surgery, he again presented to the outside facility with similar pseudarthrosis of the occipital plate, anterior subluxation of the dens, and wound dehiscence. Once again, he was taken for a revision of his occipital fusion with removal and replacement of his occipital plate. One month after his second revision, he again presented to the outside facility with pseudarthrosis of the occipital plate and anterior subluxation of the dens. At this point, he was transferred to our institution for surgical evaluation and management.

On physical examination, the patient had bilateral 4/5 weakness in hand grip strength, bilateral loss of sensation in hands, and a bilateral Hoffmann reflex but was otherwise neurologically intact. The patient also complained of a progressive worsening of balance and difficulty with ambulation over the past month. His cervical wound was intact; however, there was a large area of swelling over the upper cervical region. Preoperative radiographic imaging (Fig. 1A and B) and CT imaging (Fig. 2) showed pseudarthrosis and posterior displacement of the occipital plate as well as a chronic type II odontoid fracture with anterior subluxation of the fractured dens. There was also a fracture of the C1 vertebra with anterior displacement. In addition, there was a previous 6- to 7-cm laminectomy defect from C1 to C5. MRI (Fig. 3) demonstrated severe stenosis at the craniocervical junction, where anterior subluxation of the C1 arch and odontoid process had occurred.

FIG. 1.

A: Preoperative lateral radiograph showing posterior displacement of the occipital plate, causing deformation of the overlying skin and pseudarthrosis within the CCJ. B: Preoperative anteroposterior radiograph showing disunion of the occipital plate from the occiput. C: Preoperative lateral flexion radiograph demonstrating successful fusion. D: Preoperative anteroposterior radiograph demonstrating successful fusion.

FIG. 2.

A: Sagittal CT image through the left facets, showing pseudarthrosis of the left C2 screw, pullout of the occipital plate, and displacement of the left C1 lateral mass relative to C2. B: Midline sagittal CT image demonstrating the C1–4 laminectomy defect. C: Sagittal CT image through the right facets.

FIG. 3.

A: Preoperative sagittal T2-weighted MR image of the cervical spine, showing severe spinal cord compression at the CCJ. B: Postoperative sagittal T2-weighted MR image showing resolution of spinal stenosis.

The patient could not tolerate preoperative cervical traction and was taken to surgery for occiput-to-C7 revision posterior spinal fusion. Intraoperatively, a halo frame was used to secure the head and allow for greater intraoperative control of head positioning. Pre- and postpositioning motor evoked potentials and somatosensory evoked potentials were obtained. The patient was positioned prone, and the entire occiput-to-pelvic region was prepped and draped. The previous incision was opened, and the previous instrumentation was removed. Using intraoperative navigation, we placed subaxial cervical pedicle screws at C5 and C7, and bilateral C3 lateral mass screws were also placed. Intraoperative navigation was also used to find occipital screw trajectories that did not incorporate the previous ghost tracts. After further reduction of the position of the occiput, main rods were placed to connect the occipital plate to the subaxial instrumentation. Given the previous issues of repeated instrumentation failure, structural iliac autograft was preferred for the purpose of bone grafting. However, the cervical laminectomy was too large to be spanned safely by a single harvested iliac autograft. Therefore, we harvested two iliac structural autografts from both the left and the right posterior superior iliac spine by using osteotomes in the standard fashion. The iliac grafts were then drilled into a rectangular shape, and a cranial plating system was used to fasten the grafts end-to-end (Fig. 4). A high-speed drill was used to create a notch in the occipital plate to receive the combined graft, and the C5 and C6 lamina were drilled flat as well. At this point, one end of the graft was placed into the occipital notch, and the other was placed over the C5 and C6 lamina. A total of 2.8 mL of bone morphogenic protein–2 (1.5 mg/mL), placed on four strips of a collagen sponge, was placed between the bone and the graft, as well as superficial to the graft. Additionally, a small amount of allograft was placed throughout the exposed osseous elements. Two cross connectors were placed over the combined graft to help secure it in place. Three loops of fiber wire were tied to each main rod and placed below the combined graft (Fig. 4B) to prevent the rod from fracturing into the cervical cord in the event of repeated instrumentation failure. The wound was then closed in the standard fashion.

FIG. 4.

A: An 8-cm iliac graft after unification with a cranial plating system. B: Graft spanning previous laminectomy defect after surgical fixation.

Postoperatively, the patient did well and reported improvements in hand paresthesia. Postoperative radiographic imaging (Fig. 1C and D) showed successful fusion, and MRI (Fig. 3) demonstrated resolution of the cervical stenosis. After surgery, the patient was instructed to wear an Aspen Vista CTO4 cervical collar at all times. However, the patient was noncompliant with wearing the rigid cervical collar, and we attempted placement of a halo device with vest. The patient was unable to tolerate the halo device and self-removed the pins within a day. The halo had to be removed the following day. He was discharged on postoperative day 29. He did not return for any follow-up appointments; however, he did report to the emergency department 7 months after the operation because he was experiencing headaches. The patient reported constant pain and stiffness in his cervical spine that worsened at night, but he was neurologically intact on examination and his incision was clean, dry, and intact. CT imaging (Fig. 5) at that time demonstrated osseous fusion through the combined iliac graft, as well as osseous fusion of the dens, C1 arch, and C2 vertebral body. There were no signs of instrumentation failure, and the incision was well healed.

FIG. 5.

A–C: Sagittal CT images of the CCJ obtained at 7 months of follow-up, through the left (A) and right (B) facets demonstrating fusion between C1 and the body of C2, and demonstrating sustained osseous fusion of the autologous graft with the occiput and the C5 and C6 lamina (C). D: Graphical illustration of the iliac graft, related instrumentation, and previous instrumentation in relation to the anatomy of the cervical spine.

Informed Consent

The necessary informed consent was obtained in this study.

Discussion

Observations

OCF is indicated for a wide variety of spinal pathologies and is associated with a 90%–95% fusion rate.^1,3,4 Surgeons most commonly use rigid fixation techniques (i.e., rigid screw-rod instrumentation) with occipital plating to perform OCF, which offers immediate rigid immobilization.^1,2,5,6

Currently, there are few studies that have examined the different patterns of failure after OCF, and there remains no consensus regarding optimal techniques for revising more complex OCF failures. In a series of 100 patients undergoing rigid OCF, Bhatia et al. reported 4 cases of instrumentation failure; however, they did not report any cases of occipital plate displacement.⁹ Similarly, Martinez-Del-Campo et al. reported a series of 120 OCF patients, yet no cases in the setting of complex failure required more than one revision surgery.³ Zileli and Akıntürk presented a case series of 128 patients who underwent OCF and reported an implant-related complication rate of 32%, but the authors did not specify subtypes of failures.¹⁰ In a report of 252 patients with instrumented OCF and long-term follow-up, He et al. showed a rod fracture rate of 0.4% rate and a screw pullout rate of 1.2% rate, but they did not provide further details about the number of revision surgeries or the need for revision surgery.¹¹ Additionally, OCF instrumentation failure can be precipitated by postoperative sagittal imbalance due to increased strain on the screws.¹² For the present case, the decision was made to not correct cervical kyphosis because doing so would require a midcervical osteotomy, which would not impact his healing at the occipital cervical junction. Instead, the plan was to treat his mechanical failure and wound issues, and have him return to the clinic after medical optimization for definitive kyphosis correction. In summary, although recent research suggests a higher rate of revision surgery after OCF than historically reported, few studies have reported specific types of mechanical failure after OCF.¹³ Specifically, pseudarthrosis involving displacement of the occipital plate with anterior dislocation of the occiput, atlas, and cervical cord compression in the setting of long-segment cervical laminectomy has not been widely reported, and no case reports describing the revision of such have been published. One limitation of this case report is the patient’s socioeconomic circumstances, which may have contributed to his instrumentation failure. If our patient had greater access to care, he potentially could have been spared from undergoing such an extensive procedure.

Long-term fusion is a key aspect of successful OCF revision surgery. Although there are case reports of treating cervical fractures with temporary fixation of the occiput to the cervical spine without fusion, fusion was required in this case because of the chronic nature of the dens fracture and the upper cervical cord compression.¹⁴ Moreover, despite recent reports suggesting that allografts are noninferior to structural autografts, allografts have been shown to be most effective when placed under a compressive load, which was difficult in this case, given the long-segment cervical laminectomy defect.⁸ Therefore, a structural corticocancellous autograft was ideally required because of its osteoinductive, osteoconductive, and osteogenic properties in neutral compression.¹⁵ Given that, the technical challenge in this case was spanning the 6- to 7-cm laminectomy defect with a structural autograft. Of note, rib structural autograft can be safely harvested up to a length of 15 cm.¹⁶ However, a method to expand the length of an iliac structural autograft represents an alternative to a rib structural autograft for spanning long-segment cervical laminectomy defects. In suitable adult patients, up to 16 cm of fibular autograft can be safely harvested.¹⁷ While nonvascular fibular strut grafts can be used to span spinal defects, harvesting the fibula could lead to morbidity in the donor site leg.¹⁸ Metal meshes represent a third alternative for achieving fusion in spinal defects, and they require less autograft to be harvested compared to iliac or fibular structural autografts. Consequently, metal meshes result in less donor site complications compared to structural autografts, but are more expensive and less suitable for spanning longer defects compared to structural autografts.¹⁹ Lastly, paramedian placement of structural autograft can be utilized to span cervical defects, but the technique is more commonly utilized in the pediatric population compared to the adult population.²⁰

Lessons

In this illustrative case report, we describe a method of using bilateral autologous iliac grafts to span a laminectomy defect of 7 cm or greater by fastening the grafts together end-to-end and placing them in a notched segment of the occiput and cervical lamina. This novel technique was used to treat previous failures after OCF and led to successful fusion at 7 months of follow-up. The end-to-end bilateral structural iliac autograft may be a useful tool for surgeons treating complicated failures after OCF.

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: Burke. Acquisition of data: Burke. Analysis and interpretation of data: Burke. Drafting the article: all authors. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Burke. Administrative/technical/material support: Burke. Study supervision: Burke.

Correspondence

John F. Burke: University of Oklahoma College of Medicine, Oklahoma City, OK. john-burke@ouhsc.edu.