Revision for cage migration after transforaminal/posterior lumbar interbody fusion: how to perform revision surgery?

Spinal instability due to degenerative, traumatic, infectious and neoplastic diseases may require fusion. Minimally invasive posterior and transforaminal lumbar interbody fusion (MI-PLIF/TLIF) have become established procedures for this purpose [1‐3]. However, symptomatic pseudarthrosis and cage migration/protrusion are among the most difficult complications for MI-PLIF/TLIF. This cage migration may cause direct compression of the nerve root and/or cauda equina, pseudoarthrosis, and instrument failure. Several reports have emphasized that cage positioning in the disc space is an important factor in cage migration [4, 5]. If the patient develops severe radicular symptoms due to cage migration, removal of the cage becomes necessary. However, such removal is sometimes challenging if epidural adhesion or fibrous union has developed between the cage and vertebrae. We describe herein a new classification of cage protrusion after MI-PLIF/TLIF and a novel technique utilizing a C-arm-free navigated osteotome and the oblique lumbar interbody fusion at L5/S1 (OLIF51) technique to address these complications.

This retrospective study was approved by the institutional review board of our hospital (approval no. 283), in accordance with the research guidelines for humans. The inclusion criteria for patients were: MI-PLIF/TLIF in our hospital between July 2017 and December 2020; and follow-up for > 1 year. Exclusion criteria were: severe osteoporosis; rheumatoid arthritis; or destructive spondyloarthropathy. A total of 113 consecutive patients were enrolled in this study. We use our new classification of the cage protrusion; type 1; only low back pain without new radicular symptom, type 2; low back pain with slight radicular symptom, type 3; cauda equine syndrome and/or severe radicular symptom (Fig. 1). For each patient, direct examination, postoperative lumbar radiograms and/or computed tomography (CT) were obtained at 1, 3, 6, 12, and 24 months of follow-up. Cage migration was defined as symptomatic posterior cage protrusion > 5 mm from the posterior border of the over and underlying vertebral body compared with initial CT. We evaluated patients’ characteristics such as body mass index, smoking history, fusion level, and cage type.

We applied a new classification for cage protrusion: type 1, only low back pain without new radicular symptoms; type 2, low back pain with minor radicular symptoms; or type 3, cauda equina syndrome and/or severe radicular symptoms (Fig. 1).

Lumbar spinal fusion is recognized as an effective treatment for segmental instability. Since Cloward first introduced PLIF more than 70 years ago, this technique has been widely utilized to treat spinal instability and stenosis [1]. TLIF was initially popularized by Harms, as a modification to the PLIF procedure, and gained popularity because of the innovative surgical concept’s ability to eliminate the neural retraction injuries inherent to the traditional PLIF procedure [2]. MI-TLIF has evolved as an ideal treatment strategy for a wide variety of lumbar conditions. Complications commonly associated with MI-PLIF/TLIF include intraoperative neurological injury, interbody implant migration, dural tear, and surgical site infection. Dural tears (frequency, 4.6%) [6, 7], infection (frequency, 2%) [3, 8], screw misplacement (nerve root impingement; frequency, 1–3%) [9], retroperitoneal injury (rare) [10], wrong-level surgery (frequency, 0.03–0.04%) [11] and interbody cage migration (frequency, 2%) [12‐14] have all been reported and must be considered in the planning and execution of these surgeries. The incidence of cage migration is reportedly 2.5–6.3% [4, 15, 17]. In this study, the frequency of cage migration was 4.4%. Migration of the intervertebral cage is a relatively rare but potentially serious complication, placing adjacent neural elements at risk, and in some cases heralding nonunion. If a cage migrates forwards into the retroperitoneum or backwards into the vertebral canal, mispositioned cages can have serious clinical consequences. Of these, posterior migration is the more serious due to the risks of nerve root compression or cauda equina syndrome, intensifying neurological symptoms and making the fusion unsuccessful, as described previously [13, 15, 16].

A bullet-shaped cage, higher posterior disc height, presence of scoliotic curvature, and undersized cages have been reported as possible risk factors for cage migration [17]. In our study, risk factors for cage migration were younger age, male sex, and fusion at the L5/S1 level. The mean postoperative duration to onset of cage migration was 3.2 months (range, 2–6 months). Migration has been shown to be connected to small cage size, cage type, inadequate anterior seating of the cage, multilevel fusion procedures and osteoporosis [14, 18]. Surgeon experience less than 3 years and lumbar spondylolisthesis appear to be significantly associated with posterior cage migration [17]. Several studies have emphasized the importance of preserving vertebral bone endplates to prevent cage migration, and techniques to achieve this are very important and in high demand [19, 20]. To minimize the risk of cage migration, we should place the cage with the center located posterior to the disc center. Preparation of the disc and endplate plays an important role in avoiding cage migration. Before cage insertion, a trial cage should reach an adequate depth after disc preparation and bone graft packing. Monitoring of cage position in real time during surgery by fluorescent imaging is also indispensable. The depth ratio on lateral radiography offers a useful tool to detect malpositioned cages. Once decompression and instrumented fusion have been completed, adequate compression must be applied via the pedicle screws to prevent cage migration.

Revision surgeries for cage migration include replacement of the superior graft material, proper end-plate preparation, correction of any technical errors, enhancement of biological fusion, and improvement of the biomechanical environment [21]. However, revision surgery for cage migration is technically demanding. With the posterior approach, dural retraction and nerve root mobilization are difficult because of massive epidural fibrosis, leading to postoperative leg pain or palsy. The safety and efficacy of anterior lumbar interbody fusion (ALIF) as a salvage option for failed posterior lumbar fusion surgery have been demonstrated [22]. In this study, we have proposed a new, treatment-based classification for symptomatic posterior cage migration (Fig. 1). The most common type in our cohort was type 2, involving nerve root compression causing severe radicular pain. However, the most difficult case is type 3, which results in cauda equina syndrome and usually requires cage removal from an anterior approach (Fig. 9). If the only symptom is low back pain due to pseudoarthrosis (type 1), removal of the migrated cage may not even be necessary. Anterior approach or contralateral side approach TLIF is also indicated. For type 2 migrated cage, a removal of the migrated cage is mandatory. In this type, the same side TLIF approach is indicated. For type 3 migrated, a removal of the migrated cage from anterior approach is the best option because posterior approach is a little risky to remove the cage. To the best of our knowledge, no classification of cage migration has previously been reported. This therefore appears to be the first classification of cage protrusion in the literature.

The use of intraoperative navigation has also been shown to provide greater accuracy and less variation in device placement. In this study, we utilized a navigated osteotome, navigated high-speed burr, and the OLIF51 procedure for revision surgeries. To date, few studies have reported the clinical usage of spinal intraoperative O-arm-based navigation with ALIF for cage backout after TLIF/PLIF. Phan et al. reported one case in which ALIF and total disc replacement surgery were performed with intraoperative CT-based navigation, showing good results [23]. Park et al. described this method of navigation as safe, feasible and apparently accurate in LLIF procedures [24]. Our previous study showed improvements in the accuracy of screw positioning and a decrease in the misplaced screw rate to 3.7% among scoliosis patients with O-arm navigation [9]. O-arm-navigated surgery has been validated as a surgical intervention for lumbar revision surgery. O-arm navigation techniques offer theoretically sound advantages, appear to represent viable options for salvage operations and are safe in well-trained hands.

This case series yielded a large amount of valuable information regarding cage migration. However, this study did have some limitations that should be kept in mind. First, because of the low incidence, the total number of cage migration cases was insufficient to draw many definitive conclusions. Second, the study was conducted retrospectively without randomization. Third, in addition to cage position and cage height, whether sagittal alignment parameters for the spine represent risk factors have yet to be determined. These parameters were not included in the present study because whole-spine radiographs were not available during the period of case collection. Further studies paying attention to such aspects are needed to clarify which factors influence cage migration.

Our new classification for cage migration appears useful to create viable strategies for addressing symptomatic lumbar pseudoarthrosis requiring revision surgery due to cage protrusion after MI-PLIF/TLIF. Cage removal with a navigated curette, osteotome, and high-speed burr is an effective technique that reduces surgical time and blood loss. Anterior removal of the cage and OLIF51 should be considered for type 3 protrusion (cauda equine syndrome and/or severe radicular symptom). This new procedure reduces radiation exposure to the surgeon and operating room staff compared with conventional fluoroscopic techniques.