The incidence of cage migration varies from 0.8–23% in the literature 10,11, with the majority of studies investigating cage retropulsion with an overhang of the posterior edge of the vertebral body. In the present study, minimal and significant cage migration were detectable in 61.2% and 23.9% of the patients, respectively, resulting in an overall migration rate of 85.1%. Retropulsion of the cage was not detectable in any case.
The significantly higher migration rate in this study compared to the literature is considered to be due to the accuracy of the measurement method. We assert that minimal cage migration can be understood as a common phenomenon after interbody fusions, and it has received little attention in the scientific analysis of migration behavior. The results of the current study further demonstrate that cage migration is not equivalent to cage retropulsion.
Complications such as neural compression and pseudarthrosis are common in association with cage retropulsion. This can also be seen in the revision rate, which has been reported in the literature as 33.3–75% 10,14.
In the present study, the presence of cage migration was not associated with the clinical scores (ODI, VAS, analgesic requirement) or radiographic fusion evaluation. The revision rate of the patient group with cage migration was low at 1.5%. It is possible that minor and clinically asymptomatic cage migration is not an expression of persistent segmental instability but rather an indication of an increasing incorporation of the cage in the postoperative course and thus a component of the bony fusion process.
Most studies have described cage migration in the posterior direction 22,23. The results of the present study suggest that cage migration often occurs simultaneously in multiple directions. In the anterior–posterior direction, an almost exclusively posterior cage position change was observed.
With regard to posterior cage migration and retropulsion, numerous potential risk factors have already been identified. One study reported an association with a preoperative high disc space 10. Other studies have found that undersized cages are a risk factor for posterior cage migration 24,25. From a biomechanical perspective, this is caused by the insufficient restoration of tensile stress to the annulus fibrosus and the ligamentous apparatus—a factor that contributes significantly to the primary stability of cages 26,27.
The positioning of the cage within the disc space continues to be important for primary stability. Various biomechanical studies have shown that segmental stability increases anteriorly in a position-dependent manner, and this can be considered a consequence of a greater distance between the cage and the center of rotation 28–31. Conversely, others have demonstrated an increased risk of migration in association with a posterior cage position 7,25.
In the present study, the effects of cage position on migration were also examined. No significant correlation was observed. However, in contrast to the results in the literature, all cages were positioned anteriorly or centrally and not in the posterior region of the endplate.
The pattern of cage migration or retropulsion can be narrowed down to an average of 1–4 months postoperatively 10,11,14,24,32. Consequently, the one-year follow-up interval, as in the current study, can be considered appropriate but does not allow a conclusion to be drawn regarding the cage migration pattern over time.
To further develop standardized follow-up concepts for postoperative mobilization after lumbar interbody fusion, studies are needed to examine the progression of cage migration and subsidence over time.
The reconstruction of the original intervertebral space height to restore physiologic lumbar lordosis and the width of the neuroforamina is among the major goals of interbody lumbar fusions. Bony fusion should also preserve the position of the segment in the long term 33–35. The most common cause of secondary loss of correction is the subsidence of the cage into the adjacent endplate. Within the first eight postoperative months, there was again a loss of intervertebral space height to an average of 13.2 mm. From > 2 mm, the loss of correction was considered subsidence. Consequently, the subsidence rate was 76.7% 36.
In the present study, the incidence of cage subsidence at one year was 58.2%. As was the case with cage migration, subsidence was not shown to have a negative effect on clinical outcomes or the success of bony fusion, consistent with the literature 36–38.
In the evaluation of the radiological results of this study, the effect of cage position on subsidence was confirmed. The subsidence rate of the group with a central cage position was more than twice as high (77.14% vs. 37.5%) as that of the group with an anterior cage position. Anatomically, this position-dependent subsidence behavior is explained by the inhomogeneous nature of the endplate, the thickness of which increases from the center toward the periphery 39–42. Therefore, the endplate exhibited the highest compressive strength in the region of its cortical rim.
The patients in the present study showed an age-related increase in cage subsidence. This observation may reflect age-related changes in bone quality and an increasing prevalence of osteoporosis, as previously reported in the literature 37,43. However, because data on the bone density of the patients are not available in the current study, a relationship with osteoporosis can only be hypothesized.
The size and shape of the cage are also factors to be considered 44–46. To reduce the risk of subsidence through optimal load distribution, a cage with the largest possible bearing surface is recommended. Finally, another surgical factor to consider with respect to potential subsidence is the extent of endplate preparation as part of the disc excision procedure 47.
In this context, the increased incidence of subsidence of the posterior implant portion into the endplate in the present study suggests uneven support and higher pressure loading in the peripheral region of the cage. As the posterior implant portion was also located further centrally, subsidence in this area was also favored. Indications of increased subsidence as a result of endplate injury could be ruled out by measurement in the first postoperative CT.
Furthermore, the base plate was generally more frequently affected by subsidence than the upper endplate (49.3% vs. 37.3%). However, this result has not been confirmed in the literature. Studies that have differentiated between the base and the upper endplate in the analysis of subsidence behavior usually showed subsidence of the cage into the upper endplate 36,48−50. A possible explanation for the different results in the present study is the observation that subsidence into the inferior endplate occurred more frequently in association with an incomplete reduction of the anteriorly slipped vertebral body. The contact surface of the cage on the inferior endplate was more centrally located when the slipped vertebra was not fully corrected compared to the endplate. This observation again confirms the position-dependent subsidence risk of the cage and emphasizes the importance of reducing the slipped vertebra. In cases in which a complete correction of translational malalignment is not possible intraoperatively, the observation in this study may provide strategic guidance. Here, a more anterior positioning of the cage is recommended to ensure good support by the anterior apophyseal ring of the upper vertebra or a far lateral positioning of the cage in the lateral part of the apophyseal ring to reduce the risk of the cage subsiding into the upper endplate.
The evaluation of fusion status has been the subject of numerous studies. Nevertheless, the comparability of individual studies is generally problematic, as there is no consensus on the scientific consideration of fusion status with regard to the imaging techniques and fusion criteria to be applied 51,52. CT is the most sensitive method in the evaluation of solid fusion and the detection of pseudarthrosis 53. The main criteria for a successful interbody or posterolateral fusion are evidence of trabecular bone bridges between the endplates or articular and transverse processes and the absence of lysis fringes in the fusion region. Conversely, signs of material fatigue (screw loosening or fracture) are considered indirect evidence of non-fusion. The absence of migration and subsidence of the implant can be considered a fusion criterion 16.
The approach of standardizing fusion evaluation by measuring cage migration or subsidence was further investigated in the present study. No significant correlation between cage migration or subsidence and fusion outcomes was found. Only one of the five patients with non-fusion showed significant cage migration, and two patients showed significant cage subsidence at the same time. Therefore, the analysis of migration and subsidence behavior is not important for the evaluation of interbody fusion.
The patients in this study had a uniform radiologic follow-up through CT 12 months postoperatively. The fusion grading established by Bridwell 19 and Eck 20 was used to evaluate fusion status. The fusion rate at 12 months was 92.5%; that is, 62 of the 67 patients showed grade I (n = 51) or grade II (n = 11) fusion of the anterior and/or posterior columns.
Comparably high fusion rates are found in the literature, with fusion rates of 80–100% depending on the implants used, imaging techniques, and fusion criteria 4,53−58.
In conclusion, the incidence of cage migration was considerable. However, as cage migration and subsidence were not associated with bony fusion, their clinical significance was considered limited.