With advances in surgery and anaesthesia, posterior lumbar fusion and pedicle screw instrumentation have been used progressively more often in aged patients with lumbar degenerative disease by spine surgeons. However, screw loosening, migration, and back-out is the most common postoperative complication of the pedicle screw, which usually results in painful nonunion, progressive kyphosis, and revision surgery. Clinical studies have reported an overall instrumentation failure rate of 1–15% in ordinary patients, even 10–62.8% in patients with osteoporosis[16–19]. Reports in the surgical literature indicate that CAPSI has been widely used to enhance fixation strength to improve pedicle screw stability in osteoporotic spines [20]. However, previous reports also showed that a rigid instrument may grossly alter the physiologic load transmission at the instrumented level and has a cascading degenerative effect over the adjacent discs[21, 22].
It is generally agreed that ASD can be divided into radiologic adjacent segment degeneration (ASDeg) and adjacent segment disease (ASDis). Studies have documented a rate of clinical ASDis between 2% and 12.2% at different follow-up periods [23, 24]. For patients with ASDis who underwent revision, the satisfaction rate was approximately only 54%, which is significantly lower than that of other patients (83%) [25]. In addition, the initial disc degeneration of adjacent segments in elderly patients was more severe than that in younger patients, which has been reported to increase the incidence of ASD. Thus, for the study we report here, we planned to clarify the effect of different volumes of PMMA on ASD among aged patients and to provide a useful reference to spinal surgeons when considering CAPSI for early patients.
Finite element analysis can accurately characterize the complex biomechanical mechanism of the spine and clearly show the stress distribution of each part. Proper geometric characteristics of models are essential for the accurate outcome of FE analysis. In previous finite element studies, the three-dimensional solid models of the pedicle screw and bone cement were constructed by using SolidWorks or Hypermesh software[26, 27], which cannot accurately simulate the characteristics of the material object. Therefore, in the current study, the threaded pedicle screw (Fig. 1) and lumpy bone cement were constructed to be realistic with a 3D scanner and the patient’s CT data, respectively. In addition, lumbar degenerative diseases such as lumbar spinal stenosis occur mostly in the L4-S1 segment. Thus, this study used a model of L4-5 and L4-S1 segmental fusion to analyse the effect of volume of PMMA after CAPSI.
In this study, in order to distinguish the influence in the adjacent discs, the ROM, disc stress, and facet stress were compared with the intact value. An increase in the ROM, disc stress, and facet stress were found in adjacent segments in all loading directions that were more pronounced in the double-level CAPSI model with 2.5 ml PMMA. However, the variation in single- and double-level CAPSI models was similar for various volumes of PMMA (1 ml, 1.73 ml, and 2.5 ml). Although CAPSI increases the risk of adjacent segment degeneration, this study did not find meaningful associations between ASD incidence and the volumes of PMMA. Otherwise, experimental data have reported that a cement volume between 1.0 and 3.0 ml significantly improves screw stability, whereas a volume beyond 3.0 ml does not increase the purchasing strength linearly but results in an increase of cement leakage [20, 28]. Therefore, within a certain range, increasing the volume of PMMA does not significantly affect the stability of adjacent segments, and PMMA volumes between 1.0 ml and 2.5 ml can be selectively used according to different degrees of osteoporosis.
This computational study was based on finite element analysis and has some limitations. First, because the scanning accuracy of the 3D scanner in the uneven and subtle parts of the screw (such as the screw thread) is insufficient, further processing is required in the scanned models by Geomagic Studio 2013 and SolidWorks software, and the extra processing may result in a modicum of distortion in the model. Furthermore, several simplifications were necessary for creating the finite element model, such as the characteristics of ligaments, paraspinal muscles, and body weight, which also limited the results. Therefore, further cadaver studies and clinical observations are necessary to reach a more precise conclusion.