Following the popularity of spinal pedicle screw instrumentation, much attention has been focused on the importance of bone mineral density (BMD) for instrumented spine fusion in clinical practice [15–20]. Clinical reports have shown that sacral screw fixation has a high failure rate. The failure of sacral screw fixation may be due to several factors such as inadequate sacral bone purchase, inappropriate direction or depth of the screw insertion, and bone mineral density variations within the sacrum [13, 21–23]. Previous studies [24–26] employed BMD measurements to obtain area density (g/cm²) using dual energy X-ray absorptiometry (DEXA), or volumetric density (g/cm³) using quantitative computed tomography (QCT) (Fig. 36.2). QCT is advantageous in that it can quantify BMD along the screw insertion pathway, and is therefore a more accurate technique than DEXA for measuring the relationship between BMD and the strength of screw fixation [24, 27]. Most investigations into sacral BMD distributions have been carried out on aged specimens, with only a few studies on young cadaveric specimens. Since sacral fixations are commonly applied in patients of young to middle age, BMD measurements of young specimens provide clinically relevant data for the assessment and comparison of sacral fixation. In this text, we will focus on the BMD variations within the S1 body and ala of young patients.

Based on 13 young adult fresh cadaveric sacrum specimens, recent studies [27, 28] reported a mean BMD of the S1 vertebral body of 381.9 mg/cm³, 31.9 % higher than that of the sacral ala (mean 296.9 mg/cm³). Table 36.1 and Fig. 36.3 list the bone mineral density at different regions of the S1 body and ala. There are significant differences in BMD, with the regions near the lateral posterior and lateral anterior areas of the vertebral body having higher BMD. This again shows that these areas would provide better purchase for pedicle screw fixation. Furthermore, BMD of the posterior region closest to the spinal canal was lower than that of lateral and central areas. In the ala, the BMD of the internal anterior areas closest to the S1 body were the highest with a BMD of 326.8 mg/cm³.

The mean BMD at different layers (Fig. 36.4) of the S1 body and ala has also listed in Table 36.2. The BMD of the first layer in the S1 body, close to the superior end plate, was significantly higher than the others with a BMD of 516 mg/cm³. The BMD was lowest in the central layer of the body, and increased again towards the inferior endplate. This suggests that sacral pedicle screws directed close to or cephalad penetrating the superior sacral end plate would achieve better purchase. In the ala, the BMD of the top layer was highest with a BMD of 329 mg/cm³ and the BMD decreased caudally from the top layer.

Based on the principle that inserting the pedicle screws directly penetrating the superior endplate should increase the fixation strength, a novel sacral screw fixation technique has been developed recently with good clinical results. To support the clinical findings, the mechanical stability of this technique was conducted afterwards [29]. Following cyclic loading, the mean maximum insertion torque and mean pull-out force were significantly higher for bicortical fixation through the S1 endplate (mean 3.17 N · m and 1457N) than bicortical fixation through the anterior sacral cortex (mean 1.98 N · m and 1122N). By taking advantage of the density of the upper sacrum and the thick cortical endplate of the S1, the bicortical S1 endplate sacral pedicle screw fixation technique provided a greater screw insertion torque and a stronger screw fixation after cyclic loading than the conventional bicortical anterior cortex fixation technique. In addition, for both fixation techniques, the screw insertion torque was shown to correlate with the screw pull-out force following cyclic loading, indicating that insertion torque was a good intraoperative indicator of screw pull-out force.

Based on the BMD distribution within the sacrum [13, 27, 28], the recommended screw pathway is shown in Fig. 36.5. Further to radiographs, CT scans, and specimen observation, it is suggested that the S1 pedicle screw should be directed parallel to the S1 end plate in the sagittal plane, and the ala screw inserted to the tip of the ala along the ala slope. The insertion angle of the S1 pedicle screw in the transverse plane should be between 15° and 25° medially and the angle of the ala screw should be 30–40° laterally for better bone purchase (Fig. 36.6 from CT). If the insertion angle is less than 10° for S1 pedicle screws, the screw will not obtain good purchase on the S1 promontary, leading to a greatly reduced purchase of bone and less biomechanical stability. On the other hand, if the insertion angle for the S1 screw is larger than 25°, the potential for injury to the spinal cord is higher [30]. With respect to ala screw insertion, Edwards and Louis preferred lateral placement into the sacral ala at 35–45° [31, 32], while Mirkovic and associates suggested the screw should be angled at 30–40° in order to obtain a better purchase [7]. If the angle of ala screw is greater than 45°, the potential for lumbosacral trunk injury is high [7].

The importance of vertebral bone mineral density for screw fixation has been the subject of several studies to determine screw stability and predict lumbar vertebrae strength [7, 11–13, 20, 33–35]. Some have suggested that pre-operative measurement of BMD is necessary for transpedicular screwing in osteoporotic cases [28, 36, 37]. Other studies [20, 38] found that vertebral bodies of human lumbosacral spines had a mean BMD of 92 mg/cm³ and the mean BMD of the sacrum was 152 mg/cm³ for the elderly of mean age of 78 years old. Studies also indicated that at a QCT BMD value of less than 90 mg/ml, early loosening of the screw may be expected, while it is less likely with an BMD of more than 120 mg/cm³. The mean BMD from young adults with mean age of 31 years was approximately 2.5 times greater than that in the above mentioned group. Since the sacrum mainly consists of cancellous bone and the lumbosacral junction sustains more load than others above the sacrum, failure of sacral screw fixation can occur without significant osteoporosis.