Sacroiliac screws were placed in five patients under image guidance from 2005 through 2007 with permission from, and in accordance with the guidelines of, the Institute of Research and Ethics (Table 1). All patients were referred to the senior author (KDK) through the UC Davis Health System. The decision to perform sacropelvic fixation was mainly based on the need for additional points of fixation in the pelvis to attain successful lumbosacral fusion. All patients had received preoperative lumbosacral CT and MRI scans — and, in some cases, plain radiographs.
Case No. Age/Gender Indications for surgery Procedure 1 29/M Trauma:
S1-S2 sacral fracture w/cauda equina syndrome
S1-S3 decompression with L4-S3 fusion using bilateral S2/ sacroiliac screws and S1 and S3 pedicle screws 2 63/M Infection/Reoperation:
L4-L5 osteomyelitis fracture/dislocation
Removal of old screws and circumferential decompression and fusion using L1-S1 pedicle and S2/ sacroiliac screws 3 84/F Degeneration/Reoperation: Lumbar scoliosis and stenosis with three prior surgeries and osteoporosis L2-L5 decompression and T12-S2 fusion using S1 pedicle and bilateral S2/ sacroiliac screws 4 58/F Degeneration:
Lumbar stenosis and scoliosis
L2-S1 decompression fusion using T12-S1 pedicle and S2/ sacroiliac screws 5 54/M Infection/Reoperation:
Chronic osteomyelitis after prior surgeries
Removal of anterior instrumentation and corpectomies T10-L3 with T9-L4 cage. Staged T7-T10 and L4-L5 pedicle screws with bilateral S1/ sacroiliac screws
Table 1. Characteristics of five consecutive cases of image-guided sacropelvic fixation performed by senior author from 2005 through 2007
The Stealth Station Image Guidance System (IGS) (Medtronic Navigation, Louisville, CO, USA) was used for preoperative trajectory planning and intraoperative screw placement. Preoperative CT data were imported into the IGS computer workstation, and anatomic registration points were selected. The points selected on bony landmarks were those that were most likely to be identified once standard posterior lumbosacral exposure for posterolateral fusion was attained. Registration landmarks usually consisted of one or two spinous processes, and two to four facet joints in the lumbosacral region. In patients undergoing reoperation, any easily identifiable bony landmarks were chosen. The ideal screw trajectory traversed the cortices of the sacroiliac joint, with the tip of the screw within the ilium in the cortical bone above the sciatic notch (Fig. 1). Sacroiliac screw trajectories were planned using twodimensional navigational and three-dimensional (3D) reconstructed views (Fig. 2). Maximum ideal screw length and diameter were determined from the preoperative virtual plan.
A midline skin incision and subperiosteal dissection were performed to expose the relevant spinal levels out to the transverse processes and the sacral ala. A passive reference arc was clamped onto the spinous process of the most cephalad vertebra that was exposed. The reference arc was cranially angled away from the lumbosacral region to minimize the possibility that it could be inadvertently moved after registration. Point registration and surface merging were completed to optimize accurate registration. Good registration accuracy was consistently verified by touching observable bony landmarks with the IGS probe tip and correlating probe location, with the virtual image displayed on the monitor. Pedicle screws were typically inserted at lumbar levels using standard techniques.
Next, in keeping with the preoperative plan, IGS was used to select sacroiliac screw entry points on the sacrum. A new entry point not included in the preoperative plan may have been selected to facilitate rod alignment, contouring and attachment to screw heads. Typically, the point of entry was noted at S1 or S2. The cortical wall of the sacrum was penetrated by an IGS awl. Next, an IGS sounding probe was used to create a pilot hole for the screw. The pilot hole trajectory may have been modified from the virtual plan for placement of the screw in the ideal portion of the ilium (the cortical bone above the sciatic notch). The sounding probe was advanced with IGS tracking. The virtual navigation view was used with tactile feedback to discern when the sacroiliac joint was being traversed. Pilot hole integrity was checked with a ball-tip probe; then, the IGS tap was used to tap the same pilot hole. The tapped hole was typically probed again to assess for cortical breach. Sacroiliac screws were then inserted, with care taken to avoid deviating from the tapped trajectory. Sacroiliac screws were connected to the remaining pedicle screws with contoured rods to form a physiologic lumbosacral curve. The fusion surface was routinely decorticated, and bone graft was placed after the rod had been secured with setscrews. A postoperative CT scan was performed to check the screw position.
Seven cadaveric spines (ilio-lumbosacral) were harvested and stripped of all musculature while all of their ligamentous structures, vertebral bodies and intervertebral discs were kept intact. X-rays were obtained to assess the specimens' bone quality. Spine specimens were secured into test fixtures at L1 and the sacrum. Intact specimens were mounted onto a sixdegrees-of-freedom spine motion simulator, and nondestructive ROM flexibility testing was conducted by applying pure moments (±6 Nm) in three randomly selected physiologic planes: flexion-extension, lateral bending and axial rotation[7-8]. Segmental L5–S1 motions were collected via an optoelectronic motion measurement system with infrared light-emitting diodes placed at L5 and S1. After intact testing, 6.5-mmdiameter bilateral pedicle screws (REVERE® Stabilization System, Globus Medical, Inc., Audubon, PA, USA) were placed at L1–L5, without violation of facet joints or disc spaces. Iliac screws that were 7.5 mm in diameter were used, and a 5.5-mm-diameter rod was instrumented for all biomechanical testing to be performed. Sacroiliac fixation was performed by using (1) bilateral sacroiliac screws and (2) bilateral S1 pedicle screws and iliac screws (S + I) (Fig. 3). Lumbosacral rods were sized and contoured appropriately for the seven specimens. Additional offset connectors were used to connect iliac screws to rods (S + I construct). ROM data were normalized relative to the value of the intact spine. Statistical analysis was performed on the raw data using repeated measures analysis of variance and Tukey's post hoc test (P < 0.05). The Data Analysis Tool Pack available in Microsoft Excel (Microsoft Corp., Redmond, WA, United States) was used to run the analysis of variance, and the post-hoc test was calculated based on the equation for Tukey's honest significant difference test.