Institutional Review Board (IRB) approval was obtained (IRB No. 3-2017-0340). The researchers conducted a retrospective consecutive review of all patients who underwent sacropelvic fixation in their institution from June 2011 to May 2017. Sacropelvic fixation was performed for fusion augmentation in the cases of deformity correction operations requiring spinal fusion down to the sacrum. SAI screw fixation was performed in 79 patients and CI screw fixation in 50 patients. The inclusion criteria for this study were at least one year of a follow-up period and patients having the appropriate imaging studies such as whole spine X-ray, computed tomography (CT), and magnetic resonance imaging preoperatively and postoperatively. Based on these criteria, 43 of 79 patients with SAI screw fixation and 25 of 50 patients with CI screw fixation were included.

Patient demographics were investigated including age at operation, sex, surgical diagnosis, comorbidity such as diabetes mellitus and osteoporosis, smoking history, body mass index (BMI), bone mineral density, duration of follow-up, and preoperative and postoperative visual analog scale (VAS) for pain. Surgical characteristics were also investigated, including the history of previous lumbosacral operation, spinal level of screw fixation, postoperative complications, and reoperation. Postoperative complications included proximal junctional failure or kyphosis, distal instrument failure, rod fracture, wound dehiscence, SSI, and screw prominence. Reoperation was defined as any unplanned procedure required for the treatment of pseudoarthrosis or other complications.

Radiographic parameters were measured on the anteroposterior and lateral radiograph of the standing whole spine X-ray. Preoperative, immediately postoperative, 6 and 12 months after the surgery, and final follow-up standing whole spine X-ray were studied. The measured parameters were sagittal vertical axis (SVA), pelvic incidence (PI), pelvic tilt (PT), sacral slope, lumbar lordosis (LL), thoracic kyphosis, and cervical lordosis. Proximal junctional failure was defined as “proximal junctional sagittal Cobb angle between the lower endplate of the uppermost instrumented vertebra and the upper endplate of the two supra-adjacent vertebrae ≥10° and at least 10° greater than the preoperative measurement”⁴⁾. Unlike proximal junctional problems, there was no consensus on the definition of distal instrumentation or lumbo-sacro-pelvic fixation failure. With reference to previous studies^3,5,11), distal instrumentation failure was defined as breakage of the SAI or CI screws, halo formation around the screw, or screw pullout. If patients had focal wound dehiscence or tenderness on the buttock areas directly over the SAI or CI screw head immediately after the surgery, they were considered symptomatic screw prominence. Fusion status was assessed by CT scan one year after the surgery and the outcome was expressed as the grading system that was previously used in other studies (Table 1).

The step-by-step procedures for the CI and SAI screw fixation techniques are described in detail in previous studies^2,17,25) and will be discussed briefly herein. For the CI screw fixation technique, a simple posterior midline incision is used, followed by an additional dissection of the posterior superior iliac spine that will be the entry point of the screw. A separate skin incision was not used for the insertion of the iliac screws. Before inserting the screw, a portion of bone at the entry point was removed by performing osteotomy to create space to minimize the prominence of the screw head. The screw must be placed between the inner table and the outer table of the iliac bone and the trajectory of the screw directed with an angulation between 20° to 45° caudally and 30° to 45° laterally. For SAI screw insertion, a simple posterior midline incision was made, and the muscle layer was dissected subperiosteally to expose the sacral bone. To preserve muscle coverage over the instrumentation, subperiosteal dissection of the sacrum was limited to the entry point located 2 to 3 mm caudal and 2 to 3mm lateral to the neural foramen of S1. The SAI screw was directed toward the greater trochanter, rostral to the sciatic notch, and angulation was usually 40° lateral in the axial plane and 40° caudal in the sagittal plane; the exact trajectory was determined based on the preoperative CT and the screw was generally placed under intraoperative fluoroscopic guidance. This trajectory crosses the SI joint and the screw penetrates three cortical surfaces. The tip of the screw is engaged in the dense bone above the sciatic notch.

Intergroup comparison of categorical variables was achieved using Fisher’s exact test. The unpaired t-test was used to analyze continuous variables. Data were presented as the mean±standard deviation unless specified. Moreover, the Mann-Whitney U test was used to analyze nonparametric continuous variables. The Statistical Package for the Social Sciences software (version 20.0; SPSS Inc., Chicago, IL, USA) was used for all statistical analyses. A p-value of less than 0.05 was considered statistically significant.

The mean duration of follow-up was 640±244 days vs. 627±282 days (SAI group vs. CI group); the mean age of patient was 65.3±9.0 years vs. 60.6±13.5 years (range, 28-79 years); and the mean number of screw fixation levels was 7.3±2.1 segments vs. 7.6±3.1 segments (range, 3-14 segments) (Table 2). The predominant diagnoses were adult spinal degenerative deformity, iatrogenic flat back syndrome, and secondary acquired kyphosis. Secondary acquired kyphosis included post-traumatic and post-infectious kyphosis.

Regarding baseline patient demographics, there were no statistically significant differences in age, sex, BMI, duration of follow-up, smoking, diabetes mellitus, osteoporosis, and previous spinal surgery between the two groups. In both groups, the SVA, LL, PT, and PILL improved after surgery compared to before surgery, but no statistically significant difference was observed between the two groups (Table 3). Lumbosacral fusion rates showed no statistically significant difference between the SAI group and the CI group (90.7% vs. 92.0%, p=0.878) (Table 1, 4). There were improvements in the VAS for pain compared to preoperative values, but no statistically significant differences in the SAI and CI groups (6.7±1.6 vs. 1.6±1.4, p<0.01; 7.2±1.3 vs. 2.2±1.3, p<0.01, respectively) (Table 4).

There were no statistically significant differences between the SAI group and CI group with respect to rod fracture, wound dehiscence, and screw prominence. The rate of reoperation, one of the major complications, did not show any statistically significant difference between the two groups (16.3% vs. 28.0%, p=0.270). Additionally, the rate of proximal junctional failure and proximal junctional kyphosis was not different between the two groups statistically. The SAI group had a significantly higher rate of distal screw fracture than the CI group (16.3% vs. 0%, p=0.042) (Fig. 1). The SAI group demonstrated a good result regarding SSI compared to the CI group (0% vs. 16%, p=0.016) (Table 5).

All of the distal instrument failures occurred in the SAI group except one case of peri-screw halo formation in the CI group. Among eight patients with complications in the SAI group, seven patients experienced distal screw fractures. The screw fracture occurred in the junction between the head of the screw and the neck of the screw shaft in all patients. The duration of screw fracture was 300±30 days (range, 5-29 months). Among these seven patients, pseudoarthrosis in the lumbosacral area was Grade I in two cases, Grade II in two cases, and Grade III in three cases. Inter-group analyses between the distal screw fracture group and the non-distal screw fracture group were also performed. The rate of lumbosacral pseudoarthrosis (Grades III, IV) was higher in the distal screw fracture group compared to the non-distal screw fracture group, but there was no statistical significance (28.6% vs 6.6%, p=0.112). The improvements of SVA, LL, PT, and PI-LL after surgery did not show statistically significant differences between the two groups. All screws used in spinopelvic screw fixation were poly-axial and had diameters ranging from 7.5 to 8.5mm, and their lengths ranged from 80 to 90mm. There was no statistically significant difference between the two groups.

Regarding the two patients with wound dehiscence and the four patients with SSI in CI groups, the lumbosacral area was the main area for complications. One patient with wound dehiscence had a flap graft surgery performed by a plastic surgeon in the institute. Patients with SSI were initially treated with antibiotics including vancomycin and third-generation cephalosporin for several days. If antibiotic treatment was not working based on laboratory results, the wound was opened and irrigated with normal saline and betadine-mixed saline. Gentamycin-mixed saline was added in one patient.