Introduction

Ilio-sacral (IS) screw fixation including the transiliac–transsacral (TITS) screw placement is a technically dependent procedure because of complex posterior pelvic structures and high degree of upper sacral variability. Bone overlap, obesity, poor bone quality, and bowel gas can make it difficult to identify well-known landmarks in imaging for safe placement [1,2,3,4,5]. Accurate fluoroscopic imaging of IS screw placement is essential for verifying the distinction of sacral morphology and screw trajectory, and preventing malposition-related complications [6,7,8]. Although there have been many reports on good visualization of safe corridor and radiologic recognition of sacral variations [1, 2, 4, 9], few studies have reported the implication of IS screw placement [5, 10,11,12].

During IS screw placement, three-dimensional (3D) understanding of the safe zone is based on the surgeon’s knowledge of pelvic anatomy. This information is a prerequisite for obtaining optimal intraosseous positions in the first and second sacral segment [13]. Concerning the difference between normal and variant sacra, it is difficult to verify the degree of safe zone obliquity using conventional CT scans [14]. If the CT images were measured using picture archiving and communication system (PACS) software, its true cross-sectional area for safe IS screw placement might be overestimated or distorted because the cutting plane is not perpendicular to the sacral alar axis [4]. Thus, it might be hard to determine the expected IS screw trajectory and the possibility of TITS screw fixation for the first (S1) and second sacral segment (S2) by the usual techniques for preoperative planning. Therefore, the primary purposes of this computational simulation study were to introduce an optimal and consistent IS screw trajectory for S1 and to identify the possibility of TITS screw and its implication in Asian sacrum by simulating a virtual 7.0-mm-sized screw fixation.

Materials and methods

Human body digital data were collected from the Korea Institute of Science and Technology Information after approval. CT data of 105 adult cadavers who underwent continuous 1.0-mm slice CT scans (Pronto, Hitachi, Japan) in supine position were collected. None of the cadavers had pelvic problems based on medical records view. CT data in Digital Imaging and Communications in Medicine (DICOM) format were imported into Mimics® software (Materialise Interactive Medical Image Control System; Materialise, Antwerp, Belgium) to reconstruct 3D models of the pelvis including the sacrum and two iliac bones. Due to poor image quality, 23 cadavers were excluded. The remaining 82 adult cadavers (42 males and 40 females) were enrolled. Their mean age and height at death were 52.1 years (range 21–60 years, SD 9.2) and 161.3 cm (range 146–176 cm, SD 7.1), respectively. After generating the 3D pelvis model, the transparency mode of the model was adjusted to differentiate cancellous bone from sacral foramen and spinal canal. In addition, the safe zones of the first (SZS1) and second sacral segments (SZS2) for TITS screw fixation were verified [5, 10]. The surface area of SZS2, horizontal distance (HDS2), and vertical distance (HDS2) of the second sacral segment were measured using the Mimics® software (Fig. 1a–c).

Fig. 1
figure 1

a, b To differentiate cancellous bone from the adjacent structures, the transparency mode of the pelvis model was adjusted and the safe zone was verified. c The horizontal and vertical distance of safe zone in S2 was measured. d–f Oblique and TITS cylinders were inserted into S1 and S2

Full size image

To simulate the insertion of a conventional 7.0-mm IS screw into the ideal position of S1, a straight cylinder designated the ‘CAD object’ was applied using the Mimics® software. The insertion of oblique cylinder was performed over the midline in the first sacral body (Fig. 1d) [15]. Two 7.0-mm-sized transverse cylinders representing the TITS cylinders were inserted into S1 and S2. They were positioned from the outer cortex of iliac bone to the outer cortex of the opposite iliac bone (Fig. 1e, f) [15,16,17]. To verify the path of cylinders, the pelvic outlet view was adjusted as the cartilage of pubic symphysis overlays with the midline of the sacrum and the inferior border of first sacral foramen unlike prior reports [18, 19] (Fig. 2). The ideal position was defined as containment of a screw within bony confines without violating the first and second sacral foramen, spinal canal, or upper surface of the two iliac bones [5, 10, 14]. After inserting the three cylinders (two TITS cylinders and one oblique cylinder), their positions and entry points were fine-tuned and verified by an experienced surgeon (corresponding author). Due to high variability and complex plane, as well as technical issues, we failed to assess the surface area and anatomic features of SZS1. However, direct comparison between the two safe zones (SZS1 and SZS2) was possible by removing two iliac bones.

Fig. 2
figure 2

The modified pelvic outlet view was adjusted as the cartilage of pubic symphysis was overlying the midline of the sacrum and inferior border of the first sacral foramen

Full size image

All pelvis models were divided into two groups depending on the possibility of TITS cylinder insertion for S1. In the normal group, the transverse insertion of TITS could be possible without any violation. In the sacral variation group, the insertion could not be without any violation [4, 5, 20]. Fourteen models (six female models and eight male models) were identified as sacral variation due to angulated and narrowed SZS1. The entry point and trajectory of cylinders were assessed, and the length of each cylinder was measured to identify the maximum potential length of IS screw. All measurements are presented as mean, range, and standard deviation (SD). Chi-square test and two-sample t test were used to compare means between the pelvic normal group and the variation group. Statistical significance was set at p < 0.05. SPSS statistical software package for Windows version 23.0 (SPSS Inc., Chicago, IL, USA) and R × 64 3.1.1 (R Foundation for Statistical Computing, Vienna, Austria) were used for statistical analyses.

Results

There was no statistically significant difference in sex (p = 0.626) or height (p = 0.419) between the normal and variation group. In the normal group, the average length of the oblique cylinder in S1 was 98.8 mm (range 82.4–132.2 mm, SD 8.7 mm). The average length of the TITS cylinder in S1 was 151.4 mm (range 127.9–178.2 mm, SD 9.6 mm) and 134.2 mm (range 97.8–164.1 mm, SD 10.5 mm) in S2. In the variation group, the average length of the oblique cylinder in S1 was 101.2 mm (range 90.5–110.9 mm, SD 6.1 mm) and the average length of the TITS cylinder in S2 was 144.6 mm (range 128.0–160.3 mm, SD 8.1 mm). The average length of oblique cylinder in the sacral variation group was not significantly different (p = 0.322) from that in the normal group. However, there was a statistically significant difference in the length of TITS cylinder of the second sacral segment between the two groups (p < 0.001).

The average VDS2, HDS2, and the area of SZS2 of all models were 15.5 mm (range 8.7–24.4 mm, SD 3.0 mm), 18.3 mm (range 12.7–26.6 mm, SD 2.9 mm), and 221.1 mm2 (range 91.1–386.7 mm2, SD 68.5 mm2), respectively. The average area of SZS2 was 209.6 mm2 (range 91.1–376.2 mm2, SD 65.4 mm2) in the normal group and 276.9 mm2 (range 145.4–386.7 mm2, SD 56.0 mm2) in the variation group. When the anatomic variable of second segment was compared between two groups, the VDS2 (p < 0.001) and the area of SZS2 (p = 0.001) of variation group were higher than those of the normal group. However, the HDS2 was not significantly different (p = 0.126) between the two groups.

For TITS cylinder insertion of S1, the entry point should be placed just inferior to the iliac cortical density to some extent in the true lateral view to prevent the cortical violation. Concerning the TITS cylinder insertion of S2, the ideal entry was located inferior or slightly anterior, but not posterior, compared to the entry for S1. For safe placement of oblique cylinder in sacral variation model, the height of ideal entry point needed to be always located at the level of first sacral foramen. When the cylinder trajectory was directed toward the opposite upper corner of first sacral body, there was no violation of cortex of sacral ala, first sacral foramen, or iliac cortex regardless of the presence of sacral variation (Fig. 3).

Fig. 3
figure 3

a, b In sacral normal model, two TITS screw fixation could be inserted without cortical perforation. c, d In sacral variation, a hybrid fixation construct consisting of oblique screw for S1 and TITS screw for S2 could be performed to achieve sufficient fixation strength

Full size image

Discussion

Despite concerns of the neurovascular injuries related to IS screw fixation [21,22,23], it has been considered the most important procedure to manage unstable pelvic ring injuries. Most surgeons still perform IS screw fixation as a percutaneous procedure guided by image intensifier depending on their own experiences. We also have performed IS screw fixations for the prior 10 years without significant difference from other surgeons. Thus, the present simulation of the insertion of virtual IS screws examined the possibility of TITS screw fixation in patients of Asian ethnicity with the goal of achieving an optimal and consistent IS screw trajectory. This goal would hopefully prevent malposition-related complications, which would facilitate use of TITS screws in clinical situations [15, 17, 24] regardless of the presence of sacral variation. Our results demonstrated that the optimal and consistent screw trajectory should be obliquely directed toward the opposite upper corner of the first sacral body at the level of first sacral foramen regardless of sacral variation. It could be found that SZS2 was a sufficient osseous site for the fixation of the 7.0-mm TITS screws even in Asians and was larger than SZS1 in those with sacral variation.

A recent study [4] reported radiographic quantification and its implications for IS screw fixation depending on sacral variation. Other authors [9] reported statistically significant differences in the widths of SZS2 between normal and dysmorphic sacra in CT scans. Concerning SZS2, the cross-sectional area was more than twice as large in dysmorphic sacra compared to normal [10]. In our study, however, although SZS2 was also larger in the variation group (p = 0.001), the difference was only 24%. The dichotomy between studies could reflect differences in tools and methods used, including two-dimensional versus three-dimensional reconstructions, and PACS versus Mimics® software. Concerning the possibility of TITS screw fixation in S2, no model had insufficient osseous site for insertion in our study, in contrast to prior findings [4]. Accordingly, the TITS screw could be inserted transversely in the second sacral segment of all models. Considering that Asian people are smaller than Westerners, it has been generally accepted that their area of SZS2 has smaller osseous site for the insertion of 7.0-mm-sized screws based on personnel communications between orthopedic trauma surgeons, even though the supporting evidence is scant [10, 25]. In our study, SZS2 averaged 221.1 mm2 (range 91.1–386.7 mm2), which was larger than expected. Through the postoperative CT scans, we also realized that the TITS screw fixation could be performed without cortical perforation.

The percutaneous IS screw fixation is technically demanding because of the site’s three-dimensional anatomic complexity, being close to neurovascular structures and having frequent upper sacral morphological variations [1, 2, 4, 5, 8, 10, 24, 26, 27]. Although a safe zone and optimal IS screw trajectory were identified using preoperative planning, there is still the possibility of extra-osseous screw placement because of misinterpreted fluoroscopic imaging. For this purpose, we used a 3D rendering program (Mimics®) to allow free 360° rotations with magnification in any plane and to virtually implant IS screw in the optimal position. This computational analysis revealed that the verification for violation of sacral ala could be easily identified by the modified pelvic outlet view during the insertion of oblique cylinder (conventional IS screw), even though it was in sacral variation. Accordingly, this simply modified outlet view of pelvis might be used to check screw trajectory during fluoroscopically guided procedure if possible (Fig. 2) [19]. Concerning the maximal potential length of the optimal IS screw, the average length of the transverse cylinder for S1 was 151.4 mm (screw range 125–180 mm) and the average length for S2 was 134.2 mm (screw range 95–165 mm). Their results differ from prior findings [15]. When it was considered that 7.0-mm-sized screws are not manufactured in lengths sufficient to span between the posterior iliac bones through the upper sacral segments, the available length of 7.0-mm-sized screws should be checked preoperatively.

This computational simulation study has several fundamental limitations. First, because all measurements were from non-fracture sacrum cadavers, the usefulness of our results may be limited in practical situations with rather descriptive characteristics. Second, the number of enrolled pelvises was not enough to generalize the results to all Asian people. Nevertheless, our descriptive findings offer practical information about screw trajectory and the utility of SZS2 and have meaningful implications for conventional IS screw and TITS screw fixation for pelvic ring injuries.

Conclusions

Considering the high variability of S1, safe and consistent IS screw trajectory should be obliquely directed toward the opposite upper corner of first sacral body at the level of first sacral foramen, regardless of the presence of sacral variation. When the TITS screw fixation for S1 could not be fixed due to sacral variation and other conditions, a hybrid fixation construct consisting of oblique screw for S1 and TITS screw for S2 will be useful to achieve sufficient fixation strength, because the osseous site of the second sacral segment is large enough to place a 7.0-mm-sized IS screw.