Virtual mapping of 260 three-dimensional hemipelvises to analyse gender-specific differences in minimally invasive retrograde lag screw placement in the posterior acetabular column using the anterior pelvic and midsagittal plane as reference

Background Due to complex pelvic geometry, percutaneous screw placement in the posterior acetabular column can pose a major challenge even for experienced surgeons. Methods The present study examined the preformed bone stock of the posterior acetabular column in 260 hemipelvises. Retrograde posterior column screws were virtually implanted using iPlan® CMF (BrainLAB AG, Feldkirchen, Germany); maximal implant length, maximal implant diameter and angles between the screw trajectories and the reference planes anterior pelvic plane as well as the midsagittal plane were assessed for gender-specific differences. Results The virtual analysis of the preformed bone stock column showed two constrictions of crucial clinical importance. These were located 49.6 ± 3.4 (41.0–60.2) mm (inferior margin of acetabulum) and 77.0 ± 5.6 (66.5–95.3) mm (centre of acetabulum) from the entry point of the implant in men and respectively 43.7 ± 2.3 (38.3–49.3) mm as well as 71.2 ± 3.5 (63.5–79.99) mm in women (men vs. women: p < 0.001). The entry point of the retrograde posterior column screw was located dorsal from the transition of the lower margin of the ischial tuberosity to ramus inferior pointing to the medial margin of the ischial tuberosity. In female patients, the entry point was located significantly closer to the medial margin of the ischial tuberosity. However, 7.3 mm screws can generally be used in men and women. The angle between the screw trajectory and the anterior pelvic plane in sagittal section was 14.0 ± 4.9 (2.5–28.6) °, the angle between the screw trajectory and the midsagittal plane in axial section was 31.1 ± 12.8 (1.5–77.9) ° and the angle between the screw trajectory and the midsagittal plane in coronal section was 8.4 ± 3.8 (1.5–20.0) °. For all angles, significant gender-specific differences were found (p < 0.001). Conclusion Therefore, the anterior pelvic plane as well as the midsagittal plane can facilitate intraoperative orientation for retrograde posterior column screw placement considering gender-specific differences in preformed bone corridor, implant length as well as angles formed between screw trajectory and these reference planes.


Background
Anatomic reduction and retention are pivotal in surgical therapy of acetabular fractures to prevent posttraumatic osteoarthritis. Open reduction and stabilization via ilioinguinal or Kocher-Langenbeck-approach and in rare cases extended approaches is the internationally accepted standard [1].
Minimally-invasive osteosynthesis approaches are limited due to the demanding reduction, which cannot always be completely achieved, and because of the complex threedimensional geometry of the pelvis with limited bone corridors for screw placement. Fluoroscopy-assisted percutaneous screw osteosynthesis of the acetabulum may require several trials for correct screw placement which leads to high radiation doses [2].
Based on previous findings [3,4] suggesting that there are gender-specific differences regarding the best implant position, we first defined an optimal screw position for the retrograde placed posterior column screw and then analysed the surrounding preformed bone corridor. In order to gain statistically relevant data without the limitation of a reduced sample size of previously performed studies [3,[5][6][7][8], we decided to perform a virtual study using the technique described in 2010 [2]. We examined 260 hemipelvises virtually using data from CT scans of 130 uninjured European pelvises which were collected during clinical routine exams at our trauma centre. Our first hypothesis was that there are gender-specific differences considering the optimal implant position.
In a second step, we wanted to analyse, if there are landmark pelvic planes which could be used as reference, thus facilitating the visualization and screw placement for the surgeon. We have chosen the anterior pelvic plane as well as the midsagittal pelvic plane as reference planes since they are commonly used in hip arthroplasty [9]. We analysed the angle between the screw trajectory and these planes. Our second hypothesis was, that there are genderspecific differences considering the angles between these planes and the optimal screw position.

Methods
This study has been approved by the ethical committee of the University of Tuebingen (No 580/2012R). A patient collective of 260 uninjured hemipelvises was examined in regard to significant gender-specific differences relevant in the placement of the posterior column screw. Therefore, a CT-scan archive spanning from 06/2003 till 06/2010 was searched for exams of the pelvis to a standardized protocol (16 line helical CT Siemens Somatom Sensation 16, 120 kV, 160 mAs, slice thickness of 0.75 mm, collimation 0.75; Siemens AG, Munich, Germany). Patients with fractures of the acetabulum or pelvic ring, hip dysplasia or advanced osteoarthritis of the hip and age < 16 years or > 86 years were excluded as well as patients who underwent arthroplasty of the hip or osteosynthesis of the pelvic ring or acetabulum. Moreover, the study included only European patients.

Patient collective
The analysis of anatomically preformed periacetabular screw corridors was performed on 130 CT-Scans of uninjured Europeans (260 hemipelvises; 65 male/65 female patients). The collective showed a medium age of 50.9 ± 20.3 (16-85) years and a medium BMI of 25.8 ± 5.3 (18.2-46.1) kg/m 2 ( Table 1). Written informed consent for participation in the study was obtained from all participants.
Statistical analysis revealed that there was no significant differences between female and male patients considering age (p = 0.548) and BMI (p = 0.742). However men were significantly taller (p < 0.001) and heavier (p = 0.006) than women.

Analysis of anatomical preformed bone corridor for the posterior column screw
The analysis of the individual anatomical preformed corridor for the screw was performed on 260 hemipelvises. First, the maximal implant length and the maximal implant diameter were assessed. Therefore, the screw trajectory was increased in diameter until the implant reached the inner corticalis without penetrating it. Maximal screw length was reached when the end of the virtual screw reached the corticalis of the iliac fossa. Then, according to Ebraheim et al. [6], nine (9) parallel sections-with respect to the individual anatomy-orthogonal to the screw axis were chosen and the diameter of the bone corridor and the maximal implant diameter were determined in each section. The sectional planes were positioned according to the individual anatomy of the patient and screw length to achieve intra-and interindividual comparability (Fig. 1). The entry points were defined in respect to the anatomical landmarks-ischial tuberosity and the caudal margin of the ramus inferior of ossis pubis.

Virtual screw placement and position analysis
The virtual screw placement as well as position analysis was performed using iPlan® software (BrainLab AG, Feldkirchen, Germany). To calibrate the software and to set up the desired planes, several landmarks of the pelvis had to be manually registered. These landmarks included the anterior superior iliac spine on both sides as well as the tuberculum pubicum on both sides. With these four points the anterior pelvic plane was determined (Fig. 2). The midsagittal plane was placed vertically to the anterior pelvic plane and through the centre of the symphysis and the centre of the sacral spine.
Then, the angle between the screw axis and the both pelvic planes mentioned above was measured using also iPlan® 3.0 software. This was done in sagittal sections for the anterior pelvic plane and in axial as well as coronal sections for the midsagittal plane. To use the angle measurement tool of iPlan® 3.0, a parallel plane to the respective plane was placed through the endpoint of the screw. The angle formed between this parallel plane and the screw was then measured. An example is shown in Fig. 3a-d.

Statistical analysis
Statistical analysis was performed with IBM SPSS Statistical package Version 22. Student's t-test as well as Mann-Whitney U-test was used to analyse the anthropometric data. Correlation analysis was performed with Pearson's test for linear and Spearmen's test for non-linear correlation. A p value of <0.05 was considered  Fig. 1a shows a sagittal, Fig. 1b a coronal section. Fig. 1c shows nine axial sections which were planned according to Ebraheim et al. [5]. (8) axial section at the upper acetabular margin (6) axial section at the centre of acetabulum and (4) at the lower margin of acetabulum (9) axial section in the middle of the stretch between screw exit point and upper acetabular margin (7) axial section in the middle of the stretch between upper acetabular margin and centre of the acetabulum (5) axial section in the middle of the stretch between centre of acetabulum and lower acetabular margin (3) axial section at two thirds of the stretch between screw entry point and lower acetabular margin (2) axial section at one third of the stretch between screw entry point and lower acetabular margin (1) axial section at the screw entry point. In each of the axial sections, the maximum screw diameter is shown

Entry point of the posterior column screw
The entry point of the posterior column screw was located 12.7 ± 2.0 (6.6-18.1) mm dorsal from the transition of ramus pubis inferior and ischial tuberosity in anteriorposterior direction and 12.1 ± 1.9 (8.1-18. 2) mm lateral of medial margin of ischial tuberosity in male patients (patient in supine position). In female patients, the entry point was located significantly more medial (meaning closer to medial margin of the ischial tuberosity) than in male patients (10.8 ± 1.8 (8.1-18.2) mm; p < 0.001) (see Table 4). The patient collective (260 hemipelvises) showed a significant correlation between increasing implant diameter and increasing distance of the entry point from medial margin of ischial tuberosity (r = 0.4057).

Anatomically preformed screw corridor
The analysis of the anatomically preformed periacetabular screw corridor of the posterior column showed two sites of constriction which are important in daily clinical routine. These constrictions can be found at the inferior margin of acetabulum and at the centre of acetabulum In average, they were located 49. The detailed configuration of the anatomically preformed bone corridor for the posterior column screw in respective to the medio-lateral and superior-inferior diameter of the bone stock in relation to the maximum screw diameter is shown in Figs. 4 and 5 for male respectively female patients.
Statistical analysis showed significantly higher mediolateral (p < 0.001) and superior-inferior (p ≤ 0.001) bone  Mean ± SD 12.9 ± 1.5 13.0 ± 1. stock diameters in men compared to women. According to that, the maximally possible screw diameter was higher in male than in female patients at the first (p < 0.001) and second constriction (p < 0.001).
Angles between screw trajectory and anterior pelvic plane as well midsagittal pelvic plane and correlation of screw trajectory and anthropometric parameters The angle between the screw trajectory and the anterior pelvic plane in sagittal section was 14.0 ± 4.9 (2.5-28.6)°, a significant difference was found between both sexes (men vs. women p < 0.001). The angle between the screw trajectory and the midsagittal plane in axial section was 31.1 ± 12.8 (1.5-77.9)°with significant difference between men and women (p < 0.001). The angle between the screw trajectory and the midsagittal plane in coronal section was 8.4 ± 3.8 (1.5-20.0)°also with a significant difference between both sexes (p < 0.001). Details are shown in Table 7.

Discussion
Percutaneous stabilization as a minimally invasive treatment option is possible with non-displaced but instable fractures of the acetabulum with a risk of secondary displacement to allow early mobilization of the patient [10,11]. Moreover, displaced acetabular fractures which can be reduced via lag screw are suitable for percutaneous stabilization [3]. Patients with an elevated perioperative risk from an open approach such as old patients or patients with several injuries in which a reduced approach related morbidity outweighs perfect anatomic fracture reduction, should also be considered for percutaneous stabilization. In addition to this, patients who will need total hip arthroplasty in the short run due to the configuration of the fracture might also benefit from earlier mobilization and protection of soft tissue when undergoing percutaneous stabilization instead of open reduction. Percutaneous screw placement in the posterior column is difficult due to the complex three-dimensional pelvic  geometry. Despite improved imaging with new fluoroscopic views of the acetabulum, 3D-CT scans and the intraoperative use of computer-assisted navigation, there is a danger of perforation the bone corridor or the acetabulum [2,12]. Therefore, it is important that the surgeon visualizes the bone stock with the screw corridor in preoperative planning. The present work defined reproducible screw positions in order to analyse the narrow bone corridor of the posterior column. The study was performed on CT-scans of 130 uninjured Europeans as described by [2]. The advantage of this method used in the present work is the fact, that all screws were virtually planned using advanced 3D planning software which is used in daily clinical routine, especially in neuro-and craniofacial surgery (iPlan® CMF 3.0, BrainLAB AG, Feldkirchen, Germany). One of the main focuses of the study was to investigate gender-specific differences in minimally invasive posterior column screw placement, especially in regard to reference planes which are clinically used in pelvic surgery. So far, there is only one study with this focus [3].
The present study showed that mean implant length in men was 141.8 mm and 127.2 mm in women with significant difference between men and women (p < 0.001). Shahulhameed et al. in 2010 found a mean posterior column screw length of 138.5 mm for the whole collective of 11 pairs of adult cadaveric pelves; this is in accordance to our findings [8]. Mu et al. however found a mean screw length of 104.8 mm for the whole collective of 30 adult dried hemipelvis specimens [7]. The difference in screw length can be attributed to the fact, that Mu et al. placed antegrade screw whereas we used retrogradely placed screws in the posterior column. The percutaneous minimally invasive approach only allows for retrograde screw Table 6 Second constriction at the acetabulum centre. In retrograde posterior column screw placement, second constriction of the anatomically preformed bone corridor could be found at the centre of acetabulum in examined Europeans with uninjured pelvis  placement. We deliberately chose this approach since it offers the advantage of highly reduced soft tissue trauma compared to open approaches [1]. The entry point of the posterior column screw was located at mean 12.7 mm dorsal from the transition of ramus pubis inferior and ischial tuberosity in anterior-posterior direction and at mean 12.1 mm lateral of medial margin of ischial tuberosity in male patients. In female patients, the entry point was located significantly more medial than in male patients (mean: 10.8 mm). Similar gender-specific differences of the entry point could be found by Puchwein et al. [13] and by Dienstknecht et al. [3]. In another study, Dienstknecht also found a gender-specific entry point in posterior column screw placement when measuring the distance from the screw entry point to the landmarks ischial tuberosity, centre of symphysis, iliopectinal eminence and anterior inferior iliac spine [14].
The analysis of the anatomically preformed bone corridor of the posterior column was virtually performed using axial sections referenced to the screw axis according to the method described by Ebraheim et al. [6]. In order to account for the individual anatomy, the axial sections were adjusted to anatomical landmarks such as the walls of the acetabulum. Other workgroups did not take individual anatomical characteristics into account but used sections in defined distances along the screw axis [3,5,6]. Thus, in   [3]. In our collective, the maximally possible implant diameter was at mean 12.9 mm in men and 11.2 mm in women. This means that the established use of 7.3 mm screws in men and women would have been save. Similar results were found by Banerjee in 2011 [15] and by Shahulhameed et al. in 2010 [8]. It is also in accordance to Attias et al. [5], Puchwein et al. [13] as well as Dienstknecht et al. [3].
As mentioned above we wanted to evolve from single osseous landmarks to established reference planes which can easily be identified intraoperatively; this is why we have chosen the anterior pelvic plane and the midsagittal plane as reference plane and measured the angles between these planes in different CT scans. The anterior pelvic plane [9] and the midsagittal plane are well known from hip arthroplasty [16][17][18][19][20][21] and are commonly used and easily visualized in the OR. We measured the angle between the posterior column screw as well as the midsagittal plane and the anterior pelvic plane for each hemipelvis comparing the angles found in male and female patients.
The mean angle between the screw trajectory and the anterior pelvic plane in sagittal section was 14.0°, the mean angle between the screw trajectory and the midsagittal plane in axial section was 31.1°and the mean angle between the screw trajectory and the midsagittal plane in coronal section was 8.4°. For all angles, significant differences were found between both sexes (men vs. women p < 0.001).
Dienstknecht et al. also measured angles formed between the posterior column screw and several references lines and two different planes on a.p. radiographs. These reference lines were the tangent to the pubic tubercule, the caudal border of superior pubic ramus, the linea ilioischiadica and the tangent to acetabular rim. A defined vertical and horizontal plane were used as reference planes to measure the angles [3]; this approach has some limitations: The measurement was performed on standard a.p. radiographs, thus only in a two dimensional system compared to the three-dimensional CT scans used in the present study. Moreover, the planes chosen were defined as vertical and horizontal and are not further described. Even more important with respect to the used a.p. radiographs are the findings of Pullen et al. who could show in 2014 that the individual positional variability in pelvic standard radiographs can lead to large-magnitude changes in radiographic acetabular measures [22]. The present study eliminates these problems, since CT scans were used and the reference planes were defined by the pelvic bone landmarks and individually adopted for each pelvis examined. Moreover, the used planes are well established in daily clinical routine and are easily visualized both during procedure planning and in the OR.
However, there are limitations to the present study. The findings are limited to retrograde lag screw placement in the posterior acetabular column and the study has been performed in healthy individuals with no fracture present. Moreover, only European patients were examined, so the findings may be limited for use in Europeans. The effectiveness of the described technique in regard to safe screw placement has to be proved in clinical use.

Conclusions
The anatomical differences in the pelvic geometry of men and women lead to gender-specific differences in preformed bone corridor, implant length and diameter as well as angles formed between screw trajectory of posterior acetabular column screw and the reference planes (anterior pelvic and midsagittal plane). The angles and the reference planes are reproducible and can be easily visualized in the OR, so they can be helpful in minimally invasive retrograde screw placement in the posterior column. The effectiveness of the described technique in regard to safe screw placement has to be proved in clinical use.