Relation of the lumbosacral trunk to the sacro-iliac joint

This study aims to evaluate the relation between the lumbosacral trunk (LT) and the sacro-iliac joint (SIJ). In forty anatomic specimens (hemipelves) a classical antero-lateral approach to the SIJ was performed. The SIJ was marked at the linea terminalis (reference point A). Reference point B was situated at the upper edge of the interosseous sacro-iliac ligament. The length of the SIJ (distance A to B) and the distance between point A and the ventral branch of the fourth (L4) and fifth (L5) lumbar spinal nerves at the linea terminalis were measured. The SIJ had a mean length of 58.0 mm. The ventral branch of L5 was located closer to the SIJ in very long SIJs (mean length: ≥ 6.5 cm; mean distance: 9.8 mm) compared to very short joints (≤ 5 mm; mean distance: 11.3 mm). For the ventral branch of L4, very long SIJs had a mean distance of 7 mm and very short joints an average distance of 9.7 mm between point A and the nerve branch. A safe zone of approximately 1 cm to 2 cm (anterior to posterior) is present on the sacral surface (lateral to medial) for safe fixation of plates during anterior plate stabilization of the SIJ. Pelves with a shorter dorsoventral diameter of the most superior part of the SIJ apparently give more space for inserting plates.


Material and methods
Dissection was performed on 40 anatomical specimens of hemipelves obtained from 38 whole pelves, embalmed using Thiel's method 11 . This special embalming technique provides a life-like model through the preservation of the original tissue color, consistency and degree of transparency. None of the specimens showed signs of malformations or prior interventions in the area of interest. The anatomical specimens were donated to the Medical University of Graz. During their lifetime, donors had given their written informed consent to participate in anatomical studies.
Dissection. Dissection was performed in the supine position. After removal of the bowel content from the abdominal and pelvic cavities, the iliac muscle was released from its origin by blunt dissection and the nerves of the lumbar plexus were identified, including the femoral nerve (FN), the obturator nerve (ON) and the LT. For better visualization, parts of the iliopsoas muscle, the external and internal iliac vessels and parts of the FN and ON were removed (Fig. 1). A blunt dissection was performed to release the structures covering the sacroiliac joint. The sacro-iliac ligaments were partially removed for adequate visualization of the upper course of the SIJ from inside the true pelvis along its course lateral to the upper sacrum.
Further dissection was performed on the anterior surface of the sacrum to visualize the anterior rami of the sacral plexus.
Measurements. The superior part of the fibrous capsule covering the SIJ was marked with a cannula at the intersection with the linea terminalis (reference point A), and at the posterior-superior edge of the interosseous sacro-iliac ligament (reference point B).
The length of the SIJ was measured as the straight distance between reference point A and B as the true length is highly variable due to a usually slightly curved course of the superior joint line. All distances were measured with a digital caliper and in millimeters.
First, the distance between reference point A and the lateral border of the ventral L4 and L5 rami at the linea terminalis (Inlet plane) was measured.
Starting from reference point A, along the true superior SIJ plane, the distances between the joint line and the lateral border of L4 and L5 were measured in 1 cm increments perpendicular to the joint line starting anterior in a posterior direction. These intervals were chosen due to their convenient applicability during surgery. For better standardization, these 1 cm subsections were evaluated as percentages of the individual length of the superior aspect of the SIJ (Fig. 2a).
Additionally, the distance from reference point A to the union of L4 and L5 was measured. Starting from reference point A, 1 cm distal to the linea terminalis (Inlet plane) in a longitudinal direction the horizontal distance to the LT was measured.
Data analysis. The collected data were analyzed in the statistical software R 12 . To assess differences between side and sex, Mann-Whitney U-test were employed. Concerning descriptive statistics, continuous variables are presented as mean and standard deviation (SD), minimum and maximum, categorical data as frequencies and percentages.

Results
The hemipelves were gained from 38 body donors (15 female and 23 male donors) with a mean age of 81.6 years (SD: 8.4; range 56-96). Twenty left and 20 right hemipelves were dissected. Dissections were performed unilaterally in 36 and bilaterally in two specimens. The SIJ had a mean length of 58.0 mm (SD: 0.5; range 4.5-6.9). The average distance between point A and the union of L4 and L5 was 10.5 mm (SD: 2.6; range 5.5-17.5). In nearly all anatomic specimens the union occurred within the true pelvis at the anterior sacral surface.
The distances between the measurement points in 1 cm increments from reference point A, orientated posteriorly, and the lateral borders of L4 and L5 are presented in Table 1. In relation to the total length of the superior aspect of the SIJ, the measurement points were at a mean of 17.4% (1 cm distance), 34.8% (2 cm distance), 52.1% (3 cm distance), 69.5% (4 cm distance) and 86.3% (5 cm distance).
As the ventral ramus of L4 exits from its respective intervertebral foramen more superior when compared to L5, it was found to descend more distant to the SIJ (see Table 1).

Measurements regarding the ventral ramus of L5.
Measurements of the distance from reference point A at the linea terminalis to the ventral branch of L5 and the perpendicular distance at 17.4%, 34.8%, 52.1% Considering the measured length of the SIJ, 13 hemipelves presented with a length between 4.5 cm and 5.5 cm (mean: 5.2 cm) and were therefore ranked as pelves with a "shorter" SIJ. Thirteen hemipelves showed a length ≥ 6 cm (range 6.1-6.9 cm; mean: 6.4 cm) and were classified as pelves with a "longer" SIJ. Pelves with a "shorter" SIJ showed the respective mean distances of 9.1 mm (from reference point A), 13,7 mm (17.4%), 17,9 mm (34.8%) and 21 mm (52.1%). In comparison "longer" SIJs had mean intervals of 9.6 mm (reference point A), 14.0 mm (17.4%), 17,1 mm (34.8%) and 20.6 mm (52.1%), respectively. "Very short" SIJs were defined as articulation lines with a length ≤ 5 cm and concerned three cases. The respective distances between the SIJ and L5 were 11.3 mm (reference point A), 14.2 mm (17.4%), 16.3 mm (34.8%) and 18.9 mm (52.1%). Pelves including an SIJ with a length ≥ 6.5 cm (n = 6) were characterized as "very long" with corresponding distances of 9.8 mm (reference point A), 14.5 mm (17.4%), 16.5 mm (34.8%) and 18.9 mm (52.1%), indicating slightly more place directly at the linea terminalis level in "very short" SIJs. Analysis of sex and side differences. The superior aspect of the SIJ had a mean length of 57.8 mm (SD: 0.6; range 4.5-6.8) in males and of 58.5.mm (SD: 0.6; range 4.5-6.8) in females (p = 0.64). Male body donors were statistically significantly (p ≤ 0.001) older (mean: 87.7 years; SD: 4.8; range 75-92) when compared to females (mean: 77.9; SD: 8.2; range 56-96). The distances between the defined measurement points alongside the superior portion of the SIJ and L4 and L5 with regard to sex are displayed in Table 2. Here, no significant difference between the sexes could be evaluated for any of the distances.
The superior portion of the SIJ had a mean length of 5.9 mm (SD: 0.6; range 4.5-6.9) on the left and 5.7 mm (SD: 0.5; range 5-6.6) on the right side (p = 0.50). Distances between the measurement points and L4 and L5 concerning side are displayed in Table 3. No significant differences could be evaluated for any of the distances with regard to body side.

Discussion
The aim of the present anatomical study was to correlate the distance between the ventral rami of the fourth and fifth lumbar spinal nerves with the length of the anterior wall of the SIJ. www.nature.com/scientificreports/ The superior wall of the SIJ had a mean length of 58.0 mm. Intervals between the anterior part of the joint and the lateral border of the ventral branch of L5 were at a mean of 8.9 mm (superior part of the fibrous capsule covering the SIJ at the intersection with the linea terminalis), 13.1 mm (17.4%), 16.8 mm (34.8%) and 19.9 mm (52.1%). The distance between the superior portion of the fibrous capsule covering the SIJ at the linea terminalis and the lateral border of L5 was increased in pelves with a "very short" (length ≤ 5 cm) SIJ (mean: 11.3 mm) when compared to those with a "very long" (≥ 6.5 cm) SIJ (mean: 9.8 mm). Regarding L4, the corresponding distances between the SIJ and the ventral branches' lateral border were 7.1 mm (superior part of the fibrous capsule covering the SIJ at the intersection with the linea terminalis), 10.6 mm (17.4%), 13.9 mm (34.8%), 17.5 mm (52.1%), 20.7 mm (69.5%) and 21.8 mm (86.3%). Again, "very short" SIJs had longer (mean: 9.7 mm) distances between the reference point at the linea terminalis and the lateral border of the ventral branch of the nerve in comparison to the "very long" subgroup (mean: 7 mm). Concerning all evaluated data, no statistically significant sex or side differences could be observed.
In cases of anterior plate stabilization during pure SIJ dislocations or fracture dislocations, the LT is at risk, especially during surgical exposure of the superolateral aspect of the sacrum 7,13,14 .
Atlihan et al. 15 performed anatomical dissections of the LT in 60 specimens to evaluate the nerves at risk during approaches to the anterior wall of the SIJ. The distance between the joint line and the lateral border of L5 was at a mean of 18.4 mm at the height of its exit point from the intervertebral foramen. At the linea terminalis, the interval between the LT and the SIJ was at a mean of 5.3 mm. In the present collective this was longer with means of 8.  www.nature.com/scientificreports/ true pelvis a mean distance of 10 mm for both structures was observed. In the current study, these values were at a mean of 8.9 mm (L5) und 7.1 mm (L4) at the height of the linea terminalis. Bai et al. 9  , whereas for L5 we evaluated a slightly lower mean (19.9 mm). At the linea terminalis, we evaluated smaller average distances for both, L4 (7.1 mm) and L5 (8.9 mm).
Differences of the intervals between the nerves and the SIJ may possibly be traced back to varying embalming methods.
Waikakul et al. 16 evaluated 52 embalmed hemipelves regarding the branches of the lumbosacral plexus. With reference to the most anterior part of the SIJ, the union of the ventral branches of L4 and L5 to the LT was located above (11 specimens), directly on (12 specimens) or below (8 cases) this basing point. In our study, nearly all unions occurred inside the true pelvis at the anterior sacral surface.
Concerning sex differences, Ebraheim et al. 17 dissected the LT bilaterally in twenty anatomical specimens. Authors reported the ventral branch of L5 as the thickest branch in males (mean: 4.4 mm) and the ventral portion of S1 as the thickest branch in females (mean: 4.3 mm). Though these data haven't been evaluated in the present sample, we found no statistically significant differences between the sexes for any of our evaluated distances.
Pascal-Moussellard et al. 18 reported a 5% affection of the ventral branch of L5 and the LT in sacral zone1 fractures according to the Denis classification. However, plate positioning and fixation at the linea terminalis is Table 3. Side differences of intervals between the ventral branches of L4/L5 and respective measurements points (Point A on the linea terminalis and 17.4%/34.8%/52.1%/69.5%/86.3% of the total length of the sacroiliac joint). Data are presented in millimeters. www.nature.com/scientificreports/ only possible in direct relation to these nerve roots. Only a direct subperiosteal dissection can avoid injury to these branches. In the "very small" pelves of our sample, the distance between the anterior wall of the SIJ and the lateral borders of L5 (mean: 11.3 mm) and L4 (mean: 9.7 mm) at the linea terminalis increased to acceptable values. At the level 1 cm (17.2% of SIJ length) proximal/posterior to point A this distance was ≥ 1 cm, which allows usually placement of one screw into one plate hole. This given study is accompanied by the following limitations, which need to be addressed in the context of the here given results and interpretations. First, it has to be stated that the length of the SIJ has been measured as a straight distance between points A and B in an effort to simplify from the complex curvatures of the sacroiliac joint and adjacent substructures. This simplification will likely have resulted in marked distortion, which was however necessary to obtain comparable measurements. A further limitation has been introduced by the choice of the embalming technique for the tissues which were deployed for the measurements 11 . Modified Thiel's solution 11 includes a significant amount of osmotically active and fat dissolving chemicals. As a consequence, cell washout and tissue degradation as well as extensive degreasing are observed in the tissues. These effects have been exemplified in the tissues presented in the figures, showing a markedly different state of preservation. Given the nerve tissues investigated here consist of a significant amount of fat, the bias introduced by tissue distortion or shrinkage is unknown 19 . In consequence, further study comparing the LT with without the given embalming effects are necessary in near future. Also, we solely used anatomical geriatric specimens of Caucasian origin. Therefore, there might be a potential selection bias, as our findings might not align with morphometries obtained in other ethnicities or a younger cohort.
A safe zone of approximately 1 cm to 2 cm (anterior to posterior) is present on the sacral surface (lateral to medial) for safe fixation of plates during anterior plate stabilization of the SIJ. The results of our study indicate that pelves with a shorter dorsoventral diameter of the most superior part of the SIJ include more space for plate insertion.