1 INTRODUCTION

Understanding the biomechanics of the spine is important to understand normal locomotion, especially when considering the fact that different animal species vary in their vertebral column anatomy and related characteristics such as the vertebral shape and number. For example, the normal number of lumbar vertebrae in horses, oxen and dogs is five to seven, six and seven, respectively (Evans & De Lahunta, 2013a; Liebich and König, 2014; Levine et al., 2007).

Any variation may create changes in the function of structures within or around the spine (Bürger & Lang, 1993). Similarly, the shape, orientation and size of the joints between different bones can affect the function of the surrounding regions. The sacrum in dogs has four main joints; the lumbosacral, the sacrocaudal and the paired (right and left) sacroiliac joints (Evans & De Lahunta, 2013a).

The seventh lumbar vertebra (L.7) has a body, spinous process, cranial and caudal articular surfaces and paired costal processes (Figure 1). In dogs, there are seven lumbar vertebrae, which their bodies are generally longer than those vertebrae of the other regions. The width of the lumbar vertebrae from the 1st to the 7th usually increases as their length increases gradually from L.1 to L.5 or L.6 vertebra (Evans & De Lahunta, 2013a). Each lumbar vertebra has a spinous process, caudal and cranial articular surfaces, transverse processes, mammillary processes, caudal articular processes and body (Evans & De Lahunta, 2013a). All lumbar vertebrae are characterized by their enlarged cranial and caudal end plates, except for the L.7, which has less enlarged end plate and articulates with the sacrum (Sisson et al., 1975).

The lumbosacral joint is where the L.7 articulates with the sacrum (Evans & De Lahunta, 2013b) and this is clinically important area in dogs because it may be involved in the lumbosacral transitional anomaly in many dog breeds. In dogs, it has been reported that the L.7 is shorter than the rest of the lumbar vertebrae (Hermanson & Lahunta, 2020; Miller et al., 1979). A lumbosacral transitional vertebra is a vertebra within vertebral column that might be shown to have the characteristics of both segments; lumbar and sacral (Larsen, 1977). A lumbosacral transitional vertebra is an anomalously formed vertebra involving both lumbar and sacral vertebra, which has been reported to affect up to 3.5% of medium to large-sized breed dogs with no relation to sex (Damur-Djuric et al., 2006; Flückiger et al., 2017; Morgan, 1999). German shepherd dogs with lumbosacral transitional vertebra have a longer L.7 in relation to the L.6 vertebra than dogs with normal lumbosacral junctions (Lappalainen et al., 2012).

Two main forms of lumbosacral transitional vertebra have been reported in dogs; sacralization and lumbarization. Sacralization of the last lumbar vertebra (L.7) occurs when the L.7 attached (partially or completely) to the sacrum (Lappalainen et al., 2012), and it has been reported to cause pressure on the nerve of the L.7 root in dogs such as in German Shepherds (Fearnside & Black, 2000). Lumbarization occurs when the S1 vertebra is partially or incompletely attached to its fused sacral component (S2) (Morgan, 1968).

The lumbosacral region is under complex physiological load (Benninger et al., 2004) and the L7-S1 segment is under lower torsional stiffness compared to the rest of the lumbosacral region and this may explain the high range of spine motion at the level of the L7 and S1 vertebrae (Hediger et al., 2009). An increase in the amount of spine flexion and extension in the lumbosacral region has been reported in large dogs (Benninger et al., 2004). Such an increase in the range of the movement might be associated with the large forces applied on the bones. Variation in the anatomy and morphology of the lumbosacral junction or its stability may have clinical significance (Schmid & Lang, 1993).

A recent study by Oheida et al. (2016); on the prevalence of sacrocaudal fusion in greyhounds in Victoria, Australia, reported that 41% of specimens had a sacrum containing four vertebrae.

This study aimed to determine the possible differences in the length of the L.7 and the angle of the lumbosacral junction between greyhounds with standard and those with fused sacra. Also, this study aimed to determine the potential association of sex, body mass and type of fused sacrum (standard and fused) on the morphology of the L.7 and the angulation of the lumbosacral junction.

2 MATERIALS AND METHODS

2.2 Radiographic measurements of L.6, L.7 vertebrae and angle of lumbosacral junction

For the radiograph study relating to L.6, L.7 and LSJ angel, only samples known sex were used (n = 53). Greyhound cadavers were placed in right lateral recumbency, with the femur positioned perpendicular to the spine. For each greyhound, a set of two lateral radiographs of the lumbosacral region was taken using a portable radiographic machine (Atomscope HF80/15 UltraLight, Mikasa X-ray Co Ltd). Each cadaver was exposed to a peak kilovoltage of 80 kVp, an exposure time of 0.18 s and 2.7 milliamp seconds (mAs). The X-ray collimator was attached to a portable stand parallel to the ground at a fixed distance of one metre from the cassette.

All radiographic images were processed by a Veterinary-80 CR Scanner (3D Imaging & Simulations Corp.). Images were stored, analysed and measured with X-ray acquisition software (dicom PACS^® DX-R). Measurements and data recording the lengths of the 6th vertebra (L.6), L.7 and the angle of the lumbosacral junction for each greyhound were performed twice by a researcher and then repeated by another. The lengths of the L.6 and the L.7 were recorded by taking the midcorpus length in mm (half way between the dorsal and ventral border of the vertebral body) as described in other studies (Di Concetto et al., 2012; Lappalainen et al., 2012) (Figure 2). The angle of the lumbosacral junction is the angle formed by two lines bisecting the vertebral body of the L.7 and the sacrum as described in a previous study (Schmid & Lang, 1993) (Figure 3).

2.5 Statistical analyses

Data were analysed using the SPSS statistical package (IBM, SPSS version 23, 2013). Intra-rater and inter-rater reliability (test-retest reliability) were assessed using Lin's concordance correlation coefficient (Lawrence & Lin, 1989; Steichen & Cox, 2002). Concordance correlation coefficient values range from −1 to +1 (+1 indicating strong positive agreement and −1 indicating strong negative agreement). For continuous nominal (categorical) variables, the Kappa statistics test was used and the strength of agreement was evaluated in accordance to Landis and Koch (1977) as flowing; poor agreement if the Kappa value was less than zero, a slight agreement if the Kappa range was between 0.00 and 0.20, fair agreement when Kappa was between 0.21 and 0.40, moderate when Kappa was between 0.41 and 0.60, substantial when Kappa was between 0.61 and 0.80 and almost perfect agreement when Kappa was 0.81 and 1.00.

The descriptive statistics for the measurements of L.6 and L.7 vertebrae and the angle of the lumbosacral junction of greyhounds stratified by sacrum classification are provided (Table 1). Measurement data were plotted as frequency histograms to confirm that it follows normal distribution. The normality of each plotted distribution was assessed using the Shapiro–Wilk test. The Shapiro–Wilk test was used in preference to the Kolmogorov–Smirnov test because the number of measurements in this study was relatively small (n = 53). The equality of the variances for each of the measurements for each of the sacrum types was assessed using Levene's test.

TABLE 1. Descriptive statistics of each of the sacral measurements described in this study, stratified by sacral classification (standard and fused)

Type of Sacrum	Measurement	n	Mean ± SD	Median (Q1, Q3)	Min-Max
Standard	L.7 (mm)	33	29.3 ± 2.04	30.1 (28.0, 31.0)	24.4–32.0
	L.6 (mm)	33	34.6 ± 1.88	34.9 (33.3, 35.8)	30.4– 38.4
	Angel (LSJ) °	33	163.7 ± 5.44	164.15 (158.6, 168.3)	151.9–173.2
Fused	L.7 (mm)	22	30.98 ± 2.16	31.02 (29.29, 32.4)	27.5–34.7
	L6 (mm)	22	35.09 ± 1.70	34.9 (33.7, 36.2)	32.6–39,0
	Angel (LSJ) °	22	166.6 ± 4.13	166.9 (163.1, 170.0)	157.2–173.0
Total	L.7 (mm)	55	30 ± 2.22	30.2 (28.6, 31.4)	24.4–34.7
	L.6 (mm)	55	34.8 ± 1.81	34.9 (33.5, 35.9)	30.4–39.0
	Angel (LSJ) °	55	164.9 ± 5.12	165.5 (162.1, 169.2)	151.9–173.2

• Abbreviations: n, sample size; SD, Standard deviation.

A multiple linear regression analysis was used to quantify the association between the measurements of L.6 and L.7 vertebrae, angle of the LSJ and sacral type (standard and fused) and body mass and sex (as explanatory variables). Multiple linear regressions allowed us to provide estimates of measurements for the two sacral types, adjusting for the confounding effects of body mass and sex. Our linear regression model took the form:

(1)

In Equation (1), L.7 length_i represents the length of the L.7 for the i^th greyhound, β₀ is the intercept term, β₁ is the regression coefficient for sacrum type (a categorical variable comprised of two levels; fused and standard), β₂ is the regression coefficient for sex (a categorical variable comprised of two levels; male and female) and β₃ is the regression coefficient for body mass. Similar linear regression models were developed for the L.6 vertebra length and the angle of LSJ.

Logistic regression analysis was used to quantify the association between the length of the L.7 (as the outcome variable) and body mass and sex (as explanatory variables). This allowed us to estimate the association between sex and the length of the L.7, adjusting for the confounding effect of body mass. Our logistic regression model took the form:

(2)

In Equation (2); logit(p_i) represents the logit of the probability of i^th greyhound having a fused sacrum, β₀ is the intercept term, β₁ is the regression coefficient for sex (a categorical variable comprised of two levels; male and female) and β₂ is the regression coefficient for body mass. In all analyses, a p-value of < .05 is considered statistically significant.

3 RESULTS

3.1 Anatomy of the L.7 vertebra in greyhounds

In the greyhound, the 7th lumbar vertebra consists of a body, spinous process and a pair of costal processes directed cranially. There are caudal and cranial articular processes with mammillary process on the cranial one. (Figure 4). Apart from the length of the body, no other anatomical differences were noticed between the L.7 in greyhounds with standard sacra and those with sacrocaudal fusion.

4 DISCUSSION

Little attention has been paid to the anatomy of the L.7 in greyhounds. Direct visual observations showed that the shape and general anatomy of the L.7 in greyhounds with fused sacra were the same as those with standard sacra (Figure 4) and similar to the description of canine anatomy of the L.7 in the literature (Evans & De Lahunta, 2013b; Goody, 1997).

Investigating the radiographs, we showed a space between the sacrum and first caudal vertebra in greyhounds with standard sacrum, however, this space almost disappeared in those greyhounds with sacrocaudal fusion (Figure 2). In this study, radiographs were used to classify sacra into two types; standard and fused based on the space between S3 and Cd1 vertebrae (Figure 2), and this classification agreed with the classification of sacra after collecting the bones from cadavers and cleaning them, which validates the use of radiographs for sacra classifying by the number of fused vertebrae, (Table 4). In the previous study by Oheida et al. (2016), the correlation between classifying sacra through the radiographs of bones and classifying sacra after collecting them was very high, this supports the findings of this study (Oheida et al., 2016). However, it was not easy to use radiographs to distinguish the different types of fused sacra.

The results of this study showed a significant increase in the length of the L.7 and the angle of the lumbosacral junction in greyhounds with fused sacra comparing to standard sacra. This increase in the length of L.7 might enhance the role of this vertebra in the passive system of force transition between vertebrae (Moens & Runyon, 2002), which may increase the efficiency of force transition within the spine. The length of the L.7 and the angle of the lumbosacral junction were both influenced by the occurrence of sacrocaudal fusion. Supported the hypothesis that there are measurable differences in the length of the L.7 and the angle of the lumbosacral junction between greyhounds with standard and fused sacra (Tables 6 and 8). Also, the length of the L.7 and the angle of the lumbosacral junction were not significantly influenced by sex or body mass of the greyhounds. However, body mass and sex tend to have an association with the length of the L.6 and L.7, and the angle of the lumbosacral junction (Table 8).

Another interesting finding was that the sacrocaudal fusion significantly influenced the length of the L.7 while it also tended to increase the length of the L.6 vertebra but not significantly. This relative difference of the effects of sacrocaudal fusion between L.7 and L.6 vertebrae may be because of the direct articulation between the sacrum and L.7, but, in general, it still shows the potential of sacrocaudal fusion to influence more structures surrounding the sacral region.

This variation in the lumbosacral junction between greyhounds with standard and fused sacra may have a clinical significance. The normal shape of the intervertebral disc of L7-S1 from the lateral view is a triangle (Meij & Bergknut, 2010). The intervertebral disc is the most important structure within the spine that maintains the stability between all vertebrae (Adams & Roughley, 2006; Krismer et al., 1997; Zhao et al., 2005). The increase in the angle of the lumbosacral junction made the end plates of both L.7 and S1 vertebrae to be almost parallel and thus widened the junction dorsally. In fact, the lumbosacral junction is important because it is the site for epidural analgesic drug injections for many clinical and surgical procedures (Di Concetto et al., 2012) and has been selected for this purpose because it is easy to be located (Campoy, 2004; Halley & Riedesel, 1983). For this reason, our result suggests that an increase in the angle of the lumbosacral junction by the occurrence of sacrocaudal fusion may enhance the ease of lumbosacral epidural injections in greyhounds.

The findings of this study may have an impact on the research area of the spine, especially for the sacral region. The increase in the length of the L.7 and the angle of the lumbosacral junction may alter the stability of this segment of the spine, which could lead to many complications. It has been reported that the instability of this spinal segment might shift the load-bearing from the intervertebral disc to other parts of the spine such as ventral aspects of vertebral bodies and the articular facet joints (Meij & Bergknut, 2010) and any abnormal motion within the lumbosacral junction would cause a degeneration of the disc between L.7 and S1 vertebrae (Adams & Roughley, 2006; Kaigle et al., 1995; Tanaka et al., 2001). The findings of this study are very important because it has been reported that the motion (flexion-extension and lateral bending) within the vertebral column is the highest within the lumbosacral junction, followed by the L.5–L.6 segment during the extension and the flexion of the spine (Braund et al., 1977; Bürger & Lang, 1993). The increase in the length of L.6 and L.7 vertebrae, and the angle of the lumbosacral junction in association with sacrocaudal fusion, may alter the extension and the flexion of the spine at this lumbosacral region, which needs to be investigated in the future.

In dogs with certain complications such as degenerative lumbosacral stenosis, the flexion-extension pattern motion of the lumbosacral segment is found to be reduced comparing with healthy dogs (Schmid & Lang, 1993). Also, it has been found that dogs with abnormal lumbosacral junctions have different motion patterns comparing with healthy dogs (Gradner et al., 2007).

The variations in the angle of the lumbosacral junction have been used to differentiate between normal dogs and those with cauda equina syndrome (Schmid & Lang, 1993), therefore, it could be an advantage for greyhounds or a clinical sign of certain complications, and these need to be investigated, too. Comparing the angle of the lumbosacral junction between greyhounds with standard and fused sacra showed a significant increase in those with fused sacra and this means that the V shape of the disc space will open dorsally with the remaining space ventrally open, too, which agrees with the finding of Morgan and Bailey (1990) on their study on German Shepherd dogs, making the end plates of L.7 and S1 vertebrae more parallel. The increase in the angle of the lumbosacral joint changed the shape of the lumbosacral space (Morgan & Bailey, 1990). This variation in the angle may be related to the reduction in the flexibility of the lumbosacral junction in association with fusion. The decrease in the angle of the lumbosacral joint in females comparing with males suggests that those females have a more lordotic posture, as it is also suggested for German Shepherds and Labrador retrievers (Schmid & Lang, 1993). It remains to be investigated whether the increase in the length of L.6 and L.7 vertebrae, and the angle of the lumbosacral joint can be related to the locomotion of greyhounds or specific diseases such as cauda equina syndrome or disc degeneration.

It seems there is a clinical relevance in the increase of the length of L.7 in association with sacrocaudal fusion in greyhounds. In humans, it has been reported that the fusion between L.5 and the sacrum might cause an advanced degeneration in the intervertebral disc (Edwards et al., 2004). The increase in the length of L.7 may decrease the space between the L.7 and the sacrum causing more pressure and thus affects the intervertebral disc, which may lead to similar clinical signs to those of cauda equina syndrome, such as degenerative lumbosacral stenosis. One of the main functions of the intervertebral disc is to absorb force and stabilize the vertebrae (Chan et al., 2014). In fact, degeneration of the intervertebral disc, which is common in canine breeds, classified as chondrodysplasia (Hansen, 1951). A narrowing in the intervertebral discs between the thoracic vertebrae has been noticed in association with the thoracic vertebral fusion in Beagle puppies (Coleman, 1968).

Indeed, some variations in the morphology of the L.7 were expected to be found since it is the first bone articulating within the sacrum cranially and might suppose to reflect the influence of sacrocaudal fusion, if any exists. This increase in the length of L.7 and sacrum (length of the sacrum increased as a result of the fusion between the S3 and Ca1 vertebrae), in association with sacrocaudal fusion, might have clinical implications. An example of these implications is a reduction in the intervertebral space between the sacrum S1 and L.7 vertebrae, as a result of the increase in the sacrum and L.7 length, which might affect the lumbosacral joint function. In addition, fused sacra tend to be stiffer and more concave ventrally (increase in curvature of the pelvic surface), which tilts the sacrum cranially and ventrally. This might also reduce the intervertebral space between the sacrum and the caudal end plate of the L.7.

The results of this study highlight the need of finding the relation between the occurrence of sacrocaudal fusion reduction in the vertebral bodies and cauda equine syndrome. Moreover, the findings of this study suggest that any variations in the sacrum's anatomical features might alter the function of the sacrum and the nearby anatomical structures. Clinical diagnosis should consider the occurrence of sacrocaudal fusion before making a final diagnosis. Finally, the results of this study support our hypothesis that the occurrence of sacrocaudal fusion may favour a change in the morphology of L.7 in greyhounds.

The limitations of our study were the lack of radiographs of the whole vertebral column, the low number of greyhounds used in this study, and the age of the greyhounds under study, however, the results were statistically significant.

5 CONCLUSION

In this study, it has been found that the occurrence of sacrocaudal fusion in greyhounds has significantly increased the length of the L.7 and the angle of the lumbosacral junction. Besides, the length of the L.7 has significantly increased with sacrocaudal fusion, which has been found to increase the length and width of the sacrum, and that might affect the alignment and biomechanical properties of the lumbosacral segment and the intervertebral disc. It can be concluded that any variations in the sacrum's anatomical features may alter the function of the sacrum and the surrounding anatomical structures, such as L.7. Clinical reports and diagnosis should consider the presence or absence of sacrocaudal fusion in greyhounds, before making a final diagnosis of any related clinical complications. Finally, the radiograph of the sacral region has the potential to successfully evaluate and classify the types of sacra into two types; standard or fused.

CONFLICT OF INTEREST

No conflict of interest is declared for this work.