Optimal safe implantation corridors in feline cervical vertebrae (C2–T1): CT study in 16 domestic shorthair cats

Abstract

Objectives

The objective of this study was to define safe corridors for the optimal placement of bicortical implants in the feline cervical spine (C2–T1) using CT.

Methods

CT images of feline cervical spines (n = 16) were reviewed retrospectively. Multiplanar reconstructions were used to define the optimal safe corridors. Safe corridors were defined by their angle of insertion, width and length. The insertion point within the vertebral body was also described. Vertebral measurements were compared between vertebrae using multilevel linear regression, and left and right measurements within vertebrae were compared with the paired samples Wilcoxon signed-rank test. A P value <0.05 was considered significant for all analyses.

Results

The safe corridor insertion points were located within the caudal third of the vertebral body in C2 (mean cranial vertebral ratio 0.73) and in the cranial third of the vertebral bodies from C3–T1 (mean cranial vertebral ratios 0.34–0.38). Mean safe corridor widths ranged from 1.04 mm in C2 to 2.30 mm in C7 and T1. The mean right and left optimal angles of implantation were, respectively, 21.79° and 21.49° for C2, 45.26° and 46.19° for C3, 51.48° and 51.04° for C4, 53.52° and 54.30° for C5, 56.36° and 56.65° for C6, 63.40° and 64.92° for C7, and 53.90° and 52.90° for T1. There were statistically significant differences between vertebrae in almost every measurement.

Conclusions and relevance

Cervical vertebral safe corridors in cats are narrow and differ to those reported in dogs. Safe corridors are located in the caudal third of C2 and cranial third of the C3–T1 vertebral bodies. Current recommendations for implant sizes should be reviewed, as 1.5–2 mm implants would be oversized for bicortical implantation in most of the feline cervical vertebrae.

Introduction

Spinal stabilisation is indicated for the treatment of vertebral column disorders such as traumatic spinal fractures/luxations,^1–5 and vertebral instability following spinal decompressive surgery.^6,7

Vertebral fractures are considered common in cats,⁸ and are typically a result of external trauma.⁹ A recent study, reviewing the incidence of spinal cord injury in 92 cats, reported that 24–32% of the cases involved the cervical spine, with 8% of cases having a traumatic aetiology.¹⁰ It has been reported that 2–2.9% of feline vertebral fractures involve the cervical spine.¹¹

Intervertebral disc disease (IVDD) is relatively uncommon in cats,¹² with cervical IVDD accounting for 6–25% of all IVDD feline cases.^13,14 Cervical spinal cord decompression through a ventral slot is more challenging in cats than in dogs,¹² and it has been suggested that the feline spinal column may require stabilisation following a ventral slot procedure.¹⁵

Multiple techniques have been described for spinal stabilisation in dogs and, to a lesser degree, in cats.^{2,3,7,16–23} The use of pins/screws and polymethylmethacrylate (PMMA) remains a common treatment for vertebral body stabilisation of all regions of the vertebral column in dogs of all sizes owing to its efficacy and versatility in spinal stabilisation.^{2,19,20,24–30} In cats, 1.5–2 mm diameter pins/screws have been suggested when using a pin/screw–PMMA construct.¹⁵

Precise insertion of bicortical implants into the vertebral bodies is required to avoid penetrating either the spinal canal or the transverse foramina and injuring neurovascular structures associated with the vertebral column. Safe corridors for implant positioning have been described in the canine cervical spine.²⁴ However, information regarding the safe corridors for implant placement in the feline cervical spine is lacking.

The aim of this study was to use CT images to define safe corridors for bicortical implant insertion in the feline cervical spine (C2–T1).

Materials and methods

The data for this study were retrieved retrospectively after ethical approval from the authors’ institutional research ethics committee was obtained. The electronic records of client-owned animals were searched to identify cats that had been presented between 2013 and 2016. Inclusion criteria were skeletally mature domestic shorthair breed cats, which had undergone a cervical spine CT scan as part of their diagnostic investigations. Cats were excluded from the study when the CT acquisitions did not include a sharp bone algorithm. The retrieved data from the medical records included breed, age, sex and weight of the cats.

All acquisitions were made on a multi-detector (80-slice) CT scanner (Aquilion Prime; Toshiba Medical Systems) and the studies were retrieved from picture archiving and communication system (Visbion PACS; Visbion). Multiplanar reformatting of the cervical spines was performed in an open-source DICOM viewer (Horos 64-bit; https://horosoproject.org) using sharp bone algorithm reconstructions and displayed in a bone window (window level 700/window width 4000/WL). All measurements were made on the sagittal and transverse vertebral planes from the second cervical (C2) to the first thoracic vertebrae (T1). The orientation of the multiplanar reconstructions were standardised so that in the transverse plane the x-axis was parallel to the floor of the vertebral canal and the y-axis was parallel to the spinous process, while in the sagittal plane the z-axis was parallel with the floor of the vertebral canal. For each vertebra the transverse images were assessed by scrolling from cranial to caudal, and the image in which the distance between the vertebral canal and transverse foramina was greatest was selected to perform all measurements. The ideal implant position was calculated based on previous studies in dogs,²⁴ and was represented as line P. Line P allowed maximum implant purchase in the bone and bisected the distance between the spinal canal and the transverse foramina. Measurements were made assuming a ventral surgical approach as described in previous studies.²⁴

The following parameters were determined (Figures 1 –4):

– Safe corridor length (L): distance between the implant insertion point and the vertebral pedicle (along line P) in the transverse plane.

– Safe corridor width (W): distance between the spinal canal and the transverse foramina measured perpendicular to line P. Owing to lack of transverse foramina in vertebrae C7 and T1, the W was defined as the distance between the spinal canal and the origin of the transverse process.

– Optimal angle of insertion (α): angle between line P and the sagittal plane.

– Craniocaudal distance/ratio to the insertion point: distances from the caudal (CdIP) and cranial (CrIP) endplates to the insertion point, calculated in the sagittal plane. Measurements were recorded in millimetres and as a vertebral body length ratio.

– Mediolateral distance insertion point (MLIP): the mediolateral location for implant insertion was measured on the transverse image following previous guidelines in dogs.²⁴ In brief, for C3–C6 and T1 this was ventral midline. For C2 and C7 this was defined at the distance between vertebral body midline and the intersection of line P with the ventral vertebral cortex.

– Midsagittal vertebral body depth (msVBD): distance between the spinal canal and the ventral aspect of the vertebral body in the midsagittal plane. This measurement corresponded with the greatest vertebral body depth.

– Parasagittal vertebral body depth (psVBD): distance between the spinal canal and the ventral aspect of the vertebral body in the parasagittal plane. This measurement corresponded to the smallest vertebral body depth.

Figure 1

Optimal safe implantation corridors for C2 vertebra. P represents the implant (screw/pin). In the transverse diagram abbreviations are defined as follows: safe corridor length (L), safe corridor width (W), optimal insertion angle (α), mediolateral distance to the implant insertion point (MLIP), midsagittal vertebral body depth (msVBD) and parasagittal vertebral body depth (psVBD). In the sagittal view diagram abbreviations are as follows: craniocaudal distance to the implant insertion point (CrIP), implant insertion point (IP) and caudal distance to the implant insertion point (CdIP)

Figure 2

Optimal safe implantation corridors for C3–C6 vertebrae. P represents the implant (screw/pin). In the transverse diagram abbreviations are defined as follows: safe corridor length (L), safe corridor width (W), optimal insertion angle (α), midsagittal vertebral body depth (msVBD) and parasagittal vertebral body depth (psVBD). In the sagittal view diagram abbreviations are as follows: craniocaudal distance to the implant insertion point (CrIP), implant insertion point (IP) and caudal distance to the implant insertion point (CdIP)

Figure 3

Optimal safe implantation corridors for C7 vertebra. P represents the implant (screw/pin). In the transverse diagram abbreviations are defined as follows: safe corridor length (L), safe corridor width (W), optimal insertion angle (α), mediolateral distance to the implant insertion point (MLIP), midsagittal vertebral body depth (msVBD) and parasagittal vertebral body depth (psVBD). In the sagittal view diagram abbreviations are as follows: craniocaudal distance to the implant insertion point (CrIP), implant insertion point (IP) and caudal distance to the implant insertion point (CdIP)

Figure 4

Optimal safe implantation corridors for T1 vertebra. P represents the implant (screw/pin). In the transverse diagram abbreviations are defined as follows: safe corridor length (L), safe corridor width (W), optimal insertion angle (α), midsagittal vertebral body depth (msVBD) and parasagittal vertebral body depth (psVBD). In the sagittal view diagram abbreviations are as follows: craniocaudal distance to the implant insertion point (CrIP), implant insertion point (IP) and caudal distance to the implant insertion point (CdIP)

Statistical analysis

All statistical analyses were performed with the computer programs SPSS 22.0 (IBM) and MLwiN (Version 2.20; Centre for Multilevel Modelling, University of Bristol, UK). Dependent (outcome) variables were the vertebral parameters recorded: msVBD (mm), psVBD (mm), MLIP (mm), right and left W (mm), right and left L (mm), right and left α (mm), CrIP (mm and expressed as ratio), CdIP (mm and expressed as a ratio). Independent variables assessed included those related to the cat (weight, sex, age), and the vertebral body on which measurements were made (C2, C3, C4, C5, C6, C7 or T1), with C4 selected as the reference vertebra.

Descriptive statistics were calculated as appropriate; continuous data were summarised as median values with interquartile ranges (IQR), and categorical data as frequencies with 95% confidence intervals (95% CI). For continuous variables (age, weight), graphical assessment and a test for departure from linear trend was applied to determine whether analyses could be undertaken with an assumption of linear association. Normality of distribution for continuous variables was assessed graphically and using the Kolmogorov–Smirnov test.

As left and right vertebral parameters were measured and these measurements were made on up to eight vertebrae per cat, the resultant outcomes are not independent and so traditional analytical techniques are not considered appropriate.³¹ Vertebral measurements (level one units) were clustered within vertebrae (level two units) and these, in turn, clustered within cats (level three units) and to account for this, factors affecting the parameters were examined using multilevel linear regression models, with cat and vertebral body included as random intercept terms to account for clustering. Differences between the left and right parameters within each vertebra were assessed with the paired-samples Wilcoxon signed-rank test. P <0.05 was considered significant for all analyses.

Results

A total of 16 cats fulfilled the inclusion criteria, including five male neutered, one male entire and 10 female neutered. The median age of our population was 119.6 months (IQR 63–164.5 months) and the median weight was 4.0 kg (IQR 3.5–4.8 kg). Measurements from all seven vertebrae (C2–T1) were recorded from 14 cats; vertebrae C2, C3 and C4 were excluded in one cat owing to severe vertebral spondylosis, and T1 was excluded in another cat due to an incomplete imaging study. Results for all parameters are summarised in Tables 1 and 2.

Table 1

Mean values and interquartile range for vertebral parameters measured in the cervical vertebrae (C2–T1) of 16 cats

	L (mm)		W (mm)		α (º)
	Right	Left	Right	Left	Right	Left
C2	5.70* (4.18–7.51)	6.09* (4.48–7.43)	1.16* (0.67–1.62)	1.04* (0.71–1.36)	21.79 (12.94–32.44)	21.49 (12.09–38.74)
C3	9.09 (8.10–11.72)	9.01 (7.52–12.61)	1.21 (0.90–1.73)	1.22 (0.92–1.78)	45.26 (41.00–48.64)	46.19 (39.10–53.93)
C4	8.47 (7.71–10.24)	8.56 (7.40–9.71)	1.23* (0.80–1.59)	1.33* (0.96–1.95)	51.48 (47.63–56.54)	51.04 (45.78–57.94)
C5	8.78 (7.68–11.79)	8.79 (7.53–11.22)	1.46 (0.95–2.09)	1.44 (0.82–2.09)	53.52 (47.54–58.64)	54.30 (48.26–59.39)
C6	8.54* (6.33–11.17)	8.90* (7.15–11.00)	1.65 (0.91–2.03)	1.73 (1.30–2.29)	56.36 (49.58–67.03)	56.65 (49.22–61.91)
C7	8.88 (8.33–9.68)	9.05 (8.35–9.57)	2.03* (1.19–2.85)	2.30* (1.80–2.56)	63.40 (53.75–69.56)	64.92 (52.59–70.67)
T1	9.10* (7.86–9.82)	9.40* (8.61–10.10)	2.30 (1.80–3.14)	2.20 (1.91–2.90)	53.90* (50.16–58.83)	52.90* (48.77–58.19)

P <0.05

L = safe corridor length; W = safe corridor width; α = optimal insertion angle

Table 2

Mean values and interquartile range for vertebral parameters measured in the cervical vertebrae (C2–T1) of 16 cats

	VBL(mm)	CdIP(mm)	CrIP(mm)	CrIP(vertebral body ratio)	MLIP (mm)		msVBD(mm)	psVBD(mm)
	VBL(mm)	CdIP(mm)	CrIP(mm)	CrIP(vertebral body ratio)	Right	Left	msVBD(mm)	psVBD(mm)
C2	18.43 (15.75–21.4)	5.02 (3.43–6.17)	13.41 (12.30–15.68)	0.73 (0.62–0.78)	2.96* (2.02–3.47)	3.21* (2.60–3.99)	4.7 (3.77–5.34)	1.89 (1.60–2.54)
C3	11.67 (9.47–13.27)	7.51 (5.33–8.88)	4.16 (3.40–4.90)	0.36 (0.28–0.43)	0	0	4.07 (3.22–5.16)	2.60 (1.71–2.91)
C4	11.01 (9.44–12.08)	7.34 (5.33–9.64)	3.67 (3.11–4.72)	0.34 (0.20–0.43)	0	0	3.93 (3.34–4.58)	2.60 (2.00–3.22)
C5	10.28 (7.25–11.27)	6.67 (5.47–7.64)	3.61 (2.8–4.65)	0.35 (0.26–0.46)	0	0	3.75 (2.90–4.49)	2.60 (1.92–3.07)
C6	9.68 (8.57–11.03)	5.93 (3.97–7.66)	3.75 (2.53–4.73)	0.38 (0.29–0.61)	0	0	3.86 (3.03–4.49)	2.78 (2.16–3.34)
C7	9.72 (8.47–10.83)	5.98 (4.98–6.78)	3.74 (2.70–4.45)	0.38 (0.31–0.43)	1.34 (0.70–1.94)	1.47 (0.8–2.06)	4.45 (3.03–5.04)	2.45 (1.86–3.45)
T1	9.54 (8.68–10.76)	6.09 (4.31–6.92)	3.41 (2.66–4.08)	0.36 (0.30–0.40)	0	0	4.29 (3.30–4.77)	2.67 (1.93–3.31)

P <0.05

VBL = vertebral body length; CdIP = caudal distance to the insertion point; CrIP = cranial ratio to the insertion point; MLIP = mediolateral distance to the insertion point; msVBD = midsagittal vertebral body depth; psVBD = parasagittal vertebral body depth

Our results showed that the implant insertion point (IP) was similar between vertebrae C3 and T1 (Table 1). The IP in C2 was located in the caudal third (Figure 1), having a distance ratio of 0.73 cranially, being located, on average, 13.41 mm caudal to the atlantoaxial joint. The implant IP was located in the cranial third of the vertebral body for C3–T1 (Figures 2 –4). The average CrIP was 3.73 mm, with an average ratio of 0.36. These distances were statistically significant when comparing all vertebrae with the reference vertebra (C4). There was a significant association between the weight of the patient and distance to the caudal vertebral endplate (CdCIP, P = 0.046).

The mean MLIP on C2 was 2.96 mm on the right and 3.21 mm on the left. The mean distance to the midline in C7 was 1.34 mm on the right and 1.47 mm on the left. The difference between the right and left side was statistically significant for C2 but not for C7. The MLIP for vertebrae C3–C6 and T1 was 0 mm.

The W increased progressively in a caudal direction (Table 1), with an overall average of 1.57 mm in the right side and 1.61 mm in the left side of the vertebrae. The mean W of C2 was 1.16 mm on the right and 1.04 mm on the left, and increased to 2.3 mm on the right and 2.2 mm on the left on T1. The differences between W when compared with a named reference vertebra (C4) were statistically significant for all vertebrae with the exception of C3. The difference between the right and left side in each vertebra was statistically significant for C2, C4 and C7.

The L was similar between C3 and T1 vertebrae (Table 1), ranging from an average of 8.44 mm on the right to 9.39 mm on the left side. The L for C2 was 5.7 mm on the right and 6.09 mm on the left. The differences in L were statistically significant when comparing C2, C3, C7 and T1 to vertebra C4. The difference between the right and left side in each vertebra was statistically significant for C2, C6 and T1.

The α increased more caudally within the cervical spine (Table 1). While the mean C2 α was 21.79° on the right and 21.49^° on the left, the optimal angles of insertion increased progressively to 63.4° on the right and 64.92° on the left side in C7. Interestingly, the α were again reduced in T1 vertebra, being 53.9° on the right and 52.9° on the left side. The differences between α were statistically significant when comparing the vertebrae with the reference vertebra (C4), with the exception of T1. The difference between the right and left side was statistically significant only for T1.

Finally, results from the vertebral body depth in the cervical spine showed minor differences among the vertebrae except for C2 (Table 2). The average sVBD and psVBD for C2 were 4.70 mm and 1.89 mm, respectively. The average msVBD and psVBD depth of C3–T1 were 4.06 mm and 2.61 mm, respectively. The difference in msVBD between vertebrae C2, C7, T1 and our reference vertebra was statistically significant. The difference in psVBD between vertebrae C2, C6 and our reference vertebra was also statistically significant. There was a significant association between the age of the cat and vertebral body depth values (P = 0.003 for the msVBD and P = 0.023 for the psVBD).

Discussion

Specific recommendations for feline cervical surgical stabilisation are lacking, both in terms of implant positioning or selection of appropriate techniques. Historically, it has been recommended to use the same techniques as used in dogs.¹⁵ While stabilisation with pins/screws–PMMA is a commonly used construct, information on the challenges involved in the procedure in cats is lacking. Our results show that the safe corridor width (W) is very narrow in cats and of similar dimensions from C3–T1 vertebrae. As expected, owing to different anatomy, the values obtained for C2 differed from the values obtained in the rest of the vertebrae. To our knowledge, this is the first study to define the safe corridors for implant placement in the feline cervical column.

Our results show that in order to achieve maximum bone purchase and minimise damage to neurovascular structures, the IPs from C3–T1 should be located in the cranial third of the vertebral body (0.36 mean vertebral body ratio when calculated from the cranial endplate). This information differs from a previous report in the canine cervical vertebrae, where the safe corridors were located in the mid-point along the length of the vertebral body.²⁴ Regarding the positioning of the implants in the transverse plane, the authors followed the guidelines published for canine cervical vertebrae,^24,32 and selected the midline of the vertebral body in C3–C6 and T1. We acknowledge that other IPs are possible in the vertebral bodies, which would result in different angles of insertion of the implants. However, when selecting other IP positions, the direction of the implant may not bisect W, may have a narrower safe corridor and/or may reduce the bone purchase. Thus, the IP in C2 was modified accordingly (Figure 5). The IP for C2 was located at an average of 2.96 mm to the right from midline (for a right-side-orientated implant) and 3.21 mm to the left (for a left-side-orientated implant). Regarding C7, the IP was located at an average of 1.34 mm to the left from midline (for a right-side-orientated implant) and 1.47 mm to the right (for a left-side-orientated implant).

Figure 5

(a,b) Transverse CT images of two feline C2 vertebrae showing the difference in implant position when the insertion point is located in the vertebral body midline (implant indicated by P) and when the insertion point is located at the recommended position (implant indicated by P’). Note how both implants, P and P’, bisect the safe corridor width (W). However, P would remove part of the cortical bone of the ventral aspect of the vertebral body in (a) (achieving less bone purchase), while it would be in contact with the spinal canal cortical bone in (b) (greater risk of neurovascular damage)

Safe corridors are characterised by their length, width and angle of insertion. Our study showed that these three parameters were significantly different between C2 and the rest of the vertebrae (C3–T1). The axis had a mean L of 5.9 mm, mean W of 1.1 mm and mean α of 21.64°.

The L was very similar between C3 and T1 vertebrae, with an average of 8.8 mm on the right and 8.94 mm on the left side of the vertebrae. Regarding the W and α, there was a gradual increase in the values as we progressed caudally along the spinal column. The W increased progressively from C3–C6, from a mean of 1.2 mm in C3 to a mean of 1.7 mm in C6, whereas in vertebrae C7 and T1 these were 2.17 mm and 2.24 mm, respectively. The increase in the W of C7 and T1 is due to the lack of a vertebral transverse foramina, which enables the safe corridor to extend from spinal canal to the origin of transverse process. The use of 1.5–2 mm screws/pins has been recommended when stabilising feline cervical vertebrae.¹⁵ However, our study has shown that 2.0 mm diameter implants would be oversized in every C2, C3 and C4 vertebrae, in >87% (n = 28/32) of C5, in 75% (n = 24/32) of C6, in >40% (n = 13/32) of C7 and in >43% (n = 13/30) of the T1 studied vertebrae. Using 1.5 mm implants would be oversized in every C2 vertebrae, in >96% (n = 29/30) of C3, in 78% (n = 21/28) of C4, in >68% (n = 22/32) of C5, in >32% (n = 10/32) of C6, in >3% (n = 1/32) of C7 and in 6.6% (n = 2/30) of T1 vertebrae.

The α is arguably the most important decision to make during surgery to avoid damage to neurovascular structures. The α varied between each vertebra, with a gradual increase from C3–C6. The measurements for the axis differ from the other vertebrae, with mean angles of insertion of 21.79° on the right and 21.49° on the left side of the vertebrae. This result is related to the significant anatomical differences between the axis and the remaining cervical vertebrae. The mean angle of insertion for C3–T1 was 54.00° on the right and 54.40° on the left side. There was a progressive increase in the α between C3 and C6 vertebra – from an average of 45.7^° for C3 to 56.5° in C6. The α at C7 vertebra was greater than the rest of the vertebrae, with an average of 64.2°, which the authors believe is due to an anatomical difference (lack of transverse foramina) and its consequence in selecting the ideal implant direction and starting point. Overall, these results reflect the difference in conformation of the feline vertebrae when compared to with dogs. Therefore, following the previously reported canine guidelines²⁴ (ranging from a mean of 37.5° in C3 to 47.5° in C7) could lead to perforation of the transverse foramen/spinal canal and poor bone purchase.

In recent years there has been an increase in the use of locking plates, having been described to stabilise the cervical column of dogs and cats, using monocortical screws.¹⁷ This technique is designed to avoid laceration of the neurovascular structures associated with the vertebra. The feline cervical vertebrae do not offer much bone purchase,¹⁵ and biomechanical studies assessing the stability of monocortical constructs in the feline cervical spine are currently lacking. The msVBD and psVBD were calculated in order to quantify the vertebral body depth available when opting for monocortical screws/locking plate systems to stabilise the cervical spine. Our study has shown that the maximum depth of the vertebral body is, on average, 4 mm in the midline and 2.6 mm in a parasagittal position (where locking plates are usually placed), indicating that bone stock availability for monocortical implant purchase is very limited. It should also be noted that when locking plates and monocortical screws have been used in the feline cervical spine,¹⁷ it was not possible to prevent screw penetration into the spinal canal as even the shortest screws were too long for the feline vertebral body.

We acknowledge several limitations to this study. Our study consisted of only domestic shorthair cats, and therefore our results might not be applicable to other breeds. The study consisted of 16 cats; this not only limits the statistical power to identify significant effects, but also reduces the confidence in the values determined for the recorded measurements. The study only included one male cat and therefore it is possible that variation between the sexes may have gone undetected. Finally, the authors would like to emphasise that this study is not meant to replace individual preoperative planning. Vertebral anatomical variations may be of clinical significance, as seen in one of our cases (Figure 6), and even minor differences between individuals could be clinically significant.

Figure 6

Transverse CT image of a feline C5 vertebra showing an anatomical variation. There is no safe implantation corridor in the left side of the vertebra, and placing a bicortical implant would result in penetration into the spinal canal or the transverse foramina

Conclusions

We have defined the optimal safe corridors in the feline cervical (C2–T1) vertebrae by the use of multiplanar CT images. The implant insertion point is located in the cranial third of the vertebral body in C3–T1 and caudal third in the axis. Our study shows that safe corridors in the feline cervical spine are very narrow, and current recommendations for bicortical implants are oversized for most feline cervical vertebrae.

Footnotes

Accepted: 12 January 2018

Conflict of interest

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

References

Voss

Montavon

. Tension band stabilization of fractures and luxations of the thoracolumbar vertebrae in dogs and cats: 38 cases (1993–2002). J Am Vet Med Assoc 2004; 225: 78–83.

Sharp

NJH

Wheeler

. Trauma. In: Small animal spinal disorders. 2nd ed. Philadelphia: Elsevier Mosby, 2005, pp 281–318.

Seim

. Surgery of the cervical spine. In: Fossum

(ed). Small animal surgery. 3rd ed. St Louis, MO: Mosby, 2007, pp 1402–1459.

Weh

Kraus

. Use of a four pin and methylmethacrylate fixation in L7 and the iliac body to stabilize lumbosacral fracture-luxations: a clinical and anatomic study. Vet Surg 2007; 36: 775–782.

Eminaga

Palus

Cherubini

. Acute spinal cord injury in the cat: causes, treatment and prognosis. J Feline Med Surg 2011; 13: 850–862.

McKee

Downes

. Vertebral stabilization and selective decompression for the management of triple thoracolumbar disc protrusions. J Small Anim Pract 2008; 49: 536–539.

Danielski

Vanhaesebrouck

Yeadon

. Ventral stabilization and facectomy in a Great Dane with wobbler syndrome due to cervical spinal canal stenosis. Vet Comp Orthop Traumatol 2012; 4: 337–341.

Bali

. Diagnosis and surgical management of a fractured atlas in a cat. J Feline Med Surg 2011; 13: 280–282.

Cabassu

. Fractures of the spine. In: AO principles of fracture management in the dog and cat. Edinburgh: Thieme, 2005, pp 130–150.

10.

Goncalves

Platt

Llabre-Diaz

, et al. Clinical and magnetic resonance imaging findings in 92 cats with clinical signs of SC disease. J Feline Med Surg 2009; 11: 53–59.

11.

Besalti

Ozak

Tong

. Mangement of spinal trauma in 69 cats. Dtsch Tieraztl Wochenschr 2002; 109: 315–320.

12.

Steffen

Grunenfelder . Diseases of the spine and nervous system. In: Feline orthopedic surgery and muskuloskeletal disease. Edinburgh: Elsevier, 2013, pp 75–81.

13.

Farrell

Fitzpatrick

. Feline intervertebral disc disease. In: Advances in intervertebral disc disease in dogs and cats book. Ames, IA: Wiley Blackwell, 2015, pp 36–50.

14.

Marioni-Henry

Vite

Newton

, et al. Prevalence of diseases of the spinal cord of cats. J Vet Intern Med 2004; 18: 851–858.

15.

Voss

Montavon

. The spine. In: Feline orthopedic surgery and musculoskeletal disease. Edinburgh: Elsevier, 2009, pp 407–422.

16.

Voss

Montavon

. Tension band stabilization of fractures and luxations of the thoracolumbar vertebrae in dogs and cats: 38 cases (1993–2002). J Am Vet Med Assoc 2004; 225: 78–83.

17.

Voss

Steffen

Montavon

. Use of the compact UniLock System for ventral stabilization procedures of the cervical spine. Vet Comp Orthop Traumatol 2006; 19: 21–28.

18.

Wheeler

Lewis

Cross

, et al. Closed fluoroscopic-assisted spinal arch external skeletal fixation for the stabilization of vertebral column injuries in five dogs. Vet Surg 2007; 36: 442–448.

19.

Bruce

Barisson

Gyselinck

. Spinal fracture and luxation if dogs and cats. Vet Comp Orthop Traumatol 2008; 21: 280–284.

20.

Jeffery

. Vertebral fracture and luxation in small animals. Vet Clin North Am Small Anim Pract 2010; 40: 809–828.

21.

Agnello

Kapatkin

Garcia

, et al. Intervertebral biomechanics of locking compression plate monocortical fixation of the canine cervical spine. Vet Surg 2010; 39: 991–1000.

22.

Wilson

. Repair of sacral fractures using pins and polymethylmethacrylate (six cases). Aust Vet J 2015; 93: 311–318.

23.

Hettlich

Fosgate

Litsky

. Biomechanical comparison of 2 vertebral locking plates to monocortical screw/polymethylmethacrylate fixation in canine cadaveric cervical vertebral column. Vet Surg 2017; 46: 95–102.

24.

Watine

Cabassu

Catheland

, et al. Computed tomography study of implantation corridors in canine vertebrae. J Small Anim Pract 2006; 47: 651–657.

25.

Garcia

Milthorpe

Russell

, et al. Biomechanical study of canine spinal fracture fixation using pins or bone screws with polymethylmethacrylate. Vet Surg 1994; 23: 322–329.

26.

Schulz

Waldron

Fahie

. Application of ventral pins and polymethylmethacrylate for the management of atlantoaxial instability: results in nine dogs. Vet Surg 1997; 26: 317–325.

27.

Hawthorne

Blevins

Wallace

, et al. Cervical vertebral fractures in 56 dogs: a retrospective study. J Am Anim Hosp Assoc 1999; 35: 135–146.

28.

Wilson

. Repair of sacral fractures using pins and polymethylmethacrylate (six cases). Aust Vet J 2015; 93: 311–318.

29.

Sturges

Kapatkin

Garcia

, et al. Biomechanical comparison of locking compression plate versus positive profile pins and polymethylmethacrylate for stabilization of the canine lumbar vertebrae. Vet Surg 2016; 45: 309–318.

30.

Nei

Kat

Coetzee

, et al. Biomechanical comparison between pins and polymethylmethacrylate and the SOP locking plate system to stabilize canine lumbosacral fracture-luxation in flexion and extension. Vet Surg 2017; 46: 789–796.

31.

Sutton

Muir

Jones

. Two knees or one person: data analysis strategies for paired joints or organs. Ann Rheum Dis 1997; 56: 401–402.

32.

Wong

Emms

. Use of pins and methylmethacrylate in stabilization of spinal fractures and luxations. J Small Anim Pract 1992; 33: 415–422.