ArticlePDF Available

Evaluation of a Novel Dorsal-Cemented Technique for Atlantoaxial Stabilisation in 12 Dogs

Authors:

Abstract and Figures

Dorsal atlantoaxial stabilisation (DAAS) has mostly been described to treat atlantoaxial instability using low stiffness constructs in dogs. The aim of this study was to assess the feasibility and surgical outcome of a rigid cemented DAAS technique using bone corridors that have not previously been reported. The medical records of 12 consecutive dogs treated with DAAS were retrospectively reviewed. The method involved bi-cortical screws placed in at least four of eight available bone corridors, embedded in polymethylmethacrylate. Screw placement was graded according to their position and the degree of the breach from the intended bone corridor. All DAAS procedures were completed successfully. A total of 72 atlantoaxial screws were placed: of those, 51 (70.8%) were optimal, 17 (23.6%) were suboptimal, and 4 (5.6%) were graded as hazardous (including 2 minor breaches of the vertebral canal). Surgical outcome was assessed via a review of client questionnaires, neurological examination, and postoperative CT images. The clinical outcome was considered good to excellent in all but one case that displayed episodic discomfort despite the appropriate atlantoaxial reduction. A single construct failure was identified despite a positive clinical outcome. This study suggests the proposed DAAS is a viable alternative to ventral techniques. Prospective studies are required to accurately compare the complication and success rate of both approaches.
Content may be subject to copyright.
life
Article
Evaluation of a Novel Dorsal-Cemented Technique for
Atlantoaxial Stabilisation in 12 Dogs
Joana Tabanez 1, Rodrigo Gutierrez-Quintana 2, Adriana Kaczmarska 2, Roberto José-López 2,
Veronica Gonzalo Nadal 2, Carina Rotter 1and Guillaume Leblond 1, 2, *


Citation: Tabanez, J.;
Gutierrez-Quintana, R.; Kaczmarska,
A.; José-López, R.; Nadal, V.G.; Rotter,
C.; Leblond, G. Evaluation of a Novel
Dorsal-Cemented Technique for
Atlantoaxial Stabilisation in 12 Dogs.
Life 2021,11, 1039. https://doi.org/
10.3390/life11101039
Academic Editor: Albano Beja Pereira
Received: 24 August 2021
Accepted: 28 September 2021
Published: 2 October 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
1Fitzpatrick Referrals Orthopaedic and Neurology, Surrey GU7 2QQ, UK;
Joanat@fitzpatrickreferrals.co.uk (J.T.); Crotter@fitzpatrickreferrals.co.uk (C.R.)
2Small Animal Hospital, School of Veterinary Medicine, University of Glasgow, Bearsden G61 1QH, UK;
Rodrigo.GutierrezQuintana@glasgow.ac.uk (R.G.-Q.); A.kaczmarska.1@research.gla.ac.uk (A.K.);
roberto.joselopez.bcn@gmail.com (R.J.-L.); V.gonzalo-nadal.1@research.gla.ac.uk (V.G.N.)
*Correspondence: Guillaume.Leblond@ndsr.co.uk
Abstract:
Dorsal atlantoaxial stabilisation (DAAS) has mostly been described to treat atlantoaxial
instability using low stiffness constructs in dogs. The aim of this study was to assess the feasibility and
surgical outcome of a rigid cemented DAAS technique using bone corridors that have not previously
been reported. The medical records of 12 consecutive dogs treated with DAAS were retrospectively
reviewed. The method involved bi-cortical screws placed in at least four of eight available bone
corridors, embedded in polymethylmethacrylate. Screw placement was graded according to their
position and the degree of the breach from the intended bone corridor. All DAAS procedures were
completed successfully. A total of 72 atlantoaxial screws were placed: of those, 51 (70.8%) were
optimal, 17 (23.6%) were suboptimal, and 4 (5.6%) were graded as hazardous (including 2 minor
breaches of the vertebral canal). Surgical outcome was assessed via a review of client questionnaires,
neurological examination, and postoperative CT images. The clinical outcome was considered good
to excellent in all but one case that displayed episodic discomfort despite the appropriate atlantoaxial
reduction. A single construct failure was identified despite a positive clinical outcome. This study
suggests the proposed DAAS is a viable alternative to ventral techniques. Prospective studies are
required to accurately compare the complication and success rate of both approaches.
Keywords:
canine; spinal disorders; veterinary neurosurgery; atlantoaxial instability; craniocervical
junction anomalies; rigid dorsal stabilisation; presurgical planning
1. Introduction
Atlantoaxial instability (AAI) was first reported in dogs more than 50 years ago [
1
,
2
].
Geary et al. (1967) reported AAI in ten dogs of toy or miniature breeds; four of these
dogs were managed surgically with a dorsal stabilisation technique using a simple wire
loop [
2
]. Since then, various treatment options have been reported with ventral techniques
becoming more popular over time, likely due to lower reported mortality rates [3,4].
AAI can occur subsequent to congenital, developmental, and/or traumatic causes [
5
,
6
].
Often minor trauma in dogs with pre-existing congenital anomalies can lead to sublux-
ation [
5
]. Odontoid process malformation (aplasia or hypoplasia) is the most frequently
reported cause of AAI in dogs [
4
]. Other associated atlantoaxial congenital anomalies
include incomplete ossification of the atlas (C1), separation of the dens from the axis (C2),
and insufficient ligamentous support [
5
8
]. These malformations are more commonly
encountered in young toy breed dogs [
4
6
]. Regardless of the underlying aetiology, the
dorsal displacement of C2 into the vertebral canal leads to neurological deficits and/or
pain [5].
Various craniocervical junction anomalies can often complicate AAI cases, especially
in young small and toy breed dogs [
9
,
10
]. These malformations include atlanto-occipital
Life 2021,11, 1039. https://doi.org/10.3390/life11101039 https://www.mdpi.com/journal/life
Life 2021,11, 1039 2 of 15
overlapping, C2 dens dysplasia, caudal occipital malformations, craniocervical junction
dorsal fibrous bands, atlantoaxial incongruence, and occipito-atlantoaxial malformations
(OAAM) [
9
11
]. The latter includes occipito-atlantal fusion (often unilateral), hypoplasia
of C1 and/or C2 dens, various other C2 malformations and C1–C2 joint dysplasia with
frequent features of AAI [
12
14
]. Atlantoaxial incongruence occurs when the size of C1 is
disproportionate to that of C2 [
15
]. Recognising the complexity of these malformations is
crucial to formulate an appropriate treatment plan and prognosis. It can be argued that
complex craniocervical junction anomalies including those of C1–C2 incongruency and
OAAM are more easily accessible via a dorsal approach [7,15,16].
Many surgical stabilisation techniques and conservative methods for the management
of AAI have been reported [
3
,
4
]. Conservative management involving the use of cervical
splints or bandages is often reserved for cases with subtle clinical signs, particularly small
or skeletally immature dogs or where there are financial limitations [
3
,
17
]. Canine dorsal
atlantoaxial stabilisation (DAAS) has mainly been described using low to moderate stiffness
constructs such as orthopaedic wire, nonmetallic sutures, nuchal ligament, and Kirschner
wires maintained with polymethylmethacrylate (PMMA) cement or a metallic tension
band [
5
,
18
21
]. When compared with ventral stabilisation techniques, DAAS has been
associated with higher mortality rates [
3
,
4
]. However, studies that report the outcome in
dorsal techniques predominantly originate from several decades ago and most of them
describe techniques that required penetration of the epidural space at the level of the C1
dorsal arch [
2
,
18
]. A modified ventral approach with either threaded pins or cortical screws
embedded in cement has become the most reported technique achieving reasonably high
success rates [
3
,
22
]. Reported success rates range from 50 to 94% with a trend towards
a higher success rate in the past two decades [
22
26
]. Nevertheless, complications related
to the ventral approach such as laryngeal paralysis, dyspnoea, dysphagia, or implant-
failure-related complications are consistently reported [3,6,24,2729].
To our knowledge, there is only one recent case series describing DAAS using screws
and PMMA cement with a positive outcome reported in six dogs suffering from AAI and
C1–C2 incongruence [
15
]. Jeffrey (1996) also reported a successful outcome in a Yorkshire
Terrier following cross pinning of the spinous process of C2 to the wings of C1 and cement
embedding [
19
]. The aim of this study was to assess the feasibility and surgical outcome
of a modified rigid cemented DAAS. Here, we report a safe viable alternative to ventral
techniques. This study also demonstrates that similar implant placement accuracy can be
achieved from a dorsal approach when compared to ventral constructs [30].
2. Materials and Methods
2.1. Criteria for Case Selection and Data Collection
The medical records of all dogs treated with DAAS in two referral practices between
2019 and 2020 were retrospectively reviewed. The diagnosis of AAI was confirmed by
advanced imaging and defined as appreciable dorsal displacement of C2 relative to C1
with/or without evidence of spinal cord compression or intramedullary lesions. Descrip-
tive data collected for each dog included signalment, onset, and duration of clinical signs,
neurological examination, dog video recordings, preoperative treatments, and postoper-
ative notes including surgical complications. Using a modified Frankel scale [
24
], each
dog was neurologically graded before surgery, on discharge and on short-term (<3 months
postsurgery) and long-term (>6 months postsurgery) follow-up: grade 0 for normal gait
without pain, grade 1 for normal gait with neck pain; grade 2 for proprioceptive ataxia;
grade 3 for ambulatory tetraparesis; grade 4 for nonambulatory tetraparesis and grade 5
for tetraplegia.
2.2. Advanced Imaging
Diagnostic imaging, including MRI, CT, and plain radiography (if available), was
reviewed by at least one board-certified neurologist. Magnetic resonance (MR) images
were acquired under general anaesthesia using a 1.5 T scanner (either Tim system or
Life 2021,11, 1039 3 of 15
Magnetom Essenza 1.5 MRI; Siemens AG, Erlangen, Germany). CT scan images were
obtained either under sedation or general anaesthesia using either a 160-slice scanner
(Aquillion PRIME Toshiba, Canon Medical Systems USA, Inc., United States) or a dual-slice
scanner (Siemens Dual Slice Somatom Spirit, Siemens AG, Erlangen, Germany). CT images
were used to evaluate the osseous structure for anomalous or traumatic lesions and for
surgical planning. The images were imported and reviewed on Horos
DICOM viewer,
using bone window CT images in the 2D viewer, 3D multiplanar reconstruction, and 3D
volume rendering modes.
2.3. Surgical Planning
Preoperative surgical planning was performed using 3D Slicer software (Surgical
Planning Lab, Harvard Medical School, Harvard University, Boston, MA, USA, http:
//www.slicer.org, accessed on 27 September 2021). The optimal trajectory was deter-
mined in three planes by orientating screw models within the bone corridors of the 3D-
reconstructed bone segmentation (Figure 1). C1 and C2 segments were realigned to an
estimated anatomical location to facilitate visualisation of implant positions with respect to
the sagittal plane. Subsequently, optimal screw diameters, screw entry points, inclinations
between screw long axis and sagittal plane, and drilling depths of each implant were deter-
mined and exported to a Microsoft Excel
®
spreadsheet (Microsoft Corporation, Redmond,
WA, USA). Video recordings of the surgical plan including the 3D anatomy, entry points,
and screw directions were generated for intraoperative visualisation (Video S1). Purposed
bone corridors included C1 lateral masses and wings (4 sites), C2 cranial articular surfaces,
and cranial/caudal portions of C2 spinous process (4 sites). Occipital crest entry points
were planned to avoid the transverse venous sinus where applicable. For selected cases,
3D printed drilling guides were printed with PLA filament using Ultimaker
printer
(Ultimaker, The Netherlands) and Ultimaker Cura software.
2.4. Surgical Technique
Dogs were positioned in sternal recumbency with slight elevation and ventral flexion
of the head and secured to the table with tape and/or a vacuum cushion. A midline
dorsal approach was performed from the occipital crest to the middle of the third cervical
vertebra, elevating subcutaneous tissue and epaxial muscles until exposition of the dorsal
surface of the atlantoaxial vertebrae was obtained. Gelpi retractors were carefully placed to
allow gentle dissection around the cranial surface of C2 preserving C1 and C2 nerve roots.
The stabilisation technique involved bi-cortical screws (stainless steel or titanium) placed
in at least 4 of 8 available C1 and C2 bone corridors. The screw diameter was selected
based on the surgical plan, ranging from 1.5 mm to 2.7 mm. Using a high-speed 1 mm
burr, the entry points of each screw site were marked by burring through the cis cortex.
Custom-made stainless-steel tubes were used over the drill bits to prevent tissue damage
and to act as a drill stopper (Figure 2). Drilling direction was either estimated by visual
assessment of a video recording depicting 3D screw positions on a computer display or
using a wedge osteotomy gauge to match the calculated values of inclination angles to the
sagittal plane (Figure 2c). When significant concerns were raised about occipito-atlantal
instability, a titanium mesh and additional cortical screws were added to the construct
extending it to the occipital crest. Depending on surgeons’ preferences and accessibility,
partial articular surface drilling and bone allograft were performed within the C1–C2
synovial joint and between the C1 dorsal arch and C2 spinous process. Realignment of
C1 and C2 was achieved by cautiously applying cranioventrally directed pressure on the
spinous process of C2 whilst embedding the metal implants in polymethylmethacrylate
cement. Routine multilayered suture followed. A postoperative CT scan was performed to
determine screw placement quality.
Life 2021,11, 1039 4 of 15
Life 2021, 11, x FOR PEER REVIEW 3 of 16
2.2. Advanced Imaging
Diagnostic imaging, including MRI, CT, and plain radiography (if available), was
reviewed by at least one board-certified neurologist. Magnetic resonance (MR) images
were acquired under general anaesthesia using a 1.5 T scanner (either Tim system or
Magnetom Essenza 1.5 MRI; Siemens AG, Erlangen, Germany). CT scan images were
obtained either under sedation or general anaesthesia using either a 160-slice scanner
(Aquillion PRIME Toshiba, Canon Medical Systems USA, Inc., United States) or a dual-
slice scanner (Siemens Dual Slice Somatom Spirit, Siemens AG, Erlangen, Germany). CT
images were used to evaluate the osseous structure for anomalous or traumatic lesions
and for surgical planning. The images were imported and reviewed on Horos™ DICOM
viewer, using bone window CT images in the 2D viewer, 3D multiplanar reconstruction,
and 3D volume rendering modes.
2.3. Surgical Planning
Preoperative surgical planning was performed using 3D Slicer software (Surgical
Planning Lab, Harvard Medical School, Harvard University, Boston, MA, USA,
http://www.slicer.org). The optimal trajectory was determined in three planes by
orientating screw models within the bone corridors of the 3D-reconstructed bone
segmentation (Figure 1). C1 and C2 segments were realigned to an estimated anatomical
location to facilitate visualisation of implant positions with respect to the sagittal plane.
Subsequently, optimal screw diameters, screw entry points, inclinations between screw
long axis and sagittal plane, and drilling depths of each implant were determined and
exported to a Microsoft Excel® spreadsheet (Microsoft Corporation, Redmond, WA).
Video recordings of the surgical plan including the 3D anatomy, entry points, and screw
directions were generated for intraoperative visualisation (Video S1). Purposed bone
corridors included C1 lateral masses and wings (4 sites), C2 cranial articular surfaces, and
cranial /caudal portions of C2 spinous process (4 sites). Occipital crest entry points were
planned to avoid the transverse venous sinus where applicable. For selected cases, 3D
printed drilling guides were printed with PLA filament using UltimakerTM printer
(Ultimaker, The Netherlands) and Ultimaker Cura software.
(a).
Life 2021, 11, x FOR PEER REVIEW 4 of 16
(b).
(c).
Figure 1. Example of surgical planning screenshots used to guide screw placement
intraoperatively: (a) dorsal view of C1–C2 with planned screw insertion points depicted in red; (b)
left lateral view used to estimate the approximate orientation of the planned screws in a
craniocaudal direction; (c) caudal view of C1–C2 used to visually estimate the screw inclination
with respect to the sagittal plane which could also be obtained using numerical values depicted in
the associated table. Supplementary Video S1 further demonstrates how these images can be used
for anatomical, screw insertion, and screw orientation visualisations.
2.4. Surgical Technique
Dogs were positioned in sternal recumbency with slight elevation and ventral flexion
of the head and secured to the table with tape and/or a vacuum cushion. A midline dorsal
approach was performed from the occipital crest to the middle of the third cervical
vertebra, elevating subcutaneous tissue and epaxial muscles until exposition of the dorsal
surface of the atlantoaxial vertebrae was obtained. Gelpi retractors were carefully placed
to allow gentle dissection around the cranial surface of C2 preserving C1 and C2 nerve
roots. The stabilisation technique involved bi-cortical screws (stainless steel or titanium)
placed in at least 4 of 8 available C1 and C2 bone corridors. The screw diameter was
selected based on the surgical plan, ranging from 1.5 mm to 2.7 mm. Using a high-speed
1 mm burr, the entry points of each screw site were marked by burring through the cis
cortex. Custom-made stainless-steel tubes were used over the drill bits to prevent tissue
damage and to act as a drill stopper (Figure 2). Drilling direction was either estimated by
visual assessment of a video recording depicting 3D screw positions on a computer
display or using a wedge osteotomy gauge to match the calculated values of inclination
angles to the sagittal plane (Figure 2c). When significant concerns were raised about
occipito-atlantal instability, a titanium mesh and additional cortical screws were added to
Figure 1.
Example of surgical planning screenshots used to guide screw placement intraoperatively:
(
a
) dorsal view of C1–C2 with planned screw insertion points depicted in red; (
b
) left lateral view
used to estimate the approximate orientation of the planned screws in a craniocaudal direction;
(
c
) caudal view of C1–C2 used to visually estimate the screw inclination with respect to the sagit-
tal plane which could also be obtained using numerical values depicted in the associated table.
Supplementary Video S1
further demonstrates how these images can be used for anatomical, screw
insertion, and screw orientation visualisations.
Life 2021,11, 1039 5 of 15
Life 2021, 11, x FOR PEER REVIEW 5 of 16
the construct extending it to the occipital crest. Depending on surgeons’ preferences and
accessibility, partial articular surface drilling and bone allograft were performed within
the C1–C2 synovial joint and between the C1 dorsal arch and C2 spinous process.
Realignment of C1 and C2 was achieved by cautiously applying cranioventrally directed
pressure on the spinous process of C2 whilst embedding the metal implants in
polymethylmethacrylate cement. Routine multilayered suture followed. A postoperative
CT scan was performed to determine screw placement quality.
(a) (b) (c)
Figure 2. Photographs demonstrating the use of a drill stopper and inclination guide on a 3D printed
C1-C2 model: (a) instruments used to accurately drill through the bone corridors including a
surgical power drill, drill bits, custom-made drill stoppers, and osteotomy wedge gauge; (b) a drill
stopper in the form a stainless steel tube (marked with green and blue tape) can be used to protect
adjacent tissue and control drilling depth, the craniocaudal drilling inclination is guided by 3D
planning images (compare to green screw axis in Figure 1b); (c) osteotomy wedge gauge are simple
stainless steel triangles that can be used to approximate the inclination of drilling with respect to
the sagittal plane, the angle values for each screw site were calculated and used at the surgeon’s
discretion (see Figure 1c).
2.5. Immediate Postoperative Care
All dogs received multimodal analgesia and supportive care whilst recovering from
the procedure in the hospital environment.
2.6. Follow-Up
Short-term follow-up was assessed via physical examination, video footage, and
client questionnaire. Long-term follow-up was obtained via physical examination, video,
and/or telephone questionnaire. Where possible postoperative CT images were obtained
at one of these time points or more. A successful outcome was defined as being
ambulatory without reported or observed discomfort and without evidence of clinical
deterioration when compared to prior to surgery.
2.7. Implant Accuracy, Bone Fusion, and Implant Failure
Presurgical planned 3D screw position was compared with all available
postoperative CT studies. Registration of the different time points was performed using
3D slicer software, aligning the bone anatomy and, if necessary, the PMMA cement (when
significant bone growth occurred). Screws were classified as dangerous (vertebral canal
violation equal or greater than ½ screw diameter), hazardous (vertebral canal violation
less than ½ screw diameter or breaching intervertebral foramen or other unintended
anatomical structures), suboptimal (including monocortical placement, breaching
laterally of the bone corridor or inappropriate screw length) or optimal (bicortical and
contained within the intended bone corridor) (Figure 3). A screw ratio rather than a metric
Figure 2.
Photographs demonstrating the use of a drill stopper and inclination guide on a 3D printed C1–C2 model:
(
a
) instruments used to accurately drill through the bone corridors including a surgical power drill, drill bits, custom-made
drill stoppers, and osteotomy wedge gauge; (
b
) a drill stopper in the form a stainless steel tube (marked with green and
blue tape) can be used to protect adjacent tissue and control drilling depth, the craniocaudal drilling inclination is guided
by 3D planning images (compare to green screw axis in Figure 1b); (
c
) osteotomy wedge gauge are simple stainless steel
triangles that can be used to approximate the inclination of drilling with respect to the sagittal plane, the angle values for
each screw site were calculated and used at the surgeon’s discretion (see Figure 1c).
2.5. Immediate Postoperative Care
All dogs received multimodal analgesia and supportive care whilst recovering from
the procedure in the hospital environment.
2.6. Follow-Up
Short-term follow-up was assessed via physical examination, video footage, and client
questionnaire. Long-term follow-up was obtained via physical examination, video, and/or
telephone questionnaire. Where possible postoperative CT images were obtained at one
of these time points or more. A successful outcome was defined as being ambulatory
without reported or observed discomfort and without evidence of clinical deterioration
when compared to prior to surgery.
2.7. Implant Accuracy, Bone Fusion, and Implant Failure
Presurgical planned 3D screw position was compared with all available postoperative
CT studies. Registration of the different time points was performed using 3D slicer software,
aligning the bone anatomy and, if necessary, the PMMA cement (when significant bone
growth occurred). Screws were classified as dangerous (vertebral canal violation equal
or greater than
1
2
screw diameter), hazardous (vertebral canal violation less than
1
2
screw
diameter or breaching intervertebral foramen or other unintended anatomical structures),
suboptimal (including monocortical placement, breaching laterally of the bone corridor
or inappropriate screw length) or optimal (bicortical and contained within the intended
bone corridor) (Figure 3). A screw ratio rather than a metric measurement was used to
account for the wide variation in dog size. Bone fusion and implant displacement were
also subjectively assessed and recorded. All measurements and CT image analysis were
performed by a single observer (G.L.).
Life 2021,11, 1039 6 of 15
Life 2021, 11, x FOR PEER REVIEW 6 of 16
measurement was used to account for the wide variation in dog size. Bone fusion and
implant displacement were also subjectively assessed and recorded. All measurements
and CT image analysis were performed by a single observer (G.L.).
Figure 3. Axis 3D reconstructions depicting our proposed screw position grading system.
Examples include a dangerous screw with vertebral canal violation greater than ½ the screw
diameter (red), a hazardous screw with vertebral canal violation less than ½ the screw diameter
(orange), two suboptimal screws (yellow) including a monocortical placement (left), and an
inappropriate length (right) and two optimal screws (green).
3. Results
3.1. Signalment and Clinical Presentation
In total, 12 dogs with atlantoaxial instability (AAI) were included in this study (Table
1) with a mean age of 16.3 months (range 3.3–75 months) and a mean weight of 5.5 kg
(range 1.5–25 kg). The most represented breed was the Chihuahua (n = 4), followed by
Dachshund (n = 2). Two dogs were presented with chronic neurological signs—one with
an acute on chronic presentation, three with acute/subacute presentations and four with
paroxysmal episodes of presumptive pain and/or vestibular signs. Three dogs had a
recognised traumatic event. Grade 4, nonambulatory tetraparetic dogs (33.3%) and grade
3, ambulatory tetraparetic dogs (25%) were the more frequent neurological grades
encountered prior to surgery. Two of the four dogs with grade 4 tetraparesis had a
sustained trauma and were classified as nonambulatory by both caregivers and referring
veterinary surgeons and therefore examined in lateral recumbency. The only dog that
presented tetraplegic had acutely deteriorated one day after attempted management of
AAI using a dorsal suture technique with nonabsorbable sutures [31]. One case was
treated with a cervical bandage for 3 months until a CT scan revealed significant
worsening of the previous atlantoaxial luxation which prompted surgical intervention
(case 2).
Table 1. Clinical information, complications, and outcomes.
Case
Signalment
(breed, age in months,
gender, weight)
Onset/clinical
progression
Additional
anomalies
Surgical time (Sx),
hospitalisation (H)
and
complications (C)
Neurological
score 1 Final outcome
(time in months)
Initial Follow-up
1 Dachshund,
4 m, ME, 3.3 kg
Chronic and
progressive
Dens hypoplasia,
C1-C2
incongruence
Sx: 260 min
H: 2 days
C: none
Ad: 3
Dis: 3
ST: 3
LT: 2
19 m: improved gait,
unable to jump
2 Maltese,
17 m, ME, 1.5 kg
Paroxysmal
episodes pain None
Sx: 155 min
H: 2 days
C: regurgitation
(short lived)
Ad: 1
Dis: 2
ST: 2
LT: 0
20 m: much-improved gait;
rare episodes of mild pain,
occasional cough
when drinking
3 Chihuahua,
10 m, FE, 2 kg
Chronic
progressive
with peracute
deterioration
Dens hypoplasia,
atlanto-occipital
overlap
Sx: 300 min
H: 4 days
C: subtle torticollis.
Ad: 4
Dis: 4
ST: 3
LT: 2
11 m: improved gait;
rare episodes of 2-3 seconds
thoracic limb collapse
and limb paddling.
Figure 3.
Axis 3D reconstructions depicting our proposed screw position grading system. Examples
include a dangerous screw with vertebral canal violation greater than
1
2
the screw diameter (red),
a hazardous screw with vertebral canal violation less than
1
2
the screw diameter (orange), two
suboptimal screws (yellow) including a monocortical placement (
left
), and an inappropriate length
(right) and two optimal screws (green).
3. Results
3.1. Signalment and Clinical Presentation
In total, 12 dogs with atlantoaxial instability (AAI) were included in this study (
Table 1
)
with a mean age of 16.3 months (range 3.3–75 months) and a mean weight of 5.5 kg (range
1.5–25 kg). The most represented breed was the Chihuahua (n = 4), followed by Dachshund
(n = 2). Two dogs were presented with chronic neurological signs—one with an acute on
chronic presentation, three with acute/subacute presentations and four with paroxysmal
episodes of presumptive pain and/or vestibular signs. Three dogs had a recognised
traumatic event. Grade 4, nonambulatory tetraparetic dogs (33.3%) and grade 3, ambulatory
tetraparetic dogs (25%) were the more frequent neurological grades encountered prior
to surgery. Two of the four dogs with grade 4 tetraparesis had a sustained trauma and
were classified as nonambulatory by both caregivers and referring veterinary surgeons and
therefore examined in lateral recumbency. The only dog that presented tetraplegic had
acutely deteriorated one day after attempted management of AAI using a dorsal suture
technique with nonabsorbable sutures [
31
]. One case was treated with a cervical bandage
for 3 months until a CT scan revealed significant worsening of the previous atlantoaxial
luxation which prompted surgical intervention (case 2).
3.2. Preoperative Imaging Interpretation and Surgical Planning
All cases had a dorsal displacement of C2 causing spinal cord compression from
either congenital or traumatic aetiology (Figure 4). Further relevant craniocervical findings
are reported in Table 1. Intramedullary hyperintensity on T2-weighted images on the
region of C1–2 was reported in four dogs (33.3%). Two dogs had displaced fractures; case
4 had a cranial C2 fracture and dorsal midline C1 fracture, and case 6 had a C2 body
fracture through the cranial articular surfaces. Atlantoaxial incongruence was present
in 33.3% of the cases (n = 4), atlanto-occipital overlap was encountered in two dogs
(16.6%) and two dogs had complex occipito-atlantoaxial malformations with partial atlanto-
occipital fusion (16.6%). Dens hypoplasia was present in four dogs, while dens aplasia
was present in two dogs. Other concomitant findings were occasionally reported on MRI
including ventriculomegaly in 5 dogs (41.7%) and supracollicular fluid accumulation
in three dogs (25%). Of the three dogs with a traumatic injury, two had preexisting
congenital malformations (dog 4 had an incomplete fusion of C1 ventral arch, and dog 10
had a complex OAAM).
Preoperative surgical planning was used in all but one dog (case 4) and 3D-printed
guides were used in the first two dogs. Some neurosurgeons preferred to use visual
assessment of screw directions (R.G.Q. and R.J.L.), while others preferred to use numerical
values and osteotomy wedge gauge (G.L.).
Life 2021,11, 1039 7 of 15
Table 1. Clinical information, complications, and outcomes.
Case
Signalment
(Breed, Age in Months,
Gender, Weight)
Onset/Clinical
Progression Additional
Anomalies
Surgical Time (Sx),
Hospitalisation (H) and
Complications (C)
Neurological Score 1Final Outcome
(Time in Months)
Initial Follow-Up
1Dachshund,
4 m, ME, 3.3 kg Chronic and
progressive
Dens
hypoplasia,
C1–C2
incongruence
Sx: 260 min
H: 2 days
C: none
Ad: 3
Dis: 3
ST: 3
LT: 2
19 m: improved gait,
unable to jump
2Maltese,
17 m, ME, 1.5 kg
Paroxysmal
episodes pain None
Sx: 155 min
H: 2 days
C: regurgitation
(short lived)
Ad: 1
Dis: 2
ST: 2
LT: 0
20 m:
much-improved gait;
rare episodes of mild
pain, occasional
cough when drinking
3Chihuahua,
10 m, FE, 2 kg
Chronic
progressive with
peracute
deterioration
Dens
hypoplasia,
atlanto-
occipital
overlap
Sx: 300 min
H: 4 days
C: subtle torticollis.
Ad: 4
Dis: 4
ST: 3
LT: 2
11 m: improved gait;
rare episodes of 2–3
seconds thoracic limb
collapse and limb
paddling.
4Labrador Retriever cross,
3 m, FE, 9 kg
Trauma
C1 and C2 fracture
Incomplete
fusion of C1
ventral arch
Sx: 215 min
H: 3 days
C: none
Ad: 4*
Dis: 3
ST: 0
LT: 0
12 m: normal gait;
subtle stiffness of the
neck.
5Chihuahua cross,
23 m, FN, 2.2 kg Chronicprogressive Dens aplasia,
C1–C2
incongruence
Sx: 165 min
H: 4 days
C: none
Ad: 3
Dis: 3
ST: X
LT: 0 16 m: normal gait
6German
Shorthaired Pointer,
6 m, ME, 25 kg
Trauma
C2 fracture None Sx: 175 min
H: 10 days
C: none
Ad: 4*
Dis: 2
ST: 0
LT: 0 16 m: normal gait
7Chihuahua,
11 m, MN, 2.2 kg
Single acute
episode of pain,
collapse, and ataxia
C1 incomplete
dorsal arch
fusion, caudal
occipital
malformation,
Sx: 220 min
H: 2 days
C: none
Ad: 2
Dis: 2
ST: 2
LT: 0
10 m: improved gait;
3 further paroxysmal
episodes of syncope
vs. seizures.
Reverse sneezing and
occasional dysphagia.
Frequent neck
scratching
responding to
gabapentin
8Yorkshire Terrier,
4 m, FE, 1.7 kg Subacute
Dens agenesis,
atlanto-
occipital
overlap
(suspect
instability)
Sx: 160 min
H: 2 days
C: none
Ad: 5
Dis: 4
ST: 2
LT: 2
8 m: improved gait;
mild ataxia in all four
limbs
(suspected
occipito-atlantal
instability 5 m
postsurgery due to
acute deterioration)
9Cockapoo,
4 m, FE, 3.8 kg
Acute after
collision with
another dog
Complex
OAAM,
partial
occipito-
atlantal fusion,
C1–C2
incongruence
Sx: 220 min
H: 4 days
C: none
Ad: 4
Dis: 2
ST: 0
LT: 0
6 m: improved gait;
subtle low head
carriage.Rare
paroxysmal
vestibular episodes
(improving with diet
adjust-
ment/gabapentin)
10 Chihuahua,
34 m, MN, 2.1 kg
Paroxysmal
episodes of pain
and lateral
recumbency
C1–C2
incongruence
Sx: 165 min
H: 2 days
C: none
Ad: 1
Dis: 2
ST: 0
LT: 0
6 m: return to normal
(phone
communication only)
11 Dachshund,
75 m, FN, 4.7 kg
Paroxysmal
vestibular
episodes 2
C2–3 block
vertebrae
Sx: 245 min
H: 3 days
C: none
Ad: 0
Dis: 0
ST: 0
LT: 0
6 m: normal gait.
No episodes since
surgery
12 Lagotto Romagnolo,
4 m, FE, 8.5 kg
Acute and
progressive
Complex
OAAM,
partial
occipito-
atlantal
fusion
Sx: 200 min
H: 2 days
C: subcutaneous seroma and
subtle torticollis
Ad: 3
Dis: 3
ST: 1
LT: 0
6 m: improved gait;
mild over-reaching of
all four limbs.
Slight resistance to
cervical ventroflexion
1
Neurological grades using a modified Frankel scale: 0, normal gait without neck pain: 1, normal gait with neck pain; 2, proprioceptive
ataxia; 3, ambulatory tetraparesis; 4, nonambulatory tetraparesis; 5, tetraplegia [
24
]. Ad: admission; Dis: discharge; ST: short term; LT:
long term, X: not available; *: assessed in lateral recumbency.
2
Reproducible paroxysmal episodes of vestibular signs elicited with flexion
of the head (evaluated via video recording provided at the time of referral). ME: male entire; MN: male neutered; FE: female entire; FN:
female neutered.
Life 2021,11, 1039 8 of 15
Life 2021, 11, x FOR PEER REVIEW 8 of 16
preexisting congenital malformations (dog 4 had an incomplete fusion of C1 ventral arch,
and dog 10 had a complex OAAM).
Preoperative surgical planning was used in all but one dog (case 4) and 3D-printed
guides were used in the first two dogs. Some neurosurgeons preferred to use visual
assessment of screw directions (R.G.Q. and R.J.L.), while others preferred to use numerical
values and osteotomy wedge gauge (G.L.).
(a) (b) (c)
Figure 4. Craniocervical 3D reconstructions depicting the range of anomalies treated in this study:
(a) congenital atlantoaxial instability; (b) C2 fracture; (c) complex occipito-atlantoaxial
malformation.
3.3. Immediate Surgical Outcome
All DAAS procedures were successfully completed allowing stabilisation of C1-C2
and occasionally also involving the occipital bone due to partial occipito-atlantal fusion
(Figure 5). The median surgery time was 207 min (range 155–300 min). No major
complications were reported intraoperatively. One case had CSF leakage following a
small incision through the lateral aspect of the vertebral canal within the intervertebral
foramen. Mild-to-moderate haemorrhage was commonly observed around the lateral
foramen and C1–C2 intervertebral foramen often prolonging dissection time. On
recovery, one dog had a few episodes of regurgitation immediately postoperatively which
responded to proton pump inhibitor treatment (omeprazole, 1 mg/kg/12 h).
Based on postoperative CT images, apposition was considered optimal in all cases.
A total of 72 atlantoaxial screws were placed (Table 2), with 51 (70.8%) graded as optimal.
Four screws (5.6%) located in C1 lateral masses (n = 2) and C2 cranial articular surface (n
= 2) were graded as hazardous—two had minor vertebral canal breach, one was
excessively long, and one breached the alar foramen. The remaining 17 (23.6%) screws
were considered safe but not perfectly placed within the intended corridors, most
commonly due to a monocortical position in 11 (15.3%) screws. None of the screws were
graded as dangerous. A titanium mesh affixed to the occipital bone was added to the
construct in one dog suffering from a complex occipito-atlantoaxial malformation (Figure
5c). A case example is presented in Figure 6, depicting C1–C2 reduction and screw
accuracy.
(a) (b) (c)
Figure 4.
Craniocervical 3D reconstructions depicting the range of anomalies treated in this study: (
a
) congenital atlantoaxial
instability; (b) C2 fracture; (c) complex occipito-atlantoaxial malformation.
3.3. Immediate Surgical Outcome
All DAAS procedures were successfully completed allowing stabilisation of C1–C2
and occasionally also involving the occipital bone due to partial occipito-atlantal fusion
(
Figure 5
). The median surgery time was 207 min (range 155–300 min). No major com-
plications were reported intraoperatively. One case had CSF leakage following a small
incision through the lateral aspect of the vertebral canal within the intervertebral foramen.
Mild-to-moderate haemorrhage was commonly observed around the lateral foramen and
C1–C2 intervertebral foramen often prolonging dissection time. On recovery, one dog had
a few episodes of regurgitation immediately postoperatively which responded to proton
pump inhibitor treatment (omeprazole, 1 mg/kg/12 h).
Life 2021, 11, x FOR PEER REVIEW 8 of 16
preexisting congenital malformations (dog 4 had an incomplete fusion of C1 ventral arch,
and dog 10 had a complex OAAM).
Preoperative surgical planning was used in all but one dog (case 4) and 3D-printed
guides were used in the first two dogs. Some neurosurgeons preferred to use visual
assessment of screw directions (R.G.Q. and R.J.L.), while others preferred to use numerical
values and osteotomy wedge gauge (G.L.).
(a) (b) (c)
Figure 4. Craniocervical 3D reconstructions depicting the range of anomalies treated in this study:
(a) congenital atlantoaxial instability; (b) C2 fracture; (c) complex occipito-atlantoaxial
malformation.
3.3. Immediate Surgical Outcome
All DAAS procedures were successfully completed allowing stabilisation of C1-C2
and occasionally also involving the occipital bone due to partial occipito-atlantal fusion
(Figure 5). The median surgery time was 207 min (range 155–300 min). No major
complications were reported intraoperatively. One case had CSF leakage following a
small incision through the lateral aspect of the vertebral canal within the intervertebral
foramen. Mild-to-moderate haemorrhage was commonly observed around the lateral
foramen and C1–C2 intervertebral foramen often prolonging dissection time. On
recovery, one dog had a few episodes of regurgitation immediately postoperatively which
responded to proton pump inhibitor treatment (omeprazole, 1 mg/kg/12 h).
Based on postoperative CT images, apposition was considered optimal in all cases.
A total of 72 atlantoaxial screws were placed (Table 2), with 51 (70.8%) graded as optimal.
Four screws (5.6%) located in C1 lateral masses (n = 2) and C2 cranial articular surface (n
= 2) were graded as hazardous—two had minor vertebral canal breach, one was
excessively long, and one breached the alar foramen. The remaining 17 (23.6%) screws
were considered safe but not perfectly placed within the intended corridors, most
commonly due to a monocortical position in 11 (15.3%) screws. None of the screws were
graded as dangerous. A titanium mesh affixed to the occipital bone was added to the
construct in one dog suffering from a complex occipito-atlantoaxial malformation (Figure
5c). A case example is presented in Figure 6, depicting C1–C2 reduction and screw
accuracy.
(a) (b) (c)
Figure 5.
Craniocervical 3D reconstructions depicting the range of stabilisation constructs and screw accuracy obtained
with the proposed method: (
a
) dorsal view of a construct with screws in C1 lateral masses (n = 2), C2 cranial articular
surfaces (n = 2), and C2 spinous process (n = 1) used to stabilise a congenital AAI; (
b
) left lateral view of a similar construct
which can also be used to treat cranial C2 fractures; (
c
) dorsal oblique view of a complex construct involving screws placed
in all 8 available bone corridors and a titanium mesh affixed to the occipital crest. Yellow screws: preoperatively planned
position; blue screws: actual postoperative position; blue semitransparent areas: polymethylmethacrylate cement location.
Based on postoperative CT images, apposition was considered optimal in all cases.
A total of 72 atlantoaxial screws were placed (Table 2), with 51 (70.8%) graded as optimal.
Four screws (5.6%) located in C1 lateral masses (n = 2) and C2 cranial articular surface
(
n=2
) were graded as hazardous—two had minor vertebral canal breach, one was exces-
sively long, and one breached the alar foramen. The remaining 17 (23.6%) screws were
considered safe but not perfectly placed within the intended corridors, most commonly
due to a monocortical position in 11 (15.3%) screws. None of the screws were graded as
Life 2021,11, 1039 9 of 15
dangerous. A titanium mesh affixed to the occipital bone was added to the construct in
one dog suffering from a complex occipito-atlantoaxial malformation (Figure 5c). A case
example is presented in Figure 6, depicting C1–C2 reduction and screw accuracy.
Table 2. Accuracy of screw placement and long-term construct analysis.
Case Screw Sites (n) Screws Grading Score 1
Bone Graft C1–C2 Fusion
(Location)
Construct
Failure
CT Time
Postsurgery (Months)
C1 C2 n Sites
1LM: 2
Wi: 0
AS: 2
SP: 1
Score 0: 4
Score 1: 1
Score 2: 0
LM, AS, SP
LM Yes
Yes
(C1 dorsal arch
and C1–C2
articular surfaces)
No 19
2LM: 2
Wi: 0
AS: 2
SP: 1
Score 0: 4
Score 1: 1
Score 2: 0
LM, AS, SP
LM Yes No No 20
3LM: 2
Wi: 0
AS: 2
SP: 1
Score 0: 3
Score 1: 2
Score 2: 0
LM, AS, SP
LM, AS No No Yes 11
4LM: 2
Wi: 0
AS: 2
SP: 1
Score 0: 2
Score 1: 1
Score 2: 2
AS, SP
LM
LM, AS
No Yes
(C1 dorsal arch) No 12
5LM: 2
Wi: 0
AS: 2
SP: 1
Score 0: 4
Score 1: 1
Score 2: 0
LM, AS, SP
AS No No No 16
6LM: 2
Wi: 0
AS: 2
SP: 1
Score 0: 3
Score 1: 2
Score 2: 0
LM, AS, SP
LM, AS No
Partial bone
remodelling
(C1 dorsal arch
and C1–C2
articular surfaces)
No 16
7LM: 2
Wi: 0
AS: 2
SP: 2
Score 0: 4
Score 1: 1
Score 2: 1
AS, SP
LM
LM
No No No 10
8LM: 2
Wi: 0
AS: 1
SP: 1
Score 0: 3
Score 1: 1
Score 2: 0
LM, AS, SP
AS No No No 5
92LM: 2
Wi: 2
AS: 2
SP: 2
Score 0: 6
Score 1: 2
Score 2: 0
LM, Wi, AS,
SP
LM, AS No
Partial bone
remodelling
(C1 ventral arch)
No 6
10 LM: 2
Wi: 2
AS: 2
SP: 2
Score 0: 6
Score 1: 2
Score 2: 0
LM, Wi, AS,
SP
AS, Wi No n/a n/a n/a
11 LM: 2
Wi: 2
AS: 2
SP: 2
Score 0: 5
Score 1: 2
Score 2: 1
LM, AS, SP
LM, AS
AS
Yes No No 6
12 LM: 2
Wi: 2
AS: 2
SP: 2
Score 0: 7
Score 1: 1
Score 2: 0
LM, Wi, AS,
SP
LM Yes
Partial bone
remodelling
(C1–C2 right
articular surface)
No 6
1
Screw placement score: 0, optimal; 1, suboptimal; 2, hazardous; 3, dangerous. LM: C1 lateral masses; Wi: C1 wings; AS: C2 cranial
articular surfaces; SP: C2 spinous process; n/a: not available.
2
This dog also had 5 optimally placed self-drilling self-tapping monocortical
titanium screws placed to anchor a titanium mesh.
Life 2021,11, 1039 10 of 15
Figure 6.
MR and CT images with superimposed 3D models depicting C1–C2 reduction and screw accuracy obtained in
case 7: (
a
) preoperative sagittal MR image with dorsal displacement of C2 dens (white arrowhead); (
b
) postoperative sagittal
MR image demonstrating C1–C2 reduction and the benefit of using titanium screws allowing visualisation of the local
anatomy; (
c
) dorsal 3D reconstruction depicting the overall screw placement accuracy; (
d
) cranial view of CT transverse
image illustrating a hazardous position of the left lateral mass screw due to slight vertebral canal violation (dashed arrow);
(
e
) caudal view of a CT transverse image and C1 3D model revealing the suboptimal position of the right lateral mass screw
due to penetration of the C1–C2 synovial joint (white arrow); (
f
) cranial view of a CT transverse image and C2 3D model
depicting 3 optimal screw placements (bicortical within the intended bone corridor). Yellow screws: preoperatively planned
positions; blue screws: actual postoperative positions.
3.4. Perioperative Outcome
Mean time to discharge after surgery was 3.3 days (range 2–10 days). The dog with
the longest hospitalisation time had a C2 vertebral body fracture and presented with
nonambulatory tetraparesis. All dogs except one were discharged with one or more
of the following medications: an NSAID, prednisolone at an anti-inflammatory dose,
paracetamol, and/or gabapentin. At discharge, 6/12 dogs were graded as unchanged,
compared to admission, 4/12 improved by at least one grade, and 2/12 were considered
worse. Two dogs were discharged with subtle torticollis, one of which later developed
a subcutaneous seroma in the surgical region.
3.5. Short-Term Clinical Outcome
In total, 11 dogs were reexamined between 1 to 2.8 months after surgery (mean
2 months). Neurological grading was performed for each dog (Table 1). All owners
reported improvement in gait and/or painful episodes. One dog was reported to be unable
to jump despite his gait being much improved (case 1), and one was reported to have rare
episodes of yelping and stiffness (case 2).
Case 7 was reported to have two further episodes of syncope or seizures and to
experience reverse sneezing. Three episodes of dysphagia were reported after swallowing
entire biscuits. Repeat MRI, CT, CSF analysis, echocardiography, and a continuous ECG
(Holter monitoring) failed to identify a cause. Vagal syncope or seizure-like episodes were
the main differential diagnoses. The latter hypothesis could be related to ventriculomegaly
and supracollicular fluid accumulation. Frequent neck scratching was also reported in
this case which was considered secondary to the identified Chiari-like malformation.
Omeprazole (10 mg/kg BID) and gabapentin (17 mg/kg, BID) were prescribed.
Life 2021,11, 1039 11 of 15
Case 8 was improving until an acute deterioration 4.9 months after surgery. Dynamic
radiographs and CT imaging suggested the presence of atlanto-occipital instability which
worsened with ventroflexion of the head. This dog was managed medically with rest and
a neck brace.
Case 9 developed paroxysmal episodes that were vestibular in nature (vomiting,
horizontal nystagmus, vestibular ataxia) that lasted hours and then returned to normal.
A repeated MRI scan, CSF analysis, bile acid stimulation test, and full body CT scan failed
to identify a cause. The dog was managed with a hypoallergenic diet and gabapentin.
3.6. Long-Term Clinical Outcome
Long-term follow-up questionnaires were obtained in 11/12 dogs, while long-term
neurological examination and CT scan were obtained in 10/12 dogs at a mean time of
11.9 months (5.9–19.8 months). All owners were satisfied with the clinical outcome and
reported improvement with the gait and/or painful episodes. All dogs had a good-to-
excellent outcome and had improved neurologically by one or more grades (Table 1).
Case 2 developed rare episodes of suspected pain and weakness of the pelvic limbs
20 months after surgery. Further investigation was declined, but these clinical signs were
considered less likely to be related to the atlantoaxial surgery given the suspected location
of discomfort although an association cannot be completely excluded. The owner also
reported very occasional coughing when drinking water.
Case 3 presented rare episodes of 2–3 seconds collapse of the thoracic limbs along
with paddling without autonomic signs which could be related to persistent AAI given
that construct failure was identified on CT images.
Case 8 was treated with a neck brace for 8 weeks following an acute deterioration. The
dog improved neurologically and was only slightly ataxic in all four limbs with a subtle
hypermetric gait. A right pelvic limb lameness was also reported which was attributed to
medial patellar luxation.
Case 7 experienced another seizure-like episode 7 months after the previous. This
dog continued to be treated with omeprazole and gabapentin and was doing clinically
well in between paroxysmal episodes. Previously reported neck scratching had resolved at
long-term follow-up.
Case 9 had a decreased frequency of episodes of vestibular nature. A hypoallergenic
diet was started, and no further episodes had been observed for 3 months (at the time of
long-term follow-up). The caregivers decided not to pursue any further therapeutical trials
given that the dog had otherwise a good quality of life and was able to exercise normally.
3.7. Bone Fusion and Implant Failure
Overall, stabilisation constructs were considered appropriate and withstood the follow-
up period in all cases except one (Table 2). The construct failure was attributed to poor
cement embedding of the right lateral mass screw, leaving only one C1 screw supporting
the construct which was subsequently pulled out. This implant failure led to appreciable
C1–2 subluxation on follow-up CT; however, the dog still had a positive clinical outcome.
Bone allograft was placed in four cases, with only one having signs of C1–C2 fusion
dorsally and bilaterally on CT scan after 19.3 months after surgery. Both cases with
displaced fractures had signs of fusion of the fracture line, one also displayed C1–C2
fusion dorsally while the other displayed signs of bone remodelling of C1 dorsal arch
and C1–C2 articular surfaces without complete fusion (residual bone separation line).
Two other dogs had signs of bone remodelling ventrally. Overall, continuous C1–C2
bone fusion was observed in two dogs (16.6%), and partial bone fusion (with residual
separation line) in three dogs (25%). Significant bone growth could be appreciated in
younger dogs when overlapping the presurgical CT images with the long-term ones.
Growth mostly occurred within the C1 wings, the caudal portion of the C2 vertebral body,
and C2 transverse processes.
Life 2021,11, 1039 12 of 15
4. Discussion
The present case series suggests that DAAS cemented constructs can be safely per-
formed to treat a wide range of craniocervical junction anomalies and traumatic injuries.
All the dogs improved clinically during the initial follow-up period, and only two dogs
showed mild short-lived clinical deterioration in the immediate postoperative period. No
mortality event has been recorded at the time of writing. Whilst the number of cases re-
ported in this series is modest, our data still suggest that perioperative mortality associated
with rigid DAAS is likely to be low. The limited published data on rigid DAAS cemented
constructs also suggest a low mortality rate [
15
,
19
]. Previous retrospective studies have
associated dorsal techniques with higher mortality rates [
3
,
4
]. However, these conclusions
were based on procedures which involved penetration of the vertebral canal at the level
of C1 dorsal arch and were, therefore, more susceptible to iatrogenic trauma of the spinal
cord [
2
]. Overall, if our results can be reproduced on a larger scale, we anticipate that
success rates of rigid DAAS will likely be similar to that of ventral techniques.
Defining the parameters of a successful outcome can be difficult in veterinary medicine.
Most studies define surgical success in terms of improved subjective gait scoring and
absence of discomfort. Based on such criteria, 11 of our 12 dogs would be considered
successful. Yet, our case series also highlights the complexity of AAI cases; many dogs had
comorbidities that significantly affected the quality of life and therefore outcome. Further
complicating the assessment of outcome, it is possible that residual C1–C2 instability
may be intermittent or even subclinical, and therefore, construct failure may initially go
undetected. In this series, three of our dogs suffered from paroxysmal episodes at long-term
follow-up which may or may not be related to the DAAS procedure. One dog (case 8)
acutely deteriorated 4.9 months after surgery and dynamic imaging revealed the presence
of atlanto-occipital instability without any evidence of construct failure. Another dog
(case 3) displayed a positive clinical outcome, yet follow-up CT images revealed that the
supporting implants were failing. These cases illustrate the fact that success is difficult to
reduce to a single objective criterion in dogs suffering from AAI. When comparing AAI
studies, it may, therefore, be more relevant to compare mortality/complication rates and
technical surgical outcome variables rather than clinical scoring.
A significant aspect of spinal instrumentation procedural safety is related to the
surgeon’s ability to position stabilising implants accurately within the intended bone
corridors and away from vital structures. Our data suggest that the proposed method can
achieve a high level of screw placement accuracy with only four hazardous screws (5.6%)
and no dangerous screws identified. This result is comparatively superior to two separate
cadaveric and clinical studies assessing implant placement accuracy using ventral cemented
techniques (4.4% dangerous screws) but similar to a ventral technique using 3D-printed
drill guides (7% incomplete vertebral canal breach) [
28
,
30
,
32
]. It is difficult to know to
which extent our preoperative planning methodology optimised screw placement accuracy.
Anecdotally, the single case in which the planning method could not be used had the
poorest screw placement scores, including two of the four identified dangerous screws in
our entire population. Further investigation would be necessary to quantify the effect of
preoperative planning on screw placement accuracy. The most challenging bone corridors
proved to be the C1 lateral masses and C2 cranial articular surfaces. This could have been
anticipated considering these corridors have a narrower shape. In theory, these two bone
corridors could be completely avoided by only using the C1 wings and C2 spinous process,
as was previously reported in a single case report [
19
]. However, the remaining bone
corridors are extremely thin and may not be sufficient to sustain long-term cyclic loads.
The optimal number and distribution of screws remains to be established for both ventral
and dorsal techniques. In the absence of comparative data, our strategy was to use as many
bone corridors as possible to maximise the construct’s bone anchorage. As confidence
in the technique and surgical approach increased, it became possible to achieve screw
placement in all eight available corridors. To our knowledge, the use of the C2 cranial
articular surface corridors has not been previously reported from a dorsal approach. Based
Life 2021,11, 1039 13 of 15
on our experience, this corridor can be technically challenging in smaller dogs, but it offers
significantly more bone reserve than the C2 spinous process and is particularly valuable to
stabilise cranially located C2 fractures.
One of the hypothesised benefits of the dorsal approach is that it may prevent compli-
cations that have been historically attributed to iatrogenic injury of ventrally located vital
structures such as the vagosympathetic trunk, larynx, or oesophagus. Such injuries have
been suspected in several ventral atlantoaxial stabilisation studies reporting postoperative
complications such as Horner’s syndrome, dysphagia, dysphonia, laryngeal paralysis,
dyspnoea, aspiration pneumonia, or tracheal necrosis [
2
,
6
,
22
,
24
,
27
,
28
]. While we have not
directly observed such clinical signs, two of our cases suffer from mild dysphagia (cases
2 and 7). In humans, postoperative dysphagia and regurgitation following an anterior
approach have been linked to a craniocervical malalignment in a hyper-flexed position
which results in narrowing of the oropharyngeal space [
33
]. Such misalignment could,
in theory, also occur from a dorsal approach, and therefore, we cannot exclude that the
mild reported dysphagia may be a consequence of the DAAS surgery. Large comparative
studies would be needed to properly establish any potential benefit/detriment of the
chosen surgical approach on such infrequent complications.
Another benefit of the dorsal approach is that it offers access to the caudal occipital
bone and C1–C2 dorsal laminae, allowing surgical interventions such as occipito-atlantal
stabilisation and/or dorsal decompressive craniotomy or laminectomy [
12
,
15
,
16
,
34
37
].
This study demonstrated that craniocervical stabilisation could be successfully imple-
mented even in the presence of complex OAAM with partial occipito-atlantal fusion
(cases 9 and 12). Such anomalies are expected to cause an exacerbated fulcrum effect on the
atlantoaxial joint, and it is, therefore, important to optimise the biomechanical properties
of the associated stabilising construct [
12
]. Anatomically, the dorsal approach provides
biomechanical advantages in that it allows the placement of the implants along the tension
surface of the vertebral column [
38
,
39
]. It also offers opportunities to extend the position
of stabilising implants further rostrally than would be possible with the ventral approach
using a titanium mesh affixed to the cranium. Most of the available literature associated
with atlanto-occipital instability describes ventral stabilisation techniques, with stabilising
constructs often limited to the atlantoaxial region. These methods generally achieved
a successful outcome, but several implant failures have also been reported [
12
,
16
,
34
36
]. To
our knowledge, cases 9 and 12 represent original reports of complex OAAM solely treated
via dorsal stabilisation.
Atlantoaxial incongruence and dorsal fibrous bands are further examples of lesions
that can benefit from a dorsal approach [
15
,
34
]. Our results, along with a previous single
study, support rigid DAAS as an effective and safe method for the management of C1–C2
incongruence [
15
]. Based on our experience, DAAS allows partial resection of the C2 dorsal
lamina/spinous process when surgical reduction of a hypoplastic C2 causes dorsal spinal
cord compression. It has also been previously argued that semirigid dorsal constructs
would not be biomechanically appropriate without proper C1–C2 joint congruency [15].
Based on our experience, the main limitation of the proposed DAAS is the limited
access to the C1–C2 synovial joint for bone grafting (dorsolateral extremities) [
5
]. Despite
using bone grafting in four cases, only two cases demonstrated convincing C1–C2 fusion
on long-term CT images, and three dogs achieved partial bone remodelling consistent
with significant ankylosis. A ventral approach may offer higher arthrodesis potential
considering that the articular surfaces are readily accessible, although fusion rates have not
yet been established to our knowledge. Another technical difficulty associated with the
proposed DAAS method is the delicate dissection required, in particular around the lateral
and intervertebral foramen. Finally, the most significant limitations of this study are the
small population size and the retrospective design.
Life 2021,11, 1039 14 of 15
5. Conclusions
This study suggests the proposed DAAS is a viable alternative to ventral techniques
and can be safely used to treat a variety of craniocervical junction disorders. We believe
that this technique has the potential to reduce complication rates related to the disruption
of vital anatomical structures located ventrally. Prospective studies would be necessary to
accurately compare complication and success rates of DAAS to that of a ventral technique.
Further investigation into the role of preoperative planning and determination of the
optimal number of stabilising cortical screws would also be beneficial.
Supplementary Materials:
The following are available online at https://www.mdpi.com/article/10
.3390/life11101039/s1, Video S1: Preoperative planning screen recording used for guidance of screw
placement (case 11).
Author Contributions: Conceptualisation, G.L.; methodology, G.L. and R.G.-Q.; software, G.L. and
R.G.-Q.; formal analysis, G.L. and J.T.; investigation, J.T., R.G.-Q., A.K., R.J.-L., V.G.N., C.R., G.L.;
data curation, J.T.; writing—original draft preparation, J.T.; writing—review and editing, all authors;
visualisation, G.L.; supervision, G.L. All authors have read and agreed to the published version of
the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement:
Ethical review and approval were waived for this study due
to its retrospective nature.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Acknowledgments:
We thank Danielle Whittaker for providing insightful suggestions and support
during the writing of this manuscript. We thank Bethan Jones for helping in the gathering of the
photographs presented in Figure 2.
Conflicts of Interest: The authors declare no conflict of interest.
References
1. Downey, R.S. An unusual cause of tetraplegia in a dog. Can. Vet. J. La Rev. Vet. Can. 1967,8, 216–217.
2. Geary, J.C.; Oliver, J.E.; Hoerlein, B.F. Atlanto axial subluxation in the canine. J. Small Anim. Pract. 1967,8, 577–582. [CrossRef]
3.
Stalin, C.; Gutierrez-Quintana, R.; Faller, K.; Guevar, J.; Yeamans, C.; Penderis, J. A review of canine atlantoaxial joint subluxation.
Vet. Comp. Orthop. Traumatol. 2015,28, 1–8. [CrossRef]
4.
Plessas, I.; Volk, H. Signalment, clinical signs and treatment of atlantoaxial subluxation in dogs: A systematic review of 336
published cases from 1967 to 2013. J. Vet. Intern. Med. 2014,28, 948.
5. McCarthy, R.J.; Lewis, D.D.; Hosgood, G. Atlantoaxial subluxation in dogs. Comp. Contin. Educ. Pract. Vet. 1995,17, 215–227.
6. Slanina, M.C. Atlantoaxial Instability. Vet. Clin. N. Am. Small Anim. Pract. 2016,46, 265–275. [CrossRef] [PubMed]
7.
Warren-Smith, C.M.; Kneissl, S.; Benigni, L.; Kenny, P.J.; Lamb, C.R. Incomplete ossification of the atlas in dogs with cervical
signs. Vet. Radiol. Ultrasound 2009,50, 635–638. [CrossRef]
8.
Takahashi, F.; Hakozaki, T.; Kouno, S.; Suzuki, S.; Sato, A.; Kanno, N.; Harada, Y.; Yamaguchi, S.; Hara, Y. Epidemiological and
morphological characteristics of incomplete ossification of the dorsal neural arch of the atlas in dogs with atlantoaxial instability.
Am. J. Vet. Res. 2018,79, 1079–1086. [CrossRef] [PubMed]
9.
Waschk, M.A.; Vidondo, B.; Carrera, I.; Hernandez-Guerra, A.M.; Moissonnier, P.; Plessas, I.N.; Schmidt, M.J.; Schnötzinger,
D.; Forterre, F.; Precht, C. Craniovertebral junction anomalies in small breed dogs with atlantoaxial instability: A multicentre
case–control study. Vet. Comp. Orthop. Traumatol. 2019,32, 033–040. [CrossRef]
10.
Cerda-Gonzalez, S.; Dewey, C.W. Congenital diseases of the craniocervical junction in the dog. Vet. Clin. N. Am. Small Anim.
Pract. 2010,40, 121–141. [CrossRef]
11.
Dewey, C.; Marino, D.; Loughin, C. Craniocervical junction abnormalities in dogs. N. Z. Vet. J.
2013
,61, 202–211. [CrossRef]
[PubMed]
12.
Galban, E.M.; Gilley, R.S.; Long, S.N. Surgical stabilization of an occipitoatlantoaxial malformation in an adult dog. Vet. Surg.
2010,39, 1001–1004. [CrossRef] [PubMed]
13.
Petite, A.; McConnell, F.; De Stefani, A.; McKee, M.; Dennis, R. Abstracts from the annual conference of the european association
of veterinary diagnostic imaging. Congenital occipito-atlanto-axial malformation in five dogs. Vet. Radiol. Ultrasound
2009
,
50, 118. [CrossRef]
Life 2021,11, 1039 15 of 15
14.
Watson, A.G.; de Lahunta, A.; Evans, H.E. Morphology and embryological interpretation of a congenital occipito-atlanto-axial
malformation in a dog. Teratology 1988,38, 451–459. [CrossRef]
15.
Dolera, M.; Malfassi, L.; Pavesi, S.; Finesso, S.; Mazza, G.; Marcarini, S.; Sala, M.; Carrara, N.; Bianchi, C.; Gambino, J.M.
Computed tomography, magnetic resonance imaging and a novel surgical approach of atlanto-axial instability with incongruence
in dogs. J. Vet. Med. Sci. 2017. [CrossRef]
16.
Read, R.; Brett, S.; Cahill, J. Surgical treatment of occipito-atlanto-axial malformation in the dog. Aust. Vet. Pract.
1987
,17,
184–189.
17.
Havig, M.E.; Cornell, K.K.; Hawthorne, J.C.; McDonnell, J.J.; Selcer, B.A. Evaluation of nonsurgical treatment of atlantoaxial
subluxation in dogs: 19 cases (1992–2001). J. Am. Vet. Med. Assoc. 2005,227, 257–262. [CrossRef]
18.
Denny, H.R.; Gibbs, C.; Waterman, A. Atlanto-axial subluxation in the dog—a review of 30 cases and an evaluation of treatment
by lag screw fixation. J. Small Anim. Pract. 1988,29, 37–47. [CrossRef]
19.
Jeffery, N.D. Dorsal cross pinning of the atlantoaxial joint: New surgical technique for atlantoaxial subluxation. J. Small Anim.
Pract. 1996,37, 26–29. [CrossRef]
20.
LeCouteur, R.A.; McKeown, D.; Johnson, J.; Eger, C.E. Stabilization of atlantoaxial subluxation in the dog, using the nuchal
ligament. J. Am. Vet. Med. Assoc. 1980,177, 1011–1017. [PubMed]
21.
Pujol, E.; Bouvy, B.; Omana, M.; Fortuny, M.; Riera, L.; Pujol, P. Use of the Kishigami atlantoaxial tension band in eight toy breed
dogs with atlantoaxial subluxation. Vet. Surg. 2010,39, 35–42. [CrossRef]
22.
Aikawa, T.; Shibata, M.; Fujita, H. Modified ventral stabilization using positively threaded profile pins and polymethylmethacry-
late for atlantoaxial instability in 49 dogs. Vet. Surg. 2013,42, 683–692. [CrossRef]
23.
Thomas, W.B.; Sorjonen, D.C.; Simpson, S.T. Surgical management of atlantoaxial subluxation in 23 dogs. Vet. Surg.
1991
,20,
409–412. [CrossRef] [PubMed]
24.
Beaver, D.P.; Ellison, G.W.; Lewis, D.D.; Goring, R.L.; Kubilis, P.S.; Barchard, C. Risk factors affecting the outcome of surgery for
atlantoaxial subluxation in dogs: 46 cases (1978–1998). J. Am. Vet. Med. Assoc. 2000,216, 1104–1109. [CrossRef] [PubMed]
25.
Schulz, K.S.; Waldron, D.R.; Fahie, M. Application of ventral pins and polymethylmethacrylate for the management of atlantoaxial
instability: Results in nine dogs. Vet. Surg. 1997,26, 317–325. [CrossRef]
26.
Platt, S.R.; Chambers, J.N.; Cross, A. A modified ventral fixation for surgical management of atlantoaxial subluxation in 19 dogs.
Vet. Surg. 2004,33, 349–354. [CrossRef]
27.
Sanders, S.G.; Bagley, R.S.; Silver, G.M.; Moore, M.; Tucker, R.L. Outcomes and complications associated with ventral screws,
pins, and polymethyl methacrylate for atlantoaxial instability in 12 dogs. J. Am. Anim. Hosp. Assoc.
2004
,40, 204–210. [CrossRef]
[PubMed]
28.
Toni, C.; Oxley, B.; Behr, S. Atlanto-axial ventral stabilisation using 3D-printed patient-specific drill guides for placement of
bicortical screws in dogs. J. Small Anim. Pract. 2020,61, 609–616. [CrossRef] [PubMed]
29.
Kamishina, H.; Sugawara, T.; Nakata, K.; Nishida, H.; Yada, N.; Fujioka, T.; Nagata, Y.; Doi, A.; Konno, N.; Uchida, F.; et al.
Clinical application of 3D printing technology to the surgical treatment of atlantoaxial subluxation in small breed dogs. PLoS
ONE 2019,14, e0216445. [CrossRef]
30.
Leblond, G.; Barnard, L.; Gutierrez-Quintana, R. Proceedings 31st Symposium ESVN-ECVN. Evaluation of a preoperative implant
placement planning method for canine atlantoaxial stabilisation. J. Vet. Intern. Med. 2020,34. [CrossRef]
31.
Sanchez-Masian, D.; Lujan-Feliu-Pascual, A.; Font, C.; Mascort, J. Dorsal stabilization of atlantoaxial subluxation using non-
absorbable sutures in toy breed dogs. Vet. Comp. Orthop. Traumatol. 2014,27, 62–67. [CrossRef] [PubMed]
32.
Leblond, G.; Gaitero, L.; Moens, N.M.; Zur Linden, A.; James, F.M.; Monteith, G.; Runciman, J. Computed tomography analysis of
ventral atlantoaxial optimal safe implantation corridors in 27 dogs. Vet. Comp. Orthop. Traumatol. 2017,30, 413–423. [PubMed]
33.
Wang, X.; Chou, D.; Jian, F. Influence of postoperative O-C2 angle on the development of dysphagia after occipitocervical fusion
surgery: Results from a retrospective analysis and prospective validation. World Neurosurg.
2018
,116, e595–e601. [CrossRef]
[PubMed]
34.
Cerda-Gonzalez, S.; Dewey, C.W.; Scrivani, P.V.; Kline, K.L. Imaging features of atlanto-occipital overlapping in dogs. Vet. Radiol.
Ultrasound Off. J. Am. Coll. Vet. Radiol. Int. Vet. Radiol. Assoc. 2009,50, 264–268. [CrossRef] [PubMed]
35.
Takahashi, F.; Hakozaki, T.; Kouno, S.; Suzuki, S.; Sato, A.; Kanno, N.; Harada, Y.; Yamaguchi, S.; Hara, Y. Atlantooccipital
overlapping and its effect on outcomes after ventral fixation in dogs with atlantoaxial instability. J. Vet. Med. Sci.
2018
,80, 526–531.
[CrossRef]
36.
Fujita, A.; Nishimura, R. Surgical stabilization of the atlanto-occipital overlap with atlanto-axial instability in a dog. Jpn. J. Vet.
Res. 2016,64, 141–145.
37.
Rylander, H.; Robles, J.C. Diagnosis and treatment of a chronic atlanto-occipital subluxation in a dog. J. Am. Anim. Hosp. Assoc.
2007,43, 173–178. [CrossRef] [PubMed]
38. Evans, H.E.; de Lahunta, A. Miller’s anatomy of the dog, 4th ed.; Elsevier/Saunders: St. Louis, MO, USA, 2012.
39.
Leblond, G. Canine Atlantoaxial Ventral Stabilization: Computed Tomography Analysis of Optimal Safe Implantation Corridors
and Comparison of the Technical Outcome and Biomechanical Properties of 3 Surgical Techniques. Ph.D. Thesis, University of
Guelph, Guelph, ON, Canada, 2015. Available online: http://hdl.handle.net/10214/9354 (accessed on 27 September 2021).
Article
Full-text available
Objectives: To describe a novel transforaminal approach for surgical excision of the atlantoaxial (AA) band and examine its feasibility, safety, and mechanical advantages in an ex vivo study and clinical cases. Samples: 26 canine cadavers and 2 canine patients with AA bands. Procedures: The transforaminal approach via the first intervertebral foramen was designed to avoid damaging the dorsal AA ligament (DAAL) and dorsal laminas to maintain joint stability. The cadaveric study started on December 2020 and lasted 3 months. The ligamentum flavum (LF) was removed using a novel approach; then, gross examination was conducted to verify the potential damage to the spinal cord and associated structures and the adequacy of LF removal. Subsequently, the ex vivo tension test of the DAAL was conducted to establish whether the approach induced mechanical damage to the ligaments. Finally, 2 dogs diagnosed with an AA band were surgically treated with the transforaminal approach. Results: In the cadaveric study, postsurgical evaluation verified the subtotal removal of LF without damage to the dura mater. There were no significant differences in the mechanical properties of the DAAL, including the ultimate strength (P = .645) and displacement (P = .855), between the surgical and intact groups during the ex vivo tension test. In clinical cases, clinical signs and neurologic grades improved until the final follow-up. Clinical relevance: The described surgical procedure using a transforaminal approach appears to sufficiently permit the removal of an AA band while reducing damage to the DAAL and spinal cord. Our study highlights the feasibility of the transforaminal approach.
Article
Full-text available
Atlantoaxial instability (AAI)/subluxation commonly occurs in small breed dogs. Ventral stabilization techniques using screws and/or pins and a plate or, more commonly, polymethylmethacrylate are considered to provide the most favorable outcome. However, the implantation of screws of sufficient sizes for long-term stability becomes challenging in toy breed dogs (e.g. <2 kg). We herein report the application of 3D printing technology to implant trajectory planning and implant designing for the surgical management of AAI in 18 dogs. The use of our patient-specific drill guide templates resulted in overall mean screw corridor deviations of less than 1 mm in the atlas and axis, which contributed to avoiding iatrogenic injury to the surrounding structures. The patient-specific titanium plate was effective for stabilizing the AA joint and provided clinical benefits to 83.3% of cases (15/18). Implant failure requiring revision surgery occurred in only one case, and the cause appeared to be related to the suboptimal screw-plate interface. Although further modifications are needed, our study demonstrated the potential of 3D printing technology to be effectively applied to spinal stabilization surgeries for small breed dogs, allowing for the accurate placement of screws and minimizing peri- and postoperative complications, particularly at anatomical locations at which screw corridors are narrow and technically demanding.
Article
Full-text available
We compared clinical outcomes after ventral fixation in dogs with atlantoaxial instability (AAI) on the basis of the presence or absence of atlantooccipital overlapping (AOO). Of 41 dogs diagnosed with AAI and treated ventral fixation, 12 exhibited AOO (AOO group), whereas 29 did not (non-AOO group). The AOO group had significantly higher neurological scores before (P=0.024) and 1 month after (P=0.033) surgery compared with the non-AOO group; however, no significant differences were observed between the groups 2 months after surgery. The presence of complicating AOO affected the clinical signs for dogs with AAI, but did not directly affect the outcome of surgical stabilization of AAI.
Article
Full-text available
Atlanto-axial (AA) instability due to ligament insufficiency is a common cause of cervical spinal cord compression in toy breeds. However, in some dogs a difference in size between the atlas and the axis leads to joint incongruence that exacerbates AA subluxation and makes surgical treatment challenging. Twelve dogs with AA instability with incongruence were enrolled in a single institution prospective observational study. Computed tomography (CT) and magnetic resonance imaging (MRI) of the AA joint were compared to a retrospectively reviewed control group. A novel surgical approach consisting of a dorsal internal fixation technique was performed in six dogs. For affected dogs, the mean normalised difference between the dorso-ventral atlas canal and the dorso-ventral axis canal was 29.67% (median of 35.07%, standard deviation 25.64%), while in normal dogs a mean difference of 4.67% (median of 3.95%, standard deviation 5.21%) was observed. On MRI, 12/12 affected dogs had spinal cord compression, which was classified as reducible (3/12), partially reducible (6/12) and non-reducible (3/12). In surgically operated dogs, follow-up CT showed a partial or complete reduction of the previous spinal cord compression with a consistent amelioration or resolution of the presenting complaints. The proposed surgical technique was safe and effective in dogs with partially or completely reducible spinal cord compression.
Article
Objective To report outcome and complications following atlanto‐axial stabilisation by polymethylmethacrylate applied to screws placed using 3D‐printed patient‐specific drill guides. Materials and Methods Case series of dogs treated with this technique between May 2016 and August 2018 including pre‐ and post‐operative modified Frankel score, imaging and complications. Screw placement was graded using a modified Zdichavsky classification based on post‐operative CT. Telephone follow‐up was obtained for surviving dogs. Results Twelve cases were included. At presentation, modified Frankel score was 3 in five dogs and 4 in seven dogs on presentation. Of 61 bicortical screws placed, 57 (93%) were fully contained within the pedicle and vertebral body and four (7%) partially breached the medial pedicle wall. Post‐operative CT revealed good alignment of C1 and C2 in all planes. Reversible perioperative adverse events were described in five of 12 dogs and two dogs were euthanased shortly after discharge. At 18 to 50 days after surgery eight dogs had improved neurological status. Neurological status remained unchanged in the remaining two dogs. All dogs were reported ambulatory and pain‐free at telephone follow‐up (median 405 days post‐surgery, range 180 to 780 days). Clinical Significance This technique resulted in safe bicortical screw placement in dogs with atlanto‐axial subluxation.
Conference Paper
Sensory nerve conduction (SNC) threshold of the infraorbital nerve wasreported to be decreased in horses with idiopathic headshaking (IHS)with a stimulation threshold of≤5mA in comparison to healthy horses.In the current study, a prospective ongoing trial, horses were exam-ined using standardized diagnostic procedures to differentiate symp-tomatic headshaking (SHS) from IHS. In all patients, SNC of theinfraorbital nerve was measured bilaterally under general anesthesia.Sensory nerve conduction velocity, amplitude of the sensory nerveaction potential (SNAP) and stimulation threshold were recorded. Thetechnique described by Aleman et al. 2013 had to be modified to min-imize artefacts. All examinations were performed with written ownerconsent according to university guidelines.Twelve horses with equine headshaking were included: two with SHSand 10 with IHS. In total, 22 measurements were obtained. In thehorses with SHS, SNAP were evoked at 15mA (left side) and 20 mA(right side), as well as at 7.5mA bilaterally. In horses with presumedIHS, the median threshold was 10 mA (5 mA - 15 mA). Despite smallsubdermal hemorrhage no major side effects or deterioration of clini-cal signs were detected after the measurements.In conclusion, measurement of SNC of the infraorbital nerve is a feasibleand safe method to further support the diagnosis of IHS in horses, althoughthe displayed threshold might vary depending on the electrodiagnostic lab-oratory and laboratory specific thresholds have to be determined.
Article
Objective: The main purpose of this study was to define criteria to systemically describe craniovertebral junction (CVJ) anomalies and to report the prevalence of CVJ anomalies in small breed dogs with and without atlantoaxial instability (AAI). Methods: Retrospective multicentre matched case-control study evaluating magnetic resonance imaging and computed tomographic images of small breed dogs with and without AAI for the presence of CVJ anomalies. Results: One hundred and twenty-two dogs were enrolled (61 with and 61 without AAI). Only dogs with AAI had dens axis anomalies such as separation (n = 20) or a short-rounded conformation (n = 35). Patients with AAI were more likely to have atlantooccipital overlapping based on transection of McRae's line by the dorsal arch of the atlas (odds ratio [OR] = 5.62, p < 0.01), a transection of Wackenheim's clivus line (OR = 41.62, p < 0.01) and rostral indentation of the occipital bone (OR = 2.79, p < 0.05). Patients with AAI were less likely to have a larger clivus canal angle (OR = 0.94, p < 0.01) and larger occipital bone lengths (OR = 0.89, p < 0.05). Clinical significance: Small breed dogs with AAI are more likely to have other CVJ anomalies such as atlantooccipital overlapping or dens anomalies. The grade of brachycephaly does not differ between patients with and without AAI. Certain objective criteria from human literature were found useful for the assessment of both AAI and atlantooccipital overlapping such as McRae's line, Wackenheim's clivus line, and clivus canal angle. The classification criteria used can help to evaluate CVJ anomalies in a more systematic way.
Article
OBJECTIVE To retrospectively evaluate the epidemiological and morphological features and outcome of surgical treatment of incomplete ossification of the dorsal neural arch of the atlas (IODA) in dogs with atlantoaxial instability (AAI). ANIMALS 106 AAI-affected dogs that underwent ventral fixation of the atlantoaxial joint. PROCEDURES Medical records and CT images for each dog were reviewed. Dogs were allocated to 1 of 2 groups on the basis of the presence or absence of IODA or of dens abnormalities (DAs) in CT images. RESULTS Of the 106 dogs with AAI, 75 had and 31 did not have IODA; 70 had and 36 did not have DAs. Incomplete ossification was present in the cranialmost, central, or caudalmost portion of the dorsal neural arch of the atlas in 59, 39, and 28 dogs, respectively; 2 or 3 portions were affected in 29 and 11 dogs, respectively. The mean CT value (in Hounsfield units) for the midline of the dorsal neural arch of the atlas in dogs with IODA was significantly lower than that for the same site in the dogs without IODA. The mean age at surgery for dogs with central IODA was significantly higher than that of the non-IODA group. The severity of spinal cord injury before or after atlantoaxial ventral fixation did not differ between the IODA and non-IODA groups. CONCLUSIONS AND CLINICAL RELEVANCE Results indicated that concomitant DAs or IODA is common in dogs with AAI. In dogs with incomplete ossification in the central part of the dorsal neural arch of the atlas, surgical treatment of AAI generally occurs at a middle to advanced age.
Article
Background: Postoperative dysphagia is a known complication of anterior cervical surgery, but its incidence and possible mechanisms are seldom reported after occipitocervical fusion (OCF). Our objective was to study the relationship between craniocervical alignment and the development of dysphagia after OCF for the treatment of basilar invagination with atlantoaxial instability. Methods: The study was consisted of a retrospective series and a prospective series. A total of 78 patients who underwent OCF (30 male, 48 female) were reviewed in the retrospective series. The presence and duration of postoperative dysphagia were recorded via in person questionnaire or telephone interview. Sagittal reconstructed computed tomography images before and after the procedure were collected. The O-C2 angle and C2-C7 angle were measured. The relationship of these parameters and their influence to the incidence of dysphagia were analyzed. The patients were grouped according to whether they developed postoperative dysphagia (group A) or not (group B). A prospective case series of 27 patients (group C) were reported to verify the influence of O-C2 angle on postoperative dysphagia. Results: In the retrospective case series, 19 patients (24.4%) complained of postoperative dysphagia after OCF. The change in the O-C2 angle was significantly lower in group A than in group B (p< 0.001). In the prospective case series, only 1 patient (3.7%) complained of postoperative dysphagia. Conclusions: O-C2 angle plays an important role in the development of postoperative dysphagia after OCF procedure. Careful intraoperative alignment of the O-C2 angle may help reduce the incidence and severity of postoperative dysphagia after OCF.
Article
Objectives Ventral atlantoaxial stabilization techniques are challenging surgical procedures in dogs. Available surgical guidelines are based upon subjective anatomical landmarks, and limited radiographic and computed tomographic data. The aims of this study were (1) to provide detailed anatomical descriptions of atlantoaxial optimal safe implantation corridors to generate objective recommendations for optimal implant placements and (2) to compare anatomical data obtained in non-affected Toy breed dogs, affected Toy breed dogs suffering from atlantoaxial instability and non-affected Beagle dogs. Methods Anatomical data were collected from a prospectively recruited population of 27 dogs using a previously validated method of optimal safe implantation corridor analysis using computed tomographic images. Results Optimal implant positions and three-dimensional numerical data were generated successfully in all cases. Anatomical landmarks could be used to generate objective definitions of optimal insertion points which were applicable across all three groups. Overall the geometrical distribution of all implant sites was similar in all three groups with a few exceptions. Clinical Significance This study provides extensive anatomical data available to facilitate surgical planning of implant placement for atlantoaxial stabilization. Our data suggest that non-affected Toy breed dogs and non-affected Beagle dogs constitute reasonable research models to study atlantoaxial stabilization constructs.
Article
The atlanto-occipital (AO) overlap in combination with atlanto-axial (AA) instability was found in a dog. We hypothesized that ventral fixation of the AA junction can stabilize the atlas and prevent AO overlap by reviewing our past cases with AA instability. A standard ventral fixation of the AA junction using stainless k-wires and polymethyl methacrylate (PMMA) was performed. The dog fully recovered, and no complication was noted. The results of the postoperative CT imaging supported our hypothesis. The ventral fixation of the AA junction is a feasible treatment option for similar cases, although craniocervical junction abnormalities (CJA) including AA instability are varied, and careful consideration is required for each case.