Received: 30 December 2018 Revised: 15 February 2019 Accepted: 1 March 2019

DOI: 10.1111/vru.12753

O R I G I N A L I N V E S T I G AT I O N

Congenital malformations of the lumbosacral vertebral column

are common in neurologically normal French Bulldogs, English
Bulldogs, and Pugs, with breed-specific differences
Simon Bertram1 Gert ter Haar1,2 Steven De Decker1

1 Department of Clinical Science and Services,

Royal Veterinary College, University of London, Abstract

Hatfield, UK Congenital vertebral malformations are common findings on diagnostic imaging of the vertebral
2 Specialistische Dierenkliniek Utrecht, Utrecht,
column in “screw-tailed” brachycephalic dogs. The aims of this study were to evaluate the preva-
The Netherlands lence and anatomical characteristics of lumbosacral congenital vertebral malformations in French
Bulldogs, English Bulldogs, and Pugs presenting for problems unrelated to spinal disease, as well
Correspondence
Simon Bertram, Queen Mother Hospital for as possible associations with the degree of tail malformation, lumbosacral intervertebral disc her-
Animals (QMHA), The Royal Veterinary College, niation, or spondylosis deformans. In this retrospective cross-sectional study, CT scans of verte-
Hawkshead Lane, North Mymms, Hatfield,
brae L6 to S3 and of the coccygeal vertebrae were reviewed for type of congenital vertebral mal-
Hertfordshire, AL9 7TA, UK.
Email: [email protected] formations (hemivertebrae, block vertebrae, lumbosacral transitional vertebrae, and spina bifida),
lumbosacral intervertebral disc herniation, lumbosacral spondylosis deformans, and degree of tail
malformation. In 76 (51.0%) of the 149 included dogs (53 French Bulldogs, 37 English Bulldogs,
and 59 Pugs) at least one type of congenital vertebral malformations was found, with lumbosacral
transitional vertebrae being the most common (34.2%). There was a significantly higher preva-
lence of lumbosacral transitional vertebrae (54.2%) and lower prevalence of hemivertebrae (1.7%)
in Pugs compared to English (13.5% and 24.3%, respectively) and French Bulldogs (26.4% and
32.0%, respectively). Tail malformation was significantly more severe in dogs with evidence of
hemivertebrae. Congenital vertebral malformations are a common finding in the lumbosacral ver-
tebral column of French Bulldogs, English Bulldogs, and Pugs. These anatomical variances need
to be considered when interpreting diagnostic studies and when planning for neurosurgical and
neurodiagnostic procedures. Furthermore, this study suggests a possible association between the
degree of tail malformation and lumbosacral hemivertebrae.

KEYWORDS
brachycephalic, butterfly vertebra, hemivertebra, screw-tail, spina bifida, transitional vertebra

1 INTRODUCTION most often incidental findings, they may cause a progressive myelopa-
thy or contribute to the development of other clinically relevant
Congenital vertebral malformations are common findings on diagnos- pathologies.2–10
tic imaging of the vertebral column in “screw-tailed” brachycephalic Several studies have demonstrated a high prevalence (80.7% to
dogs. There is controversy about the most appropriate terminology 87.7%) of thoracic congenital vertebral malformations in brachy-
and classification of vertebral malformations in veterinary medicine.1 cephalic breeds, such as the English Bulldog, the French Bulldog, and
The most commonly reported abnormalities are: block vertebrae, the Pug.2–4,8,10–12 Thoracolumbar (30.9%) and cervicothoracic (46.0%)
hemivertebrae (this includes “true” hemivertebrae, butterfly verte- transitional vertebrae as well as caudal articular process dysplasia
brae, and wedge-shaped vertebrae), transitional vertebrae, and neural (97.0%) are considered common especially in Pugs,10,12,13 hemiver-
tube defects.1,2 Although congenital vertebral malformations are tebrae in French Bulldogs (78.0- 93.5%),3,8,10 and lumbosacral spina
bifida in English Bulldogs.14–16
While the cause and pathogenesis of these malformations remain
Presented in poster form at the 31th Annual Symposium of the European Society of Veterinary
Neurology–European Conference of Veterinary Neurology, Copenhagen, September 2018.
unknown, most studies support multifactorial etiologies.7,8,17–21 A

400 
c 2019 American College of Veterinary Radiology wileyonlinelibrary.com/journal/vru Vet Radiol Ultrasound. 2019;60:400–408.

Augusto Cavalheiro Vieira Schonfelder - [email protected] - IP: 189.101.142.197

 BERTRAM ET AL . 401

recent study has identified a mutation in the gene DVL2 linked with used. After transverse images were obtained, multiplanar sagittal and
thoracic and caudal vertebral column malformations in English Bull- dorsal reformatted images, as well as three-dimensional reformatted
dogs, French Bulldogs, and Boston Terriers but not in Pugs.22 In French images, were created and reviewed. The vertebrae from L6 to S3 were
bulldogs, an association has been shown to exist between the degree of individually assessed for presence of hemivertebra, block vertebra,
tail malformation and the severity of thoracic hemivertebrae and it has spina bifida, and lumbosacral transitional vertebrae for each dog.
been suggested that the selection of dogs with a “screw-tailed” phe- Additionally, evidence of lumbosacral intervertebral disc herniation
notype for breeding will cause a further increase in the prevalence of and spondylosis deformans was recorded for each dog.
hemivertebrae.8 Hemivertebrae were defined as a defect in vertebral body for-
No studies have yet focused on prevalence and anatomical char- mation leading to a part of the vertebral body being deficient
acteristics of lumbosacral congenital vertebral malformations. There- (Figure 1A).1,2 Block vertebra was defined as a vertebral segmentation
fore, this study aims to evaluate the prevalence and anatomical char- defect with evidence of complete or partial fusion of two adjacent ver-
acteristics of congenital vertebral malformations in French Bulldogs, tebral bodies without an intervertebral disc space between them.1,2
English Bulldogs, and Pugs presenting for problems unrelated to spinal Spina bifida was defined as a congenital defect in neural tube closure
disease. Additionally, an association between these congenital verte- leading to varying degrees of dorsal cleft formation affecting the
bral malformations and the degree of tail malformation, lumbosacral vertebral arch and the spinous process (Figure 1B).1 A lumbosacral
intervertebral disc herniation, and spondylosis deformans were eval- transitional vertebra was defined as a vertebra showing features of
uated. It was hypothesized that (1) lumbosacral congenital vertebral both lumbar and sacral vertebrae with or without an intervertebral disc
malformation would be common in all three evaluated breeds, but that space between the lumbosacral transitional vertebra and the first true
(2) breed-specific differences would exist concerning the prevalence sacral vertebra (Figure 2).1 Further classification of the lumbosacral
and anatomical localization of the individual types of malformation. transitional vertebrae was done by adapting a proposed radiographic
Furthermore, it was hypothesized that there would be (3) a significant scheme individually for each side23 : Type 1 transverse process resem-
association between the severity of tail malformation and the presence bles a lumbar process; Type 2 transverse process is partly in contact
of lumbosacral hemivertebrae and (4) between the presence of lum- with the ilium with a varying degree of the tip resembling a free trans-
bosacral transitional vertebrae and intervertebral disc herniation and verse process; Type 3 transverse process is completely in contact with
spondylosis deformans. the ilium, or resembles a sacral vertebra; and Type 4, normal formation
of the vertebra with evidence of separation of at least the S1 and S2
spinous processes.23 To take asymmetrical changes into account lum-
bosacral transitional vertebrae were typed overall as type 1 (lumbar
2 MATERIAL AND METHODS type) if being bilateral type 1 (1/1) or a combination of type 1 and 2
(1/2, 2/1; Figure 2A and B), as type 2 (intermediate type) if bilateral
This retrospective descriptive cross-sectional study was approved type 2 (2/2) or a combination of type 1 and 3 (1/3, 3/1), as type 3 (sacral
by the Clinical Ethical Review Board of the Royal Veterinary College, type) if bilateral type 3 (3/3) or a combination of type 2 and 3 (2/3, 3/2)
University of London (URN SR2017-1144). The electronic medical (Figure 2C and D), and as type 4 if bilateral type 4 (4/4; Figure 2E and
database of the Small Animal Referral Hospital, Royal Veterinary F). Intervertebral disc herniation was defined as evidence of material
College was searched for French Bulldogs, English Bulldogs, and Pugs hyperattenuating to epidural fat signal in the vertebral canal dorsal to
that underwent CT from October 2010 to November 2016 for reasons and confined to the intervertebral disc space and was classified as with
unrelated to spinal disease. Dogs were included if CT included the lum- or without evidence of compression or displacement of neural struc-
bar vertebrae six and seven, the sacrum, and the coccygeal vertebrae, tures (spinal cord, dural sac, or nerve roots). Spondylosis deformans
and if the medical files were available for review. Dogs were excluded was defined as presence of osteophytes on the edge of the epiphysis up
if presented for further evaluation of orthopedic or neurological to osteophytes bridging the intervertebral disc space.24 No differenti-
disease, if a gait abnormality was recorded in the medical file, or if the ation between grades of intervertebral disc herniation or spondylosis
CT studies were not available for review or of insufficient quality. was implemented in this study. Tails were assigned an overall grade
Information retrieved from the medical files included signalment, considering prevalence and degree of spina bifida, block vertebrae,
presenting clinical signs, and the final diagnosis of the investigation. hemivertebrae, sacrocaudal transitional vertebra, vertebra curva
Computed tomographic images were reviewed by one investigator (normal vertebral anatomy with generalized curvature), degree of
(S.B., Veterinary specialist-in-training in Veterinary Neurology and resulting tail kinking, and length of the tail from 0 to 3 (0: normal verte-
Neurosurgery under direct supervision of a board-certified neurolo- bral anatomy with or without fluent curvature of the tail and vertebra
gist, S.D.D.) in a randomized order using a random sequence generator curva, more than 15 vertebrae; 1: presence of sacrocaudal transitional
(random.org) and blinded to any patient data. The investigator was vertebra, presence of focal spina bifida, block vertebrae, hemiverte-
aware that none of the included dogs had recorded orthopedic or brae, and/or abrupt kinking of the tail, more than 15 vertebrae; 2: pres-
neurological signs. After retrieval from PACS to a workstation (Mac- ence of multifocal spina bifida, block vertebrae, hemivertebrae, and
Book Pro 13 inch, 2015, Apple Inc.), a commercially available DICOM abrupt kinking of the tail, less than 15 vertebra; 3: presence of diffuse
viewing software (Horos, version 1.1.7., www.horosproject.org) was spina bifida, block vertebrae, hemivertebrae, and abrupt kinking of the

Augusto Cavalheiro Vieira Schonfelder - [email protected] - IP: 189.101.142.197

 402 BERTRAM ET AL .

F I G U R E 1 Transverse CT image at the level of the: A, S1vertebral body showing hemivertebra (butterfly vertebra formation; arrow marks the
ventro-median aplasia); B, S2 vertebra showing spina bifida (Arrow marks dorso-median cleft in the vertebral arch and the missing spinous
process)

tail, less than 10 vertebra; Figure 3). The included caudal pathologies obstructive airway syndrome (n = 115), neoplastic disease (n = 14),
were based on previously published radiological studies.25,26 gastrointestinal disease (n = 8), ear disease (n = 5), other respiratory
Data were recorded using a spreadsheet (Microsoft Excel for Mac, disease (n = 4), trauma (n = 2), and cardiac disease (n = 1). All CT scans
Version 15.33). Statistical tests were performed by the first author were acquired using a 16-slice scanner (Mx8000 IDT, Philips, Best, the
(S.B.) and a statistical analysis software (SPSS Statistics for OSx, Ver- Netherlands) with dogs in sternal recumbency under anesthesia or
sion 24.0, IBM Corp, Armonk, NY) was used for data analyses. Data sedation. The following parameters were used: 16 × 1.5 mm collima-
were tested for Gaussian distribution using Kolmogorov-Smirnov test tion, helical scan mode, 2 or 3 mm reconstruction slice thickness with
to enable the selection of the appropriate statistical test. Interbreed 1 to 1.5 mm overlap, 2 s tube rotation time, 120 to 180 mA depending
differences between the prevalence of abnormalities were tested on patient size, 120 to140 kVp, 500 mm acquisition field of view, with
using Fisher’s exact test, corrected for multiple comparison according reconstruction field of view dependent on patient body size (varying
to Bonferroni. Fisher’s exact test was also used to compare prevalence between 200 and 250 mm), and 512 × 512 matrix.
of intervertebral disc herniation and spondylosis deformans between Of the 149 dogs, 76 (51.0%) had evidence of one or more lum-
dogs with and without evidence of lumbosacral transitional verte- bosacral congenital vertebral malformations. Detailed results about
bra or hemivertebra, and to compare the degree of tail malformation the prevalence of the different congenital vertebral malformations and
between dogs with and without evidence of lumbosacral transitional their subtypes, the degree of tail malformation and the prevalence of
vertebra or hemivertebra. Mann-Whitney U test and a binary logistic lumbosacral intervertebral disc herniation and spondylosis deformans
regression model were used to test age and breed as covariates for, are listed in Tables 1 and 2. Lumbosacral transitional vertebra was
and its effects on, the prevalence of intervertebral disc herniation and the most common type of congenital vertebral malformation (51/149
spondylosis deformans. Statistical significance was defined as P < .05. dogs [34.2%]) followed by hemivertebra (27/149 [18.1%]) and spina
bifida (15/149 [10.1%]). There was no evidence of block vertebrae in
any of the dogs. The first sacral vertebra was most frequently affected
by hemivertebra (16/37, 43.2%) followed by the second (10/37,
3 RESULTS 27.0%) and third (8/37, 21.6%; Figure 1A). Spina bifida was most
commonly seen at the third sacral vertebra (12/24, 50.0%) followed
3.1 Included animals by the second (9/24, 37.5%) and first (3/24, 12.5%; Figure 1B). The
most severe degree of tail malformation (grade 3) was seen in 62
A total of 149 dogs were included in the study, consisting of 53 French
of 149 (41.6%) whereas normal tail morphology (grade 0) was seen
Bulldogs, 37 English Bulldogs, and 59 Pugs. Included animals under-
in 49 (32.9%) dogs. Lumbosacral intervertebral disc herniation was
went CT imaging for a variety of reasons, including brachycephalic
seen in 72 dogs (60.5%) causing displacement of neural structures in

Augusto Cavalheiro Vieira Schonfelder - [email protected] - IP: 189.101.142.197

 BERTRAM ET AL . 403

F I G U R E 2 Three-dimensional reformatted CT studies showing a dorsal and ventral view of the lumbosacral vertebral column and pelvis. The
arrow marks the gap between the spinous processes of the lumbosacral transitional vertebra and the first sacral vertebra. The arrowhead marks
the gap between the lumbosacral transitional vertebra and the first sacral vertebral body. A and B, Type 1 lumbosacral transitional vertebra (left:
type 2, right: type 1). Furthermore, there is evidence of ventral and median aplasia of the vertebral body of S1. C and D, Type 3 lumbosacral
transitional vertebra (left: type 2, right: type 3). E and F, Type 4 lumbosacral transitional vertebra

13 of them (18.1%). Detailed breed-specific results are also listed in (P < .0001) with Pugs being less often affected compared to French
Tables 1 and 2. (P < .0001; odds ratio = 0.036; 95% confidence interval, 0.0034-0.23)
and English Bulldogs (P < 0.001; odds ratio = 0.053; 95% confidence
interval, 0.0047-0.37). There was no significant difference between

3.2 Comparison between breeds French and English Bulldogs (P > .05). Breed did not have a signifi-
cant influence on the spatial distribution of hemivertebra (P > .05).
The group of 53 French Bulldogs consisted of 34 males (21 neutered) The prevalence and spatial distribution of spina bifida did not signifi-
and 19 females (nine neutered), aged between 5 and 133 months cantly differ between breeds (P > .05 for both; Table 1). Lumbosacral
(median 16 months). The group of 37 English Bulldogs consisted of transitional vertebra was the most common congenital vertebral mal-
23 males (18 neutered) and 14 females (11 neutered), aged between formation in Pugs (32/59, 54.2%) and the prevalence was significantly
4 and 132 months (median 14 months). The group of 59 Pugs con- different in the three analyzed breeds (P < .0001) with Pugs being
sisted of 34 males (15 neutered) and 25 females (17 neutered), aged more often affected than French (P < .01; odds ratio = 3.30; 95% con-
between 7 and 151 months (median 37 months). Pugs were sig- fidence interval, 1.51-7.35) and English Bulldogs (P < 0.0001; odds
nificantly older compared to English (P < .05) and French Bulldogs ratio = 7.59; 95% confidence interval, 2.53-19.55). There was no sig-
(P < .001). nificant difference between French and English Bulldogs (P > .05). The
Breed did not have a significant influence on the overall preva- prevalence of specific subtypes of lumbosacral transitional vertebra
lence of congenital vertebral malformations (P > .05). Hemivertebra (P > .05) or the lumbosacral transitional vertebra being symmetrical
was the most common congenital vertebral malformation in French or asymmetrical did not significantly differ between breeds (P > .05;
(17/53, 32.0%) and English Bulldogs (9/37, 24.3%; Table 1) and breed Table 2). The degree of tail malformation was significantly different in
did have a significant influence on the prevalence of hemivertebra the three analyzed breeds (P < .0001 for each), with Pugs having the

Augusto Cavalheiro Vieira Schonfelder - [email protected] - IP: 189.101.142.197

 404 BERTRAM ET AL .

FIGURE 3 Three-dimensional reformatted CT studies showing the different grades of tail malformations. A, grade 0. B, grade 1. C, grade 2. D,
grade 3

lowest degree and French Bulldogs the highest degree of tail malfor- 3.3 Associations between congenital vertebral
mation (Table 1). Lumbosacral intervertebral disc herniation was most malformations and the degree of tail malformation,
commonly seen in English (24/37, 64.9%) and French Bulldogs (28/53, lumbosacral intervertebral disc herniation, and
52.8%) and the prevalence was significantly different in the three ana-
spondylosis deformans
lyzed breeds (P < .01), with English Bulldogs being more often affected
than Pugs (P < .01; odds ratio = 3.6; 95% confidence interval, 1.56- There was no significant association between the presence of lum-
8.20). There was no significant difference between English and French bosacral transitional vertebra and the presence of lumbosacral
Bulldogs (P > .05) or between Pugs and French Bulldogs (P > .05). There intervertebral disc herniation with or without neural tissue displace-
were no significant associations between breed and displacement of ment, degree of tail malformation, or spondylosis deformans (P > .05
neural tissue or spondylosis deformans (P > .05). for each). There was a significant association between the presence

Augusto Cavalheiro Vieira Schonfelder - [email protected] - IP: 189.101.142.197

 BERTRAM ET AL . 405

TA B L E 1 Breed-specific prevalence of individual types of congenital vertebral malformations

French Bulldogs (n = 53) English Bulldogs (n = 37) Pugs (n = 59) Total (n = 149)
Total CVM 26 (49.1%) 15 (40.5%) 35 (59.3%) 76 (51.0%)
Hemivertebra 17 (32.0%)1 9 (24.3%)2 1 (1.7%)1,2 27 (18.1%)
Multiple hemivertebra 6 (11.3%) 4 (10.8%) 10 (6.7%)
Spina bifida 5 (9.4%) 5 (13.5%) 5 (8.5%) 15 (10.1%)
LTV 14 (26.4%)3 5 (13.5%)2 32 (54.2%)2,3 51 (34.2%)
Tail score
Grade 0 4 (10.8%) 45 (76.3%) 49 (32.9%)
Grade 1 1 (1.9%) 3 (8.1%) 13 (22.0%) 17 (11.4%)
Grade 2 3 (5.7%) 17 (45.9%) 1 (1.7%) 21 (14.1%)
Grade 3 49 (92.5%) 13 (35.1%) 62 (41.6%)
LS IVDH 28 (52.8%) 24 (64.9%)3 20 (33.9%)3 72 (60.5%)
NT displacement 5 (9.4%) 3 (8.1%) 5 (8.5%) 13 (8.7%)

Abbreviations: CVM = congenital vertebral malformation; LTV = lumbosacral transitional vertebra; LS = lumbosacral; NT = neural tissue.
Superscript numbers indicate a significant difference between two breeds (1 P < .0001; 2 P < .001; 3 P < .01).

TA B L E 2 Prevalence of specific subtypes of lumbosacral transitional vertebra

French Bulldogs (n = 53) English Bulldogs (n = 37) Pugs (n = 59) Total (n = 149)
Total 14 (26.4%) 5 (13.5%) 32 (54.2%) 51 (34.2%)
Type 1 2 (40.0%) 9 (28.1%) 11 (21.6%)
Type 2 6 (42.9%) 2 (40.0%) 5 (15.6%) 13 (25.5%)
Type 3 6 (42.9%) 4 (12.8%) 10 (16.6%)
Type 4 2 (14.3%) 1 (20.0%) 14 (43.8%) 17 (33.3%)

4 DISCUSSION

This study evaluated and compared different types of lumbosacral

congenital vertebral malformations in three brachycephalic dog
breeds and investigated possible associations with the degree of
tail malformation, lumbosacral intervertebral disc herniation, and
spondylosis deformans. Our results revealed an overall high preva-
lence of lumbosacral congenital vertebral malformations (51.0%) and
lumbosacral intervertebral disc herniation (60.5%). In agreement with
previous studies, there were breed-specific differences concerning the
prevalence of different types of congenital vertebral malformations.
F I G U R E 4 Degree of tail malformation of the included dogs with Furthermore, our results suggest an association between the presence
(HV) and without (no HV) evidence of hemivertebra of lumbosacral hemivertebrae and degree of tail malformation and the
presence of lumbosacral intervertebral disc herniation.
A high prevalence of lumbosacral congenital vertebral malfor-
of hemivertebra at L7 or S1 and the presence of lumbosacral interver-
mations and of degenerative changes in neurologically normal
tebral disc herniation (P < .0001; odds ratio = 7.20; 95% confidence
French Bulldogs, English Bulldogs, and Pugs poses a risk for over-
interval, 2.57-20.17) as well as to a higher grade of tail malformation
interpretation of imaging studies, resulting in inappropriate clinical
(P < .0001; Figure 4). There was no significant association between the
decision making. It therefore confirms the importance of consider-
presence of hemivertebra at L7 or S1 and the presence of neural tissue
ing the clinical presentation and results of a complete neurological
displacement or spondylosis deformans (P > .05 for each). There was
examination, including determining the neuro-anatomical localization,
a significant association between age and the presence of lumbosacral
when evaluating advanced imaging studies of the vertebral column.
intervertebral disc herniation (P < .01) and spondylosis deformans
Intervertebral disc degeneration, intervertebral disc herniation, and
(P < .0001). Each month of age increased the odds of intervertebral
compression of neural tissues have previously been demonstrated
disc herniation by 2.1% and the odds for lumbosacral spondylosis
in the cervical and lumbosacral vertebral column of clinically normal
deformans by 2.5% (P < .0001 for both).

Augusto Cavalheiro Vieira Schonfelder - [email protected] - IP: 189.101.142.197

 406 BERTRAM ET AL .

dogs.27–29 In agreement with findings in asymptomatic humans and dogs stated between 0.2 and 0.5.20,32 Heritability of transitional ver-
dogs, intervertebral disc herniation and spondylosis deformans were tebrae has not been investigated in small brachycephalic breeds, but
associated with higher age and can therefore be considered part of the can be considered likely in Pugs. Although transitional vertebrae are
normal ageing process.28,30 It needs to be considered that the included unlikely to cause neurological signs, they have been associated with
Pugs were significantly older than the two other breeds and that there alterations in biomechanical properties of the vertebral column.33
is a chance that a group of age-matched Pugs would show a lower An association between the presence of lumbosacral transitional
prevalence of degenerative lumbosacral changes. The demonstrated vertebra and lumbosacral intervertebral disc herniation has previously
association between lumbosacral hemivertebrae and intervertebral been suggested but could not be confirmed in this study.18,34 More
disc herniation is in agreement with previous studies assessing the specifically, German Shepherd dogs with evidence of lumbosacral
association between thoracic hemivertebra and intervertebral disc transitional vertebra were shown to be eight times more likely to
degeneration and herniation.11,31 Vertebral body malformations develop clinical signs of degenerative lumbosacral stenosis compared
and especially associated kyphosis are suggested to change spinal to dogs without evidence of lumbosacral transitional vertebra.34
biomechanics that can lead to premature degeneration of adjacent Spina bifida was a rather uncommon finding in this study but was
intervertebral discs and a higher prevalence and different anatomical still markedly higher than previously suggested with 0.007% indepen-
distribution of thoracolumbar intervertebral disc extrusion.4,11,31 Our dent of the breed.13 Although it is possible that the three assessed
results seem to confirm that not only thoracic but also lumbosacral brachycephalic breeds are predisposed for lumbosacral spinal bifida,
hemivertebrae are common in French and English Bulldogs but not it is likely that the use of cross-sectional imaging has contributed to
in Pugs.10 Furthermore, our data showed lumbosacral hemivertebrae the high prevalence in this study as previously suggested.35 Analysis of
were associated with a higher degree of tail malformation, whereas previous reports suggests a predisposition of English Bulldogs for lum-
Pugs, therefore, only showed a low degree of tail malformation. This bosacral spina bifida,14 which could not be confirmed in our study. It
is in agreement with a previous study demonstrating a negative cor- should, however, be emphasized that previous reports have focused on
relation between the number of coccygeal vertebrae and the number clinically affected dogs, while this study included only apparently neu-
and degree of thoracolumbar hemivertebrae in French Bulldogs.8 The rologically normal dogs. Spina bifida in clinically affected dogs occurs
shortened and screwed or kinked tails, which are considered a desir- most often at L7 or S1,14–16,36 while spina bifida in our study occurred
able phenotypic trait in several brachycephalic breeds such as French most often at S2 and S3. It is possible that differences in anatomical
and English Bulldogs, are predominantly caused by coccygeal congeni- distribution predispose animals to develop neurological signs associ-
tal vertebral malformations and especially hemivertebrae.7,8,25,26 On ated with spina bifida. Compared to the L7 and S1 vertebrae, there is a
the contrary the smooth curvature of the less extreme tail phenotype decreased volume of neural tissue present at the level of S2 and S3.37
in Pugs has been hypothesized to be caused by vertebra curva that is A more caudal localization of the neural tube defect along the verte-
caused by asymmetric tension of the soft tissue structures of the tail bral column could therefore potentially decrease the risk of concur-
rather than by true congenital vertebral malformations.25 A recent rent meningocele, meningomyelocele, or tethered cord syndrome. The
study identified a mutation in the gene DVL2 in English Bulldogs, demonstrated high prevalence of sacral spina bifida can potentially be
French Bulldogs, and Boston Terriers but not in Pugs that is linked with explained by the curly tail mutant mouse, which is considered a model
thoracic and coccygeal vertebral column malformations in a recessive for human neural tube defects.38 Neural tube defects as well as the
manner.22 This is in agreement with the breed-specific differences atypical curly tails are caused by delayed closure of the caudal neu-
demonstrated in this and additionally in previous studies.10,12 In ropore due to an increased ventral curvature.38 Among other factors,
summary our results agree with the suggestion that Pugs should dysregulation of the expression of Wnt5a and RAR𝛽 genes have been
therefore not be considered a “screw-tailed” brachycephalic breed like associated with this process.39
French and English Bulldogs.22 This study was limited by its retrospective design. Most of the
A breed predisposition of lumbosacral transitional vertebra in included dogs did not undergo a neurological examination and there-
Pugs is in agreement with previous studies demonstrating a higher fore it cannot be guaranteed that dogs with intermittent or very mild
prevalence of thoracolumbar transitional vertebrae and homeotic neurological deficits were included. Although the occurrence of degen-
transformation of C7 in Pugs compared to other breeds.10,13 The erative lumbosacral stenosis in the assessed breeds is very rare,40 it
stated prevalence of lumbosacral transitional vertebra in the literature is possible that some of the included dogs might develop clinical signs
varies and depends on the applied inclusion criteria and selected later in life. Additionally, this study was limited to CT scans with no
imaging modality. A prevalence of 3.45% was stated for a mixed cohort MRI studies available. It therefore cannot be excluded that changes of
of 4000 dogs, ranging from 0.0% to 9.4% depending on the breed.18 soft tissue structures and subtle myelopathies were overlooked. The
An extensive review of the literature in the same study stated a breed lack of comparison to breed-matched dogs affected by clinical appar-
dependent prevalence between 0.0% and 22.0%.18 This study did how- ent sequelae due to lumbosacral congenital vertebral malformation
ever not include any of the breeds evaluated in our study. Based on could be considered another limitation. This was, however, not the aim
the variable interbreed prevalence, a genetic predisposition has been of this study and clinically relevant lumbosacral congenital vertebral
suggested,18,19,21 with heritability estimates for German Shepherd malformations can be considered very rare in the assessed breeds and

Augusto Cavalheiro Vieira Schonfelder - [email protected] - IP: 189.101.142.197

 BERTRAM ET AL . 407

warrant further investigation. Furthermore, further studies are nec- ORCID

essary to investigate if selection of dogs with a true “screw-tailed”
Simon Bertram https://orcid.org/0000-0001-8225-8504
phenotype for breeding purposes is indeed associated with a higher
likelihood of hemivertebrae and other malformations and to evaluate
if dogs with a curled or screwed tail and concurrent sarcococcygeal
REFERENCES
spina bifida indeed have similar genetic abnormalities than the curly
tail mutant mouse. 1. Westworth DR, Sturges BK. Congenital spinal malformations in small
animals. Vet Clin North Am Small Anim Pract. 2010;40:951-981.
2. Gutierrez-Quintana R, Guevar J, Stalin C, Faller K, Yeamans C, Pen-
deris J. A proposed radiographic classification scheme for congenital
thoracic vertebral malformations in brachycephalic “screw-tailed” dog
5 CONCLUSION breeds. Vet Radiol Ultrasound. 2014;55:585-591.
3. Moissonnier P, Gossot P, Scotti S. Thoracic kyphosis associated with
This study demonstrated that neurologically normal French Bulldogs, hemivertebra. Vet Surg. 2011;40:1029-1032.
English Bulldogs, and Pugs have a high prevalence of lumbosacral 4. Aikawa T, Shibata M, Asano M, Hara Y, Tagawa M, Orima H. A compar-
ison of thoracolumbar intervertebral disc extrusion in French Bulldogs
congenital vertebral malformations and associated intervertebral disc
and Dachshunds and association with congenital vertebral anomalies.
herniation. In agreement with previous studies, Pugs demonstrate Vet Surg. 2014;43:301-307.
significantly more often lumbosacral transitional vertebra and less 5. Aikawa T, Kanazono S, Yoshigae Y, Sharp NJ, Munana KR. Vertebral
often hemivertebrae compared to French and English Bulldogs. Lum- stabilization using positively threaded profile pins and polymethyl-
methacrylate, with or without laminectomy, for spinal canal steno-
bosacral hemivertebrae are associated with a higher likelihood of
sis and vertebral instability caused by congenital thoracic vertebral
lumbosacral intervertebral disc herniation and more severe tail mal- anomalies. Vet Surg. 2007;36:432-441.
formations. These anatomical variances support the thesis that true 6. Jeffery ND, Smith PM, Talbot CE. Imaging findings and surgical
“screw-tailed” brachycephalic breeds such as French and English Bull- treatment of hemivertebrae in three dogs. J Am Vet Med Assoc.
dogs are affected by different vertebral malformations than Pugs and 2007;230:532-536.
7. Done S, Drew R, Robins G, Lane J. Hemivertebra in the dog: Clinical and
that these anatomical variances need to be considered to avoid misin-
pathological observations. Veterinary Record. 1975;96:313-317.
terpretation, complications, or failure when evaluating imaging studies 8. Schlensker E, Distl O. Heritability of hemivertebrae in the French Bull-
and when planning for neurosurgical and neurodiagnostic procedures. dog using an animal threshold model. Vet J. 2016;207:188-189.
9. Meheust P, Robert R. Surgical treatment of a hemivertebra by partial
ventral corpectomy and fusion in a Labrador puppy. Vet Comp Orthop
Traumatol. 2010;23:262-265.
10. Ryan R, Gutierrez-Quintana R, Ter Haar G, De Decker S. Prevalence
LIST OF AUTHOR CONTRIBUTIONS
of thoracic vertebral malformations in French Bulldogs, Pugs and
English Bulldogs with and without associated neurological deficits. Vet
Category 1 J. 2017;221:25-29.
11. Faller K, Penderis J, Stalin C, Guevar J, Yeamans C, Gutierrez-Quintana
(a) Conception and Design: De Decker, ter Haar, Bertram R. The effect of kyphoscoliosis on intervertebral disc degeneration in
(b) Acquisition of Data: Bertram dogs. Vet J. 2014;200:449-451.
12. Bertram S, Ter Haar G, De Decker S. Caudal articular process dysplasia
(c) Analysis and Interpretation of Data: Bertram of thoracic vertebrae in neurologically normal French Bulldogs, English
Bulldogs, and Pugs: Prevalence and characteristics. Vet Radiol Ultra-
sound. 2018;59:396-404.
Category 2 13. Brocal J, De Decker S, Jose-Lopez R, Manzanilla EG, Penderis J, Stalin
C, et al. C7 vertebra homeotic transformation in domestic dogs -
(a) Drafting the Article: Bertram, De Decker are Pug dogs breaking mammalian evolutionary constraints? J Anat.
2018;233:255-265.
(b) Revising Article for Intellectual Content: Bertram, ter Haar, De
14. Wilson JW, Kurtz HJ, Leipold HW, Lees GE. Spina bifida in the dog. Vet
Decker Pathol. 1979;16:165-179.
15. Fingeroth JM, Johnson GC, Burt JK, Fenner WR, Cain LS. Neuroradio-
graphic diagnosis and surgical repair of tethered cord syndrome in an
Category 3 English Bulldog with spina bifida and myeloschisis. J Am Vet Med Assoc.
1989;194:1300-1302.
(a) Final Approval of the Completed Article: Bertram, ter Haar, De 16. Shamir M, Johnston D, Rochkind S. Surgical treatment of tethered
Decker spinal cord syndrome in a dog with myelomeningocele. Vet Rec.
2001;148:755-756.
17. Kramer JW, Schiffer SP, Sande RD, Rantanen NW, Whitener EK. Char-
acterization of heritable thoracic hemivertebra of the German Short-
ACKNOWLEDGEMENTS
haired Pointer. J Am Vet Med Assoc. 1982;181:814-815.
No third-party funding or support was received in connection with 18. Damur-Djuric N, Steffen F, Hassig M, Morgan JP, Fluckiger MA. Lum-
bosacral transitional vertebrae in dogs: Classification, prevalence,
this study or the writing or publication of the manuscript. The authors
and association with sacroiliac morphology. Vet Radiol Ultrasound.
declare that there were no conflicts of interest. 2006;47:32-38.

Augusto Cavalheiro Vieira Schonfelder - [email protected] - IP: 189.101.142.197

 408 BERTRAM ET AL .

19. Larsen JS. Lumbosacral transitional vertebrae in the dog 1. Veterinary 31. Inglez de Souza M, Ryan R, Ter Haar G, Packer RMA, Volk HA, De
Radiology. 1977;18:76-79. Decker S. Evaluation of the influence of kyphosis and scoliosis on inter-
20. Ondreka N, Amort KH, Stock KF, Tellhelm B, Klumpp SW, Kramer vertebral disc extrusion in French Bulldogs. BMC Vet Res. 2018;14:5.
M, et al. Skeletal morphology and morphometry of the lumbosacral 32. Wigger A, Julier-Franz C, Tellhelm B, Kramer M. Lumbosakraler
junction in German shepherd dogs and an evaluation of the possible Übergangswirbel beim Deutschen Schäferhund: häufigkeit, formen,
genetic basis for radiographic findings. Vet J. 2013;196:64-70. genetik und korrelation zur hüftgelenksdysplasie. Tierärztl Prax Ausg K.
21. Moeser CF, Wade CM. Relationship between transitional lumbosacral 2009;37:7-13.
vertebrae and eight lumbar vertebrae in a breeding colony of Labrador 33. Buerger R, JL. Kinetische studie ueber die lendenwirbelsaeule und
Retrievers and Labrador Crosses. Aust Vet J. 2017;95:33-36. den lumbosakralen uebergang beim Deutschen Schaeferhund. Schweiz
22. Mansour TA, Lucot K, Konopelski SE, Dickinson PJ, Sturges BK, Ver- Arch Tierheilk. 1992;134:411-416.
nau KL, et al. Whole genome variant association across 100 dogs iden- 34. Flückiger MA, Damur-Djuric N, Hassig M, Morgan JP, Steffen F. A lum-
tifies a frame shift mutation in dishevelled 2 which contributes to bosacral transitional vertebra in the dog predisposes to cauda equina
Robinow-like syndrome in Bulldogs and related screw tail dog breeds. syndrome. Vet Radiol Ultrasound. 2006;47:39-44.
PLoS Genet. 2018;14:e1007850. 35. Brocal J, De Decker S, Jose-Lopez R, Guevar J, Ortega M, Parkin T,
23. Lappalainen AK, Salomaa R, Junnila J, Snellman M, Laitinen- et al. Evaluation of radiography as a screening method for detection
Vapaavuori O. Alternative classification and screening protocol and characterisation of congenital vertebral malformations in dogs. Vet
for transitional lumbosacral vertebra in German Shepherd dogs. Acta Rec. 2018;182:573.
Vet Scand. 2012;54:27. 36. Parker AJ, Byerly CS. Meningomyelocoele in a dog. Vet Pathol.
24. Langeland M, Lingaas F. Spondylosis deformans in the Boxer: Esti- 1973;10:266-273.
mates of heritability. J Small Anim Pract. 1995;36:166-169. 37. Fletcher TF. Spinal Cord and Meninges. In: Evans HE, de Lahunta A,
25. Schawalder P, Dietschi E, Stich H. Congenital and acquired malfor- eds. Miller’s Anatomy of the Dog. 4 ed. St. Louis, MO: Elsevier Saunders
mations of coccygeal vertebrae in dogs [in German]. Wien Tierärztl Co; 2013:589-610.
Monatsschr. 2010;97:185-202. 38. van Straaten HW, Copp AJ. Curly tail: A 50-year history of the mouse
26. Paninarova M, Stehlik L, Proks P, Vignoli M. Congenital and acquired spina bifida model. Anat Embryol. 2001;203:225-237.
anomalies of the caudal vertebrae in dogs: Radiographic classification 39. Finnell RH, Gould A, Spiegelstein O. Pathobiology and genetics of neu-
and prevalence evaluation. Acta Vet Hung. 2016;64:330-339. ral tube defects. Epilepsia. 2003;44:14-23.
27. da Costa RC, Parent JM, Partlow G, Dobson H, Holmberg DL, Lamarre 40. De Decker S, Wawrzenski LA, Volk HA. Clinical signs and outcome
J. Morphologic and morphometric magnetic resonance imaging fea- of dogs treated medically for degenerative lumbosacral stenosis: 98
tures of Doberman Pinschers with and without clinical signs of cervical cases (2004-2012). J Am Vet Med Assoc. 2014;245:408-413.
spondylomyelopathy. Am J Vet Res. 2006;67:1601-1612.
28. De Decker S, Gielen IM, Duchateau L, Van Soens I, Bavegems V,
Bosmans T, et al. Low-field magnetic resonance imaging findings of the
caudal portion of the cervical region in clinically normal Doberman Pin-
schers and Foxhounds. Am J Vet Res. 2010;71:428-434. How to cite this article: Bertram S, ter Haar G, De
29. Jones JC, Inzana KD. Subclinical CT abnormalities in the lumbosacral
Decker S. Congenital malformations of the lumbosacral
spine of older large-breed dogs. Vet Radiol Ultrasound. 2000;41:19-26.
30. Cheung KM, Karppinen J, Chan D, Ho DW, Song YQ, Sham P, vertebral column are common in neurologically normal
et al. Prevalence and pattern of lumbar magnetic resonance imaging French Bulldogs, English Bulldogs, and Pugs, with breed-
changes in a population study of one thousand forty-three individuals. specific differences. Vet Radiol Ultrasound. 2019;60:400–408.
Spine (Phila Pa 1976). 2009;34:934-940.
https://doi.org/10.1111/vru.12753

Augusto Cavalheiro Vieira Schonfelder - [email protected] - IP: 189.101.142.197