Prevalence, MRI findings, and clinical features of lumbosacral intervertebral disc protrusion in French Bulldogs diagnosed with acute thoracic or lumbar intervertebral disc extrusion

Introduction Intervertebral disc protrusion (IVDP) is a neurological disorder commonly observed at the lumbosacral junction of old, medium-to-large breeds, non-chondrodystrophic dogs. Although uncommon, lumbosacral IVDP can also be seen in chondrodystrophic dogs, among them French Bulldogs (FBs) and could be associated with congenital vertebral malformations in this breed. This study aims to evaluate the prevalence, clinical features, and MRI characteristics of lumbosacral IVDP and congenital vertebral malformations in FBs diagnosed with thoracic or lumbar intervertebral disc extrusion (IVDE) and to evaluate the possible interference of the neurologic deficits related to chronic IVDP on neurological examination. Materials and methods This is a single-center, retrospective case series. A search for FBs diagnosed with IVDE affecting the thoracic or lumbar regions is done on the database of the AniCura I Portoni Rossi Veterinary Hospital (Zola Predosa, Bologna, Italy). Eligible dogs have a complete medical report and a high-field MRI of the lumbosacral junction. MRIs of the lumbosacral junction are evaluated to determine the position of IVDP, cranial intervertebral foraminal stenosis, and signs of nerve root involvement. Radiographs, when available, are reviewed to identify the presence of lumbosacral congenital vertebral malformations. Results Eighty FBs are included in the study. The prevalence of lumbosacral IVDP among FBs is 91.3%. Among FBs with lumbosacral IVDP, 45.0% show concurrent cranial intervertebral foraminal stenosis, 28.8% exhibit concurrent nerve root involvement, 56.2% appear to be asymptomatic for lumbosacral changes, while 15.1% manifest a decreased or absent withdrawal reflex as a supposed consequence of chronic lumbosacral IVDP. Congenital vertebral malformations are detected in 10 dogs. Conclusion The results of this study support the hypothesis that lumbosacral IVDP is frequent in FBs presenting with thoracic or lumbar IVDE. In over half the dogs lumbosacral IVDP appears to be asymptomatic; however, in other cases, chronic lumbosacral IVDP seems to cause neurological deficits that may lead to erroneous localization of acute IVDE, representing a confounding factor for clinicians.

Introduction: Intervertebral disc protrusion (IVDP) is a neurological disorder commonly observed at the lumbosacral junction of old, medium-to-large breeds, non-chondrodystrophic dogs.Although uncommon, lumbosacral IVDP can also be seen in chondrodystrophic dogs, among them French Bulldogs (FBs) and could be associated with congenital vertebral malformations in this breed.This study aims to evaluate the prevalence, clinical features, and MRI characteristics of lumbosacral IVDP and congenital vertebral malformations in FBs diagnosed with thoracic or lumbar intervertebral disc extrusion (IVDE) and to evaluate the possible interference of the neurologic deficits related to chronic IVDP on neurological examination.
Materials and methods: This is a single-center, retrospective case series.A search for FBs diagnosed with IVDE affecting the thoracic or lumbar regions is done on the database of the AniCura I Portoni Rossi Veterinary Hospital (Zola Predosa, Bologna, Italy).Eligible dogs have a complete medical report and a high-field MRI of the lumbosacral junction.MRIs of the lumbosacral junction are evaluated to determine the position of IVDP, cranial intervertebral foraminal stenosis, and signs of nerve root involvement.Radiographs, when available, are reviewed to identify the presence of lumbosacral congenital vertebral malformations.
Results: Eighty FBs are included in the study.The prevalence of lumbosacral IVDP among FBs is 91.3%.Among FBs with lumbosacral IVDP, 45.0% show concurrent cranial intervertebral foraminal stenosis, 28.8% exhibit concurrent nerve root involvement, 56.2% appear to be asymptomatic for lumbosacral changes, while 15.1% manifest a decreased or absent withdrawal reflex as a supposed consequence of chronic lumbosacral IVDP.Congenital vertebral malformations are detected in 10 dogs.

Conclusion:
The results of this study support the hypothesis that lumbosacral IVDP is frequent in FBs presenting with thoracic or lumbar IVDE.In over half the dogs lumbosacral IVDP appears to be asymptomatic; however, in other cases, chronic lumbosacral IVDP seems to cause neurological deficits that may lead to erroneous localization of acute IVDE, representing a confounding factor for clinicians.
Albeit uncommon, IVDP can occur in CD dogs.IVDPs affecting the LS junction are poorly reported in CD dogs (4).A recent study found 72/149 dogs with LS IVDH within a population of neurologically normal FBs, English Bulldogs, and pugs; among FBs alone, the same study recorded a prevalence of 52.8% for LS IVDH (13).Another study reported a prevalence of 77.4% within a population of FBs, both with and without neurological disorders (14).
Furthermore, the LS junction has been reported to be affected by congenital vertebral malformations (CVMs) in CD dogs based on CT examination and a possible association between CVMs and LS IVDP has been suggested (13,14).
To the best of our knowledge, no studies have focused on MRI characteristics of LS IVDP and the resulting involvement of nerve roots in FBs.It is the authors' opinion that LS IVDP in FBs is a frequent, underdiagnosed, chronic condition that can cause neurological deficits possibly interfering with neurological evaluation when these dogs are presenting with acute thoracic or lumbar IVDE.Therefore, this study aims to evaluate (1) the prevalence and MRI characteristics of LS IVDP, cranial intervertebral foraminal stenosis, root involvement, and CVMs in FBs diagnosed with thoracic or lumbar IVDE and (2) the possible interference of the neurologic deficits related to chronic IVDP on neurological examination performed when dogs are evaluated for acute IVDE.
It is hypothesized that a high prevalence of LS IVDP may be a confounding factor at the time of neuroanatomical localization of acute myelopathy and an association between LS CVM and IVDP may exist.

Case recruitment criteria
This is a single-center, descriptive, retrospective case series.The database of the AniCura I Portoni Rossi Veterinary Hospital (Zola Predosa, Bologna, Italy) was searched for FBs that were diagnosed through MRI with IVDE affecting the thoracolumbar (T3-L3) or LS (L4-S3) regions between April 2018 and February 2023.All the animals used in the study were client-owned and underwent MRI examinations of the vertebral column as part of their diagnostic workup.Dogs were eligible if they had (1) a medical report including signalment and a neurological examination performed by a European College of Veterinary Neurology (ECVN) board-certified veterinary neurologist or an ECVN resident and (2) an MRI of the vertebral column extended to the LS junction, imaged in the T2-weighted sagittal plane and at least one other plane.If multiple MRI studies of the same dog were performed at different times, only the oldest one including the LS junction was examined.Patients with multiple IVDEs, LS IVDE, or other concurrent pathological conditions affecting the spinal cord or acquired disease processes of the vertebral column between T3 and S3 were excluded.Radiographs of the LS junction were also evaluated when available.

Medical records review
Details regarding signalment (sex, age, and neutered/spayed status), neurological examination, and location of IVDE were collected.Information obtained from neurological examination included: (1) analysis of the gait graded from 1 (spinal hyperesthesia only) to 5 (paraplegia without deep pain) according to a grading system described elsewhere (15); (2) assessment of the withdrawal reflex of pelvic limbs (normal/increased or reduced/absent); (3) level of cutaneous trunci muscle reflex (CTMR) cut-off; and (4) assessment of thoracic/lumbar or LS pain (present or absent).Dogs were divided into two groups based on the localization of IVDE: group A (cranial to the fourth lumbar vertebra [L4]), and group B (caudal to L4).Dogs in both groups were further subdivided into cases with normal withdrawal reflexes and cases with reduced/absent withdrawal reflexes.MRI scans of the LS junction were evaluated for LS IVDP, cranial intervertebral foraminal stenosis, and involvement of spinal nerve roots.IVDP was classified as canalar, left or right intervertebral foraminal, bilateral intervertebral foraminal, canalar + right/left intervertebral foraminal, and all positions (canalar + bilateral intervertebral foraminal) (Figure 1).The degree of LS 1 www.horosproject.comIVDP was scored as grade 1 when obstruction of the spinal canal was <25%, grade 2 when obstruction was 25-50%, and grade 3 when the obstruction was >50% (10, 16).Cranial intervertebral foraminal stenosis was defined as an abnormal conformation and/ or signal changes of the edge of the cranial part of the LS intervertebral foramen leading to a reduction in size (Figure 2).Spinal nerve root involvement was defined as occurring when nerve roots were compressed, enlarged, showed MRI signal changes, or were not detected due to the presence of degenerative material along the nerve root path (Figure 3) and estimated as absent, left/right lateral, or bilateral.Radiographs were reviewed for the presence of LS CVM such as butterfly vertebra, transitional vertebra, block vertebrae, and spina bifida.When radiographs were not available, we analyzed MRI for the same alterations.The above  3 Results

Case recruitment criteria and medical records review
Initially, 173 dogs were considered for inclusion.Seven cases were excluded because of an incomplete medical report, 77 cases because the LS junction was either not included in the MRI or not imaged as required, and 9 cases because of concurrent pathological conditions affecting the spinal cord, multiple IVDEs, or LS IVDE.Consequently, 80 FBs fulfilled the inclusion criteria.These FBs comprised 54 males (67.5%) of which 10 were neutered and 26 females (32.5%) of which 21 were spayed.Age at presentation ranged from 15 months to 9 years, with a mean of 2.9 years and a median age of 4 years.
Neurological examination findings and sites of IVDE are summarized in Table

Imaging analysis
Radiographs were available for review in 70/80 cases (87.5%), with both lateral and ventrodorsal projections available in 52/70 cases (74.3%), and only 1 projection in 18/70 cases (25.7%).MRI was conducted under general anesthesia, with dogs placed in dorsal recumbency.All studies were performed with a 1.5T MR scanner (Vantage Elan, Canon Medical Systems Europe B.V., Netherlands) with the best-fitting 16-channel flexible receiving coil from the 2 available.MRI of the LS junction was available in 2 planes in 53 cases (66.3%) (48 cases on sagittal and dorsal planes and 5 cases on sagittal and transverse planes) and in all 3 spatial planes in 27 cases (33.7%).MRI sequences available for review are summarized in Supplementary Table 1.
In both group A and group B, IVDPs affected at least one position (43/49, 87.8% and 30/31, 96.8%, respectively).In both groups, most IVDPs affected all the positions, while IVDPs at both canalar and unilateral intervertebral foraminal level, or bilateral intervertebral foraminal level, or canalar position or unilateral intervertebral foraminal level were seen less frequently.The distribution of IVDP, cranial intervertebral foraminal stenosis, and nerve root involvement within the population and within each group is summarized in Table 2. CVMs were detected in 10 cases (12.5%).A butterfly vertebra affecting the last lumbar and the first sacral vertebra was seen in 3 and 4 cases, respectively.An LS transitional vertebra was observed in 3 cases.No dogs presented with spina bifida or block vertebrae.All the 10 dogs with CVM had a concurrent LS IVDP.

Association between clinical and MRI findings
Among the 34 dogs in group A with a bilaterally normal withdrawal reflex, 30 (88.2%) showed LS IVDP with or without concurrent LS alterations.Of the remaining 4 dogs, 2 (5.9%) revealed cranial intervertebral foraminal stenosis without spinal nerve root involvement, and the other 2 (5.9%) displayed no LS abnormalities.Among the 15 dogs in group A with decreased/ absent withdrawal reflex, 13 (86.7%)manifested LS IVDP with or without concurrent LS alterations; in this subgroup, 7 dogs were neurologically graded 2, 4 dogs were neurologically graded 3, and 2 dogs were neurologically graded 4. One of the two remaining dogs exhibited only cranial intervertebral foraminal stenosis without spinal nerve root involvement and the other showed no LS abnormalities; these two dogs were neurologically graded four and five, respectively.Among the 12 dogs in group B with bilateral normal withdrawal reflex, 11 (91.6%)showed LS IVDP and 1 showed no LS abnormalities.All 19 dogs in group B with decreased/absent withdrawal reflex showed an LS IVDP with or without concurrent LS abnormalities.Details about LS changes within each group are summarized in Table 3.
Among the 73 dogs with LS IVDP, only 4 (5.5%) were found to have LS spinal pain upon neurological examination.All the 4 dogs revealed bilaterally decreased/absent withdrawal reflex; among these dogs, 3 were in group A and 1 was in group B. Of the remaining 69 dogs, 44 (63.8%) showed hyperesthesia at the caudal thoracic or lumbar levels that matched with the IVDE site upon MRI examination, 20 (29%) revealed no spinal pain and a bilaterally normal withdrawal reflex or a CTMR cutoff that could suggest thoracolumbar localization and 5 (7.2%) exhibited no spinal pain, decreased/absent withdrawal reflexes, and no CTMR cut-off.

Discussion
The results of this study support our hypothesis that LS IVDP is frequent in FBs presenting with thoracic or lumbar IVDE.LS IVDP affecting at least 1 position (canalar, intervertebral foraminal, or both) was observed in 91.3% of the FBs in our population, a percentage much higher than that reported in 2 previous studies [77.4%; (14) and 52.8% (13)] and more than half of the FBs in our population were graded ≥2.This difference may be due to several factors.First, the FBs in our population were older (median age = 4 years) than the dogs described in the paper written by Bertram et al. (median age 16 months) (13).Second, the positioning of the patient for MRI may have played a role.Our patients were placed in dorsal recumbency while in the study of Bertram et al. patients were examined under sternal recumbency (13).The supine position has the effect of extending the LS junction, leading to an exacerbation of IVDP with respect to the neutral position (1,16,17).Thus, positioning may have partially led to an overestimation of the real severity of LS IVDP in our population.However, in our study, 45% of the FBs with LS IVDP showed concomitant cranial intervertebral foraminal stenosis, a degenerative condition frequently reported with LS IVDP, and 28.8% revealed concurrent nerve root involvement.These findings decrease the likelihood of positioning-related bias (17,18).Third, the contrast between epidural fat and a degenerated disc is stronger in MRI than in CT, allowing MRI to offer a precise detection of mild IVDPs.This property of the MRI may also be useful in defining the spatial distribution of IVDPs (3,19).The latter point may explain the greater prevalence (87.6%) of large LS IVDPs affecting both the vertebral canal and least one foramen in our population compared to a previous study (14).The presence of LS CVMs has been considered to be a predisposing factor for LS IVDH.An association between LS IVDH with both LS transitional vertebra in a population of non-CD dogs (20) and hemivertebra at L7 or S1 (13) has been described.The prevalence of LS CVMs in our population study was much lower (12.5%)than that reported in recent papers based on CT examination (49.1 and 56.3% in neurological and non-neurological FBs, respectively) (13,14) and similar to that reported in studies based on radiographs (0 to 17.5%) (21-23).We can suppose that LS IVDPs in FBs are a common finding even in absence of LS CVMs.However, MRI and radiography are not the best imaging modalities for diagnosing CVMs.CT is reported to be far more sensitive in the identification of the number of CVMs when compared to both radiography (24, 25) and MRI (1).Therefore, it is likely that the diagnostic modalities used in the present study may have underestimated the real prevalence of CVMs in our population.
The vertebral column outgrows the spinal cord in length, and the degree of elongated growth of spinal cord segments varies regionally (26).Thus, positions of most spinal cord segments reside in the vertebral canal cranial to the vertebra of the same number.This phenomenon is most pronounced in the caudal lumbar and sacrocaudal segments of the spinal cord.Generally, the three sacral spinal cord segments lie within the fifth lumbar  vertebral foramen, albeit with some variation between breeds: in small breeds, these segments extend approximately one further caudally (2,26).As a result, the segments involved in the withdrawal reflex of the pelvic limbs (L6-S2) likely reside in the vertebral canal caudal to the L4 vertebra.Therefore, the withdrawal reflex could be decreased/eliminated by an IVDE at the L4-L5 intervertebral space or caudally.For this reason, in order to compare clinical findings (normal, decreased, or absent withdrawal reflexes) with the presence/absence of LS changes, the L4 vertebrae were taken as a cut-off point to classify the population into 2 groups: 61.3% of dogs were diagnosed with an IVDE cranial to L4 (group A) and 38.8% with an IVDE caudal to L4 (group B).Forty-one out of the 73 FBs with LS IVDP (30 in group A and 11 in group B) had a bilaterally normal withdrawal reflex and none of them showed LS pain upon neurological examination.Therefore, in all these dogs, LS IVDP appears to be asymptomatic.This finding is in accordance with the 52.8% of neurologically normal FBs with LS IVDH reported in the literature (13).
Thirty-two out of 73 dogs with LS IVDP (13 in group A and 19 in group B) had a decreased or absent withdrawal reflex.In group A, 2 dogs were paraplegic (one graded 4 and the other graded 5); and the remaining 11 dogs showed variable degrees of paraparesis.Paraplegic dogs have increased odds of spinal shock, a condition characterized by impaired spinal reflexes and muscle tone caudal to an injury to the spinal cord, compared to dogs with persistent motor function (2,27).Therefore, the impaired withdrawal reflex in the two paraplegic dogs was likely a consequence of the spinal shock caused by IVDE rather than due to the LS IVDP.In contrast, the 11 paraparetic dogs are more likely to have shown a decreased or absent withdrawal reflex as a consequence of LS IVDP.This is an important fact from a clinical point of view.
When performing a neurological examination in FBs, clinicians should be aware that chronic neurological deficits caused by LS IVDP might interfere with the neurological deficits due to acute IVDE, potentially leading to a wrong localization.In group B, a decreased or absent withdrawal reflex could be due to both acute IVDE and chronic LS IVDP.Both types of IVDH could contribute to a clinical presentation.This fact might not have an important impact in making immediate surgical decisions, since IVDE is likely to play a major role.However, a possible persistent flexor impairment in the follow-up could be erroneously attributed to irreversible spinal cord damage at the level of IVDE rather than to IVDP-especially in dogs where an L7 nerve root involvement has been seen using MRI.
LS pain is reported to be the most frequent and first presenting clinical sign in dogs diagnosed with LS degenerative stenosis (8).Approximately half of the FBs with LS IVDP in our population were graded at least 2 regarding the severity of IVDP, but only 5.5% of the FBs with LS IVDP in our population exhibited LS spinal pain upon neurological examination.One possible explanation could be that most dogs presented with severe thoracic or lumbar hyperesthesia related to acute IVDE and the pain possibly arising at the LS junction could be masked or overlooked under neurological examination.Therefore, it is advisable to carefully evaluate FBs with thoracic or lumbar IVDE for the presence of LS hyperesthesia.
The main limitations of this study are due to its retrospective nature.The first concerns the different MRI protocols used to investigate the LS region.This lack of standardization may have influenced the assessment of LS changes, especially with regard to nerve root involvement.However, as an inclusion criterion, all dogs were at least examined in the sagittal plane, which is considered an accurate approach in evaluating canalar and foraminal involvement.Another limitation is the lack of clinical follow-up.A long-term assessment of the hind limb withdrawal reflex after recovery from IVDE might have allowed for gathering of additional information about the role of LS IVDP in FBs with an IVDE caudal to L4.
In conclusion, LS IVDP is frequently reported in FBs, with or without cranial intervertebral foraminal stenosis, nerve root involvement, or CVMs, which often causes neither deficits upon neurological examination nor evident LS pain.In some dogs, chronic LS IVDP may cause neurological deficits that may lead to incorrect neurolocalization, representing a possible confounding factor for clinicians even in young and middle-aged subjects.Therefore, FBs need to be carefully examined at the level of the LS junction even when a patient is presenting with thoracic or lumbar IVDE.

FIGURE 1 Parasagittal
FIGURE 1Parasagittal (A) and sagittal (C) T2W images of the lumbar vertebral column and the LS junction; transverse T2W at the level of L4 vertebra (B) and at the level of LS junction (D).A severe extrusion of the L4-L5 IVD in seen on the right side of the vertebral canal (long arrow).An IVDP affecting both the vertebral canal and the intervertebral foramina is seen at the LS junction (short arrows).

FIGURE 3 Sagittal
FIGURE 3 Sagittal T2W image (A) of the T11-S1 vertebral column; transverse T2W images at the level of L4 caudal epiphysis (B) and at the level of LS junction (C); dorsal STIR (D) image of the L6-S1 vertebral column.A T12-T13 IVDE is seen (long arrows); a LS IVDP (short arrow) is associated with an enlargement and signal change of the right L7 nerve root (arrowheads).

TABLE 1
Distribution of thoracic and lumbar IVDE and clinical neurological findings in all dogs included.

TABLE 2
LS changes within the entire population.

TABLE 3
Prevalence of LS changes within each group.