Long-term follow-up of spinal segmental stabilization for surgical treatment of dorsal hemivertebrae associated with kyphosis in brachycephalic dogs

Abstract

The aim of this study is to report chronic complications (> 2 mo after surgery) following spinal segmental stabilization (SSS) to treat myelopathy associated with thoracic congenital vertebral malformations in brachycephalic dogs. Follow-up medical records (years 2006 to 2020) of 12 cases that underwent SSS at 3 university hospitals were retrieved and analyzed with a minimum follow-up period of 1 y. Five dogs showed no chronic complications and 7 dogs had chronic complications which are reported here. This case series demonstrates that the rate of chronic complications associated with SSS was high (58%) but most of these were minor and did not require revision surgery.

Introduction

Congenital vertebral malformations (CVMs) are common incidental findings on diagnostic imaging of the thoracic vertebral column in “screw-tailed” brachycephalic breeds (1,2). Although the true prevalence of CVMs in such breeds is unknown, previous studies have indicated this could be as high as 96% in pugs (2). These malformations are associated with moderate to severe kyphosis and scoliosis of the thoracic vertebral column (2). Most cases are incidental and non-clinical but those associated with neurological deficits are considered a consequence of vertebral instability and vertebral canal stenosis (3).

Optimal treatment for CVMs has not been established. Nonsurgical treatment has been reported to result in most patients having progression of clinical signs (4). Surgical treatment to stabilize the abnormal region and relieve compression can be challenging, with published techniques including stabilization with pins and polymethylmethacrylate (PMMA), implant placement with 3D printed patient-specific drill guides and spinal segmental stabilization surgery (SSS), all with or without decompressive laminectomy or corpectomy (5–9).

Spinal segmental stabilization surgery is a relatively easy surgical method for treating symptomatic hemivertebrae and carries minimal risk of iatrogenic damage to the spinal cord, compared with other stabilization methods (8). We aimed to provide further evidence regarding this technique by reporting chronic complications. In addition, as SSS can lead to long-term implant failure with associated overall poor outcomes, there is a clear need to assess the rate of long-term, clinically important implant failure in cases in which the technique has been used.

Materials and methods

Follow-up medical records (years 2006 to 2020) of 12 cases that underwent SSS at the Queen’s Veterinary School Hospital, University of Cambridge, the Small Animal Teaching Hospital, University of Liverpool, and the Faculty of Veterinary Medicine, Ghent University were retrieved and analyzed. These were either extracted from the universities’ databases or obtained by contact with the owners and referring veterinary surgeons. Five of the cases (Dogs 1, 3, 7, 8, and 9) were included in a previous study (8).

The data collected included signalment, duration of clinical signs pre-operatively, results of initial neurological examination, results of investigations including radiographs, computed tomography (CT), and magnetic resonance images (MRI), neurological status immediately after surgery and at follow-up periods of > 12 mo as well as complications.

To maintain consistency for the follow-up analysis among the 5 cases from the previous study and the new cases, the same neurological status graded scale was followed as in Charalambous et al (8). Grade 0 was considered a neurologically normal status, Grade 1 was mild and inconsistent ataxia, Grade 2 was mild and consistent ataxia, Grade 3 was moderate ambulatory paraparesis, Grade 4 was severe ambulatory paraparesis, and Grade 5 was non-ambulatory paraparesis.

Acute complications were defined as those seen within 8 wk of initial surgery and chronic complications were defined as those seen at more than 8 wk after surgery.

Results

The breeds included pugs (n = 10), a French bulldog (n = 1), and a Pomeranian (n = 1). There were 10 males and 2 females with a median age at the time of surgery of 9 mo (mean: 18 mo; range: 2 to 64 mo) (Table 1).

Table 1.

Clinical details of 12 cases.

Dog	Breed	Age at surgery	Gender	Surgery performed	Neuro status pre-op (grade)	Neuro status post-op (grade)	Duration: surgery to neurological deterioration	Presentation	Pathology due to complication	Action required	Outcome and further details
1	Pomeranian	9 mo	M	SSS	3	0	2 y	Spinal pain, possibly in association with a traumatic fall.	Implant loosening.	Resolved with rest and NSAIDs.	Euthanasia 5 y post-op due to unrelated condition.
2	Pug	4 y, 1 mo	MN	SSS	3	1	2 y and 5 y	2 y — Intermittent swelling over the thoracic vertebral column. 5 y — Swelling re-occurred.	5 y — Osteolysis surrounding implant and loosening of implant.	2 y — Rest and NSAIDs. 5 y — Non-responsive to NSAIDs and rest. Cytology: mixed eutrophilic/ macrophagic inflammation, cultured Staphylococcus epidermidis. Surgery to remove pins and caudal wires.	Cranial wires left in-situ due to difficulties retrieving them. Significant fibrosis noted around implants. Swelling resolved.
3	Pug	9 mo	F	SSS	5	0	5 y	Neurological deterioration (Grade 4).	n/a (no imaging).	Investigations declined declined by owners.	Euthanasia due to deterioration to non-ambulatory status and non-responsive to NSAIDs and rest.
4	Pug	4 mo	M	SSS	4	1	7 y	Severe ambulatory paraparesis.	Implant failure.	Surgery declined and placed on cart (for 2 y) as eventually non-ambulatory and urinary incontinent.	Euthanasia due to concurrent condition.
5	Pug	5 y, 4 mo	MN	SSS, HL	3	1	2 mo		Osteolysis surrounding implant.	No neurological deterioration.
6	French bulldog	1 y 8 mo	M	SSS, HL, D	4	2	n/a	No complications.	n/a	n/a
7	Pug	6 mo	F	SSS	3	0	n/a	No complications.	n/a	n/a
8	Pug	7 mo	M	SSS, DL	4	3	3 mo and 1 y	1 y — Mild neurological deterioration.	Implant failure (broken pin).	No intervention as improved by the time radiographs were taken.	Remained ambulatory paraparetic.
9	Pug	2 mo	M	SSS, DL	5	1	n/a	No complications.	n/a	n/a
10	Pug	2 y	M	SSS	3	1	1 y	Neck pain.	Implant failure (broken pin) and migration.	Non-responsive to NSAIDs and rest. Surgery to remove all pin fragments and and most wires.	Significant fibrosis noted around implants. Pain-free and no neurological deterioration 5 wk post-op.
11	Pug	10 mo	MN	SSS, HL	5	1	n/a	No complications.	n/a	n/a
12	Pug	8 mo	M	SSS, HL	5	4	1 y	Mild neurological deterioration (still Grade 4) and purulent fistula over the region of the implants suggestive of infection.	Implant loosening.	Surgery to remove implants.	Fistula resolved and neurological status reverted to that before the deterioration Pain-free 3 mo post-op.

M — Male; F — Female; N — Neutered; n/a — Not applicable; SSS — Spinal segmental stabilization surgery; HL — Hemi-laminectomy; DL — Dorsal laminectomy; D — Durotomy.

All dogs had SSS as previously described (8), with 2 also having a dorsal laminectomy, 3 a hemilaminectomy and 1 a hemilaminectomy and durotomy. Pre-operatively, CVMs were identified with plain radiographs and MRI in all dogs except Dog 4 who only had plain radiographs. Dogs 6, 11, and 12 had both MRI and CT performed. Hemivertebrae were detected at T5–T6 in Dog 1, T7–T8 in Dogs 3, 9, and 12, T6–T7 in Dogs 4 and 5, T7 and T9 in Dog 6, T5–T8 in Dog 7, T8 in Dog 8, T6–T8 in Dog 11. In Dog 2, multiple vertebral malformations in the thoracic spine were detected, associated with marked scoliosis. A left-sided hemilaminectomy was performed in Dog 5 due to a T6–T7 disc protrusion and a right-sided hemilaminectomy and durotomy in Dog 6 at T7–T9 due to spinal cord compression caused by a combination of severe dorsal subarachnoid diverticulum, vertebral abnormalities, and mild protrusions. Dogs 8 and 9 had a dorsal laminectomy and Dogs 11 and 12 had a hemilaminectomy (left- and right-sided, respectively) to alleviate static compression at areas of vertebral canal stenosis (T8, T7 T8, T6–T7, and T6–T8, respectively). Most dogs presented initially with a moderate ambulatory paraparesis (Grade 3, 5 dogs), 3 dogs had severe ambulatory paraparesis (Grade 4), and 4 were non-ambulatory paraparetic (Grade 5). All dogs had improved post-operatively (Table 1). Specifically, dogs were either neurologically normal (Grade 0) or had mild ataxia (Grade 1) except for 1 patient that had mild and consistent ataxia (Grade 2), 1 with moderate ambulatory paraparesis (Grade 3) and another with severe ambulatory paraparesis (Grade 4).

Five out of the 12 dogs (Dogs 5, 6, 7, 9, and 11) showed no chronic complications with follow-up periods of 1, 2, 3, 9, and 12 y after surgery. These were 4 pugs and a French bulldog which also varied in age and gender. Two dogs (Dogs 5 and 8) showed acute complications, one of which was reported previous (8) and 7 dogs (Dogs 1, 2, 3, 4, 8, 10, and 12) showed chronic complications as follows.

Chronic complications

1. Spinal pain was noted in 2 dogs: Dogs 1 and 10. Dog 1 presented with mild spinal pain but no neurological symptoms which resolved with non-steroidal anti-inflammatories and rest. Dog 10 (Figure 1 A) presented with neck pain and implant breakage and required revision surgery to remove the entire pin and most of the wires.

2. Two dogs presented with swelling over the region of implant placement: Both Dogs 2 (Figure 1 B) and 12 presented with swellings over the region of the implants suggestive of bacterial infection. In Dog 2 this occurred 2 y after surgery and resolved with rest and NSAIDs but reoccurred at 5 y after surgery and was non-responsive to rest and NSAIDs; thus, culture and sensitivity was performed confirming infection. Dog 12 presented with a purulent fistula over the region of the implants 1 y post-operatively. Both resolved with surgical removal of the implants. Dog 2 also had lytic lesions of the dorsal spinous processes surrounding the implants seen on radiography. Large amounts of fibrosis and scar tissue were noted around the implants during surgery to remove these.

3. Neurological deterioration was noted in 4 dogs. Dogs 8 and 12 (concurrent spinal swelling as above) had mild neurological deterioration, and no intervention was undertaken for Dog 8 as he improved by the time the radiographs were taken. Dogs 3 and 4 presented with marked neurological deterioration. Dog 4 developed severe ambulatory paraparesis 7 y after surgery progressing to a non-ambulatory status with urinary incontinence. This dog did well with a cart for 2 y until he was eventually euthanized due to a concurrent condition. Implant (Steinman pin) breakage was noticed on imaging in Dogs 4 and 8. Dog 3 was not investigated; she presented with frequent pelvic limb collapse and severely delayed or absent postural reactions in both pelvic limbs which developed 5 y after surgery. Investigations were declined by the owners and due to continued progressive deterioration to non-ambulatory status and no response to conservative treatment, euthanasia was performed. One dog presented with no neurological deterioration (Dog 5) but had focal lysis of the dorsal spinous processes surrounding the implants with no associated clinical signs.

Figure 1.

A — Radiograph of Dog 10. B — Radiograph of Dog 2.

Most chronic complications were similar in nature with overlapping features. All chronic complications which were investigated were associated with the implants. Interestingly, fibrosis was noted around the implants in 2 of the 3 dogs in which surgical removal was required. Infection was suspected in both dogs (Dogs 2 and 12), although only 1 was confirmed with bacterial culture and sensitivity (Dog 2). Removal of all (Dog 12) or most of the implants (Dog 2) led to resolution of the infections.

In summary, out of the 7 dogs with chronic complications, 2 improved with conservative treatment (Dogs 1 and 8), 3 required surgery (Dogs 2, 10, and 12), 1 did well with a cart as revision surgery was declined (Dog 4), and 1 was euthanized as there was no investigations and subsequent neurological deterioration (Dog 3).

Discussion

The rate of chronic complications was 58% (n = 7); 4/7 of these dogs were managed without further surgical intervention, although 2 of these dogs were severely impaired. Studies examining complications of other surgical techniques for the stabilization of abnormal regions in the spinal cord such as with pins, screws, and PMMA have also been published (10,11). Although this technique was performed for vertebral fractures and luxations in 1 study, notable long-term complications were observed such as PMMA reactions requiring implant removal occurring at 9 to 18 mo after surgery (4/5 cases) and an infection rate of 60% suggesting possible association to PMMA exposure (10). Another study observing complications from this technique (2 to 41 mo after surgery) for the use of atlantoaxial instability indicated promising results with 8 out of 12 dogs showing no post-operative complications (11).

Most dogs in this study were pugs, which may reflect the fact that they are diagnosed with CVMs as the cause of neurological signs more often than other breeds (2). Male dogs were also overrepresented. Although this study did not aim to examine the effectiveness of this method to treat dogs presenting with myelopathy associated with CVM, further work should look at whether signalment factors may influence treatment method.

The results of this study showed that SSS led to initial neurological improvement in all dogs with myelopathic signs associated with CVMs, kyphosis, and spinal cord compression. Acute complications have already been published and are relatively common (8). In this follow-up study, the chronic complications associated with SSS, with or without decompression, were examined. The occurrence of chronic complications varied from 1 to 7 y after surgery, which could be attributed to various factors, including different activity levels among dogs (e.g., traumatic fall of Dog 1) and variations in surgeon’s experience and choice of surgical techniques and implants (12). Two of the dogs included in this series with chronic complications (Dogs 3 and 4) had late neurological deterioration. Neither of these were investigated further with imaging and they are assumed to have been complications of SSS surgery. Dorsal spine lytic lesions were seen on post-operative radiographs in 2 cases, which may suggest that this is a common complication of the procedure which may remain undetected. It may be that a “cheese-wiring” effect of wire movement within the relatively thin bone of the dorsal spinous processes is the cause of this, perhaps suggesting that pins should be cerclaged to as many processes as possible to minimize loss of stability. Three cases which developed implant loosening requiring revision surgery had lytic lesions, but no further stabilization was performed and no deterioration in clinical or neurological status was seen following implant removal. At revision surgery, the finding of fibrosis around the implants may contribute to long-term stability.

This study did not aim to examine the effectiveness of this technique, nor its suitability as a treatment method for any specific spinal cord condition. Anecdotally, we have seen good results when using the technique in cases of arachnoid space disorders associated with pug dogs, but we were concerned enough about chronic complications to want to examine those in more detail.

The limitations of our study are that it is retrospective, has a small study population, no treatment control group, and different surgeons were involved in the treatment of these cases. Furthermore, 50% of the cases reviewed also had a concurrent decompressive technique performed which may have influenced the results by worsening instability. In addition, we cannot determine if the neurological deteriorations seen (e.g., in Dogs 4 and 8) were the result of implant failure or if pin breakage had occurred in dogs that did not deteriorate as not all cases had further imaging or surgery performed. The spinal stabilization technique was also not standardized across all cases, with cases having different diameters of Steinman pin and different numbers of cerclage wires employed to perform the stabilization. Moreover, the unconfirmed nature of the neurological deterioration in 3 dogs (Dogs 1, 3, and 8) highlights an important drawback with SSS. This is because advanced imaging such as MRI cannot be performed in SSS-treated cases to determine the exact cause of the neurological deterioration. Nevertheless, complications were quite homogeneously distributed within both groups (Table 1).

In conclusion, this case series showed that SSS carried a high rate of chronic complications but many dogs did not require revision surgery and most dogs had improved with a good quality of life before these developed. The 3 dogs which did require revision surgery did not deteriorate post-operatively following implant removal, suggesting that long-term improvement does not require the permanent presence of the implants. Despite the limitations, the findings of this study support the continued use of SSS in selected cases of CVM in which neurological deficits are relatively mild and there is owner awareness of potential chronic complications. Future randomized controlled clinical trials assessing the efficacy and rate of complications of SSS in symptomatic hemivertebrae are required for establishing solid conclusions. CVJ