Neurologists and researchers working on SM in cavaliers gathered in London at the Royal Veterinary College on November 11, 2006, to share research papers and exchange views. This is a summary I made of of each of the 10 talks given at the event. Please note the researchers also issued this summary of their roundtable discussion that concluded the meeting.

Paper 1: Current Approach to Syringomyelia in the Netherlands
Paul J.J. Mandigers, DVM, MVM, PhD, DECVN; DRNVA-Internal Medicine, Utrecht University
Key points:
Both SM and MVD are known problems within the Dutch CKCS population that have reduced life expectancy from 13-15 years (for dogs of similar weight) to 8-10 years
Because SM has only been addressed on a consulting basis (eg when people have brought in an affected dog) it is impossible to make estimates of prevalence. However SM is expected to become more prevalent because asymptomatic cavaliers have been allowed to breed
Due to a private initiative from a Dutch breeder that started in 2002, 196 CKCS have been MRI-screened for SM. The point of screening is to identify the best possible breeding animals. Of these:
30% were graded A/B; 14% were C; 39% D, 13% E and 4% F, using the Rusbridge grading scheme. Overall 56% had SM. Blood samples were taken for all dogs for DNA research.
In a two year period, 2004-2006, Dutch breeders have increased the number of dogs scanning as A/B from 10% to about 25% by concentrating on screening relatives of clear dogs rathe
r than random dogs
Early observations: 14 Breeders demonstrating best use of their clear dogs
Coping with large number of affected dogs
Initially A x D crosses predominated (not enough As)
Now A x A mating is an option
Up to 4 generations “SM clear”
SM Clear dogs related to other SM clear dogs
No SM affected dogs from A x A crosses

Summary of talk:
(Clare Rusbridge read this paper as Paul Mandigers could not attend)
Of a population of 74 screened for one project, 75 per cent had SM. There is a wide range of affectedness -- not all cases are painful (about 30%). The worst pain is experienced when the syrinxes are wide, and early onset cases tend to be the worst as well.
It is unrealistic to expect to eliminate SM as there are too many affected dogs (similar to the case with MVD) and some dogs with SM are free of clinical signs making it hard to pinpoint dogs with it by breeding age unless MRId. So the goal must be to increase the number of symtom-free CKCS.
Breeders are in a difficult place because they are told, don’t breed from affected dogs. However: individual breeders may not have any clear dogs; and affected dogs may be valuable champions and valuable in other respects eg good hearts. Breeeders also really need a screening test for young dogs, but because the condition is progressive, we don;t currently have the ability to pick an ideal age for screening. The best compromise (and that used by the Dutch breeders in this project) are the breeding guidelines suggested by Clare Rusbridge which allow some affected dogs with other valuable qualities to be bred to clear dogs, and which uses the age of 2.5 yrs to fit with the heart protocol (NB CR noted the age was arbitrary but was intended to fit breeders’ use of the heart protocol).
In summary: don’t breed dogs with clinical signs (F dogs); don’t breed dogs with early onset SM (E dogs); breed asymptomatic SM dogs (D)to clear (A/B) dogs; breed at risk (C) dogs with clear (A/B) dogs only.
In the Netherlands, 14 breeders set up the Cavalier Guild for Health and have been screening their dogs since 2000. This is a breeder-led selection of dogs (so no researcher control over which dogs are scanned), and dogs are selected in order to identify the best animals for breeding and with the aim of producing pain-free dogs.
In 2004, 70 dogs were scanned showing a 70% incidence of SM and only 10% were clear A/B dogs. Scanning showed that clear dogs tended to be closely related to other clear dogs while affected dogs were closely related to other affected dogs. By concentrating on screening relatives of clear dogs and screening early -- at 7-18 months to identify good candidate offspring to know which were worth running on -- by 2006 196 different dogs had been screened and about 25% were clear A/B dogs. All sires and dams are screened; puppies are homed with a one year health guarantee.
This breeding programme demonstrates that significant successes can be achieved in only a relatively short period period. Breeders were coping with large numbers of affected dogs -- initially they had very few A dogs and had to do lots of AxD matings. Now, AxA is an option for them. They have had up to four generations “SM clear” (A/B offspring) and no AxA mating has yet produced an affected offspring. SM clear dogs are definitely related to other SM clear dogs and those are the lines the breeders are focusing on.
He hopes to do follow up screenings with dogs to see how they progress.
Paper 2: Update on Chiari Malformation/Syringomyelia Genetic Research
Clare Rusbridge, BVMS, DipECVN, MRCVS; European and RCVS Specialist in Veterinary Medicine
Key points:
Identifying genes responsible for the Chiari-like malformation (CM) with or without syringomyelia (SM) will help a better understanding of the underlying pathogenic mechanisms for better diagnosis, prognosis and clinical management of the condition
DNA research should also help unravel the embryonic development of the affected structures that create the malformation
We have constructed a 10,000 dog genealogy of related CKCS spanning 24 generations across three continents (N America, Australia and Europe) and established a DNA collection of 100 samples mainly from CKCS but including 6 other affected breeds (King Charles Spaniels, Yorkshire terrier, Bull terrier, Boston Terrier, Brussels Griffon, Chihuahua)
10 dogs were selected for genotyping with 122 markers distributed among the 38 autosomes and X chromosomes. Next, we recently completed the genotyping of 173 CKCS over 249 microsatellite markers distributed over the 28 autosomes and the X chromosome
Because CKCS dogs are so closely related and from a closed gene pool, linkage disequilibrium (LD) analysis is a promising strategy for gene mapping.
Preliminary results of LD analysis identified six statistically significant regions on six associated chromosomes. Possible candidate genes are in the Hox and Pax gene families
Summary of talk:
In humans CM/SM is rare: current estimates are that 1 in 1,290 have the Chiari malformation. SM occurs in 65-80% of humans with Chiari, and 80% experience pain. Genetics are thought to be important but the mode of inheritance is unknown. However, researchers have observed co-segregation with known genetic “bony” (i.e. mesodermal) syndromes e.g. achondroplasia, Klippel-Feil syndrome, primary basilar impression and Goldenhar syndrome. They have also reported familial clustering with vertical and male-to-male transmission consistent with an autosomal dominant mode of inheritance. A human genome scan was recently started and candidate genes are being investigated.
Early observations in the CKCS are:
CM/SM is not simple recessive, but a complex trait -- either polygenic and/or dominant incomplete penetrance or variable expressive is possible.
Clinically affected offspring often have affected (clinically or sub-clinically) parents
Often worse with each generation
At present confirmation only by MRI
All clinically affected dogs share small number of common ancestors (6 out of 8 great-grandparents of all affected dogs can be traced back to two female ancestors)

What is the Chiari malformation? It is thought to involve the somatic mesoderm at the basicranium and craniovertebral junction. There may be para-axial mesoderm insufficiency after the closure of the neural folds. This leads to underdevelopment of the basichondrocranium. The consequence is overcrowding and subsequent herniation of cerebellum into / below the foramen magnum.

Possible candidate genes are:

The Hox family, which controls development of the occipital bone. Ectopic expression of Hox-2.3 results in dysplasia / deficiency of occipital, basisphenoid and atlas bones in transgenic mice (McLain et al. 1992).

The Pax group of genes
Pax-1: which affects somatic segmentation and sclerotomal differentiation in the cervico-occipital transitional zone
Noggin: which affects growth /differentiation somites in the paraxial mesoderm. But: in 33 cases of CM that were analysed, no variants were identified (Speer et al. 2003).

The goals of the genome project are:
Genetic mapping of the CM and SM gene(s) by linkage disequilibrium analysis in the CKCS breed and in other related breeds
Identification of candidate genes for CM and SM using the positional candidate gene approach
Molecular characterization of the gene(s) mutated in CM and associated SM.

The DNA collection programme is an ongoing collaboration between:
Stone Lion Veterinary Centre, UK
DNA Archive for Companion Animals, University of Liverpool, UK
Centre Study of Brain Diseases, Notre Dame Hospital, Montreal, Canada
Department of Genetics, Hospital for Sick Children, Toronto, Canada
DNA contributions from many referral centres, breed clubs and general practitioners (Worldwide)

Because CKCS dogs are so closely related and from a closed gene pool, linkage disequilibrium (LD) analysis is a promising strategy for gene mapping. Association studies are not possible in this breed because CM is presenting in variable degrees of expression in nearly 100% of the CKCS dogs, so comparison to other breeds is necessary -- especially as closely related breeds are more likely to share ancestral chromosomes and carry the same disease allele.
We have constructed a 10,000 dog genealogy of related CKCS spanning 24 generations across three continents (N America, Australia and Europe) and established a DNA collection of 1000 samples mainly from CKCS but including 6 other affected breeds (King Charles Spaniels, Yorkshire terrier, Bull terrier, Boston Terrier, Brussels Griffon, Chihuahua)
A first goal was to evaluate the informativeness of genetic markers. Ten dogs were selected for genotyping with 122 markers distributed among the 38 autosomes and X chromosomes. In the second genetic analysis, we recently completed the genotyping of 173 CKCS over 249 microsatellite markers distributed over the 28 autosomes and the X chromosome. We looked at multiple affected and unaffected siblings and half siblings and parents. We used Marshfield genotyping services from the NIH (National Institute of Health).
Preliminary results of LD analysis identified six statistically significant regions on six associated chromosomes. Further investigation is ongoing. We are also identifying family clusters of CM/CM in breeds other than CKCS. We have identified CM/SM in nine breeds but we need 10.
Q&A on first two talks:
The earliest case of SM CR has diagnosed was in a 12 week old puppy.
The reason ventricle dilation is recorded on MRI scans is because is may be useful for research as it might indicate an obstruction: ie the start of a syrinx. We don’t really know what to do with that information now but we record it as it may be significant for later research.
CR says she doesn’t know if selecting for a modern head shape has anything to do with the development of SM but says “I think it must have something to do with it -- but the problem is how do you know what factors are relevant. I see all kinds of head shapes in dogs with SM but I can’t find any typical head shape and say, there’s one head shape you should select for.”
CR says in original pedigree analysis all the originally clear dogs did not trace back to the ancestor dogs identified as probable origin of SM in the breed and now a few do but not too many. “Unfortunately a lot of the old, clear dogs are now gone and the lines are completely gone.”
All nine breeds they have found with SM related to CM are short-skulled. Neck and shoulders could be a factor. It hasn’t been seen in cockers. It is rumoured to be in pekes but they have not seen evidence. In King Charles spaniels some are clear of both CM and SM but have hydrocephalus.

Paper 3: Incidence of caudal occipital malformation syndrome and ultrasound, computed tomographic, magnetic resonance imaging findings in 16 clinically normal Cavalier King Charles Spaniel genitors
Laurent Cauzinille, DMV, Ecole Nationale Veterinaire de Maisons Alfort, Paris, France
Neurology and Neurosurgery Department, Centre Hospitalier Fregis, Arcueil, France
Authors: Jerome Couturier, Delphine Rault, Laurent Cauzinille
Note: a rough translation from French of his published article on this study is here. The French version is here.

“We had two years ago, what we called ‘the French exception’. There was ‘no SM in France’,” Cauzinille said. The French CKCS club referred to SM as ‘the British CKC disease.’ Then, a couple of cases appeared and now there are more than 25 known cases.
In November 2004 he gave the first lecture and produced an article on SM in cavaliers. In June 2006, the French club accepted his proposed study, which was small but with two key French breeders, conducted at the the Vet Imaging Centre in Paris.
* 16 dogs in the study
* all were clinically and neurologically ‘normal’ according to their breeders (who filled in a questionnaire) and verified by neurologist’s tests (but not MRI)
*all were LOF (French kennel club pedigree) and breeding dogs
* Dr Couterier performed the neurological clinical testing, which included checking for: cervicalgia, hyperesthesia, scratching, ataxia, paresia, abnormal head posture, scoliosis, strabismus, facial paralysis, cranial nerve deficit
A number of study tests were then done.
*Ultrasound was performed on awake dogs with the goal of trying to find a less expensive screening test. Diameter of the spinal cord was measured at C1/C2 and at the level of the foramen magnum with the head flexed 90 degrees.
* Dogs were then anaesthetised, and a spiral CT scan was performed from the tuberculum sellae to the dens of the axis. Sagittal, transverse and dorsal images were taken (which is quite extensive). Linear measurements and area of the caudal fossa were obtained in the sagittal plane; height of the foramen magnum; area of the caudal fossa. In the transverse plane at the level of the widest part of the caudal fossa, linear measurement of the maximum width of the fossa was obtained.
*Then, dogs were MRId in a dorsal recumbent position using a 0.3T MR system -- the skull to the 1st thoracic vertebrae. They measured hydrocephalus, ventricles, cerebellar size and degree of herniation.
As a result of these tests, the population was divided into groups: 1) normal; 2) syrinx; 3) hydrocephalus but no syrinx.
1) normal dogs: 8, 7 males and 1 female -- but had Chiari-like malformation
2) syrinx: 7 had SM, 3 males and 4 females, “a big shock to the breeders”
3) hydrocephalus but no syrinx: 1 male
In all the dogs, no statistically significant difference was found in the measurements taken except in the hydrocephalic dog. With this group, 43% of clinically normal champion breeding dogs (genitors) have SM “so there’s no more ‘French exception’.” But this is only 16 dogs so researcher cannot give a percentage affected for France, but these were were all pedigree dogs representing 6 key breed lines all with common ancestors.
* There was no association between the degree of herniation and SM. Doesn’t know if size of caudal fossa matters.
*They were able to pick up syrinxes with ultrasound
* They will repeat MRIs on the French dogs in 1 year and 3 years. He hopes King Charles Spaniels can be included in future research.
* The doctor doing the ultrasound was impressed with how much of the cerebellum could be seen with ultrasound.
* He thinks the next step is to keep looking at CSF dynamics -- “the thing that will really help us to progress a lot.”
Paper 4: Association between cervical and intracranial dimensions and syringomyelia in the Cavalier King Charles Spaniel
Harvey Carruthers, BVMS MRCVS

The aim of this study was to establish if intracranial and vertebral measurements can predict the occurrence of SM. The study tested the hypothesis that SM results from caudal fossa overcrowding and is influenced by cervical vertebral abnormalities.
MRI scans of 85 CKCS were reviewed. Seven dogs were excluded because of missing or incomplete MRI data.
* Of 78 dogs, 59 had syrinxes; average age of dogs at time of MRI= 3 years+/-2.05. “SM in cavaliers is predominantly a young animal problem.” But older cavaliers were more likely to have a syrinx.
* dogs without syrinxes were used as controls
* MRIs were made anonymous when reviewed (so no one knew the specific dogs they were viewing) and read by three people at two sites
* the three paper authors studied sagittal and transverse T2 -- weighted images -- for the presence or absence of syrinxes
Cervical measurements: they took top to bottom height measurements of C2-C3 vertebrae, then measured the angles on the C2-C3. Specifically:
* the caudal fossa base length and height were combined to give a measure of caudal fossa volume (“the caudal fossa triangle”).
*the widest vertical diameter across the spinal canal, at C1/C2, at C2, at C3; the widest point across the C2/C3 joint; and the angulation of the C2/C3 spine were measured
* These measurements were compared to those of dogs without SM
There were no significant correlations between the caudal fossa volumes measured and incidence of SM, except for reduced volume. But there were some interesting measurements in cervical width of C2-C3 junction associated with the presence of syrinxes, and also between the width of the widest point of C3 and the presence of syrinxes.
* A greater width at C2/C3 is associated with a syrinx
* A greater width at C3 is associated with a syrinx: this was a surprise finding fro the group
These results may contribute to the predictability of syrinx occurrence. They could possibly also help in the development of testing for SM not reliant on MRI (eg x-rays). Further study is planned to see if cervical findings are significant in the parthenogenesis of SM.
There was a significant difference in the C1/C2 diameter between dogs showing pain and those without pain.
* The presence of pain is associated with the narrowest point from C1/C2 to the dens
* Dogs with pain had the narrowest measurement of .93cm versus .99cm for no pain
Paper 5: Association between spinal cord dorsal involvement and pain in syringomyelia secondary to canine Chiari malformation
Nick Jeffery, BVSc PhD CertSAO DSAS (ST), DipECVN, DipECVS, FRCVS

This study was designed to test the hypothesis that pain associated with SM is related to spinal cord dorsal horn damage, and look at the relationship between pain and syrinx dimension.

CM/SM Pain:

* Commonly localized to head or neck
* May be difficult to define or intermittent
worse at night
when first getting up
may vary with atmospheric conditions
when excited
related to posture
* May be sensitive to touch
* Can scratch at shoulder, ear, neck or sternum
typically one side only
whilst moving, often without skin contact

Is pain associated with damage to the dorsal horn? To find out, researchers need to correlate signs of pain with syrinx dimension. The top part of the spinal cord is where pain is processed.

Looked at the size and location of syrinxes in 85 CKCS from measurements made over a two year period (60 from breeder screening programs). 11 dogs were excluded leaving 74, 60% female. Images were coded and viewed independently by three of the authors of this paper.

Is there a syrinx? If so:
* measured the maximum width of the transverse image (any angle)
* is there dorsal or ventral asymmetry?
* if syrinx is asymmetrical, how long was the asymmetry (the reason to look for this is that it has been associated with pain in humans with SM)

* 55 of 74 CKCS had syringomyelia
35% SM were painful
27% SM had scratching behaviour
* Comparison pain SM and no pain SM
No correlation sex or age
Strong association with maximum syrinx width (p<0.0001)
Dogs in pain - mean maximum width 0.58cm
Dogs without pain - mean maximum width 0.32cm
95% of CKCS with SM greater/equal to 0.64cm were painful

There is a very strong association between maximum syrinx width and pain:

* Asymmetry of syrinx is only found in the dorsal half of the spinal cord, and is associated with pain. 79% of dogs with pain had such a syrinx.
* syrinx length is also associated with pain.

Results – dorsal asymmetry:
* Syrinx asymmetry only found dorsal half spinal cord
* Dogs in pain more likely to have dorsally asymmetrical syrinx
15/19 (79%) - dogs with pain
16/33 (49%) - dogs without pain (p=0.0419)
* Mean length of asymmetrical syrinx
5.15cm - dogs with pain
2.8cm - dogs without pain (p=0.0039)

But the association of length with pain is not significant when correlated with width. The strongest predictor of pain is syrinx width.

Disordered neural processing in the damaged dorsal horn is a likely consequence of SM and can cause neuropathic pain that often responds poorly to conventional analgesics.
Clinical significance:
* Syrinx width strongest predictor of pain
95% of CKCS with SM greater or equal to 0.64cm are painful
* SM likely to cause neuropathic pain
Signs of SM suggest
Dysaesthesia - unpleasant abnormal sensations
Allodynia - pain due to a stimulus that does not normally provoke pain e.g. simple touch
Hyperalgesia - increased response to a painful stimulus

* Dogs tend to scratch to the side of syrinx asymmetry. This is due to an inbalance in input to the thalamus
* Doesn’t think it is possible to pick out heads more likely to have SM
Paper 6: Thermographic imaging of the normal canine
Catherine Loughin,
Long Island Veterinary Specialists

Thermography is a non-invasive technique that detects abnormal functions in the sensory and sympathetic nervous system. Thermographic machines were introduced in 1959 and vets started using them with horses in 1965.

LIVS use a focal plane array detector scanner, and can scan in real time; they get a high speed image and spacial resolution and software analysis is done on the image.

Thermography is now used in human and equine medicine and can be used to diagnose such things as injuries, cancer and arthritis. It hasn’t really been used on dogs, which (unlike humans and horses) are hairy – so there’s a need to establish a protocol for imaging dogs. This is a project LIVS has been working on.

They used 10 normal, young adult Labradors. They imaged the dogs before clipping, after clipping, and at 15, then 60 minutes after clipping and 24 hours after clipping.

• full hair coat results differed significantly
• difference: 6 degrees C. –plus for unclipped to clipped
• but clipping hair does not change the thermal image
• found that after one hour, they can image the dog as temperatures normalize

Now, LIVS are doing a major study using thermography with elbow dysplasia, cruciate disease, and COMS/CM/SM.

Thermographic images of humans with SM have shown some changes in the trunk and extremities and in dogs with COMS/CM/SM, you have the same effect.

At the moment, LIVS are trying to prove whether you really need to clip the dogs as obviously it would be better for owners, especially breeders who show, if you didn’t have to clip them.

Paper 7: Cause and effect: reflections on the presence of syringomyelia and the relationship to presenting signs
Dr Jim Anderson,
University of Glasgow

At U. of Glasgow, they have a small animal neurology unit.

From a clinical perspective, this is an extremely important disease for us, he says. We have this imaging phenomenon of the MRI, but clinically, SM presents us with a number of problems. Lu and others in a study in 2003 found that there are a large number of clinical signs associated with MRI findings. They were looking mostly at the skull and not the spine. In some cases, you can see the correlation between the image and the clinical signs but with others, it is hard to tell. So: is it possible to refine the image findings to the clinical findings? What are the important imaging findings? What about occipital dysplasia without a syrinx?

What clinical symptoms are associated with SM? Some are likely to be cause and effect.
• Abnormalities in the spinal cord seem to be strongly associated with signs of pain.

• Seizures: an association with SM has been described, but it seems unlikely in his opinion.
• PSOM: Lu saw this in 32% of SM cases. In 2004-05 of the cases they saw at Glasgow, 8 of 33 had significant ear disease. Many with vestibular disease had an alternative explanation for symptoms other than SM. Are people taking enough consideration of PSOM? In our experience it is an important consideration in pain considerations. In one dog they treated, it had the Chiari-like malformation and glue ear (PSOM). They treated the dog for the PSOM and the pain went away. Thus PSOM is probably a real concern and is presenting as a real problem.

He fears minimal MRIs used by some neurologists for assessing SM are “risky for diagnostic evaluation as they may miss ear disease”.

At Glasgow they are sometimes seeing a referral to a neurologist from inexperienced vets and there is an owner expectation of SM as well. In some cases these vets missed other problems because they made assumptions about SM without checking for other causes of symptoms.

Most of the cavaliers they have coming into Glasgow are queries on SM. Of six recently, five had other explanations for their clinical problems. In situations like their neurology unit where the dog will be properly screened, the causes of the problem will be caught. “What does concern me is that there are assumptions by some vets who don’t know much about SM that SM isn’t really treatable (and hence dogs might remain untreated or be euthenised) and therefore these symptoms aren’t being investigated and other diseases are probably being missed. While SM is a very serious concern in cavaliers, it is also in danger of becoming a ‘bucket diagnosis’. A major risk is that no investigation is done because of a vet’s assumptions that a dog has SM.” CKCS suffer from other diseases too, some similar to SM in symptoms, which are reviewed by Clare Rusbridge in 2005. Vets need to be aware of these.


He says breeders have sometimes put cavaliers in for MRIs on the basis of clinical signs and get the MRI done on their insurance, then they aren’t diagnosed with SM. He thinks they are doing this to diagnose for their own breeding purposes.

Most of the cavaliers they see at Glasgow come from pet owners whose vets have given them a possible SM diagnosis.

He says neurologists need to be including the ears to check for PSOM, and that dog owners should not rely on a local (vet) practitioner making a diagnosis based on an assumption of SM but need to get accurate data and an expert diagnosis (eg need to get a diagnosis from a neurologist or similar expert and an MRI). Also that vets need to eliminate other causes first.

Clare Rusbridge and other neurologists attending -- eg Dr Jeffery and Dr Marino -- pointed out at this point that all the MRIs they know of being done for CM/SM, and all those at the low cost screening clinics, do diagnose for PSOM and include the ears and the top vertebrae. PSOM is listed as an alternative diagnosis for symptoms on Rusbridge’s Introduction to Syringomyelia, which is available on several CKCS club websites and other SM information websites.

Paper 8: MRI assisted morphometric and anatomic-histological examination of the Chiari malformation in the Cavalier King Charles Spaniel, a brachycephalic breed of dog
Martin Deutschland,
Chestergates Animal Hospital, Cheshire

This work shows the use of MRI imaging in the diagnosis and morphometric evaluation of anatomical changes of the caudal fossa in CKCS with Chiari malformation.

The goal is:

1) to achieve a reduction in symptoms for affected dogs. At Chestergates their current approach for diagnosis is to perform a neurological exam, which possibly includes an MRI. For the individual dog, treatment aims to reduce the amount of CSF with diuretics as a first step. If further intervention is needed, Geoff Skerritt prefers placing a shunt in the syrinx to release pressure.

2) to exclude severely affected dogs. This can be done in two ways: a) selecting by severity of symptoms; b) selecting by anatomical changes.

His focus was to select by the degree of anatomical changes and study that aspect.

For the study, a total of 77 CKCS and 32 other breed dogs were obtained and evaluated morphometrically and graphically: 37 CKCS without a syrinx and 40 CKCS with a syrinx; 30 non-CKCS without a syrinx and 2 non-CKCS with a syrinx.

Gender of CKCS: 42% male
Age: .5 to 11.5 years old, with an average of 2.5 yrs and studies going back to 1999.
Coat colour: In this sample, blenheims had the most syrinxes; tricolours had lowest number of syrinxes.

• cerebellar herniation had a significant impact on whether there was a syrinx. NB he took the same approach as Lu had in his study, but Lu found the opposite, that there was no relation between herniation and syrinx.
• a normal dog cerebellum has a square shape. In cavaliers, it is more of a wedge, a triangular shape, because it is being compressed even in ‘normal’ cavaliers.
• He did measurements of the fossa. The most interesting measurement was the height and length of the cerebellum, what he calls the height and the’half-length’ of the cerebellum.
• He couldn’t find any relationship between fossa volume and the presence of syrinxes. BUT: the ratio of cerebellar length/height was less than .63 in 7.5 % of the syrinx group but 51.4 % of the non-syrinx group.
• This might be a beginning, to have a measurement that is significant for predicting the likelihood of SM.

One particularly interesting CKCS he studied: Carlina
• died at age 10
• had three litters of puppies
• after her last whelping, she developed severe scoliosis and scratching

He did a post mortem examination. (Here he showed a number of slides from the post mortem). The bone was very thin on the back of her skull and there was a convex bulge at the point of her herniated cerebellum. This is very unusual and is probably an indication of the high pressure within the skull at some stage. It doesn’t look like there was hydrocephalus. She probably had the bulge from the time she was born. (He noted here that he would encourage people to donate bodies of cavaliers because such information is invaluable in better understanding SM).

The whole spinal cord had shifted to the right around her syrinx, and the syrinx has actually ruptured the spinal cord.

He says that it is in some cases possible to see syrinxes even after death. Hydromyelia probably dries out but syrinxes probably stay longer and so can be seen in a postmortem at least some of the time.

Even after surgery –either shunt or decompression – you would be able to tell where the syrinx was.

Conclusions in his paper:

• there were no significant age or gender related differences related to syrinx formation, but there were a larger number of blenheims affected in the syrinx forming group
• the CKCS cerebellum seems generally greater ventrodorsally and more caudally elongated that that of the control group.
• The length of herniated cerebellar tissue and the quotient derived from the caudal cerebellar rostrocaudal dimension and the cerebellar ventrodorsal dimension both seem to have a statistically significant effect on syrinx formation
• The volume of the syrinx does not seem to be influenced by morphometric cerebellar changes
• There are only small discrepancies in the measurements of the caudal fossa of the CKCS group and the control group. But an exceptionally large number of dogs with a large foramen magnum and a short squama occipitalis are CKCS. This does not seem to influence syrinx formation however
• The parthenogenesis of Chiari malformation remains poorly understood. But the shape of the cerebellum, the dorsally enlarged foramen magnum with an absent dura mater at this point and the convex and the squama occipitalis allow the formation of a high CSF pressure gradient between the ventricles and the cerebellomedullary cistern during embryonic development
• Syrinx formation seems influenced by a cardiac systole-dependent piston-like effect created by the herniated cerebellar tissue. It is likely that high blood pressure would cause transparenchymal osmosis of CSF
• This research suggests some parameters through MRI screening that may guide the choice of breeding stock that are less likely to perpetuate Chiari malformation and this should lead to gradual change in the cerebellar rostrocaudal dimension and the length of herniated tissue

Paper 9: Long term follow up after surgical management of canine Chiari malformation with syringomyelia
Clare Rusbridge,
BVMS, DipECVN, MRCVS, European and RCVS specialist in veterinary neurology
Stone Lion Veterinary Centre

The purpose of the study was to evaluate the long term success for foramen magnum decompression for management of CM/SM in 15 CKCS.

15 cases of dogs that have had decompression surgery after diagnosis of CM/SM by MRI, were tracked over a six year period. The clinical outcome was monitored for at least 12 months.

Before surgery, dogs were allocated a pain score.
0: no pain
1: occasional pain (eg one or fewer scratching sessions per week)
2: mild neuropathic pain (eg occasional scratching)
3: moderate (eg consistent shoulder scratching)
4: severe (eg consistent scratching; body parts can’t be touched, frequenting screaming)

Other signs of pain can include scoliosis, signs of ventral horn damage, seizures, long tract signs (but she doesn’t think this is associated with SM).

All 15 dogs had cranial cervical decompression. When operating, she doesn’t take off the C2 vertebrae or ligaments as she was worried that all of C2 could be weakened. She removes the dura, which is “very tough and probably contributes to compression”.

• 8 dogs had durotemy alone
• 7 dogs had durotemy with a Biosist patch
• dogs were reassessed at 2, 6 and 12 months post-op
• if assessment was missed, it was made by phone, email and reading the primary vet’s comments
• End point: hasthe dog improved, deteriorated or remained stable

• 6 male, 9 female dogs, were diagnosed with CM/SM between the ages of 12 weeks to 4.5 years (mean: 2.2 yrs), and had surgery between .8 to 8.1 years of age (mean 3.3 years)
• duration of symptoms before surgery, from .1 to 4.2 years (mean, 1.1 year)
• reason for surgery: 10/15 dogs, severe symptoms
• 1 epileptic dog had epilepsy after surgery and was treated
• 3 dogs were slightly hypermetric and alaxic (for 48 hours)
• all exercised normally within circa 4 weeks

Of 6 dogs with scoliosis, 5 improved (she thinks they were improving anyway)
Of 4 dogs with long tract signs, all improved
Of 6 dogs with ventral horn signs, 2 improved, 4 persisted

Of 10 dogs that started with a pain score of 4:
4 dogs dropped to a 3
6 dogs dropped to a 2

Of 4 dogs with a pain score of 3:
2 dogs remained at 3
2 dogs dropped to a 2

Of 1 dog with a pain score of 2:
1 dog remained a 2

Two months after surgery, 80% (12) had improved.
7 out of 12 maintained their improvement
1 of the 12 had tonsillar carcinoma and was euthenised
4 of 12 deteriorated between .2 to 2 years post-op
2 of those were euthenised
Of the 3 out of 15 that were unchanged, 2 deteriorated between .8 to 2.3 years post-op

In summary:
• 7 (47%) out of 15 dogs deteriorated to some degree, between .2 to 2.3 years post-op; the mean is 1.3 yrs; and two were eventually euthenised as a result
• Of the 15, 12 are still alive. One is 6.5 years post-op (mean = 2.5 years)

On their MRIs: they show clinical improvement but not radiographical improvement. The clinical improvement is probably related to the relief of cerebellar pressure. But syrinxes are not collapsing in dogs post surgery.

By contrast, in humans, 90% see collapse of their syrinxes post-op, so why is the syrinx not collapsing in dogs? It is possible that:
• CSF movement remains inadequate (perhaps not enough bone is removed to restore movement?)
• herniation of the vermis rather than cerebellar tonsils?
• there’s a narrower subarachnoid space?
• Also: the atlas is just a narrow ring in dogs, not broad as in humans
• There’s reduced spinal cord compliance
• Parthenogenesis involves other uncorrected factors

Lots of surgeons say to remove the tonsils in humans – but in dogs, the vermis controls movement and obviously do not want to remove it!

Also, in humans, the longer you’ve had a syrinx, the less likely it is to collapse.

Some remaining unknowns:
• don’t know if progression is delayed by performing surgery
• don’t know best time to operate after 1st observance of clinical signs. Dr Dewey believes if you operate early, you have greater success. But from her small sample, dogs operated on early were as likely to have recurrence

Problems with the study:
• pain is subjective
o clinical history may be interpreted incorrectly
o we need a better pain scoring system
• with new signs of pain: owners may be more tolerant of clinical signs

Conclusions regarding the use of decompression surgery:
• there’s a low mortality and morbidity – no dog died
• for dogs with a high pain score, 87% had improvement in chronic pain
• improved quality of life – CAN be sustained for many years

• there may be no radiographic improvement
• improvement may not be sustained and dogs can deteriorate as close as 2 months post-op
• surgery may not delay progression
• surgery doesn’t address the primary cause of SM (whatever that may be)

We need:
• modified surgery
• better analgesic techniques
• better understanding of the pathology of CM/SM


Q: why not put a shunt in as well?
A: Shunts have more long term complications – tethering, etc. She isn’t comfortable ‘stabbing into spinal cord’

Durotemy alone seems as successful as using the Biosist patch.

Q: (from expert surgeon in human SM): In humans with SM and hindbrain herniation, they have headaches. But is the scratching due to headaches in dogs? The one thing most likely to improve for humans after a decompression is the headaches.
A: Scratching is persistent (after surgery in dogs).

Paper 10: Foramen magnum decompression with cranioplasty for treatment of caudal occipital malformation syndrome in dogs – an update on 18 cases
Dr Dominic Marino,
Long Island Veterinary Services

Foramen magnum decompression (FMD) is the preferred mode of therapy for humans with Chiari 1 malformation and is often successful but 8-30% of human patients require re-operation, due to excessive scar tissue formation at the surgery site. The scar tissue results in compressions of nervous tissue and abnormal CSF dynamics, perpetuating the secondary formation of a syrinx and neurological deterioration.

In humans, there are six forms of CM/SM and even more subforms. We are still not completely sure we are dealing with the exact same disorder in dogs.

FMD in dogs has an 81% success rate and similar to humans about 25% of dogs require re-operation due to excessive scar tissue formation post-op. In LIVS’ experience clinical signs of worsening associated with scar tissue typically occur within 3 months of surgery. LIVS have developed a cranioplasty technique that they say “has essentially eliminated the problem of post operative scar tissue compression at the FMD site”.

CM/SM affects small breed dogs, rarely is seen in dogs over 30lbs and unfortunately cavaliers are over-represented.

• anatomical variations may account for the range of symptoms
• there’s a loss of collagenous integrity with secondary cerebellar ‘slouching’, but normal volume measurements
• abnormal physiognomy of cavaliers (hind brain compression as ‘normal’ plus high incidence of malformation) can cause the direct compression of nervous tissue and the formation of syrinxes

In clinical presentation only about 6.7% scratch but that generally increases. CSF flow explains a lot of this, 87% of COMS dogs they see have a syrinx, primarily found in the cervical region. He notes: “We found this out by accident and maybe if we were looking for it, syrinxes would be a lot more prevalent. Syrinxes are often in the neck but then, that is the only place we’re usually looking. As we collect more data, the numbers change.”

The Venturi effect may explain why we are seeing COMS and syrinx formation in dogs. Just the obstruction alone (from herniation) isn’t enough to explain syrinx formation, he believes. At ASAP’s 18th conference in 2006 (American Syringomyelia Alliance Project) there was a lot of exchanging of information with experts in human SM. There was an engineer trying to explain why syrinxes form. The heartbeat causes a piston effect, so that CSF fluid squirts out. The malformation is causing turbulent CSF flow. The engineer is trying to figure out how bad is bad, in terms of the obstruction and turbulence. He did 3D reconstructions of Chiari patients and found significant differences between them.

The velocity magnitude is higher at the neck, but why is a syrinx forming? Further down, abnormal currents occur. This may explain why syrinxes also form further down.

Medical therapy: LIVS stance is ‘somewhat controversial”. They don’t recommend long term medication, and use glucosteroid therapy (steroids). Humans with SM don’t use this. LIVS are shying away from narcotics and towards gabapentin. Surgical treatment options are shunting or decompression. “Our treatment of choice is surgical decompression”. They like to restore normal CSF flow as soon as possible.

• FMD is 81.25% effective right now
• Re-operation is needed in 25% of canine patients and seems to be due to scar tissue
• In such cases, dogs usually are fine 3-6 months then start to show signs of recurrence

LIVS is only 3-4 miles down the road from the Chiari Institute so have benefited from a close relationship and the proximity to human SM experts.


LIVS is using a titanium mesh – “it makes no sense” but seems to work quite nicely. They make what they call a “Lay’s potato chip” shape out of polymethylmethacrylate (PMMA) for the mesh, which works like rebar. The titanium mesh is moulded to the dog’s head.

They have done the surgery on dogs as small as 4lbs. He feels most surgeons doing decompression are not taking enough bone out: “The tendency is to not take away enough bone which leads to scar tissue formation.” Then they do meningeal decompression, to remove the band of fibrous tissue under the skull. At first they didn’t know it was importent to remove this – but in smaller dogs he says this membrane is compressing the brain more than the bone itself.

They then sink five titanium screws into the occipital bone around the edge of the FMD defect, which will anchor the ‘potato chip’ mesh array. He says surgeons need to use a substantial amount of mesh to ‘tent out’ to allow you to maintain the decompression inside; then you cement into place. The chip slips under the screws so both screws and cement are holding it in place. At first he thought the movement of the dog’s head might snap off the mesh but it has held very well in all dogs.

There were no intra-operative complications with any dogs. Post operative complications include a mild head tilt in two dogs (both re-operative cases) and need for narcotic pain medications for a week in another dog for persistent neck pain.

In this study:

• 18 dogs (now 21-22 as a few more have had surgery and will be included). 15 were CKCS, 1 pug, 1 maltese, 1 chihuahua. 7 males, 11 females, all neutered/spayed. In two dogs, the surgery was a re-operation due to scar tissue formation. Median age 26 months (range, 12 months-98 months)
• Mean follow up time = 6 months; 6 have a longer than 6 month follow up
• 94%, or 17 of 18, have had significant improvement or complete resolution of clinical signs
• none have yet required re-operation
• one dog with PSOM was euthenised because of chronic aural pain; it isn’t clear whether recurrent pain was due to COMS or PSOM
• one dog – scar tissue formed distal to the surgical site at C1/C2 which needed surgery
• one dog has needed to continue to take gabapentin, 1 occasionally needs carprofen

Conclusions: Cranioplasty with titanium screws and titanium/PMMA plates appears to be well tolerated in dogs with COMS. Although one dog needed reoperation in this small case series, further cases with more follow-up information will be needed to ascertain to what level, if any, the cranioplasty procedure mitigates comprehensive scar tissue compression at the FMD site.

Questions still to be answered:
• we would like to know what medical therapy to use and for how long; or when to go for surgery
• we need to do genetic work
• we need screening tests that are low cost and minimally invasive
o they did look at BAER testing (but there’s still a need to anaesthetize)
o all 10 dogs with COMS had abnormalities in BAER testing, while normal dogs were normal
o so BAER may be useful – but, what if dog isn’t showing clinical signs of COMS?
o THERMOGRAPHY: perhaps will be the tool of choice? Wouldn’t have to anaesthetize; don’t know yet if you’d have to clip; but it’s way too early yet to tell if it is viable for screening

Could SM/CM be related to collagen and MVD? Doesn’t think so. Also a need to talk about the difference between CM and SM. There are five or six cases where we’ve found syrinxes with no apparent reason.

Close of day general Q&A:

Q: is there a way to set a general standard for MRIs?
A: Marino: On the schemes available, there’s been a great amount of work done to standardize. There’s also a need to standardize as we go.

Q: How does someone make the decision to choose surgery, or shunt?
A: Marino: For me, it’s a work in progress. The goal, as we understand it, is to relieve pain. We find the syrinx is resolving in new surgeries with the cranioplasty technique. Whereas without cranioplasty, we are seeing problems and no resolution of syrinxes.

Q: Why should people MRI screen?
A: Marino: Because it’s the best tool we have.

Q: What about progeny experiments?
A: With the Dutch breeders dogs with no SM are producing a higher ration of clear dogs. There are no clear dogs that have no related clear dogs. Maybe a need to try to get some funding for breeders to do progeny testing.

Q: Could oversupplementation contribute to bone malformation?
A: Dominic Marino: There is some possibility though there’s a very weak link to cranial formation. Maybe hormones would affect development but would need to be extreme doses. COMS is also restricted to a certain number of breeds. Either there’s a specific element in the breeds involved – but that’s unlikely.
Martin Deutschland: What is the chance that all these breeders began feeding and oversupplementing all over the world, all at once? He says he is looking at whether a genetic defect could lead to lack of vitamin A, etc, but thinks it unlikely.

Q: Has incidence dropped since MRIing began by breeders?
Clare Rusbridge: No, it’s far too early to tell from scanning.
Rodolfo Cappello, DVM, PhD, DipECVN, MRCVS, Royal Veterinary College: As knowledge has increased, the case load has increased.

Dominic Marino: current understanding of how syrinxes form is the Venturi effect. And is pressure from the outside forcing fluid into the canal due to this effect? If this is the case, it may push people to operate sooner because scar tissue starts to form from that effect. Cine-MRIs will help researchers to get a movie video type of MRI which would enable them to see the turbulent flow. Also, ultrasound might be helpful.

Q: Why do some dogs have dilated ventricles?
A: Clare Rusbridge: It may have to do with nervous system compliance.
Martin Deutschland: It may just be a variation – bulldog studies have shown variations.

Q: What about risks of doing a myelogram?
A: Rodolfo Cappello: It is one of the tests that is being abandoned because of complications. Specialists and vets would refuse to do that kind of test. A myelogram is always risky but perhaps more risky with dogs in this condition.
Laurent Cauzenille: We are still doing 9-10 myelograms a week. We want to see the whole image. We did find three cases of SM with myelograms. It is not the preferred tool for diagnosing SM.

Q: Could there be two forms of the disease? Some get it early and some get it late. Could there be medical reasoning behind that?
A: Nick Jeffery: There could be, but we need to see more dead animals.
Jim Anderson: It is not uncommon for dogs with neck syrinxes to have them go all the way to the tail.
Dominic Marino: It could also be symptoms of the same disease at different times. In conclusion, we don’t know, but that’s part of the research effort.
Dr Flint, neurosurgeon for human SM: Human syrinxes come in all shapes and forms. In a sense, it’s not a single disease.
Nick Jeffery: We have developed treatment methods without the parthenogenesis, which is kind of scary. We’re almost finding the causes by trying treatments.
Dr Flint: That’s why it is time to join together. SM is enigmatic. What causes it? We’ve been asking that question for years and we still don’t understand it.

End of general sessions