A bad case of Chiari

March 2012

The case for this month is slightly different – a clinical problem presented for some discussion. This is the information we have:

Presenting complaint

The patient is a 13 year old wheelchair-bound girl born with a thoracolumbar myelomeningocele. She was managed elsewhere but presented to us with progressive deterioration in bulbar and upper limb function over the past 18 months.


She was born by normal delivery at term. The myelomeningocele was not diagnosed antenatally. She underwent closure of the lesion and insertion of a ventriculoperitoneal shunt at the same sitting at another centre. After that she had several revisions following a shunt infection, but in the intervening years she has had no further shunt-related problems requiring surgery.

She was managed by Orthopaedics for a hip dislocation (surgery at age 4) and scoliosis (no surgery), and has had regular urological follow-up but no neurosurgical review for many years. She attends a special school as she has mild cognitive impairment.

Current symptoms

Over the past 18 months, her parents have noted a progressive deterioration in her speech, difficulty with feeding and episodes of apnoea. She has also experienced a steady deterioration in her left hand function. She has no headache or visual dysfunction.

Clinical examination

On examination she is alert, follows commands but is unable to speak. She has marked scoliosis, with her right shoulder held much higher than the left, and her head tilted to the right and rotated to the left. She has marked nystagmus in all planes (lateral, upwards on upgaze and rotatory on downgaze) and an absent gag reflex.

She has an upper motor neuron weakness in both upper limbs, worse on the left (power 4/5 vs 4+/5). She has no active lower limb function.

She has a left-sided ventricuoperitoneal shunt in situ (unknown make), the shunt appears intact clinically, and she has no signs of raised intracranial pressure.


The most relevant pictures are included below. In summary, her imaging show that her ventricles are not dilated, she has marked herniation of her cerebellum through her foramen magnum, with compression of the lower brainstem and an expanded supracerebellar CSF space. There is a syrinx in the cervical region, ‘exophytic’ in parts. The cord is low-lying and there is a small distal lipoma attached to the terminal end.

Craniocervical Junction




So, how should this patient be managed optimally?

  1. What do you think is the reason for the progressive cerebellar herniation?
  2. What would you do first – and what next if that doesn’t work?

Do you think the shunt could be a problem? What about the tethered cord and expansion of the CSF space above the cerebellum? What about craniocervical surgery, how do you prevent progressive slump?

Comments from some colleagues (edited):

Charlie Teo says: Perfect case for discussion because there is probably no right way to treat this child. The capacious CSF space above the cerebellum would suggest slumping rather than herniation ie. this may not even be a Chiari but slumping of the cerebellum secondary to the craniocervical junction dysmorphology. There is significant basilar invagination and moderate platybasia resulting in severe ventral compression from the dens. If the family were against surgery and you needed to do the least invasive first, I would section the cord at the level of the MM and do a terminal ventriculostomy at the same time. This would be low risk and may just improve her enough for all to be happy. However, I suspect this would only give temporary relief and within a few weeks/months she would return with neurological deterioration. The next least invasive procedure would be to do an endonasal dentectomy with arch of C1. However, this would likely make her CCJ even more unstable and you would then be forced to fuse her from behind. If you were going to do this, it would be tempting to decompress her foramen magnum and do a tonsillectomy and wide opening of the fourth ventricular foramina at the same time…..I would avoid this like the plague as it is likely to increase the slumping and create more tethering around the fourth ventricular outlets. I do NOT believe there is a shunt problem but my old boss, Rick Boop, always said to explore the shunt as the initial treatment and many times he was right despite no obvious ventriculomegaly. GREAT CASE!!

John Kestle says: It sure sounds like her current problems are from the craniocervical junction region. On some of the slices provided, her ventricles look larger than the typical MMC child with shunted hydro. You say the suprasellar space is “expanded” – are you implying that this is progressive (ie do you have old films?). The first question before opening a can of worms is whether she needs treatment. Sounds like she does from your description. I doubt the shunt is the problem but in the absence of old films I would probably do a shunt revision first. If that doesn’t help then I’d likely consider the anterior compression of the cervicomedullary region. I don’t personally do craniocervical surgery but looks to me like she needs anterior decompression. I’d defer to the spine gurus to say whether she also needs posterior stabilization. Thanks for posting this.

Jack Walker says: The case is interesting and obviously complex, but in the famous words of Dave McLone “it’s always the shunt”. I would do a shunt revision as the first step. Even if the ventricles are stable and she is otherwise free of typical shunt malfunction symptoms shunt revision almost always improves the symptoms and the syrinx. Assuming that she is stable after shunt revision, I would repeat her MRI in 2-3 months. Operating on her posterior fossa without being sure of the shunt status is a setup for disaster. Following this protocol, I have not had to decompress a Chiari or shunt a syrinx in a myelomeningocele patient in the past 15 years. Thanks for the interesting case. You are welcome to use my comments so my colleagues can rant on me.

KC Wang says: First of all, I think we have to make sure whether the shunt is really working properly or not. Previously imaging and/or shunt tap may be helpful. Frequently shunt malfunction is the cause of deterioration of Chiari symptoms. During early childhood progressive Chiari may occur in cases with ventricular shunting. If there is shunt malfunction, it may aggravate morphological and symptomatic Chiari. I would like to recommend detailed shunt evaluation first and shunt revision if it malfunctions or is questionable. Then decompression of Chiari should be considered.

Shlomi Constantini says: I have seen several cases over the years with an increase in tonsilar decent with optimally draining supra-tentoreal shunts. Never to such an extent though. I always thought that shifting forces in a small calvarium may cause cerebellar problems and not only a supra-tentoreal slit-like clinical syndrome. So.. I think it is not a primarily fluid problem but some disproportion between posterior fossa volume and content in a VERY bizarre anatomy. The anterior angle/pressure of the CVJ may be a contributing factor. A dynamic study to RO instability is mandatory. If it is not stable, you may be forced to fuse in the optimal angle (maximal reduction) to be determined……..Since the clinical syndrome is bulbar, you are forced to address the pressure point at the foramen magnum. This seems to be a case where only bone and dura may not be sufficient to get a relief…. Some tissue resection may be required. If following such a foramen magnum decompression-like procedure, there will be no improvement, a fluid diversion procedure may help, e.g. syringo-sub-arachnoid shunt etc.

Rick Abbott says: All of my mentors always reminded me to start with the shunt so I definitely would look there first even though ventricles are not changed by your description. The ventricles in myelomeningocele patients can be stiff, especially with an infection history and it is quite possible that pressure from a partial shunt failure is being relieved by the herniation of the cerebellum so the ventricles are not enlarging. If that is not the problem then I would do a chiari decompression. I suspect that the lipoma is a red herring with regards the neurological deterioration. I would not forget it though as it needs to be followed as it may be, in fact, a dermoid associated with the closure. If it is, then there is growth potential.  Good luck. I hope you work it out so I can send you more.

Paul Steinbok says: She has a Chiari II – symptomatically getting worse related to the Chiari II and syrinx. No evidence that the anatomic descent of the cerebeallar tissue is progressive. Hydrocephalus supposedly controlled but no information about prior scans to show the ventricles. A significant supracerebllar CSF collection – again we do not know if this is longstanding. I would make sure shunt is really working. The ventricles are not collapsed and it may be that the child has a malfunctioning shunt. If shunt working – would proceed to decompress the Chiari II and see what happens. I would avoid surgery to drain syrinx directly. What about untethering the spinal cord? Something to consider – but I would not do this as my first option – I would decompress the Chiari first.

Dominic Thomson says: The clinical history of new and progressive bulbar change I think is a compelling argument that treatment is indicated. (A baseline sleep study might be useful as might further strengthen the evidence base for bulbar dysfunction and it can add some objectivity to the follow up). I would favour a stepwise approach:

1. I agree with the view that a shunt exploration would be my first step. Ventricular size alone is such a poor indicator of adequate shunt function in the MM group and surgery to the Chiari in the presence of actual or incipient shunt malfunction could precipitate a significant deterioration.

2. Having confirmed shunt function, if the symptoms remained I would treat the Chiari II. I would not be concerned about the degree of ventral compression so long as stability had been confirmed on flexion extension plain films. I would give serious thought to starting off by performing a bony decompression only. In this case I think that you would have little alternative but to remove the posterior arch of C2 as well as some foramen magnum and C1. A dura sparing procedure I think would carry less risk and just might be sufficient to alleviate the compression around the brainstem and afford some clinical relief.

3.  If the symptoms persist you can always go back to perform a definitive dural opening and duaraplasty.

Clearly the number of opinions reflect the unique nature of this case, but it has resulted in very interesting debate. I would definitely be interested to hear the outcome.