Acute onset of postoperative syringohydromyelia.

Syringohydromyelia is a frequent finding in cases of tethered cord syndrome. The classical teaching is that the development and progression of a syrinx is a chronic process. We present a case report of an acute onset syringomyelia in an infant, who underwent an excision of a lumbosacral transitional lipoma and detethering of the cord. Immediately after recovery, the infant was found to have flaccid paraplegia. An emergency magnetic resonance imaging revealed a large acute onset syringomyelia for which he underwent an emergency midline myelotomy and release of fluid from the syrinx. Though the eventual recovery was good, this made us re-visit our understanding of the concept of syringohydromyelia. The case details and a plausible hypothesis for the rapid development of the syrinx are presented.

Introduction

Syringomyelia with tethered cords have been seen pre-operatively or chronically with myelomalacia years after surgery. What prompted us to write this article is the sheer rapidity with which the syrinx developed post-surgery.

Case Report

An 8-month-old male child, with no medical problems save a swelling in the lower back was referred to the neurosurgical department for evaluation and further management. Clinical examination revealed the absence of any neurological deficits and a development normal for the infant's age. There was no history of any difficulty in passing urine or stools. He had a midline lumbosacral cutaneous swelling which was consistent with a lipoma. Radiological evaluation by magnetic resonance imaging (MRI) [Figure 1] revealed a conus lying at L5 and a lumbosacral transitional lipoma. There was a minimal dilation of the central canal, but there was no evidence of any syrinx. The rest of the neuraxis, including the craniovertebral (CV) junction, was normal. The child underwent an L4–L5 laminotomy, total excision of the lipoma and untethering of the cord. After the lipoma was excised, the plaque was re-neurulated with 8-0 prolene pial sutures. A primary watertight dural closure was done with 5-0 vicryl after ensuring hemostasis. The wound was closed meticulously in layers. Blood loss was very minimal negating the need for transfusion. During reversal from anesthesia, the priapic response, which is seen normally in male children, was absent. Postextubation, the baby was found to have flaccid paraplegia. The neurological examination revealed a sensory level and motor level at L1. There were no reflexes elicitable. An emergency MRI was done with the requisite sequences to rule out bleeding at the operated site [Figure 2].

Figure 1

(a-c) Preoperative magnetic resonance imaging (note the absence of a syrinx and a normal craniovertebral junction. The central canal is dilated but this is commonly found in many cases)

Figure 2

(a and b) Immediate postoperative magnetic resonance imaging: Syringohydromyelia from D12-L5 causing ballooning of the cord (note the central canal regains the preoperative size above D12)

The MRI revealed the presence of a large syringohydromyelia in continuum with the central canal extending from L1 to L5 correlating with the clinical findings. There was no evidence of any hematoma or infarct. The syrinx appeared to be an acute enlargement of the existing minimal dilation of the central canal. There was a very clear distinction between the central syrinx and the cord in all MRI sequences, which ruled out cord edema. As the imaging findings correlated with the clinical examination, it was decided re-explore in keeping with the departmental policy to intervene in all cases of symptomatic syrinx. The patient underwent an emergency D12-L3 laminotomy and midline myelotomy at three levels, where the spinal cord was found to be ballooning out. The presence of a large new onset syrinx was thus confirmed. Mildly blood stained cerebrospinal fluid (CSF) was drained. Postsurgery, the child was stable but continued to be in spinal shock for 12 h, following which he slowly improved in his sensations. By the time of discharge, a week later, he had a good appreciation for nociceptive stimuli, had no urinary or fecal incontinence and had a flicker of limb movements. At follow-up 6 weeks later, the child was normal, and his motor milestone achievement was normal and is continuing normally even at last follow-up. It must be however, mentioned that no intra-operative neurophysiological monitoring was used in this case. Follow-up imaging showed the disappearance of the syrinx [Figure 3]. The cord sac ratio was less than 30%.

Figure 3

(a and b) Follow-up magnetic resonance imaging: Showing resolution of syrinx 1-month postoperatively

Discussion

Syringomyelia or “tubular cavitations” of the spinal cord has been a poorly understood pathological entity. Brain[1] remarked that it was “relentlessly progressive.” Simon,[2] first defined “hydromyelia” as spinal intramedullary cavity lined by ependyma way back in 1857. He distinguished it from syringomyelia, which is a fluid-filled cavity lined by glial tissue. The distinction between the two is an academic exercise at best. The amalgamated term syringohydromyelia is used in neurosurgical parlance commonly. We have used the terms interchangeably since there is no consensus on the correct terminology.

Numerous attempts to classify syringomyelia have existed in literature the most frequently quoted being that of Milhorat et al.[3] who classified syringomyelia as communicating (anatomically in continuum with the fourth ventricle), noncommunicating (which is separated from the fourth ventricle by a segment of normal spinal cord) and atrophic syrinx (associated with myelomalacia). Recently, Batzdorf[45] proposed a simpler classification whereby syringomyelia was classified into those which had problems at the CV junction abnormalities and those due to problems lower down in the spine. Oi et al.[6] have recently proposed a new clinical category of hydrocephalus wherein central canal dilation is classified as a type IV hydromyelic hydrocephalus - with an isolated central canal dilation.

Numerous theories exist to explain the etiology of syringomyelia, none of them are perfect. There is still no grand unified theory for this. The earlier theories were obsessed with the fact that it occurred so frequently with the Chiari malformation, and so stressed upon various embryological mechanisms. These include the theories of Chiari (persistent embryonal state), Gardner's hydrodynamic theory and William's modified hydrodynamic theory (venous pressure changes being responsible for syringomyelia formation). These theories need an anomaly at the CV junction to make them workable.

There are theories, however, which seem more suited to explain the problems in our case. Oldfield et al.[7] proposed that a block at the foramen magnum resulted in an engorgement of cervical veins in the subarachnoid space thereby decreasing space for pooling of CSF during normal systole. What gathers significance is that he went on to propose, that the pulsatile pressure waves force the CSF into the cord through the perivascular and interstitial spaces and these are responsible for the origin and propagation of the syrinx. Thus, for the first time, a theory was proposed wherein syrinx formation was not due to pressure from within the cord but from without, which forced fluid into the central canal. It is also different in that; the syrinx formation is by the systolic pressure wave and not by increased venous pressure from valsalva manoeuver. Stoodley et al.[8] with horseradish peroxidase tracer studies in sheep proved that reducing arterial pulsations decrease the fluid in the central canal. He went on to demonstrate the unidirectional flow of CSF, from the perivascular spaces across the interstitial spaces and into the central canal driven by arterial pulsation. This explains the formation of a non-communicating syrinx. Stoodley et al.[9] proposed two distinct mechanisms for normal migration of fluid through perivascular spaces: (i) Systolic expansion of arteries in the perivascular space may force fluid through the basement membrane while in diastole, fluid may enter the perivascular space from, the subarachnoid space. (ii) Pulsation in the subarachnoid space transmitted through perivascular spaces acts an impetus for flow. Milhorat et al.[10] proposed that CSF was continually produced by the ependyma of the central canal and expansion occurred in those segments isolated by occlusion or stenosis at each end. Ball and Dayan[11] were of the view that an occlusion at the CV junction forces subarachnoid fluid to dissect along the Virchow Robin space and this intramedullary fluid would slowly enter the spinal cord. Aboulker[12] proposed a similar theory with the CSF entering along the dorsal nerve roots.

The spinal-spinal pressure dissociation model[13] proposes that subarachnoid scarring after trauma, in a fashion similar to hindbrain descent, impedes rapid pressure equilibrium in the subarachnoid space proximal and distal to the scar, which causes CSF to move into the low- pressure environment of the central canal of the cord. The exact cause of this is unknown, but what is agreed upon is that the subarachnoid space scarring and obstruction prevents instantaneous pressure equilibrium during coughing or straining. Dayan and Aboulker proposed that the fluid is driven into the cord parenchyma by the arterial pulsations, coalesce and then finally rupture into the central canal.

We have thus far, reviewed the theories of syringomyelia relevant to our case. On reviewing the aforementioned theories of the syrinx, we cannot propose a definite theory to explain what caused the formation of the acute syrinx in our child. The pial suturing might have trapped the central canal rather than allow a free communication between the canal of the cord, and the subarachnoid space, or it might have been that the dural closure was tight thereby decreasing the subarachnoid space. The drainage of the CSF might have set up a state of “dysequilibration” in the local fluid dynamics forcing CSF into the central canal. The presence of blood admixed with CSF would have further lessened the potential space available. We can only hypothesize. Furthermore, the absence of intraoperative spinal monitoring worked to our disadvantage. However, intraoperative electrophysiologic studies would have indicated pathology, but would not have shown us the formation of the syrinx. It must be remembered that the absence of the priapic response during reversal showed that the pathology had already occurred. To quote Sir Arthur Conan Doyle “Once you eliminate the impossible, whatever remains, no matter how improbable must be the truth.” In our case, unfortunately; the impossible, the improbable and the truth seem unclear no matter how hard we try to use logical reasoning. Nevertheless, we proffer an explanation as follows: The child had a dilated central canal preoperatively which may have already been acting like a low-pressure sump as per Bernoulli's theorem. Postlipoma resection, the cord was reconstituted, and the dura was closed by primary closure without a graft. This together with the blood in the CSF might have decreased the available subarachnoid space for pressure equilibrium during recovery. Applying the theories of Milhorat and Stoodley we conclude that during recovery from anesthesia, systolic pressure might have forced the CSF in the subarachnoid space into the central canal which due to the absence of CSF (lost during surgery) might have served as a low pressure sink resulting in the formation and propagation of a syrinx. Also, the handling of the cord would have rendered it edematous and facilitated the dissection of the spinal matter by the extramedullary fluid on their way to the central canal.

A literature search revealed a very similar case described by Post et al.,[14] wherein they report a progressive paraplegia 48 h after surgery for a lipomyelomeningocele. Re-exploration revealed no hematoma. MRI revealed a dorsal syrinx. They however, managed their patient conservatively. Their patient too did well. What differed our case from theirs was the rapidity of syrinx onset in our case and the fact that the syrinx was within the confines of the operative field. In their case, the syrinx was above the level of the operative field. The child in their case was managed conservatively for the syrinx while in our case the infant underwent an emergency myelotomy. The final neurological outcomes in both cases were very good and normal. The reduction in the cord sac ratio[15] from more than 50% in the immediate postoperative MRI prior to the surgical intervention for the syrinx to <30% in the follow-up MRI at 6 weeks together with the normalization of the patient's neurological status goes to prove that we were perhaps wise in immediately investigating[16] and intervening surgically for the syrinx.

Both the cases are indeed red herrings, and so a guideline for management cannot be offered. Acute transient worsening of neurological deficits following surgery for the tethered cord syndrome has been seen in our department; we have not however seen immediate complete paraplegia previously, and this prompted us to image the child. We have also not seen long term deteriorations after successful untethering of cords as described by Cochrane et al.[17] Though the various theories can explain the location and formation of the syrinx, there is nothing even remotely available to explain the rapidity of syrinx formation in our case.

Conclusion

This case illustrates the fact that acute onset of syringomyelia though rare can occur. It also throws light upon the fact that despite a century and three-quarters having elapsed since the description of the syrinx, there are still several mysteries regarding it that need to be unlocked. Sir Issac Newton once remarked; “For hypothesis should be employed only in explaining the properties of things but not assumed in determining them.” Our attempt in reporting this case and trying to deduce a reason for its occurrence is, but an effort to shed some light upon the mysteries of syringohydromyelia. We would like to reiterate two important lessons from this experience of ours namely, the need for intraoperative spinal monitoring to enhance patient safety, and the need to ensure a lax dural sleeve reconstruction with water tight closure after surgery for the tethered cord syndrome.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.