Tao Fan, XinGang Zhao, HaiJun Zhao, Cong Liang , YinQian Wang, QiFei Gai, Fangyi Zhang

Abstract

Summary of background data: It is well established that syringomyelia can cause neurological symptoms and deficit by accumulation of fluid within syrinx cavities that lead to internal compression within the spinal cord. When other intervention treating the underlying etiology failed to yield any improvement, the next option would be a procedure to divert the fluid from the syrinx cavity, such as syringo-subarachnoid, syringo-peritoneal or syringo-pleural shunting. The indications and long term efficacy of these direct shunting procedures are still questionable and controversial.

Objective: To investigate the clinical indication, outcome and complication of syringe-pleural shunt (SPS) as an alternative for treatment of syringomyelia. Study design: We reported a retrospective 26 cases of syringomyelia were found to have indication for a diversion procedure. SPS was offered. Patients’ symptoms, mJOA score, and MRI were collected to evaluate the change of the syringomyelia and prognosis of the patients. 2-tailed wilcoxon signed-rank test was used to perform the statistical analysis of the mJOA scores.

Methods: All 26 patients underwent SPS. The clinical information was collected, the mean follow-up time was 27.4 months, 2-tailed wilcoxon signed-rank test was used to perform the statistical analysis of the mJOA scores. The key surgical technique, outcome and complications of SPS were reported in detail.

Results: No mortality and severe complications occurred. Postoperative MRIs revealed near-complete resolution of syrinx in 14 patients, significant shrinkage of syrinx in 10 patients, no obvious reduction or unchanged in remaining 2 patient. Postoperatively, the symptoms improved in 24 cases (92.3%). Statistical analysis of the mJOA scores showed a statistical significance (P < 0.001) between the preoperative group and the 2-week postoperative group. No further significant improvement between 2 weeks to the final follow up at 27 months.

Conclusion: Collapse or remarkable shrinkage of the syrinx by SPS could ameliorate or at least stabilize the symptoms for the patient. We recommend small laminectomy and a less than 3mm myelotomy either at PML or DREZ. The SPS procedure can be an effective and relatively long-lived treatment for the idiopathic syringomyelia and those that failed other options.

Keywords: Syringomyelia，Cerebrospinal fluid homeostasis, Syringopleural shunting, CSF homeostasis, Myelotomy

1.Introduction

Syringomyelia, as a simple or complex phenomenon, can be either a primary or a secondary condition associated with other pathologies, such as Chiari malformation, basilar invagination posttraumatic, postoperative subarachnoid adhesion, spinal cord tethering, intramedullary tumors, or scoliosis. The etiology of syringomyelia is often multifactorial and thus should be evaluated carefully and managed by a variety of strategies. It is generally assumed that the accumulation of fluid within syrinx cavities can cause worsening neurologic symptoms or deficits by progressive internal compression of the neural elements within the spinal cord. Thus, when all other surgical options aimed at treating the underlying etiology are exhausted, the remaining surgical treatment should target at the reduction of the size of syrinx or the collapse of the cystic cavity to ameliorate or at least to halt the deterioration of the symptoms.

The advancement in understanding of the pathophysiological mechanism of syringomyelia has suggested that restoration of cerebrospinal fluid (CSF) flow dynamics is the ultimate goal for surgical intervention and this can be achieved by decompression, reconstruction of spinal subarachnoid space, or diversion of the intra-syrinx cerebrospinal fluid. Among the various effective surgical treatments, the posterior fossa decompression surgery for Chiari malformations,detethering the tethered spinal cord, or release of the subarachnoid adhesion of the spinal cord at the previous surgical site or posttraumatic vertebral level mainly aim at restoration of the CSF flow dynamics. However, once these surgical interventions failed to lead to any improvement, the direct fluid diversion of the syrinx are recommended to reduce the size of syrinx and prevent further accumulation of CSF inside the cord. The options　include　syringo-subarachnoid,syringo-peritoneal, or syringo-pleural shunting. Reports of these shunting procedures are few in the literature. Even though some authors have reported sustained improvement following direct CSF shunting from the syrinx cavity, unfortunately, the indications and efficacy of the different shunting procedures are still questionable and controversial.

During the past 4 years, we have treated about 160 cases of syringomyelia-related patients. After individualized analysis and evaluation of each patient by clinical physical examination and MRI, we have selected syringo-pleural shunting as a treatment of 26 cases of syringomyelia. The indications, technique points, clinical results, and patients’ prognosis are reported and discussed as follows.

2 Methods

2.1. Patients’ information

From 2008 to 2013, the author performed syringo-pleural shunting (SPS) in 26 of 160 patients with syringomyelia. Male 16 cases, female 10 cases, age from 18 to 61 years, mean: 41.5 years. All these patients underwent preoperative workup and were evaluated by physical examination and imaging study with MRI. When the previous operation aimed at treating the potential underlining etiology failed or any other possible canal/arachnoid space reconstruction surgery was not feasible, the SPS procedure was selected (as last resort) for the direct CSF diversion from the syrinx cavity.

This series included Chiari I syringomyelia (failure of previous posterior fossa decompression surgery) 7 cases, posttraumatic or postoperative adhesion 6 cases, syringomyelia was found during the scoliosis work up in 5 cases, post meningitis 5 cases, and idiopathic (no evidence of obvious causes) 3 cases. The patient’s main symptoms were in accordance with classic syringomyelia neurological patterns, such as pain or numbness, sensory loss, and weakness.of the different shunting procedures are still From 2008 to 2013, the author performed syringo-pleural shunting (SPS) in 26 of 160 patients with syringomyelia. Male 16 cases, female 10 cases, age from 18 to 61 years, mean: 41.5 years. All these patients underwent preoperative workup and were evaluated by physical examination and imaging study with MRI. When the previous operation aimed at treating the potential underlining etiology failed or any other possible canal/arachnoid space reconstruction surgery was not feasible, the SPS procedure was selected (as last resort) for the direct CSF diversion from the syrinx cavity.

Fig. 1. After duratomy, we normally detect the surface of the spinal cord under the surgical microscope, sometimes we could find that there is only an arachnoid membrane covered on the posterior midline (Blue Arrow), or on the DREZ, then the myelotomy was made at this thinnest zone to minimize the effect on spinal cord matters. Otherwise, we would like to choose the posterior midline for the myelo- tomy, which is 3 mm in length. (For interpretation of reference to color in this figure legend, the reader is referred to the web version of this article.)

This series included Chiari I syringomyelia (failure of previous posterior fossa decompression surgery) 7 cases, posttraumatic or postoperative adhesion 6 cases, syringomyelia was found during the scoliosis work up in 5 cases, post meningitis 5 cases, and idiopathic (no evidence of obvious causes) 3 cases. The patient’s main symptoms were in accordance with classic syringomyelia neurological patterns, such as pain or numbness, sensory loss, and weakness.

2.2. Syringo-pleural shunting technique

The syringo-pleural shunting device utilized is the 2-piece CSF lumbo peritoneal T-tube shunt unit (Medtronic PS Medical), 0.7-mm inner diameter, and 1.5-mm outer diameter. The proximal end (Syrinx side) is a T-tube with a length of 8 cm and 57 holes spiral configuration. The distal tubing has multiple perforations at the end, which is positioned in the pleural space (Fig. 1).

Patients are all placed in left lateral recumbent position, according to the preoperative MRI evaluation; an upper or mid-thoracic midline incision was made at the level of the shunting site of the obvious syrinx cavity. Routinely, for the syrinx involving the entire cord, a T1–T2 or T2–T3 level is chosen for the myelotomy and shunt tube entry site, while in cases of segmental syrinx, the segment with the largest syrinx cavity and thinnest wall was chosen.

In early three cases of this series, a single-level complete laminectomy was made at the shunting site. For the following 23 cases, a modified bilateral partial hemilaminectomy was performed and only portion of the spinous process and medial inferior lamina bilaterally was removed to minimize the structure loss of the posterior elements of spinal column. After opening the dura and exposing the dorsal surface of the spinal cord, we identified the area of the thinnest wall of the syrinx for myelotomy (Fig. 1) under the magnification of a surgical microscope. Myelotomy was performed at midline in 16 cases, while Dorsal root entry zone (DREZ) was chosen for myelotomy in the rest 10 cases. The two short ends of the T-tube were trimmed to 8–15 mm in length, and carefully inserted into the syrinx cavity via a 2–3 mm myelotomy entrance (Fig. 2). The dura was sutured to watertight degree. The catheter was connected to the distal tubing and secured to the spinous process through the fixture clip, which is included in the T-tube unit (Medtronic PS Medical). Then, the distal tubing was tunneled to a separate right-side sub-axillar paramedian incision, which was usually made at the eighth-ninth intercostal space and right above the superior margin of the ninth rib. The distal end of the catheter was then fed through a stab incision into the pleural space and the length of the catheter within pleural space was about 12 cm.

Fig. 2. After insertion of the catheter, the distal end of the Ttube (Arrow) was left out and fixed on the spinous, the dura was sutured water tight.

Illustration: The location of the drainage-tube in the thoracic cavity

2.3. Patient evaluation, follow–up, and statistical analysis

Patients’ clinical symptoms and signs were documented preoperative, 2 weeks after SPS, and at the final follow-up visit. The length of follow-up varies from 16 to 53 months and the mean final follow-up time is 27.3 months. The changes of preoperative symptoms were documented at these follow-up points. Magnetic resonance imaging (MRI) with T1, T2, or Cine protocol was also performed at the same time points to observe the changes in the size of syrinx cavity.

To maximize the accuracy and consistency, all

2.3. Patient evaluation, follow–up, and statistical analysiscavity

Fig. 3. A 57-year-old male patient who had an 8-month history of distending pain and numbness of his left limb. (A) Preoperative magnetic resonance imaging scan showed central canal of spinal cord expansion and whole syringomyelia, the mJOA scale score is 7. (B) MRI scan obtained 2 weeks after SPS surgery showing marked decrease in size of the syrinx cavity and the mJOA score is 9. (C) MRI obtained 23-month postoperatively demonstrating that syrinx of cervical spinal cord totally disappeared and the remarkable shrinkage of the thoracic syrinx. The postoperative mJOA score is 10.

patients’ neurological functions were assessed using the modified Japanese Orthopedic Association (mJOA) Scale.

Analysis on statistical significance was conducted employing a 2-tailed Wilcoxon signed-rank test (SPSS13.0) between preoperative and 2-week postoperative group, and 2-week postoperative and the last follow-up group. A statistical significance was recorded when the P value was less than 0.01.

3 Outcomes

Among 26 patients who underwent SPS procedures, there are idiopathic syringomyelia (3 cases), post-meningitis syringomyelia (5 cases), failed to improve (both clinical symptoms and size of syrinx) after posterior fossa decompression for Chiari malformation (7 cases), syringomyelia secondary to subarachnoid adhesion following trauma or previous surgery (6 cases), and syringomyelia identified during the work up for scoliosis (5 cases). The patient’s general clinical data and mJOA scores are summarized in Table 1. One patient underwent a second exploratory surgery 1 month after the initial SPS due to a sudden deterioration of neurological function with increased weakness of bilateral lower extremities. Infection was found around the epidural space of the durotomy entrance site. A surgical debridement and wash out were performed epidurally without reopening of the dura. The syringomyelia shrinkage was eventually demonstrated on MRI with the reduction of syrinx to less than 50% of that preoperatively. The declination of his neurological function was reversed after the washout, but still not able to reach his baseline prior to the SPS surgery. Another patient turned out to have a relevant history of pleurisy and, unfortunately, it was not disclosed by the patient (No. 5, Table 1) prior to the surgery. Postoperatively, the patient’s symptoms were not altered and little changes of the syrinx cavity were found on postoperative MRI and this was the only case in this series that has no change in the size of syringomeylia after SPS. In the other 24 cases (92.3%), there are significant shrinkages (more than 50% reduction of syrinx), or even near-complete collapse of the syrinx cavity as well as improved symptoms either during immediate postoperative hospital stay and/or at the final follow-up visit (Figs. 3 and 4).

MRI was performed 2 weeks postoperatively and at the final follow-up time points. The remarkable shrinkage of the syrinx cavity was found in 24 patients (92.3%) 2 weeks after SPS, unchanged in one patient with a history of pleurisy and little reduction of syrinx in one patient who underwent the second operation for the postoperative infection (7.7%). At the final follow-up visit, near-complete collapse of the syrinx cavity occurred in 14 cases, while significant shrinkage in other 10 patients, and little reduction and unchanged in the remaining 2 patients as mentioned above.

No mortality or CNS (intradural) infection occurred in this series. A majority of the patients in the series experienced transient minor chest pain or discomfort on the side of shunt insertion into the pleural space. This complaint invariably resolved 3–7 days after the surgery.

A small amount of pleural effusion could be found on the postoperative MRI in some patients, but the patients are asymptomatic and none of them required any treatment. New onset of truncal numbness was observed in 3 patients. No proximal catheter migrating out of the cavity or new onset of cord tethering at the shunt entry site was observed in our series. Preoperative mJOA score of the series is 10.7 ± 1.7 (Mean ± -SD). At 2 weeks after SPS surgery, mJOA score of the series increased to 12.8 ± 2.1 (Mean ± SD). The mJOA score of the series at the final follow-up was 12.6 ± 2.4 (Mean ± SD). 2-Tailed Wilcoxon signed-rank test (SPSS13.0) was performed between preoperative group and 2-week postoperative group (P < 0.001), 2-week postoperative group, and the latest follow-up group (P = 0.48). Therefore, there is a statistical significance between preoperative group and 2-week postoperative group, but no statistical significance exists between 2-week postoperative group and the final follow-up group.

Fig. 4. A 40-year-old female patient who underwent a suboccipital decompression on account of numbness and weakness of her both arms 10 years ago. After the first decompression, her symptoms showed little improvement and deterioration of the neurofunction before SPS. (A) Preoperative magnetic resonance imaging scan,reveal in cervicothoracic and thoracic syrinx with large cavities and the mJOA score is 11. (B1, B2) MRI scans obtained 2-weeks after SPS surgery, showing remarkable shrinkage of the cervical syrinx (B1) and almost disappearance of the thoracic syrinx (B2). The postoperative mJOA scale score is 14.

4 Discussion

Since the first description of syringomyelia, various modalities of surgical treatment have been advocated for the shrinkage of the syrinx cavity and thus internally decompressing the spinal cord, which could eventually lead to neurological improvement or at least symptom stabilization. With the improved understanding of the pathophysiological mechanism of syringomyelia, the surgical intervention mainly consists of two kinds of operations: (a) One is to restore normal CSF flow dynamics/pattern either at the craniocervical junction or at the site where the obstruction of arachnoid space occurs, such as posterior fossa decompression, transoral odontoid resection, or dural reconstruction/adhesion lysis at the obstruction site 1, 3–4, 6. (b) The other kind of operation is to establish a CSF drainage system from the syrinx cavity, which is also called shunt procedure 5. Careful scrutiny of each individual patient to determine the true etiology of the syringomyelia is the key for selection of the appropriate surgical procedure. It is important to recognize that a procedure to treat the underlying etiology, e.g. structural anomalies, of the syrinx offers the patient the best opportunity to gain a long-term improvement. Most recent studies have reported on the importance of improving CSF flow dynamics regardless of the treatment strategy employed.

When all options aimed at the restoration of physiological CSF flow dynamics are not feasible, or such a surgical intervention failed, the shunting procedure usually is the only last resort left. Many authors have reported a neurological improvement following a shunting surgery that diverts CSF from the syrinx cavity.

The shunting surgery options for syringomyelia include syringo-subarachnoid, syringo-peritoneal, and syringo-pleural shunting. There is one study in which the effectiveness of syringo-pleural and syringo-subarachnoid shunting are directly compared and revealed that shunt system draining the syrinx to the pleural cavity appears to be associated with better outcomes. The substantial difference in pressure radients accruing in the above three shunting systems has a bearing on both the CSF drainage rate and thus the maintenance of a shunt system patency. Normally, intrapleural pressure is subatmospheric. Depending on the respiratory cycle, pleural pressure fluctuates from −5 cm H2O to −8 cm H2O. The negative pressure of the thoracic pleural cavity appears to be ideal for draining of syrinx fluid. Although the syringo-subarachnoid and syringo-peritoneal shunts remain to be other viable options, the subarachnoid space surrounding the cord or nerve roots has a hydraulic pressure that is almost comparable with that of intra-syrinx cavity, whereas the peritoneal cavity demonstrates a positive pressure with respect to the syrinx most of the time and further, the presence of the omentum increases the distal catheter obstruction rate. Based on the above reasons, the pleural cavity has become the most favored place to host the distal catheter of a syringomyelia shunt system.

Isik et al. treated 32 cases of Chiari malformation syringomyelia by SPS. In our study, 7 patients with Chiari malformation patients experienced a failed initial posterior fossa decompression surgery. Fearing that the manipulation around the medulla oblongata area during the second redo surgery through the scar tissue might lead to some fatal complications, we opted SPS as a safer approach. For other syringomyelia cases, individual analysis of the patient to figure out the true etiology causing the syringomyelia plays quite an important role. Ghobrial et al. use arachnolysis for treatment of post-infectious and post-traumatic syringomyelia. In our group, we treated 6 cases of syringomyelia secondary to subarachnoid adhesion (including post-operative and post-traumatic) by SPS. After careful assessment of these particular patients for subarachnoid lysis, it was believed that arachnolysis can potentially be more dangerous to cause severe complications. Under such a situation, we choose SPS as a second-line treatment to compromise the deterioration of the patients’ symptoms related to syringomyelia. In 5 post-meningitis cases and 3 idiopathic cases, it was speculated that a diffused adhesion or arachnoiditis formed along the entire neural axis and spinal canal may contribute to the progressive expansion of the syrinx. Therefore, the only option for the surgical approaches is to drain and shrink the syrinx cavity directly, and thus the SPS was offered. SPS was straightforwardly performed to shrink the syrinx cavity in 5 cases of syringomyelia, which was found during the scoliosis workup.

Complications such as shunt obstruction/dislocation, spinal cord tethering, truncal numbness, or pain have been reported after syringo-pleural shunting surgery in the literature. Through the length of our follow-up period, we have not seen any severe complications such as shunt failure, further spinal cord injury, or other fatal complications in this series. Only one young patient experienced a worsening of neurological function 4 weeks after SPS because of an epidural infection around the shunting site. The most common complaint of the patients was right chest pain and discomfort, which lasted only 3–7 days after SPS. Other complications include truncal numbness in 2 patients. A small amount of pleural effusion could be found on the postoperative MRI in some patients and they are asymptomatic and thus none of them required any treatment.

Prestor et al. provided a strong argument in favor of DREZ myelotomy as the choice of the approach for syringomyelia shunting surgeries by electrophysiologic study. Our experience suggests that since only 3mm incision is needed for the T-tube insertion, the myelotomy site should be determined based on where the thinnest dorsal syrinx wall locates. Sometimes, this thinnest portion of the syrinx wall is semi-transparent, just like a layer of arachnoid mem-brane and it should not contain any meaningful neural elements. Therefore, the myelotomy should be made at this spot irrespective of whether it is at PML or DREZ. However, if this thin wall cannot be identified after opening of the dura, one can make the myelotomy at PML since it is very easy to be accurately located under the magnification of a surgical microscope. The PML myelotomy should be of little structural damage of the neural elements inside the cord theoretically. The key is to keep the myelotomy small in order to minimize the damage to vital spinal cord structures and to reduce the subsequent complications, such as truncal numbness or pain, which could even prevent over-drainage of the overall CSF from intrathecal space to pleural cavity. Further, we did not observe any spinal cord tethering at the shunt entry site. This possibly benefits from such a small myelotomy of 3mm and small bony window with partial hemilaminectomy in most of the cases. The development and use of less-invasive procedures may further reduce the occurrence of such complications related to syringomyelia shunting procedures.

Although there are a number of series reports with different syringomyelia shunting procedures, such as syringo-subarachnoid, syringo-peritoneal, and syringo-pleural shunt in the literature, the relative small number of the patients and/or mixed cohorts of the patients in each of these papers make it difficult to draw con-clusions regarding which system is more logical, more effective, and with less complications. Since Phillips et al. first described the syringo-peritoneal shunt for syringomyelia in 1981, Lund-Johansen et al. and Kunert et al. reported this procedure with a series with a small number of patients. The presence of the omentum within the peritoneal cavity may increase the incidence of the shunt distal catheter obstruction. The nature that a physiological positive pressure presents during most of the time with respect to that of the syrinx may hinder the rate of shunt drainage. Intestine and abdominal wall motions may lead to relative high risk of dislodging of the peritoneal catheter. All these have been reported as the shortcoming for the syringo-peritoneal shunting procedure. Vernet et al. reported two groups of patients, comparing the outcome of syringo-subarachnoid and syringo-pleural shunting, finding a favorable outcome in the later group. The pressure within subarachnoid spaces is believed to be similar to that of the syrinx intra-cavity compartment and further, the likelihood of developing arachnoid adhesion surrounding the distal opening of the shunt may account for the short lives of the syringo-subarachnoid shunt, and thus the long-term efficiency of this shunting system was suspected. Recently, more and more authors have discovered a more favorable result of syringo-pleural shunting echnique for treatment of syringomyelia. In our series, he SPS could lead to a substantial shrinkage, or even near-complete resolution of the syrinx, and thus lead to a progressively improved neurological function, even occur-ring immediately after the procedure. However, there is no further improvement between 2 weeks to the final follow-up visit ( 27 months). Although we were not able to set another follow-up time point in between for all our patients, based on our observation on some of the individuals, no further improvement was detected between 2 weeks postoperatively to the final follow-up visit. Therefore, this may suggest a natural course of the syringomyelia treated with syringo-pleural shunt and our data indicates that the recovery should reach the plateau 2 weeks after the operation. Although no relapse of syrinx was found during an average 27-month follow-up, a longer follow-up may reveal some potential shunt system failure. Nevertheless, in contrary to the common belief that many syrinx shunt systems are short lived, our data proves that a well-done syringo-pleural shunt can survive at least 2 years without any need for revision.

5.Conclusions

Collapse or remarkable shrinkage of the syrinx by SPS could ameliorate or at least stabilize the symptoms for the patient. We recommend small laminectomy and a less than 3mm myelotomy either at PML or DREZ. SPS is a safe and rather straightforward technique for those syringomyelias that are unresponsive to causal treatments. SPS has good results both clinically and radiologically and the result is to be expected within the first few weeks and may be long lasting for the idiopathic syringomyelia and those that failed other options.

Conflict of interest None declared.

We have no relationships with pharmaceutical companies, biomedical device manufacturers, or any other corporation whose products or services may be related to the subject matter of this article or who have sponsored the study. We certify that there is no actual or potential conflict of interest in relation to this article.

Financial support

Funding for this study was provided by Construction Project of National Clinical Key Specialties of P.R.China [Ministry of Health of P.R. China: 873(2011)] and The capital Health Research and Development of Special: 2014-2-8011. And the corresponding author Tao Fan received the support of those funding.

Declaration

We certify that we have participated sufficiently in the work to take public responsibility for the appropriateness of the experimental design and method, and the collection, analysis, and interpretation of the data. We hereby certify that this paper consists of original, unpublished work which is not under consideration for publication.