Minimally Invasive Approach for the Treatment of Intradural Spinal Pathology

Ravi H. Gandhi, M.D; John W. Germ an, M.D.


Neurosurg Focus. 2013;35(2):e5 

In This Article


Since Foley and Smith[3] popularized the microendoscopic technique, there has been a wealth of literature addressing the utility and potential benefit of that technique. Experience with tubular retractor systems and MIS technique has allowed surgeons to expand the application of these techniques to treat various spinal pathologies. The technique has been well described for the purposes of decompression, stabilization, and various extradural pathologies. This series further adds to the literature showing that MIS techniques can prove useful for all intradural pathology, intra- and extramedullary pathology as well as spinal dysraphism, syringomyelia, and treatment of Chiari malformation.

A traditional approach to intradural pathology of the spine involves a midline incision, bilateral muscle dissection, and removal of the posterior spinal elements. A rate of instability from this approach requiring a second operation has been reported as high as 60%.[5] In a series by Seppälä et al. looking at schwannoma resection in Finland, there was a 10% postoperative complication rate and a 6% delayed spinal instability rate.[9] There have been improvements in open technique using a hemilaminotomy approach. This approach has been associated with fewer complications, better outcomes, and shorter hospital stays.[2,7,10,13,14] In our series, only 1 patient underwent fusion at the time of surgery (Case 3) because the entire facet was removed to reach the tumor. One patient required anterior cervical discectomy and fusion for degenerative disc disease, but this was to treat a preexisting condition. Although dynamic imaging was not customarily used in our series, no other patients required subsequent surgery for the treatment of instability or symptomatic kyphosis.

In an attempt to further advance minimally invasive approaches to the spine, tubular retractors with muscles-plitting techniques have been applied to the treatment of intradural pathology. This minimizes muscle destruction and affords an adequate corridor to tumor resection. There is also a decrease in the dead space at the end of surgery, theoretically limiting the possibility of postoperative CSF leakage. By minimizing tissue destruction and unnecessary bone removal, the risk of postoperative instability is decreased. Haji et al.[4] have described the use of a minimally invasive technique to remove both intra- and extradural neoplasms including 2 intramedullary lesions without added neurological risk. Treway et al.[11] presented a series of 3 patients in whom they used a minimally invasive approach for the treatment of tethered cord syndrome, with reduced blood loss and length of hospital stay.

Our series further demonstrates that a wide variety of intradural pathologies can be treated with an MIS approach. Of the 14 oncological lesions that were treated, 12 were completely resected. The 2 tumors that were not completely resected would have been incomplete using any surgical approach. None of the tumors have recurred. Of the 4 patients who underwent suboccipital craniectomy and dural augmentation for Chiari I malformation, all had an improvement in their headaches and gait instability. Although postoperative imaging showed a 75% rate of pseudomeningocele, higher than the 5% reported in the literature, there was no need for reoperation in our series. Of note in these cases is that the dura was loosely reconstructed using a dural substitute (Duragen, Integra Life Sciences) and a few tack-up sutures. Three patients underwent surgical fenestration of arachnoid cysts. One patient had to return to the operating room 18 months later for a repeat fenestration at another level. Although the rate of reoperation for arachnoid cyst was higher than expected, the need for reoperation was not a result of limitations of the exposure, but rather due to an underestimation as to the extent of the arachnoid cyst septations on preoperative imaging. The 3 patients who underwent spinal cord untethering did not have any complications from surgery: 2 patients remained stable from a chief complaint standpoint and 1 patient had a significant improvement in his incontinence 6 weeks after surgery. One patient had a significant decrease in the size of the syrinx after an MIS approach to syringomyelia-arachnoid shunt placement with significant improvement in his myelopathic symptoms. One patient had a punctate midline myelotomy and cordotomy for the treatment of chronic visceral pain. Although there was no change in her functional status due to her cancer, she was able to leave the hospital with a significant reduction in her requirement for continuous pain medications.

Although the average length of hospital stay of 3 days in this series is comparable to that of series of open resection,[2,10,13] the heterogeneity and characteristics of our population may have artificially lengthened that average. Twenty-five patients were ambulating within 24 hours of surgery and 48% of patients were discharged to home by postoperative Day 2. In our anecdotal experience and in other series,[10] patients undergoing an open excision are often kept on bed rest for 24–72 hours, thus increasing the risk of pressure ulcers and deep vein thrombosis. Although our average blood loss of 197 ml (range 15–900 ml) is higher than other series,[4,6] only 1 patient in our series required transfusion. This may represent our learning curve.

We have demonstrated that this technique can be applied to a wide array of intradural pathologies; however there are added complexities that must be addressed. To deal with pathology at the midline or the spinal cord, a pedicle-to-pedicle exposure is preferred. To facilitate obtaining this exposure, we add resection of the ventral portion of the spinous process to maintain a medial trajectory. It remains important to carefully review the preoperative images to appropriately plan the levels that need to be approached and a trajectory that best suits the pathology. Although a smaller postoperative dead space may help reduce CSF leakage, we believe that a watertight dural closure is often preferable. One exception is following treatment of a Chiari malformation where our preference has been to keep the dura loosely approximated to optimize decompression. Dural closure can be difficult with a smaller working space. Park et al.[8] have described dural closure with a U-clip (Medtronic). In our experience, the 4.0 Vicryl stitch on a U-shaped needle has been adequate. Meticulous muscle and fascial closure are also important in preventing CSF leak.

Although we've shown that there is a wide application of an MIS approach, there are limitations to its use. Lesions extending more than 2 spinal segments may not be best treated via an MIS approach, as the expandable tubular retractor is often limited. Changing the angle of the retractor is feasible; however, doing so may increase the risk of surgery. Pathology with a large ventral component may not be best addressed by an MIS approach. This series demonstrates our experience with the application of MIS approaches to intradural pathology, but there is a significant learning curve when adopting such an approach. The senior author (J.W.G.) has had a robust experience with minimally invasive surgery for other spinal pathology, and the transition to applying these techniques for intradural pathology was gradual. There does not seem to be a significant increase in operative time, but this is anecdotal, and considering our heterogeneous group it is difficult to compare. There was also no improvement in our operative time with added experience (the cases in Table 2 are organized in chronological order). After obtaining experience with MIS approaches, the combination of those techniques with traditional microsurgical techniques is natural. It is our recommendation that surgeons should have adequate experience with MIS approaches prior to attempting resection of intradural pathology with this method.