11 Minimally Invasive Resection of Intradural Extramedullary Lesions within the Thoracic Spine

Abstract

Keywords: extramedullary, intradural, laminectomy, meningioma, minimally invasive, schwannoma, spinal dural arteriovenous fistula, thoracic

You cannot depend on your eyes when your imagination is out of focus.

Mark Twain

11.1 Introduction

It dawned on me that I was looking at the spine through a distinctive minimally invasive lens when a colleague of mine asked me to review magnetic resonance imaging (MRI) of a thoracic spine. The study demonstrated a plum-sized enhancing intradural extramedullary lesion severely compressing the spinal cord in the lower thoracic spine. By this time in my career, I had converted my practice almost entirely to minimally invasive approaches. I felt comfortable with the instruments and confident with the exposures that I was able to achieve using the various minimally invasive access options. As I looked at the MRI, I began to envision how I would access the lesion with a paramedian minimal access approach. I was familiar enough with the anatomy that I could visualize an exposure with the rostrocaudal dimension of 28 to 30 mm. I knew that angling the blades of the access port would provide me with an additional 5 to 7 mm of rostrocaudal exposure. I could imagine exposing the midline of the canal by a combination of mediolateral blades and rotation of the bed. Undercutting the spinous process and contralateral lamina would provide me access to the contralateral recess. All these techniques are similar to those used in accomplishing those same objectives when working on a medial decompression for a minimally invasive transforaminal lumbar interbody fusion or a lumbar laminectomy. The only difference was that I was in a different region of the spine, and there was more work to do once the dura was exposed. I recognized that while the canal dimensions and interpedicular distances are unique in the thoracic spine, the experience with the management of extradural metastatic disease in the thoracic spine that I described in Chapter 10 provided me the familiarity with the thoracic landscape to feel confident with accomplishing the necessary exposures.

I snapped out of my trance when my colleague asked me for a hand. When he mentioned a midline approach, I plainly remember asking, “Why not a minimally invasive one?” Up until that time, I had never performed a resection of an intradural extramedullary lesion through a minimally invasive approach. But I could not find a compelling reason why we should not.

Yet another learning curve presented itself with this approach. However, the learning curve was distinct from the initial learning curve of minimally invasive spine surgery. Instead of developing new skills working with bayoneted instruments in the access port and becoming accustomed to new perspectives for maintaining orientation, this learning curve built upon the skill set that I had already refined by the microdiscectomies, laminectomies, cervical foraminotomies and instrumented lumbar fusions that I had performed over the years. I soon discovered the nuances of positioning the access port, the exposure of the spinal cord and the closure of the dura. As these cases began to accumulate and my experience continued to grow, I recognized that the operation reaches a point, specifically when the dura is open and the lesion exposed, where it becomes indistinguishable from an open midline operation. In certain circumstances, the operation actually becomes easier. A focused and highly efficient exposure fosters that facility. In the end, whether the exposure is through a midline approach or a minimally invasive one, you will employ the identical microsurgical techniques to remove the lesion from the spinal cord.

Time and again in this book, I referenced the concept of the efficiency of the exposure in minimally invasive approaches, a relationship that I have referred to as the Caspar ratio. In its most elemental form, the Caspar ratio translates into the percentage of the surgical target relative to the surgical exposure. I hope that I have been able to convince the reader at this point in this Primer that minimally invasive exposure is highly efficient, utilizing a higher proportion of the surgical exposure relative to the surgical target than its open counterpart.

The resection of an intradural extramedullary lesion is the quintessential example of the Caspar ratio. The skill of precise placement of the minimal access port directly over the lesion and the proficiency in securing every millimeter of exposure afforded by the minimal access port are of even greater importance in the management of intradural extramedullary lesions than in any other procedure that I have described earlier. It is essential to recognize that acquiring these skills does not require the management of dozens of intradural extramedullary lesions. Instead, those proficiencies develop with every microdiscectomy, laminectomy and instrumented lumbar fusion that you perform minimally invasively. Management of thoracic metastatic lesions further refines that skill set. These are the cases where your mind will effortlessly connect the lines of what cannot be seen to the limited exposure that you can see. Synthesis of these images inside of your mind, built on the experience of managing the characteristic pathology of the lumbar and cervical spine, is what you will harness to perform minimally invasive resections of intradural extramedullary lesions in the thoracic, lumbar or even cervical spine. The management of intradural extramedullary lesions through minimally invasive techniques is the zenith of a highly developed skill set. It is only after mastery of the various other minimally invasive procedures in the lumbar, thoracic and cervical spine presented in the earlier chapters that you should endeavor on this approach.

As will be discussed in the coming pages, master surgeons of yesteryear described the resection of intradural extramedullary lesions with preservation of the midline structures decades ago.1,2,3 You need not jump immediately to a paramedian transmuscular approach through a minimal access port. There is tremendous value in performing resections of intradural extramedullary lesions with a midline incision but with preservation of the midline structures, as described by Yaşargil³ and Seeger.¹ Understanding the amount of access to the entire canal through a hemilaminectomy exposure with the preservation of the spinous process is a component of the stepwise progression to a purely minimally invasive approach. It is important to recognize that these lesions have all of the characteristics that lend themselves exceptionally well to a minimal access resection. The very nature of a spherical lesion growing within a canal gives it laterality. The lesion always displaces the spinal cord to one side as it occupies the majority of the other side of the canal. The dimensions of the thoracic canal impose an upper limit of 25 mm in any dimension. The combination of laterality and limited dimensions plays to the strength of the minimally invasive approach.

It has been several years since that initial resection of a thoracic meningioma through an expandable minimal access port. The experience that I have gained through the management of the subsequent 26 intradural extramedullary lesions that I have encountered has led me to manage practically all of them using minimally invasive techniques. I have reached a point where my mind has a difficult time conceiving another manner to perform these operations. As mentioned time and again in this Primer, the mind once enlightened cannot again become dark.

In this chapter, I present the anatomical basis and then describe the technique for the resection of intradural extramedullary pathology using a minimally invasive approach. I highlight some of the subtle differences with the use of the minimal access port, along with the difference in the approach and bone work. At the end of the chapter, I provide case illustrations where intradural extramedullary pathology, including meningiomas and dural arteriovenous (AV) fistulas, are managed with minimally invasive techniques. However, it would be difficult not to begin the chapter with the remarkable history of a paramedian technique that would eventually evolve into the minimally invasive technique described in this chapter.

11.2 Historical Perspective

Concern for the stability of the spinal column after resection of spinal tumors has been on the mind of spine surgeons for decades. It was the unease caused by the potential risk of progressive kyphosis and scoliosis in the years and decades of life that followed the resection that prompted a few pioneering surgeons to ask themselves whether a complete laminectomy was, in fact, necessary to resect intradural lesions. Similar to the surgeons who did not view the spinous process as an obstruction for the management of lumbar stenosis, Seeger and colleagues¹ and Yaşargil and colleagues³ viewed the lesion within the dura accessible without disruption of the posterior tension band or even sacrifice of the spinous process. The postoperative thoracic kyphosis they saw in their early patients drove these surgeons to explore surgeries that preserved the midline structures. The logical question was whether the same operation could be accomplished with less disruption of the native spine. The rational answer was a unilateral hemilaminectomy approach with preservation of the midline elements.^1,2,3 But could such a limited exposure provide the surgeon with the requisite anatomy needed for complete resection of an intradural extramedullary lesion?

Despite this text supposedly being a modern-day primer on minimally invasive spinal surgery, the techniques that I describe in the following pages were routinely used by surgeons over 35 years ago. Although the retractors used to access the spine may be different, the techniques are surprisingly the same. In fact, the figure in the wonderful manuscript by Yaşargil and colleagues³ on unilateral hemilaminectomy for resection of spinal tumors could easily replace some of the figures created for this chapter ( ▶ Fig. 11.1).³ The same could be said for the illustrations in the manuscript by Eggert and colleagues ( ▶ Fig. 11.2).²

Fig. 11.1 A figure from Yaşargil et al¹ on the use of unilateral hemilaminectomy for the resection of intradural extramedullary lesions. The technique described by Yaşargil³ is indistinguishable from the techniques described decades later by modern-day minimally invasive spine surgeons. Access to the spine is the only difference. (a) Axial illustrations of the thoracic spine demonstrating the bone work described by Yaşargil.³ Note the use of the drill to undercut the spinous process in the middle axial image in order to provide access to the contralateral recess. (b) Posterior view of the thoracic spine with a 15-degree rotation where multiple hemilaminectomies have been performed and the dura has been opened and tacked up with sutures. Note the preservation of the spinous processes at all levels. (Reproduced with permission from Yaşargil MG, Tranmer BI, Adamson TE, et al. Unilateral partial hemilaminectomy for the removal of extra- and intramedullary tumours and AVMs. Adv Tech Stand Neurosurg. 1991; 18:113–132.)

Fig. 11.2 Illustrations from the manuscript by Eggert and colleagues 2 demonstrate the use of hemilaminotomy for the management of intradural extramedullary lesions. (a) The trajectories shown in the figure (axial view of the spine) are indistinguishable from those used in a minimally invasive approach today. (b) This drawing shows a posterior view of the spine with a laminotomy used for resection of the lesion. The article was published in 1983, but the bone work demonstrated is as relevant today for minimally invasive approaches as it was over 30 years ago.² (Reproduced with permission from Eggert HR, Scheremet R, Seeger W, et al. Unilateral microsurgical approaches to extramedullary spinal tumours. Operative technique and results. Acta Neurochir (Wien). 1983; 67:245–253.)

Dr. Wolfgang Seeger¹ also championed this technique for intradural pathology, reserving full laminectomies for intradural intramedullary lesions. Again, with the exception of the retractors used then and the minimal access ports used now, there is essentially no difference between the approach described in this chapter and Seeger’s approach three decades ago ( ▶ Fig. 11.2).

The common theme among all of these surgeons is consistent with the common theme that resonates through this book: preservation of the midline elements. Poletti⁴ and Lin⁵ saw no reason to remove the spinous process and disrupt the posterior tension band for a lumbar laminectomy; likewise, Seeger¹ and Yaşargil³ did not view the spinous process as an obstruction to the central canal for resection of a tumor. It is this philosophy of accessing pathology within the spinal canal with minimal disruption to the native spine itself that culminated in Dr. Richard Fessler and colleagues’s publication in 2006, where they reported their experience with the resection of intradural extramedullary lesions using modern-day minimally invasive techniques.⁶ Today, the minimally invasive resection of intradural extramedullary lesions has become routine in several centers around the country due in large part to the efforts of these pioneering surgeons.

11.3 Rationale for a Minimally Invasive Approach: An Observation

The first step for establishing the rationale for a minimally invasive approach for the resection of an intradural extramedullary lesion is a thoughtful analysis of the dimensions of the thoracic canal. Based on the anthropometric measurements of the thoracic spine by Panjabi and colleagues, the transverse dimension of the thoracic canal (the distance from pedicle to pedicle) is seldom more than 22 mm in the mid-to-upper thoracic spine (T1–T9) and seldom more than 24 mm in the lower thoracic spine (T10–12). The anteroposterior (AP) dimension tends to be more constant throughout the entire thoracic spine and ranges between 16 and 18 mm.7 Unlike a metastatic lesion, which has a less predictable and more destructive growth pattern, an intradural extramedullary lesion has a more subtle presentation and more predictable growth pattern. There is an inherent limit that the lesion reaches within the thoracic canal before that patient becomes symptomatic. Therefore, the boundaries of the thoracic canal place an intrinsic limit on the dimensions of a thoracic intradural extramedullary lesion. Applying Panjabi’s measurements, the limit of these lesions should be no greater than 20 mm in the AP and lateral dimension ( ▶ Fig. 11.3).⁷ I tested that hypothesis with a dimensional analysis of 26 consecutive intradural extramedullary lesions.

Fig. 11.3 Cross-sectional anatomy of the thoracic spine. Anteroposterior and lateral dimensions of the central canal in millimeters from T1 to T12. As patients tend to become symptomatic when an intradural extramedullary lesion reaches the boundaries of the thoracic canal, the dimensions of the canal define the dimensions of the lesion. Abbreviations: SCD, spinal canal depth; SCW, spinal canal width.

Benign intradural extramedullary neoplasms that grow within the thoracic canal are indolent lesions with slow doubling times. The slow growth rate allows for a remarkable amount of accommodation by the spinal cord to take place. It is always astonishing to see these patients in clinic who present with only subtle neurologic findings when a lesion occupies their entire central canal. Early in my experience, I would peer at the axial T1-weighted gadolinium-enhanced MRI and look aghast at the thoracic spinal cord completely flattened into a ribbon by a tumor within the canal. Meanwhile, the patient would be standing alongside me, looking at the same image, asking me what I thought. It is these times that I can only but marvel at the resilience of the central nervous system ( ▶ Fig. 11.4).⁷

Fig. 11.4 Intradural extramedullary lesion at T3. This patient presented with a subtle gait disturbance and decreased proprioception but was otherwise intact. She was walking upward of 3 miles a day at the time of presentation. A neurologist discovered the lesion when evaluating the patient for an ill-defined neuropathy. (a) Sagittal T2-weighted magnetic resonance imaging (MRI) demonstrating an intradural extramedullary lesion at T3, remarkably without signal change abnormality on the spinal cord. (b) Sagittal T1-weighted MRI with gadolinium showing uniform enhancement of the lesion. (c) Axial T1-weighted MRI with gadolinium showing the lesion occupying almost the entire canal. The spinal cord can be seen flattened against the lateral aspect of the canal. Consistent with the observation that lesions become symptomatic when they approach the boundaries of the dimensions of the canal, this lesion measured 10 mm in the transverse dimension, 16 mm in the anteroposterior dimension and 20 mm in the rostrocaudal dimension. Those are consistent with the range of dimensions at T3 reported by Panjabi et al.⁷ More importantly, those dimensions fall within the range of what may be safely resected through minimal access ports.

When patients become symptomatic, they have reached the tipping point of what their spinal cords can tolerate. At that point, neurologic symptoms begin to arise. Still, measurement of these lesions tends to be no more than 15 to 20 mm in any dimension, and the lesions typically have a remarkably spherical shape. A dimensional analysis of the 26 intradural extramedullary lesions identified the mean dimensions of 18.6 mm (range, 10–25 mm) in the rostrocaudal dimension, 13.0 mm (range, 7–18 mm) in the lateral dimension and 13.6 mm (range, 9–17 mm) in the AP dimension. Three additional observations were made from that series. First, all lesions displaced the spinal cord to one side, which plays to the strength of a paramedian approach. Second, the fact that the rostrocaudal dimension is the only dimension not bound by the bony structures of the canal made it always the largest dimension of the lesion. However, that dimension still never exceeded 25 mm, which falls within the range accessible to a minimally invasive approach. Finally, no lesion remodeled the lamina, foramen or pedicles of the central canal for expansion beyond the expected reported dimensions of the canal.⁸ The dimensional analysis of these 26 lesions validated the prediction made earlier regarding the inherent dimensional limitation.

The first component in establishing a favorable Caspar ratio is defining the surgical target. The average dimensions from these 26 cases provide insight into the precise dimensions of any thoracic intradural extramedullary lesion. A near-spherical lesion in the thoracic canal with a maximum diameter of 20 mm that displaces the spinal cord to one side has all of the aspects that play to the strengths of a minimally invasive approach: focal and lateral. Defining the surgical exposure is the second component of the Caspar ratio that establishes the rationale for a minimally invasive unilateral approach to thoracic intradural extramedullary lesions.

When the average dimensions of a thoracic lesion, the average dimensions of the thoracic canal and the diameter of the minimal access port are superimposed one over another, the anatomical basis for a minimally invasive technique becomes self-evident.▶ Fig. 11.5 brings together all of these measurements into one image, which firmly establishes the anatomical basis for such an approach.

Fig. 11.5 Anatomical basis for a minimally invasive approach to intradural extramedullary lesion. Superimposing an expandable minimal access port over the intradural extramedullary lesion with the average measurements from the dimensional analysis of 26 lesions. (a) Illustration of an intradural extramedullary lesion with the following average dimensions: 18.6 mm in the rostrocaudal dimension, 13.0 mm in the lateral dimension and 13.6 mm in the anteroposterior dimension. As can be seen in this illustration, the lesion displaces the spinal cord to the right and occupies the left side of the canal. The laterality of the lesion plays to the strength of a minimally invasive approach. A 22-mm diameter minimal access port readily encompasses the entire lesion. (b) With a minimum expansion of the access port, 35-mm rostrocaudal exposure may be achieved, which is more than adequate to visualize the rostral and caudal poles of the lesion for resection.

11.4 Intradural Extramedullary Lesions of the Lumbar versus Thoracic Spine

There is a limitation to the minimally invasive approach with lesions in the lumbar spine. That limitation is due in part to the cauda equina being more accommodating than the spinal cord. The absence of the spinal cord provides lumbar lesions with the potential to grow larger in the rostrocaudal dimension before they become symptomatic. I have observed lesions in excess of 30 mm in the rostrocaudal dimensions in the lumbar spine only. Based on my experience, I have assigned a rostrocaudal limit for the exposure of 35 mm before the benefit of a minimally invasive approach I feel has been exhausted. Under these circumstances, a minimally invasive approach may become more of a liability than an asset when attempting to visualize the entire lesion at once.

Whether I manage a lesion through a minimally invasive paramedian approach or a traditional midline approach, it is imperative for me to visualize the rostral and caudal poles of the lesion simultaneously. Such visualization becomes a tall order once a lesion begins to exceed 30 mm and untenable after 35 mm ( ▶ Fig. 11.6). Although certainly possible, a limited exposure would result in the need for piecemeal resection, which is not a viable option with lesions such as myxopapillary ependymomas. If I encounter a lesion that I am not confident that I can visualize in its entirety through an expandable minimal access port, I forgo the paramedian transmuscular approach but still use a hemilaminectomy and spinous process-sparing approach as described by Yaşargil et al.³ Having said that, I have found this situation to be a rare exception and limited to the lumbar spine. The growth pattern in the thoracic spine tends to be more predictable, and the rostrocaudal dimension is unlikely to exceed 25 mm.

Fig. 11.6 Lumbar ependymoma. (a) Axial T1-weighted magnetic resonance imaging (MRI) with gadolinium enhancement demonstrating an intradural extramedullary lesion with an anteroposterior dimension of 11 mm and a transverse dimension of 14 mm consistent with the measurements in the thoracic spine. (b) Sagittal T1-weighted MRI with gadolinium enhancement demonstrating a rostrocaudal dimension of 31 mm. This lesion is approaching the upper limit of what may be safely resected through a minimally invasive approach; however, a paramedian, hemilaminectomy, spinous process sparing approach is still feasible.

11.5 Anatomical Considerations: Variations in the Lumbar and Thoracic Spine Topography

The majority of our experience in minimally invasive approaches occurs in the lumbar spine for the simple reason that degeneration of the spine that requires surgical intervention occurs predominantly in the lumbar region more so than the thoracic region. As a result, we achieve a great familiarity and sophisticated understanding with the subtleties of the lumbar spine anatomy. After a period of time, the subtle progression of the facets from their sagittal to coronal orientation, as we descend from the upper lumbar levels toward the sacrum, becomes intuitive.

Conversely, most of the intradural extramedullary lesions that we will manage in our careers occur in the thoracic spine. In the series by Seegar et al,1 152 of 256 intradural extramedullary lesions occurred in the thoracic spine. In the series by Yaşargil et al,³ 141 of 250 occurred in the thoracic spine. Such ratios have been consistent with my experience with the management of these lesions, with 26 out of 35 intradural extramedullary lesions harbored in the thoracic spine. For this reason, the anatomy of the thoracic spine from a minimally invasive perspective is the focus of this chapter.

When turning your minimally invasive skill set to the thoracic spine, it is worthwhile to highlight the unique topography of the thoracic spine that distinguishes it from that of the lumbar spine. There are elements of the anatomy in this region that can be harnessed into an advantage in minimally invasive approaches. The transverse process of the thoracic spine is one of them. We begin our discussion of the anatomy with that bony prominence.

11.5.1 Thoracic Transverse Processes and Laminae

The lumbar spine transverse process projects more in the lateral projection than in the posterior projection ( ▶ Fig. 11.7). As a result, lumbar transverse processes have a small role in minimally invasive exposures and approaches, with the medial aspect of the lumbar transverse process limited to serving as a reference point for the pedicle screw insertion point. Instead, the lumbar facet joint is the main minimally invasive target in the lumbar spine. However, in the thoracic spine, the roles of these anatomical structures are reversed. The thoracic facet joint is not prominent enough to be palpable with the tip of a dilator or index finger in the thoracic spine. A review of the axial illustration in▶ Fig. 11.7 demonstrates that it is recessed to the level of the lamina. The posteriorly projecting thoracic transverse process, however, is palpable and prominent. Therefore, the thoracic transverse process takes the place of the lumbar facet as the target for the initial dilator in minimally invasive approaches into the thoracic canal. This valuable prominence serves as a beacon for the location of the pedicle and the canal while assisting in docking the access port.

Fig. 11.7 The differences between the thoracic and lumbar facet joints and transverse processes. (a) In the lumbar spine, the transverse process serves as one of the three landmarks to identify the entry point for the pedicle, but it is not a landmark for docking the minimal access port. As the lumbar facet is at a shallower depth (blue plane) than the transverse process (magenta plane), it is readily palpable and an ideal target for the initial dilator. The angle of the projection as determined from the central canal to the tip of the transverse process is more obtuse than the angle in the thoracic spine (angle created by the green line). (b) In the thoracic spine, the more acute angle of projection of the transverse process (angle formed by the green line) makes it more prominent and in a higher plane (magenta plane). Therefore, in the thoracic spine, the transverse process is an ideal landmark for a docking target. The thoracic facet, on the other hand, is in a deeper plane (blue plane). The posterior projection and higher plane of projection are the unique attributes that make the thoracic transverse process an attractive prominence to target for initially dilating and securing a minimal access port.

The transverse process in the thoracic spine projects more posteriorly than laterally. That orientation makes it a tremendously valuable target for minimally invasive approaches. After the incision and opening of the fascia, an index finger can immediately palpate the transverse process and use it to guide placement of the initial and subsequent dilator for the eventual placement of the minimal access port. Docking on the transverse process in the thoracic spine may be analogous to docking on the facet in the lumbar spine. It is a reliable and safe starting point for dilatation.

Another key difference between the lumbar and thoracic spine is the lamina. The thoracic lamina tends to be steeper than the lumbar lamina as it joins the spinous process. That steepness makes the corridor to work through narrower than the lumbar spine laminar corridor ( ▶ Fig. 11.8). At the same time, the steepness of the thoracic lamina allows a better trajectory to readily undercut the spinous process and expose the midline of the thecal sac. The steeper lamina is an indication of the narrower canal in the thoracic spine and therefore a more constrained working area for intradural work. The canal may be enlarged to compensate for this constrained working channel by drilling down the medial aspect of the pedicle, a technique that I emphasize in the section on operative technique (see Section 11.9, Operative Technique).

Fig. 11.8 Anatomy of the (a) lumbar and (b) thoracic spinous process and lamina. Compared to the lumbar spinous process and lamina working channel, the steeper thoracic spinous process and narrower lamina offer a more modest working corridor into the canal. However, that same steepness allows ready access to the entire canal by facilitating undercutting the lamina.

11.6 Patient Positioning and Localization

Localization in the thoracic spine is a challenging endeavor under any circumstance; however, the stakes are significantly heightened when attempting an intradural lesion resection through a 30-mm incision. Being off by one level with localization changes the entire dynamics of the operation. Unlike an open procedure where the incision need only be extended in one direction or another and the relevant anatomy exposed, in a minimally invasive surgery case, the entire access port has to be removed, another incision planned or extended, sequential dilatation performed, the access port docked onto the appropriate level and another laminectomy performed. Midline open approaches in the thoracic spine are simply more forgiving. Although one always strives for correct identification of the thoracic level in any case, in minimally invasive approaches, there is no room for error. The segment has to be confirmed by counting upward from the sacrum for lower thoracic lesions or downward from the C7–T1 interspace for upper thoracic lesions. Distinct from destructive metastatic lesions in the thoracic spine, which can alter the vertebral body anatomy in a manner that allows for confirmation of the affected level on a fluoroscopic image or lends itself to identifying the altered bony anatomy on computer-assisted navigation, the anatomy of the vertebral bodies in these particular cases is not altered in the least bit. There is no identifiable characteristic appearance on either fluoroscopy or intraoperative computed tomography (CT).

In the operating room, only radiolucent tables are used to minimize any obstruction to seeing the bony anatomy. I avoid the use of standard operating tables and Wilson frames. The upper components of a Wilson frame have large circular gears, which can block visualization of the pedicles in an AP view. The base of a standard operating table prevents the free passage of the fluoroscope. My preference, as with all minimally invasive cases, is a Jackson table (or equivalent) with standard chest, hip and thigh pads. All of these components are radiolucent.

I confirm the localization of the level in two views, both AP and lateral. All patients have standard radiographs of the lumbar spine and confirmation of five non–rib-bearing vertebrae in the lumbar spine on the AP image. These radiographs also serve as a valuable reference in the operating room when looking at the fluoroscopic images.

For mid and lower thoracic cases, I localize twice counting upward from the sacrum. Prior to draping the patient for the actual surgery, I perform a preliminary localization with a series of spinal needles from the lumbosacral spine into the thoracic spine. For the first phase of level confirmation, I prep the entire lumbar and thoracic spine, not for surgery but for localization. I put on a pair of sterile gloves and approximate (by palpation) and mark the L4–5 segment. I employ five spinal needles, alternating an 18-gauge with a 20-gauge needle. Alternating the needles facilitates keeping track of them on AP and lateral imaging. I place the first spinal needle in the vicinity of the L4–5 facet. Depending on the level of the lesion, I pass two or three additional spinal needles in approximately 5-inch increments, docking them on the lateral aspect of the bony anatomy away from the canal. Lateral fluoroscopic images are then taken, and the levels of the spinal needles confirmed. The ideal position of the spinal needle is to have the tip of the needle pointing directly toward a pedicle, which is a position that minimizes any confusion on AP and lateral images. I strive to have the second spinal needle pointing directly toward the T12 pedicle and adjust it to ensure that I am pointing directly at the first rib-bearing vertebral body. Once I have adjusted the position of all the spinal needles to point directly toward the pedicle, I confirm the position with AP and lateral fluoroscopic images until I am satisfied that I am at the correct level. I mark the level and plan the incision 25 mm off the midline on the side of the lesion (more on that later). Then, I mark the entry points of all the spinal needles. Those marks are essential for the second phase of localization, where I repeat the entire process with the patient prepped and draped before I make an incision ( ▶ Fig. 11.9).

Fig. 11.9 Localization for resection of a T9 intradural extramedullary lesion. (a) Intraoperative photograph with spinal needles confirming T9 and T12 pedicles. The patient has been prepped but not draped for the localization process. The T12 pedicle was confirmed on both lateral and anteroposterior imaging. (b) Corresponding fluoroscopic image of the photograph in a. Note the spinal needle pointing directly toward the pedicle. The lesion is immediately medial to the pedicle of T9 on the right. The patient is prepped and draped again, and the localization process repeated.

For the second phase of localization, I prep and drape the proposed incision along with the entire lumbar and thoracic spine once again. I use the previously marked entry points for the spinal needles to once again confirm the level of the incision in the AP and lateral planes. An incision is made only after I have verified the correct thoracic level with absolute certainty.

11.7 Planning the Incision: An Observation on the Location of the Lesion

Localization of the level is only the first component of planning the incision. The precision of planning an incision immediately over the lesion is the second component. Doing so involves an AP fluoroscopic image along with a thoughtful analysis of the exact location of the lesion within the thoracic segment. When only 35 mm of the spinal cord is exposed, the lesion must be precisely in the center of that exposure to optimize access to the lesion and closure of the dura. Review of the 26 intradural extramedullary lesions identifies a consistent position of the lesion relative to the disc space and the pedicle. The lesions in all of these cases, from T2–3 through T11–12, are consistently alongside the thoracic pedicle ( ▶ Fig. 11.10). The unfailing location, relative to the pedicle, reveals a great deal about the origin of the lesion and how it grows.▶ Fig. 11.11 illustrates how a lesion that arises off a thoracic nerve root at the level of the thoracic pedicle grows into the exact location relative to the pedicle seen in all of the lesions in▶ Fig. 11.10. Assuming a circumferential growth pattern, the majority of the lesion grows at the level of the pedicle. From there, the lesion either grows in a rostral direction toward the disc space or in the caudal direction toward the inferior aspect of the pedicle. The circumferential growth pattern extends into the canal and displaces the spinal cord to the opposite side ( ▶ Fig. 11.11). As seen in▶ Fig. 11.10 and▶ Fig. 11.11, the majority of the lesion resides immediately medial to the pedicle in every case. There is a pattern of the rostral component of the lesion reaching the disc space; however, on occasion, the lesion may grow in the caudal direction and occupy the level of the vertebral body. Therefore, the incision is planned relative to the center of the pedicle that corresponds to the center of the lesion ( ▶ Fig. 11.12). Bone work extends either rostrally or caudally from the pedicle, depending on the direction of the growth seen on the MRI. Understanding the origin and growth pattern also serves as the basis for the laminotomy work described below.