MIS: Luiz Pimenta, Luis Marchi, and Leonardo Oliveira
Open: Christopher M. Bono

Open surgery for the treatment of ASDis should be considered the default technique for addressing virtually any type of pathology that can be encountered. This is particularly true for patients with adjacent level stenosis with or without some form of instability. The principles of revision decompression are well known to most spine surgeons.

While minimally invasive surgical techniques such as LLIF purport advantages of avoiding previous areas of surgical scar, thoughtful posterior surgical maneuvers can be performed to work within a previously operated field. Of most importance is appreciating that the anatomical plane between the dural sac and operative scar is tightly adherent. In our practice, this plane is left intact when possible in order to avoid dural injury and cerebrospinal fluid leak.

In cases of adjacent level stenosis, as is presented in this chapter’s case, usually there has not been decompression previously performed at the new (adjacent) area of stenosis. Thus, any previously decompressed areas (provided that decompression remains adequate) do not require extensive epidural dissection. Instead, these areas are left “buried” within the scarified field, while the areas of new stenosis are carefully dissected free.

This consideration highlights a distinction in posterior revision decompression surgery that can be helpful. Cases of adjacent level stenosis are usually well addressed by a so-called fake revision laminectomy. In other words, despite the presence of operative scar, the interlaminar space with adjacent level stenosis has not been previously operated. Thus, once the laminae have been exposed (albeit through scar), the interlaminar window should be relatively virginal, with little reason to expect that the ligamentum flavum will be scared to the underlying dural sac. Illustratively, one can imagine the very typical case of adjacent level stenosis that occurs at L3–L4 years following an L4–L5 laminectomy and fusion. Provided that the previous decompression maintained the superior aspect of the L4 and did not enter the L3–L4 interspace, the surgeons should expect a “fake” revision at the L3–L4 level.

This can be distinguished from “true” revision laminectomy. Had the patient above undergone a full L4 laminectomy (from the bottom of L3 to the top of L5) with fusion only of L4–L5 and subsequently developed recurrent stenosis at L3–L4 secondary to ASDeg, the surgeon must dissect within the scarred dural sac from the bony borders of the previous decompression prior to revision laminectomy or facetectomy.

The primary advantages of posterior open surgery are versatility and direct decompression of the neural elements. As discussed above, regardless of the extent and location of the epidural scar, the bony borders can be defined and revision decompression can be performed. Moreover, decompression does not rely on any type of realignment. Whereas LLIF increases disc space height, which then is believed to indirectly decompress the spinal canal by tensioning redundant ligamentum flavum, posterior open revision laminectomy (whether “fake” or “true”) does not require anatomical realignment. While proponents of LLIF extol its ability to reduce low-grade spondylolistheses via insertion of a tall interbody device, one should keep in mind that this can also be performed through a posterior approach (e.g., transforaminal lumbar interbody fusion [TLIF]). What is contestable, however, is the clinical benefit of reducing such mild deformities, particularly when the overall global balance of the spine is satisfactory. It is in fact our routine practice to not perform any substantial deformity correction unless indicated based on global sagittal alignment measurements. In further demonstration of open surgery’s superiority, one can consider the common case of isolated foraminal stenosis in an area of previous decompression that is not associated with any significant deformity or misalignment. This can arise from hypertrophy of an unfused facet joint. In this case, it would be difficult to imagine how performing an LLIF procedure would affect any neural decompression. It is, however, easy to envision how an open revision foraminotomy and extension of the fusion can be an effective procedure.

The above arguments have only considered the advantages of open surgery for decompression. Similar arguments can be made for open surgery being the better approach to address issues involving the fusion. By nature of the diagnosis of ASDis, a previous fusion has been performed adjacent to a level of new pathology. In most cases encountered, fusion has been stabilized with pedicle screws. With the fear of sounding too obvious, what has been inserted through a posterior approach (i.e., pedicle screws) can be revised through an open posterior approach. While there might be a few, technically gifted and adventurous surgeons who would venture to extend a previous posterior pedicle screw construct through a minimally invasive approach, this is by no means something that is widely accepted. Justifiably, this is also not what has been proposed by Dr. Pimenta and his colleague in this chapter. It appears they would be satisfied with the stabilizing effects of the interbody cage alone, or perhaps with a lateral plate. It would only be in the rare cases in which no previous posterior instrumentation had been placed that one could expect the average spine surgeon to insert percutaneous pedicle screws at the adjacent level for stabilization.

In the operating room, the first step is the placement of surface electrodes from an electromyography system that monitors the lumbar plexus during the psoas traverse, which is mandatory in this procedure. Four muscle groups per side are monitored as they represent spinal nerve distributions from L2–S2: vastus medialis, anterior tibialis, biceps femoris, and medial gastrocnemius. Also, a reference electrode is placed upper to the lateral thigh, and a return electrode is placed superior to the operative site, such as on the latissimus dorsi muscle. Proper skin preparation ensures good electrical conductivity. The patient is transferred onto a radiopaque bendable surgical table in a direct lateral decubitus position (90°), perpendicular to the table, with the greater trochanter directly positioned over the table break and with legs and knees slightly bent. The four adhesive strips to attach the patient are: (1) torso, (2) iliac crest, (3) leg and knee, and (4) knee and foot. This configuration increases the space between iliac crest and ribs, especially relevant when accessing thoracolumbar junction or L4–L5 level.

The ideal positioning is confirmed by fluoroscopy, ensuring that when at 0° the C-arm provides a true anteroposterior (AP) image and when at 90° a true lateral image. It is substantial that the lateral fluoroscopic images show both vertebral plateaus and superior pedicles aligned, presented as a single line, and that the AP image reveals the spinous processes in a middle position, and pedicles as circumferences.

Over the skin, the iliac crest, the transition between the last rib and the posterior abdominal wall, and the quadratus lumborum muscles must be marked. After skin asepsis, the central position of the targeted disc can be identified using two Kirschner wires and lateral fluoroscopic images, making a marking that covers the center of the affected disc space. Afterward, a longitudinal skin incision is made, over the intersection between the post-erolateral muscles of the abdominal wall (abdominal internal oblique, abdominal external oblique, and transverse abdominis). A first fascia incision is made to permit the surgeon to introduce the index finger into the retroperitoneal space and gently create a pathway and releases all attachments of the peritoneum, providing a safe lateral entry. Once the retroperitoneal space is identified, a second fascia incision is made below the first skin mark to introduce the initial dilator. The index finger will safely guide all dilators up to the psoas muscle, protecting abdominal structures.

The first dilator is placed upon the posterior third of the L2–L3 disc, as confirmed by AP and lateral fluoroscopy. Then, the fibers are gently separated by the initial blunt dilator, with concomitant EMG monitoring for assessing the closeness to the lumbar plexus. The dilator must be rotated in position to determine proximity and spatial distribution of nerves. The dilators in sequence are placed over the previous, always checking the EMG, until the final placement of the closed retractor. The working portal is connected to a suspension arm in order to prevent unwanted movement. After confirming the ideal position by fluoroscopy, the blades can be selectively adjusted to the desired diameter. A bifurcated optical fiber cable is attached to the retractor for optimal direct visualization of the exposure. Moreover, the retractor opening must be minimal, with the duration of muscle spreading as short as possible, since the lumbar plexus must be compressed during psoas traverse.