Revision Laminectomy: Indications and Techniques

Transcription

1 CHAPTER 43 Peter D. Angevine Revision Laminectomy: Indications and Techniques INTRODUCTION The progressive nature of degenerative diseases of the spine makes it inevitable that, even with optimal patient selection and surgical execution, a proportion of patients who have previously undergone a decompressive lumbar laminectomy for herniated disc or lumbar stenosis will, at some point, require revision surgery. The reported likelihood of reoperation after lumbar spine surgery may vary depending on the index procedure (discectomy or decompressive laminectomy), the size of the series, and the length of follow-up. Reoperation rates of 5% to 10% are commonly reported. 4,5,10 Studies with long-term follow-up (4 to 11 years) have reported that up to 28% of patients may undergo reoperation after a lumbar decompressive procedure for lumbar spinal stenosis. 6,8,9 A European population based study of reoperation following lumbar discectomy found a 9-year cumulative risk of reoperation of 18.9%. 7 A reoperation may be necessary because of persistent or recurrent symptomatic stenosis at the previously treated vertebral levels or because of a recurrent disc herniation ipsilateral or contralateral to the previous surgery. The reoperation may consist solely of further de, or it may include a fusion procedure with or without instrumentation. In some cases, such as a retained disc fragment, an early ( 30 days) herniation of additional disc material, or a symptomatic epidural hematoma, the presentation occurs acutely after the initial surgery, and the decision to proceed with reoperation is straightforward. Other patients who have previously undergone a laminectomy with or without an arthrodesis may develop symptoms clearly due to degenerative pathology adjacent to the previously operated level (adjacent segment degeneration). These circumstances are discussed elsewhere in this textbook and will not be covered further in this chapter. Instead, the following discussion focuses on the diagnostic and therapeutic problem of delayed symptoms after previous decompressive spinal surgery. Reoperations pose challenges in planning and execution that are distinct from those presented by primary surgery. Adhering to fundamental principles of patient evaluation and selection, surgical technique, and close follow-up evaluation will help to ensure optimal outcomes for patients undergoing revision laminectomy (Fig. 43.1). EVALUATION AND INDICATIONS GENERAL CONSIDERATIONS When evaluating a patient who might be a candidate for a revision laminectomy, it is important to take a careful history regarding the current symptoms and the symptoms that were present prior to the previous surgery. The frequency, timing, quality, and intensity of the patient s complaints are explored along with any exacerbating or ameliorating factors. It should not be assumed that the patient s symptoms are of spinal origin even if they are similar to those that led to the index procedure. All potential etiologies for the patient s symptoms should be considered and excluded prior to concluding that the underlying cause is spinal and that revision surgery is potentially indicated. The patient s response to the previous surgery or surgeries should be assessed to the extent this is possible. The patient is closely interviewed about the degree, adequacy, and duration of any improvement that occurred after the previous procedure(s). This is particularly important if the current symptoms are similar to those present before the prior operation. In this case, recurrent or residual stenosis at the previously operated level(s) is suspected and some judgment may be made about the likelihood of response to reoperation based on the outcome from the first procedure. A thorough neurological and directed physical examination is performed to determine the presence of any objective neurological findings. The patient s overall and regional spinal alignment is assessed. If there are any relevant findings such as weakness, reflex asymmetry, sensory abnormalities, or straight leg raise, the patient is carefully and specifically asked if these had been noted on previous examinations. Incisional scars are noted and correlated with prior operations. A focused musculoskeletal examination of the lower extremity is performed, with particular attention to the hip and knee examination. Peripheral pulses are assessed and cutaneous signs of peripheral vascular disease, arterial or venous, are noted. The patient s imaging studies are thoroughly reviewed. When possible, all imaging studies that have been performed, not just the most recent ones, should be reviewed. It is not uncommon for a patient to have had multiple imaging studies over a long period of time and to have undergone more than 402

2 Recurrent/ persistent Neurological deficit No neurological deficit Reoperate Nonoperative management Improve Reoperate Recurrent leg pain Early Late MRI No neurological deficit New neurological deficit Neuropathic pain No evident Consider other etiologies Vascular claudication Diskitis/ osteomyelitis Improvement Nonoperative management No improvement Arachnoiditis MRI Consider pain management/ spinal stimulation No definite No red flags Red flags present Redicular Urgent evaluation and treatment Radiographs (AP, lateral, supine and upright, flexion and extension) EMG/NCS No instability Instability Pain management Vascular surgery evaluation Medical/ surgical treatment No definite Revision laminectomy Revision laminectomy with fusion Myelo /postmyelo CT Nonspecific/ chronic radiculopathy Consider other sources/ pain management Figure A schematic of the treatment decision-making process for radicular or claudicative symptoms in a patient with prior decompressive surgery. 403

3 404 Section V The Thoracic and Lumbar Degenerative Spine TABLE 43.1 Diagnostic Studies for the Revision Laminectomy Candidate Patient Study Strengths Limitations Radiograph Best study for overall alignment; Shows only osseous structures flexion/extension views help evaluate stability MRI Excellent soft tissue/neural detail; Potentially significant metal artifact addition of gadolinium can help distinguish recurrent herniated disc from postoperative scar CT Excellent bony detail; less metal Limited soft tissue detail artifact than MRI Post-myelography CT Excellent resolution of central canal Limited ability to define foraminal and lateral recesses; less artifact from narrowing other than that due metal than MRI to bony hypertrophy EMG/NCV May help to differentiate radicular Operator dependent; may be from peripheral pathology; can help difficult to interpret in patients to determine involved levels with multiple chronic complaints one previous procedure. A careful timeline should be established to try to establish a correlation between the symptoms present at the time of each imaging study, the surgical procedure, if any, that was subsequently performed, and the response of the patient s symptoms to the surgery. Available paper records including office and operative notes are also reviewed. The imaging and electrodiagnostic evaluation of patients who may benefit from revision laminectomy may differ from that of candidates for primary surgery and may involve multiple modalities (Table 43.1). Metallic implants produce artifacts on magnetic resonance imaging (MRI) that prevent the thorough evaluation of the spinal canal and the neural foramina. Plain radiographs, including supine, upright, lumbar, and long cassette studies provide information about the extent of prior bony resection and the regional and global alignment and stability of the spine. MRI obtained with and without gadolinium can help the clinician to differentiate between epidural scarring, a ubiquitous postoperative finding, and recurrent herniated disc or other causes of neural. In some cases computed tomography (CT) imaging or myelography with post-myelography CT may be necessary. Myelography, however, has limited ability to demonstrate foraminal nerve root impingement. Evidence of a pseudomeningocele should be noted, particularly if a reoperation may be indicated. Also, radiographic signs of arachnoiditis should be noted, and the significance of this finding should be discussed with the patient. 3 Electromyography (EMG) and nerve conduction velocity (NCV) studies may be difficult to interpret in the patient with a history of previous spinal surgery. Chronic denervation changes may mask acute changes responsible for current symptoms. If acute findings are present that parallel the clinical scenario, the surgeon may be reassured by this apparently confirmatory evidence. On the other hand, the absence of clear-cut findings on neurophysiological studies may not supersede strong clinical data. The importance of establishing, to the extent possible, the causes of a patient s symptoms, and discussing frankly with the patient the possible outcomes of revision surgery cannot be overemphasized. Patients considering revision surgery often have had symptoms for a long time; this may reduce the likelihood of improvement or resolution even with an adequate de. A patient who had a disappointing result from a first operation or who initially had a good result and subsequently developed symptoms may be susceptible to unrealistic expectations regarding the outcome of revision surgery. Helping the patient to develop reasonable, realistic expectations for surgery is an important part of the therapeutic relationship. The probability of a surgical complication during revision surgery is likely greater than that for an index procedure, and patients should be frankly informed of this, too. RECURRENT HERNIATED LUMBAR DISC Occasionally, the patient s history makes the diagnosis of a symptomatic recurrent lumbar herniated disc fairly straightforward (Fig. 43.2). In these cases, the patient may describe symptoms in a similar or identical location to those experienced during a previous episode that resolved significantly or completely with surgery. Alternatively, symptoms similar to the previous episode but involving the other leg may indicate a contralateral disc herniation at the previously operated level (Fig. 43.3). If an appropriate lesion is demonstrated on imaging studies, the diagnosis may be made with a high degree of certainty. A more difficult situation arises in which a patient had a previous discectomy and obtained partial or temporary relief. A return or increased intensity of symptoms often leads to the presentation for evaluation. Imaging may be equivocal. A comparison of the T1 MRI images with and without gadolinium enhancement may help to differentiate scar tissue, which enhances uniformly, from a recurrent disc herniation, which shows only peripheral enhancement. Patients in whom only epidural scar and fibrosis is seen generally do not respond as readily to surgery as those with a definite disc herniation. In the absence of a recurrent herniation, the subjacent level should be examined carefully to rule out a far lateral herniation involving the same nerve root as the previous posterolateral lesion. As for index lesions, nonoperative management options should be pursued when appropriate (Fig. 43.4). Nonsteroidal anti-inflammatory drugs (NSAIDs) have a potential role in the initial management of recurrent herniated disc in the absence of a progressive neurological deficit or cauda equina syndrome. Physical therapy may help to ameliorate mild symptoms in a

4 Chapter 43 Revision Laminectomy: Indications and Techniques 405 A B C Figure A 48-year-old man presented 9 months after an L4-5 microdiscectomy with persistent back pain and right radicular pain in a distribution similar to his initial symptoms. Sagittal T2 MRI (A) demonstrated a recurrent L4-5 disc herniation. On T1 axial imaging without contrast (B) the disc is demonstrated; it enhanced peripherally with gadolinium (C), consistent with a recurrent disc herniation. He underwent a reoperation lumbar microdiscectomy with significant improvement in his symptoms. motivated patient. A selective nerve root injection, if it produces a symptomatic relief, is informative and may be sufficiently therapeutic for the patient to avoid surgery. Recurrent disc herniations, however, seem to be less amenable to nonoperative management than first-time lesions in general, and the clinician s threshold for recommending, and the patient s for undergoing, surgery may be correspondingly lower. RECURRENT LUMBAR STENOSIS The certainty with which one may expect symptomatic improvement after revision laminectomy for stenosis may be directly related to the degree of residual stenosis and the correlation between the location of the narrowing and the location of the patient s symptoms. Although recurrent stenosis can occur, samelevel symptomatic lumbar stenosis is often the result of insufficient de as evidenced by only a brief or nonexistent period of improvement following index surgery (Fig. 43.5). The paramedian sagittal T2 images through the neural foramina bilaterally can be particularly informative regarding the adequacy of the previous foraminal de. A narrow central laminectomy may not adequately decompress the lateral recesses and the foramina. Although the intervertebral discs and foramina are at their greatest height on the supine MRI and may be expected to be narrower when the patient is upright, the utility of MRI images obtained in the upright position has not been demonstrated. Two potential limitations are the smaller magnet size on current machines that limits image resolution and that

5 406 Section V The Thoracic and Lumbar Degenerative Spine A B C D Figure A 45-year-old woman presented with severe, intractable left-sided S1 radiculopathy and had imaging demonstrating a large left L5-S1 herniated intervertebral disc (A, B). She recovered well and was asymptomatic for 2 years until she suddenly developed right-sided pain, also in an S1 distribution. MRI showed a right L5-S1 herniated disc (C, D). Nonoperative measures failed to ameliorate her symptoms, and she underwent a successful right L5-S1 microdiscectomy. many machines obtain images in the sitting position, in which many patients are relatively asymptomatic. A myelogram and post-myelography CT scan may be necessary for the evaluation of the nerve roots in patients who have had a previous fusion with instrumentation or in patients with a significant spinal deformity. These studies are most useful for demonstrating lateral recess or proximal foraminal nerve root ; it can be more difficult to assess distal to the termination of the root sleeve in the foramen. ARTHRODESIS One important treatment decision is whether an arthrodesis with or without instrumentation is necessary at the time of revision laminectomy. Typically this decision is based on several factors, primary among which is the presence of radiographic or clinical signs of instability. Supine, upright, flexion, and extension radiographs are used to assess the stability of the segment(s) to be addressed. A spondylolisthesis at the level requiring de is a finding of potential concern, particularly if excessive mobility is demonstrated on the dynamic radiographs. Other factors considered in deciding whether or not an arthrodesis is indicated include the nature of the patient s complaints, the location and degree of stenosis, and the degree of degeneration of the disc space. Similar to patients who have not had previous surgery, a significant component of axial back pain may indicate clinical instability or discogenic pain that may improve with instrumentation and arthrodesis. The surgeon should also consider the expected extent of de necessary to achieve symptomatic relief. Recurrent or persistent stenosis often involves the lateral recess and the neural foramina rather than the central canal. Adequate de may require significant lateral bone removal, particularly of the facet. Even with care taken to limit the de to the residual ligamentum flavum, ventral facet capsule, and impinging superior facet, sufficient stabilizing structures may be removed to potentially induce instability, particularly if a bilateral de is planned.

6 Chapter 43 Revision Laminectomy: Indications and Techniques 407 A B C D Figure A 63-year-old physician presented with symptoms consistent with neurogenic claudication. Imaging revealed lumbar spondylosis with stenosis at L3-4 (A) and L4-5 (B). He underwent decompressive lumbar laminectomy of L4 with de of the L3-4 and L4-5 levels with significant improvement of his symptoms. He presented 3 years postoperatively with new left leg radiculopathy without weakness. MRI showed adequate central de (C) and a left-sided L4-5 herniated intervertebral disc (D). With nonoperative treatment, his symptoms improved significantly and he continues to be followed expectantly. In some cases of recurrent disc herniation, also, an arthrodesis may be appropriate. If a simple sequestrectomy is necessary and minimal additional bony removal is anticipated a fusion may not be necessary. If, however, the intervertebral disc is extensively degenerated and further facet removal and intradiscal dissection is expected, an arthrodesis may treat axial back pain or minimize the likelihood of developing it by immobilizing the segment and arresting the degenerative process. If it is decided to perform an arthrodesis after rede, consideration should be given to an interbody fusion, either using a posterior lumbar interbody fusion (PLIF) or transforaminal lumbar interbody fusion (TLIF) approach. In the setting of multiple des, there may not be dorsal bone other than the transverse processes available for decortication and grafting. Furthermore, disc degeneration may lead to loss of disc height and relative hypolordosis; properly sized and positioned interbody structural grafts may restore segmental lordosis, which may help to reduce abnormal forces on adjacent levels. These considerations must be weighed, obviously, against the possible potential of the arthrodesis itself to induce adjacent segment degeneration. TREATMENT Although the treatment of patients with recurrent or persistent symptoms of lumbar neural frequently begins with nonoperative interventions, management of these patients may differ from that of patients who have not had previous surgery.

7 408 Section V The Thoracic and Lumbar Degenerative Spine A B C D Figure A 73-year-old man underwent L3-4 and L4-5 decompressive lumbar laminectomies for progressive neurogenic claudication unresponsive to nonoperative management. He had a brief period of incomplete relief and then had recurrence and continued progression of his symptoms. MRI showed persistent lateral recess and foraminal stenosis at L3-4 and L4-5. He underwent revision surgery with wide des and an instrumented fusion. AP radiograph following the first operation (A) show limited lateral bony removal; axial MRI at L3-4 (B) and L4-5 (C) show persistent lateral recess stenosis. Paramedian sagittal MRI (D) demonstrates persistent foraminal nerve root. Some patients, particularly those who obtained only partial relief or no significant relief from the primary surgery, will have limited patience for a protracted course of nonoperative treatment. It may be helpful, however, to have a patient who is suspected of having symptomatic recurrent or persistent stenosis undergo either an epidural steroid injection or a selective nerve root injection for therapeutic and diagnostic purposes. Other patients, including those with stenosis adjacent to a previous fusion, may be considered less likely than patients who have not had previous surgery to respond to nonoperative care. In general, however, most patients undergo a trial of nonoperative management prior to repeat surgery. SURGICAL TECHNIQUES Reoperations are generally more difficult than primary procedures and more likely to involve complications. By following a few fundamental surgical principles, however, the surgeon can minimize the likelihood of intraoperative complications and maximize the probability of an optimal clinical outcome. Working from regions of normal anatomy toward the previously operated segment helps the surgeon to avoid an unintentional durotomy. As with all lumbar spinal surgery, great care should be taken to avoid or minimize the disruption of potentially stabilizing structures at levels that will not be instrumented

8 Chapter 43 Revision Laminectomy: Indications and Techniques 409 and arthrodesed. This includes the facet capsules as well as the supra- and interspinous ligaments at the proximal and distal ends of the exposure. The bony margins surrounding the previous de are defined and cleaned of soft tissue. Sharp curettes are used to define the bony edges and to establish the dural margin. A small amount of bone removal helps this process by exposing normal dura. A high-speed drill with a small matchstick or round burr quickly thins the bone, which is then removed with curettes and Kerrison punches. While a diamond-tipped burr may offer a slight advantage over a cutting burr in terms of dural protection, it generates significant heat and generally takes longer to achieve similar bony removal. For these reasons, a standard cutting bit is used. Once the dural margin is developed, the dura is exposed over the extent of the previous de. There is generally a dissectable plane between the dura and overlying scar tissue. With firm, constant upward tension on the scar with a Leksell rongeur this plane is developed with a Cobb elevator and the thick scar tissue is removed safely and quickly. If as the plane is developed a thinning of the dura is noted dissection in that direction is halted. If possible, the tissue plane is reestablished elsewhere and scar removal proceeds toward the thinned region. If necessary, a thin layer of scar is left on the dura to avoid a cerebrospinal fluid leak. Once the bony margins have been defined and the scar thinned or removed the de proceeds as for an index procedure. REVISION DISCECTOMY The critical landmark for performing a safe, effective reoperation for discectomy, as for primary surgery, is the pedicle immediately distal to the operative disc. For an L4-5 discectomy, for example, the L5 pedicle is the key structure. Sufficient bony removal is performed to positively identify the medial pedicle. The dissection and exposure then proceeds proximally and laterally. The safe zone is immediately proximal to the pedicle and lateral to its medial border. The distally exiting nerve root (L5 in this example) will not be displaced laterally to the L5 pedicle; the L4 root will be proximal to the adjoining disc space in the medial foramen. By dissecting ventrally along the proximal aspect of the pedicle to its base the distal margin of the intervertebral disc is encountered. Working in the safe zone proximal and lateral to the medial border of the distal pedicle minimizes the probability of encountering the exiting nerve root. Care must be taken, however, if severe disc degeneration and collapse is present; as the interpedicular distance decreases the rostrocaudal dimension of the safe zone decreases, also. Soft tissue overlying the annulus is swept proximally and the dorsal annulus is identified. In the presence of significant scarring, bipolar cautery and sharp dissection are used to free the dorsal annulus of soft tissue to minimize the possibility of injury to the nerve root or dorsal root ganglion from tension and stretch. Once the dorsal annulus is identified and exposed it is incised, if necessary, and the discectomy proceeds as for a primary procedure. If a free fragment is present, unusual in a reoperation unless the herniation is contralateral to the original site of pathology, the disc space may not need to be entered. The dissection of the herniation is facilitated by staying within the disc space. The dura does not necessarily need to be dissected clean of all scars and disc materials, but any tissue remaining must be sufficiently thin to confirm that there is no residual neural. REVISION LAMINECTOMY/FACETECTOMY Once the bony margins have been established and the bone dura interface dissected carefully, the removal of additional lamina, ligamentum flavum, and facet proceeds as for an index procedure. Often the full extent of the lateral dural margin cannot initially be dissected free of the adjoining bone. In this situation, inroads are made proximally and distally to the scarred region, and lateral bony removal is performed. The dissection then proceeds toward the region of the adhesions working laterally to the adherent portion. An island of scarred or thickened dura may be created in this way. This may be shaved down and left in place if it is not causing neural. Direct palpation or visual confirmation must be used to determine that the symptomatic nerve roots are adequately decompressed. Removal of bone and soft tissue are continued until this is unquestionably established. The surgeon should avoid, however, excessive dissection around the nerve root and dorsal root ganglion in the foramen. Devascularization or stretch or injury to the nerve root or dorsal root ganglion may result in postoperative radiculitis, which may be transient or permanent. FUSION In some cases an arthrodesis may be performed as an adjunct to the de. This part of the procedure generally does not differ significantly from index surgery. The more extensive bony removal that is often performed in reoperations may, however, have two important effects with ramifications for surgical planning. First, there may be a greater degree of iatrogenic instability than following a primary procedure; the use of instrumentation may therefore be advantageous. Second, as discussed above, an interbody procedure may be appropriate to address segmental malalignment and the relative paucity of dorsal bony surface area for fusion. OUTCOMES CLINICAL OUTCOMES The comparison of outcomes of patients after primary and revision lumbar laminectomy is fraught with potentially confounding factors. It is essentially impossible to assess the effect of revision surgery independently from the potential effects of the longer duration of pain, higher proportion of worker s compensation claims, and other psychosocial factors, measured and unmeasured. In the Maine Lumbar Spine Study (MLSS), Atlas et al found a 10-year reoperation rate of 25% after lumbar discectomy. Comparing the outcomes of patients who underwent at least one reoperation to those who did not demonstrated that the former cohort had lower proportions of patients with symptomatic improvement and individuals who were satisfied with their treatment. 2 A recent case control study, however, seemed to show that patients with recurrent disc herniations within 1 year of the index surgery may have outcomes not dissimilar to those of patients without recurrence. 12 With 8- to 10-year follow-up after index surgery for lumbar stenosis, 23%

9 410 Section V The Thoracic and Lumbar Degenerative Spine of surgical patients in the MLSS underwent reoperation. Fortythree percent of those having a reoperation reported satisfaction with their health state at follow-up compared with 60% of the patients that did not undergo repeat surgery. 1 A pain-free interval after the initial surgery is associated with a higher likelihood of improvement after reoperation. 11 COMPLICATIONS Revision surgery in general is associated with a higher likelihood of surgical complications than do primary procedures. Abnormal anatomy and the presence of scar tissue increase the risk of nerve root injury or unintended durotomy during revision laminectomy. Patients should be informed preoperatively of these increased risks. Two parts of the operation during which a durotomy may occur is during the dissection of the scar off of the dura and when removing bone and ligament with Kerrison punches. Firm, steady tension on the scar tissue and careful dissection at the interface of the scar with the dura minimizes the risk of this part of the procedure. Ensuring that the dura is free of overlying bone and ligament and keeping the Kerrison orthogonal to the dura will reduce the risk of causing a durotomy during the de. If a durotomy does occur it is closed primarily, if possible, with interrupted 6-0 nonabsorbable monofilament sutures. Prior to final incision closure, the area of the durotomy is covered with a small collagen sponge and fibrin glue or another dural sealant. The most difficult dural openings to treat are often laterally or ventrally located. In these cases, primary closure is generally not possible and a small collagen sponge is placed and dural sealant applied. Routine spinal drainage is not necessary but may be indicated for large durotomies or if the closure is suboptimal, at the surgeon s discretion. Neural injury may also be expected to occur in a higher proportion of reoperations compared with first-time procedures. Neurophysiologic monitoring has not been demonstrated to reduce the risk of neural injury in decompressive lumbar procedures and its use is at the discretion of the surgeon. As for any operation, meticulous dissection and the establishment of clear anatomic landmarks are essential for a safe procedure. CONCLUSION Revision laminectomy for the treatment of lumbar disc disease or lumbar stenosis can significantly improve a patient s quality of life by ameliorating or eliminating his or her symptoms. A thorough clinical and radiographic evaluation is essential to determine the appropriateness of a reoperation for an individual patient and to help the surgeon to plan the procedure with the greatest likelihood of producing a long-term benefit. The increased risk of complications and lower proportion of patients with good outcomes with revision surgery compared with index surgery are cautionary statistics. REFERENCES 1. Atlas SJ, Keller RB, Wu YA, Deyo RA, Singer DE. Long-term outcomes of surgical and nonsurgical management of lumbar spinal stenosis: 8 to 10 year results from the Maine Lumbar Spine Study. Spine 2005;30: Atlas SJ, Keller RB, Wu YA, Deyo RA, Singer DE. Long-term outcomes of surgical and nonsurgical management of sciatica secondary to a lumbar disc herniation: 10 year results from the Maine Lumbar Spine Study. Spine 2005;30: Bernard TN. Repeat lumbar spine surgery: factors influencing outcome. Spine 1993; 18: Hu RW, Jaglal S, Axcell T, Anderson G. A populatio-based study of reoperations after back surgery. Spine 1997;22: Javid MJ, Hadar EJ. Long-term follow-up review of patients who underwent laminectomy for lumbar stenosis: a prospective study. J Neurosurg 1998;89: Katz JN, Lipson SJ, Chang LC, Levine SA, Fossel AH, Liang MH. Seven- to 10-year outcome of decompressive surgery for degenerative lumbar spinal stenosis. Spine 1996;21: Keskimaki I, Seitsalo S, Osterman H, Rissanen P. Reoperations after lumbar disc surgery: a population-based study of regional and interspecialty variations. Spine 2000;25: Martin BI, Mirza SK, Comstock BA, Gray DT, Kreuter W, Deyo RA. Reoperation rates following lumbar spine surgery and the influence of spinal fusion procedures. Spine 2007; 32: Oertel MF, Ryang Y-M, Korinth MC, Gilsbach JM, Rohde V. Long-term results of microsurgical treatment of lumbar spinal stenosis by unilateral laminotomy for bilateral de. Neurosurgery 2006;59: Silvers HR, Lewis PJ, Asch HL, Clabeaux DE. Lumbar diskectomy for recurrent disk herniation. J Spinal Disord 1994;7: Stewart G, Sachs BL. Patient outcomes after reoperation on the lumbar spine. J Bone Joint Surg 1996;78-A: Wera GD, Marcus RE, Ghanayem AJ, Bohlman HH. Failure within one year following subtotal lumbar discectomy. J Bone Joint Surg 2008;90-A:10 15.