Percutaneous Nucleotomy with CT and Fluoroscopic Guidance1 Rainer M . M. Seibel, M D Dietrich H. W. Gronemeyer, M D Reingard A. L. Sorensen, M D
Index terms: Computed tomographic guidance, 33.1211 Spine, interventional procedure, 33.1299 Spine, intervertebral disks, 336,1299,336,783
JVIR 1992; 3:571-576 Abbreviation: PNT nucleotomy
=
percutaneous
From the Institute of Diagnostic and Interventional Radiology, University Witteni Herdecke, Schulstr 10, D 4330 Mulheim a.d. Ruhr, Germany. From the 1991 SCVIR annual meeting. Received February 27, 1991; revision requested August 31; revision received February 18, 1992; accepted February 24. Address reprint requests to R.M.M.S. "
SCVIR, 1992
Automated percutaneous diskectomy was performed with use of computed tomographic (CT)and fluoroscopic monitoring. Degenerative disease of the intervertebral disk was treated with local administration of anesthesia and use of a nucleotome. One hundred ten patients with neurologic symptoms and morphologic changes of one segment were selected for treatment. Previous conservative therapy had been unsuccessful. Patients with completely prolapsed and sequestered fragments of herniated disks ("uncontained disk"), narrow intervertebral spaces, posterior osteophytes, diseased facet joints, and spinal stenoses were not considered candidates for percutaneous nucleotomy (PNT).After PNT, 82%of the patients had complete remission of their neurologic symptoms; Lasegue sign was negative or improved in 92%.In 18%(20 patients), the symptoms did not improve sufficiently; 11%(12 of 110)of these patients underwent surgical nucleotomy. There were no serious complications, in particular, no injuries to vital structures (nerves, thecal sac, arteries, veins), except for one case of spondylodiskitis. Guiding PNT with CT and fluoroscopy provides a safe procedure with good clinical results. The addition of CT has shortened the operation but increased overall procedure time. In the future, a shift to outpatient treatment may offset the additional time and cost of including CT guidance. S U R G E R Y of the herniated intervertebral disk was introduced by Oppenheim and Krause in 1909 (1). They performed complete transdural laminectomies to remove the herniated segment, which they interpreted as being an enchondroma. Lasegue (2) described the clinical symptoms of the disease, and Mixter and Barr (3) recognized the cause of sciatica as a herniation of the nucleus pulposus. Oppenheim's surgical technique was modified by Love (hemilaminectomy, minilaminectomy) (4); the procedure was performed extradurally. The introduction of microsurgical techniques (5) reduced the injury to the surrounding tissue, scar formation, and vertebral instability following laminectomy. Intraoperative complications, however, ranged between 7% and 14%, and surgical exploration had to be repeated in 14%-28%. Postoperative complications (delayed
healing, diskitis, early recurrence, pulmonary embolism) are described to occur in between 0.8% and 7% (6,7). The risk of complications and the recuperative time are directly related to the duration of the surgical intervention and the age of the patient. Chemonucleolysis is a n alternative to laminectomy. For the chemical dissolution of the intervertebral disk, chymopapain is injected into the nucleus pulposus (8,9). The injection is done under local anesthesia and appears to be almost without risk. A comparative study demonstrated that in patients with a small herniation, results of chemonucleolysis were slightly better than those of surgical interventions (75% vs 62% a t 1 month, 84% vs 82% a t final followup). The results were not as good in patients with large herniations (surgical success rate 89%, chemonucleo-
572
Journal of Vascular and Interventional Radiology
August 1992
lysis success 50%) (10). In selected patients with small herniations, 70%-75% had instant relief of ~ a i n and a very short hospital stay. ?'he procedure, however, was limited to small herniations. Results were very poor in patients who had previous surgery, spinal stenoses, lateral recess stenoses, or sequestered disks (9). Chemonucleolvsis has three advantages over surgery: It is a minor procedure that can be performed under local anesthesia; it avoids injuries to the soft tissue and local scarring.: it requires a shorter hospital stay\nd, thus, is less expensive. Chemonucleolysis fails in 15%-30% and must be followed by surgery if unsuccessful (10,11), it has a recurrence rate of 8%, and it accelerates degeneration (12,131. Occasionally, complications such as anaphylactic reactions (0.2%) and transverse myelitis occur (9). To reduce the invasiveness of surgery and to overcome the complications and limitations of chemonucleolysis, percutaneous lumbar diskectomy was developed in 1975 (14-19) and was later combined with diskoscopy (19). A main advantage of percutaneous diskectomy over minilaminectomy was preservation of the stability of the vertebral spine and the soft-tissue structures and the possibility of a controlled removal of the nucleus pulposus. After administration of a general anesthetic, an introducer 11mm in diameter was used bilaterally. The disk was removed by means of a forceps through a cannula. The necessary long manipulation time increased the danger of infection; the size of the instrument also created injuries to vital structures. Excellent results were reported in 70%-87% of cases (7,15,18,20-26). There were, however, difficulties in treating diskopathies a t L5-S1 due to anatomic reasons. Punctures at this level were only successful in one of the three patients (15). Automated percutaneous diskectomy was introduced by Onik and others (14,27,28). A cannula with a 2.8-mm outer diameter was advanced under fluoroscopic guidance into the
intervertebral disk. The trocar was replaced by a 2-mm nucleotome, which automatically aspirated the nucleus pulposus through a side hole. Onik et al reported a success rate of 75.2%. They treated 327 patients at segments L-1 to S-1. The follow-up was 12 months. The complication rate was 1%. Com~licationsincluded infection, psoas hematoma, vasovagal reactions, and postoperative spasm (27,28). With fluoroscopic guidance alone (29-31), it is still possible to lacerate vital structures because soft-tissue structures cannot be identified at the time of the procedure. Since 1989, computed tomography (CT) has been used as a guide for automated percutaneous diskectomy. I t has enabled visualization of small nerves; the thecal sac; the center of the disk; and the medial border of the pedicles, arteries, and veins within 1cm. CT also provides a three-dimensional method for determination of the angle necessary for the puncture. The present study assesses the feasibility of CT monitoring for reducing the duration of the procedure, decreasing the manipulation of the soft tissue, and avoiding infection and laceration of vital structures. The CT-guided procedure was also assessed for its role in rendering heretofore contraindicated cases amenable to percutaneous treatment.
PATIENTS AND METHODS Patient Population In 1988, we began performing nucleotomy with the combination of a C arm with an integrated fluoroscopy unit and CT. One hundred ten patients were treated until January 1991. Their mean age was 45 years (range, 25-66 years). Fifty-eight patients underwent PNT at the intervertebral disc spaces of L4-5, 50 at L5-S1, and two at L3-4. Clinical Criteria for Patient Selection Protrusion and herniation of the intervertebral disk of the lumbar
spine, followed by sciatica, is the primary complaint of all patients. If this is combined with segmental neurologic and morphologic signs and symptoms, and if the discomfort into the lower extremity is more pronounced than back pain, treatment is indicated and the patient is considered a potential candidate for percutaneous nucleotomy (PNT).
Physical Findings The physical findings are pain when the leg is raised, crossover pain and a positive bowstring sign, wasting and weakness of the musculature, and the loss of sensitivity and reflexes. Conservative Care Conservative therapy prior to PNT includes bed rest, medication for pain relief, and anti-inflammatory drug therapy. A course of physical therapy for at least 6 weeks usually is successful. In our institution periradicular therapy is performed with injection of 1-2 mL of mepivacaine 0.5% and 10-40 mg of triamcinolon (Volon A soluble; Squibb-Heyden, Munich) prior to PNT. This long-acting local anesthetic agent and the corticosteroid are injected around the segmental nerve under CT guidance (30,311. Only if those treatments fail is PNT applied. Radiographic Criteria The pathologic condition must be verified by means of CT or magnetic resonance (MR) imaging. Acquisition of an additional myelogram is not necessary if neurologic symptoms and the disease extent correspond. CT is performed at the lumbar region of L-3 to S-1 with acquisition of 2-mm sections. Radiographs of the lumbar spine in anterior-posterior and lateral positions are required to identify bone abnormalities outside of the recognized findings. MR imaging usually demonstrates a herniated nucleus pulposus. We distinguish between stages of disk disease: Only "contained disks" are amenable to PNT. A "contained disk" is (a)the protrusion of the intervertebral disk
Seibel et a1
573
Volume 3 Number 3
Figure 1. CT- and
fluoroscopy-guided PNT. The C arm is used with CT for three-dimensional imaging.
fragments were not considered suitable for PNT. Uncontained disks were usually not treated by means of PNT. To avoid bleeding complications, PNT was not performed in patients with coagulopathies. Patients who had undergone previous surgery were not excluded.
Outcome Criteria The interval of follow-up examinations was 3 weeks, 6 weeks, 3 months, 6 months, and 1year after PNT. The criteria for a successful ~rocedurewere the im~rovementof physical and neurologic signs (pain, Lasegue sign). The disappearance of physical signs and neurologic symptoms was correlated with CT findings 3 weeks and 6 months following PNT and with MR findings of the lumbar spine 6 weeks after PNT. Technique PNT was ~erformedwith the automated nuclebtome (Nucleotome R; Surgical Dynamics, San Leandro, Calif) (27,28). Punctures were made with the Seldinger technique. Just prior to treatment, a CT scan of the repion of interest was obtained with the patient prone to demonstrate recent pathologic changes (Fig 1). The intervertebral disk space was scanned in an axial plane. The angle for the puncture and the distance from the region of interest to the skin were calculated and marked on the CT section (Fig 2). The puncture site was chosen such that the most efficient suction effect could be applied to the site of the herniated disk. This was done by means of a three-dimensional technique (combination of fluoroscopy with a C arm, Siremobil, and Somatom DRG CT scanner, Siemens; Fig 1). The angle chosen for puncture was oblique, traversing several CT sections (Fig 3). After local administration of anesthesia to the skin and musculature with a coaxial needle set, a 22-gauge needle, 20 cm long, was advanced to the anulus fibrosus with fluoroscopic and CT monitoring. Local anesthesia of the nerve root was avoided so that the patient could respond to pain. Theou
a.
D.
Figure 2. PNT at L4-5. (a)Axial CT scan of the intervertebral structures obtained with the patient in the prone position. The direction and angle of the puncture are
planned with CT and fluoroscopic guidance ( 1 ) .The distance is marked on the patient's skin (2).(b) Under the guidance of three-dimensional imaging, the 22-gauge needle is advanced toward the segmental nerve root. The nerve is visualized as a linear structure at the tip of the needle. With the blunt tip, the nerve can be gently shifted anterior to the instruments.
with the beginning of a dislocation of the anulus fibrosus, (6) the protrusion of parts of the nucleus pulposus with compression of the dural sac, (c)the herniation of the intervertebral disk with the destruction of the anulus fibrosus, and ( d )the herniation with an extension beneath the ligament. An "uncontained disk" is (a) the herniation through the posterior longitudinal ligament with the destruction of its structure and (b) the herniation with a fragment (sequestrum) of the disc. PNT is not performed in patients with a
dislocated sequestrum. PNT of a herniated nucleus pulposus with the destruction of the posterior longitudinal ligament is possible if there is no fragment visualized at CT or MR imaging.
Exclusion Criteria Patients with extremely narrow intervertebral spaces, large vacuum phenomena, posterior osteophytes protruding into the spinal canal, diseased facet joints, narrow spinal canals, large herniations with compression of the spinal cord, or free disk
574
Journal of Vascular and Interventional Radiology
August 1992
retically, anesthesia might increase the possibility of injury to the nerve. The needle was guided in such a way that it passed the facet joint and the structures underneath the segmental nerve. CT was used to control the position. A dilator was advanced over the needle to the anulus fibrosus together with a straight or curved cannula. The position was documented with fluoroscopy and CT (Fig 3). If there was no interposition of tissue between the cannula and the anulus fibrosus, the trocar was advanced into the nucleus pulposus. After another CT scan was obtained to check the position, a channel 2 mm in diameter was drilled into the anulus fibrosus with a trephine. The nucleotome could now be introduced into the nucleus pulposus (Fig 4). The puncture and the introduction of the instrument can be especially dangerous because of the proximity of the segmental nerves. Control of every step of the procedure was accomplished with CT and fluoroscopy. If segmental neurologic symptoms occurred during introduction of the instruments, an injury of the segmental nerve was suspected and accurate documentation of the level of a possible laceration was necessary. The nucleotome aspirated the nucleus pulposus. During suction, its position was changed manually to remove anterior, posterior, and lateral parts of the nucleus pulposus as well as the herniated portion. The mean apiration time was 32 minutes; the entire procedure lasted between 50 and 60 minutes.
satisfactory; 11% of all patients (12 of 110) underwent conventional surgery within 1year.
Physical Findings after PNT Positive results were seen in 92% immediately following PNT. Lasegue sign was negative in 67% or improved (by 30" or more) in 25%. CT and MR Findings Immediately after PNT, CT signs of herniated disks were gone or significantly reduced in only 56% of cases. At MR imaging 3 months after PNT, reduction of the herniation was seen in only 53% of patients.
Complications Complications of CT-guided and fluoroscopic-guided PNT are rare. We encountered one case of spondylodiskitis, which was cured with antibiotic therapy. Nerve root irritation occurred at the level of the segments treated in 23% of the patients. Pain occurred 2 days following PNT and lasted up to 5 weeks. The symptoms were considered due to mechanical irritation from Dressure on the nerve root bv the nucleotome. Spondylodiskitis was excluded as a cause of pain by means of MR imaging. The symptoms subsided spontaneously in most of the patients; physical, conservative, and periradicular therapy were necessary in 6.6%. Pseudoradicular pain mimics the symptoms of sciatica. The pain, however, is a diffuse discomfort confined to the back or radiating to the pelvis, gluteal area, groin, and thigh. Pain increases with movement. Neurologic RESULTS signs are absent; pain is not confined to a segment. Pseudoradicular pain is Among all patients, 82% were free caused by changes in the alignment of pain during or immediately after of the articular facets following any PNT. An average of 2.9 g of material kind of mechanical stress. Among per patient (range, 1.9-7.9 g) was patients treated with PNT, 20% had removed from intervertebral disks. symptoms of pseudoradicular disBy 42 days after the procedure, 85% of the patients had returned to work. ease. This was treated by means of CT-guided blockade of the articular Three months after PNT, 80% were facets bilaterally. A periarticular infree of pain and 84% had no neurologic symptoms. In 18% (20 patients), jection with a local anesthetic gave relief. improvement in symptoms was not
Figure 3. The straight cannula has
been introduced and advanced through the intervertebral foramen into the disk. Suction can be directed toward the region of herniation under CT guidance to keep the procedure most efficient. To completely demonstrate the instruments, which are in an oblique position, several CT sections are usually necessary.
DISCUSSION Onik et al(20) were the first to use automated percutaneous diskectomy. The main advantage of automated over manual percutaneous diskectomy is its safety (8,14,15,19,20,23, 27,30,31). Use of CT for planning the procedure helps avoid puncturing organs within the peritoneum (33). Immediate detection of pain during manipulation of an instrument alerts the operator to the possibility of a nerve injury. To minimize nerve root lacerations, local anesthetic is not used at the nerve roots. The center of the disk and the thecal sac can be located accurately with CT. The anulus fibrosus is not cut but is only traversed by the trephine, so suction of the anulus into the side port can be avoided. If the instrument is placed into the center of the disk, its blunt end remains there during the procedure. It does not cut in the forward direction, so trauma to vital structures is avoided. Compared with the percutaneous manual method, the procedure described by Onik et a1 is less time-consuming and the cannulas necessary to remove the material are much smaller. There is less danger of injury to or contamination of the surrounding soft tissue.
Seibel et a1
575
Volume 3 Number 3
Figure 5. Follow-up CT scan obtained after nucleotomy. The nucleotome has been removed. Scan shows a vacuum phenomenon indicating the area where the nucleus pulposus had been removed.
b. Figure 4. Percutaneous CT- and fluoroscopy-guided PNT. The tip of the nucleotome is seen in the middle of the intervertebral disk. (a)Position a t CT. (b)Position a t fluoroscopy.
Percutaneous diskectomy guided by means of fluoroscopy alone poses a potential danger of lacerating softtissue structures, as these cannot be identified at the time of the procedure. Prevention of nerve root and other injuries, as described by Epstein (29), involves avoidance of traversing nerves; the thecal sac; the center of the disk; and the medial border of the pedicIes, arteries, and veins. The ability to localize these
structures 1cm from the tip of the instrument is a distinct advantage of CT. The fluoroscopic unit in front of the gantry is useful for rapid identification of anteroposterior and lateral bone structures during puncture and of the instrument in the center of the intervertebral disk. The technique offers the advantage of an optimal placement of the nucleotome close to the herniated nucleus pulposus. Direct visualization of softtissue structures avoids the thecal sac, too. Therefore, injury of the cauda equina should be eliminated. The special difficulties of treating the disk at L5-S1 have been overcome by using curved instruments. In contrast to results reported in the literature (30,311, almost 50% of our patients had herniations at L5-S1 and were treated successfully. CT monitoring at L5-S1 is necessary to prevent nerve root injury. In about half the cases, the morphologic changes following removal of the herniated disk can be visualized on CT scans immediately following PNT (Fig 5). However symptoms may also disappear after PNT, even if only minimal change is demonstrated on CT scans (34). This may be due to the relief of pressure on the nerve. PNT procedures under the guidance of CT and fluoroscopy last
50-60 minutes. This includes a CT examination prior to and following the actual operation. The average time of aspiration was only 32 minutes, which is almost identical to the time the nucleotome is in place within the nucleus pulposus. This compares with the time described in the multi-institutional study of Onik and others (27,28). Their group reported that the procedure required less than 40 minutes to complete. This, however, does not include the CT examination before and following PNT. The risks of infection from contamination are not increased. Acquisition of additional CT scans increases the cost of the entire procedure. In the future, a shift to outpatient treatment may offset the additional cost of the CT portion of the procedure. References 1. Oppenheim H, Krause F. ijber einklemmung bzw: strangulation der cauda equina. Dtsch Med Wschr 1909; 35:697-708. 2. LasBgue C. Considerations sur la sciatique. Arch Gen Med 1864; 4:558-563. 3. Mixter WJ, Barr JS. Rupture of the intervertebral disc with involvement of the spinal canal. N Engl J Med 1934; 210-211. 4. Love JG. Removal of protruded intervertebral discs without laminectomy. Proc Staff Meet, Mayo Clin 1939; 14:800-807. 5. Yasargil MG. Microsurgical operation of herniated lumbar discs. In: Advances in neurosurgery. Berlin: Springer Verlag, 1977; 81-82. 6. El-Gindi S, Aref S, Salama M, Andrew J. Infection of intervertebral discs after operation. J Bone Joint Surg [Brl 1976; 58:114-116. 7. Stolke D, Sollmann WD, Seifert V. Intra- and postoperative complications in lumbar disc surgery. Spine 1989; 14:56-59. 8. Smith L. Enzyme dissolution of the nucleus pulposus in humans. JAMA 1964; 187:137-140. 9. McCulloch JA. Chemonucleolysis. J Bone Joint Surg [Br] 1977; 59:4552. 10. Postacchini F, Lami R, Massobrio M. Chemonucleolysis versus sur-
576
Journal of Vascular and Interventional Radiology
August 1992
gery in lumbar disc herniations: correlation of the results to preoperative clinical pattern and size of the herniation. Spine 1987; 12247-96. Edwards WC, Orme TJ, Orr-Edwards G. CT discography: prognostic value in the selection of patients for chemonucleolysis. Spine 1987; 12:792-795. Huckmann MS, Clark JW, McNeil TW. Chemonucleation and changes observed on lumbar MR scan: preliminary report. AJNR 1987; 8:l-6. Nelson KL, Runge VM. Lumbar spine. In: Runge VM, ed. Clinical magnetic resonance imaging. Philadelphia: Lippincott, 1990; 269-278. Hijikata S, Yamlagishi M, Nakayama T, Oomori K. Percutaneous discectomy: a new treatment method for lumbar disc herniation. J Toden Hosp 1975; 5:5-13. Hijikata S. Percutaneous nucleotomy: a new concept technique and 12 years experience. Clin Orthop 1989; 238:9-23. Friedman WA. Percutaneous discectomy: an alternative to chemonucleolysis? Neurosurgery 1983; 13: 542-547. Kambin P, Sampson S. Posterolatera1 percutaneous suction-excision of herniated lumbar intervertebral discs. Clin Orthop Re1 Res 1986; 207:37-43. Kambin P, Schaffer JL. Percutaneous lumbar discectomy: review of 100 patients and current practice. Clin Orthop 1989; 238:24-34. Suezawa Y, Jacob H. Percutaneous nucleotomy: an alternative to spinal surgery. Arch Orthop Trauma Surg 1986; 105:287-291.
Onik G, Helms CA, Ginsberg L, Hoaglund FT, Morris J . Percutaneous lumbar discectomy using a new aspiration probe. AJNR 1985; 153:290293. Bocchi L, Ferrata P, Passarello F, et al. La nucleoaspirazione second0 Onik nel trattamento dell'ernia discale lombare analisi multicentrica dei primi risultati su oltre 650 trattamenti. Riv Neuroradiol 1989; 2 (suppl l):119-122. Monteiro A, Lefevre R, Pieters G, Wilmet E. Lateral decompression of a pathological disc in the treatment of lumbar pain and sciatica. Clin Orthop 1989; 238:56-63. Schreiber A, Suezawa Y, Leu H. Does percutaneous nucleotomy with discigraphy replace conventional discectomy? Eight years of experience and results in treatment of herniated lumbar disc. Clin Orthop 1989; 238:35-42. Shepperd JAN, James SE, Leach AS. Percutaneous disc surgery. Clin Orthop 1989; 238:43-49. Wilson DH, Harbaugh R. Lumbar discectomy: a comparative study of microsurgical and standard technique. In: Hardy R, ed. Lumbar disc disease. New York: Raven, 1982; 147-156. Stern MN. Early experience with percutaneous lateral discectomy. Clin Orthop 1989; 238:50-55. Onik G, Mooney V, Maroon JC, et al. Automated percutaneous discectomy: a prospective multi-institutional study. Neurosurgery 1990; 26:228-233. Onik G, Helms CA. Automated percutaneous lumbar discectomy. AJR 1991; 156:531-538.
29. Epstein NE. Surgically confirmed cauda equina and nerve root injury following percutaneous discectomy at an outside institution: a case report. J Spinal Disorder 1990; 3:380383. 30. Seibel RMM, Gronemeyer DHW, Grumme TH. Neue verfahren der therapie degenerativer wirbelsaulenerkrankungen in der interventionellen radiologie. In: Gronemeyer DHW, Seibel RMM, eds. Interventionelle computertomographie. Vienna: Blackwell Scientific, 1989; 92130. 31. Seibel RMM, Gronemeyer DHW, Grumme TH. New methods of treatment of spinal column diseases using interventional radiological techniques. In: Seibel RMM, Gronemeyer DHW, eds. Interventional computed tomography. Boston: Blackwell Scientific 1990; 89-133. 32. Dilke TFW, Burry HC, Grahame R. Extradural corticosteroid injection in the management of lumbar nerve compression. Br Med J 1973; 2:635637. 33. Hopper K, Sherman J, Luethke J . Retrorenal colon in the supine and prone patient. Radiology 1987; 162: 443-446. 34. Bonneville JF, Runge M, Issaadi K, Tang YS, Guetarni S, Boulard D. Percutaneous discectomy using aspiration. Radiologie 1989; 9: 19951996.
Index terms: Computed tomographic guidance, 33.1211 Spine, interventional procedure, 33.1299 Spine, intervertebral disks, 336,1299,336,783
JVIR 1992; 3:571-576 Abbreviation: PNT nucleotomy
=
percutaneous
From the Institute of Diagnostic and Interventional Radiology, University Witteni Herdecke, Schulstr 10, D 4330 Mulheim a.d. Ruhr, Germany. From the 1991 SCVIR annual meeting. Received February 27, 1991; revision requested August 31; revision received February 18, 1992; accepted February 24. Address reprint requests to R.M.M.S. "
SCVIR, 1992
Automated percutaneous diskectomy was performed with use of computed tomographic (CT)and fluoroscopic monitoring. Degenerative disease of the intervertebral disk was treated with local administration of anesthesia and use of a nucleotome. One hundred ten patients with neurologic symptoms and morphologic changes of one segment were selected for treatment. Previous conservative therapy had been unsuccessful. Patients with completely prolapsed and sequestered fragments of herniated disks ("uncontained disk"), narrow intervertebral spaces, posterior osteophytes, diseased facet joints, and spinal stenoses were not considered candidates for percutaneous nucleotomy (PNT).After PNT, 82%of the patients had complete remission of their neurologic symptoms; Lasegue sign was negative or improved in 92%.In 18%(20 patients), the symptoms did not improve sufficiently; 11%(12 of 110)of these patients underwent surgical nucleotomy. There were no serious complications, in particular, no injuries to vital structures (nerves, thecal sac, arteries, veins), except for one case of spondylodiskitis. Guiding PNT with CT and fluoroscopy provides a safe procedure with good clinical results. The addition of CT has shortened the operation but increased overall procedure time. In the future, a shift to outpatient treatment may offset the additional time and cost of including CT guidance. S U R G E R Y of the herniated intervertebral disk was introduced by Oppenheim and Krause in 1909 (1). They performed complete transdural laminectomies to remove the herniated segment, which they interpreted as being an enchondroma. Lasegue (2) described the clinical symptoms of the disease, and Mixter and Barr (3) recognized the cause of sciatica as a herniation of the nucleus pulposus. Oppenheim's surgical technique was modified by Love (hemilaminectomy, minilaminectomy) (4); the procedure was performed extradurally. The introduction of microsurgical techniques (5) reduced the injury to the surrounding tissue, scar formation, and vertebral instability following laminectomy. Intraoperative complications, however, ranged between 7% and 14%, and surgical exploration had to be repeated in 14%-28%. Postoperative complications (delayed
healing, diskitis, early recurrence, pulmonary embolism) are described to occur in between 0.8% and 7% (6,7). The risk of complications and the recuperative time are directly related to the duration of the surgical intervention and the age of the patient. Chemonucleolysis is a n alternative to laminectomy. For the chemical dissolution of the intervertebral disk, chymopapain is injected into the nucleus pulposus (8,9). The injection is done under local anesthesia and appears to be almost without risk. A comparative study demonstrated that in patients with a small herniation, results of chemonucleolysis were slightly better than those of surgical interventions (75% vs 62% a t 1 month, 84% vs 82% a t final followup). The results were not as good in patients with large herniations (surgical success rate 89%, chemonucleo-
572
Journal of Vascular and Interventional Radiology
August 1992
lysis success 50%) (10). In selected patients with small herniations, 70%-75% had instant relief of ~ a i n and a very short hospital stay. ?'he procedure, however, was limited to small herniations. Results were very poor in patients who had previous surgery, spinal stenoses, lateral recess stenoses, or sequestered disks (9). Chemonucleolvsis has three advantages over surgery: It is a minor procedure that can be performed under local anesthesia; it avoids injuries to the soft tissue and local scarring.: it requires a shorter hospital stay\nd, thus, is less expensive. Chemonucleolysis fails in 15%-30% and must be followed by surgery if unsuccessful (10,11), it has a recurrence rate of 8%, and it accelerates degeneration (12,131. Occasionally, complications such as anaphylactic reactions (0.2%) and transverse myelitis occur (9). To reduce the invasiveness of surgery and to overcome the complications and limitations of chemonucleolysis, percutaneous lumbar diskectomy was developed in 1975 (14-19) and was later combined with diskoscopy (19). A main advantage of percutaneous diskectomy over minilaminectomy was preservation of the stability of the vertebral spine and the soft-tissue structures and the possibility of a controlled removal of the nucleus pulposus. After administration of a general anesthetic, an introducer 11mm in diameter was used bilaterally. The disk was removed by means of a forceps through a cannula. The necessary long manipulation time increased the danger of infection; the size of the instrument also created injuries to vital structures. Excellent results were reported in 70%-87% of cases (7,15,18,20-26). There were, however, difficulties in treating diskopathies a t L5-S1 due to anatomic reasons. Punctures at this level were only successful in one of the three patients (15). Automated percutaneous diskectomy was introduced by Onik and others (14,27,28). A cannula with a 2.8-mm outer diameter was advanced under fluoroscopic guidance into the
intervertebral disk. The trocar was replaced by a 2-mm nucleotome, which automatically aspirated the nucleus pulposus through a side hole. Onik et al reported a success rate of 75.2%. They treated 327 patients at segments L-1 to S-1. The follow-up was 12 months. The complication rate was 1%. Com~licationsincluded infection, psoas hematoma, vasovagal reactions, and postoperative spasm (27,28). With fluoroscopic guidance alone (29-31), it is still possible to lacerate vital structures because soft-tissue structures cannot be identified at the time of the procedure. Since 1989, computed tomography (CT) has been used as a guide for automated percutaneous diskectomy. I t has enabled visualization of small nerves; the thecal sac; the center of the disk; and the medial border of the pedicles, arteries, and veins within 1cm. CT also provides a three-dimensional method for determination of the angle necessary for the puncture. The present study assesses the feasibility of CT monitoring for reducing the duration of the procedure, decreasing the manipulation of the soft tissue, and avoiding infection and laceration of vital structures. The CT-guided procedure was also assessed for its role in rendering heretofore contraindicated cases amenable to percutaneous treatment.
PATIENTS AND METHODS Patient Population In 1988, we began performing nucleotomy with the combination of a C arm with an integrated fluoroscopy unit and CT. One hundred ten patients were treated until January 1991. Their mean age was 45 years (range, 25-66 years). Fifty-eight patients underwent PNT at the intervertebral disc spaces of L4-5, 50 at L5-S1, and two at L3-4. Clinical Criteria for Patient Selection Protrusion and herniation of the intervertebral disk of the lumbar
spine, followed by sciatica, is the primary complaint of all patients. If this is combined with segmental neurologic and morphologic signs and symptoms, and if the discomfort into the lower extremity is more pronounced than back pain, treatment is indicated and the patient is considered a potential candidate for percutaneous nucleotomy (PNT).
Physical Findings The physical findings are pain when the leg is raised, crossover pain and a positive bowstring sign, wasting and weakness of the musculature, and the loss of sensitivity and reflexes. Conservative Care Conservative therapy prior to PNT includes bed rest, medication for pain relief, and anti-inflammatory drug therapy. A course of physical therapy for at least 6 weeks usually is successful. In our institution periradicular therapy is performed with injection of 1-2 mL of mepivacaine 0.5% and 10-40 mg of triamcinolon (Volon A soluble; Squibb-Heyden, Munich) prior to PNT. This long-acting local anesthetic agent and the corticosteroid are injected around the segmental nerve under CT guidance (30,311. Only if those treatments fail is PNT applied. Radiographic Criteria The pathologic condition must be verified by means of CT or magnetic resonance (MR) imaging. Acquisition of an additional myelogram is not necessary if neurologic symptoms and the disease extent correspond. CT is performed at the lumbar region of L-3 to S-1 with acquisition of 2-mm sections. Radiographs of the lumbar spine in anterior-posterior and lateral positions are required to identify bone abnormalities outside of the recognized findings. MR imaging usually demonstrates a herniated nucleus pulposus. We distinguish between stages of disk disease: Only "contained disks" are amenable to PNT. A "contained disk" is (a)the protrusion of the intervertebral disk
Seibel et a1
573
Volume 3 Number 3
Figure 1. CT- and
fluoroscopy-guided PNT. The C arm is used with CT for three-dimensional imaging.
fragments were not considered suitable for PNT. Uncontained disks were usually not treated by means of PNT. To avoid bleeding complications, PNT was not performed in patients with coagulopathies. Patients who had undergone previous surgery were not excluded.
Outcome Criteria The interval of follow-up examinations was 3 weeks, 6 weeks, 3 months, 6 months, and 1year after PNT. The criteria for a successful ~rocedurewere the im~rovementof physical and neurologic signs (pain, Lasegue sign). The disappearance of physical signs and neurologic symptoms was correlated with CT findings 3 weeks and 6 months following PNT and with MR findings of the lumbar spine 6 weeks after PNT. Technique PNT was ~erformedwith the automated nuclebtome (Nucleotome R; Surgical Dynamics, San Leandro, Calif) (27,28). Punctures were made with the Seldinger technique. Just prior to treatment, a CT scan of the repion of interest was obtained with the patient prone to demonstrate recent pathologic changes (Fig 1). The intervertebral disk space was scanned in an axial plane. The angle for the puncture and the distance from the region of interest to the skin were calculated and marked on the CT section (Fig 2). The puncture site was chosen such that the most efficient suction effect could be applied to the site of the herniated disk. This was done by means of a three-dimensional technique (combination of fluoroscopy with a C arm, Siremobil, and Somatom DRG CT scanner, Siemens; Fig 1). The angle chosen for puncture was oblique, traversing several CT sections (Fig 3). After local administration of anesthesia to the skin and musculature with a coaxial needle set, a 22-gauge needle, 20 cm long, was advanced to the anulus fibrosus with fluoroscopic and CT monitoring. Local anesthesia of the nerve root was avoided so that the patient could respond to pain. Theou
a.
D.
Figure 2. PNT at L4-5. (a)Axial CT scan of the intervertebral structures obtained with the patient in the prone position. The direction and angle of the puncture are
planned with CT and fluoroscopic guidance ( 1 ) .The distance is marked on the patient's skin (2).(b) Under the guidance of three-dimensional imaging, the 22-gauge needle is advanced toward the segmental nerve root. The nerve is visualized as a linear structure at the tip of the needle. With the blunt tip, the nerve can be gently shifted anterior to the instruments.
with the beginning of a dislocation of the anulus fibrosus, (6) the protrusion of parts of the nucleus pulposus with compression of the dural sac, (c)the herniation of the intervertebral disk with the destruction of the anulus fibrosus, and ( d )the herniation with an extension beneath the ligament. An "uncontained disk" is (a) the herniation through the posterior longitudinal ligament with the destruction of its structure and (b) the herniation with a fragment (sequestrum) of the disc. PNT is not performed in patients with a
dislocated sequestrum. PNT of a herniated nucleus pulposus with the destruction of the posterior longitudinal ligament is possible if there is no fragment visualized at CT or MR imaging.
Exclusion Criteria Patients with extremely narrow intervertebral spaces, large vacuum phenomena, posterior osteophytes protruding into the spinal canal, diseased facet joints, narrow spinal canals, large herniations with compression of the spinal cord, or free disk
574
Journal of Vascular and Interventional Radiology
August 1992
retically, anesthesia might increase the possibility of injury to the nerve. The needle was guided in such a way that it passed the facet joint and the structures underneath the segmental nerve. CT was used to control the position. A dilator was advanced over the needle to the anulus fibrosus together with a straight or curved cannula. The position was documented with fluoroscopy and CT (Fig 3). If there was no interposition of tissue between the cannula and the anulus fibrosus, the trocar was advanced into the nucleus pulposus. After another CT scan was obtained to check the position, a channel 2 mm in diameter was drilled into the anulus fibrosus with a trephine. The nucleotome could now be introduced into the nucleus pulposus (Fig 4). The puncture and the introduction of the instrument can be especially dangerous because of the proximity of the segmental nerves. Control of every step of the procedure was accomplished with CT and fluoroscopy. If segmental neurologic symptoms occurred during introduction of the instruments, an injury of the segmental nerve was suspected and accurate documentation of the level of a possible laceration was necessary. The nucleotome aspirated the nucleus pulposus. During suction, its position was changed manually to remove anterior, posterior, and lateral parts of the nucleus pulposus as well as the herniated portion. The mean apiration time was 32 minutes; the entire procedure lasted between 50 and 60 minutes.
satisfactory; 11% of all patients (12 of 110) underwent conventional surgery within 1year.
Physical Findings after PNT Positive results were seen in 92% immediately following PNT. Lasegue sign was negative in 67% or improved (by 30" or more) in 25%. CT and MR Findings Immediately after PNT, CT signs of herniated disks were gone or significantly reduced in only 56% of cases. At MR imaging 3 months after PNT, reduction of the herniation was seen in only 53% of patients.
Complications Complications of CT-guided and fluoroscopic-guided PNT are rare. We encountered one case of spondylodiskitis, which was cured with antibiotic therapy. Nerve root irritation occurred at the level of the segments treated in 23% of the patients. Pain occurred 2 days following PNT and lasted up to 5 weeks. The symptoms were considered due to mechanical irritation from Dressure on the nerve root bv the nucleotome. Spondylodiskitis was excluded as a cause of pain by means of MR imaging. The symptoms subsided spontaneously in most of the patients; physical, conservative, and periradicular therapy were necessary in 6.6%. Pseudoradicular pain mimics the symptoms of sciatica. The pain, however, is a diffuse discomfort confined to the back or radiating to the pelvis, gluteal area, groin, and thigh. Pain increases with movement. Neurologic RESULTS signs are absent; pain is not confined to a segment. Pseudoradicular pain is Among all patients, 82% were free caused by changes in the alignment of pain during or immediately after of the articular facets following any PNT. An average of 2.9 g of material kind of mechanical stress. Among per patient (range, 1.9-7.9 g) was patients treated with PNT, 20% had removed from intervertebral disks. symptoms of pseudoradicular disBy 42 days after the procedure, 85% of the patients had returned to work. ease. This was treated by means of CT-guided blockade of the articular Three months after PNT, 80% were facets bilaterally. A periarticular infree of pain and 84% had no neurologic symptoms. In 18% (20 patients), jection with a local anesthetic gave relief. improvement in symptoms was not
Figure 3. The straight cannula has
been introduced and advanced through the intervertebral foramen into the disk. Suction can be directed toward the region of herniation under CT guidance to keep the procedure most efficient. To completely demonstrate the instruments, which are in an oblique position, several CT sections are usually necessary.
DISCUSSION Onik et al(20) were the first to use automated percutaneous diskectomy. The main advantage of automated over manual percutaneous diskectomy is its safety (8,14,15,19,20,23, 27,30,31). Use of CT for planning the procedure helps avoid puncturing organs within the peritoneum (33). Immediate detection of pain during manipulation of an instrument alerts the operator to the possibility of a nerve injury. To minimize nerve root lacerations, local anesthetic is not used at the nerve roots. The center of the disk and the thecal sac can be located accurately with CT. The anulus fibrosus is not cut but is only traversed by the trephine, so suction of the anulus into the side port can be avoided. If the instrument is placed into the center of the disk, its blunt end remains there during the procedure. It does not cut in the forward direction, so trauma to vital structures is avoided. Compared with the percutaneous manual method, the procedure described by Onik et a1 is less time-consuming and the cannulas necessary to remove the material are much smaller. There is less danger of injury to or contamination of the surrounding soft tissue.
Seibel et a1
575
Volume 3 Number 3
Figure 5. Follow-up CT scan obtained after nucleotomy. The nucleotome has been removed. Scan shows a vacuum phenomenon indicating the area where the nucleus pulposus had been removed.
b. Figure 4. Percutaneous CT- and fluoroscopy-guided PNT. The tip of the nucleotome is seen in the middle of the intervertebral disk. (a)Position a t CT. (b)Position a t fluoroscopy.
Percutaneous diskectomy guided by means of fluoroscopy alone poses a potential danger of lacerating softtissue structures, as these cannot be identified at the time of the procedure. Prevention of nerve root and other injuries, as described by Epstein (29), involves avoidance of traversing nerves; the thecal sac; the center of the disk; and the medial border of the pedicIes, arteries, and veins. The ability to localize these
structures 1cm from the tip of the instrument is a distinct advantage of CT. The fluoroscopic unit in front of the gantry is useful for rapid identification of anteroposterior and lateral bone structures during puncture and of the instrument in the center of the intervertebral disk. The technique offers the advantage of an optimal placement of the nucleotome close to the herniated nucleus pulposus. Direct visualization of softtissue structures avoids the thecal sac, too. Therefore, injury of the cauda equina should be eliminated. The special difficulties of treating the disk at L5-S1 have been overcome by using curved instruments. In contrast to results reported in the literature (30,311, almost 50% of our patients had herniations at L5-S1 and were treated successfully. CT monitoring at L5-S1 is necessary to prevent nerve root injury. In about half the cases, the morphologic changes following removal of the herniated disk can be visualized on CT scans immediately following PNT (Fig 5). However symptoms may also disappear after PNT, even if only minimal change is demonstrated on CT scans (34). This may be due to the relief of pressure on the nerve. PNT procedures under the guidance of CT and fluoroscopy last
50-60 minutes. This includes a CT examination prior to and following the actual operation. The average time of aspiration was only 32 minutes, which is almost identical to the time the nucleotome is in place within the nucleus pulposus. This compares with the time described in the multi-institutional study of Onik and others (27,28). Their group reported that the procedure required less than 40 minutes to complete. This, however, does not include the CT examination before and following PNT. The risks of infection from contamination are not increased. Acquisition of additional CT scans increases the cost of the entire procedure. In the future, a shift to outpatient treatment may offset the additional cost of the CT portion of the procedure. References 1. Oppenheim H, Krause F. ijber einklemmung bzw: strangulation der cauda equina. Dtsch Med Wschr 1909; 35:697-708. 2. LasBgue C. Considerations sur la sciatique. Arch Gen Med 1864; 4:558-563. 3. Mixter WJ, Barr JS. Rupture of the intervertebral disc with involvement of the spinal canal. N Engl J Med 1934; 210-211. 4. Love JG. Removal of protruded intervertebral discs without laminectomy. Proc Staff Meet, Mayo Clin 1939; 14:800-807. 5. Yasargil MG. Microsurgical operation of herniated lumbar discs. In: Advances in neurosurgery. Berlin: Springer Verlag, 1977; 81-82. 6. El-Gindi S, Aref S, Salama M, Andrew J. Infection of intervertebral discs after operation. J Bone Joint Surg [Brl 1976; 58:114-116. 7. Stolke D, Sollmann WD, Seifert V. Intra- and postoperative complications in lumbar disc surgery. Spine 1989; 14:56-59. 8. Smith L. Enzyme dissolution of the nucleus pulposus in humans. JAMA 1964; 187:137-140. 9. McCulloch JA. Chemonucleolysis. J Bone Joint Surg [Br] 1977; 59:4552. 10. Postacchini F, Lami R, Massobrio M. Chemonucleolysis versus sur-
576
Journal of Vascular and Interventional Radiology
August 1992
gery in lumbar disc herniations: correlation of the results to preoperative clinical pattern and size of the herniation. Spine 1987; 12247-96. Edwards WC, Orme TJ, Orr-Edwards G. CT discography: prognostic value in the selection of patients for chemonucleolysis. Spine 1987; 12:792-795. Huckmann MS, Clark JW, McNeil TW. Chemonucleation and changes observed on lumbar MR scan: preliminary report. AJNR 1987; 8:l-6. Nelson KL, Runge VM. Lumbar spine. In: Runge VM, ed. Clinical magnetic resonance imaging. Philadelphia: Lippincott, 1990; 269-278. Hijikata S, Yamlagishi M, Nakayama T, Oomori K. Percutaneous discectomy: a new treatment method for lumbar disc herniation. J Toden Hosp 1975; 5:5-13. Hijikata S. Percutaneous nucleotomy: a new concept technique and 12 years experience. Clin Orthop 1989; 238:9-23. Friedman WA. Percutaneous discectomy: an alternative to chemonucleolysis? Neurosurgery 1983; 13: 542-547. Kambin P, Sampson S. Posterolatera1 percutaneous suction-excision of herniated lumbar intervertebral discs. Clin Orthop Re1 Res 1986; 207:37-43. Kambin P, Schaffer JL. Percutaneous lumbar discectomy: review of 100 patients and current practice. Clin Orthop 1989; 238:24-34. Suezawa Y, Jacob H. Percutaneous nucleotomy: an alternative to spinal surgery. Arch Orthop Trauma Surg 1986; 105:287-291.
Onik G, Helms CA, Ginsberg L, Hoaglund FT, Morris J . Percutaneous lumbar discectomy using a new aspiration probe. AJNR 1985; 153:290293. Bocchi L, Ferrata P, Passarello F, et al. La nucleoaspirazione second0 Onik nel trattamento dell'ernia discale lombare analisi multicentrica dei primi risultati su oltre 650 trattamenti. Riv Neuroradiol 1989; 2 (suppl l):119-122. Monteiro A, Lefevre R, Pieters G, Wilmet E. Lateral decompression of a pathological disc in the treatment of lumbar pain and sciatica. Clin Orthop 1989; 238:56-63. Schreiber A, Suezawa Y, Leu H. Does percutaneous nucleotomy with discigraphy replace conventional discectomy? Eight years of experience and results in treatment of herniated lumbar disc. Clin Orthop 1989; 238:35-42. Shepperd JAN, James SE, Leach AS. Percutaneous disc surgery. Clin Orthop 1989; 238:43-49. Wilson DH, Harbaugh R. Lumbar discectomy: a comparative study of microsurgical and standard technique. In: Hardy R, ed. Lumbar disc disease. New York: Raven, 1982; 147-156. Stern MN. Early experience with percutaneous lateral discectomy. Clin Orthop 1989; 238:50-55. Onik G, Mooney V, Maroon JC, et al. Automated percutaneous discectomy: a prospective multi-institutional study. Neurosurgery 1990; 26:228-233. Onik G, Helms CA. Automated percutaneous lumbar discectomy. AJR 1991; 156:531-538.
29. Epstein NE. Surgically confirmed cauda equina and nerve root injury following percutaneous discectomy at an outside institution: a case report. J Spinal Disorder 1990; 3:380383. 30. Seibel RMM, Gronemeyer DHW, Grumme TH. Neue verfahren der therapie degenerativer wirbelsaulenerkrankungen in der interventionellen radiologie. In: Gronemeyer DHW, Seibel RMM, eds. Interventionelle computertomographie. Vienna: Blackwell Scientific, 1989; 92130. 31. Seibel RMM, Gronemeyer DHW, Grumme TH. New methods of treatment of spinal column diseases using interventional radiological techniques. In: Seibel RMM, Gronemeyer DHW, eds. Interventional computed tomography. Boston: Blackwell Scientific 1990; 89-133. 32. Dilke TFW, Burry HC, Grahame R. Extradural corticosteroid injection in the management of lumbar nerve compression. Br Med J 1973; 2:635637. 33. Hopper K, Sherman J, Luethke J . Retrorenal colon in the supine and prone patient. Radiology 1987; 162: 443-446. 34. Bonneville JF, Runge M, Issaadi K, Tang YS, Guetarni S, Boulard D. Percutaneous discectomy using aspiration. Radiologie 1989; 9: 19951996.
