SURGICAL ANATOMY

Microsurgical Anatomy of the Lateral Approach to Extraforaminal Lumbar Disc Herniations

Department of Neurosurgery, Ludwig-Maximilians-University, Klinikum Grosshadern, Munich, Germany (H-JR, AM), and Department of Neurosurgery, Berne University Hospital, Berne, Switzerland (UE)

OBJECTIVE: During the "lateral" approach to extraforaminal lumbar disc herniations, the surgeon may be confronted with considerable variations in anatomy, making this approach extremely difficult in some patients. An anatomic study, therefore, was undertaken to examine the bony boundaries of the operative target, the medial intertransverse space.
METHODS: In 31 lumbar spine specimens taken from cadavers of people who had been between 30 and 93 years old at death, the relevant distances and proportions of the operative window were measured at the levels L1­L2 to L5­S1.
RESULTS: Measurements revealed that the operative window in a systematic fashion becomes progressively smaller as the approach moves from L1­L2 toward L5­S1: 1) from L1 to L5, the medial boundary, the isthmus laminae, gradually extends farther laterally and eventually covers the waist of the respective vertebral body; 2) the lower boundary, the facet joint, gradually overlaps the disc space in an upward and lateral direction; 3) the upper boundary, the transverse process, gradually moves downward. Anatomic variations and abnormalities are found particularly often at the L5­S1 level.
CONCLUSION: The anatomic findings led to important conclusions regarding the microsurgical approach to extraforaminal lumbar disc herniations; at levels L1­L2 to L3­L4, the midline approach with lateral retraction of the paraspinal muscles allows for efficient exposure of the lateral neural foramen and avoidance of trauma to the facet joint. Often at level L4­L5, and nearly always at level L5­S1, a tangential route through a paramedian transmuscular approach offers many advantages. (Neurosurgery 39:345­351, 1996)

Key words: Extraforaminal lumbar disc herniation, Microsurgical anatomy

Extraforaminal lumbar disc herniations (ELDH) constitute ~4 to 7% of all lumbar disc herniations (1, 3, 5, 14, 16, 19). Only 20% of all ELDHs are found at the level of the disc space; in the remaining cases, the free fragment has slipped upward and is found in the posterolateral waist of the respective superior vertebral body (Figs. 1 and 2) (8, 14, 15). The nerve root and ganglion are displaced cranially or dorsally and pressed against the lower edge of the pedicle. Downward migration of a fragment is prevented by the lower pedicle.

Between 1983 and 1987, three groups independently described the microsurgical "lateral" approach to ELDHs (2, 5, 18, 19), and, since that time, several other groups have advocated this approach or some variants of it (1, 4, 7, 9, 10, 12, 14, 15, 20, 22). The target of this lateral approach, the intertransverse process region, is well defined by several bony structures. While accumulating experience in managing ELDH in nearly 200 patients, we realized that those bony structures vary considerably between L2­L3 and L5­S1, depending on the level. The surgical approach may become complicated by such variations and, in particular, become extremely difficult at level L5­S1. A study, therefore, was undertaken to systematically analyze the anatomy of the bony structures of the lateral approach. From the resulting observations, clear implications can be derived for the surgical management of ELDH.

MATERIALS AND METHODS

RESULTS

The descriptions of three-dimensional anatomy that follow may best be understood by reading them with a specimen of the lumbar spine at hand. Average values and ranges (minimum and maximum values) for measurements in the 31 specimens are presented in Tables 1 and 2. Based on the average values for the 31 specimens, schematic dorsal views of the spine at levels L2­L4 and L4 and L5­S1 (Figs. 4 and 5) were constructed. For the height of the disc space, values of 4.5 mm (L2­L3), 4.5 mm (L3­L4), 5.5 mm (L4­L5), and 4.5 mm (L5­S1) were used (24). The site of pedicle origin, as analyzed in a previous study (17), is lightly outlined in the figures. The nerve root runs downward in close contact with and along the inner side and lower edge of the pedicle. At levels L1­L2 to L3­L4, the nerve root leaves the intervertebral canal in a slightly posteroanterior direction, but the root takes a more tortuous and anterior course in the lower lumbar spine (17). The outer aperture of the intervertebral canal is formed by the upper and lower pedicle, the dorsal aspect of the vertebral body, and the lateral edge of the interarticular portion of the lamina, the "isthmus laminae." From the surgeon's dorsal view, the root leaving the intervertebral canal becomes visible outside the isthmus laminae. In the clinical situation, it is hidden by the intertransverse muscle and ligament.

The "operative window" for the lateral approach to an ELDH is well defined by several bony structures (Figs. 3 and 4); medially, the approach is limited by the isthmus of the respective lamina, superiorly by the lower edge of the pedicle and the transverse process and (if present) by a prominent accessory process, and inferiorly by the apophyseal joint of the respective motion segment. Because the distances and proportions of those bony structures differ at the various levels, the surgical anatomy is presented separately for each of the most important levels.

Levels L1­L2 to L3­L4

Surgical anatomy at levels L1­L2 to L3­L4 is very similar, so those levels will be described together. In the upper lumbar spine, the pedicles arise more or less vertically from the dorsal aspect of the vertebral bodies (17) (Fig. 4) and the mean distance from the midline to the lateral margin of the isthmus laminae is 13, 14, and 15 mm, respectively, for L1­L2, L2­L3, and L3­L4. The intervertebral canal is correspondingly short (7.0­10 mm). The nerve root, after leaving the canal, courses a variable distance (4­11 mm) along the posterior aspect of the vertebral body, where it can be seen from the surgeon's dorsal view. The distance between the inferior margin of the transverse process and the superior boundary of the facet joint averages 10 mm (5­16 mm). The facet joint in our specimens partially or completely covered the intervertebral space, but overlapped it only rarely. Thus, the "operative window" at those levels is large, allowing the surgeon easy access to nerve root and herniated disc in most cases. The maximum and minimum values indicate some variations, which occasionally may render the operative situation more difficult.

Levels L4­L5 and L5­S1

In the lower lumbar spine, especially at level L5­S1, the operative window progressively becomes smaller (Fig. 5). The pedicle originates more obliquely (17), and the isthmus laminae is wider by an average of 18 to 22 mm (¾28 mm), which results in considerable lengthening of the intervertebral canal, ¾ 18 to 25 mm. The distance between the inferior border of the transverse process and the superior boundary of the apophyseal joint progressively decreases, from 7.9 mm (3­14 mm) at L4­L5 to 5.1 mm (0­11 mm) at L5­S1 (Tables 1 and 2). At those two levels, the facet joint may considerably overlap the intervertebral space, sometimes by 7 to 10 mm. At both levels, a prominent accessory process may further decrease the area of the operative window (Table 2).

In summary, our measurements show that the bony operative window for the approach to an ELDH gradually becomes smaller from L2­L3 to L5­S1 as a result of the following anatomic factors: 1) from L2 to L5, the medial boundary of the operative window, the isthmus laminae, gradually extends farther laterally and finally covers the waist of the respective vertebral body; 2) the transverse process gradually moves downward (17); and 3) from L1­L2 to L5­S1, the facet articulation gradually overlaps the disc space in a superior and lateral direction.

DISCUSSION

The approach most surgeons used until the mid-1980s to remove an ELDH was an upward and lateral enlargement of the interlaminar fenestration (3­6, 9, 11, 12). In the majority of patients, most, if not all, of the base of the inferior articular process had to be removed, leading to medial or complete facetectomy, to decompress the root (Figs. 1 and 2). To reach the medial margin of the ELDH, 10 to 15 mm of the base of the inferior articular process, together with the facet, must be removed in a lateral direction and 5 to 10 mm in a cranial direction. This base, however, has a width of only 12 to 18 mm (average, 14 mm) for vertebrae L1 to L4 (Fig. 4), and 12 to 23 mm (average, 17 mm) for vertebrae L5 and S1 (Fig. 5). This reflection clearly demonstrates that removing so much bone will result, in many patients, in the destruction of the base of the inferior articular process, or, at least, in thinning it so much that it will break when heavily loaded. Although the consequences of damage to, or resection of, the articular facets on one side are often considered negligible, there is now increasing agreement that partial resection of the facet joint on one side in conjunction with discectomy may, in 5 to 10% of such patients, be the source of a new disease, an irritable joint with persistent severe low back pain and pseudoradicular pain (13, 14).

For these reasons, the "lateral approach," after its first description (2, 5, 18, 19), has found many advocates (5, 10, 12, 15, 21, 22, 25). Essentially, the lateral approach to an ELDH targets the lateral margin of the isthmus laminae, in particular, the upper boundary of the apophyseal joint (19). One group of authors advocates a standard midline incision, dissection of the muscle insertions from the spinous processes and the facet joint, and retraction of the paraspinal muscles laterally until the isthmus, the facet articulation, and the base of the transverse process become visible. To visualize the disc space, resection of the cephalic and/or lateral portion of the apophyseal joint may be necessary (5, 10, 12, 15, 19, 21, 22). Others have found a more tangential route, either by paramedian muscle splitting (14, 18, 20) or even by retroperitoneal exposure (23), a more satisfactory way to approach an ELDH. Proponents of these tangential approaches claim that they minimize or avoid trauma to the facet joint. The results of the microanatomic study reported here may help to clarify why one or another approach may be more successful for operative management of ELDHs at various levels.

Levels L2­L3 and L3­L4

At the L2­L3 and L3­L4 levels, the isthmus laminae is small and the lateral edge of the ELDH has been found on average 30 to 32 mm from the midline (8); thus, a large part of an ELDH at these levels lies outside the isthmus (Figs. 1 and 4). At these levels, only a few millimeters of the isthmus need to be removed with a microdrill to identify the oblique course of the nerve root, the ganglion, and the disc fragment. The cephalocaudal distance that must be exposed between the inferior border of the transverse process and the superior rim of the facet joint, which was our operative window (c in Fig. 3B), is relatively large, averaging 10 mm. Even in patients with the minimum of 6 mm (Tables 1 and 2) between these two structures, the operative window is adequate.

Because the surgeon views the operative area dorsally, the disc space lies underneath the upper half of the facet joint (Figs. 1 and 4). The cranial rims of the facet joints L2­L3 and L3­L4 are, on average, at the level of the upper edge of the disc space, although they may lie from 4 mm below to 5 mm above this level.

The lateral boundary of the facet joint never reaches as far laterally as the vertical body of the L2­L3 or L3­L4 disc space. This means that in almost all patients, there is free access to the disc space, either from above or from a lateral direction. In our experience, resection of the cephalic or lateral portion of the apophyseal joint at the L2­L3 or L3­L4 levels is unnecessary. Recently, we have been removing only free fragments rather than clearing the disc space routinely, as we had been previously. We open the disc space laterally and clear it medially only in those rare cases in which there is a significant disc protrusion.

Level L4­L5

For the majority of patients with ELDH at the L4­L5 level, our operative approach is the same as for an ELDH at L2­L3. In some patients, however, the cephalocaudal dimension of our operative window may be as small as 6 mm, the lateral rim of the isthmus may be as far as 23 mm from the midline, and the facet joint may overlap the disc space cephalolaterally (Figs. 2 and 5). In addition, at this level, we may encounter a prominent accessory process (Figs. 2 and 5). In elderly patients, the situation may be further complicated by a degenerative reduction of the disc space and spondylotic appositions along the facet joint. Consequently, the facet joint moves upward and diminishes the operative window further.

How can we recognize this particular and, for the surgeon, unfavorable situation? The crucial factor, in our opinion, is the position of the facet joint relative to the intervertebral space. The lateral extent of the isthmus is less important, because a larger operative window can be created by removing part of the isthmus in an arcuate fashion, as indicated in Figure 2. If the facet joint overlaps the intervertebral space considerably in the cranial and lateral direction, then the standard midline approach (19) requires cranial and/or lateral resection, i.e., partial destruction of this articulation to unroof the root, fragment, and disc space. It is important to remember that the position of the apophyseal joint can be recognized before surgery, either on a high-quality plain x-ray film or on a computed tomographic or magnetic resonance imaging scan.

In this particular situation, the tangential approach through a paramedian transmuscular route presents obvious anatomic advantages, because it offers an oblique view of the neural foramen and intervertebral space below the facet joint. Facet joint integrity can usually be preserved with this technique (14, 18, 20).

Level L5­S1

A special, and sometimes extremely difficult to manage, situation may be encountered at level L5­S1:

1) The isthmus may reach as far lateral as 28 mm from the midline and sometimes even extend beyond the waist of the vertebral body (Fig. 6A). In such a case, the isthmus must be reduced until the nerve root and disc herniation become visible. At this level, the nerve root and ganglion lie in an intervertebral canal 20 to 25 mm long. A disc herniation, to be mainly extracanalicular, then must reach farther laterally than at any other lumbar level. Our measurements have shown that the lateral edge of an ELDH at L5­S1 can lie 32 to 44 mm from the midline (8). Intracanalicular disc herniations, probably because of the long intervertebral canal at L5­S1, are relatively more frequent than extracanalicular disc herniations at this level.

2) The cephalocaudal dimension of the operative window at L5­S1 is very short, sometimes as small as 0 to 3 mm. The reasons for this are that the facet joint overlaps the disc space considerably in a cephalad and lateral direction and that the transverse process is of a lower origin. In addition to these anatomic particularities, two variations of normal anatomy can be encountered at L5­S1. First, the transverse process of L5 usually is directed slightly upward, but in a small percentage of patients, it may be horizontal or may even bend downward. Downward bending of the transverse process thus narrows the cephalocaudal aperture significantly (Fig. 6B). Second, a similar situation occurs if the massa lateralis of the sacrum, instead of being in a horizontal plane, forms a steep slope. Such a sloped massa lateralis significantly reduces the operative window and conceals the course of the nerve root.

Occasionally, a patient will be encountered who has both abnormalities. Whether one or both abnormalities are present, good quality preoperative computed tomographic or magnetic resonance imaging studies will allow the surgeon to recognize most of these anatomic variations.

With experience, we have become convinced that, at level L5­S1, the paramedian transmuscular and tangential approach (14, 18, 20) has definite advantages, if destruction of the facet joint is to be avoided. The lateral extension of the isthmus, as well as the lower part of the transverse process, together with the lower part of the pedicle, can be removed with a microdrill (Fig. 2). The nerve root can then be carefully displaced upward, so that the disc fragment can be reached. This technique provides access to the nerve root, disc fragment, and interspace in almost all special situations.

A few remarks are necessary concerning the soft tissues found within the operative window. The intertransverse fascia and the lateral extension of the ligamentum flavum are removed only when drilling and bone resection are complete. It seems that the ganglion is very vulnerable and thermal (drilling, cautery) or mechanical trauma may lead to painful postoperative dysesthesia in the distribution of the affected nerve root (13).

CONCLUSION

Often at level L4­L5, and nearly always at level L5­S1, the operative window may become very tight, and the tangential route through a paramedian transmuscular approach offers many advantages. A preoperative plain x-ray film is recommended to study the various possible anatomic variations in advance and tailor the operative route to the anatomy of the individual patient.

ACKNOWLEDGMENTS

Received, November 11, 1995.
Accepted, February 12, 1996.
Reprint requests: Hans-Jürgen Reulen, M.D., Ludwig-Maximilians-University, Klinikum Grosshadern, D-81377 Munich, Germany.

REFERENCES

1. Abdullah AF, Wolber GH, Waefield JR, Krist Gunadi I: Surgical management of extreme lateral lumbar disc herniations: Review of 138 cases. Neurosurgery 22:648­653, 1988.
2. Ebeling U, Reulen HJ: Der laterale lumbale Bandscheibenvorfall. Nervenarzt 54:521­524, 1983.
3. Ebeling U, Mattle H, Reulen HJ: Der extrem laterale lumbale Bandscheibenvorfall: Häufigkeit, Symptomatik und Therapie. Nervenarzt 61:208­212, 1990.
4. Epstein EN, Epstein JA, Carras R, Hyman RA, Vishnubakhat SM: Far lateral lumbar disc herniation: Diagnosis and surgical management. Neuro-Orthopedics 1:37­44, 1986.
5. Fankhauser H, de Tribolet N: Extreme lateral lumbar disc herniation. Br J Neurosurg 1:111­129, 1987.
6. Garrido E, Connaughton PN: Unilateral facetectomy approach for lateral lumbar disc herniation. J Neurosurg 74:754­756, 1991.
7. Godersky JC, Erickson DL, Seljeskog EL: Extreme lateral disc herniation: Diagnosis by computed tomographic scanning. Neurosurgery 14:549­552, 1984.
8. Huber P, Reulen HJ: CT observations of the intra- and extracanalicular disc herniation. Acta Neurochir (Wien) 100:3­11, 1989.
9. Jackson RP, Glah JJ: Foraminal and extraforaminal disk herniation: Diagnosis and treatment. Spine 12:577­585, 1987.
10. Jane JA, Haworth CS, Broaddus WC, Lee JH, Malik J: A neurosurgical approach to far lateral disc herniation: Technical note. J Neurosurg 72:143­144, 1990.
11. Kornberg M: Extreme lateral lumbar disc herniations: Clinical syndrome and CT recognition. Spine 12:586­589, 1987.
12. Kunogi J, Hasue M: Diagnosis and operative treatment of intraforaminal and extraforaminal nerve root compression. Spine 16:1312­1320, 1991.
13. Markwalder TM, Reulen HJ: Translaminar screw fixation in lumbar spine pathology. Acta Neurochir (Wien) 99:58­60, 1989.
14. Maroon JC, Kopitnik TA, Schulhof LA, Abla A, Wilberger JE: Diagnosis and microsurgical approach to far-lateral disc herniation in the lumbar spine. J Neurosurg 72:378­382, 1990.
15. Melvill RL, Baxter BL: The intertransverse approach to extraforaminal disc protrusion in the lumbar spine. Spine 19:2707­2714, 1994.
16. Patrick BS: Extreme lateral ruptures of lumbar intervertebral discs. Surg Neurol 3:301­304, 1975.
17. Pfaundler S, Ebeling U, Reulen HJ: Pedicle origin and intervertebral compartment in the lumbar spine and upper sacral spine: A biometric study. Acta Neurochir (Wien) 97:158­167, 1989.
18. Recoules-Arche D: La chirurgie de la hernie discale du canal de conjugaison lombaire. Neurochirurgie 31:61­64, 1985.
19. Reulen HJ, Pfaundler S, Ebeling U: The lateral microsurgical approach to the extracanalicular lumbar disc herniation: A technical note. Acta Neurochir (Wien) 84:64­67, 1987.
20. Schlesinger SM, Fankhauser H, de Tribolet N: Microsurgical anatomy and operative technique for extreme lateral lumbar disc herniations. Acta Neurochir (Wien) 118:117­129, 1992.
21. Segnarbieux F, Van de Kelft E, Candon E, Bitoun J, Frèrebeau PH: Disco-computed tomography in extraforaminal and foraminal lumbar disc herniation: Influence on surgical approaches. Neurosurgery 34:643­647, 1994.
22. Siebner HJ, Faulhauer K: Frequency and specific surgical management of far lateral lumbar disc herniations. Acta Neurochir (Wien) 105:124­131, 1990.
23. Sturm PF, Armstrong GW, O'Neil DJ, Belanger JM: Far lateral lumbar disc herniation treated with anterolateral retroperitoneal approach: Report of two cases. Spine 17:363­365, 1992.
24. Tibrewal SB, Pearcy MJ: Lumbar intervertebral disc heights in normal subjects and patients with disc herniation. Spine 10:452­454, 1985.
25. Zindrick MR, Wiltse LL, Rauschning W: Disc herniations lateral to the intervertebral foramen, in White AH, Rothman RH, Ray CD (eds): Lumbar Spine Surgery. London, CV Mosby Co., 1987, pp 195­207.

COMMENTS

It is my personal preference to approach the disc herniation from the lower transverse process and proceed in a cephalad direction. At the level of the lower transverse process, it is easier to identify the lumbar disc. Moreover, the nerve root is separated farther from the spinal canal and there is greater safety because retraction of the nerve root is usually not needed. Once the disc is identified (and confirmed radiographically), it is easier to retrieve the disc herniation from the axilla of the root. Despite the many variations in approaching these lesions, the authors provide a helpful study.

Stewart B. Dunsker
Cincinnati, Ohio

Reulen et al. meticulously measured the dorsolateral bony surfaces, foramina, and processes of 31 cadaveric lumbar spines in an attempt to provide the reader with clinically relevant anatomic data to aid in the operative management of patients with extraforaminal lumbar disc herniations. They have succeeded. While the title of the article is misleading in that this text is not a review of the microsurgical anatomy of the lateral approach to a far-lateral disc herniation in the traditional sense (1), it does represent a thoughtful, original description of the relevant bony anatomy the surgeon faces in the approach to an extraforaminal disc herniation, sequestered or otherwise.

The authors documented a variety of bony dimensions of the dorsolateral human lumbar spine in Figures 4 and 5 and Tables 1 and 2 of the article that will serve as an anatomic reference data base. The less numerous and particularly clinically relevant measurements and dimensions are well described in the text. The text is useful because of the thoughtful correlation of these measurements with the authors' clinical experiences. I agree with the authors' insistence that significant facet disruption need not occur during microdiscectomy surgery, reducing the potential for postoperative spinal instability.

My decision-making process to employ one surgical approach over the other is similar to that espoused by the experienced authors. I too use preoperative radiographic studies to help decide which approach to employ. I favor a midline incision for lateral exposure at levels above L5­S1 and use both approaches at L5­S1, depending on individual patient anatomy. If the patient is referred for failure to improve after far lateral discectomy attempted via a medial, medial facetectomy-foraminal approach, then the paramedian posterolateral approach is used invariably. The lateral, sagittal magnetic resonance study is excellent for documenting the extent of cephalad migration of the far lateral HNP.

My philosophy and surgical technique differs slightly, however, from those of the authors. In my experience, the extreme cephalad free-fragment depicted by the authors in Figures 1 and 2 of the article is less common than the contiguous lateral and cephalad herniation at a degenerated interspace. For these latter patients, I instruct the residents and fellows to respect the bony and ligamentous structures of the facet capsules during dissection at the interspace in question but, when necessary (occasionally at L4­L5 and often a L5­S1), to shave the superolateral aspect of the facet complex in addition to the isthmus. I prefer to begin the extraforaminal disc dissection inferior to the dorsal and cephalad displaced root, at the inferior aspect of the interspace, rather than to begin directly downward on the nerve or even cephalad to it. In this fashion, facilitated by only a modest amount of lateral bony drilling if needed, the pathological finding is addressed and encountered first, not the compromised nerve root and ganglia.

Mark N. Hadley
Carl Lauryssen
Birmingham, Alabama

1. O'Brien MF, Peterson D, Crockard HA: A posterolateral microsurgical approach to extreme lateral lumbar disc herniation. J Neurosurg 83:636­640, 1995.

Reulen et al. meticulously studied the anatomy of the dorsal lumbar spine. Although we, as spine surgeons, profess to and indeed have, a significant knowledge of spine anatomy, it is evident that there is always more to learn.

The spine is a very complex structure, the nuances of which are brought to the forefront by researchers such as Reulen et al. The information they provide us is of extreme value when a surgical consideration for the management of lateral or far lateral lumbar disc herniation is entertained. The anatomic information provided by Reulen et al. should be incorporated into all spine surgeons' "cerebral data bank," so that we can develop management schemes for our patients more appropriately.

Edward C. Benzel
Albuquerque, New Mexico

Return to the Neurosurgery Home Page

© 1996 Williams & Wilkins