Written 1996

Thoracolumbar Junction and Thoracolumbar Spinal pain syndromes

Jean-Yves Maigne, MD, Physical Medicine, Hotel-Dieu Hospital, 75181 Paris Cedex 04, France


Home page English Section / Home page sofmmoo.com

The thoracolumbar junction (TLJ) is comprised of the T10-11, T11-12 and T12-L1 motion segments. This transitional area, interposed between the thoracic and lumbar spine, is often the source of a characteristic pain syndrome characterized by a referral of the pain in the related dermatomes (T10 to L1).
Although Judovich and Bates [Judovich et al, 1950] were the first to report on low back and groin pain referred from the TLJ in 1952, this syndrome was fully described and studied by Robert Maigne in France, as early as 1974 [Maigne, 1974] [Maigne, 1980]. He coined the term "Thoracolumbar Junction Syndrome" and described the semiology and stressed the responsibility of the thoracolumbar facet joints. He emphasized the frequency of this syndrome among the low back pain sufferers and advocated treatment by spinal manipulation. He has authored fifteen papers on this topic, with the majority of these publications found in the French medical journals.

Course of the thoracolumbar nerves

Standard pattern

The thoracolumbar nerve roots divide into two rami, ventral and dorsal, after exiting the intervertebral foramen. We report here the findings of two previous studies dealing with the course of these nerves [Maigne et al, 1986], [Maigne et al, 1989].

a) The thoracolumbar ventral rami

The T10 and T11 ventral rami are intercostal nerves. They run under the ribs and end in the abdominal wall. They supply the intercostal and abdominal muscles as well as a strip of skin parallel to the ribs parallel to each vertebral level representing the dermatome. The T12 and L1 ventral rami are the subcostal and iliohypogastric nerves, respectively. They run parallel to the iliac crest. The subcostal and iliohypogastric nerves supply the lower muscles of the abdominal wall and the skin of the groin area. When they pass over the lateral aspect of the iliac crest, they give rise to a lateral cutaneous branch descending along the lateral surface of the hip. In most cases, this branch terminates at the level of the greater trochanter (Fig. 4). At times, it courses 5 to 10 centimeters distally. Interestingly, the branch of the iliohypogastric nerve becomes superficial by passing through a rigid fibro-osseous tunnel formed by the fibres of the muscle obliquus externus and the superior rim of the iliac crest. As seen in our own dissections, this orifice may occasionally entrap the nerve. The branch of the subcostal nerve passes through a purely muscular, weaker orifice.

b) The thoracolumbar dorsal primary rami

The thoracolumbar primary rami are of smaller diameter than the ventral rami. They are very short, dividing after a few millimeters into medial and lateral branches.
The medial branch runs dorsally along the angle between the transverse and zygapophyseal processes of the corresponding vertebra and gives off branches supplying the facet at that level. These nerves are very thin and very difficult to study macroscopically. A second inconsistent branch runs caudally to supply the facet at the level below. The medial branch then passes along the spinous process, supplying the periosteum of both the lamina and the spinous process prior to terminating at the tip of the latter. It also innervates the multifidus muscle, one or two levels caudal to their vertebral exit [Hayashi et al, 1992].
The lateral branch is directed caudally, laterally and dorsally, supplying the erector spinae and passing through the thoracolumbar fascia two to four levels caudal to their exit [Maigne et al, 1989], where it becomes superficial. This branch gives cutaneous innervation to the subcutaneous tissues of the lumbar and buttock area as distal as the greater trochanter in some cases (Fig. 6). The skin covering the sacral area is innervated by the S1 to S4 or S5 dorsal rami.
The most common pattern presents with the T12 branch lying laterally and the L1 branch medially as they traverse the iliac crest. At this level, the distance between the two branches varies from 1 to 5 centimeters. The L1 branch crosses the crest at a very consistent distance of seven centimeters from the midline, in all patients. Of particular interest is the fact that the L1 dorsal ramus becomes superficial by passing over the crest through a fibro-osseous tunnel formed by the thoracolumbar fascia and the superior rim of the iliac crest. This fibro-osseous tunnel is a rigid structure, which makes the nerve prone to compression. An entrapment neuropathy is thus possible at this level.


The iliac crest is usually traversed by 2 or 3 branches of the dorsal rami which supply the cutaneous region of the buttock. Data from our dissections indicate some anatomical variations in that the L1 dorsal ramus supplies a significant area of skin in 65% of the cases. In the remaining 35%, the L2 ramus, (occasionally receiving a contribution from L3 (10%)), is the major cutaneous branch for the buttock, although this relationship was variable. L4 and L5 have no significant dorsal cutaneous territory.

Pain patterns

The different pain patterns have been studied in healthy volunteers and in patients. The TLJ appears as a pain generator with a special feature: the radiation of the pain downward in the corresponding dermatomes. This was first illustrated by Kellgren, who injected the thoracolumbar interspinous ligaments in normal volunteers. He noticed a referral of the pain to the iliac crests [Kellgren ,1939]. McCall noticed the same type of referral when injecting the upper lumbar zygapophysial joints [McCall et al, 1979].
In patients, it is common to reproduce the radiation of the pain by needling or injecting around the facets of the TLJ.

Pain syndromes of the "thoracolumbar synddrome"

The Thoracolumbar Junction Syndrome (also known as the Maigne syndrome, as described by Robert Maigne) is defined by a dysfunction of the thoracolumbar junction (TLJ) referring pain in the whole or part of the territory of the corresponding dermatomes (eg. T11 to L1 or L2). Depending on the branch involved, the pain could refer to the low back (cutaneous dorsal rami), to the groin (subcostal or iliohypogastric nerve) or in the lateral aspect of the hip (lateral cutaneous rami of the subcostal or iliohypogastric nerve) (Fig. 7). All combinations of these clinical presentations are possible with one, two or three involved territories. Furthermore, even if the patient is unaware of symptoms in the above regions, the clinical examination may reveal tenderness of the cutaneous and subcutaneous tissues in the involved dermatome.

Low back pain

a) Symptoms

Low Back Pain is certainly the most frequently encountered pain complaint in the TLJ syndrome. The pain is distributed in the dermatomes of T11, T12, L1 or L2. Because the limits of these dermatomes are ill defined, due to overlapping and anastomosis, the pain is usually spread in the lateral part of the low back without corresponding exactly to a specific dermatome. Rarely, the pain is bilateral (the sacral area being pain free); more often, it is unilateral. Oddly, the right side is more commonly involved than the left one. In a personal series of one hundred cases, the pain was on the right in 62% of the cases (and 38% on the left). We have no explanation for this fact. Vernon?Roberts, studying the course of the degenerative process at the disc T12-L1, pointed out that the annular radial tears were twice as frequent on the right than on the left part of the annulus. He related this to the fact that most people were right?handed [Vernon-Roberts et al, 1995].
The pain is usually acute, of less than 2 or 3 months duration, often appearing after a false rotatory movement of the trunk, prolonged strenuous posture, lifting and occasionally, without any obvious precipitating factors. Repeated attacks are of course possible. Less commonly, the pain may have a more chronic course, but the disability is always less than that seen with lumbar pain syndromes. The pain is frequently increased by contralateral side bending, whereas ipsilateral side bending is pain free. This is likely due to the stretching of the cutaneous nerves, between the TLJ and the iliac crest where they are more or less fixed and appears to be fairly characteristic of the TLJ syndrome, as compared to the lumbosacral junction pain syndromes, where ipsilateral side bending increases the pain. This mechanism could be compared to a Lasegue sign of the trunk. Extension is often painful, as is ipsilateral rotation. Flexion is normal or painful, but without stiffness. This could be due to a mild protective guarding of the lumbar paraspinal muscles. Interestingly, this syndrome occurs more frequently after the sixth decade, as younger low back pain patients usually have a lumbar discogenic origin.

b) Clinical Signs in the Low Back

When examining the lumbar spine, one must always look carefully for two clinical signs. The first is the presence of a "posterior iliac crestal point", the second a positive “pinch-roll” test.

- The posterior iliac crestal point.

Pain and deep tenderness are located at the level of the iliac crest at a point which is consistently located seven centimeters from the midline. Pressure at this point causes a sharp pain similar to the patient's complaint (Fig. 8). The pain is usually excruciating, whereas the crest is much less painful even one centimeter left or right from this point. A small bony groove can be palpated at this level, corresponding to the passage of the cutaneous branch of L1. This sign requires careful and precise localization. This is facilitated by placing the patient in a forward flexed position, across the examining table, in order to open up the spine into flexion and gap the posterior elements. This is a very convenient and comfortable position to examine the spine from the TLJ to the sacrum, because the lordosis is reversed. The examiner places his (her) finger along the iliac crest with a moderate pressure every half centimeter, in an attempt to isolate the tender point. The examiner moves his (her) finger slightly laterally, medially and vertically in a probing manner. Once the irritated nerve is located, deep pressure and gentle movement produce marked tenderness, that is clearly demonstrated by the patient’s reaction. The opposite iliac crest is examined in a similar manner and is commonly unaffected.

- The pinch-roll test

Referred pain is accompanied by hyperalgesia of the skin and subcutaneous tissues in the involved dermatomes. This hyperalgesia or hypersensitivity can be revealed by gently grasping a fold of skin between the thumbs and forefingers, lifting it away from the trunk and rolling the subcutaneous surfaces against one another in a pinch and roll fashion. On the involved side the skin overlying the buttock and iliac crest is found to be tender when compared to the opposite side. This sign is difficult to elicit if the patient is obese or if the examination is hurried.

c) Examination of the thoracolumbar junction

Examination of the TLJ should be systematic in patients presenting with low back pain, especially when the pain is unilateral, located in the area of the iliac crest and buttock, and when an iliac crestal point is found.

- Clinical Examination of the thoracolumbar junction

The patient remains in the same position (lying across the examining table). We use two maneuvers. The first is longitudinal friction pressure over the facet joints; the second is lateral pressure against the spinous process.

Friction pressure over the facet joints. Pressure is applied deeply, and longitudinally approximately 1 cm lateral to the spinous process and is followed by a slow, gentle friction movement by the palpating finger. Tenderness is elicited over one or two joints ipsilateral to the lower back pain. Interestingly, clinical examination under fluoroscopic control has shown that the tender spot always corresponded to a facet joint, provided the palpation was slow and careful. T11?12 was the most frequently involved joint, followed by T10-11 and T12-L1. This is likely due to the orientation of the articular processes. T11-12 has a thoracic orientation in 58.6 % of the cases [Maigne et al, 1992] and the widest range of rotation : 5.2 +/? 2 degrees on each side as we had demonstrated it in a previous study using normal volunteers, positioned in rotation at the TLJ and then undergoing CT scan [Maigne et al, 1988]. This could lead to unusual stress and overuse of the joints at this level. In 40% of the cases, T11-12 has a lumbar, sagittal orientation, thus restricting the range of rotation at this level to 0.5 degree. T10-11 then undergoes the greatest amount of rotation instead of T11-12. When T10?11 is tender, it is often due to this transitional abnormality.

Pressure against the lateral aspect of the spinous process. The pressure is applied with the thumb slowly and tangentially at each level. This maneuver imparts rotation to the vertebra. Ipsilateral rotation is frequently tender at the involved level.

- Imaging of the thoracolumbar junction

Routine X-rays, CT or MRI scans of the TLJ are unremarkable in the majority of cases. They are often considered normal although, in anatomic studies, degenerative discs are frequent at this level. These imaging studies have no predictive value regarding the diagnosis or the response to treatment. In a previous study, we demonstrated the frequency of ossification at the attachments of the ligamentum flavum at this level, as compared to the others level of the thoracic spine, by using CT scan cuts. We hypothesized that this frequency was due to the high level of rotational strain involving this zone [Maigne et al, 1992].

- Establishing the diagnosis : the diagnostic block

Confirmation that the pain is referred from the TLJ to the iliac crest and the buttock can be demonstrated by an anesthetic block of the dorsal ramus and the painful facet joint. The needle is inserted in the centre of the tender spot overlying the facet, 1 to 1.5 centimeters from the midline. The point for injection is generally located on a horizontal line crossing the interspinous ligament. The needle is inserted until periosteum is contacted. After aspiration for blood, three cc. of anesthetic (Lidocaine 1%) are injected around the facet joint, and more laterally around the dorsal ramus. The cutaneous branch of the dorsal ramus can also be blocked as it crosses the iliac crest (iliac crestal point). The technique is easy : the needle is inserted at the level of the tender point and directed toward the superior rim of the crest. The anesthetic is then injected around the nerve. Whatever the technique, the injection should, within minutes, suppress the pain and discomfort previously produced by the patient's rotation, flexion or extension movements, and diminish the tenderness over the iliac crestal point, thus confirming the diagnosis.

Pain in the groin

a) Symptoms

Because it is acknowledged that the dermatomes covering the groin are T12 and L1 groin pain is easily related to a TLJ origin, provided that hip pathology has been ruled out. Groin pain may accompany low back pain or be an isolated complaint. The pain may sometimes be located above the groin, in the T10 or T11 dermatomes, depending on the involved level of the TLJ.

b) Clinical signs in the groin

Two clinical features are characteristic of the involvement of the ventral rami (subcostal and iliohypogastric nerves) : a positive pinch-roll test and tenderness over the superior aspect of the pubis.

- The pinch-roll test

The maneuver has been described above. The patient lies supine. The test has to be performed on both sides.

- Tenderness over the superior aspect of the pubis

Friction over the periosteum of the superior aspect of the pubis is tender on the involved side. This hypersensitivity is likely due to a lowering in the pain threshold.

c) Examination of the thoracolumbar junction

The examination of the TLJ is conducted as described above. The findings are basically the same. A diagnostic block may be useful for establishing the diagnosis. However, blocking the facet and the dorsal ramus is insufficient. It is mandatory to infiltrate around the ventral ramus as well. It can be easily done without fluoroscopic control, by inserting the needle as described above for the block of the dorsal ramus and by conducting it in a forward direction, toward the intervertebral foramen and the ventral ramus. The injection of three cc. of Lidocaine blocks the nerve and relieves the pain.

Pain over the lateral aspect of the hip

The third feature of the TLJ syndrome is pain over the lateral aspect of the hip. It is a referred pain in the territory of the lateral cutaneous branch of either the iliohypogastric, or the subcostal nerve.

a) Clinical signs in the lateral aspect of the hip

Referred pain in this territory is characterized by its distribution, a positive pinch-roll test and a lateral iliac crestal point. The lateral cutaneous branch usually reaches the trochanteric area, but can sometimes descend 5 to 10 cm distally. A shorter variety may be found, ending a few centimeters below the iliac crest [Maigne et al, 1986]. When the pain is referred in this area, the pinch-roll test is positive, as compared to the other side.
The lateral iliac crestal point has the same characteristics as the posterior iliac crestal point (see : low back pain of thoracolumbar origin). The crest has to be carefully palpated by the index or middle finger with the patient in the lateral decubitus position, painful side up, to reveal the tender point. It is located on the lateral part of the iliac crest, 10 to 13 cm from the anterosuperior iliac spine, at the intersection of the crest by a vertical line drawn from the greater trochanter. This location corresponds to the crossing of the iliac crest by the nerve, where a bony groove is often palpable.

b) Examination of the thoracolumbar junction

The examination of the TLJ is conducted as described above. The findings are basically the same.
As for both other pain syndromes, a diagnostic block may be useful for establishing the diagnosis. The block can be performed at the TLJ level by injecting around the ventral ramus, according to the same technique as described above. But it is also possible to block the nerve itself when it crosses the crest. One only has to needle the iliac crestal point, down to periosteum and to inject at this level and a little bit above the nerve. A positive response consists of resolution of symptoms within a few minutes.

Clinical aspects of the thoracolumbar junction syndrome

Each of the different pain syndromes characterizing the TLJ syndrome can appear in isolation or in combination in a given patient. Furthermore, when the pain complaint is isolated to only one of the regions, (often in only one precise area), the clinical examination may reveal a positive pinch-roll test in the other territories, an iliac crestal point, or tenderness over the superior aspect of the pubis, independent of the patient’s primary complaint. Another very common pattern is low back pain originating from the TLJ associated with pain emanating from the lower lumbar discs or facets.

Causes of the thoracolumbar junction syndrome

Causes at the TLJ

The most common cause of the TLJ syndrome is a minor intervertebral dysfunction at the TLJ (T10?11, T11?12 or T12?L1). One, two or three levels can be involved. Patients are often unaware of symptoms at the level of the TLJ. Pain is usually felt distally in the corresponding dermatomes. The nature of this dysfunction remains unknown, although the involvement of either the facets or the disc is very likely. More than any other part of the spine, the TLJ is involved in rotatory movements. At the lumbar levels, the total amount of rotation is limited because of the orientation of the facets in the sagittal plane. Above T10, despite a more favorable disposition, the rotation is restricted by the rib cage which is fixed to the thoracic spine. The major part of the rotation is thus concentrated between T10 and T12. This may lead to an overuse of the motion segment which could initiate disc or facet degeneration. On the other hand, the frequent facet asymmetry at T11?12 could disrupt the smooth rotation and initiate painful blockages or hypomotility in case of false motions (particularly if rotation is combined to forward flexion). Some other causes are possible, although very rare, such as a disc herniation or a collapse of the vertebral body of T11, 12 or L1 referring pain only in the low back.

Other causes

There are other possible causes for referred pain in the cutaneous nerves of T11, T12 or L1. Although the primary cause for pain may not be located at the TLJ, the symptomatology and the clinical signs are very near the TLJ syndrome. Moreover, both (TLJ and non TLJ) causes can be associated in a given patient and have thus to be diagnosed and treated simultaneously.

a) Nerve entrapment

- Entrapment of the cutaneous dorsal ramus of L1 [JY Maigne et al, 1989]

When the cutaneous dorsal rami crosse the iliac crest, the most medial among them (L1 in the majority of cases, sometimes L2) become superficial by perforating a rigid fibro-osseous tunnel formed by the thoracolumbar fascia above and the rim of the crest below. This orifice, always located 7 centimeters from the midline, may entrap the nerve, leading to pain in its cutaneous territory (Fig. 9). The clinical signs are very similar to those observed in the TLJ syndrome, except for the fact that the TLJ is normal to palpation. The major feature is the iliac crestal point, whose pressure reproduces the actual pain. The anesthetic block of this point must abolish all signs and symptoms to establish the diagnosis. We have at the present time a series of 21 patients with this syndrome. All were older patients (mean age : 67) and all were operated on (neurolysis), allowing a confirmation of the diagnosis and a prompt relief in the majority. None of them had hypesthesia in the territory of distribution of the nerve, possibly due to overlapping of dermatomes. It seems likely that this entrapment might be associated in many cases with a TLJ syndrome, thus reinforcing the symptoms in the low back.

- Entrapment of the lateral cutaneous branch of the iliohypogastric nerve [JY Maigne et al, 1986]

A similar arrangement may be observed for the lateral cutaneous branch of the iliohypogastric nerve (L1) which becomes superficial by passing through the same sort of fibro-osseous tunnel located at the intersection of the lateral part of the iliac crest and a vertical line passing over the greater trochanter (that is 10?13 cm from the anterosuperior iliac spine). This orifice is constituted by the rim of the crest below and the aponeurosis of the obliquus externus muscle above and may sometimes entrap the nerve (Fig. 10) . The pain is located on the crest and radiates downward, to the trochanter or even lower. The pinch-roll test is positive in the dermatome and the pressure over the lateral iliac crestal point reproduces the actual pain. Here too, the TLJ is normal to palpation.
This entrapment neuropathy is by far less frequent than the other one (cutaneous dorsal ramus of L1). In our current series, only six patients have required surgical decompression in five years.

b) New advances : Pain from the lumbosacral junction projecting in the TLJ dermatomes

Japanese authors have recently addressed the question why sciatic pain is often accompanied by a radiation in the groin area. They demonstrated (in the rat) a possible link between the L5?L6 disc and the L2 root. This disc could be partially innervated by dichotomizing sensory C?fibres present in the L2 spinal nerve in rats, or the higher levels, which also innervate the groin skin. [Takahashi et al, 1993]. Furthermore, by blocking the L2 spinal nerve in low back pain patients with degenerated lumbar discs, they were able to temporarily relieve the pain, that is to say that, according to their study, a discogenic pain from the lower lumbar spine could project in the L2 dermatome [Nakamura et al, 1995].

Treatments options for the thoracolumbar junction

The treatment of the TLJ syndrome is at first the treatment of the TLJ itself. Complementary treatments on the nerves may be done in case of failure.

Spinal manipulation

a) Definition and mechanism of action

The TLJ syndrome is particularly responsive to spinal manipulative therapy. Manipulation is a forced movement applied to a joint within the anatomic limits. This movement is characterized by a cracking sound due to a vacuum phenomenon as the facets separate. The vacuum phenomenon, or cavitation, makes the separation of the articular surfaces very sudden, even more so than the movement which initiated it. Thus, the cavitation appears as a motion accelerator, which could play a role by stretching hypertonic muscles. This is true not only for the TLJ but also for any part of the spine. The separation of the facets could also unblock the motion segment. Manipulation may also act on the disc. In a previous study using intradiscal pressure transducers into the lumbar discs, we demonstrated that the manipulative thrust initiated a sudden and temporary negative intradiscal pressure. This could alter the load transmission through the disc, thought to be one of the factors transforming a pain free degenerated disc into a painful one.

b) Technique

The first session is very important, because a good result is often obtained after one or two maneuvers, confirming the diagnosis. The most frequently used manipulative techniques are illustrated in figure X. One to five sessions are necessary to treat the patient, with one to four maneuvers in each session. If there is no improvement after the second one, the treatment and the diagnosis should be reevaluated.

Therapeutic injections

a) Technique

Facet injections are performed without fluoroscopic control, according to the technique described above for the anesthetic block above. We use a steroid, such as Hydrocortancyl*, 3 cc. In case of an obese person, or after a first failure, it is preferable to infiltrate under fluoroscopy. Injections are also the treatment of choice in cases of nerve entrapment at the iliac crest (posterior or lateral part). The needle is inserted in the centre of the iliac crestal point, until contact is made with periosteum. The needle is then directed upwards to the rim of the crest and around the nerve. Fluoroscopic guidance is not required.

b) Indications

The facet injections can be performed as a first attempt, especially in elderly patients, where osteoarthritis is likely, or after the failure of a first manipulative treatment. One or two injections are usually sufficient. A negative result should lead to reconsideration of the diagnosis. The same applies for injection of the cutaneous branch, which may be performed as a primary treatment when the TLJ appears normal or the iliac crestal point is very sensitive.


Legends for figures
Figure 5 : The lateral cutaneous branches of the subcostal (short arrows) and iliohypogastric (long arrows) nerves. Both nerves become superficial as they cross the iliac crest. The branch of the subcostal nerve passes through a muscular orifice (oblique muscles), the branch of the iliohypogastric nerve through a rigid fibro-osseous tunnel (superior arrow). Left lateral view. EI : Anterosuperior Iliac spine. T : Greater Trochanter.

Figure 6 : Dorsal cutaneous rami of T12 (1) and L1 (2) supplying the lower lumbar area. The medial nerve crosses the iliac crest (*) seven cm from the midline (arrow 2) Left side.

Figure 7 : Areas of pain and tenderness in the thoracolumbar junction syndrome.
a - Unilateral low back pain (Cutaneous dorsal rami of T10 to L1 or L2 roots).
b - Pain on the lateral aspect of the hip area (Lateral cutaneous branch of the subcostal and iliohypogastric nerves).
c - Pain on the groin area (Subcostal and iliohypogastric nerves).

Figure 8 : Iliac crestal point, always located seven cm from the midline.

Figure 9 : Left side. Superficial emergence of the dorsal cutaneous nerve of L1 though a fibro-osseous tunnel (arrow). FTL : thoracolumbar fascia. Dotted line = iliac crest.

Figure 10 : Left side. Superficial emergence of the lateral cutaneous branch of the iliohypogastric nerve though a fibro-osseous tunnel (arrow).