ACDF vs EndoscopePri_vs_MRNEndoscopic ACDF
APLDCentral SensitizationChemical RadiculitisNucleoplasty
Antibiotics for LBP discogel CMT 4 HNP Discography dangers?

| The Study | The Experiment | Results | Study Downfalls | Conclusions | Doug's Conclusions |

Milette PC, et al. “Radiating Pain to the Lower Extremities Caused by Lumbar Disk Rupture without Spinal Nerve Root Involvement.” AJNR Am J Neuroradiol 1995; 16:1605-1613

It is common knowledge that compressive disc herniation can result in both low back pain and radicular pain down the lower limb (sciatica) (43,252). What is less commonly known is the fact that radicular-like symptoms can also occur in the absents of compressive disc herniation (7); however, the exact mechanism for the latter is still controversial.


Milette et al. proposed that annular tears through the substance of the disc (aka: Internal Disc Disruption) may not only cause low back pain (discogenic pain) but also cause radicular like pain (discogenic sciatica) down the lower limbs. Others have also proposed this same theory (1,7,440).

In the mid 1990s, Milette et al. attempted to prove that structures from within the disc could act as pain-generators and were responsible for producing back pain as well as radiating lower extremity extremity pain.  

To prove this theory of ‘discogenic sciatica,’ the team first gathered a group of suspected IDD patients who all suffered both back and lower extremity pain yet had no signs of nerve root compression on MRI or CT. Next they pressurized the suspect discs via 'provocative discography' in an attempt to recreate the patient’s exact pains. Patients that were able to be provoked into their usual pain were placed into the final phase of the study. With the discography needle still in place, a powerful anesthetic (lidocaine) was injected into the center of the disc, in hopes that it would eradicate the patients back and lower extremity pain.

Their theory was this: if the patient's back and leg pains originated from structures within the disc (i.e., irritated sinuvertebral nerve fibers within the outer annulus), then the anesthetic would numb those structures and relieve the patient’s pain as it drained from the center of the disc, through the radial annular tear and into that outer pain-sensitive annulus. 


The results of the lidocaine intradiscal injection demonstrated that 76% of the patients obtained 75% to 100% relief of their back pain, and 94% of the patients obtained 75% to 100% relief of their low extremity pain!  Further more, all of the patients obtained at least 50% relief from their back and leg pain.   A control group that was injected with a placebo obtained no relief of pain at all.


Unfortunately, this study suffered some major flaws and technical difficulty that must bring into question the validity of the results; however, if we can believe the fact that the 1ml of anesthetic did not drain out of the disc through a grade V annular tear and numb any irritated posterior nerve roots, then this study certain yields evidence that both discogenic pain and discogenic lower extremity pain occur as the result of irritated structures within the disc. These structures are probably the Sinuvertebral nerve fibers that lace the outer 1/3 of the posterior annulus and posterior longitudinal ligament. Personally, I'm having a hard time believing that the lidocaine didn't ‘leak’ from the disc and contaminate the study. (see my comments under Dr. Gillard's Comments)


Over a three year period, Milette et al. gathered a group of 235 patients of whom were believed to be suffering from ‘discogenic’ back and discogenic radiating lower extremity pain.  

The entry criteria for this study were as follows: 1) they had to be suffering from both lower back pain and radiating lower extremity pain that traveled below the gluteal fold; 2) there could be no past history of lumbar surgery; 3) no spondylolisthesis could be present; and 4) no obvious nerve root compression could be present, as seen on MRI or CT.

A second set of criteria was next applied by to the dwindling group of patients: Provocation discography and CT discography (which are the only techniques we have for making the diagnosis of IDD (351-353,355)) would be applied to each patient.   For the patient to remain in the study group, the following criteria had to be met:   1) injection of only one disc had to recreate the patient’s typical lower back pain and lower extremity pain. 2) The pain intensity had to be at least a sustained ‘5’ on a scale of 0 to 10 where 10 is severe pain.  There was no mention made of throwing out Grade V annular tears (which of course leak injected contrast {or lydocaine} into the epidural space.

The reason that only one-level IDD patients were used in the study was that, methodologically speaking, they could only anesthetize one disc at a time.

Only 33 (14%) patients met the strict discographic criteria of this investigation; although, 74 bi-level IDD patients also had reproduction of both back and leg pain during provocation.   It’s unfortunate that these 74 patients couldn’t have been included in the study as well.   


Next, the center of each qualifying patient’s disc was injected with 1 milliliter of an antiseptic (lidocaine) in hopes of eradicating the patients back and leg pain.   After one minute, the patient was asked to grade their level of back pain reduction and leg pain reduction on a scale of 1 to 100, where 100 indicated complete pain relief.   Unfortunately, due to the fact that many of discs’ had become ‘supersaturated’ with contrast material, and would NOT accept the injection of anesthetic.   Sixteen more patients were lost since their discs couldn’t be tested.

After the experiment was over, all of the patients were given a steroid injection into the nucleus, annulus, and anterior epidural space.   The 2ml of depo-medrol was simultaneously injected as the needle was pulled out through the various regions of the disc and epidural space.

Below are the results for the 17 patients that were able to be injected with lidocaine:


17 Patients with single level discogenic pain:

Back Pain Relief following lidocaine:

Leg Pain Relief following lidocaine:




100% pain relief via VAS:

6 (35%)

10 (59%)

75% pain relief via VAS:

7 (41%)

6 (35%)

Good pain relief (>75%):

13 (76%)

16 (94%)

50% pain relief via VAS:

4 (24%)

1 (6%)

< 50% pain relief:



Ave. Pain relief via VAS:



Of the 17 patients with a single level painful disc, 100% of them got at least 50% relief of their back pain from the injection of lidocaine.   76% of the patients experienced 75% to 100% relief of their low back pain; 35% of them got 100% relief of their low back pain.  

*The most amazing statistic was that 16 of the 17 patients (94%) experienced a 75% to 100% relief of their lower limb pains.

Therefore, it appears that the intradiscal injection of an anesthetic was more effective at relieving lower limb pain (94%), than the lower back pain (76%).

A semi-control group (made-up from those 74 patients who fit the criteria with more than one disc) was injected with a placebo, which gave none of the patients any relief in pain.


94% (16 of 17) of the patients demonstrated a full thickness radial tear or a radial tear into its outer layers.   Only one patient had a tear that reach the “inner layers of the external part of the annulus.

*I wish they would have told us if any of those full thickness tears were leaking into the epidural space!!! That was a key flaw in this study.


  • No detailed description of the lower extremity pain!   Radiating thigh pain, leg, and foot were all grouped into two categories: right lower extremity pain or left lower extremity pain.
  • A very small study--17 patients
  • Not a very good control group (although it was certainly better than nothing).
  • Not very well written--lots of grammar errors.
  • I bet the incomplete annular torn discs’ were the ones that were supersaturated so they couldn’t hold any lidocaine, and were left out of the study.   On the other hand, the grade 5 discs, which probably had leaked contrast out of the disc and into the anterior epidural space, were able to take the injection of lidocaine and possible increased the probability that lidocaine also leaked into the anterior epidural space as well.   If lidocaine did get into the epidural space, it may have numbed any pain generation from irritated posterior nerve roots and/or dura.  

    The authors counter this possible ‘flaw’ in their study design by arguing that by limiting the lidocaine injection to only 1.0 ml, this ‘spillage-out-the-back-effect’ would not have occured, for there wasn’t enough lidocaine to make it this far.   They also noted that when the intradiscal and epidural steroid injection was given, only the anterior epidural space was ‘painful’ to the steroid and not the inside of the nucleus or annulus, which indicated that this region had not been numbed by extravagating lidocaine. (since the inside of the disc has no nerve-endings, this doesn’t surprise me – not the best argument)
  • Should have used the Dallas Discogram Classification system!


“The results of our study suggest that the pain sensitive structures responsible for the radiating pain to the lower extremity are located somewhere inside the disk, probably in the external part of the annulus fibrosus and in the longitudinal ligaments.” The existence of such sensory nerve fibers and endings has recently been demonstrated in rats (439), and it is reasonable to infer that they probably also exist in humans.” The typical low back pain itself is probably reproduced by the sudden increase in intradiscal pressure with stretching of these nerve endings as a result of the initial contrast injection (discography), rather than by irritation from the contrast medium.


This was a good attempt to prove the existence of discogenic back pain and discogenic sciatica; however, the experiment really failed due to the fact that almost one half of the group was lost to technical error, i.e., many patient discs were supersaturated with contrast and failed to take the experimentally-crucial injection of the anesthetic (lidocaine).   I worry that the remaining discs - that were able to be tested - were inflicted with grade 5 anular tears (via modified Dallas Discogram Scale) that allowed contrast to leak out of the disc and into the epidural space, hence making room for the upcoming injection of anesthetic.   If so, then the odds of the anesthetic also leaking from the disc and onto the posterior neural structures--numbing them as well as the sinuvertebral nerves of the posterior annulus--ruined the study by ‘numbing’ any pain-generating nerve roots or dura.

The authors say that because of the small amount of anesthetic used, it would be unlikely that any anesthetic had leaked from the disc and numbed any symptomatic nerve-roots, but there is just too much uncertainty for me to completely believe these results. Its really too bad that some of the partially torn discs (which were supersaturated) couldn't have been tested!     

The study also confirmed, once again, that discography is certainly not a perfect science: The chances of CT provocation discography confirming the diagnosis of discogenic back and leg pain are only about 50 percent when using the author’s strict entry criteria.   Specifically, of the 235 patients whom met the clinical diagnosis of discogenic back and leg pain, only 107 (45.5%) of them experienced both low back pain and lower extremity pain during provocation.  

The investigators confessed that reproducing the patients radiating lower limb pain was “frustratingly” difficult.   Because of this difficulty, they concluded that the lower extremity pain that often accompanies internally disrupted discs must be a “referred” type of pain and that “the mechanism for the production of this referred pain is not obvious.”

If we were to loosen their criteria and include the 95 patients who were excluded because provocation only reproduced low back pain WITHOUT the associated lower limb pain, then the accuracy of discography confirmation would rise to 86% (202/235).


By recreating a patients painful symptoms with an injection of contrast medium in a disk demonstrating an annular tear, and then relieving these symptoms by injecting a local anesthetic with a range of action that did not seem to extend beyond the disk margins, our study supports the existence of discogenic pain and leads to the conclusion that a simple disk rupture, without direct nerve root compression by disc material, can account for low back pain with radiating pain to the leg.   At these disc levels, CT may show no abnormality, or an unimpressive broad or narrow based bulge of the disc margin, whereas MR is likely to show, in addition, a partial loss of disk signal intensity on a T2-weighted image.  


1) Crock HV. “Internal disc disruption: A challenge to disc prolapse fifty years on.” Spine 1986; 11:650-653

7) Ohnmeiss DD, et al “Degree of disc disruption and lower extremity pain” Spine 1997; 22(14):1600-1665

43) Mixter WJ, Barr JS. “Rupture of the intervertebral disc with involvement of the spinal canal.” New Engl. J Med 1934; 211:210-215

252) Kuslich SD, et al. "The Tissue Origin of Low Back Pain and Sciatica: A report of pain response to tissue stimulation during operations on the lumbar spine using local anesthesia." Orthop Clinics North Am 1991 ;22 (2):181-187

439) McCarty PW, et al. “Immunohistochemical demonstration of sensory fibers and endings in lumbar intervertebral discs of the rat.” Spine 1991; 16:653-655.

440) King JS. “Sciatica viewed as a referred pain syndrome.” Surg Neurol 1976;5:46-50