1Assistant professor, MM Institute of Nursing, India
2Specialist Pain International Clinic, Singapore
3Department of Anesthesiology, Radboud University Medical Centre, Netherlands
Submission: January 18, 2018; Published: February 26, 2018
*Corresponding author: Damian Ong, Department of Anaesthesiology, Intensive Care and Pain Medicine, Tan Tock Seng Hospital, 11 Jalan Tan Tock Seng, Singapore 308433, Tel: +65 98163454, Email: firstname.lastname@example.org
How to cite this article: Damian O, Nicholas H C, Kris V. Discogenic Low Back Pain: A Topical Review. Ortho & Rheum Open Access 2018; 10(4): 555795. DOI: 10.19080/OROAJ.2018.10.555795
Background: Discogenic low back pain (LBP) is a significant medical condition that carries a heavy socioeconomic burden. The pathophysiology, diagnosis and treatment modalities are reviewed in this article.
Methods: A review of the English literature addressing discogenic LBP in PubMed until 2015 is undertaken.
Results: The various methods of confirming the diagnosis of discogenic pain is elucidated in this article. With regards to the effectiveness of percutaneous treatments, intradiscal methylene blue (MB) and biacuplasty have a weakly positive recommendation. The evidence for intradiscal electrothermal therapy (IDET) is equivocal. Nucleoplasty, intradiscal pulsed radiofrequency (pRF) and rami communicans radiofrequency (RF) have too few studies to allow them to be assessed. Intradiscal RF, steroids, DiscTRODE and L2 block/RF have negative recommendations.
Conclusion: More evidence is needed for to allow for a better assessment of the effectiveness of treatment for discogenic pain. Future studies need to adhere to the IASP definition of discogenic pain.
Keywords: Low Back Pain; Intradiscal; Intradiscal Electrothermal Annuloplasty; Radiofrequency; Discogram; Back Pain
Chronic Low back pain (LBP) remains a significant medical problem and is a major cause of disability worldwide. In a study published in 2014, the global age-standardized point prevalence of LBP was 9.4%, with it being higher in men at 10.1% compared with women at 8.7% . The prevalence increases with age, peaking at 80 years old with a prevalence of up to 40% in males and 35% in females. In a retrospective chart review by Depalma et al. , the intervertebral disc contributed to 41.8%. The aim of this topical review is to give an overview on the pathophysiology, diagnosis and the evidence for the various treatment modalities for discogenic pain. A search in Pubmed for English articles up to 2015 was used to provide the literature contained in this review.
From the rationale that a structure must have a sensory nerve supply for it to be capable of causing pain, the innervation of the intervertebral disc has been extensively studied and documented. Bogduk described the innervation of the lumbar intervertebral disc in 1980 . The anterior and lateral portions of the annulus fibrosis (AF) are supplied by branches of the grey rami communicantes of the sympathetic trunk. The posterior aspect of the AF is supplied by the sinuvertebral nerve, which is
a combination of a branch from the ventral ramus and a branch of the grey ramus communicantes of the corresponding segment.
In a normal disc, only the outer one third of the AF is innervated. However, in patients with discogenic pain, it is found that there is innervation all the way into the inner third of the AF and even into the nucleus pulposus (NP) [4-6]. The tears in the AF, which are commonly in the posterior part, caused the formation of a vascularized granulation tissue starting from the outer part of the AF into the disc. Within this granulation tissue, nerve fibers are found and these abnormal nerves are the likely cause for the nociception of the intervertebral disc . In addition to the abnormal sensory innervation described above, inflammation within the intervertebral disc has also been implicated in the pathogenesis of discogenic pain [8,9].
According to the International Association for the Study of Pain (IASP), the definition of lumbar discogenic pain is lumbar spinal pain, with or without referred pain to the lower limb girdle or lower limb, stemming from a lumbar intervertebral disc .
The diagnosis hinges on the fact that the pain must be shown
to be conclusively from the intervertebral disc. This can be done
in 3 ways.
i. Selective anesthetization of the putatively symptomatic
intervertebral disc completely relieves the patient of the
ii. Selective anesthetization of the putatively symptomatic
intervertebral disc substantially relieves the patient of the
accustomed pain for a period consonant with the expected
duration of the local anaesthetic used.
iii. Provocation discography of the putatively symptomatic
intervertebral disc reproduces the patient’s accustomed
pain but provided that provocation of at least 2 adjacent
intervertebral discs clearly does not reproduce the patient’s
Of note, none of the above criteria is based on MRI findings.
Physiological aging of the intervertebral disc will appear
abnormal on MRI even though that disc is not symptomatic
. As such, an MRI by itself cannot be used to diagnose the
symptomatic disc and the above interventional diagnostic
procedures are indicated. Though the IASP states that selective
anaesthesia of the disc is a valid method of diagnosing discogenic
pain, only a few studies attempt to use selective blockade of the
disc [12-17] as part of their inclusion criteria.
The most commonly used method of diagnosing discogenic
LBP hinges on a positive provocative discography. The procedure
is briefly outlined and the criteria for a diagnosis of discogenic
LBP are highlighted. The Spine Intervention Society (SIS),
formerly known as the International Spine Intervention Society
(ISIS), recommends the following criteria to diagnose discogenic
pain during provocative discography .
i. Provocation of the suspect disc induces concordant
ii. The pain is at least 7/10 on a numeric rating scale.
iii. The pain is provoked by less than 50 psi over opening
iv. Provocation of at least 1 adjacent disc does not induce
The rate of infusion should not exceed 0.05 ml/s as this
allows for uniform pressurization within the disc. If too high
an infusion rate is used, false positives may occur because of
the pressure peaks. These pressure peaks may cause vertebral
end plate compression and distention of the adjacent facet joint
. The patient should be blinded as to which disc is being
stimulated and the disc which is expected to be most painful
should be stimulated last.
During provocative discography, the following events should
i. Opening pressure (OP) which is defined as the pressure
in which contrast is first seen entering the disc.
ii. Provocation pressure which is defined as the pressure
greater than the OP in which the patient first complains of
iii. The peak pressure or final pressure at the end of the
procedure which is greater than the OP.
The procedure is terminated if the following is reached.
a. Concordant pain is reproduced to a level of 7 out of 10
on a numeric rating scale
b. The volume injected reaches 3 ml (or up to 4 ml in very
degenerated disc in which the pressure remains less than
c. The pressure rises to 50 psi above OP in discs with a
grade 3 annular tear
d. If contrast leaks through the outer annulus fibrosis or
through the endplates in which it might not be possible to
pressurize the disc sufficiently to test its sensitivity.
The first study to use selective blockade of the disc was
performed in 1996 . 1.5-2 milliliters of a mixture of two
milliliters of 2% lignocaine and 1 ml of contrast media were
injected into the disc and pain relief was assessed 30 minutes
later. Pain relief of >50% was judged to be a positive test for
discogenic LBP. The same group of authors then used the
same criteria in a subsequent study in 2001 . The term
“Discoblock” was first used by Ohtori . 0.75ml of 0.5%
bupivacaine was injected into the intervertebral disc and if pain
was reduced thereafter, the diagnosis of discogenic pain was
made. This method was then compared to that of conventional
provocative discography in patients who underwent surgery
based on the diagnosis of discogenic pain obtained from the 2
different diagnostic methods. They found that patients whose
diagnosis was obtained from Discoblock were more satisfied
2 studies from the same authors on intradiscal pulsed
radiofrequency (pRF) used Discoblock in conjunction with
provocative discography as part of their inclusion criteria [15,16].
They required a 70% reduction in pain for it to be positive. A
modification of this technique is used by Alamin et al.  to
confirm the diagnosis obtained from provocative discography.
In patients who had a positive provocative discography, they
inserted a catheter into the nucleus pulposus of the symptomatic
disc. They then randomized the patient into either an infiltration
of 0.6ml of 4% lignocaine or normal saline via the catheter. The
results of the 2nd part of the test were then compared with the initial provocative discography findings and the results of the
functional disc block and provocative discography were different
in 46% of the patients. This is the only study that has used this
protocol and in the absence of a gold standard diagnostic test, it
is difficult to comment on its utility as compared to provocative
ISteroids were introduced into the disc to decrease the
inflammatory reaction in patients with discogenic pain. As early
as the 1960s, clinicians were already introducing steroids into
the disc . The last 3 studies published were randomized
controlled trials (RCT) and these will be reviewed in this article.
In 1992, Simmons et al conducted a randomized double blind
active control study (Jadad scoring = 3, no mention of method
of randomization or blinding) . Either methylprednisolone
80mg (14 patients) or Bupivacaine 7.5mg (11 patients) was
injected intradiscally after a positive single-level provocative
discogram. A follow-up of only 10-14 days showed that 21%
of patients who had methylprednisolone injected versus 9% of
bupivacaine reported improvement. There was no statistical
superiority in the steroid group.
The RCT by Khot et al. in 2004 (Jadad score = 4, method
of blinding inadequate) compared methylprednisolone 40mg
against 1ml of normal saline with 60 patients in each group. Only
the patient was blinded and follow-up was for 1 year. There was
a high loss to FU rate of 14 patients in the steroid group and 8
patients in the saline group. The results of this trial showed no
difference between the 2 groups in terms of disability or pain
scores. There was a median change of 0 in pain scores in both
Peng et al.  published the latest RCT in 2011 in which
they looked at patients with positive provocative discography
and end plate Modic changes (Jadad score = 4, method of
randomization not stated). They divided the group into 2 based
on Modic change Type I or II then subsequently randomized them
into 3 treatment groups. The first group received intradiscal
saline, the 2nd group intradiscal betamethasone, and the 3rd
group intradiscal betamethasone and songmeilie (Chinese
herbal medicinal ingredient that reduces inflammation). At
6-month follow-up, in both types of end plate Modic changes, the
intradiscal betamethasone and intradiscal betamethasone with
songmeilie groups had significant drops in VAS and ODI scores.
In contrast to the previous RCTs, intradiscal steroids showed
some short-term benefit in patients with discogenic pain and
end plate inflammatory changes. This has been suggested in
another study .
Based on the above RCTs, there is a limited role of intradiscal
steroids for patients with discogenic LBP. Older observational
studies were not included in this review. Based on the results of
the RCTs, only in patients with inflammatory end plate changes
would intradiscal steroids have a possibility of short-term
benefit and this still needs further clarification with more trials.
The IDET procedure utilizes a navigable intradiscal catheter
(Spine CATH, Oratec Interventions Inc., Menlo Park, CA, USA)
with a 6-cm active tip that is introduced through a 17G introducer.
The 17G introducer is inserted into the disc in a similar fashion
to performing a discography (Table 1). The flexible catheter is
then inserted and it directed to make a loop such that the active
tip lies against the posterior AF. In which layer the catheter lies
exactly is a matter of debate as some studies state it should lie in
the NP-AF junction whereas others say that it should lie within
the AF itself [24-26]. It is then postulated that the RF current
which is applied across the catheter causes heating of the active
tip which ablates the aberrant nociceptive nerve endings in the
In the first observational study, Saal et al.  reported that
80% of the patients had at least a 2 point reduction on VAS with
a mean reduction of 3.74 at 7 months follow-up. There were
also improvements in analgesic use, SF-36 scores and sitting
tolerance. Multiple prospective studies were then published and
selected studies are presented in Table 1. Outcomes at 2 years
were also published [25,28] and these were in favour of IDET
with 57-72% having a decrease of at least a 2-point decrease in
VAS at 2 years. However, several studies showed disappointing
results with IDET [29-31] and so there was a need for randomized
studies to be performed to better ascertain the utility of IDET.
The first RCT was performed by Pauza et al. . In this
well-conducted study, patients either had IDET treatment or
a sham procedure. It must be noted that the inclusion criteria
for this study required patients to only have a loss of <20%
of disc height. The results of the study showed that both the
placebo and IDET group improved at 6 months but more marked
improvement was seen in the IDET group (56% vs. 38%). IDET
was also more effective in patients who had poorer function preprocedure.
Even in this select group of patients with relatively
preserved disc height and reasonable function, almost 50% did
not benefit from IDET.
The results from a 2nd RCT were even less impressive.
Freedman et al.  conducted a placebo controlled trial in
which placebo consisted of inserting the introducer needle and
the catheter into treatment position but not heating the catheter.
They included patients with up to 50% loss in disc height and
patients on worker’s compensation. Compared to the Pauza
study, the patients had poorer ODI and SF-36 scores. In this
study, both the IDET and placebo groups had no improvement at
6 months compared to baseline. It should be noted that in both
the RCTs on IDET, the targeted sample size for 80% power were
Complications stated in the literature includes cauda equina
syndrome [30,34,35], increased disc herniation [30,36,37],
vertebral osteonecrosis [38,39], a broken catheter that migrated
intradurally causing radiculopathy , nerve root injury
[30,37] anterolisthesis and discitism . Though severe
complications can occur, the overall rate of complications is
low. Based on multiple observational studies that show positive
results, an RCT that showed fair results in selected patients and a
RCT that showed a negative result, there is still clinical equipoise
over this modality.
The discTRODE (Radionics, Burlington, MA, USA) is also
based on a similar working principle as the IDET. First, a 17G
curved and electrically insulated introducer needle is inserted
into outer annulus from the contra-lateral side to the annular
tear. Electrical impedance is checked and the needle is further
advanced in the annulus until an impedance of 300-400Ω is
obtained, which usually corresponds to that of a mid-annulus
depth. The catheter is then inserted and it is navigated through
the posterior annulus and upon contacting the lateral wall of the
annulus, it travels anteriorly. As such, the electrode covers the
posterolateral portion of the AF. The catheter is then heated to a
maximum temperature of 65ºC for 10 minutes.
The first published study on DiscTRODE was in 2005 by Finch
et al. . 31 patients underwent the procedure and another
15 who were refused funding became the control group. At
12-month follow-up, VAS scores had dropped 37% in the treated
group with a 3% increase in the control group. 9 out of 31 patients
had >50% drop in VAS scores. The ODI decreased significantly
over the same period in treated patients. Interestingly, though
VAS and ODI scores dropped in the treatment group, medication
use did not decrease accordingly.
A comparative non-randomized study between IDET and
DiscTRODE was published in 2005 by Kapural et al. . 21
patients underwent the DiscTRODE procedure, in which the
heating protocol consists of 55ºC for 4 minutes, 60ºC for 5
minutes then 65ºC for 5 minutes. Another 28 patients underwent
IDET and the patients in both groups were matched against
each other. Patients in the DiscTRODE group had a decrease in
VAS from 6.6 to 4.4 at 1-year follow-up but patients in the IDET
group had a larger drop of 7.4 to 1.4. Though DiscTRODE had a
significant drop, it was much smaller compared to IDET. 81% of
patients in the IDET group had a 60% improvement in VAS as
compared to 29% in the DiscTRODE group. As such, based on
the results of this study, DiscTRODE is inferior to IDET for the
treatment of discogenic LBP.
A well conducted (Jadad score = 5) randomized, double
blinded, placebo controlled trial was then conducted by
Kvarstein et al. . The placebo was a sham procedure in which
the introducer needle and electrode were both inserted into
the appropriate position but the RF current was not delivered.
In the treatment group, incremental heating started at 50ºC
and this was increased by 5ºC every 2 minutes. At 65ºC, the
heating was maintained for 4 minutes. The authors intended to
have 25 patients in each group but planned an interim analysis when 10 patients in each group had reached 6-month follow-up.
Unfortunately, the interim analysis did not show any significant
benefit to the patients who underwent DiscTRODE and so the
study was discontinued. As such, this study did not attain the
predetermined sample size and analysis of this study was based
only on the 20 patients recruited. The VAS scores between the
sham and treated group were not different. The treatment group
had only a slight reduction of 1 compared to 0 in the sham group.
There were more patients in the sham group who reported a 2 or
greater decrease in VAS. In terms of complications, there is a case
report of Complex Regional Pain Syndrome after the DiscTRODE
procedure  but nothing else can be found in the literature.
As such, based on a well-conducted underpowered RCT
which showed no significant difference between DiscTRODE and
sham procedure, a comparative study that showed inferiority
of DiscTRODE to IDET and a positive prospective study, the
evidence is not supportive of DiscTRODE for the treatment of
Intradiscal RF first appeared in the literature in 1996. The
hypothesis for its effectiveness in treating discogenic pain was
that the heat produced would destroy the nerves in the AF.
The intervertebral disc has poor internal circulation and the
vertebral endplates act as insulators to trap the heat generated
by the RF current within the disc . The procedure consists of
inserting a cannula with a 10mm active tip into the centre of the
NP. An RF electrode is then inserted and a lesion is made. The
studies are presented in Table 2.
The first single cohort prospective observational trial was
published by van Kleef et al. . After positive anesthetic
discography, the patients had RF of the intervertebral disc to
achieve a temperature of 70°C for 90 seconds. This study showed
positive results with 70% of patients having pain reduction
at 8 weeks and 55% on longer follow up (mean of 16 months)
if they have not had previous surgery on the affected disc.
The results for patients who have had surgery on the affected
intervertebral disc were worse with only 37% of patients having
pain relief at 8 weeks and 27% on longer follow-up. A good
quality RCT was then conducted based on the positive results
from the previous prospective study . Barendse et al. 
randomized 28 patients with positive anaesthetic discography
into a treatment group which used the same settings for RF as
above, and a control group which had a needle inserted into the
intervertebral disc but no RF applied. In contrast to the previous
study, only 2 patients in the control group and 1 patient in the
treatment group had a positive response at 8 weeks. By 1 year,
only 1 patient in the treatment group still had a successful
outcome. This RCT showed no difference between the treatment
group and the control group.
Another group conducted a RCT of moderate quality in 2003.
Ercelen et al.  randomized 39 patients into 2 groups, the first
of which had RF delivered to obtain 80°C for 120 seconds and
the second group had it maintained for 360 seconds. At 6-month
follow-up there was no difference between the 2 groups. Both
the groups had significant pain reduction up to 1-month post
procedure but they returned to pre-treatment levels by 6
months. With regards to complications, there was 1 case of
discitis reported in the trial by Ercelen et al. .
In summary, there are 2 RCTs which showed that intradiscal
RF has no long-term utility in the treatment of discogenic pain.
Both the RCTs did not show sample size calculation and so it is
uncertain if the lack of efficacy of intra-discal RF is due to a lack
of power in the trials or that it is truly ineffective. It is unlikely
that any more trials would be conducted for this modality as
results from both the RCTs were disappointing. In addition,
IDET and biacuplasty delivers a targeted thermocoagulation of
the posterior AF as compared to intradiscal RF which should
theoretically produce better results. Based on current evidence,
there is no role for intradiscal RF.
pRF has been gaining popularity in recent years and several
studies has used it intradiscally for discogenic pain (Table 3).
The highest temperature attained with pRF is limited to 42°C
and so thermal destruction of the aberrant nociceptive nerves
in the disc does not occur with pRF. The most likely pathway for
the therapeutic effect is that the electric field generated induces
changes in the nerves targeted and so generate a long term
analgesic effect . The technique of inserting the cannula is
similar to that for provocative discography. An RF electrode is
then inserted and the current delivered. Teixeira et al.  was
the first to publish an article on the use of pRF intradiscally.
There were significant decreases in NRS in the 8 patients in his
study and at least a 4-point drop in NRS at 3 months.
Jung YJ et al.  then used intradiscal pRF in 26 patients .
More than half the patients had multiple levels treated. There
were significant but small drops in VAS and ODI at 12-month
follow-up. Only 35% of the patients had a reduction of >50% in
VAS at 12 months though. A relatively large prospective study of
76 patients was done by Rohof . At the 3-month follow-up
after intradiscal pRF, patients whose pain reduction was <50%
may have additional diagnostic and therapeutic interventions
like pRF of the dorsal root ganglion or a RF ablation of the medial
branches of the spinal nerves. At the 12-month follow-up, 56.5%
had a reduction of >50% in NRS and this includes the group
that just had intradiscal pRF alone and those in whom multiple
procedures had been done. For the patients who only had
intradiscal pRF, only 30% had a successful outcome at 1 year.
A comparative study of IDET against intradiscal pRF was
then done by Fukui et al. . The first group of patients
underwent IDET from 2003 to 2009, then the 2nd group of
patients underwent intradiscal pRF from 2009 to 2011. The
inclusion and exclusion criteria for the 2 groups were the same.
This study showed that pRF was comparable to IDET with
significant decreases in NRS and in the Roland-Morris Disability
Questionnaires. The same group of authors then performed
intradiscal pRF for a prospective study of 23 patients .
82.6% had at least a drop of 2 in the NRS, with 65.2% having
>50% reduction at 12-month follow-up.
In summary, there is limited evidence for the efficacy of
intradiscal pRF. Only prospective observational studies have
been done and the results differ substantially between studies.
More studies are needed and at the current time intradiscal pRF
cannot be routinely recommended.
There has only been 1 study published for this technique and
the aim of the technique is to disrupt the nociceptive input from
the AF by creating a thermal lesion in the rami communicans.
In 2004, Oh et al.  published a RCT (Jadad score = 2)
comparing radiofrequency ablation of the rami communicans
above and below to the pathological disc versus a lignocaine
2% infiltration. No blinding or method of randomization was
mentioned in the article. The study group consisted of patients
who had a poor outcome following IDET at a single disc after a
positive provocative discography. A diagnostic block of the rami
communicans is first performed. If pain decreases >50%, the
RF procedure is repeated one day later. A curved needle with a
10mm active tip is directed to the inferior third of the vertebral
body. The needle should then lie in the postero-lateral 1/3 of the
vertebral body where the rami communicans runs. The control
group receives 2 millilitres of 1% lignocaine but the treatment
group gets RF thermocoagulation at 65ºC for 60 seconds. 26
patients were in the treatment group and 23 patients in the
control group. All patients completed 4 months of follow-up.The patients in the treatment group had significantly better
improvements in the VAS and SF-36 scores. Mean VAS dropped
from 7.1 to 3.8 in the treatment group.
Though the results from this RCT were positive, it was not
well conducted and the inclusion criteria of patients who failed
IDET are not generalizable to all patients with discogenic pain.
Another possible route of nociceptive transmission of
a painful intervertebral disc back to the spinal cord is via the
lumbar sympathetic chain. It is postulated that nociceptive
fibers from the AF of the lower lumbar intervertebral discs run
through the rami communicans to the lumbar sympathetic chain.
It then ascends the sympathetic chain, travels via the L2 rami
communicans which then joins the ventral root of the L2 spinal
nerve before reaching the spinal cord. Nakamura et al.  were
the first group to investigate this pathway in 1996. 33 patients
with LBP of at least 1 month and diagnosed with discogenic pain
based only on physical examination, plain radiographs and MRI
had a unilateral L2 nerve block with 1.5ml of 1.5% lignocaine.
After 15 minutes, VAS scores (max of 20 points) and pain
provocation were assessed. 26 patients had complete relief of
ipsilateral pain and the remaining 7 had some relief. Mean VAS
scores dropped from 10 to 1.7. The average duration of pain
relief was 20.7 days.
Simopoulos et al.  then reported on a case series of 5
patients who underwent RF of the L2 rami communicans with
good results. Patients had to have both a positive provocative
discography and greater than 50% pain relief from diagnostic
L2 rami communicans block to be considered for a RF treatment.
This consisted of reaching a temperature of 80ºC for one minute.
There was improvement in the VAS scores, sitting tolerance and
medication use. However, the pain relief only lasted for about
4 months following which the procedure had to be repeated to
maintain the pain relief.
A negative prospective cohort study was then published by
Richardson et al. . Following positive manual provocative
discography, 12 patients underwent bilateral L1 and L2 dorsal
root ganglion (DRG) blocks using methylprednisolone 80mg,
clonidine 75 mcg and 4 ml of 0.5% bupivacaine in total. After
1 month, there were no differences in the pain scores and pain
interference scores regarding daily activities pre- and postprocedure.
The study was thus terminated. No complications
were reported in any of the above studies. Conflicting results
are seen in the studies on L2 nerve blockade and the numbers
of patients are small in each study. It is uncertain if there is
any utility in treating discogenic LBP with a L2 nerve root
intervention. It is likely that the sympathetic chain is not the
main route for the transmission of discogenic pain and that
the traditional route via the cauda equina contains most of the
nociceptive pathways. As such, there is no role for it in routine
The Spine Wand (Arthrocare Inc., Sunnyvale, CA, USA)
utilizes coblation technology to decrease intradiscal pressure
by removing a small amount of volume from the NP . The
coblation is meant to result in a decrease of about 1 millilitre of
NP volume and thus causing a subsequent large fall in intradiscal
pressure. Though it was initially used for radicular pain
secondary to a herniated disc, it was also utilized in patients
with discogenic LBP.
Only observational studies are found for the use of
nucleoplasty in discogenic LBP (Table 4). Sharps et al.  first
published this in 2002 where they had positive results up to 1
year in duration. Singh et al then published 3 prospective studies
where all showed positive results with about 50% of patients
obtaining greater than 50% pain relief at 1 year [56-58]. Reddy
et al.  published a retrospective analysis of 49 patients with
a mix of axial and radicular pain. The criteria for inclusion for
axial back pain seemed to suggest that a positive provocative
discography is not a requirement. As such, this study was not
included in the Table. Similarly, the study by Masala et al. 
had similar problems in their study and was not included in the
table. Both these studies though had positive results. Another
study by Kumar et al.  had positive results but concordant
pain during discography was not an inclusion criteria though it
was performed prior to nucleoplasty. They found that concordant
pain had no predictive effect on the success of the procedure.
Lastly, He et al.  published an interesting study in
which they modified the technique of nucleoplasty and named
it coblation annuloplasty. Instead of creating channels in the
NP, they purposely left the introducer needle at the edge of the
outer annulus, and created 6 channels in the annulus instead.
The selection criteria were good as only patients with discogenic
LBP confirmed by provocative discography was included. In
17 patients, the mean VAS scores dropped from 6.5 to 3.2 at
6-month follow-up. In addition, 58.8% of patients had greater
than 50% pain relief.
The complications of nucleoplasty have been presented
in another article  and it includes increased back pain,
increased radicular pain, new onset transient lower limb
neurological symptoms and epidural fibrosis. No complications
were reported in the articles for discogenic pain. In summary,
based only on observational studies in which some did not
diagnose discogenic pain per IASP criteria before the procedure,
nucleoplasty showed a consistently positive result. Based on the
current literature, it is not possible to recommend this technique
for routine clinical practice.
Intradiscal biacuplasty is essentially a cooled radiofrequency
ablation of the posterior annulus. Two 17G introducers are
inserted into the posterior annulus on both sides of the disc
using fluoroscopic guidance. The electrodes, which are 18G
in size with a 6mm active tip, are then inserted through the
introducers and these electrodes are cooled internally using
circulating water. A bipolar current is then created between the
2 electrodes, thus creating a strip lesion between and around the
electrodes, ablating the posterior and posterolateral annulus.
The aim of biacuplasty is to destroy the abnormal nociceptors
that are responsible for discogenic pain in the posterior aspect
of the disc. A porcine study  and a cadaveric study  have
shown that there is minimal thermal damage to the anterior disc
or posterior longitudinal ligament.
2 prospective cohort studies and 1 RCT with a follow-up
report at 1 year is found. Biacuplasty was first described in a
case report by Kapural et al. . The main author then enrolled
15 patients into a pilot study with a 6-month follow-up  with
positive results. Another group then conducted a prospective
study in 2011, again with 15 patients and 6 months follow-up,
and improvements in the VAS and ODI were seen .
The only RCT on biacuplasty was done by Kapural et al.
. In this double blinded sham-controlled study consisting
of 32 patients in each group with a follow-up of 6 months,
positive results were seen in the treatment group in terms of
statistically significant improvements in SF-36, NRS and ODI
scores with no complications reported. However, 9 patients
were lost to follow-up and the results stated were from a per
protocol analysis. The same authors recently published a followup
study to the RCT mentioned above . The patients who
underwent the treatment had follow-up to 1 year and the sham
patients who elected to cross over to the treatment arm had
follow-up to 6 months. Sustained improvements were seen at
12 months in the NRS and SF-36 scores but the improvement
in ODI score became insignificant. For the cross-over patients,
significant improvement in the SF-36 score was obtained but
the improvements in NRS, ODI and opioid usage did not attain
statistical importance. Similar to the criticism of the RCT, there
was a high drop-out rate and a per-protocol analysis was used.
In summary, there is 1 RCT of moderate quality and 2
prospective studies that show positive results for intradiscal
biacuplasty. No complications have been reported that is directly
related to this procedure. There is certainly a need for more
studies to be done on this promising procedure. There is some
evidence that this modality may be useful for the treatment of
The latest addition to the treatment of discogenic LBP is
intradiscal MB. The postulated mechanism of action is that MB
is neurolytic and it denervates the small nociceptive fibers that
grow into a degenerated disc’s annulus fibrosis [71,72]. MB is
also a direct inhibitor of nitric oxide which is involved in the
inflammatory process in the intervertebral disc [73,74].
5 studies have been published on it with the first being
reported in 2005. In this prospective study  with a follow up
of 18 months, Peng et al had very good results with intradiscal
MB in 24 patients. The authors then went on to perform a well
conducted multi-centre placebo controlled RCT  (Jadad
score = 5) in which similarly impressive results were replicated.
Patients were followed for 24 months and large decreases in
NRS and ODI were seen. These improvements were seen by 6
months and sustained up to 24 months.
Unfortunately, these results were not seen when other
authors utilized MB in their own centers. Gupta et al.  had
a small retrospective case series of 8 patients in which only 1
patient has 100% relief at 6 months. 3 patients had relief for
shorter durations of between 2 weeks to 5 months. The other 4
had no pain relief at all. Kim et al.  conducted a prospective
study in 20 patients with a follow-up of 12 months. At 3 months’
post-procedure, 55% had an improvement but this reduced to
25% at 12 months. The difference in this study as compared to
the previously mentioned ones is that there was a 1 week lapse
between PD and MB injection, and that a significant number of
their patients (75%) had treatment at 2 or more levels. There is a
possibility that the patients in this study had more severe disease
as compared to the other studies and as such their patients could
not attain the same results as in Peng’s study. The most recent
study published was by Kallewaard et al.  as a pilot study
to assess the efficacy of MB, and if it was effective to carry on
and conduct a larger placebo controlled RCT. In their study, they
defined success as a 30% decrease in pain at 6 months which 6
out of 15 patients (40%) managed to achieve.
In addition to questions over its overall efficacy, it is a
relatively recent development of intradiscal therapy and so it is
unknown if there is any utility or harm in repeating intradiscal
MB injection in situations in which it provided short term pain
In summary, there is a well-conducted RCT that showed very
good results for intradiscal MB therapy but conflicting results
were then shown in the other trials. There is some evidence that
it may be efficacious but more studies will be needed to confirm
A recent article describes a new navigable catheter, L’DISQ
(U&I Co. Ltd., Uijeongbu, Korea) that uses coblation technology to
ablate the torn annulus. L’DISQ allows coblation to be performed
directly to the annular tear as the catheter is navigable so it
should theoretically be more effective. In this pilot study of 20
patients, which was conducted by the inventor of the device, 11
had greater than 50% pain reduction at 48-week follow-up .
More studies are needed to properly assess this modality.
Several articles have already been published utilizing
percutaneous lumbar disc decompression (PLDD) for lumbar
radicular pain but the literature of its effectiveness on discogenic
LBP is scanty. A prospective study was published in 2011 
which described its usage in discogenic pain. 11 patients with
LBP secondary to a single level disease that had a selective
anaesthetic blockade of 1 ml of 1% lidocaine were included. At
a follow-up of 2 years, there were significant improvements in
ODI and VAS scores. 1 patient developed a discitis that needed
surgery. The other study on the use of PLDD for discogenic pain
was a retrospective report in which patients were included either
based on clinical findings or a discography . 32 patients
responded to a telephone interview in which 28 had a good or
fair response. More studies will be needed before this modality
can be considered a viable treatment for discogenic LBP.
There are 2 articles that have been published investigating
the efficacy of using etanercept, a TNF-α inhibitor, intradiscally.
The first study was a double blind, placebo controlled, dose
response RCT (Jadad score = 5) in which intradiscal saline was
compared against a range of etanercept doses with the maximum
being 1.5mg . At 1-month follow-up there was no difference
between any of the etanercept groups and placebo. The 2nd
article, also an RCT (Jadad score = 4), compared 2 groups. The
control group had 10mg bupivacaine infiltrated intradiscally and
the other group had 10 mg etanercept introduced in addition
to that . The study showed lower NRS scores from one to
4-week follow-up in the etanercept group but this difference
disappeared at 8-week follow-up. Intradiscal etanercept seems
to provide short term pain relief at best.
As presented in our topical review, most of the studies on
percutaneous treatments to discogenic pain are observational in
nature. Only 7 out of more than 40 articles presented are RCTs.
Though all are of reasonable to good quality (Jadad score all ≥3),
5 out of 7 either did not calculate a sample size or did not meet
it at the end of the study [14,32,43,46,83]. As such, the negative
results obtained in the study could have been due to a lack of
power. The difficulty of obtaining a proper blinded control group
may also be a reason for the lack of RCTs as subjecting patients to
a sham interventional procedure is neither palatable to patients
and ethical boards.
The assessment of the various modalities is dependent
largely on the results of prospective observational studies with
all their inherent biases. As such, it is difficult for any modality
to have a strong positive recommendation and none were found
in our review. Even among the modalities with a weakly positive
recommendation i.e. biacuplasty and MB, more well conducted
studies are certainly needed. IDET has the most number of
publications though most of it is observational in nature. As
mentioned above, 1 RCT was a negative study and the other RCT
showed a slight advantage of IDET over placebo in well selected
MB and biacuplasty are still in its infancy as compared to
IDET but the initial results show promise. More evidence will be needed to ascertain their role in the treatment of discogenic pain.
Though the mechanism of action of nucleoplasty for discogenic
LBP is still in doubt and with some of the trials not fulfilling IASP
criteria, all the observational studies show a positive result.
More literature specifically addressing nucleoplasty for the
management of discogenic LBP are needed to evaluate if it is a
viable treatment option.
For the other modalities, it has been shown in this article
that intradiscal steroids, DiscTRODE, intradiscal RF and L2
blockade/RF are not effective and there are better alternatives in
the treatment of discogenic LBP. Intradiscal pRF has conflicting
results and the methodology differs substantially between the
studies, making it difficult to evaluate and should only be used in
the context of a study. Rami communicantes RF only has 1 study
in a very select group of patients and so no conclusions can be
made on this modality.
It should be remembered that in this group of patients
that have failed conservative management of discogenic
LBP, the options for treatment are limited. Fusion surgery
for this condition is not without controversy [84,85] and its
complications significant. However, allowing the patients
to continue on long-term analgesics with its attendant side
effects is not optimal, and many of these patients will still have
significant disability . Percutaneous treatments may allow a
subset of these patients to have a significant improvement and
the risk of complications as compared to surgery is much lower
[87-90]. It is therefore critical to appraise the evidence for such
modalities and adhere to strict patient selection. The limitations
of this article are that only literature in PubMed and published
in English were considered [91,92]. A broad overview, though
not exhaustive, of all available percutaneous treatments for
discogenic back pain has been discussed in sufficient detail to
identify a need to have better designed studies and modalities
for the indication of discogenic back pain [93-95].
Biacuplasty and intradiscal MB have a weakly positive
recommendation for the treatment of discogenic pain.
Nucleoplasty may potentially be a viable treatment though
more evidence and basic research is needed. The evidence for
IDET is equivocal. There are negative recommendations for
intradiscal steroids, intradiscal RF, DiscTRODE or L2 blocks/
RF. Future trials on discogenic pain should ensure that the IASP
definition of discogenic pain is adhered to and the methodology
of confirming the diagnosis of discogenic pain clearly elucidated.