Abstract
Objective:
To evaluate the feasibility of sensory mapping of lumbar facet joint pain in patients scheduled to undergo radiofrequency (RF) denervation.
Design:
Prospective cohort study.
Setting:
University teaching hospital.
Subjects:
A total of 15 participants listed for RF denervation of lumbar facet joint.
Method:
After written informed consent, participants were recruited to the study. Participants completed a pain diagram prior to their procedure. After successful image-guided placement of RF cannulas, the sensory detection threshold using 50 Hz stimulation was obtained, followed by application of suprathreshold stimulation. Participants mapped their stimulated area in comparison to their pre-procedure pain diagram.
Results:
All 15 participants had previously undergone diagnostic blocks. All participants were able to report either pain or paraesthesia during suprathreshold stimulation. In total, 14 out of 15 participants reported complete coverage of their usual painful area with suprathreshold stimulation of nerves scheduled for RF denervation. In one of the participants, an area of upper lumbar pain was not covered during suprathreshold stimulation. Nearly two-thirds of the participants (n = 9), reported either pain or paraesthesia, outside their normal painful area during suprathreshold stimulation. A total of 71 nerves were scheduled for RF denervation. Sensory electrical stimulation was successfully achieved in 68 out of 71 nerves (96%). The average sensory detection threshold was found to be 0.3 V while the suprathreshold stimulation was 0.6 V.
Conclusion:
Lumbar facet joint pain can be mapped using suprathreshold sensory stimulation, which has the potential to introduce objectivity during RF denervation.
Keywords: Low back pain, electrical stimulation, facet joint, radiofrequency neurotomy, nerve stimulation
Introduction
Low back pain is a very common chronic pain condition subjected to different treatment modalities including somatic and psychological approaches. The somatic approach seeks to establish a diagnosis. A variety of structures such as facet joints, intervertebral discs, sacro-iliac joints and muscles are known to be involved in low back pain. The well-established pathoanatomical approach employs precision diagnostic blocks to identify these anatomical structures as source of nociception. Of them, facet joints have been found to be one of the most common structure contributing significantly to back pain.1 In clinical practice, the medial branches from posterior primary rami or dorsal rami itself are anaesthetised to establish the diagnosis of facet joint pain. Radiofrequency (RF) denervation of these nerves is usually offered as evidence-based interventional pain management option to the sufferers.2
The validity and reliability of the medial branch blocks have been well established. However, their clinical utility has been a subject of discussion due to the false positive rates, post-procedure evaluation and cost-effectiveness.3 As the false positive rates of single diagnostic blocks are high at around 30%, controlled diagnostic blocks with a stringent post-procedure evaluation with 75% improvement in the index pain are recommended by the Spine Intervention Society.4 However, pragmatic trials highlight that such strict criteria are not often followed.5 A survey of pain physicians in UK also revealed significant variation in practice including the number of diagnostic blocks.6 Due to low morbidity associated with the procedure, RF denervation has also been carried out without prior diagnostic blocks.7 In addition, the currently accepted diagnostic paradigm follows a standardised approach of blocking two nerves for each lumbar facet joint but this approach does not account for any anatomical variations. Thus, if we develop a strategy for identifying the culpable nerves–medial branches and/or dorsal rami, it would account for any anatomical variation and introduce objectivity. Hence, we investigated the feasibility of sensory mapping of the medial branches using suprathreshold sensory stimulation during RF denervation.
Methods
The study was prospectively registered on the ClinicalTrials.Gov database: Identifier: NCT 03323775.
Participants
Following prospective ethical approval from the North West-Greater Manchester West Research Ethics Committee, consecutive male and female patients aged 18–80 years who were scheduled for RF denervation at the Aberdeen Royal Infirmary were invited to participate in the study. Participants who were deemed able to complete a pain diagram as well as understand and comprehend instructions given in English were recruited. Participants who had requested or required sedation during the procedure were excluded. This was a pilot feasibility study which aimed to recruit 15 patients.
Procedure
The standard local practice is to carry out a single diagnostic block with lidocaine before being considered for RF denervation. The target joints are identified based on the presence of paraspinal tenderness under imaging. Two nerves are blocked for each facet joint as advocated by the Spine Intervention Society guidelines. For denervation, RF cannulas are placed parallel to the nerve in the groove between superior articular and transverse processes using fluoroscopic pillar views as advocated by the Spine Intervention Society.4 Our local practice is to use 50 Hz stimulation for sensory detection and 2 Hz motor stimulation up to 2 V as safety check before anaesthetising the medial branches and/or dorsal ramus for RF denervation.
After written informed consent, participants completed a pre-procedure pain diagram–a body map of their usual back pain and rated their pain on a visual analog scale of 0–10. The standard procedure for RF denervation was undertaken. After adequate fluoroscopic confirmation of the RF cannula placement, lumbar medial branches and dorsal rami were stimulated using 50 Hz electrical stimulation. The voltage at which participants felt any new sensation (tingling, pressure or back pain) was considered as the sensory detection threshold. The voltage was deliberately increased to a maximum of threefold of the sensory detection threshold, considered as the suprathreshold stimulation. After suprathreshold stimulation of each nerve, participants were asked to describe the area where they were feeling pain in a body map, considered as sensory mapping. They were asked to compare this against the pain diagram completed prior to the procedure. For each stimulated nerve, the collected data included sensory detection threshold voltage, suprathreshold stimulation voltage, whether the stimulation was within the pain area or extended beyond the pain area, as well as the percentage of usual pain area covered. The standard procedure for RF denervation was then completed in accordance with local practice.
Other participant data collected included age, sex, occupation, duration of pain, pain co-morbidities, the number of nerves scheduled for lesioning and total coverage of pain after stimulation.
Data analysis
Anonymised data were entered into case record forms prior to transfer to a spreadsheet. Descriptive analysis was undertaken.
Results
In total, 24 potential participants were invited to take part in the study and 18 were recruited. The CONSORT diagram is shown in Figure 1. Three participants were excluded due to the requirement for sedation and data from 15 participants are presented. Baseline participant data are presented in Table 1. A total of 11 participants stated that they had other pain comorbidities, such as chronic widespread pain or other neuropathic pain, in addition to their lower back pain. The median (range) duration of low back pain was 7 (2–30) years. The median pain score on the day of RF procedure was 7.5 (3–8.5). Two participants had previous RF denervation.
Figure 1.
Study flow diagram.
Table 1.
Baseline demographics.
| Demographics | |
| Patients (M = 7, F = 8) | 15 |
| Age (years) | 62 (30–80)a |
| BMI (kg/m2) | 26.8 (21.9–39.1)a |
| Pain duration (years) | 7 (2–30)a |
| Pain Score Average (0–10) | 7.5 (3–8.5)a |
| Previous RF denervations | 2/15 |
M: Male; F: female; BMI: Body mass index; RF: radiofrequency.
Data presented as median (range).
A total of 71 nerves were scheduled for RF denervation. Sensory stimulation was successfully achieved in 68 out of 71 nerves (95.8%) using 50 Hz electrical stimuli. All 15 participants reported either pain or paraesthesia during suprathreshold stimulation and 14 of 15 participants (93.3%) reported complete coverage of their usual painful area with suprathreshold stimulation. The median coverage of pain areas for L3, L4 medial branches and L5 dorsal rami was found to be 50%. In five participants, suprathreshold stimulation of one nerve reproduced all their pain, thereby providing 100% coverage of the pain area. In only one of the participants, an area of upper lumbar pain was not covered with suprathreshold stimulation.
Nearly two-thirds of the participants (n = 9, 60%), reported pain/paraesthesia outside their normal painful area during suprathreshold stimulation. This was particularly evident in four out of five patients who had 100% coverage with a single nerve stimulation. Of the stimulated 68 nerves, five (7.4%), produced pain/paraesthesia completely outside the participant’s painful region while the pain/paraesthesia extended beyond the normal pain region in another 11 nerves (16%). The most common nerve, not contributing to pain, was found to be L5 dorsal ramus (n = 9).
The details of painful area covered by individual nerves, sensory detection threshold and suprathreshold stimulation can be found in Table 2. The sensory detection threshold ranged from 0.1 to 0.8 V with an average of 0.3 V. The suprathreshold stimulation voltage ranged from 0.3 to 1.5 V with an average of 0.6 V. In 60% (n = 41), a threefold increase in sensory detection voltage was needed while a twofold increase was needed in 19% (n = 13) to achieve an adequate clinical response. For a further 13% (n = 9), the voltage increase lay between two and three times the sensory detection threshold and in 7% (n = 5), adequate clinical response was achieved with increasing the voltage up to twice the sensory detection threshold. No adverse events were noted during this study.
Table 2.
Individual nerve characteristics.
| Nerve | Number (n) | Sensory detection threshold (V) | Suprathreshold stimulation (V) | Contribution to painful area (%) |
|---|---|---|---|---|
| L2 | 2 | 0.16 (0.15–0.18) |
0.44 (0.42–0.45) |
65 (50–80) |
| L3a | 19 | 0.3 (0.15–0.8) |
0.8 (0.3–1.4) |
50 (0–100) |
| L4a | 24 | 0.3 (0.1–0.7) |
0.8 (0.3–1.5) |
50 (10–100) |
| L5 | 23 | 0.3 (0.12–0.6) |
0.9 (0.3–1.5) |
50 (0–100) |
Data expressed as median (range).
Excluding nerves that were not stimulated (one L3 and two L4 medial branches).
Discussion
In our study, we chose to electrically stimulate the nerves which have been scheduled for RF denervation to explore acceptability of suprathreshold stimulation, feasibility of sensory mapping of lumbar facet joint pain and to evaluate whether there is concurrence with the nerve blocks. We showed that it is feasible to reproduce the pain in patients with chronic back pain using 50 Hz electrical stimulation in all participants.
A total of 93% of participants had complete coverage of the back pain with suprathreshold stimulation implying concurrence with the diagnostic nerve blocks and introducing objectivity during RF denervation. However, suprathreshold stimulation produced pain/paraesthesia outside the normally painful area in 60% of the participants. Non-painful region mapping was noted more with participants who had full coverage with single nerve stimulation alone. Thus, it can be inferred that all the stimulated nerves were not involved in pain transmission. In addition, one participant’s area was not fully covered, implying that all the appropriate nerves were not included for RF denervation. In our study population, 7.5% of the denervated nerves did not contribute to pain transmission as detected by suprathreshold stimulation.
Electrical stimulation of lumbar medial branches and dorsal rami was carried out in chronic low back pain patients as early as 1997. A study conducted by Fukui et al.8 is very similar to our study, as they electrically stimulated the medial branches of the dorsal rami and mapped out the ‘geographical’ distribution of each single nerve before RF denervation. However, it is unclear to what extent they related this back to the patients’ original pain. Windsor et al.9 demonstrated that a reproducible perceptible stimulation of all lumbar (L1–L4) medial branches and L5 dorsal ramus can be obtained between 0.05 and 0.25 V using 50 Hz stimulation in healthy volunteers. Recently, O’Neill et al.10 demonstrated that increasing the sensory detection threshold by a factor of three, on average, induces local pain of the stimulated lumbar facet joint. So, we used up to a threefold increase in the sensory detection threshold and successfully induced/reproduced the back pain. The average suprathreshold stimulation in our study was 0.6 V.
Initial selection of the lumbar facet joints for diagnostic blocks is subjective. Lower lumbar facet joints are commonly affected and tenderness of these joints under imaging is conventionally used in clinical practice to choose the joints. This dictates the number of nerves to target diagnostic blocks. For a given facet joint (L4/5), the medial branch from the same level (L4) and one level above (L3) are anaesthetised. If the patients respond positively by 50% or 80% relief, RF denervation of these nerves is carried out. In the event of inadequate relief from blocking nerves of the selected joints, other joints may subsequently be targeted for blocks, thereby exposing patients to further invasive treatments with inherent risks.
The diagnostic blocks have shortcomings in terms of high-false positive rates, post-procedural evaluation and cost-effectiveness.3,11,12 In addition, anaesthetising two nerves for a given lumbar facet joint ignores the well-established anatomical variations or aberrant medial branch innervation.13–15 All of these have the potential to negatively influence the success rate of RF treatment.16
Thus, suprathreshold electrical stimulation of medial branches has the potential to offer objectivity by reproducing patients’ back pain and improving safety by limiting RF denervation to nerves involved in pain transmission.
This was a pilot study to explore the feasibility and acceptability of suprathreshold electrical stimulation in a small number of participants. Furthermore studies are needed to explore the utility of sensory mapping using suprathreshold stimulation during RF denervation and the clinical outcomes.
Conclusion
Electrical sensory stimulation was able to recreate lumbar pain in almost all patients indicating that it may complement the diagnostic anaesthetic blocks by introducing objectivity in identifying nerves during RF denervation, improved patient safety and a better patient experience.
Acknowledgments
The authors would like to thank Dr Jeremy Weinbren, Consultant Anaesthetist, The Hillingdon Hospitals NHS Foundation Trust for his statistical advice on this paper.
Footnotes
Conflict of interest: The author(s) declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.
Contributorship: S.K. conceived the study. S.K., H.F.G. and S.D. were involved in protocol development, gaining approval, data analysis, and reviewing of manuscript. K.F. was involved in patient recruitment and data collection. S.K. drafted the manuscript. All authors reviewed and approved the final version
Disclosure: H.F.G. was the intercalated degree programme director at the University of Aberdeen during the conduct of this research.
Ethical approval: Ethical approval was obtained from theNorth West – Great Manchester West Research Ethics Committee, Barlow House, 3rd Floor, 4 Minshull Street, Manchester (REC Ref: 17/NW/0587).
Funding: The author(s) disclosed receipt of the following financial support for the research, authorship and/or publication of this article: K.F. received a John Snow Anaesthesia Intercalated BSc bursary. No funding was obtained for the running costs of the project.
Guarantor: S.K. is the Guarantor for this article.
Informed consent: Written informed consent was obtained from all participants.
Trial registration: The study was prospectively registered on the ClinicalTrials.Gov database. Identifier: NCT 03323775.
ORCID iD: Saravanakumar Kanakarajan 
https://orcid.org/0000-0002-8497-4982
References
- 1. Bogduk N. Management of chronic low back pain. Med J Aust 2004; 180: 79–83. [DOI] [PubMed] [Google Scholar]
- 2. National Institute for Health and Care Excellence (NICE). Low back pain and sciatica in over 16s: assessment and management (NICE guideline NG59), 2016, https://www.nice.org.uk/guidance/ng59 (accessed 30 November 2017). [PubMed]
- 3. Van Zundert J, Mehhail N, Vanelderen P, et al. Diagnostic medial branch blocks before lumbar radiofrequency zygapophysial (facet) joint denervation: benefit or burden? Anesthesiology 2010; 113: 276–278. [DOI] [PubMed] [Google Scholar]
- 4. Bogduk N. Practice guidelines for spinal diagnostic and treatment procedures, 2nd edn. San Francisco, CA: International Spine Intervention Society, 2013. [Google Scholar]
- 5. Juch JNS, Mass ET, Ostelo RWJG, et al. Effect of radiofrequency denervation on pain intensity among patients with chronic low back pain. JAMA 2017; 318: 68–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Manickam S, Krishna S, Kapur S. Survey of variation in the practice of interventional management of facet joint pain. Br J Pain 2014; 8: 84–89. [Google Scholar]
- 7. Cohen SP, Williams KA, Kurihara C, et al. Multicenter, randomized, comparative cost-effectiveness study comparing 0, 1, and 2 diagnostic medial branch (facet joint nerve) block treatment paradigms before lumbar facet radiofrequency denervation. Anesthesiology 2010; 113(2): 395–405. [DOI] [PubMed] [Google Scholar]
- 8. Fukui S, Ohseto K, Shiotani M, et al. Distribution of referred pain from the lumbar zygapophyseal joints and dorsal rami. Clin J Pain 1997; 13(4): 303–307. [DOI] [PubMed] [Google Scholar]
- 9. Windsor RE, King FJ, Roman SJ, et al. Electrical stimulation induced lumbar medial branch referral patterns. Pain Physician 2002; 5(4): 347–353. [PubMed] [Google Scholar]
- 10. O’Neill S, Graven-Nielsen T, Manniche C, et al. Ultrasound guided, painful electrical stimulation of lumbar facet joint structures: an experimental model of acute low back pain. Pain 2009; 144(1–2): 76–83. [DOI] [PubMed] [Google Scholar]
- 11. Derby R, Melnik I, Lee JE, et al. Cost comparison of various diagnostic medial branch block protocols and medial branch neurotomy in a private practice setting. Pain Med 2013; 14: 378–391. [DOI] [PubMed] [Google Scholar]
- 12. Bogduk N, Holmes S. Controlled zygapophysial joint blocks: the travesty of cost-effectiveness. Pain Med 2000; 1(1): 24–34. [DOI] [PubMed] [Google Scholar]
- 13. Bogduk N, Wilson AS, Tynan W. The human lumbar dorsal rami. J Anat 1982; 134(Pt. 2): 383–397. [PMC free article] [PubMed] [Google Scholar]
- 14. Kaplan M, Dreyfuss P, Halbrook B, et al. The ability of lumbar medial branch blocks to anesthetize the zygapophysial joint: a physiologic challenge. Spine (Phila Pa 1976) 1998; 23(17): 1847–1852. [DOI] [PubMed] [Google Scholar]
- 15. Kanakarajan S. Anatomy of the lumbar dorsal rami and the nerve supply to lumbar facet joints. Br J Pain 2016; 10: 33–34. [Google Scholar]
- 16. Bogduk N, Dreyfuss P, Govind J. A narrative review of the lumbar medial branch neuroanatomy for the treatment of back pain. Pain Med 2009; 10: 1035–1045. [DOI] [PubMed] [Google Scholar]