Image-guided lumbar facet joint infiltration in nonradicular low back pain

Abstract Objective: To assess the efficacy of facet joint infiltrations for pain relief in 44 selected patients with chronic nonradicular low back pain (LBP). Materials and Methods: Forty-four patients with chronic LBP of more than 3 months′ duration were selected for facet joint infiltration. The majority (n = 24) had facetal pain with no evidence of significant facetal arthropathy on imaging. Fifteen patients had radiological evidence of facetal arthropathy, one had a facet joint synovial cyst, three were post-lumbar surgery patients, and two patients had spondylolysis. Facet joint injections were carried out under fluoroscopic guidance in 39 patients and under CT guidance in 5 cases. Pain relief was assessed using the visual analog scale at 1 h post-procedure and, thereafter, at 1, 4, 12, and 24 weeks. Results: A total of 141 facet joints were infiltrated in 44 patients over a 2-year period. There was significant pain relief in 81.8% patients 1 h after the procedure, in 86.3% after 1 week, in 93.3% after 4 weeks, in 85.7% after 12 weeks, and in 62.5% after 24 weeks. No major complications were encountered. Conclusions: Facet nerve block was found to be a simple, minimally invasive, and safe procedure. With meticulous patient selection, we achieved long-term success rates of over 60%. We conclude that this method represents an important alternative treatment for nonradicular back pain.


Patient selection
All patients were initially assessed by a neurosurgeon before referral to the anesthesiologist-run pain clinic. A total of 44 patients were included in the study on the basis of the following criteria: a. Chronic LBP of more than 3 months' duration, not responding to conventional drugs, exercise, and physiotherapy b. Symptoms suggestive of facetal pain; i.e., LBP with or without radiation to the butt ocks, thigh, or groin; pain increasing on hyperextension; and pain when initiating movement c. Focal tenderness over the facet joint elicited by digital pressure d. Post-lumbar disc surgery patients with persistent pain and no MRI evidence of arachnoiditis or recurrent disc disease Chronic low back pain (LBP) has assumed endemic proportions, with an annual prevalence of 5-20% in the industrialized world. [1] It is not always possible to pinpoint the exact structure or pathology responsible for LBP. In 2001, a large study on patients with chronic LBP showed that in as many as 15-45% of patients the pain was due to pathology of the facet joints and only in 13-20% was the pain due to herniated discs. [2,3] With the realization by surgeons that not all backache is a 'disc' and not all patients with LBP will respond to surgery, there is increasing awareness of the role of percutaneous injection techniques in the nonoperative management of chronic LBP. The use of image-guidance with ß uoroscopy or CT scan has increased the precision and safety of these procedures. [4] However, despite the increasing popularity of these procedures, there are few studies exploring their therapeutic effi cacy. We report our experience of facet joint inÞ ltration in 44 selected cases of chronic nonradicular LBP. To the best of our knowledge, this is the Þ rst sizeable series documenting the therapeutic use of facet joint injections in the Indian context.
All the patients had a preprocedure MRI for categorical exclusion of a discogenic cause for the LBP. Evidence of facetal arthropathy on MRI was noted but was not considered an inclusion criterion by itself.
The exclusion criteria were: a. A neurological deÞ cit in the lower limb or a positive sciatic nerve stretch sign (i.e., radicular pain radiating below the knee, elicited by a passive straight leg raise of 60-90°) b. Evidence of nerve root compression at the expected level on MRI c. Clinical or imaging evidence of infection or neoplastic disease d. Possible pregnancy, bleeding diathesis, or anticoagulant therapy e. History of sensitivity to local anesthetics There were 23 men and 21 women included in the study; the ages ranged from 20-74 years. The duration of symptoms varied from 3 months to 2.5 years (mean duration: 11 months). In our patient population, the largest subgroup (23/44) had clinical signs and symptoms of facetal pain but no evidence of signiÞ cant facetal arthropathy on imaging (radiography / CT / MRI). Fift een patients had clinical as well as radiological evidence of facetal arthropathy [ Figure  1], one patient had a facet joint synovial cyst [ Figure 2], three were post-lumbar surgery patients, and two patients had spondylolysis at the L 4-5 and L 3-4 levels, respectively.

Procedure
The procedure was explained to the patient in detail and writt en consent was obtained. Injections were performed under ß uoroscopic guidance in most (39/44) patients. CT guidance was used in Þ ve cases. The levels and side(s) to be injected were selected by the treating pain physician on the basis of the tenderness elicited over the joint, correlated with imaging Þ ndings, if any.
Fluoroscopic procedure: The patient was placed in the prone position with a pillow under the abdomen to correct the lumbar lordosis. The joint to be injected was located and marked. The x-ray tube was then slowly rotated till the joint appeared in proÞ le as two parallel lines. Aft er cleaning and draping, and administration of local anesthesia, a 22-G spinal needle was inserted in line with the x-ray beam till it contacted bone at the lip of the facet joint. With Þ ne movements the needle tip could be made to enter the joint with a distinct 'give.' In the early cases, we conÞ rmed the intra-articular position of the needle by injecting 0.5 ml iohexol (Omnipaque ® ; Amersham Health, New Jersey, USA) under ß uoroscopy. Later, with increasing experience, we found that conÞ rmation of correct needle placement could be made by feel and by viewing the joint in the lateral oblique projection [ Figure 3]. Once the needle was in place, 0.5 ml of 0.25% bupivacaine (a long-acting local anesthetic) and 0.5 ml (20 mg) of methylprednisolone acetate were injected into the joint. The patient was observed for 1 h aft er the procedure to document pain relief and to monitor for allergic reactions.
CT-guided procedure: The patient was placed in the prone position and 5-mm axial sections were obtained at the level of interest to determine the entry site and the angle of approach. The entry site was marked on the skin and a 22-G needle was advanced into the joint [ Figure 4]. The drug injection protocol was identical to the one used with ß uoroscopic guidance.

Assessment of pain relief
Pain relief was assessed using a visual analog scale (VAS), with a score of 0 denoting 'no pain' and a score of 10 the 'worst pain possible' [ Figure 5]. The VAS score was assessed before the procedure, 1 h aft er the procedure and, thereaft er, at 1, 4, 12, and 24 weeks. A reduction in the VAS score of Chaturvedi et al.: Lumbar facet joint infi ltrations   50% or more from the pretreatment score was considered as signiÞ cant pain relief and the patient was labeled a 'responder.'

Results
A total of 141 facet joints (75 bilateral and 66 unilateral) were inÞ ltrated in 44 patients, over a period of 2 years. The maximum number of inÞ ltrations were at L 4-5 (31.9%) followed by L 3-4 (26.5%) [ Table 1]. Only a small proportion of the patients (6/44) had an injection at a single level; most required injections at multiple levels: 14 at two levels, 16 at three levels, and 8 at four or more levels.

Pain relief
The number of patients with signiÞ cant postprocedure pain relief on day 1 and at 1, 4, 12, and 24 weeks is detailed in Table 2. Immediately postprocedure, 81.8% reported signiÞ cant pain relief. This number increased slightly at 1 week and reached a peak at 4 weeks, by which time as many as 93.3% patients had responded. However, the number of patients with pain relief declined to 62.5% at 24 weeks.
At 4 weeks, the three patients who did not respond were re-injected at the aff ected levels. Of these, one responded to the reinjection; the other two remained non-responders throughout the period of follow-up. Four of the patients who had responded well at 4 and 12 weeks required repeat injections between 16 and 20 weeks to maintain the pain relief. Another four patients who had good relief till 12 weeks received repeat injections at 24 weeks as the eff ect waned.
A further analysis of the pain relief in the different subgroups of patients is shown in Table 3. There was no diff erence in the number of responders in the group of patients with imaging evidence of facetal arthropathy as compared to those with no imaging Þ ndings. The single patient with a synovial facet cyst and both the patients with spondylolysis showed good pain relief in the immediate postprocedure period and remained pain-free for the entire duration of follow-up. The three patients with failed back surgery syndrome (FBSS) showed a poor response, with only one patient responding in the short term; none had any signiÞ cant pain relief aft er 4 weeks, despite re-injections.

Complications
We saw no major complications. Five patients had minor undesired eff ects in the form of soreness or local skin bruising; these symptoms lasted 2-3 days and subsided without treatment.

Discussion
Facet joints are true synovial joints which are innervated by the medial branches of the dorsal rami. The presence of nociceptive nerve Þ bers in the synovium and Þ brous capsule of the facet joints suggests that these joints may be a cause for LBP when they are stressed due to segmental instability, inß ammatory synovitis, degenerative arthritis, or a combination of all of these. [5,6] Based on studies using controlled diagnostic blocks, it has now been conclusively proved that facet joints are a source of pain in as much as 15-45% of patients with LBP. [2,7] Unfortunately, there are no clinical or imaging Þ ndings to deÞ nitively diagnose facetal pain and very oft en the term 'facet syndrome' is used as a 'dustbin diagnosis' when nothing else Þ ts. However, there are some features that are characteristic of facetal arthropathy. These include diff use referred pain over the butt ock and posterolateral thigh, exacerbation of pain with hyperextension or lateral bending, tenderness localized over one or more facet joints on deep pressure, and absence of root pain or neurological deÞ cits. [4] Imaging is not reliable for the diagnosis of facetal osteoarthritis since the changes seen on x-ray, CT, and MRI are equally common in patients with and without LBP, and most studies have failed to show a correlation between radiologic imaging Þ ndings and facet joint pain. [8][9][10] Facet joint injection with local anesthetic and steroid is the simplest and most common procedure for facet jointmediated pain. These inÞ ltrations are diagnostic as well as therapeutic and the choice of guidance-whether CT or ß uoroscopic-is largely a matt er of personal preference and experience, as both are equally effective. [4,9,10] The immediate pain relief aft er the injection is att ributed to the eff ect of the long-acting local anesthetic which interrupts the pain-spasm cycle. The corticosteroid begins to act by 1 week and by about 3 weeks the peak eff ect sets in. There may be a nonspeciÞ c synovitis present in many of these joints that is relieved by the anti-inß ammatory action of corticosteroids. In many cases, rupture of the articular capsule during injection results in the drugs diff using into the neural foramina too, thus, acting on the adjacent nerves as well [ Figure 6]. A simple physical eff ect, whereby inß ammatory exudates or adhesions are cleared from the joint and the nerve root sleeve, may also play a role. [11][12][13] Accurate assessment of pain is a prerequisite for the effective management of patients with LBP. Of the numerous tools available to assess pain, we adopted the numerical visual analog scale (VAS) since it is an objective measure and can be used to track serial changes. [13] Many recent studies of facet joint injections have also used the VAS as the scoring system for pain, thus allowing us to compare our Þ ndings with other studies. Also in accordance with these studies, we considered a 50% reduction in the VAS score from the preprocedure level as indicating signiÞ cant pain relief. [12,14] Literature de scribing the effectiveness of facet joint inÞ ltrations is as abundant as it is controversial. Early studies of facetal infiltrations showed poor results and led to the conclusion that this is a nonspeciÞ c and ineffective method of treatment. [14,15] It is possible that these disappointing results were due to improper patient selection, poor localization of the site for injection, or inadequacies in the volumes and types of drugs used. [ 4] Other studies in recent years have reported encouraging results with facet joint inÞ ltrations and the results of the pain relief obtained in these studies are shown in  4. These results demonstrate signiÞ cant short-term (1-12 weeks) pain relief in 62-74% of patients. Though the pain response gradually declines over time, even in the medium term (up to 24 weeks) as many as one-third of patients still experience signiÞ cant pain relief. [10,12,17,18] Our results showed short-term relief in 86-93% of patients and medium-term relief in 62%, which compare favorably with the results of these earlier studies. In our opinion, the high number of responders in our study could be due to meticulous adherence to the patient selection criteria, with elicitation of paraspinal tenderness over the facet joint being the most important inclusion criterion. The presence or absence of facet joint arthropathy on imaging was not related to pain relief in any way, and the main factor associated with a successful outcome of facet joint interventions was clinically elicited paraspinal tenderness. [10,12,14] Injections for spondylolysis are a modiÞ cation of facetal injections, providing good response in most patients. Some workers feel that it is the fracture site which is painful and, accordingly, inÞ ltrate the break in the pars interarticularis. [4] Others have demonstrated that injection  of the adjacent facet joint also involves the spondylolytic area and the technique can thus be used for therapeutic injections. [10,19] On ß uoroscopy, we could demonstrate the passage of contrast from the facet joint into the defect in the pars interarticularis [ Figure 7], with good and sustained pain relief. The uniformly poor results with facet joint injections in patients with FBSS is because post-lumbar surgery pain is due to an interplay of numerous causes and facet joint injection addresses only one of them. [14,20] Due to the small number of FBSS patients in our series we could not derive any statistically signiÞ cant information from our data.
The major complications of facet joint infiltrations are related to improper needle placement, bleeding, or infection. Complications include dural puncture, hematoma formation, spinal cord or neural trauma, spinal anesthesia, septic arthritis / spondylitis, and chemical meningitis. [21] Despite this long list and numerous anecdotal reports in literature, [22][23][24] it is our experience that with meticulous att ention to technique and sterility, these major complications can be eliminated.
We acknowledge the following limitations of our study. Firstly, we did not have a control group receiving placebo injections. Secondly, we did not carry out a preliminary diagnostic block for patient selection prior to giving the therapeutic injection. This was because we found it diffi cult to justify a preliminary diagnostic block followed by a Chaturvedi et al.: Lumbar facet joint infi ltrations  Schulte et al. (2006) [17] 39 Not assessed 62% 41% 36% Gorbach et al. (2006) [11] 42 74% 33% Destouet et al. (1982) [9] 54 Not assessed 54% 38% subsequent therapeutic injection; such a protocol would have delayed treatment and exposed the patient repeatedly to the risk of complications and radiation exposure. Therefore we selected our patients on the basis of a good clinical examination.

Conclusions
Recent evidence-based guidelines, published in 2007, with respect to lumbar facet joint injections have convincingly demonstrated moderate evidence for short-and mediumterm (up to 6 months) relief and limited evidence for longterm relief. [11] Facetal injections are not curative; however, by abolishing pain for periods of up to 6 months they can decrease dependence on oral medications and facilitate early return to work. Since their clinical eff ect is for a limited duration and wanes aft er 6 months, they need to be repeated to maintain the pain relief. It is also important to adopt stringent criteria for diagnosing facet joint pain in order to avoid unnecessary and unwarranted injections. In conclusion, we have found that in carefully selected cases, lumbar facet block is a relatively simple, safe, and minimally invasive procedure that can be a valuable adjunct in the treatment of LBP.