Fluoroscopic Epidural Steroid Injection: Pain Relief in Discogenic Sciatica Versus Lumbar Spinal Stenosis. A Study on Middle Eastern Patients

Objective: To compare the effect of epidural steroid injections (ESI) in patients with discogenic sciatica (Sci) versus patients with lumbar canal stenosis (LSS), not controlled by conservative treatment. Materials and methods: In our study, 80 patients with Sci and 66 with LSS were included. A single ESI (10 mg dexamethasone in 3 cc 0.25% bupivacaine) was applied under fluoroscopic control: one level above the highest stenotic level, in the posterior epidural space, via interlaminar approach in LSS and at the prolapse level, in the anterior epidural space, via transforaminal route in Sci. Pain intensity was assessed by VAS at baseline and on days 1, 15 and 30 after intervention. Results: The procedure was successful in 78 Sci and 63 LSS patients. Patients with Sci responded significantly better. At one month, pain reduction over 50% was achieved in 63% (52.3–73.7% at p = 0.95) of Sci but only in 35% (23.2–46.8%) of LSS (p = 0.03). Return to preintervention level happened in 47% (34.7–59.3%) of LSS versus 14% (6.3–21.7%) of Sci patients (p = 0.01). In 5 patients the procedure failed, without resulting morbidity. Conclusion: ESI are more effective in patients with Sci than in single level LSS. In multiple level LSS, results are disappointing


INTRODUCTION
Vertebrogenic low back pain (LBP) and lower extremity pain is very common and a high socioeconomic burden (1). About 15% of patients develop chronic LBP that persists throughout life (2). Surgically treated patients develop chronic LBP and/or neuropathic pain ("failed back surgery syndrome") in at least 10% of cases with discogenic sciatica and in up to 40% of LSS cases (3).
Early control of vertebrogenic pain may prevent the development of chronic pain and the associated negative outcomes (analgesic abuse/dependence, loss of employment, psychosocial problems etc.) (2,4). It enables early physiotherapy/rehabilitation that improves non-surgical treatment results (5). However, pain control is not always possible with non-invasive means only (6).
Epidural steroid injection (ESI) is a minimally invasive procedure that introduces the medication (often combined with local anesthetic) close to the pathology site, presumably enhancing its local action while minimizing the systemic effects (7). Numerous studies recommend ESI in low back and radicular pain, but others dispute their utility (8)(9)(10).
We compare the short-term efficacy of epidural steroid and anesthetic injection for pain relief in Sci and LSS patients, not controlled by conservative therapy alone. We involved Middle Eastern patients only as there are observations of reduced pain tolerance in this population compared with other groups (11,12).

STUDY POPULATION
This prospective case-control study involves 80 patients with discogenic sciatica (Sci) and 66 patients with lumbar spinal stenosis (LSS), recruited at a tertiary Spinal Clinic. The study was approved by the institutional ethics committee and required informed consent from the partic-ipants, in accordance with the Helsinki Declaration. The diagnosis was established by clinical examination and MRI obtained within the 3 months before intervention. Included were patients with low back pain and radicular pain who did not achieve satisfactory pain control and functional improvement after 6 weeks of medical treatment and physiotherapy, as this length of time is considered a transition point from acute to subacute pain stages (13).
Excluded were: 1. Cases of disc prolapse with extruded disc sequester within the spinal canal (who were offered surgical treatment). 2. Patients who had already undergone operative treatment or epidural injection. 3. Patients with motor deficit (they were offered surgical treatment). 4. Patients with absolute contraindications for corticosteroid treatment (peptic ulcer, uncontrolled hypertension, uncontrolled diabetes, etc.), hypersensitivity to local anesthetics or contrast.
The demographic and clinical features of the two groups are summarized in Table 1.
The clinical assessment and the invasive procedures in this study were personally performed by the authors.

PAIN INTENSITY ASSESSMENT
Pain was quantified using the visual analogue scale (VAS) (14) at baseline (the day of the procedure) and one, 15 and 30 days thereafter. Non-steroid analgesics were withheld 3 days before ESI and during the observation period.

EPIDURAL INJECTION
The procedure was performed in the operating theatre. The patient was positioned prone on a radiolucent table, with intravenous catheter inserted and monitoring of the ECG and the blood pressure. The level of the injection and the approach was determined according to nosology (see below). We first infiltrated the subcutaneous tissues at the injection site with 1 ml 1% lidocaine solution. The epidural space was reached using an 80 mm, 18G spinal cannula (Spinocan, Braun™). Epidurography was first performed by injecting 1 ml iohexol 300 mg/ml into the epidural space. After fluorographic verification of the needle position, 3 ml 0.25% bupivacaine and 10 mg dexamethasone was applied.
Injection in the posterior epidural space via the inerlaminar approach was done 1) one level above the highest level of stenosis in all LSS patients and 2) at the level of the disc lesion in 10 Sci patients, who had root symptoms in 2 contralateral dermatomes. We applied the following technique. The level of injection is determined under X-ray guidance. The point of skin penetration is 1-1.5 cm away from the midline, with the needle directed about 30 degrees towards the midline in the axial plane and 15 degrees rostral in the sagittal plane. The tip of the needle is directed towards the interlaminar space under X-ray control. When reaching the ligamentum flavum we used the "loss of resistance" technique, instilling about 5 cc of air. After entering the epidural space, we always performed aspiration to ensure the needle is not situated intrathecally or within a vessel. Then epidurography was carried out (Fig. 1). Finally, the anesthetic/steroid preparation was applied.    The 70 patients with unilateral Sci received anterior epidural injection via the transforaminal approach. In single disc pathology, it was done at the same level, on the side of the affected root. In multiple disc levels and clinical involvement of 2 or more roots, we approached the rostral disc level on the side of maximal clinical radiculopathy. For visualization with this approach we used oblique left or right projections. The C-arm was inclined to 30 degrees in the axial plane until obtaining the "Scotty Dog" image of the pedicle and adjacent bony structures. To avoid conflict with nerve structures, the needle tip should be directed below the pedicle, into the "safety triangle" defined by the tangential lines of the vertical and horizontal contours of the pedicle, as illustrated in Figure 2.
After reaching the safety triangle, the C-arm is positioned for frontal and lateral projections to verify the proper position of the needle tip. Then epidurography is performed and finally the anesthetic/steroid solution is injected.
All patients were observed for 2 hours after the injection.
The procedure was cancelled in cases of intrathecal or intravasal penetration; these patients were treated conservatively and excluded from further participation.
Statistical methods included descriptive, alternative, variance and non-parametric analysis. Statistical level of significance was accepted at p < 0.05.

RESULTS
In 5 patients (3.5%, 0.2-5.8%) the procedure was unsuccessful. Three had intrathecal penetration of contrast (2 after interlaminar and one after transforaminal approach). In the other two, aspiration yielded blood, indicating a vessel penetration. There was no associated morbidity.
In the remaining 141 patients, the procedure was uneventful. Their total hospital stay (including 2 hours' observation) was 6 hours (3.5-6.8). All completed the scheduled follow-up.
The patients with LSS had significant pain relief. Their reported pain intensity was 6 +/− 1.8 VAS score immediately before the intervention, 3.2 +/− 0.4 on the day after the injection, 4.3 +/− 0.6 two weeks and 5 +/− 0.4 at one month respectively. The best results were achieved in the 25 patients with single level stenosis, with 3.8 +/− 0.9 VAS score one month after the procedure.
One-way ANOVA confirmed the significant difference between groups at all comparison points, as illustrated in Figure 4.
The early effects, lasting days to weeks were quite significant in both groups. To better compare the effects of the procedure between LSS and Sci patients at one month we stratified the response at that point of time as good (VAS score decrease with more that 50%), satisfactory (VAS improvement less than 50%) and poor (return to preoperative values). Results are summarized in Table 2.
The higher effectiveness of the intervention in Sci patients is evident.

DISCUSSION
ЕSI produced major short-term improvements, significantly greater in Sci that in LSS patients. After one month, nearly half of the LSS but only about 10% of Sci patients returned to their pre-intervention condition. The procedure was particularly effective in LSS patients with a single level of stenosis and Sci patients with monoradicular involvement. The percent of failed interventions was very low.
ESI are the most popular interventional technique in low back and lower extremity pain (at least in the United States) and their use continues to increase (15). While their long-term efficacy and cost-effectiveness remain controversial, most experts agree on the short-term pain relief provided, particularly in sciatica (8)(9)(10)15). We assessed systematically the efficacy of epidural steroid injections in a Middle Eastern population as significant differences in pain tolerance between Middle Easterners and other cultures have been demonstrated (11,12).
Our study has weaknesses like is the lack of a placebo group (not allowed by the IRB for ethical considerations) that places it into the third class in terms of quality of evidence (16). Another disadvantage is focusing on pain and relying on VAS only. We intended to study functional outcomes using a standard tool (the Oswestry Disability Index) in its Arabic validated translation (17) but it was not well accepted by patients (questions regarding sexual life, hygiene were considered intrusive). As our referrals in a Military Hospital were predominantly male, the sexes proportions in our study cohort are not representative of the incidence of LSS and Sci in the general population.
Despite these shortcomings, we present some significant results.
The efficacy of ESI at one month in our patients was lower than the reported in some studies but similar to or higher than other patient series, originating from very diverse countries and settings and methodologically comparable with our study (9)(10)(11)(18)(19)(20)(21). While a direct comparison across different studies is methodologically not appropriate, we should note that our results tend toward a "median value" and suggest that the differences are not due to lower tolerance to pain in our population but rather to patient selection, to dose of medication used, different study design (blinded, open) and/or other confounding factors.
The significantly higher utility of ESI in Sci compared to LSS in our study confirms some previous reports (11,22,23). It may reflect the different pain mechanisms prevailing in either pathology. In LSS, pain is mostly nociceptive or mixed (nociceptive and neuropathic) (24), determined by mechanical factors. The neuropathic pain component in LSS is considered a consequence of ischemia causing repolarization disturbance of neuronal membranes that manifests clinically as hyperesthesia and neurogenic claudication (24). The membrane-stabilizing effect of steroids with hyperpolarization and inhibition of C-fibers conduction may explain the extended analgesic effect in LSS (25). However, the mechanical and vascular factors involved in LSS would not be corrected by the steroid application.
In contrast, in discogenic radiculopathy the pain is mostly neuropathic and inflammation likely plays a central role (26). Penetration of nucleus pulposus into the superficial layers of annulus fibrosus triggers an immune response, in particular -tumor necrosis factor alpha (TNF-α) from activated macrophages (27). Inflammation leads to disturbances in microcirculation with alteration of the myelin sheets in the nerve root that produce the clinical manifestations of radiculopathy (26,27). Neuropathic pain may be also related to prostaglandin cascade activation with release of substance P and other mediators in the affected dorsal ganglion (28).
If inflammation has a much greater significance in the pathogenesis of discogenic sciatica, the anti-inflammatory action of steroids would explain their higher efficacy in Sci versus LSS. This is supported indirectly by the effects of epidural application of TNF-a antagonists -etanercept, adalimumab -that alleviate sciatica (29,30). The efficacy of epidural clonidine in neuropathic radicular pain may also partially depend on its anti-cytokine actions (31).
The superior effect of epidural steroid/anesthetic in single-level LSS and in monoradicular Sci is intuitively expected. This finding may be helpful in patient selection.
Complications of ESI, from mild to even fatal have been reported (32), but we didn't observe procedure related   morbidity neither in the 141 patients with successful injection nor in the minimal percent of cases where we had technical failure.
In conclusion, ESI with fluoroscopic guidance is safe, effective in short term pain relief and may be routinely recommended for patients with Sci and single level LSS, if conservative measures are not sufficient. In multiple level LSS, well-motivated patients should be selected for ESI.

DISCLOSURE
The authors declare no conflicts of interest related with this article.