Access to the PPM Journal and newsletters is FREE for clinicians.
14 Articles in Volume 9, Issue #7
Anomalous Opiate Detection in Compliance Monitoring
Anticipating Biotechnological Trends in Pain Care
Continuous Lumbar Epidural Infusion of Steroid
Disordered Sacroiliac Joint Pain
Efficacy of Stimulants in Migraineurs with Comorbidities
Hand Tremor with Dental Medicine Implications
Helping Patients Understand the
Non-surgical Spinal Decompression (NSSD)
Pain Management in Nursing Homes and Hospice Care
Patients Who Require Ultra-high Opioid Doses
Relief of Symptoms Associated with Peripheral Neuropathy
Share the Risk Pain Management in a Dedicated Facility
The Multi-disciplinary Pain Medicine Fellowship
Thermal Imaging Guided Laser Therapy: Part 2

Continuous Lumbar Epidural Infusion of Steroid

Administering a continuous lumbar epidural infusion of corticosteroid and normal saline via an indwelling epidural catheter/infusion pump over a 3-day period offers a minimally-invasive option for treatment of lumbar radiculopathy.

Lumbar epidural corticosteroid injections have been the treatment of choice in treating a variety of low back conditions, including radiculopathy, for over 50 years. Recent improvements in technique utilizing fluoroscopy with use of a contrast medium, in conjunction with understanding of potential additional risk factors with the procedure,1 have served to both improve efficacy and reduce complications. Unfortunately, despite these ad-vances, there remain several pitfalls with the procedure.

One potential pitfall of therapeutic epidural steroid injections (ESIs) may involve the rapid re-uptake of medication within the epidural space. Over the years of performing ESIs, I have noticed in some cases the contrast may be com-pletely reabsorbed within several minutes of being injected within the epidural space. This being the case, one could postulate the same mechanism of rapid re-absorption takes place with the medication being utilized for the epidural, namely the therapeutic corticosteroid.

This may explain why many physicians I have met over the years believe particulate steroids may be more effective then more water-soluble steroids. The particulates in the steroid may not get reabsorbed as quickly and, by “sticking around” the suspect segment, may give a more timed release of the medication to the inflamed site. Whereas, the more water-soluble, smaller particulate steroid is completely reabsorbed into the bloodstream within minutes of being injected into the epidural space. I have talked personally with several spine surgeons over the years who have seen first hand these steroid particulates used in epidurals when operating on the spine. Unfortunately, I do not know of a study that compares the “efficacy” of utilizing certain steroid preparations over others. Some physicians believe more water-soluble steroids are as effective as more particulate steroids. Some do not. There may be, however, a difference with respect to safety of which steroid is used. Derby et al demonstrated in his recent study1 that, when compared to other particulate steroid solutions, dexamethasone phos-phate had particle size smaller than red blood cells, had the least tendency to aggregate, and the lowest density. These characteristics should significantly reduce the risk of an embolic infarct should intra-arterial injection occur.

Although some of the particulate may “stick around” in the epidural space, the majority is most likely reabsorbed rather quickly, just as the contrast is. This being the case, it may explain in part, why ESIs are not equally effective on all patients. Some patients get no relief, some get partial relief (i.e. 20% better after two ESIs), some get transient relief (i.e. 90% better for three weeks, but back to baseline by week 4), and some get complete relief after 1-3 ESIs. It may be that patients who have a more rapid reabsorption of the steroid may not respond as favorably as those patients who have a slower reabsorption.


What if a more prolonged, continuous infusion of the steroid to the inflammed segment were administered via an indwelling epidural catheter using an infusion pump? Would the patient achieve a greater and more prolonged clinical response versus a single bolus ESI? This is the premise behind the continuous infusion of lumbar epidural corticosteroid to reduce inflammation and pain using the Infusion Catheter Epidural (ICE) (see Figure 1). Prior to further discussion, I believe a short review of the scientific literature would be indicated.

Figure 1. Example of catheter with infusion pump (ICE).

Review of Literature

There have been several studies over the years that have studied continuous infusion of medications for treatment of radiculopathy. In a recent study by Pasqualucci et al,2 one hundred and forty one patients were involved in a clinical trial in which continuous infusion of anesthetic and corticosteroid utilizing an epidural catheter was compared to single bolus injections for treatment of cervical radiculopathy. “The results of this randomized study show a statistically significant efficacy of the treatment of cervicobrachial pain with epidural local anesthetic plus corticosteroids in continuous infusion rather than in single injections, in patients with chronic pain who did not respond to conservative therapies.” Worth noting, the epidural catheter remained in place, on average, for 24.23 +/- 4.82 days in those chronic patients. Six month follow up occurred in this study. Patients received on average 320mg of methylprednisilone over 30 days. No significant complications were reported in the study.

In another clinical trial, Kim et al assessed 46 patients with lumbago and sciatica using continuous epidural block with steroids and local anesthetics using an epidural catheter connected to a multiday infusor, with an infusion rate of 1 ml/hr.3 At the time of discharge, 69.5% of all cases showed excellent or good results. The catheter was removed after 1-2 weeks, on average. Their conclusion was “continuous epidural block is simple and safe in the treatment of lumbago and sciatica, especially in acute phase.” Once again, no significant complications were cited.

Pauza et al, also studied treating radiculopathy with an indwelling lumbar epidural catheter and infusion pump.4 There was no patient morbidity or mortality associated with this treatment. Dr. Pauza noted that “safe and effective treatment of lower extremity radiculopathy may be obtained with this new method.” Two patients in the study had previously experienced brief significant relief with prior transforaminal injections of lidocaine and corticosteroid, but failed to experience sustained relief. In this study, however, they experienced appreciable benefit with an indwelling catheter. In conclusion, Dr. Pauza states,” While affording satisfactory symptom relief with comparatively few surgical facility encounters, administering corticosteroid through an indwelling epidural catheter over three days may be a viable treatment option for radicular pain.” Once again, no significant complications were cited.

Another study looked at continuous epidural block for cervicogenic headache.5 In this study by. HE Ming-Wei et al, a retrospective analysis of 37 patients was performed to investigate the safety and efficacy of a continuous cervical epidural block utilizing a combination of dexamethasone, lidocaine, and saline for 3-4 weeks. Patients received approximately 1mg of dexamethasone per day. Epidural catheters were placed into the cervical epidural space under fluoroscopic or CT guidance utilizing a contrast medium for verification. The catheter was then hooked up to an infusion pump with a flow rate of 5ml/hour over a 3-4 week period. The catheter was removed after 3-4 weeks of treatment. In addition, triamcinolone acetomide 5mg was administered once weekly during this 3-4 period. This retrospective study demonstrated that the continuous epidural administration of corticosteroids and anesthetic is effective and can control chronic cervicogenic headache for at least 6 months. No corticosteroid related adverse events were reported.

One additional study worth mentioning is one by Sugita et al with the Department of Neurosurgery, Shinshu University School of Medicine, Matsumoto, Japan.6 Dr. Sugita studied 20 cases of sciatica of various causes and 200 post traumatic visual disturbance. Patients were given 1-2 intrathecal injections weekly, using 8mg of dexamethasone per injection, until a total of 80-100mg was given (over an approximate 1-2 month period per patient). They also cite in the article an additional 200 cases of intrathecal steroid therapy for radiculapathy, myelopathy, and optic neuropathy. Although intrathecal use of ICE mentioned earlier is not indicated, ICE does utilize dexamethasone, with 8-12mg administered in the infusion pump over a 72-hour period. In Dr Sugita’s study, the patients received a far larger dosage over a more prolonged period of time, and apparently tolerated this dosage quite well. I believe this is worth noting because one rare possible complication of ICE could be catheter migration intrathecally. For this to happen, one would postulate the physician would have had to unknowingly puncture the spinal sac and enter the subarachnoid space, not have gotten aspiration of spinal fluid, not have seen this myelographic pattern using contrast dye and flouro, and then placed the catheter intrathecally (or it could migrate there through prior puncture) and, once again, not visualized the flow intrathecally with contrast under flouro. Although this scenario is highly unlikely in the hands of a skilled interventionalist, it is nonetheless theoretically possible. Using preservative-free medications in the pump should minimize the risk of arachnoiditis in this scenario and using dexamethasone as described above by Dr. Sugita, apparently is safe should inadvertent intrathecal administration of medication occur. One would expect the patient to develop a positional headache and the catheter should be immediately removed should this occur.

One final study by Antonio Aldrete7 cites 551 patients with severe, low back pain due to a variety of non-malignant causes and previously unresponsive to epidural injections who underwent temporary lumbar epidural catheter infusions using low dose bupivicaine and fentanyl via disposable infusion pumps. All but a few treatments resulted in good to excellent pain relief and most permitted patients to increase their physical activities to near normal levels. Treatment duration averaged 19.8 days per patient. Two epidural abscesses occurred, five cases of cellulitis, and two cases of fasciitis. Seven of these infections occurred after the 20th day of catheter infusion, and two occurred after the 11th day. No infections were noted prior to the 11th day of catheter insertion. Patients did receive prophylatic antibiotics, but they were not started until day four of treatment, and were given for six days on, then six days off, for the course of treatment. Of these nine infections, all were treated successfully with IV antibiotics, and none of these patients required surgical intervention. No neurologic sequelae were noted. Overall, the cost of this approach was substantially less than that associated with insertion of an implantable device or surgery. Although this approach appears to be more of a palliative procedure until the patient’s condition resolves, it does demonstrate safety, ease of use of long- term, continuous lumbar epidural catheter placement via an infusion pump for treatment of low back conditions unresponsive to traditional conservative care, as well as relatively low cost compared to other more aggressive treatments.

In summary, these six clinical trials studied approximately 1,004 patients utilizing continuous infusion/multi-injection use of steroids for a prolonged period of time, with five out of six studies using long term placement of indwelling catheters in the epidural space. Both safety and efficacy were established, with no significant complications reported.

Thus with this backround, I would like to discuss a case report of continuous epidural infusion of steroid and saline via an indwelling epidural catheter utilizing an infusion pump (ICE) in an ambulatory setting for a patient with lumbar radiculopathy.

Indications For ICE

The ideal candidate for this technique is a patient who had prior partial or transient success with lumbar/caudal epidurals. This would seem to indicate the patient is a responder to corticosteroids, but did not achieve long-term success. From a historical perspective, ESIs are typically the final conservative measure prior to surgical intervention. Although there are new non-surgical procedure options available, ESIs are typically the last stop prior to surgical consideration. In the recent study by Weinstein et al8: “both operated and non-operated patients with intervertebral disc herniation improved substantially over a 2-year period. The intent to treat analysis in this trial showed no statistically significant treatment effect for the primary outcomes.” Although spinal surgery is a viable and necessary option in many situations, additional conservative measures are needed since the “gold standard” lumbar epidural is not effective in all cases, and spinal surgery may not offer a statistically significant improvement in outcome after a two-year period.8

Contraindications For ICE

Similar to accepted standards for therapeutic ESIs. This technique should not be performed on pregnant patients, those with compromised immune system, those with active infections, those anti-coagulated, those with history of anaphylaxis to contrast dye, or with bleeding diathesis. Caution should be utilized in those patients with significant co-morbidities, particularly those susceptible to adverse reactions related to corticosteroids.

Case Report

The patient presented as an 81 year old white female with a diagnosis of Right L5 radiculopathy and Post Lumbar Laminectomy Syndrome. She had undergone multiple prior lumbar and caudal epidural injections with minimal or short-term success. The last epidural done prior to the continuous epidural gave her 90% relief of pain, but only lasted three weeks, and she then returned to baseline status. She was thus felt to be a good candidate for continuous infusion because of this transient response to prior ESIs.

Following informed consent, a 3.5 inch Touhy spinal needle was slowly advanced utilizing a retrograde interlaminar approach at L4 on the symptomatic right side. Following verification utilizing both loss of resistance, and direct visualization of contrast in the epidural space under real time AP and lateral fluoroscopic guidance, an initial bolus of 2cc of Kenalog 40mg/cc, 2cc of 1% preservative free lidocaine, and 2cc of saline were injected into the epidural space. Following this, a 20-gauge styletted radiopaque epidural catheter was advanced through the spinal needle in a caudal direction, to the L5 level. Verification of catheter placement was obtained by injection of the contrast through the catheter, which demonstrated epidural spread of the dye on AP and lateral real time fluoroscopic images.

Following this, an infusion pump which was pre-filled with 8mg of dexamethasone in normal saline for a total volume of 72cc, was connected to the epidural catheter. The pump was pre-set with an infusion rate of 1cc/hour, for a total of 72 hours of infusion. Antibiotic ointment was placed over the injection site. A tegaderm patch secured the catheter laterally along with additional surgical tape. The infusion pump was attached to the patient via a belt clip and harness.

The patient tolerated the procedure well. There were no complications to report. Discharge instructions were dispensed prior to her departure from the clinic. Prophylactic Ceftin 250mg BID for 4 days was prescribed.

At time of catheter removal 72 hours later, the patient reported a subjective 99% improvement with no reported radicular pain down her right leg. Straight leg raising was negative on the affected right side. Her post op course was otherwise uneventful, without complications or adverse reactions. She subsequently followed up with her primary treating physician.

I next saw her over seven months later. She was referred back by her primary treating physician to undergo the same procedure. At that time, the patient reported complete relief of her right leg radicular pain for over six months, with recurrent pain in the three weeks prior to this visit. She again denied any complications or adverse reactions to the procedure over the prior 7 plus months. Consequently it was felt indicated to repeat the continuous epidural procedure in light of her successful response. The identical procedure as described above was repeated. An over two years follow-up questionnaire is presented in Table 1.

Table 1. Follow Up Questionnaire to ICE Procedures

After over a two year period, the patient was given the following questionnaire. Her responses are in bold. Informed consent was obtained to utilzize this data in a published review.


  1. How many Infusion Catheter Epidural (ICE) procedures were performed on you by your physician?
  2. a. 1
    b. 2
    c. 3
    d. 4 or more
  3. Approximately how many lumbar epidural injections have you had performed? (includes caudal, interlaminar, transforaminal, nerve root block, etc.)
  4. a. 1
    b. 2
    c. 3
    d. 4 or more
  5. In general, was ICE more effective than an epidural?
  6. a. yes
    b. no
    c. not sure
  7. If you answered yes to question 3, how much more effective was ICE than an pidural with respect to degree of pain relief?
  8. a. 10%
    b. 20%
    c. 30%
    d. 40%
    e. 50%
    f. 60%
    g. 70%
    h. 80%
    i. 90%
    j. 100%
  9. If you answered yes to question 3, approximately how much longer did your pain relief last with ICE compared to a single shot epidural?
  10. a. it lasted less than an epidural
    b. it lasted about the same amount of time as an epidural
    c. it lasted 2x as long as an epidural
    d. it lasted 3x as long as an epidural
    e. it lasted 4x or longer than an epidural
  11. Did you have any significant complications or adverse reactions with ICE?
  12. a. yes
    b. no
  13. Would you recommend this technique?
  14. a. yes
    b. no
  15. Would you have ICE performed on you again if it was indicated?
  16. a. yes
    b. no
  17. Rate your overall satisfaction with ICE: (circle best response)
    0 1  2  3  4  5  6  7  8  9  10
    unsatisfied.................................... extremely satisfied
  18. Please feel free to make any additional comments regarding ICE:
    (She did not give any additional comments.)


These findings are very encouraging for administering a continuous lumbar epidural infusion of corticosteroid and normal saline via an indwelling epidural catheter/infusion pump over a three-ay period for treatment of lumbar radiculopathy. This continuous bathing of “inflamed” tissue may offer an advantage over a single epidural bolus in that the therapeutic medication is exposed to the suspect segment for a more prolonged period of time. This is presumed to increase the therapeutic effect of the medication and offers the patient a potentially greater and more lasting response without significant increased risk. This procedure may give the patient one additional safe, relatively low cost and easy to use option prior to consideration of lumbar spine surgery or other more invasive treatments and give the interventionalist one more conservative weapon in his/her arsenal to fight this difficult condition. Although the literature cited above appears to further demonstrate both the safety and efficacy of continuous epidural infusion with catheter placement, additional scientific research is necessary and would be welcomed.

Author Disclosure

Dr. Shapiro is the inventor of the continuous epidural technique described above (aka: ICE) and sold this technology to I Flow Corporation. He has received royalties and consulting fees totaling over $10,000. Dr. Shapiro holds no position in I Flow Corporation nor does he receive any other additional support.

Last updated on: January 28, 2012
close X