Adhesions in the epidural space caused by frequent epidural blocks

Repeated long-term single or continuous epidural blocks are considered to cause adhesions in the epidural space, resulting in decreased effects of the epidural block. Few reports have detailed epidural space anatomical findings and treatment for such cases [1]. We report our observations of the epidural space in a patient who underwent frequent epidural blocks.

A 41-year-old man with no significant medical history presented with a chief complaint of pain in his left leg. The top of his left foot had been run over by a car tire. No fractures or joint injuries were noted in the toes. He was diagnosed with complex regional pain syndrome (CRPS).

He underwent twice-weekly lumbar epidural blocks at his local hospital and a month-long continuous epidural block each winter, which allowed him to return to work. The needle used for a single epidural block was 20 gauge, and the catheter used for the continuous epidural block was 20 gauge. The degree of pain in his left leg according to the numerical rating scale (NRS) was 8 during activities of daily living and 3 after a lumbar epidural block was performed.

Five years after the initiation of epidural blocks, greater resistance was felt while injecting the medication into the epidural space, and it was difficult to insert the catheter. The degree of pain in the leg after the epidural block became NRS 6, indicating that effects of the epidural block were insufficient. The degree of pain in the leg during activities of daily living was not exacerbated, remaining at NRS 8. Adhesions in the epidural space caused by frequent epidural blocks were suspected, and the patient was referred to our department.

The patient had undergone epidural blocks many times at his local hospital. These comprised single epidural blocks with 10 ml 0.375–0.75% ropivacaine at the L4/5 and L5/S1 levels performed over 200 times as well as continuous epidural blocks using 0.2% ropivacaine for approximately 1 month performed multiple times. During examination, the patient exhibited spontaneous pain from the left leg to the foot accompanied by allodynia. Muscular atrophy and edema were also noted at the same site. Manual muscle testing results ranged from 2 to 3 in the tibialis anterior muscle, triceps surae muscle, gastrocnemius muscle, long extensor muscle of the thumb, and flexor hallucis longus muscle. There was no decreased perception and no other abnormal neurological findings. In the area from the left leg to the foot, plain X-rays revealed mild bone atrophy and magnetic resonance imaging (MRI) revealed mild bone and muscle atrophy. No abnormalities were noted in the lumbar spine on plain X-rays, and MRI did not reveal any abnormal findings in the intervertebral discs, vertebrae, spine, or nerve roots. T1-weighted fat-suppressed images showed uneven contrast effects of gadolinium throughout the back of the epidural space from L4 to S1 (Fig. 1). Epidurography performed from the sacral hiatus revealed blockage in the L5/S1 vertebral area. Lower limb thermography revealed decreased temperature throughout the entire left leg. We diagnosed the pain in his left foot as CRPS because it applied to continuing pain which was disproportionate to any inciting event, hyperesthesia, allodynia, temperature asymmetry, motor dysfunction, and trophic changes among the diagnostic criteria for CRPS by International Association for the Study of Pain, 2005 [2]. However, the attenuated effects of the epidural block were determined to be due to adhesions in the epidural space rather than CRPS exacerbation.

Regarding further treatment, the patient was advised that, although it might be possible to regain the effect of the epidural block by performing adhesiotomy in the epidural space, this was not a radical cure for CRPS of the left leg, and it would be difficult to achieve good-quality continuous epidural blocks over long periods if adhesions in the epidural space recurred following adhesiotomy. Therefore, he was advised to undergo spinal cord stimulation rather than long-term epidural blocks as a means of treating the pain in his leg. However, the patient requested to undergo adhesiotomy in the epidural space.

Therefore, epiduroscopy was performed. We used video-guided catheter and 0.9 mm fiberopticscope (Kobamed system, Japan Medicalnext Co., Ltd., Osaka, Japan). Contrast enhancement of the area from the sacral hiatus to the epidural space performed before epiduroscopy revealed blockage of the L5/S1 vertebral area, with no contrast of the epidural space any further toward the cranial side of this area (Fig. 2a). Endoscopic findings revealed adhesions in the epidural space in the L4 to S1 vertebral area. The adhesions were mainly formed from the soft connective tissue, and little hard scar tissue was observed. The adhesions were distributed throughout the L4 to S1 vertebral area, with many adhesions observed on the left side of the L4 and L5 vertebral area. There were few areas of hyperemia, reddening, or blood vessel proliferation (Fig. 3a). Adhesiotomy was carefully performed using the catheter tip while injecting the physiological saline solution. We were able to easily detach the adhesions, and contrast effects were achieved in the epidural space (Fig. 2b). In the epidural space toward the cranial side from the L3 vertebral area, dura mater and fatty tissue could be clearly observed and appeared to be normal (Fig. 3b). The operation time was 58 min, and the physiological saline solution used for adhesiotomy was 250 ml.

After epiduroscopy, although the degree of pain in his leg after an outpatient epidural block had recovered to NRS 4, the effects of once-weekly epidural blocks were only sufficient for 1 month. At this point, the patient realized that the treatment of the pain in his leg with epidural blocks would be difficult and requested to undergo spinal cord stimulation.

For spinal cord stimulation, an 8-pole electrode (Octrode™, Medtronic, Inc., Minneapolis, MN, USA) was percutaneously placed into the epidural space so that the tip was in the Th9 vertebral area and a nerve stimulator (Eon Mini™, Medtronic, Inc) was placed in the right hypogastric region. Thereafter, the degree of pain in his leg during activities of daily living decreased to NRS 5.

We described the epiduroscopy findings in the epidural space in which adhesions had formed because of frequent epidural blocks. Adhesions in the epidural space were noted at the epidural block puncture sites. Most adhesions involved the growth of the soft connective tissue, and little hard scar tissue was observed.