Occurrence of a spinal intradural arachnoid cyst after epiduroscopic neural decompression

Abstract

Epiduroscopy is a relatively new technique that enables visualization of structures directly within the epidural space to detect pathology, improve adhesion, and to administer local anesthetics and steroids to affected nerve roots [1,2]. However, several complications such as visual impairment, epidural abscess, and encephalopathy with rhabdomyolysis have been reported. We report a case of a symptomatic spinal intradural arachnoid cyst induced by an accidental dural tear during epiduro scopy in a patient with failed back surgery syndrome (FBSS). A 65-year-old woman presented with chronic low back pain and radiating leg pain. She complained of hip and thigh pain, and claudication improved with rest. She had undergone a discectomy and laminectomy twice at L4, L5, and S1 2 years ago. On exami nation, the straight leg raising test was positive bilaterally. The Patrick test was negative, and no vascular abnormalities were observed. Magnetic resonance imaging (MRI) revealed an epidural adhesion with postoperative changes, degenerative changes of the L4-5 disc, and a laminectomy state at L5 and S1. A variety of pain therapies were ineffective. Her illness was diag nosed as FBSS, and our plan was epiduroscopic neural decom pression. She was placed in the prone position with routine moni toring in the operating room. An 18.5-G Tuohy needle was introduced into the epidural space through the sacral hiatus under aseptic conditions and with local anesthesia. An intro ducer sheath was advanced into the sacral epidural space. A 0.9 mm flexible endoscope (3000E, Myelotec, Roswell, GA, USA), covered with a 2.9 mm steering catheter (2000, Myelotec), was introduced through the sheath and advanced cephalad into the epidural space under fluoroscopic guidance. Adhesions in the epidural space were mobilized with the tip of the instrument under careful direct vision. However, sufficient epidural adhesiolysis could not be achieved above the S1 level due to the dense adhesions and scar tissue. Epidural fat and scars were visualized, but the image was fuzzy. Further adhesiolysis was attempted. Despite slow advancement of the endoscope, there was a sudden loss of resistance at the level of L5-S1. The S1 nerve roots appeared with small blood vessels, but they were unclear. Iohexol contrast (IOBRIX 300 Ⓡ , 640 mg/ml, Accuzen) instillation under fluoroscopy demonstrated a myelogram, which confirmed that the endoscope had entered the subara chnoid space. The endoscope was retracted from the hole in the dura, and a catheter was placed into the ventrolateral aspect of the left L5-S1 intervertebral foramen under fluoroscopy via the epiduroscope working channel. Contrast agent administration revealed good ventrolateral epidural spread along the L5 nerve roots without connection to the subarachnoid space after negative aspiration of cerebrospinal fluid and blood. A test dose of 2 ml 0.4% lidocaine with 5 mg triamcinolone was injected. The patient had no adverse hemodynamic or neurologic sequelae. Thirty ml of 0.4% lidocaine and 30 mg of triamcinolone were injected into the epidural space at the end of the epiduroscopy. The total volume of saline used for epidural irrigation was 200 ml. The operation was terminated after 40 min due to her frequent complaints of low back pain and movement during the procedure. No complications developed, and the patient was