Ankylosing spondylitis is an inflammatory disease affecting multiple organ systems that commonly presents with chronic back pain beginning in the third decade of life and results in decreased axial mobility that can lead to severe functional limitations.1 Characteristics of inflammatory back pain include age of onset younger than 40 years, improvement with exercise, duration of back pain longer than 3 months, insidious onset, and morning stiffness.2 Ankylosing spondylitis is associated with arthrodesis of spinal articular and ligamentous structures. These changes result in significant axial and hip pain, and in severe cases, complete obliteration of the interlaminar spaces.
Complex regional pain syndrome (CRPS) is a neuropathic pain condition typically affecting a single extremity that can occur spontaneously but usually arises after trauma or surgery. Trophic, motor, and vasomotor signs confirm the diagnosis.3 The treatment algorithm of CRPS includes spinal cord stimulation (SCS) for patients remaining refractory to more conservative management, such as physical therapy (PT), medical management, and sympathetic blockade.4
We present a case of successful SCS in a patient with ankylosing spondylitis and midthoracic to sacral spinal calcification, the so-called bamboo spine. He also suffered from CRPS of the left lower extremity. After failure of conservative therapy (PT, medical management, and lumbar sympathetic blocks), a trial of SCS was performed with 2 epidural electrodes advanced to the thoracic level via the sacral hiatus. Although successful treatment of axial pain associated with ankylosing spondylitis has been achieved with SCS,1 use of the sacral hiatus for epidural access for electrode placement or for its concurrent treatment of CRPS pain has not been described. In addition, while SCS electrodes have been threaded through the sacral hiatus for sacral nerve stimulation, there are no reports of thoracic electrode placement using this approach.5
Our institution’s IRB reviewed the case report and gave written permission for the authors to publish this report. Verbal consent was obtained before this manuscript preparation by author ACY; in addition, IRB approval was given before manuscript preparation.
A 42-year-old man with a history of ankylosing spondylitis, CRPS type 2 of the left lower extremity, bilateral sacroiliac arthropathy, and chronic bilateral hip pain (status after bilateral total hip arthroplasty [THA]) presented for management of his chronic pain. This patient’s CRPS developed after his left THA. After THA, the patient developed hyperesthesia, color changes, and decreased left lower extremity strength and demonstrated allodynia and muscular weakness on examination, satisfying the criteria for CRPS. An electromyogram indicated that he had a left femoral neuropathy that further classified his CRPS as type II. He had undergone PT and a series of left lumbar sympathetic blocks that provided moderate, short-term relief from his CRPS-mediated pain. His medical management included topiramate, celecoxib, hydrocodone/acetaminophen, and a compounded formulation of topical lidocaine, gabapentin, ketamine, and clonidine for management of his neuropathic pain. Despite these therapies, he continued to suffer from allodynia and hyperalgesia of the left leg. Spine radiographs (Fig. 1) revealed extensive interspinous ligament calcifications with syndesmophyte formation, straightening of lumbar lordosis, and facet fusion extending from the midthoracic spine to the sacrum. Preoperative psychological assessment revealed no barriers to a favorable SCS outcome. Given the extent of his spinal pathology, the high risk of aborting the procedure was communicated to the patient before commencement of the trial.
A wide area, including the back and buttock area, was prepped. Prophylactic antibiotics were administered. After failure to access the epidural space at the L1-L2, L2-L3, and L5-S1 lumbar levels despite fluoroscopic guidance, entry was achieved at the sacral hiatus via the 14-gauge Boston Scientific introducer needle. A 70 cm subcompact, 8-electrode lead was placed at the superior end plate of the ninth thoracic vertebra (Fig. 2, A and B). Satisfactory response to stimulation of the left lower extremity and low back was obtained intraoperatively.
The lead was then tunneled from the entry site at the sacral hiatus to the right flank and anchored using tapered anchor and 2-0 silk suture. The midline wound was approximated with 3-0 nylon suture. SurgiSealTM (Adhezion Biomedical, Wyomissing, PA) was placed over the midline entry site, followed by a sterile dressing of TegadermTM (3M, St. Paul, MN). The lead exit site was dressed with a chlorhexidine gluconate-impregnated sponge (BioPatch®, Johnson & Johnson, New Brunswick, NJ) and Tegaderm™.
During the 7-day trial period, he received antibiotics and was instructed to not shower to minimize the risk for infection. On return to the clinic on postoperative day 5 for evaluation, the patient reported paresthesias in the area of pain and a 50% to 75% improvement in his left lower extremity pain, as well as his chronic low back and hip pain. The trial lead was removed under fluoroscopy on postoperative day 7 without incident.
He returned 1 week later for permanent implantation of a SCS system. Surgical positioning was as before. The gluteal crease and area overlying the sacral hiatus were sterilely prepped and draped, and prophylactic antibiotics were administered. A mixture of 1% lidocaine and 0.5% bupivacaine was used to anesthetize the skin and subcutaneous tissue over the sacral hiatus, and a 14-gauge Boston Scientific introducer needle was placed through the sacrococcygeal ligament. A soft guidewire was then advanced through the needle and fluoroscopically confirmed to be in the sacral epidural space. A 70 cm 16-contact Boston Scientific Infineon™ (Boston Scientific, Natick, MA) lead was threaded over the guidewire in the dorsal epidural space to the level of T-8, slightly offset intentionally to the left of midline. A second introducer needle was brought onto the field and a second Infineon™ lead placed to the same level, slightly right of midline (Fig. 3, A and B). Intraoperative complex neuroanalysis indicated the patient had stimulation coverage of both lower extremities and coverage of his low back. Consequently, a stab incision was made at the entry sites of both introducer needles through which the leads would be tunneled.
A location in the left flank was chosen for implanting the pulse generator (IPG). The skin was anesthetized with local anesthetic, and a subcutaneous pocket was then created via a transverse incision. A Boston Scientific 14-gauge 6.5-inch introducer needle was used to tunnel a tract from the pocket to each stab incision. The leads were passed retrograde through the needles, and the needles were removed. Once in the pocket, each lead was connected to a Boston Scientific SpectraTM IPG (Fig. 3C). After connection, impedances were confirmed to be within acceptable limits. Both the IPG pocket and stab incisions over the sacral hiatus were irrigated with a solution containing antibiotics. The IPG was secured in the pocket with nonabsorbable suture and the transverse incision closed by suture in 3 layers. The 2 stab incisions were closed with a single interrupted nylon suture. All incisions were covered with SurgiSeal™ and Telfa™/Tegaderm™. Of note, no anchoring mechanism was used at the entry site of the leads. There were no dense fascial attachments to use as an anchoring site, and the subcutaneous tissue overlying the lead entry site was thin. Implanted hardware would likely cause discomfort and could result in erosion through the skin. The patient was admitted overnight for observation and acute pain management. He was discharged the following day with a prescription for 10 days of oral cephazolin for postoperative infection prophylaxis.
On subsequent follow-up to the pain clinic, he has reported near complete resolution of his CRPS, chronic hip, and low back pain. He has not developed any signs or symptoms of infection postoperatively. The pattern of stimulation captured in the operating room had not significantly changed, and LeadSync™ technology has indicated that lead migration had not occurred.
Before a trial of stimulation, our patient was sedentary, as evidenced by considerable weight gain, after development of CRPS. He ambulated with crutches or used a wheelchair. He ventured little outside his home, because he was limited by pain that was uncontrolled with conventional therapies. During his trial, he was able to ambulate for both greater distances and for longer durations with and without crutches. He did not use a wheelchair during his trial. He was able to decrease his oral opioid requirement by 50% during the trial. His verbal pain scores decreased by 70%. After permanent implant, functional improvements were maintained, and oral opioids were further tapered to a lower dose. The writing of the original manuscript was 6 months after his permanent SCS implant.
Hardware-related failures are the most prevalent complications associated with SCS. In a retrospective review of 707 cases, Mekhail et al.6 recorded an incidence of 38% that included lead migration (22.6%), connection failure (9.5%), and lead breakage (6%). Lead migration occurs when mechanical stress causes lead displacement from its original intended position, often resulting in loss of paresthesia coverage.7 Technical considerations that decrease the risk of lead migration include epidural entry at or above the L1-2 intervertebral space for placement of thoracic leads. Below this level, the lumbar spine has increased mobility that may cause stress along the length of the lead. It is also likely that the longer the lead segment distal to the anchor, the greater the risk that the lead will migrate. The decreased mobility of our patient’s spine may have been of benefit in this regard and may have reduced the risk of lead migration. However, because anchors were not placed at the epidural exit sites before tunneling to the IPG, our patient was otherwise at a greater risk for lead migration. These issues were considered before lead placement and led to the decision to use 16-contact leads. It was assumed that some degree of lead migration would occur at some point, and with optimal stimulation achieved with the contacts in the center of the lead, any movement of the leads, cephalad or caudad, could be mitigated with reprogramming rather than reoperation.
The risk of infection after lumbar and thoracic lead placement ranges from 3.3%8 to 4.5%.6 Our sacrococcygeal entry site for the leads increased the likelihood of exposure to gut flora, thus presumably increasing the risk of infection. Meticulous preparation, adherence to sterile technique, use of antibiotic irrigation, and perioperative antibiotics were the safeguards used to minimize this risk.
Limited access to the epidural space poses a technical challenge to pain medicine physicians for SCS lead implantation. This first report of the functioning SCS thoracic epidural electrodes inserted via the sacral hiatus for CRPS brought additional improvement in our patient’s chronic axial pain related to ankylosing spondylitis and suggests this approach may be beneficial in patients with this challenging combination of pain disorders.
The exact mechanism of analgesia from SCS is unknown. At its conception, stimulation of the dorsal columns of the spinal cord were thought to alter resting membrane potentials, resulting in paresthesias in the area stimulated or hyperpolarizing neurons resulting in decreased afferent pain signaling. However, current research points to more complex processes that may account for analgesia. These include receptor activation/deactivation, altering chemical mediators, and modifying vascular tone in stimulated areas.9–12 In our patient, it is hard to say what exactly led to the alleviation of both of his pain generators, ankylosing spondylitis, and CRPS. What is interesting is that each process has differing pain mechanisms: ankylosing spondylitis is primarily nociceptive, and CRPS primarily neuropathic. In this case, SCS was able to successfully treat both types of pain. Our initial goal was to treat his CRPS pain; it came as some surprise that his ankylosing spondylitis pain was also relieved with stimulation.
1. Deshpande KK, Wininger KL. Spinal cord stimulation for pain management in ankylosing spondylitis: a case report. Neuromodulation. 2009;12:54–9
2. Calin A, Porta J, Fries JF, Schurman DJ. Clinical history as a screening test for ankylosing spondylitis. JAMA. 1977;237:2613–4
3. Bruehl S. An update on the pathophysiology of complex regional pain syndrome. Anesthesiology. 2010;113:713–25
4. Stanton-Hicks M, Baron R, Boas R, Gordh T, Harden N, Hendler N, Koltzenburg M, Raj P, Wilder R. Complex Regional Pain Syndromes: guidelines for therapy. Clin J Pain. 1998;14:155–66
5. Park CH, Kim BI. Sacral nerve stimulation through the sacral hiatus. Korean J Pain. 2012;25:195–7
6. Mekhail NA, Mathews M, Nageeb F, Guirguis M, Mekhail MN, Cheng J. Retrospective review of 707 cases of spinal cord stimulation: indications and complications. Pain Pract. 2011;11:148–53
7. McGreevy K, Williams KA, Christo PJ. Cephalad lead migration following spinal cord stimulation implantation. Pain Physician. 2012;15:E79–87
8. Kreis P, Fishman S Spinal Cord Stimulation: Percutaneous Implantation Techniques. 2009 New York, NY Oxford University Press
9. Linderoth B, Gazelius B, Franck J, Brodin E. Dorsal column stimulation induces release of serotonin and substance P in the cat dorsal horn. Neurosurgery. 1992;31:289–96 discussion 296–7
10. Linderoth B, Stiller CO, Gunasekera L, O’Connor WT, Ungerstedt U, Brodin E. Gamma-aminobutyric acid is released in the dorsal horn by electrical spinal cord stimulation: an in vivo
microdialysis study in the rat. Neurosurgery. 1994;34:484–8 discussion 488–9
11. Sato KL, King EW, Johanek LM, Sluka KA. Spinal cord stimulation reduces hypersensitivity through activation of opioid receptors in a frequency-dependent manner. Eur J Pain. 2013;17:551–61
12. Tanaka S, Barron KW, Chandler MJ, Linderoth B, Foreman RD. Role of primary afferents in spinal cord stimulation-induced vasodilation: characterization of fiber types. Brain Res. 2003;959:191–8