CHAPTER 117 Surgery for Sacroiliac Joint Syndrome
INTRODUCTION
Surgical treatment for sacroiliac pain can be considered when other therapeutic options have failed, and when the patient’s symptoms or functional limitations are of sufficient magnitude to balance the attendant risks. In the course of a general pain or spine practice a relatively small number of patients will be encountered who fall into this group. For patients with recalcitrant symptoms, it is the practitioner’s responsibility to educate patients about all potentially effective treatments, including those that lie outside the scope of a particular physician’s practice. Although not widely appreciated by most practitioners, extensive literature exists that reports satisfactory results from surgical treatment of sacroiliac pain in appropriately selected patients. This chapter will provide a review of the relevant literature and will describe the approaches to surgical treatment of patients with pain of sacroiliac origin.
PATIENT SELECTION
As is true for most therapeutic interventions, the most important factor in the ultimate success of surgical treatment of sacroiliac pain is proper patient selection. A patient for whom surgical treatment is a consideration should undergo a thorough evaluation that clearly identifies the sacroiliac joint as a pain generator. As coexistent spinal pathology is not uncommon,1 the preoperative evaluation must include a detailed search for other pain generators that may mimic or overlap symptoms arising from the sacroiliac joint. Plain films and magnetic resonance imaging (MRI) studies of the lumbar spine should be obtained, and additional studies such as provocative discography, selective root blocks, median branch blocks, and other studies described elsewhere in this text should be utilized when appropriate. Physical examination maneuvers alone have been demonstrated to be inadequate in the diagnosis of a sacroiliac origin of pain.2,3 Radionuclide imaging is also of limited usefulness in confirming a diagnosis of sacroiliac-mediated pain, due to its very low sensitivity.4 Diagnosis of pain of sacroiliac origin requires an unequivocally positive result from a flouroscopically or computed tomography (CT)-directed injection of local anesthetic into the synovial portion of the sacroiliac joint. False-positive results can be minimized by the use of a double block as described by Maigne et al.5 The author believes that the criteria of Maigne et al. should be modified slightly in that these authors recommended that the patient should report at least 75% reduction of pain for an injection to be considered positive. While percentage reduction of pain is by its nature a very subjective ‘measurement,’ in the author’s experience the response that best predicts a positive outcome from surgical treatment is a response of ‘near complete’ or ‘greater than 90%’ relief of pain. Coexistent pathology can also make any attempt at quantification of pain relief by an intervention that removes only a component of the patient’s symptoms very difficult to interpret, as some patients are more able than others to discriminate different components of their pain complaints. Ultimately, clinical judgment plays a crucial role in the assignment of all diagnoses relating to pain generators in any particular patient. The practitioner should, however, avail himself or herself of all diagnostic information necessary to formulate the best conclusions regarding the sources of pain generation before making a recommendation for surgical treatment of any type.
Only patients who have failed to improve after undergoing a thorough trial of conservative treatment should be considered candidates for surgical treatment. No consensus exists, however, on what constitutes a thorough trial of conservative treatment. Mooney has described a defined course of physical therapy with demonstrated effectiveness in many patients.6 Prolotherapy has advocates, as does repeated corticosteroid injections. The practitioner should be acquainted with the various alternatives for nonsurgical treatment, and be satisfied that nonsurgical measures have been exhausted before making a recommendation for surgical treatment.
SURGICAL ALTERNATIVES
Posterior arthrodesis
Historical review
Goldthwait and Osgood7 reported on the association of low back pain and sacroiliac laxity in pregnant women. They drew attention to the fact that the sacroiliac joints demonstrate motion in nonpathologic conditions and hypothesized that ligamentous laxity associated with pregnancy might lead to hypermobility and associated pain. Albee8 reported on his dissections of 50 pelvic specimens and drew attention to the fact that the sacroiliac joints were synovial joints and not ‘synchondroses’ as many physicians of that period assumed. Baer9 reinforced the fact that the sacroiliac joints were synovial joints and recommended manipulation under anesthesia as treatment for chronic sacroiliac pain. Gaenslen reported on his surgical treatment first in 1921,10 and later reported on 9 patients’ outcomes at varying intervals after surgery.11 Gaenslen’s operation consisted of a posterior approach to the joint by splitting the ilium to reflect a bone flap with the gluteal musculature attached. A window through the remaining cancellous portion of the ilium was then made to expose the synovial portion of the joint. Curettage of the interior of the joint was carried out. He reported good or very good results in 7 of 9 patients (78%). Smith-Petersen and Rogers12 reported on 26 patients who had undergone sacroiliac arthrodesis for chronic sacroiliac pain (one of whom was Smith-Petersen’s wife). Smith-Petersen’s technique differed slightly from Gaenslen’s technique in that the gluteal musculature was reflected and a transiliac window was made to gain access to the synovial portion of the joint. Decortication of the interior of the joint was carried out with curettes and gouges, and the window was replaced and countersunk across the joint after removal of the cartilage and subchondral bone from the window fragment, allowing opposition of cancellous surfaces between the sacrum and ilium. He reported clinical success in 23 of 26 patients (89%), with radiographic union in 95%. Twenty-five of 26 patients (96%) returned to their prior occupation. Willis Campbell13 reported an extra-articular method of fusion, in which the posterior ilium and adjacent sacrum were exposed and decorticated, and bone graft was placed into the gutter dorsal to the joint proper. He reported that five of seven patients had a successful outcome, with two patients being too close to surgery to assess the outcome at the time of his publication.
After this series of publications describing differing techniques, all with successful outcomes, there was a surprising absence of publications on both the diagnosis and the surgical treatment of sacroiliac pain. The most likely reason for this was the popularization of the diagnosis of disc pathology and its association with sciatica. Although Mixter and Barr14 are frequently given credit for ‘discovering’ the herniated lumbar disc, several publications discussing both the pathology15,16,17,18 and surgical treatment19 pre-date the Mixter and Barr article. Nonetheless, the influence of the Mixter and Barr article was apparently of sufficient magnitude to largely eliminate the sacroiliac joint from attention of authors in neurosurgical and orthopedic journals. In the 40 years following the Mixter and Barr article, only a handful of publications in the surgical literature make mention of the sacroiliac joint as being a source of low back pain and sciatica, with the exception of papers reporting on septic arthritis and sacroiliitis arising from rheumatoid disorders. Extending the work of Steindler and Luck,20 Haldeman and Soto-Hall21 recommended diagnosis of sacroiliac pain by injection of procaine, although their injection technique involved injecting large volumes (20–30 cc) of 1% procaine into the ligamentous portion of the joint. No mention of surgical treatment was made in their article. Norman and May22 brought attention to the fact that sacroiliac pain could mimic symptoms arising from intervertebral disc lesions, a fact largely unrecognized until the refinement of spinal diagnostic techniques in the 1990s.23
Orthopedic and neurosurgical journals remained silent on the issue of surgical treatment of sacroiliac pain until 1987, when Waisbrod and colleagues reported on 21 patients who underwent posterior uninstrumented sacroiliac arthrodesis for chronic sacroiliac pain.24 Diagnostic evaluation of this cohort included a provocative test, radiographic evaluation with plain films and CT images of the sacroiliac joint, radionuclide imaging, and several psychological questionnaires. The psychological screening was not utilized in the early part of the study period (1981–84). The provocative test consisted of injection of 10% NaCl solution into the ‘dorsocaudal’ portion of the joint. To be considered a positive test, this intervention was required to exactly reproduce the patient’s typical pain pattern. One to 2 cc of local anesthetic was also injected, and presumably this alleviated the previously aggravated symptoms, although the authors were not specific on this point. Follow-up ranged between 12 and 55 months. To be considered for surgery, the patients had to have positive findings on radiographic, nuclear medicine, and provocative testing. Patients with psychological disturbances were excluded regardless of other findings in the later part of the study (1984–86). Patients were divided into satisfactory and unsatisfactory outcomes. A satisfactory outcome required reduction of symptoms of at least 50%, no need for analgesics, and resumption of their preoperative occupation. They reported 50% satisfactory results overall, although this included six patients who would not have been operated upon in the later portion of the study based on psychological exclusion criteria. When they excluded the six patients with psychological disturbances, they concluded that a satisfactory result could be expected in 70% of cases. Moore reported on 13 patients treated with instrumented posterior sacroiliac arthrodesis for chronic sacroiliac pain at the 1992 North American Spine Society meeting.25 Patients were offered surgical treatment when conservative treatment had failed and the patients had unequivocal relief of their typical symptoms after injection of local anesthetic into the synovial portion of the joint under CT or fluoroscopic guidance. Six of the patients had prior failed lumbar spine surgery. He reported 10 excellent, one fair, and two poor results. This same group of patients was followed up again, with successively larger groups in 1995, 1997, and 1998.1,26,27 In the 1998 review, Moore reported on 59 patients who had isolated sacroiliac joint pain and no coexistent spinal pathology or prior spinal surgery.27 Chart reviews and phone interviews were carried out, and outcome variables included subjective pain relief, satisfaction with outcome, surgical complications, and incidence of pseudoarthrosis. All phone interviews were carried out by an independent research assistant (i.e. the surgeon did not carry out any of the interviews). Of these patients, 53 (89.8%) had a satisfactory outcome, and six (10.2%) were considered failures. Of the six patients who were failures, four had a pseudoarthrosis. The other two patients were clinical failures despite evidence of solid arthrodesis on fine-cut CT scanning. Complications were limited to one superficial wound infection, one intraoperative fracture into the sciatic notch during creation of the transiliac window, necessitating intraoperative repair which healed without incident. One patient required reoperation to retrieve a screw that penetrated the anterior cortex of the sacrum, causing mild radicular dysesthesia. Except for this latter case, there was no incidence of neurologic complications and there were no cases of visceral or vascular injury.
Keating et al.29 reported on follow-up of 28 patients who underwent posterior arthrodesis of the sacroiliac joint for chronic sacroiliac pain. Statistically significant improvement between preoperative and postoperative pain was observed and patients remained improved up to the minimum 1 year follow-up period. Berthelot et al.30 reported on posterior sacroiliac arthrodesis in the treatment of two patients with recalcitrant sacroiliitis related to spondyloarthropathy. Both patients were significantly improved at follow-up of greater than 2 years. Belanger and Dall31
