Unstable posterior pelvic ring fractures and dislocations are potentially life-threatening injuries associated with highenergy trauma. Early definitive management decreases the risk of immediate complications caused by lengthy decubitus and long-term complications caused by chronic pain and gait dysfunction. Management might include a conservative, nonsurgical approach that requiring the patient to stay in bed for 1 month. The classic operative therapy option for sacroiliac joint disruptions consists of open reduction and internal fixation. It allows for early ambulation but carries substantial risks, such as significant blood loss and iatrogenic nerve and vascular injury [1]. Secondary to the progress in interventional imaging, new therapeutic options have been developed that decrease potential complications because they are minimally invasive and they decrease patient rehabilitation time. As a diagnostic modality, computed axial tomography (CAT) allows precise evaluation of the degree of sacroiliac reduction that must be performed. Moreover, the use of CAT enables easy positioning of screws across the sacroiliac joint, thus avoiding nerve and vascular damage. Several reports [2–6] have already described the use of CAT for guidance and fixation of sacroiliac disruptions. We report our clinical experience of 20 patients using CT-guided percutaneous fixation for posttraumatic unilateral sacroiliac disruption. Our purpose was to investigate our technique as well as its safety, outcomes, and long-term results. The long-term outcome of exclusively CAT-guided sacroiliac joint disruptions has not yet been evaluated in others studies. Most of the existing publications deal with fixation of the pelvic ring, which may include associated injuries, such as fracture of the acetabular roof and sacrum or lesions of the iliopubic or ischiopubic rami. Materials and Methods Unstable fractures of the pelvic ring and, more particularly, unstable sacroiliac joint disruptions result from highenergy trauma (road-traffic accident, self-inflicted injury). These fractures are generally managed with fluoroscopyguided reduction and internal fixation, which had tended to replace the traditional open reduction and internal fixation (ORIF). Fluoroscopy-guided fixation offers the advantages of minimally invasive procedure: decreased blood loss and early ambulation. However, there is also a significant limitation: limited visualization of the sacral neuroforamina, particularly in obese patients.This procedure was first described in 1987 by Ebraheim et al., [3] who reported early results obtained in three patients with stabilization of pelvic fractures. It was not until 4–5 years later that Nelson and Duwelius [2] and Gay et al. [7] confirmed the real feasibility of the procedure and reported consistently good results in unstable fractures of the pelvic ring. Although many articles on a minimally invasive approach are available in the orthopaedic and emergency literature, surprisingly few reports have been published by interventional radiologists on CAT guidance. More recently, in 2007, Sciulli et al. [6] reported on their own experience with this procedure. All of these studies demonstrate that CAT guidance is a safe and effective method to guide screw placement in sacroiliac joint disruption and results in consistent outcomes and low complication rates. To obtain a more homogenous study population, only patients with sacroiliac joint disruption without associated acetabular, sacral, iliopubic, or ischiopubic lesions were included in our study. Study Method The Institutional Review Board approved the study, and informed consent was obtained from all patients who underwent the procedure. Between November 2001 and May 2004, pelvic ring fixation was performed in 74 patients (45 men and 29 women) age 14–94 years. Only 20 patients of these patients who had sacroiliac joint disruption without associated pelvic injuries (i.e., fractures) were included in our study (Table 1). Classification of Pelvic Disruption Pelvic disruption was classified according to the Tile classification [8] as stable (type A), vertical stable with rotational instability (type B), and concurrent vertical and rotational instability (type C).Patient-Related Data Only patients with unilateral type B (n = 9) or C disruption (n = 11) were candidates for CAT-guided fixation and were included in our study. Patients with associated acetabular fracture or bilateral sacral injury were excluded. In total, 8 women and 12 men, ranging in age from 18 to 54 years old (mean 38.6), were treated. Before intervention, every patient was assessed by a team of orthopaedic surgeons and interventional radiologists specialized in spine interventions. Patient selection was made by referring orthopedic surgeons and the referring interventional radiologist, who decided which patients could safely be treated using CAT-guided fixation. For all patients, CAT guidance was chosen for the advantages it offered with regard to accurate screw placement in contrast to C-arm fluoroscopic guidance. None of our patients had significant morbid obesity, which would have justified the use of CAT guidance in and of itself. In fact, CAT guidance is routinely used by our orthopaedic team for the stabilization of sacroiliac joint disruption. Time from diagnosis to CT-guided surgery was 2–4 days. Data acquisition was prospective, and clinical evaluation of the patients was performed by the orthopaedic surgeons. Follow-up was accomplished through retrospective review of medical charts notes from the follow-up visits to the rehabilitation centre. The follow-up visits took place between 36 and 48 months after the procedure (average 3.8 years). All patients underwent high-resolution CAT scan (Lightspeed Ultra; GE Medical Systems, Milwaukee, WI) of the sacrum after 2 months and again 3 years. Technique Before the image-guided procedure, sacroiliac joint disruptions are reduced with closed traction. All reductions are performed by preoperative skeletal traction or manually by the surgeon. The disruption is defined as being adequately reduced if vertical displacement between the sacrum and the iliac bone is\10 mm. The patients in this series had exclusively unilateral sacral screws placed. The procedure took place in an interventional CAT scan room, which is situated in the radiology department, with dual guidance (CAT scanner and C-arm fluoroscope positioned at the gantry of the CAT scanner). As described previously, all of the interventions are performed in tandem by a radiologist and an orthopaedic surgeon. Contraindications to CAT-guided iliosacral screw placement are the same as those for ORIF and fluoroscopically guided placement: haemodynamic or severe cardiopulmonary instability, local infection or skin breakdown, and severe soft-tissue damage at the insertion site. The procedures require standard prone positioning of the patient in the CAT unit under conscious sedation. CAT scan data are obtained using an eight-row CAT scanner. The sedated patient undergoes high-resolution CAT acquisition of the sacroiliac joint before intervention to evaluate the geometry of the disruption and to plan the procedure. Radio-opaque catheters taped to the skin served as a reference point for defining the entry site. Our scanning protocol includes 1-mm slice acquisition of the pelvis, volume rendering, and multiplanar reconstructions of the sacroiliac joints. An axial CAT image is used to plan the procedure, i.e., the distance and angle from the skin to the target structures were defined, and the optimal path for placement of the screws across the sacroiliac disruption is traced. Three measurements are obtained from the entry point: the angle of approach measured from the horizontal, the depth of soft tissue to be penetrated from the skin surface to the posterior aspect of the iliac bone, and the distance from the outer surface of the iliac bone to the desired depth of penetration of the surgical screw. All measurements are obtained directly from the CAT monitor. The entry points are located and marked on the patient’s skin according to the previously positioned radio-opaque catheters. At this point, the patient’s skin is vigorously disinfected. Surgical preparation of the patient and surgical dressing of the operators and the table are mandatory. This procedure can be performed using simple devices employed daily by surgeons and interventional radiologists alike. After administration of local anaesthesia to the entry points, a 20-gauge needle is placed along the previously defined path on the planning image. Subsequent CAT scans guide needle position and progression. After making a small skin incision, an 11-gauge vertebroplasty guiding cannula (Osteosite Bone Biopsy Needle Set; Cook, Bjaeverskov, Denmark) is placed over the needle shaft inside the needle and moved down to the bone (lateral surface of the iliac bone; Fig. 1). After removing the 20-gauge needle, a 2.0-mm Kirschner guidewire (Synthes United States, Paoli, PA) is placed inside the cannula and drilled manually through the sacroiliac disruption (Fig. 2). Successive CAT scans performed during different steps of the procedure confirm good progression and position of the guidewire and insertion at an appropriate angle. Once the Kirschner wire transfixes the sacroiliac joint, a pilot hole is drilled through the articulation and into the body of the S1 vertebra (Fig. 3). A screw tract is then drilled over the guidewire. The correct length of this was calculated on the initial planning CAT image. The optimalscrew position is achieved when it reaches at least the midbody of the S1 vertebra. While leaving the guidewire in situ, the trocar is then removed and replaced by a hollow screw (8-mm Stryker Trauma Asnis III cannulated screws; Switzerland), which is pushed through the sacroiliac joint, thus fixing the disruption (it is pushed through the sacroiliac joint) (Fig. 4). The same procedure was performed to introduce a second Kirshner wire and screw, thereby allowing better fixation of the disruption (we did all work at one level and then moved to the next level; Fig. 5). At the end of the procedure, a final CAT scan confirms the correct position of the two screws, which enables us to evaluate the reduction of the disruption (Fig. 6A, B). Because each patient received 2 screws, a total of 40 screws were placed in 20 patients.Results The average time for screw placement in this study was 65 min (range 35–90). Time was measured from the moment the patient was placed on the CAT table to the moment at which the patient was ready to be removed from the CAT table. The most time-consuming part of the procedure was the acquisition of multiple test images to ensure proper screw placement. The effective dose per the CAT scanner in SmartStep mode was measured at 15 nGy/series of three slices, and the average SmartStep duration per procedure is 12.89 seconds (range 8–36).The average SmartStep-related effective dose per procedure is 193.35 nGY (range 90–540, i.e., approximately 0.0002 mSv). All patients in that series had unilateral sacroiliac disruption, and 19 reductions (95%) of disruption were satisfactory (\1 cm). One patient (5%) had superior displacement of the left hemipelvis[10 mm after fixation. The required skin incision was approximately 1.0 cm in length. Wound healing was excellent, and no scarring was noted. No complications due to anesthesia were encountered. Furthermore, no other complications—such as death, dural tears, neurologic worsening, deep wound infection, deep veins thrombosis, or screw breakage— occurred during or after surgery. Estimated blood loss was considered minimal because neither suctioning nor cautery nor surgical ligation of vessels was required during the interventions. Moreover, the hospitalization time was very short, 4 days on average. All patients tolerated the procedure well and without incident. Clinical follow-up, including visual analog pain scale scores and use of pain medication, took place at 2 and 6 months and again at 1, 2, and 3 years. All patients in this study had a successful outcome, which was judged according to how much pain they experienced and how quickly they resumed normal activity after the procedure. Twelve of 16 patients were able to return to work at postoperative month 2. The other 4patients were unemployed. In this series, all of the patients, including the patient who had a residual displacement [1 cm, had VAS scores of zero as early as postoperative month 2. Pain medication was unnecessary because after fixation of the sacroiliac disruption, patients were totally pain free. One patient (5%) had degenerative sacroiliac joint syndrome, which was confirmed by CAT scan 6 months after surgery (Fig. 7). He was successfully treated by conservative treatment (i.e., nonsteroidal antiinflammatory medication; NSAIDS). No patient showed radiologic or clinical evidence of instability of the sacroiliac joint or screw migration. Postoperative follow-up, which was performed at 1, 2, and 3 years in our rehabilitation department, showed stable results with time. All pain disappeared in 19 patients (95%) without the use of medication. The patient who had a diastasis [1 cm wore orthopaedic insoles to correct the limb length discrepancy and did not need any medication. The patient who had a degenerative sacroiliac joint disease was treated with NSAIDs and cortisone injections at the sacroiliac joint, which provided excellent pain relief. CAT scan of the whole pelvis was routinely performed at 2 months and again at 3 years in all patients. Imaging was not used during periodic follow-ups except in the patient who had painful sacroiliitis. None of the patients showed radiologic evidence of screw migration (at 2 months or at 3 years). Discussion Currently, treatment of sacroiliac disruption generally consists of either conservative treatment by bed rest or definitive management by ORIF. Because of the lengthy period of immobilization, conservative treatment can result in complications such as deep venous thrombosis and pulmonary embolism. Because of prolonged immobilization and thereby delayed rehabilitation, chronic pain and gait dysfunction are often associated with conservative treatment of pelvic disruption. ORIF, the current treatment of choice for posterior pelvic ring disruption with instability has significant disadvantages. These include relatively ‘‘blind’’ placement of the fixation screws, infection, exsanguinating haemorrhage, and high wound complication rates. We believe that fluoroscopy does not offer significant clarity in defining the posterior structure. Advantages of CAT-guided sacral fixation are direct visualization of the course of the screws and absence of significant wound complications. This technique provides superior visualization of the nerve roots and sacral canal compared with fluoroscopic methods. Because open surgical reduction allows fast and definitive management, the risks of prolonged decubitus are decreased. Because surgery permits early ambulation, patients can enter faster into rehabilitation programmes, thereby decreasing further long-term complications associated with pain and functional disabilities. Classical surgical intervention is associated with intraoperative risks, such as extensive blood loss, scar formation caused by the open surgical approach, dissection of muscles with associated wound-healing problems, pain at the posterior scar, infection, and potential nerve and vascular damage [9]. Use of fluoroscopy for guidance makes it extremely difficult to define the posterior structures of the pelvis. A critical determination in this regard is the location of the sacral foramina, through which the sacral nerves travel. This problem is compounded in extremely obese patients. Another common feature in previous reports was the relative lack of complications associated with the CAT-guided procedure. Complications from conventional surgical exposure and screw placement have been reported to result in mortality rates as high as 10% and morbidity rates as high as 52%. Nelson and Duwelius [2] reported that ORIF was associated with a high incidence of paresthesia in the ipsilateral extremity, gait disturbances, lower back pain, and neurologic abnormalities from nerve damage. Because of the progress made in interventional radiology and the rapidity of multislice CAT, new therapeutic options have been proposed, including percutaneous CATguided fixation of pelvic disruptions. This technique has been described by Gay et al. [7] for screw fixation of acetabular fractures and by Nelson and Duwelius [2] for sacroiliac disruption, both with excellent immediate results. Similarly, Ziran et al. [4] showed that the only significant complication was fracture of a screw and 5-mm displacement in a noncompliant patient who began immediate weight bearing. There were no other significant complications.There were no technical difficulties, no misplaced screws, and no cases of infection or nonunion. All patients stated that they would choose to have the CAT-guided procedure again rather than a procedure requiring general anaesthesia. CAT-guided intervention allows the radiologist and surgeon to obtain a real-time view of the operative site, thereby permitting a safe and successful procedure. The screws are placed precisely, thus injury to the sacral nerves and spinal canal is avoided. Preoperative CAT and CAT guidance during the intervention enable us to plan the procedure, to measure directly the required screw length, and to determine accurate position of the operative material. Different types of guidance have been studied on human cadavers: three-dimensional navigation for the simulation of screw fixation (stereotactic CAT) on pelvic fractures [10] or fluoroscopically based navigation systems [11]; however, those techniques are not realized in real time. We believe that CAT guidance permits more perfect placement of the screw with precise localization of the sacral nerves. Other investigators have used the scanner in continuous fluoroscanner mode and reported that the irradiation doses are high ([1 mGy/s). Thus, using the scanner in Smart- Step mode seems to be more advantageous than the conventional scanner use. It allows faster simultaneous acquisition of several reconstructed slices, which are displayed almost instantly on the examination room screens. Thus, guidance is more accurate given that the radiologist can adjust the needle’s trajectory more easily. It is true that this technique does not allow real-time guidance of the needle, as is the case for the use of the fluoroscanner in continuous mode; however, this seems to be a good compromise in view of the differences in patient exposure to radiation. Tonetti [12] showed that computer-assisted guidance allows better placement of iliosacral screws, with no outside- bone trajectories and a lower radiation exposure (an average radiation time of 0.35 min/patient and 0.14 min/ screw in the computer-assisted guidance group and an average radiation time of 1.03 min/patient and 0.6 min/ screw in the fluoroscopic group). The presented work shows that the procedure is performed promptly and safely. The percutaneous approach avoids muscle dissection to approach of the site of interest. Thus, scar-formation is minimal and intraoperative and postoperative risks are decreased. Because of the minimal blood loss, transfusions and the associated risks become unnecessary. Because patients can stand up 24 h after the procedure, rehabilitation can begin early, and thus the loss of productivity in this often-young patient group is decreased. To our knowledge, this is the first study to analyze the long-term outcome of patients with unilateral sacroiliac disruption treated by percutaneous CAT-guided screw positioning. This study evaluated CT-guided percutaneous screw fixation of sacroiliac joint disruption with follow-up of patients 36 months after injury and fixation. The results of this study show long-term follow-up regarding the effectiveness of this treatment. Based on our results, we conclude that the use of this minimally invasive imageguided technique is safe with reproducible results, thus decreasing the risks of other treatment options. The longterm results are promising and have encouraged our team to further apply this technique. An important factor ensuring a good outcome is a multidisciplinary team, including radiologists and orthopaedic surgeons. Well-functioning cooperation is mandatory between members of the team for accurate patient selection, procedure planning, and procedure performance. All materials used are currently available in most hospitals at which interventional radiology and orthopaedic surgery are performed. Perfect screw positioning in all patients suggests that the technique can be performed easily and reproducibly, and we believe that this intervention can be performed in any radiology department. Further prospective studies investigating long-term functional results and including larger patient groups are required to assess the definitive merits and larger applications of percutaneous instrumentation of the sacroiliac joint.