Minimally Invasive Removal of Failed Sacroiliac Fusion Implants

Sacroiliac joint (SIJ) pain is a leading cause of disability and chronic pain that is treated by both surgeons and other non–surgery-trained specialists. Surgical management of sacroiliac joint pain is often reserved for patients who have failed multi-modal pain regimens, physical therapy programs, and failed persistent improvement in pain after sacroiliac joint (SIJ) injections.1 Surgical management traditionally involved an open posterolateral approach to the SIJ, removal of cartilage, and application of bone graft into the joint.2 Several variations of this approach, with or without the addition of trans-articular screws, have also been described. Over the last 20 years, minimally invasive techniques to perform SIJ arthrodesis have become more popular and often involve limited removal of cartilage and placement of a permanent implant with the goal of promoting fusion. The most commonly used implants, such as iFuse (SI-BONE, Santa Clara, CA), do not denude the entire cartilage of the SIJ and rely on interference fit, reducing motion, and bony on-growth to the implant. The reported revision rate at 2 years of 3.4% to 3.8% is likely underestimating the actual incidence of implant failure due to study bias, and the large proportion of MIS SIJ fusion procedures being done by non–surgery-trained specialists.3 Furthermore, with increasing numbers of these procedures being performed, there will be more revisions and failed implants being referred to orthopedic surgeons and neurosurgeons for removal. Reasons for revision or removal of the implant include painful nonunion at the fusion site, implant malposition, and persistent pain at the SIJ with confirmed improvement after SIJ injection. The removal process of these implants often depends on the manufacturer’s removal set, whose technique guide is not readily available online in most instances.4 Commonly, these will involve a threaded pin that mates with the implant, but often bony overgrowth may prevent threading into the implant. Use of osteotomes to cut out the implant en bloc can also be challenging due to the tendency for the osteotome to take a different path away from the sclerotic bone around the implant, and lead to more bone loss. A preoperative CT may show sclerotic bone around the implant (Fig. 1). We have successfully applied an alternative removal technique to remove iFuse implants that limits bone loss and allows for additional control of blood loss and ease of bone graft application into the void left behind by the implant.FIGURE 1: Preoperative CT demonstrating sclerotic bone around the implant.TECHNIQUE To begin the implant removal, we recommend utilizing the previously made incision from the index procedure. This is followed by bluntly dilating down to the location of the implant using conventional drill sleeves or other blunt cylindrical instruments. Next, we use a drill-tipped guide pin to cannulate the implant with the use of fluoroscopic guidance. This step removes bone away from the center of the implant and provides the angle needed to easily find the implant. The next and key step in the removal process is to use an 15 mm Arthrex OATS Donor Harvester, a cylindrical drill large enough to accommodate the 14.2 mm diameter implant, to drill around 80% of the length of the implant (Fig. 2). We recommend stopping short of the entire implant to be able to access it without pushing it further into the sacrum. If the bone around the entirety of the implant is cut out, the implant may be pushed deeper into the sacrum when the threaded pin is attempted to be inserted.FIGURE 2: The OATS Harvester is advanced over the implant using the drill tip guide wire as a guide.Finally, with the bone around 80% of the implant cut away, we are able to insert the threaded insertion rod into the device and mallet the implant out (Fig. 3) using a slap hammer. There should be minimal bone left around the implant using this technique. This same sequence of steps can be used to remove additional implants as needed. Once completed, this technique will leave behind uniform cylindrical defects in the SIJ, which are predictable in size and depth. We recommend inserting allograft bone dowels that are sized to fit the diameter and depth of the defect to encourage healing of the bony defect and reduce bleeding as this is introduced into the gluteal musculature. Figure 4 demonstrates a postoperative CT scan showing the bone dowels filling the bony defects.FIGURE 3: With 80% of the bone around the implant cut away, a threaded extraction pin can be used to back slap the implant out, while removing minimal bone.FIGURE 4: Bone dowels sized to fit the bone voids can be placed to reduce bleeding and promote filling of the bone defects.Expected Outcomes Utilizing this technique, we expect consistent ease in removing these implants, as this method allows for controlled bone removal around the implants and facilitates ease of removal with the threaded pin that mates with the implants. Before this technique was used at our institution, we experienced longer operative times, blood loss, and bone loss due to the difficulty in removing these implants. Although these implants do not always lead to fusion of the SIJ, they are commonly covered with sclerotic bone, which makes extraction difficult unless osteotomes are used to cut out bone around the implant, which is difficult to control. Utilizing bone dowels to fill the bone voids can also be helpful in minimizing blood loss from cancellous bone oozing and possibly encouraging healing at the bone voids in the SIJ. We have only used this technique to remove the iFuse implants, but the same technique could be used for implants of different diameters, using a different size Arthrex OATS Donor Harvester. Complications Common complications such as bleeding, pain, and possible infection are unavoidable with this technique. In addition, damage to gluteal neurovascular structures is possible and could be minimized by ensuring blunt dissection down to the implant and only using the drill when over the implant or on bone, and not turning the drill on when out of bone. Possible complications that we have not encountered but could be related to the use of the OATS Donor Harvester are being off-axis with the implant and potentially causing metal-on-metal wear, producing free metallic fragments in the SIJ or surrounding soft tissues. However, we have not encountered this as the cylindrical drill is 0.8 mm larger in diameter than the iFuse device and allows for a small margin of error without creating metal debris. Careful attention to matching the angle of the drill tip guide wire is an important aid in staying in line with the implant, and only remove a small amount of sclerotic bone around the implant to facilitate removal.