"Thank you for all taking time out. My name’s Michael Moore. I’m an orthopedic spine surgeon practicing in Bismarck, North Dakota. Prior to coming to North Dakota, ten years ago to start a new spine program, I was in practice in the Denver area for about ten years. I’ve had an interest in the sacroiliac joint since beginning practice, and have acquired a fair amount of experience for various reasons in the evaluation and treatment of patients with sacroiliac pain. I’ve done about 210 open sacroiliac fusions over the years, and I’ve had the opportunity at various times to review the results on these, and report on them. What I’d like to do today is share that experience with you, and hopefully, you’ll find this useful in evaluating and treating patients of your own with similar problems.
While my experience is mainly with open arthrodesis, I believe this experience is applicable to the newer minimally invasive approaches. I’ll just go into how I got interested in this to begin with. When I was in training, and in a residency, I kind of came away with the notion that they used to think that sacroiliac joints did hurt in the old days, and they used to fuse them, and it didn’t work, so we don’t do them anymore. That was kind of the mainline dogma of the time, which was in the ‘80s. I had one attending, however, who was trained in England. He was British, and they had a different point of view, and he actually had a few patients when I was resident that he did sacroiliac fusions on to the extent he could follow things as resident, I recalled them doing well.
So when I got into fellowship and into practice, I started looking for these problems, and recalled the experience from residency. And I thought well, first I’ll take a look at the literature, and see if I can find this information that was sort of off-handedly quoted to me that sacroiliac joint fusions weren’t done any more because they didn’t work. If you go back in the literature, in the early part of the century, you’ll find quite a bit written about the sacroiliac joint being a source of low back pain. And if you go back to the beginning of the century, Goldthwaite and Osgood talked about its association with pregnancy, Albee carried out 50 dissections of pelvises in cadavers, and reemphasized the point that the sacroiliac joint was, in fact, the synovial joint, which had been described by anatomists previously. But Albee observed in 1909 that most of his colleagues thought that the joint was a synchondrosis, and did not move.
When you get to reports of surgical series, the main ones are Gaenslen 1921, Smith Petersen and Rogers 1926, and Campbell 1927, all reporting about 80 to 90 percent success. Smith Petersen and Rogers reported on 26 patients, and one of these was smith Petersen’s own wife. And I could not find any articles reporting on surgery that reported negative results. I also came across an early article that attempted to describe the entity that they called sacarthrogenetic telalgia. The term, obviously, didn’t catch on, but what they were referring to was the fact that disorders of the sacroiliac joint could refer pain into the low extremity, and even into the foot. And on retrospect, they could look very similar to patients with herniated discs.
Now, I think most people are familiar with this article, Mixter and Barr 1934, which is frequently misidentified as the first article that described the herniated disc. There were descriptions of herniated discs going back to Virchow in the 1850s, and the earliest surgical report was by Dandy in 1929. Nonetheless, the article by Mixter and Barr captured the world’s attention, and subsequent to the publication of that article, it seems that most of the medical world tried to turn every back problem into a disc problem. And so if you look at the literature between the late 1930s and the 1970s, there’s almost no mention of the sacroiliac joint in orthopedic or neurosurgical literature. And as we arrive at the time of my training, despite the fact that there had not been any negative press, so to speak, on sacroiliac joints, at least in the form of literature, recognition of the sacroiliac joint as a pain generator, and certainly, surgery directed at it was viewed with disfavor.
In looking at the articles that you could find, this one was interesting. Norman and May writing in 1956 in a fairly obscured journal had an article entitled, “Sacroiliac Conditions Simulating Interververtebral Disc Syndrome,” and what they discuss in this short article is their experience seeing a number of patients, who had been operated upon for reported disc lesions, who did not improve after surgery. And when they looked into the reason for this, they diagnosed these patients, as having sacroiliac mediated pain based on injections with local anesthetic into the sacroiliac joint. The first surgical article that I came across wasn’t until 1987, and this was by Waisbrod, who’s German, I believe, and this is in an European journal.
He reported on 21 patients followed from a year to a little over five years. Nine of these patients had prior spine operations. I’d like to draw your attention to that number because that comes up frequently in a series of sacroiliac patients, the issue of prior surgery, and possible prior misdiagnosis. They only reported 50 percent good results overall after they instituted psychological profiling, their success rate went up to 70 percent, which is not spectacular, but it was promising. And if you look in the article at how they were diagnosing sacroiliac joint mediated pain, they did it by a provocative test using an injection technique. In other words, they attempted to pressurize the interior of the sacroiliac joint, and provoke the patient’s typical pain, such as you would do with a provocative discogram.
Most people would not endorse that as a valid method of diagnosis at this time, and that may have accounted for some of their failures. As I began looking for these patients, I started operating on selected ones, and in 1992, at NASS, I was able to report on my first 13 patients, who were treated with open arthrodesis. All these patients had symptoms for greater than a year, and they had failed conservative treatment. They were all diagnosed with CT or fluoroscopically guided injection. They had to have near complete relief of their typical symptoms during the anesthetic phase of that injection. Intraconal pathology was excluded by other imaging studies. There were ten female, and three male patients, mean symptom duration 23 months. Ten of thirteen were able to identify a specific episode of trauma, after which their systems immediately began.
And this was usually something like a slip and fall or a lifting injury, such as lifting something heavy out of the trunk of a car. It was not major pelvic trauma or pelvic disruption. Six of the thirteen, once again, had prior lumbar spine surgery, and I don’t have a complete characterization of those, but I remember my first patient was a – about a 24-year-old lady, who had previously been an aerobics instructor. She’d been involved in a motor vehicle accident, in which she was a passenger, and her knee had struck the dashboard applying a posterior force to her femur, and she had pain on the right side of her lower back since that time. When she sought attention for this, someone did an MRI, and thought they saw a herniated disc, did a discectomy, and she wasn’t improved. And thereafter, she was cycled through a number of pain clinics, and psychiatric evaluations, and so on. We did a sacroiliac joint injection on her. It completely eliminated her pain, and carried out a fusion, and she did very well.
So after that, I got pretty excited about looking for this. I have recorded just in terms of physical findings in that initial series, two of thirteen had a positive straight leg raise, and ten of thirteen had a positive FABER. I had nine excellent, two fair, and two poor results. I tried to discern why the two patients had poor results. I thought it might be because of pseudarthrosis, but with the CT scan, I was able to demonstrate a solid arthrodesis, so I was not able to identify what the reason was. I suppose it was a misdiagnosis. There was one pseudarthrosis, which was repaired, in which ended up doing well. This shows the operation I was doing at the time. A curved linear incision is made centered on the posterior superior iliac spine, and I placed two or three AO cancellous screws across the joint with washers, and then cut a transiliac window into the synovial portion of the joint.
This shows an intraoperative photograph to orient you to the left of the screen is the patient’s head, to the right is their feet. Their prone on chest rolls. This is the posterior superior iliac spine. I’ve placed two screws across the joint using a floral. Then used a Midas Rex Bone scalpel to score the outer cortex of the osteotome to complete the bone window. And if you do this right, you pull out a plug of bone that has some kind of bone and cartilage from the iliac side of the joint, and you’re looking right at the cartilage on the sacral side of the joint. Again, trying to stay out of the ligamentous portion, and trying to target the synovial portion. I then dequarticated the base of this, and excavate the margins. I harvested some bone graph from the posterior superior iliac spine, and pack in the depths of that, and then countersink the window across the joint.
This is what the postoperative appearance would typically look like, and this is what a healed fusion by that technique looks like. You can see cancellous bones spanning the sacroiliac joint there. As a result of that paper, it received mixed reviews to say the least at the meeting. About half the audience thought I was crazy, and some of them demonstrated their lack of knowledge about the joint by saying well, it’s a synchondrosis, and why are you fusing a synchondrosis? It doesn’t move. The same problem that Albee encountered. The other group of people responded enthusiastically because they said they had patients with sacroiliac pain, and they didn’t know what to do with them. They were glad to see that there was something to do. Well, what happened is I started getting referrals from around the country, so I acquired quite a bit of the experience fairly quickly.
And I was able in 1998 to report on 110 patients with a two-year minimum follow-up. Only three patients required bilateral procedures and that is a figure that I found to hold up fairly well. A lot of patients will have pain on both sides, but they’re able to identify one side as being the worst, and if you make the diagnosis, views that side. The other side, most of the time, ceases to be a problem, in my experience. Diagnosis was all by CT or fluoroscopically guided injection of a low volume. This is very important of local anesthetic and steroid into the synovial portion of the joint. What I’m doing with that injection is I’m trying to confirm the diagnosis with the use of the local anesthetic, and I’ll show you how we evaluate these injections a little later on. But I’m looking to see if they have pain relief during that phase.
If the steroid happens to relieve their symptoms over a longer term, fine. If not, then we can talk to them about other options, including surgery. Average operative time was an hour and fifteen minutes, blood loss 200 ccs. I found that in patients who had isolated sacroiliac joint pathology meaning that they had no prior surgery, and no coexisting diagnoses, such as spinal listhesis or degenerative disc disease, and so on, I was able to get about 90 percent success. If they had coexistence spinal pathology, success was harder to come by. It dropped to 80, 84 percent rate. I had 9 percent pseudarthrosis, which I wasn’t happy with. Eight of the ten patients with pseudarthrosis that I knew about were smokers. I didn’t evaluate everyone with the CT forced arthrosis. I only evaluated patients with continued symptoms.
Seven of these underwent repeat surgery, and five ended up with clinical and radiographic success. No vascular or visceral injuries, no permanent nerve damage. Now, what I was not able to comment on, on the basis of that experience was how common is sacroiliac pain because my experience was clearly bias because of the referrals I was getting. And so I was unable to comment on the true incidents of this, but people have looked at Schwarzer and April using an injection technique, such as I described, looked at all patients referred to their diagnostic center. And found that of patients with pain below L5, 30 percent of these were attributed to the sacroiliac joint on the basis of their injection studies. I think if they took all comers, it was like 13 percent of all low back pain patients.
Bernard and Kirkaldy-Willis reviewed 1,295 patients, and their conclusion was 22 percent of all low back pain was related to sacroiliac joint. It also noted that 30 percent of patients with radiographic spondylolisthesis had sacroiliac pain. And demonstrates that someone can have coexistent problems or someone can have a radiographic abnormality, which catches your attention, but which is not necessarily the actual pain generator. Maigne recommended using a technique involving two separate injections to try and improve the precision. He concluded 18 percent of all low back pain is preferable to the sacroiliac joint.
A more recent article by Sembrano and Polly. They looked at patients referred for evaluation of low back pain, and were trying to study how often it was related to the sacroiliac joint or the hip. They found that in their review, 14.5 percent of all patients referred to them did, in fact, have sacroiliac pain. So a variety of sources sort of giving you a span of estimates of somewhere between 15 percent and 30 percent of low back pain is due to sacroiliac mediated pain. That’s not to say that that percentage requires surgery, but at least that it’s not an uncommon diagnosis or uncommon entity. Who are the patients who has sacroiliac joint dysfunction or sacroiliac mediated pain? In my experience, and in the experience of some other people, who have been interested in this, about 70 to 80 percent of patients are post-traumatic.
And it’s usually not a major trauma, such as a pelvic disruption, amalgating fracture that type of thing, although I have seen those. Most of the time, it’s something that does not sound like it should lead to a long-term problem, something like a slip and fall. De novo or if you wanna call it idiopathic is another category, and post-lumbar fusion, less agreement on what percentage these represent. But I would like to focus on the post-lumbar fusion group for a second. NASS in 1995, Kleiner and Weingarten reported on sacroiliac pain as a complication of spine fusion. And they reviewed anterior posterior fusions, posterior fusions, and tried to relate that to a level of fusion, and so on.
They found about 9 percent of their anterior posterior fusions; about 11 percent of their posterior fusions alone; who developed back pain post-operatively had sacroiliac mediated pain. They found an incidence of 33 percent in patients, who were fused to L3 or above. They did not find any patients who developed sacroiliac pain, unless the fusion involved the sacrum. And they also found no relation to the bone graph site. Katz, Schofferman and Reynolds in 2003, Journal of Spinal Disorders, reported on the sacroiliac joint as a potential cause of pain after lumbar fusion. They had 34 patients presenting with low back pain after lumbosacral fusion. The sacroiliac joint was the cause in about a third of them.
Once again, they found no correlation with the side of bone graft harvest. So the trick in this, I think most would agree like all things in spine surgery, it’s a matter of identifying the correct patient group making the correct diagnosis, and picking the right patients. So what do you do to diagnose this? There are a number of physical examination maneuvers that are reported to be useful in diagnosing sacroiliac pain. I’ll just run through these quickly, as I’m sure you’re familiar with them. Distraction, you can attempt to compress the joint, reproduce pain, Patrick’s Test or a Faber maneuver can be done in a supine position or a seated position. You can press directly on the sacrum. You can apply a posteriorly directed force on a flex tip while stabilizing the sacrum, and see if you can reproduce the pain.
There are a number of functional tests or motion tests, such as the Gillet test. You ask the patient to raise, in this case, the right lower extremity with the hip and knee in a high degree of flexion, and the normal response is said to be an inferior movement of the posterior superior iliac spine relative to the other side. If it does not move, that’s considered to be a positive test reflecting lack of mobility. Well, there’s been a number of attempts to see how good these tests are. Although they’re handed down in textbooks and so on, as being useful in diagnosing this, it turns out when you compare it with something of a gold standard, if you accept an injection study as a gold standard; it’s hard to correlate these very well.
Dreyfuss, in an article in 1996, attempted to correlate 12 tests with the results of intra-articular diagnostic block of the joint requiring 90 percent relief of their typical symptoms to make the diagnosis of sacroiliac joint mediated pain or sacroiliac joint dysfunction. Now, none of these tests allowed you to predict with any accuracy the result of the injection study. And in fact, no combination of these was found to be useful. Others have attempted similar studies, and replicated these results. There’s also a number of papers that will dispute this finding, and do things like standardized training in the test to try and reduce inner observer reliability. But I would just summarize by saying this, the reports are mixed, but many reports in my experience is that there’s fairly poor reliability in physical examination maneuvers for forecasting a positive sacroiliac joint injection study. And we can spend more time on that in the question period, if you like.
It bothers some people, and it bothers me, as far as that goes that there’s not something that is equivalent of a straight-leg raise to diagnose sacroiliac joint pathology, but if you look around other areas in medicine, there are lots of things that cannot be diagnosed on physical exam that nonetheless are real entities. Occult visceral neoplasms of pancreas, lung, colon, liver, and so on can be present, can be diagnosed with a completely normal physical exam, early renal failure, many brain tumors, parasitic diseases, and so on. And the sacroiliac joint is a unique joint, both in terms of its function, topology, and geometry, and ligamentis anatomy, so I’ll share with you what my sort of algorithm is for deciding who to send for a diagnostic injection.
I find the history, and the patient’s description, and location of their symptoms to be the most valuable thing. About 80 percent of patients will have a history of specific trauma, which seems to often involve a twisting injury. The pain is not in the midline, and contrast to patients with discogenic pain, who I find essentially always have pain in the midline, patients with sacroiliac joint immediate pain point directly at their posterior superior iliac spine or just medial to it. I can’t give you chapter and verse on this, but something I’ve noticed that if you ask about, they frequently will describe using a non-reciprocal gait in ascending stairs. In other words, they prefer to go up one-step at a time, advancing the good leg, and bringing the systematic leg up afterwards.
They also described pain rolling over in bed. Now, these latter two things, of course, are non-specific, and you have to exclude hip arthritis, stroke enteric bursitis, other things that might do the same thing, but this is part of what I use in trying to screen people. Have everybody fill out a pain diagram. Now, seeing these three patterns pretty commonly, the first one I call the Pseudo S1 pattern because they will trace out a pattern that looks very much like what you would expect someone to mark if they had S linear radiculopathy from an L5, S1 disc herniation. There are a handful of patients, who will give you a drawing like this, which you might immediately identify as non-anatomic or hysterical pattern. But I wouldn’t throw those patients on the ash heap until you’ve taken a look at them because some sacroiliac patients will have this type of pain pattern. I think it’s around 4 or 5 percent.
I think this is the most common, C., where there’s pain directly over the sacroiliac joint. Any of these in combination with groin pain; I believe forecast a positive injection study. Diagnosis, if they’ve gone pain over their sacroiliac joint, point to their posterior superior spine, and they’ve got plus or minus groin pain. If I have a confirmatory injection, we’ll describe how that’s done and interpreted in a second, and if I have excluded or identified alternative or co-existent pain generators, then I’ve made the diagnosis of sacroiliac joint mediated pain. I require that the injection results should be unequivocal meaning the patient should not be wishy-washy about whether the anesthetic alleviated their pain. I’ll show you what regards a positive injection in a second. It also must be technically successful. I like to see a fluoroscopic view or a CT scan showing the needle in the joint, in order to know that the person doing the injection had it in the correct place.
How do you interpret this? Well, you do one or two low-volume injections with flouro or CT guidance. You have the patients keep a pain diary, and we give them this sheet of paper, so they fill out specific times before and after the injection. I tell them that they should have maximal symptoms prior to the injection. If there’s anything they can do to get their pain up as high as they can that will make it easier for them to discriminate whether or not the injection is positive or not. They have to have 90 percent pain relief. By this I mean, near complete. I don’t calculate a percentage number or ask them for a percentage. I try and get a feel from talking to them exactly how much of their pain was removed. What I’m looking for is near complete.
I would suggest that you have partial relief, 50, 60, 70 percent relief with a technically successful sacroiliac injection, I’d recommend that you look for another pain generator whether a set block, medial branch block, selected nerve root block. If you do discography, you might wanna consider that. But I think if you don’t have near complete relief, you need to look for alternative or coexistent pain generators. You also will come across some patients nowadays, who have an expectation bias. They show up announcing that their diagnosis is sacroiliac joint, and they’ve read about it, and they know you’re gonna send them for an injection. And so there could be some bias due to that. If I have a patient I suspect that in, I’ll tell them you know what, we’re gonna do two or three injections, and I’m not gonna tell you what we’re injecting.
You have to get the patient’s permission to do this, of course, but I tell them we might inject a long-acting local anesthetic, a short-acting one or we might inject saline. And then if their pain diagram doesn’t make sense, then I would certainly avoid operating on that patient. If they’re all consistent, I think you can be fairly confident. Now, recently, I started using the I-fuse implant for a minimally invasive approach to this. I just have two cases I’ll share with you. These are fairly typical patients, who I would’ve done an open arthrodesis on in the past. The first lady was a 49-year-old lady with two years of pain after a twisting injury, and here was her intake pain diagram. Right smack dab over the posterior superior iliac spine, and this gets me thinking about the sacroiliac joint right off the bat.
She did, in fact, have pain and tenderness over the PSIS. She happened to have a positive FABER. Here was her MRI, which shows some degenerative disc disease at L4, 5. Some people may have thought of discography, however, because of the pain diagram, and the history, I thought the sacroiliac joint pathology was much more likely. Another marker for patients, who may have sacroiliac joint pathology is someone who’s been to – it seems like a legitimate character, who’s been to a number of practitioners without a satisfactory diagnosis. For example, this patient had been to the family doctor, who had prescribed physical therapy sort of generically for low back pain. A neurologist, who did the MRI, diagnosis was degenerative disc disease.
The neurologist mentioned in their note the sacroiliac joint, but didn’t go anywhere with it. Didn’t know where to send the patient, didn’t know what to do with it, and I think this is just a knowledge deficit that that person had. And if they had some more information about it, they may have followed that up because it did occur them. Neurosurgeon diagnosed degenerative disc disease. Did not recommend surgery, and chiropractor was no help. This shows a fluoroscopic view with outlining the joint, and this shows her pain diary. She started off with a pain level of 8, and at 30 minutes, she’s down to a zero or 2. Over the next four to six hours, she gets back up to a 4. She’s a little better on Day 2, and by Day 3 is back up to an 8, so very short, if any, relief from the steroid. But this, in my mind, was a positive study.
This shows her intraoperative appearance with the I-fuse implants that shows her follow up. At four months, she’s reporting zero to 1 out of 10 pain, so this is early follow up. I keep people essentially on weight bearing on crutches for two months, so she’s been ambulating for two months on that. At this point, she’s happy, and I’m happy, and we’ll see if the result is durable. The next patient is 39-year-old self-employed gentleman, financial consultant, 15 years of right greater than left-sided low back pain. He had extensive previous evaluations, and in the past year, have been diagnosed with sacroiliac joint mediated pain at a pain center. He had repeated corticosteroid injections with no long-term improvement, and he had also undergone a recent radio frequency denervation without significant improvement.
This shows his plain films. You notice there might be a suspicious area there in the parse. This shows his MRI with normal disc calculation at all levels. And just to make things interesting, he did have a unilateral parse defect. Now, this had been injected multiple times, but I didn’t’ have any fluoroscopic views that demonstrated that, so I sent him for another one because I wanted to be absolutely sure that this was not the culprit. And he had no relief with that. He had significant relief for the duration of local anesthetic with multiple sacroiliac joint injections, so my conclusion was he did, in fact, have pain arising from the sacroiliac joint. I recommend getting a CT on all these patients that you’re considering or planning surgery on.
There’s a great deal of variability in the anatomy of the joint between individuals, and also, in a particular individual between the right and left sides. I have not been able to identify any features that mark a painful joint looking at these, other than the relatively unusual case where you see extensive degeneration on one side, and not on the other. He has a fairly planer joint, and I’ll just show you – I’ll study the contrast with that. This shows the first patient, and this is higher up in the joint, but she’s got what is an anatomic variation where she’s got this excavated area in the sacrum, which shows up both on the transverse views and the sagittal view. And the importance of this is that if you place an implant across at this level, and seat it to this area here on all your intraoperative fluoroscopic views, it will look like it’s well under the sacrum, but in fact, you’re not capturing any bone at all.
And it’s good to know this going ahead with surgery. You can see she has a similar finding on the other side, and there are all types of anatomic variations like this. You also get information on how thick the ilium is, which can be useful. So this shows the male patient’s intraoperative lateral. He had a fairly short joint, and the sacrum got quite narrow down inferiorly here, so I elected to place a smaller implant for the third implant. He is a little over three months now post-op, and his only complaint is pain on the contralateral side. He’s not having any pain on the side that we operated on. He’s actually requesting a fusion on the contralateral side, but he’s only been ambulating a month. I told him we’ll treat that conservatively. We’ll see how you feel in six months. If you’re happy with the site we did, and the other site still hurts, and hasn’t responded to treatment, then we’ll discuss it.
Now, I was asked to comment on how do you keep patients from disappearing into what I call injection world because some of the interventional pain people have a lot of ideas on how to treat this without – there was some justification to be fair. But when these things repeatedly fail, I don’t think it makes sense to have people continue with them. Sometimes people disappear into this group where repeated and repeated and repeated corticosteroid injections are carried out. There’s prolotherapy, radiofrequency denervations, neuroaugmentation, spinal cord stimulators. What I do to kind of keep track of these patients is I’ll order a specific study.
And then I’ll have them follow up with me in a month or something, and that way I can review the results rather than just referring them to the pain clinic for a trial of conservative therapy or treatment, which may go on for a long, long time. I’m able to get them back and discuss what happened with the injection, if they did have substantial relief with the steroid injection, then great, continue that if that continues to work. Another way to keep track of these patients is do the injections yourself. This requires you to go take a course, but it’s not too hard to learn how to do. Some joints are kind of difficult to get into, but most aren’t, and that’s another option.
Finally, if you find yourself in a debate with some of your pain colleagues on whether surgery is a legitimate consideration for these patients, this is a recent textbook edited by Curtis Slipman, who’s a physiatrist at the University of Pennsylvania, and that’s over 120 chapters of big, thorough discussion of all aspects of interventional spine treatment. And he asked me to write the chapter on surgery for sacroiliac joint syndrome, so your pain colleagues probably have this book on their shelf, and if you need to, you can refer them to that chapter for more information. So thank you for your attention. I’ll take any questions that you have."