Dr. Daniel Kaplan:
Welcome to the Arthroscopy Journal Podcast. I'm Dr. Daniel Kaplan from NYU Langone Medical Center in New York City. Today, I have the pleasure of speaking with Dr. Alan Zhang, professor in the Department of Orthopedic Surgery at the University of California, San Francisco.
Dr. Zhang was the senior author of the article titled Preoperative Hip Injection Response Does Not Reliably Predict Two-Year Postoperative Outcomes Following Hip Arthroscopy for Femoroacetabular Impingement, which is currently in press at the Arthroscopy Journal. Welcome Dr. Zhang, and thank you for joining me.
Dr. Zhang:
Thank you for having me. It is such an honor to talk about our research on the Arthroscopy Journal podcast.
Dr. Daniel Kaplan:
As a special note, this is my first one doing it, so it's special. Thank you for making this an easy one for me. I appreciate it.
Dr. Zhang:
Well, I didn't say I was going to make it easy for you.
Dr. Daniel Kaplan:
To get going, as a young hip arthroscopist, I do find that decision to inject a hip ambiguous and determining which value to ascribe to an injection response even murkier. I'm really excited to speak to you about this topic.
Dr. Zhang:
Great.
Dr. Daniel Kaplan:
To begin, can you tell us just a little bit about your workup and approach to FAI patients including key elements of the history physical exam and imaging you're looking for?
Dr. Zhang:
Yeah. Well, thanks again for having me. I'm really excited to talk about this study. This had some very interesting results that we weren't expecting per se, so happy to talk a little bit more about it. To answer your question, we always start with a thorough history for the patient, including symptom duration, any inciting events to their symptoms, any trauma. The location of their pain is really important to us, any activities that exacerbate their symptoms, and then, of course, any prior treatments that they have. We'll really kind of take in the whole story and try to figure out what might be the source of their issues. Then after the history part, we'll go into the physical exam where we'll check the hip range of motion, focusing on any differences from side to side, especially with internal rotation if we're looking for FAI in particular.
We'll also, assess the patient's gait if they're limping, if there are any particular types of gait patterns that they're exhibiting. Then we'll, of course, check the Fadir tests, F-A-D-I-R or flexion adduction, which is a very well-known test for hip impingement. We'll also use our version of the labral stress test or a scour test where you're abducting and flexing the hip and then ranging it and then going into an internal rotation extension, kind of like a McMurray's of the hip, so to speak, to try to load the labrum and see if that causes any pain or even any mechanical symptoms.
Then, of course, we'll readily use our radiographs and AP pelvis as well as a Dunn lateral view at 90 degrees. We found that the Dunn lateral view is the most specific view for assessing for CAM deformities compared to other ones like the cross-table lateral or the frog leg. The Dunn is the preferred one for us. Then lastly, we'll use a false profile view on radiographs to look for any joint space or early signs of joint space narrow.
Dr. Daniel Kaplan:
That was a great overview. Thanks. I hadn't actually heard of that test as frequently described, but that's definitely something I'll look to incorporate, so that's great.
Dr. Zhang:
Great.
Dr. Daniel Kaplan:
Just building off your work up a little bit, so can you tell us a little bit more about your injection algorithm? From your study, it seemed like about 40% of people got a preoperative injection. How do you determine who would be a good candidate?
Dr. Zhang:
Right, so this is kind of the reason we did this study is just who should be getting these injections and do people need to get them if they're a good candidate for surgery already. In terms of our algorithm, it depends on what the patient has tried already. Some patients will already come in having had injections by their prior physicians or from previous workups. And in those patients, I typically don't repeat those injections to start.
Dr. Daniel Kaplan:
Sure.
Dr. Zhang:
Ideally, as a surgeon, patients come to us already having had worked up and had a full gamut of conservative treatments like physical therapy and injections, and then they're just basically in the office to discuss whether surgery is appropriate or not. That isn't always the case. Sometimes if we see patients that have not had any previous work up, I'll start with physical therapy, of course, and then after they've had physical therapy, if there's no improvement after six to eight weeks of physical therapy, some patients will opt to have an injection for treatment. This is where I talk to them about the option of having an injection.
I don't make every one of my patients have an injection, but there's definitely patients that would like to avoid surgery or delay surgery as long as possible. Those are the patients that will really support doing preoperative or injections for, and in those cases, they can have, usually we'll treat those patients with steroid injections to start. Now, there are some other patients where we're not sure where their source of pain is coming from. If they have pain that's both in the groin, which would be consistent with FAI, but they also have pain in the back of the hip or in the glute muscle or another part that's more atypical, then we may use an injection to try to localize their pain. The classic teaching has been that if they have a positive response to the injection, then that will indicate the source of their pain.
Dr. Daniel Kaplan:
Got it, so I do laugh a little bit. Yes, as surgeons, we would love for those patients all to come in treated, and maybe when you're Dr. Zhang, professor of orthopedics, you get a lot of those patients already worked up when you're Dan Kaplan just starting out, and most of those patients, they're your first visit. I guess a little different algorithm.
Dr. Zhang:
Yeah, no, not always even for us who have been in the game for a little while, we definitely see patients that have not had any treatment at all.
Dr. Daniel Kaplan:
Your branch point for a local anesthetic versus a steroid injection, the local anesthetic is usually just that diagnostic to determine to localize basically joint versus extraarticular, right?
Dr. Zhang:
Exactly, so for patients that are, we're trying to actually treat their pain or their symptoms, then we'll use steroid injections typically. These are patients who, again, either don't want surgery or want to delay it, or some of them may not be good surgical candidates. It might be the patient who has a little bit too much degeneration or they're a little bit older where the data shows they may not do as well with arthroscopy, so we're not as aggressive about recommending surgery for them. Then we'll treat them with corticosteroid injections, see what kind of benefit they'll get and how long that lasts For. The patients who get local anesthetics are typically the patients who are trying to just find the source of the pain, but not necessarily treat them with it because the local anesthetic, as we all know, it won't last for more than a few hours.
If they have temporary relief, that typically indicates that it's localizing their source of the pain to the location of the injection, whether that's in the joint or sometimes we'll use the injections in the hip flexor tendon, the iliopsoas, just to see if that's a source of pain as well. Those are the patients. I get local anesthetics. Then the other thing I'll add is that some patients will get a local anesthetic just because the insurance is requiring it. That's another reason for us to have done the study because there are a lot of patients that were well-indicated for FAI surgery, but the insurance would not approve the surgery until the patient had a diagnostic injection. Then we would just have the patient do a local anesthetic injection to verify the source of their pain before their surgery. That's also what we'll get into later about the results, which are surprising in that.
Dr. Daniel Kaplan:
Yeah, I saw that line in the discussion and I loved it, and we'll talk more about it later. Can you tell us a little bit about the logistics? Let's say once the decision is made to proceed with an injection, what's the workflow in your office? Are you referring these patients to IR? Are you and your team doing it in clinic? Are these under fluoro, under ultrasound?
Dr. Zhang:
Right, so I know a lot of other providers who will do their own injections, whether it's under fluoro in the back office or under ultrasound. My clinic flow is a little bit tight, I'll say, so that I usually don't have time to do injections in the office, and I've not acquired the skill of using an ultrasound proficiently myself. I'll refer out all of my injections. We have a great group of primary care sports medicine doctors that are experts in doing injections, whether it's biologics or various other forms and all sorts of hard to reach places, so they're very nimble with the ultrasound. I usually refer to our primary care sports medicine specialists for ultrasound-guided injections. Sometimes if they're backed up and the patients can't get in, then we'll refer to radiology where the musculoskeletal radiologist will either do it under ultrasound or floral guided for the intraarticular injections.
Dr. Daniel Kaplan:
Got it, so once that patient gets the injection, now they're back in your clinic, what are you asking them about their response? I know in the paper you guys used sort of a binary improve or didn't improve, but are you asking them specifics like duration and percentage of improvement, or is it more of just a yes/no?
Dr. Zhang:
Right, so the first thing we asked was how was the response to the injection? Did it help at all? Then it's first just a binary response, and we try to record that response in our notes, and it'll either be, "Yeah, it helped," or a lot of patients will say it didn't help at all, and then I'll try to follow that up, but, "Oh, not even for a few hours later that day or nothing," or they'll say, "I really couldn't tell if it changed anything at all," so that would be a negative response. Then if they say, "Yeah, it did help," and I'll ask them how long. If it's a local injection, it usually wouldn't be for more than a few hours or a day or something. Cortical steroid, that's probably a little bit more interesting. That's quite variable. Some patients after a steroid injection, they'll say, "Oh, it lasted for a couple of days, or a couple of weeks, or even a couple of months."
Then if they do have improvement, I'll try to ask them to quantify. Then I'll say, "How much better was your pain? Was it like 50% better, 100% completely gone," and I found that most patients had a really hard time doing this. They would just say, "I'm not sure," and then they would just kind of throw up a random number for me. We didn't try to use that in our paper because we just felt that it was a very inaccurate number for the patients to try to quantify it, and it was kind of all over the place based on their responses.
Dr. Daniel Kaplan:
Got it. With that background in mind, can you tell us a little bit then about the impetus for this study and touch on the state of the literature prior to your work?
Dr. Zhang:
Right, so this study kind of was based on some observations in my own practice and my clinical practice, where we've seen many patients with classic FAI syndrome, large impingement deformities, who were given a diagnostic preoperative injection for probably most of these that had the lidocaine were done for insurance purposes just so they could get to the surgery that was needed. I found that a lot of those patients told me they did not have improvements from those injections, even though the classic teaching is that they should have.
In that case, we wanted to study this a little bit more. For the patients that had steroid injections where I wasn't as surprised that they didn't always get improvement, because some patients will have a steroid flare up for the first couple of days, and some patients just don't respond to steroids at all. For the patients that had a local injection who said, "No, it didn't help me at all, I still need to get the surgery," that's where we wanted to investigate exactly what that correlation was. That led us to look at our outcomes after a surgery based on these patients who had preoperative injections and what their response to that injection was.
Dr. Daniel Kaplan:
Yeah, I'm glad you said that because especially just starting out practice, I had always been trained and thought that link was there too, where if they responded, they should be a good candidate. I can distinctly remember two patients, maybe the worst CAMs I've seen in practice, and neither of them had any response. One actually had a negative response to the injection and I just couldn't wrap my head around it. I'm glad you guys sought to clarify that a little bit.
Dr. Zhang:
Yeah, that's exactly what I saw too. Some patients that were very well indicated for surgery, just complete impingement syndromes by the textbook, but just did not have good responses to those injections. Now, I'll say that the ones that did have positive responses, those were still patients that were well-indicated for the surgery, but it's really the patients that didn't have good responses that may still be good candidates is what we're trying to get to with this study.
Dr. Daniel Kaplan:
Yeah, so with that, what were you guys expecting to find based on this? Based on either your anecdotal experience or the literature?
Dr. Zhang:
Right, so at least anecdotally, my hypothesis coming in was that patients who had severe FAIs syndrome and were well indicated for surgery, getting a preoperative diagnostic injection response may not always show benefit or give even temporary improvement. Essentially, the preoperative injection would not correlate with how well they did after surgery.
Dr. Daniel Kaplan:
Now, let's get into the actual study. Can you tell us a little bit about the design?
Dr. Zhang:
Right, so this is a retrospective study where we match the cohorts of a prospectively collected data on patients. We try to collect PROs on all of our patients undergoing hip arthroscopy at UCSF, both at pre-op and regular post-op time points. Then we also collect physician forms where we fill out very different variables such as physical exam information, prior treatments such as their injection and their response to the injection. I could basically select out the patients in the database who had injections and what their response to that injection was, and then track them to see what their response was after they had surgery two years later.
In our database, we found about 209 patients who had preoperative injections and who had completed PROs, both pre-op in two years afterwards. Then out of these patients, 42 of these patients have reported that they had no improvement at all with their injections. These are our non-responders group. Then to kind of compare this group, we matched them one to two with 84 patients who did have improvement after the injection. That kind of gave us a little bit of a comparison group to see the results of these non-responders versus the responders.
Dr. Daniel Kaplan:
Got it. I'll just want to give you guys one shout out. I loved as a stats nerd that you calculated your own MCID and included floor and ceiling analysis. That's definitely a huge issue in sports orthopedic surgery research. I think that's something that hopefully we seem in more and more papers.
Dr. Zhang:
Yeah.
Dr. Daniel Kaplan:
Great. Yeah, so with that, what did you guys actually find? Tell us about some of these results.
Dr. Zhang:
Right, so based on my intuition, it was that essentially these patients who had no response to the pre-op injections might still do well. What we saw is that the 42 non-responders, after we matched them to the 84 responders, they had the same demographics, same intraoperative variables like what we did during the surgery in terms of thermoplasty, laboral repair, and the same kind of preoperative variables in terms of severity of their impingement lesions. After matching them, we looked at their two-year post-op outcomes, and then we also separated based on injection types. There's the group that had injections with corticosteroids and then the group that had injections with local anesthetics. Then when we tracked all of those groups together, we found that there's no difference for the groups that had corticosteroids between the responders and the non-responders in amount of patient reported outcome improvement, and the percentage that achieved MCID or minimum clinically important difference after their surgeries at two years.
Regardless if they had improvement from their steroid preoperatively, either group would have the same results at two years. Then we looked at the local injection group, and there was also no difference, for the most part, in the PRO improvement as well as the MCID achieved except for the VAS pain score, which the non-responders to the local anesthetic group actually had a better improvement in their VAS pain score, which means that the groups that did not respond to the lidocaine actually ended up doing a little bit better after their surgeries than the group that did respond to the lidocaine, but all four branches of this study, all four groups did improve significantly from pre-op to post-op.
Dr. Daniel Kaplan:
Yeah, I noticed that paradoxical larger improvement, at least for the local anesthetic patients, any idea why that was? They seemed to even have a bigger Delta from pre-op to post-op. Was it just because they were starting from a worse place maybe?
Dr. Zhang:
Exactly, so when we kind of looked farther into the data, the patients that did not respond to the local anesthetic actually started off of lower pre-op scores than the patients that did have a positive response. These are the patients that are just hurting really badly before surgery. Then the post-OP scores all kind of maxed out because we did see a significant ceiling effect of over 30% in some of the PRO scores. Pretty much you can only max out to a certain degree of improvement. For the responders, they started off a little bit better at baseline. Their Delta from pre-op to post-op was smaller than the Delta for the non-responders, and that's why there's a bigger change or more significant improvement in the non-responders group.
Dr. Daniel Kaplan:
Got it, and then relatedly, it seemed like that difference between responders and non-responders only for the lidocaine local anesthetic injection, not the steroid injection. Any idea why the steroid responders and non-responders seemed more equivocal?
Dr. Zhang:
I think that's also just based on where they started at baseline. For the steroid groups, they all started at pretty much the same place baseline-wise.
Dr. Daniel Kaplan:
Got it.
Dr. Zhang:
Just the Delta was less of a change than the lidocaine non-responders.
Dr. Daniel Kaplan:
Mm-hmm, and then your results are, I mean, certainly contrary to my expectations and certainly, classic teaching, but they're also in contrary to at least a good amount of pre-existing research, like you guys mentioned, a systematic review by Martins et al. Any idea the difference between yours and theirs, just different designs or any explanation?
Dr. Zhang:
Yeah, I mean, this was surprising to us too. We didn't think we would find this big of a difference. With these prior studies, especially that Martins et al is a systematic review. They looked at different studies where there's multiple surgeons, multiple different people doing different surgeries, so it's not as well-controlled. I guess in our case, being a single surgeon study, we know all the patients are getting the same surgery, having the same type of rehab. Those variables are a little bit more matched in terms of what's actually happening during the surgery. I remember a few years ago when some of these other studies are being brought around, some of the other members of the audience were saying that this study is very obvious because how the patient does to an injection is not going to be correlated with how they do after the surgery because it's going to be what you do during the surgery that makes them better or worse.
I remember Dr. Thomas Byrd talked about this one example where he said, "If you inject a patient with arthritis, they're going to get better from that injection, but then if you do surgery on them for hip scope, they're not going to do well from that surgery." I thought that was a really good example. But our study, the difference with our study is that it's not that the patients who did well with the injections did bad after the surgery. It's that the patients that did not do well with the injections still did well after surgery. In our case, I think that's the novelty of this, where we saw that even if they had a negative response to the pre-op injection, they could still have a positive response after surgery. Of course, we're not doing surgery on arthritic patients in cohort.
Dr. Daniel Kaplan:
That makes a lot of sense. The one thing that I have trouble with is I see your results, and I've seen in practice a couple times now, people not respond to, I think or have done well with surgery, but it's just hard for me to wrap my head around, you're dumping a bunch of lidocaine into someone's intraarticular space. How can we reconcile them not having improvement from that, but then improving from surgery? It just seems so counterintuitive. Do you have any idea?
Dr. Zhang:
Yeah, I think we alluded to this earlier in that these are probably really severe cases of FAI where they're not responding to even local anesthetic. In those cases, the FAI is just causing a lot of damage, a lot of inflammation. There's synovitis throughout the whole joint. It's possible that the lidocaine is just not reaching all the spaces or getting to those nerves or air of inflammation, or there's just so much synovitis around the capsule, around other areas where the lidocaine is just not reaching it. Then when you do the surgery, you correct the FAI, repair the labrum that eventually leads to improved outcomes because you're correcting the underlying deformity, but it might just be that the FAI is so severe that the lidocaine is not really getting to all the spots.
Dr. Daniel Kaplan:
Yeah, that makes sense. As we start to conclude, how are you going to incorporate the results of this study into your practice? Is there a value for either a therapeutic or diagnostic local anesthetic injections based on these findings?
Dr. Zhang:
Right. So I think first, I hope that the study is picked up by some of the insurance carriers that are requiring mandatory injections before hip arthroscopy, as the point of that is that the patient has tried conservative treatment like physical therapy, and they're well-indicated for the surgery. Doing an injection is not really going to change the algorithm of whether they need the surgery or not. Hopefully, that's they can get rid of that practice. I'm not holding my breath on that, but that's one thing I would like to hopefully see in terms of some different thoughts on that practice.
Now, in my practice though, I do still use diagnostic injections, especially when, well, first the patients want to have the injections for therapeutic reasons, definitely, I won't argue with that at all. If the patients are, I'm not sure where the source of the pain is for the patient, I will still use a diagnostic injection because if there's still a good positive predictive value in that, if they have a good response to the injection, it still helps you localize it. If they have a poor response to the injection or no response to the injection, that's where you want to think a little bit more about whether they could still be a candidate for surgery in that case.
Dr. Daniel Kaplan:
Got it. Then as an extension of that, what's your spiel for patients in the office when you're describing plus or minus an injection? You did a workup, you're offering an injection. Can you tell us a little bit about how you're counseling a patient?
Dr. Zhang:
Right, so for patients, so if there's any contraindications to surgery or if they're trying to avoid surgery, I'll basically recommend the injection because the injection will help with their pain. It'll help decrease inflammation in their hip, and then I can at least give them some temporary relief. Then hopefully with that they can do other things, either modifying their activities or doing more physical therapy to strengthen all their muscles to try to offload the joint a little bit better to avoid getting to surgery, so that's the first part. If the patients are looking to not have surgery, the injections are an important part of that treatment algorithm.
Then if we do recommend an injection, sometimes I'll describe to them that we use it to try to localize if that's a source of the pain, because some patients may have iliopsoas tendonitis as well as a labral tear. If we want to see which one's causing more pain, we'll probably inject both spots, but not at the same time at different times. We'll see if there is better improvement with one or the other. For example, if they have no improvement from the intraarticular injection, but the lidocaine injection helps with their iliopsoas tendonitis, that's probably the area we want to target, whether with their rehab or other types of injections. That's kind of how I use the injection treatment algorithm in my practice at this point.
Dr. Daniel Kaplan:
Awesome, and then as we wrap up, do you have any closing thoughts you want to share with the listeners?
Dr. Zhang:
Yeah, I think the main thing with this study was that it came from a clinical observation that did not necessarily agree with prior literature. I saw some of these patients and we did take them to surgery, even though prior literature would probably indicate that they may not do well, but we saw that they still had great improvement even though their preoperative injections were not very helpful. I encourage all of our clinicians out there, all of our listeners, to investigate any interesting observations and trends that you see in your practice because not everything may align with prior research, and some of these acute observations may lead to wrinkles and previously accepted data, and then it makes for great future research and helps the community a lot.
Dr. Daniel Kaplan:
Dr. Zhang, thank you so much again for your time and your insights. It was a pleasure speaking with you as always.
Dr. Zhang:
Thank you for having me. Great job on your first podcast, Dr. Kaplan.
Dr. Daniel Kaplan:
Dr. Zhang's article titled Preoperative Hip Injection Response Does Not Reliably Predict Two-Year Postoperative Outcomes Following Hip Arthroscopy for Femurocetabular Impingement is currently in press and available at the Arthroscopy Journal website at www.arthroscopyjournal.org. Thank you for joining us. This concludes this edition of the Arthroscopy Journal Podcast. The views expressed in this podcast do not necessarily represent the views of the Arthroscopy Association or the Arthroscopy Journal. Thank you for listening. Please join us again next time.