• LecturehallOsteomyelitis: Should Hardware be Removed?
  • Lecture Transcript
  • TAPE STARTS – [00:00]

    Harold Schoenhaus: Our next speaker is Dr. Michael Troiano who comes from Philadelphia. I have some personal knowledge about Michael. I trained him, taught him as a student, did a fellowship with me. Has been in my practice and has pretty much kick me the hell out and said I’m old and over the hill. He is a professor at the Temple School of Podiatric Medicine. He’s on staff at Penn Presbyterian Medical Center, Temple University Hospital, Abington Hospital. He does the majority of work at our Penn Wound Care Center. He’s an excellent lecturer and surgeon. And it’s with great pleasure that I introduce Mike Troiano to share some thoughts with you on Osteomyelitis: Should Hardware Be Removed? So please welcome Dr. Troiano. Thank.

    Michael Troiano: Alright. Thank you for having me. So, Dr. Schoenhaus and I were having a discussion, a conversation. And for the first time in a long time it wasn’t about scotch or cigars or him leaving because he’s over the hill. But what we were talking about was we were talking about this Mako. Have you guys heard of the Mako? Not double M-A or double, whatever, M-A-C-C-O. But there’s a robot, M-A-K-O called the Mako which Stryker has recently purchased the company of. And this robot is making hips, knees, and soon to be total ankles I anticipate very, very, reproduceable. The surgeon literally gets like an iPad and you set up your robot on the person’s knee or hip or what have you. And that surgeon literally takes a pen and just draws and the robot cuts away what the robot needs to. And each implant is reproducible over and over again. So, it begs the question; what’s the purpose of the doctor? Right? The purpose of the doctor, the surgeon, the doctor is not made in the operating room. It’s made when he or she is out of the operating room. It’s the decision to do that particular case, the decision to make that particular orthotic or the skive or the width or what have you, the experience. And what I’m going to talk about today is what happens oftentimes post-operatively, which is infection and how to handle it. Because if you pull a myriad of people, you’ll get a myriad of answers. And I don’t know that there’s a definitive answer in each way that we handle a situation, so we’ll just go through some of these and see how we would handle situations. This is a paid grant by Stryker with limited advertisement. So, this is our overview and objectives. The first thing we have to understand is osteomyelitis and how it’s created. And just as a refresher, osteomyelitis has two major ways of being communicated. Number one is direct bone inoculation trauma. So patients stepping on a nail, open fracture, car accident. You know, at the time of that trauma the bone is inoculated. The second is a contiguous spread. So if I can call your attention to one thing, we have to identify here clotted blood. So oftentimes, we’ll see someone in our practice that comes in with a hematoma and we say no big deal. We’ll stick a needle in it, we’ll prep it, aspirate it, take it out. The end. But that clotted blood is actually a nidus for infection, right? Blood is very sugary sweet and bacteria love it as we know. And it will oftentimes grow and use that blood as fodder or food for the bacteria and seed a nice healthy osteomyelitis. So we have to be a little bit more aggressive in some cases than the old Keflex and more on compresses that we may have used 20, 30 years ago. Surgical site infection is another way to get osteomyelitis. And surgical site infections make up about one to four percent of each surgery that we do. So this surgical site infection actually at 4% does not include those surgeries which you would expect to become infected. So if you have an open pylon fracture, calcaneal fracture, metatarsal fracture, that what we would expect to be infected is not included in this 4%. Surgical site infection is just by definition a positive culture from a surgical site. It can be deep, it can be superficial and develops within six months of the surgery.


    And then finally, prosthetic joint infections. There is no great idea, say, guideline on how to handle hardware and infection. So by putting together methods of treatment of osteomyelitis, by putting together methods of treatment of surgical site infections, and by putting together methods of treatments of prosthetic joint infections we get an amalgamation and in some cases an algorithm of how to treat our patients. So with prosthetic joint infections, the definition involves a presence of a sinus track that communicates with the prosthesis. So that in your case may be your calcaneal fracture plate. It doesn’t necessarily have to be the total ankle replacement or the knee replacement. Acute inflammation with a sample of tissue evaluated, purulents with no known etiology, and two or more intraop or combo of preop and intraoperative cultures culturing out one viral organism. So this is how we identify a prosthetic joint infection. Orthopedic device-related infections is 1% to 2% of all surgeries. So if we look at those 4%, two of those percent are associated with implantation of hardware or joint. The PJI’s, the prosthetic joint is separated into early, delayed and late onset of injuries. So early is less than 3 months after surgery, delayed is 3 to 12, and then late is 12 months after surgery usually being seated hematogenously. So these are the people that will call you up, “Hey, doctor I had my bunion done and I’m going to the dentist, do I need amoxicillin preoperatively?” Right? Early onsets are really the ones that we worry about and delayed onsets are the ones that we are really scratching our heads about. As podiatrists, we’re probably not doing as many total joint replacements as our orthopedic colleagues. But it gets a little bit easier in our 3 to 12-month period here because if you’ve done a bunion, obviously, after three months if that bunion gets infected, the hardware comes out and no big deal. Right? The three-month period is the one that keeps us up at night. Because in all reality, we’re sitting there wondering, do I take the hardware out and sacrifice the positive surgical outcome? Do I leave it in and risk that the osteomyelitis spreads or turns into a bacteremia or septicemia? What do I do? So we go ahead and we consult infectious disease. So at Presby, we have an infectious disease doctor, her name is Judy. We won’t bring up her last name. And the residents affectionately call her Judge Judy. Because Judge Judy believes that everyone should be born with bilateral BKA’s. Any time you put in hardware or someone gets an infection, just remove that part of the foot. No problem. And obviously, the phenomenon exists whereby this patient has come to your office to get better and you’re fixing their bunion or doing their first MPJ implant or whatever you’re doing, and you have to have a realistic discussion with them saying, “Look, there’s a chance that you’re going to lose a portion of your foot.” So I think informed consent is very, very important. That this person is going to understand that even in the most minor soft tissue procedure, things go wrong. Toes go into vasospasm, hardware gets infected. You know, the scrub tech may not have had a great day and just glances at the drapes, not seeing the hole. And this infection is a train going down the track which sets you up for certain failure. So these people have to know what the possibility is before you get into the surgical intervention. In actuality, the initial diagnosis of osteomyelitis open and closed trauma, surgical site infection, only gives the surgeon one objective finding which is a positive culture. One objective finding; positive culture. Everything else is subjective. So when you document that this patient has fever and redness and erythema and warmth and pain, all of these are signs of osteomyelitis. And I think we have to quantify them by saying, you know, in training with Dr. Schoenhaus, he had a very important term that he puts in his notes; as per primam, within normal limits. What you would expect at this time post-operatively, right? So there is erythema, there’s edema, as per primam. Meaning, that this is normal healing. You’re going to get some inflammation. As opposed to documenting the erythema streaking cellulitis, this person has a pylon fracture that an orthopedist fixed in our area. And whether the person walked on it or whether it wasn’t a very good job, I don’t really remember. But you can see radiographically this kind of involucrum occurring here.


    This is osteomyelitis with broken hardware. So someone is not healing after a period of time. We have to think to ourselves, do they have an underlying osteomyelitis? Is there a reason that they’re not healing? Again objectively, we can order some tests, but objective tests are oftentimes subjective. Andrew Meyr at Temple University School of Podiatric Medicine and Temple University Hospital, he’s a residency director there did an amazing study whereby he took bone from bunions, from osteomyelitis and from cadavers. And he said, “Here pathologists, which one has the osteomyelitis?” And don’t you know every one of the esteemed pathologists which he pulled gave a very different answer. So even osteomyelitis under a microscope is a subjective determination. Right? Plasma cells found osteomyelitis, found in Charcot, found in inflammation. Nobody really knows the answer, so we rely often times on our subjective findings. So we can order an ultrasound or CT scan to look for a fluid collection. Because that fluid collection is probably going to be some walled-off fluid consistent with an infection. We can order an MRI which has up to 100% specificity in feet but can’t tell timeline. In other words, if you’ve operated in three months you’re going to see bone marrow edema. So is that bone marrow edema on MRI the process of the operative procedure or is it the process of infection? No one knows. We can order PET scans which are nonspecific but relatively sensitive. And then we can order bone scans, and bone scans are twofold. And that we have the white blood cell bone scan which will accumulate those white cells. And we have the technetium or the three-phase which will just accumulate activity of bone, osteoclastic and blastic activity. So in our bag of tricks, in our armamentarium that we have, we have all these options, none of which are great. What we do have is we do have inflammatory markers in the bloodstream. Meaning the, ESR and CRP. And we’re going to get into those in a few minutes. But ESR and CRP are an absolute necessity any time you think someone is developing an infection. And not the level of the ESR or CRP, but the trending level. So it’s not enough to order an ESR or CRP post-operatively because, of course, it’s going to be elevated due to the nature of the surgery. But if you order one every couple days, every week, you can begin to see that blood count fall for improvement and not fall for lack thereof. Tibia is the most common site of an infected nonunion primarily because of the limited and anastomotic blood supply. There’s also not a lot of meat in between the tibia and in the skin. So a wound dehiscence here or an infection can be quickly out of control. For prosthetic joint infections, the IDSA, again, like the SSI, surgical site infections, gives us basically subjective findings, do a thorough H&P, order some labs. Right? Imaging, they identify the imaging, bone scan, leukocyte scans should not routinely be used for diagnosis, IDSA. But I think in practicality, you need something to hang your hat on. And you need a reason where this person is not just treated with your subjective findings. Right? How many times have you seen a patient from another doctor and he or she identified no pulses or one plus pulses and they’re normal or nonexistent conversely in your examination. So we need things that we can hang our hat on to this patient where we can say, “Look, these are the things that support that you do have an infection. These are things that support the nonexistence of the infection.” And be able to intelligently talk with them so that they can make the best decision with you kind of as a quarterback of their care. Frozen sections. Your hospital has to be prepared to do frozen sections. Because intraoperative frozen sections are paramount with periprosthetic soft tissue identification of infection or inflammatory markers. So when we take these people back to the operating room, it’s important to take a little sample of that tissue and make sure on the frozen section within minutes that there inflammatory markers are decreasing and you’ve gotten all the infection out. This is a little algorithm that is used in the orthopedic world whereby, if you see a sinus tract acute onset of painful prosthesis after it’s been implanted for a long time or a chronic painful prosthesis, meaning, it didn’t work from day one, the person did not have any significant relief of time, we have to refer to the orthopedist or the podiatrist.


    And then he or she should do, again, subjective examination. If there is no suspected infection, you’re done. If there is, we go down to cell count, differential, and aerobic and anaerobic cultures. And this can be as simple as a core needle biopsy or, you know, stab incision culture. Obviously, people need to be off of antibiotics for two weeks before you do a culture as an aside. And then we plan for surgical intervention going from there and we’ll see the continuation of this algorithm a little bit later. Preimplantation, ESR and CRP’s are recommended. So if you’re going to take your hardware out or your prosthetic joint out, make sure you start ordering ESR’s before you put it back in just to make sure that that ESR is trending downward. Again, the changes in the CRP value are more important than the actual number. Now, what pathogens are we going to see? As we know diabetics, all bets are off. They’re going to get polymicrobial infections, right? So many times, I will see someone come to me with a horrendous infection and the doctor has been treating them with Keflex or with Cipro or Clinda. Diabetics get polymicrobial infections and therefore, you need to cover anaerobes, gram positives, and gram negatives until you have a direct identification of the bacteria which exists. In the healthy patient or non-immunocompromised, non-diabetic patient, it’s usually a gram positive with the number one infecting organism being methicillin resistant staphylococcus epidermidis, MRSE. So when we talk about staph epi, all right, people say well that’s a normal skin contaminant. But in prosthetic joints, in hardware, it can really run amok on our patient. Less common, anaerococci anaerobes, fungi, and mycobacterium. But that does not mean that you should not culture for them. You want to culture, pan culture when you take this person to the operating room. So when the nurse asks what do you want to send for culture? The whole gamut. Staph aureus in a foreign body versus regular. So we’re taught in school that the definition of an infection is one times ten to the eighth colony forming units per sample of tissue taken. Anything over that is an infection, under that is normal. But once you introduce the phenomenon of hardware, then that one times ten to the eighth drops to one times ten to the second. So you’re only talking about one thousand units needed to cause infection and form a material presence. And the reason being is because of the glycocalyx or the bacterial biofilm that we’re going to get into in a minute. Biofilms takes about three to four weeks to form. They have antibiotic resistance because they set a nice kind of barrier across the hardware. Such that the only antibiotic that can really penetrate is an IV antibiotic that’s tough to tolerate, which we’ll get into. Infection can become apparent even weeks or months after the surgery. And sometimes there is no clinical response for days or weeks until this kind of hard layer of ice is penetrated. This is a little slide that shows how a biofilm results. You have a single free-floating bacteria that lands on to the surface of your plate or your joint. And the bacteria begins to multiply, aggregate and attach and then you have a nice little shelf here. As it grows and grows and it fragments off, more bacteria which reorganize and basically colonize the whole hardware, so you have a huge barrier to infect to antibiotic penetration to your hardware or bone. Again, this is where intraoperative frozen section can be useful and bone biopsy can be useful, decide between hardware removal or retention. So you want to culture the hardware when you take it out. And then at the same time, underneath the hardware you want to biopsy that bone in two separate pathology specimens to ensure that that bio burden hasn’t made it into the bone. Treatment. Again, the only thing I would remove from this slide is infectious disease. You have to consult them, but I would put a star in say begrudgingly so. Make sure you partner up with an infectious disease doctor who doesn’t just want to cut everything off. Because, you know, surgical site infections, prosthetic joint infections, there are positive outcomes. It’s not just a death sentence for the joint or that infected extremity each time. IDSA guidelines. There’s a study coming out in 2017 that involves children.


    This is a poorly researched area, orthopedic hardware implantation and should it be removed or left in. So there’s a lot of research that needs to be done in our literature search and we identified this as a possible outcome. But again, the child is obviously a lot different than the adult population. Then there’s some open fracture management literature. Irrigation and debridement, this IDSA actually identifies that within your debridement of an open fracture you should not culture. I think that that’s a bad decision. But it says that if you need to go back for a second wash or a debridement, at that point, you should culture. They cite only a 25% correlation between initial cultures in the pathogen which later causes the osteomyelitis. But boy, would I like to know what it is from day one. Because at the same time, I’m going to just go blow everything away with every antibiotic I can choose. Much of this are of anybody who follows Warren Joseph. SSI. Again, there is limited data and therefore, our algorithm should be set on extrapolation of data from PJI guidelines as to whether to remove or replace or leave it in. And then there’s the study in 2012, prosthetic joint infection and what the IDSA recommends to do. With that study, they talk about three different types of ways to remove our prosthetic joints. The first is a two-stage which is probably what we do in practice more than often. Somebody has a bunion; the first stage is take out the hardware, put in a smooth K wire as not to propagate the infection. Put on a little mini X fix to hold the hardware out to length and weight. Treat with antibiotics for four to six weeks. Hyperbaric oxygen if you have a strong suspicion of osteomyelitis. And then after the six weeks once the inflammatory markers fall, go ahead, repeat the hardware and hopefully everything is all said and done at that point. It’s routinely done in hip and knee arthroplasty. In Europe, they have a one-stage resection arthroplasty. This doesn’t play so much into our normal everyday surgery, but into those of us that are doing total ankle replacements or our first MPJ replacements with polys, the poly is removed and then treated for two to six weeks and then reinserted. Not the joint. The joint itself is left, but the poly is taken out. And the exception to that rule would be someone who can’t tolerate a more invasive surgery or someone with a looser component to that implant such that at that point the whole implant would be removed. We’re going to begin to talk about rifampin. Rifampin is really the only antibiotic that can cross through that film, that glycocalyx. Rifampin is poorly tolerated for people. It can become synergistic with vancomycin and hit the levels that we need to in theory with vancomycin. Although there’s been some studies to dispute that synergy lately, the classic answer is rifampin and vancomycin work in concert together. Albeit with some side effects to the host that you’re administering to. But certainly, has a place in prosthetic joint infection. Duration of symptoms. If it’s less than three weeks or the joint age is less than thirty days, then you’re going to go and ask yourself, is there a well-fixed prosthesis? Is there sinus track? Is it susceptible to oral anti-microbial agents? If the answer is no, remove the prosthesis. If it’s yes, which most of the cases that we see, they debride, retain the actual implant and just exchange the poly when it’s time to re-implant. Of course, no re-implantation of hardware is going to be based on, this is on the side of our fusion, All right? So we’re going to take out the implant at this point. If the person has limited bone stock, highly resistant pathogens, or unstable for multiple surgical interventions, once the hardware is out, they’re on long-term antibiotics, now we’re going to get into our septic fusions. Which is a whole lot easier to fuse an ankle and a knee and have that person still move along. And of course, amputation is obviously the final choice, although Judge Judy’s first going forward. And amputation. Once we have a clearing margin, we want to treat this person for 24 to 48 hours as per IDSA most of our literature supports about two weeks of oral and IV antibiotics if we’ve gotten a clear margin.


    If not, we’re looking at four to six weeks. Be surprised how many people when they do an amputation don’t send a clearing fragment of osteomyelitis. Send the clearing fragment, it tells you if you’ve got it all out. Takes two seconds. You take another little slice of bone. You can even take it off of the slice that you’ve already removed and end it pathology, pathologist looks at it under a microscope and they can compare the osteomyelitis to the non-osteomyelitis. And did you get it out? Yes, no, maybe so. And that determines your antibiotics. So you’re not guessing, should I treat this person for two weeks or for six weeks? Rifampin again, 90% effective is what the literature identifies when used in combination. Although there are studies to refute that. Without rifampin, only 50% to 60% of antibiotics are effective. That’s again, a classic answer. I don’t know that I necessarily 100% believe that. Downsides, there are significant drug-to-drug interactions and high toxicity. But in a place where you can use it, it has a nice penetration to that biofilm and should be used probably a little bit more than we are using it. Antibiotic bone cement. We’re going to just spend a couple minutes on this. It is a podiatrist’s dream to take out hardware and pack either cement or calcium sulfate or what have you. And we’re going to spend some time. There’s never been any data to determine whether or not it works. Cement can be very, very dangerous and that it creates exothermic reaction. So it heats up so hot that it actually can necros the bone. And if you want to pack it into a defect, you have to pack it when it’s still moldable. As it begins to set, it sets with its high, high lability of heat killing a lot of the antibiotics. So we have to be cognizant that not only the antibiotic is killed but the bone is killed. So that leaves us with two or three really heat stable antibiotics. One being vancomycin one being tautomycin and the other one being gentamicin. Bimodal allusion is 5% in the first twenty-four hours with cement. Again, it does impede bony growth. But you also risk the fact that a biofilm formulates over the actual cement. Biodegradable have zero orthokinetic. Meaning, that there is a secondary antibiotic release during degradation. So as it begins to dissolve it will allude out more antibiotics. There’s different types; ceramic, calcium, lactic acid. And there’s no need for removal. I will caution you that a lot of these when they degrade, degrade with a white chalky substance that looks a lot like purulents. So the patient should be aware, you know, preoperatively, hey, I’m going to put the stuff in it’s going to seep through your incision a little bit. So when you start to see it, you can anticipate that this is pretty normal and par for the course. Stryker actually makes a PMMA with tobramycin already in it. So it sets nice and easily, moldable into blocks and beads. Vancomycin ceases alluding at 12 days, so if you’re going to start to use bone cement or tobramycin recognize that you should have plans to change the cement every week or so. So pack it in, let it sit, kill, seven days later change it over until there’s no signs of infection anymore. PMMA again, has by bimodal allusion properties. Five percent in the first twenty-four hours, less than fifty percent is alluded at four weeks. And the diffusion rate is based basically upon the brand of PMMA, the coefficient, the fusion, the fusion coefficients and the surface area of the bead. So how small did you make it? How large did you make it? Obviously, smaller beads have more surface area per antibiotic, larger beads can’t really penetrate what you need them to. And then, of course, the blood flow over that bead. If it’s a poorly vascularized area, that blood flow needs to wash the antibiotic out of the bead. So if it’s a poorly blood flow area, you’re not going to get a great dilution. Calcium sulfate will disappear at about eight weeks. It is osteoconductive. So in theory, as you’re killing the bacteria you’re also allowing bone to grow new bone. And then a support angiogenesis and osteogenesis. Calcium sulfate is kind of an off-label use with antibiotics, so that patient needs to be cautioned or at least identified. And most of the times the rep that’s with you in the room, he or she has seen this so many times that they know the cookbook recipe. Add a gram of vanco to this much, take a little bit of water out. But they can’t tell you that cookbook because it’s off label. So see me in the back and I’ll sell you some recipes. Just kidding. Finally, antibiotic selection and MIC.


    If we go to this slide, an MIC is a minimum inhibitory concentration. Basically, this disc is dropped with antibiotics over this bacteria. So let’s say this is all staphylococcus aureus plated here. We’re going to drop a disc and we look at how much is killed from this antibiotic on this particular bacteria. And this area is the zone of inhibition. So this would be a control, no antibiotic. And this would be the antibiotic that we would like to choose. And then you might see other different antibiotics and we’ll look and see how much that zone of inhibition is. So your antibiotics selection should be to reduce the toxicity of the individual agents, prevent emergence of resistance, and of course, treat polymicrobial infections. Finally, this should have minimal systemic response. In other words, you’re using these beads because you don’t want them to allude into the periphery. Right? We want them concentrated at the area. So in conclusion, evidence-based treatment should be based on IDSA recommendations as well as PGI guidelines and clinical judgment until we get a better answer as to what we should do. There should be definitive culture data, debridement is very important, antibiotics are very important. And in general, if the hardware has been in for less than three weeks, you should leave it in if it’s stable. And if more than three to four weeks, you can go ahead and take it out. Thank you very much.

    TAPE ENDS - [32:32]