It's an honor to be able to talk on this subject. As mentioned, this is primarily a technique discussion here for patellar osteochondral allografts. I do have disclosures listed here as well as online. The patella has a complex surface topography more so than the distal femoral condyle, and so it might seem a little daunting to get appropriate matching. As you're trying to get grafts, you don't want to be overly restrictive, so you can get access to them, and it turns out two of the probably most important factors to match correctly is the width of the patella and the length of the articular cartilage of the patella. If you have a very large defect, you need to be more accurate with this and get a graft that's the same or bigger than the recipient or the patient. If it's a smaller defect, you have much more variability that you can tolerate here, and you'll see that in some of the examples that I'll show in this video. To start, you have to get into the knee, and the approach here for me is always lateral. Many times I'm doing a tibial tubercle osteotomy, so we do a lateral lengthening. Here's a Z lengthening taking apart the transverse and oblique fibers of the lateral retinaculum. If you start looking at this as you cut through it, you can really see the definition of the LPFL, and that's very helpful to repair at the end of the surgery. Lateral side also, the vastus lateralis is more proximal, so you're less likely to get into muscle, and you do really get nice patellar eversion with this approach. This is probably one of the two most important things to me for doing this procedure. So we do almost like a dental impression, and I do it on the donor and the recipient before we do anything else. So I make sure that my graft can handle the size that I'm going to take. Every once in a while your graft is a little smaller than you expect, and you don't want a hole that's bigger than the plug you can create. Then when you see a central defect like this, you're going to get good contact proximal and distal, and you will not get contact with the cylindrical guide medial and lateral. And that's where some variability comes in where you want to try to match the distance that it's off on both sides. The good news is that if you're off a little bit on each one, it will either match, and if you were wrong, you could flip it 180 degrees, and it doesn't necessarily have to be orthotopic in the exact same position that it was natively in the donor. This matching will be different than when it's on the facets. So here's an example of a smaller medial facet defect. You can see that we have really nice matching. The imprint is almost exactly the same in both. So as long as we don't make a mistake while we're doing the harvest, we should get really perfect matching in that setting. And the same thing in a lateral facet. So lateral facet, you can sit the guide directly down on the cartilage. You have almost complete contact, and then you just move it over until you have a similar amount of overhang if it goes through the central ridge. And we know that that cartilage will be pretty thin by the time it gets over to that medial facet, but when you have a large lateral facet, they're usually dysplastic, and a lot of times that becomes non-articulating in that area. Once you have that part figured out, the rest is a little bit more rote. So you can place our guide pin. We then use a scoring reamer followed by a coring reamer that has a depth stop if you use certain systems. And then this is usually a little bit thicker than other grafts. It's usually 7-9 millimeters because the patellar cartilage is much thicker than any other area in the body. And then we measure the depth at 12, 3, 6 and 9 o'clock so that we have an idea of the depth in a circular fashion. We then had that same imprint on our donor, and then we just placed the guide directly on that marking that we made. This should be done under significant irrigation, and lately we're realizing now that maybe even submerging the graft completely actually helps decrease the viability issues that we have and the thermal necrosis that happens with these reamers. Once you have it harvested, if your imprint matched in one area, then you're going to have a good graft in that area. If you have an area where it didn't match, and in this setting the donor it touched and in the recipient it didn't, that means your donor is going to be proud in that area. You have to expect things like this in the patella because it's hard to get a match that's perfect every time. And so you should just be very expected that that's going to be a little bit proud, and we'll show you what you can do if that's the case. And this is where your final double-check step is. So now that you have the plug harvested and you're taking those concordant points at 12, 3, 6, and 9 o'clock, you just make those dots around the graft in a clock face, and then you draw a line that's through at the first dot that you're at that's perpendicular to the long axis of the plug. If that line, you're not connecting the dots, you're trying to draw a perfect line around the graft that's perpendicular. If you do that and it hits all the dots, then you're going to have a graft that's perfectly confluent throughout the surface. If you do it and two of the dots are off, then that one area will be proud and the other area will be recessed. This is a bit of an exaggeration, but just to make a point, these are cylindrical grafts, and if you have a graft that's long enough, when you put it in, there's no variability in how that graft can go in in a coronal or sagittal plane. So if you bevel the back of the graft to try to change its orientation, you're really just going to create a hole in the bottom of the graft. If it's very thin, then you might be able to get away with that, but the thicker it is, the less wiggle room you'll have. So the only variability you can really alter is rotation in the long axis of the graft, and that's something that you can do as you're placing those dots around if you're not getting good matching. And then once you've done that and you've prepared it, chamfering the edge or doing a bevel around the edge of the defect or edge of the graft is helpful because that'll help you get it started. The patellar cartilage is thick and sometimes soft when it's damaged, so it's very easy to get caught up on that with the bone of the graft itself, and depending on your system, which you should be very aware of the amount of offset that's present, some of them are up to a millimeter difference. You really have to dilate these. And on the patella it can be hard to do that because it's a mobile sesamoid bone. So you really want to make sure that you're doing that multiple times to make sure that that dilator sits totally flush down with the bone. Once that's complete, we want to try to improve the biology, so this is part of any OA graft technique. So we use pulsatile irrigation using high-pressure CO2. And then many times we're recommending to patients bone marrow aspirate concentrates so we can get better integration of the graft, less cyst formation, and hopefully a better clinical outcomes and less revisions. And then the final step is press fit. So we know that impaction causes cell death. If you impact less, you get some less death, but a lot of these end up with the same amount of death after eight days of culture. So it's better to not hit it at all. And then we can use a suture that's in the back that we pull out once we're happy with it. And you want to try to push this in either with an impactor that you're using without a mallet, or you can flip the patella down onto the trochlea and push it from anterior to posterior. And once you're happy with that, you can go ahead and take it out. If you don't put the graft in initially collinear, it will get tight. It will not seat. So take it out, put it back in, keep doing that until you're pretty sure it's collinear. Then it will go in much easier. Then here you can see that area that we were saying it didn't match. We knew it was going to be off in that area. We just take a knife and trim that part down. It always hurts me a little bit inside to be cutting off the surface layer of cartilage because it's very important biologically. But it's probably more important to have a smooth congruent joint for patients. Because you can see on the arthroscopy here, if you leave that part prominent, that's probably going to be more likely to catch and delaminate. So I think you're better off trimming it, making it flush. These patients really do usually notice an improvement in the smoothness of their joint afterwards because now you've restored the normal articular hyaline cartilage. Finally this is not an outcomes talk, but something technically that is important. Jack Farr in our group looked at the outcomes of patellar osteochondral allografts and found that the overall failure rates were significantly higher with shell-based grafts. A lot of times those are salvage-type situations, but you really can do a very large plug that covers 80% of the patella, or at least the main central aspect. And sometimes you can do marrow stimulation on the far lateral aspect. And that's probably preferable to a shell when possible. Finally as Christian mentioned, there's different algorithms here. This is how OA grafts fit into my algorithm for the patellofemoral joint, but a lot of factors matter, and Christian covered that very well. Thank you very much for your time and attention.

patellar osteochondral allografts

graft selection

surgical approach

patellar eversion

graft integration