Barrack v. Lombardi: Patella Resurfacing: Rarely, if Ever Necessary
OTW Staff • Tue, July 11th, 2017
This week’s Orthopaedic Crossfire® debate was part of the 17th Annual Current Concepts in Joint Replacement® (CCJR), Spring meeting. This week’s topic is “Patella Resurfacing: Rarely, if Ever Necessary.” For the proposition Robert L. Barrack, M.D., Washington University School of Medicine, St. Louis, Missouri. Opposing is Adolph V. Lombardi, Jr., M.D., Mt. Carmel New Albany Surgical Hospital, New Albany, Ohio. Moderating is Thomas S. Thornhill, M.D., Harvard Medical School, Boston, Massachusetts.
Moderator Thornhill: I’ve always said there are somethings you never criticize. People’s children, people’s dogs, or whether or not they resurface the patella. This debate is to talk about that.
Dr. Barrack: So a legendary football coach from my opponent’s hometown said that three things can happen when you pass the ball in football and two of them are bad. So there’s a corollary in knee replacement—a lot of things can happen when you resurface the patella and all but one of them are bad. Patella resurfacing is not a benign or simple procedure. You do it at the end of the case. Most surgeons do it freehand and the incidence of problems with this is frightening and I don’t think it’s accurately reflected in data, particularly from registries.
Under-resection; over-resection; oblique resection and disrupting the blood supply. And there are a number of negative sequelae of all of these errors that are very common—more common than you think. How often do these occur? One of our faculty members reviewed X-rays of resurfaced patellas and estimated a 4% failure rate in only 2.6 years. Very few of these were revised, but you know in the long run that a high percentage of these are going to present problems, and many of them are going to be unsolvable problems.
So we have a unique laboratory to look at these in St. Louis because most of the patellas are not resurfaced, because one of our esteemed faculty, Leo Whiteside, hardly ever resurfaced patellas and all of his trainees didn’t resurface patellas. So in St. Louis, most patellas are not resurfaced. In our total joint clinic, what do we see most? Do we see unresurfaced patellas that are having problems?
There are tens of thousands of patients in our area that have unresurfaced patellas. I’ve never resurfaced one of Leo’s unresurfaced patellas. What I have done are patella revisions, patella fragmentations, loose components and AVN [avascular necrosis]. We see late stress fractures. We see lateral facet pain syndrome, tilted patella, oblique resurfacing. A lot of these don’t get counted in registries as failures, but they’re all failures. They’re painful, resurfaced patellas. Oblique resurfacing was second most common to lateral facet syndrome. Occasionally you can re-operate on these and make a flat surface, but usually there’s not enough bone to work with.
If you look at randomized trials…we followed a group of patients over 10-15 years. Early on we saw no difference, even in bilateral cases. Beyond 10 years the resurfaced patellas were doing worse. So over time, the unresurfaced patellas tend to remodel, while the resurfaced patellas become more symptomatic.
Virtually identical results were reported by the group from London, Ontario, and there’s most compelling information recently published from the UK, 34 centers, over 1,700 knees. Their conclusion was that all the scores were the same. We see no difference in any score, and if there is any difference, it’s too small to be of any clinical significance.
The major determinant of the clinical result and the presence of anterior knee pain is a surgical technique, getting the components aligned and rotated properly, not whether the patella is resurfaced or not. Is it necessary? Rarely. For rheumatoids with aggressive inflammatory disease or a mal-tracking patella, but for the vast majority, it’s simply not necessary.
Dr. Lombardi: If you look across the world, there is no consensus on this topic. In Norway they do about 2% and then in England and Wales, they do about 70%. If you look at the Americans, we do about 80% and that has been recently published in our new registry that’s been formulated.
We know there are three philosophies. Either resurface in all cases. Never resurface. Or resurface in selective cases. Then I have to go through, ‘What am I going to select?’ I look at patient factors, implant designs…I should resurface probably in inflammatory, maybe not in non-inflammatory; maybe if they have anterior pain I should resurface. Maybe if they’re young I should resurface. Maybe if they’re too heavy I should resurface. I think Robert hit the nail on the head. Surgical technique is the most important thing.
I think you should restore the mechanical axis. I think you should have an implant that restores the anatomical femoral size. And by virtue of the fact if you go out there and look at the vendors, yes there are more sizes now available so that we can do this very appropriately.
Don’t overstuff the patellofemoral articulation. That means, measure the patella before you resect it and measure it afterwards, and measure the thickness of the anterior femoral resection. I wonder how many of you really do that or think about how much bone you’re taking away anteriorly when you’re doing a joint replacement.
It’s very important to align the femur correctly with the transepicondylar axis and the AP axis. Use all landmarks to set the component correctly. And when you resect the patella, obviously, you want to get a smooth, flat, symmetrical surface. Remember that the thickest portion of the patella is medial, so if you center the patellar component on the patellar bone, you’ve actually lateralized it, so you’ve got to medialize that patellar component. You can also lateralize the femoral component, which is going to enhance your patellofemoral tracking.
I went to literature to see, indeed, if surgical denervation of the unresurfaced patella is important. There are three papers that have been published in the past two years, and there is some consensus. Maybe denervation has less anterior pain is what I got out of this; better function; complications were similar.
Now let’s go to literature and see what the meta-analyses tell us. One from 2004 showed re-operation rates as 5% resurfaced and 13% non-resurfaced. And they had a fancy algorithm on when to resurface, but they resurfaced 90% of their patients.
And one from 2005. Unresurfaced patellas had increased risk for re-operation; significant anterior pain; significant pain during stair climbing.
Another one from 2005 again…14 articles looked at …and non-resurfaced patellas had a greater incidence of anterior knee pain; 18.7% required secondary resurfacing. Non-resurfaced patellas resulted in less patient satisfaction.
In 2009, a critical appraisal of literature looking at meta-analyses; systematic review and randomized controlled trials, patella resurfacing had decreased anterior knee pain and decreased patella related re-operation.
In 2011, sort of the same theme. Absolute risk of re-operation was reduced by 4% with patella resurfacing.
I think I’m seeing a theme here.
And another one where re-operation for any reason was 3.9% in resurfaced versus 7.8% in non-resurfaced.
In 2011, a higher rate of re-operation was observed in the non-resurfaced group. No difference in the incidence of anterior knee pain.
In 2012, equivalent incidence of anterior knee pain and level of satisfaction, and resurfaced patients underwent significantly fewer additional surgeries.
Another one from 2013, resurfacing reduces the risk of re-operation.
Let’s move on to the registry data. What does the registry data tell us? The Norwegian Arthroplasty Register: 11,887 total knees. Resurfaced 9 years median follow-up, non-resurfaced 7 years. Similarly reduced overall risk of revision with resurfaced patellas. Resurfaced patellas had a lower risk due to pain alone, but a higher risk due to tibial loosening for defective poly insert.
If we look at the Australian Orthopaedic Association National Joint Registry, we see 5-year results and resurfaced patellas had a cumulative revision rate of 3.1% versus 4%. Patellofemoral pain was 1% resurfaced versus 17% in non-resurfaced. Interestingly enough, if we look at their rate of resurfacing it has actually increased and it’s up to 59% now. Higher revision rates occur in non-resurfaced patellas due to patellofemoral pain, pain itself and patellar erosion.
In the New Zealand Registry, 76% patellas were unresurfaced; 33% resurfaced and 1% required a secondary resurfacing. This is the one registry, Robert, which does show that the resurfaced patella may have a little higher re-operation rate currently. So they’re selecting the worst patients to resurface the patella and they’re telling us those worst patients now may have a little higher revision rate.
At the end of the day, pain is the primary reason patients come to see us. Right? And our primary goal is to relieve pain and I think I’ve shown you enough literature that says there is less anterior knee pain and, if you don’t want the patient to have another operation…resurface the patella.
Moderator Thornhill: You know, Dr. A. Seth Greenwald tells us to make these things evidence-based and we have done that. But it reminds me of the statement that orthopedists use data like drunks use lampposts—much more for support than illumination. So much of the stuff here was, in fact, the same articles making the other point.
Robert, if you get rid of the metal backing, you get rid of the small lugs which loosened; the large lugs which fractured the patella; and you have a 3-lug all poly patella…then you have the technical issues. If you got rid of all that stuff, would you resurface more patellas, or how many patellas do you resurface and when?
Dr. Barrack: Ten percent or less. I would like to borrow some of Adolph’s slides because I would actually use them in my talk…because a lot of the data he was showing were these meta-analyses that went back to 1966, so you’re looking at components that were universal and didn’t even have rights and lefts. So there are clearly some components that do poorly with patella resurfacing. They do just as bad without patella resurfacing. If you resurface a patella you just tell patients to live with the pain. And Adolph’s well familiar with that phenomenon.
Moderator Thornhill: Adolph, I think we could probably say that you resurface the patella in all cases, and I’m supposed to have no bias as the moderator, but I resurface all patellas as well. I think the biggest issues when I see failure is sort of what you eluded to, and so did Robert, technical issues. Overstuffing the patella, I think, is one thing. My concern is I see people with lateral bone. Could you give us your pearls for resurfacing the patella to do it correctly?
Dr. Lombardi: I use a freehand technique. I evert the patella and that’s usually after I’ve made all my bone cuts. And the reason I wait until the end of the operation, Robert, is that that’s when it’s easiest to evert the patella. During the rest of the operation, I just sort of sublux it. I don’t put a lot of stress in my own mind on the extensor mechanism. I identify where the patellar tendon is; where the quad tendon is; I use a wide saw blade. I go from inferior to superior and then I come across lateral to medial and I make sure they’ve got a smooth, flat, symmetrical surface. I think you can feel that with your fingers. And then when I put the sizing guide on, I try to get as much bony coverage as possible and I put it as medially as I can and then I usually nibble away the extra overhanging lateral facet. I usually take the electrocautery and release the tissue on the lateral side. I don’t go around the whole patella. I’d be interested to know if everybody thinks that’s a good technique or not. I’ve always thought it was not, but this presentation made me look at the literature more closely and now I’ve found there are a number of articles on that which say that might be beneficial.
Dr. Barrack: What do you do when you over-resect the patella and take too much bone that first time?
Dr. Lombardi: Robert, you name the complication and I’ve done it. You say, ‘Oh, (shoot)!’ and you move on. Put a patella button on it, what are you going to do? Yeah, have I done that? Of course I’ve done it.
Moderator Thornhill: I think the easy thing is to go from chondro-osseous junction to chondro-osseous junction at the quad tendon above and the nose of the patella below. I would rather overstuff the patella than to take away too much. You have resurfaced some patellas in people that had unresurfaced patellas and had pain, correct?
Dr. Lombardi: Correct.
Moderator Thornhill: What percentage of those get better, do their symptoms go away when you resurface them?
Dr. Lombardi: I would say about 25%. I don’t think that our message should be here that if you have a painful knee, and have an unresurfaced patella, to resurface the patella.
Moderator Thornhill: Thank you very much. Great discussion.
Please visit www.CCJR.com to register for the 2017 CCJR Winter Meeting, – December 13 – 16 in Orlando.
Senior Editor: Jay D. Mabrey, M.D., whose 35 year career in orthopedics included residency at Duke University Medical Center, service in the United States Army Medical Corps, Fellowship at the Hospital for Special Surgery and a long, distinguished career at Baylor University Medical Center where, in addition to his overall leadership at that institution, developed the Joint Wellness Program that helped patients get up after surgery more quickly, developed the first virtual reality surgical simulator for knee arthroscopy and chaired the FDA Orthopaedic Device Panel, is Orthopedics This Week’s newest contributing writer and editor.