Dr. Malchau won the coin toss and came out swinging!
Dr. Malchau: “Highly cross-linked polyethylene (PE) has been used for nearly a decade and with good results. Michael Ries will claim that the problem is crescent shaped fracture fragments (as was noted in Orthopedics Today in October 2008). But if you look at the peer reviewed literature on these fractures what do you find? One publication…an article published in the Journal of Bone and Joint Surgery in 2007 that did not focus on fractures but on surface cracking, mechanical properties, and one case.”
Dr. Malchau also cited work by Harry McKellop, Ph.D., who said that data from laboratory hip simulation studies conducted 10 years ago accurately predicted the clinical results seen as of 2008. “In a database study up to May 2008, McKellop found that only 22 fractures of the 141 reported failures of hips using highly cross-linked PE cups were attributed to liner fractures. McKellop told Orthopedics Today that after several million implantations the total number of fractures to date was small.”
“Dr. Ries cautioned that the tradeoffs for wear resistance are decreases in the mechanical properties. In his own study however, Dr. Ries found four fractures that occurred in patients with highly cross-linked PE…all had crescent shaped fracture fragments and the breaks started on the outer surface of the liner. I’m quoting Dr. Ries: ‘Some of the cups were malpositioned, but there were also some implant design factors such as 2mm thickness in the rim and 7.5mm of overhanging PE, which likely contributed to the fracture rate.’”
“So, Mike, ” said Dr. Malchau, “is this a typical case?” “There was a 65- degree latero-version. Some of the four cups were malpositioned, Mike, and some were thin and some had overhang…how many are ‘some’ of 4…more than 1 and less than 4? And what about p-values? If we estimate that there have been four million re-melted highly cross-linked PE cups implanted worldwide then the incidence of liner fracture is close to zero.”
“At MGH [Massachusetts General Hospital] we are looking at the status of re-melted highly cross-linked PE with a focus on wear and osteolysis. We have been using the Martell Hip Analysis Suite which uses edge detection of the femoral head and shell to measure 2-D head penetration from AP radiographs. After analyzing 80 hips and with more than 7 years of follow up, we have had no components revised due to PE fracture or wear. There have been no indications of osteolysis (on plain radiographs) in either the standard or large (femoral) head size groups.”
“There was a Swedish study of a cemented cup with 61 hips, all of which were osteoarthritic and evaluated by RadioStereometric Analysis (RSA) at 3 months, 6 months and 1, 2, 3, 5, and 7 years. We have 0.1 mm head penetration, which, at up to 5 years was no problem. All of a sudden past 5 years the highly cross-linked PE runs parallel with the conventional materials.”
“So, ” summarizes Dr. Malchau, “what is the clinical implication of this? Could highly cross-linked PE suffer from accelerated wear past 5 years or does this represent a change in creep properties? We don’t know…and part of the reason that we don’t know is because most series on wear of highly cross-linked PE are underpowered and have too short of a follow up. The problem thus far is not related to liner fractures—it’s the increased head penetration that is causing concern. Bottom line…we need more data, which we are collecting.”
The diplomatic Dr. Ries waited patiently while Dr. Malchau built his case. And when he was done, Dr. Ries came back with vigor. Dr. Malchau had at the outset gone after Dr. Ries’ crescent theory, but Dr. Ries did not take the bait.
Dr. Ries: “Henrik has shown that cross linking is associated with a decrease in wear, but cross linking also reduces the mechanical properties which can occasionally lead to in vivo fracture. Wear is reduced more with increased doses of radiation, but this reduction reaches a maximum at 10 Megarads; most materials we use are in the 5 to 10 Megarad range.”
“There have been reports in the literature of rim or crescent shaped fractures initiating at the locking mechanisms or on one of the corners that is used to provide rotational control of the liner. The fracture may occur along the rim and then propagate inside producing a crescent shaped fracture—which is different from fractures that occurred in gamma irradiated PE with earlier cup designs.”
“The processing methods affect the mechanical properties as well as the wear; increases in radiation result in decreased wear but also a decrease in mechanical properties. The irradiation used to produce highly cross-linked PE also produces free radicals, which can oxidize in vivo. If the material is annealed above the melt temperature, then the free radicals are neutralized so no oxidization can occur in vivo. But there is also a decrease in crystallinity and mechanical properties. Alternatively, it can be annealed below the melt temperature. This reduces the mechanical properties less, but leaves behind free radicals that can oxidize later.”
“The series that Henrik is referring to is the one we presented at AAOS [American Academy of Orthopaedic Surgeons]. There were four fractures in four different cup designs and with different material processing methods. There are some similarities: they seem to produce a crescent shaped fracture; three out of four were malpositioned; all had some evidence of impingement which contributed to the fractures; all had areas of thin PE and PE protruding above the rim.”
“As (femoral) heads have gotten bigger in order to maintain adequate thickness at the dome, many heads have moved outward, which results in contact with the PE above the rim. This is a relatively thin area that—if impingement occurs—can result in cantilever stresses and a crack initiating at the locking mechanism.”
Dr. Ries’ ultimate solution? Newer, better, implants with enhanced material properties.
A Melancholic Swede and a Laconic Californian
Moderator Dr. Thornhill: “We’ve got a melancholic Swede and a laconic Californian here! Henrik, for your routine total hip what is the head size you prefer and the minimum thickness of PE that you allow?
Dr. Malchau: “I prefer 6mm PE and whatever head size that would accommodate. I am trying to do a 32 mm head, but I won’t do that and trade off going down to 3mm PE.”
Dr. Ries: “We agree on thickness…5 or 6mm. But, I will ream more bone to get a larger cup and accommodate a larger head when appropriate. And I don’t use an elevated lip liner. If you have a 36mm head why would you need an elevated lip liner these days?”
Moderator Dr. Thornhill: “And the verticality of the cup…is it important, and what do you shoot for? Henrik?
Dr. Malchau: “That is the most important issue today…how to put the cup in. It doesn’t matter what kind of bearing you use, you’re going to get squeak, stripe wear, and fatigue failure of the PE. And there is little data saying how good we are at this. In our work at MGH I thought we’d have a narrow shaped bell curve, but we haven’t, so maybe navigation could do something for us.”
But don’t blame malpositioning of implants, says Dr. Ries: “Whatever cup you put in you want to optimize the position and the bearing surface. Even with navigation we’re not going to be perfect. Most retrieval studies show a high amount of impingement so I don’t think we can blame malpositioning for the adverse effects. Our materials and implant design should be able to accommodate some impingement that occurs in vivo even in well positioned cups.”
Moderator Dr. Thornhill: “Let’s talk about impingement…What is the most common cause of this—is it over-rotation of the acetabular shell, head to neck/ratio, is it prosthesis specific?”
Dr. Ries: “It’s probably primarily related to head/neck ratio and implant positioning, but it’s also affected by range of motion [ROM]. Some female patients have wide ROM of the hip and they’re going to impinge more…even in well positioned cups they’ll impinge.”
Dr. Malchau wanted to stay on his earlier point: “That means that nobody really knows what the problem is because there is no data supporting that. My fear is that malpositioning does have a prominent role. If you look at the Scandinavian registries you see that recurrent dislocation as a reason for admission is the second most frequent issue. There must be a way to prevent that—maybe through optimal implant design or going to larger heads.”
Moderator Dr. Thornhill: “Let’s discuss material properties. Mike, you showed that wear goes down as you increase the Megarads up to about ten. The downside is the mechanical properties, so how critical is it to be oxidatively stable?”
Dr. Ries: “Definitely important. The first generation submelt annealed cross-linked PE material is no longer available and retrieval studies show more oxidation than previously utilized PE. It requires an above melt annealed process for the stabilizing effect to reduce the concentration of free radicals and we just don’t have long-term in vivo data.”
Moderator Dr. Thornhill: “If we want to crush free radicals using antioxidants or triple annealed material, is that going to solve this problem, Henrik?”
Dr. Malchau: “Well there are still free radicals. What you need is to prevent the oxidation because what’s been detrimental prior to highly cross-linked PE is the incidence of osteolysis. We haven’t seen osteolysis so far up to 7 to 9 years…even 10, but what’s happening beyond that point I don’t know. Will it stay stable? We don’t know what is going on, including why we see that increase in head penetration in the Swedish study. I think it’s real but we don’t understand what it is.”
Moderator Dr. Thornhill: “So if you look at 5 to 7 years you begin to see an increase in head penetrations. Mike, can you speculate as to what causes that?”
Dr. Ries: “I’m not really sure. Perhaps if the femoral head roughens in vivo and that increases the wear rate due to counterface roughening that could occur…or maybe it’s creep. Henrik knows more than I’ve mentioned on this topic.”
Moderator Dr. Thornhill: “Thank you both for a nice and informative debate.”
Please visit www.CCJR.com to register for the upcoming 2011 CCJR Winter Meeting, December 7-10 in Orlando, Florida and the 2012 CCJR Spring Meeting, May 20-23 in Las Vegas, Nevada.
