CORR Insights®: Molded, Gamma-radiated, Argon-processed Polyethylene Components of Rotating Hinge Knee Megaprostheses Have Lower Failure Hazard Revision Rates Than Air-sterilized, Machined, Ram-extruded Bar Stock Components

Where Are We Now?

Belzarena et al. [3] compared two differently manufactured polyethylene components based on a competing risk analysis and found that polyethylene damage was reduced in recent molded, gamma-radiated, argon-processed polyethylene components for rotating-hinge knee megaprostheses. Within 10 years, the risk of polyethylene failure was decreased by 17% compared with traditional air-sterilized, machined ram-extruded bar stock components.

We need to be mindful that the latest implants and techniques are not always successful. For example, no improvement has been seen with the use of different polyethylene components in posterior cruciate-sparing TKA [7]. In addition, knee kinematics and polyethylene damage of rotating hinge megaprostheses may be different from that of conventional TKA to treat osteoarthritis. Polyethylene inserts are likely to wear or break in diverse ways at different times, while megaprostheses must survive longer in young patients with tumors.

The key in understanding this study [3] is its use of a competing-risk analysis [5]. To help explain this statistical method, which might be unfamiliar to some readers, it is worth knowing a bit about the basic concept of a survival analysis and censoring. Some people may be lost to follow-up or may drop out of a study in a survival analysis; this incomplete observation of time to failure is known as censoring [2]. Censoring in a survival analysis is supposed to be noninformative [2]; that is, in principle, it should not influence the likelihood that an event of interest (such as revision for polyethylene wear) might occur. However, this assumption is often violated. For example, if some patients died of cancer before polyethylene failure occurred, that would distort the results of a conventional survivorship approach such a Kaplan-Meier approach, since it is not possible for a patient who has died to subsequently experience polyethylene failure [2, 10]. In this article [3], in which the primary outcome was time to revision for polyethylene damage, death because of a tumor served as a competing event because the patient will not have polyethylene failure once he or she has died of cancer. Generally, the presence of competing risks leads to an overestimation of the cumulative incidence of the event [2, 5], so the incidence of polyethylene failures might have been overestimated if the authors did not adjust for the presence of competing risks. I congratulate them for doing so [3].

The authors [3] used a Fine-Gray subdistribution hazard model to compare the time to failure of two different polyethylene components. This model considers those who are currently event-free as well as those who have experienced a competing event and is preferred for estimating the incidence or predicting prognosis in the presence of competing risks [1, 2]; methodologically, the authors chose wisely in using it. Absolute percentages of competing events of more than 10% should be accounted for [2]; in this study [3], 10% to 20% of patients in each group died of cancer during the follow-up period. This helps us trust their survivorship estimates. Based on their findings, surgeons will be able to use the latest polyethylene components without reservation and patients will enjoy long survivorship of their megaprostheses.

Where Do We Need To Go?

The ultimate goal is to develop durable megaprostheses that do not result in revisions caused by polyethylene damage in patients with tumors, who are likely younger than those who undergo TKA for osteoarthritis. Therefore, the duration of follow-up must be much longer than that for conventional TKA because many patients with osteosarcoma are teenagers; 10-year follow-up is apparently insufficient. Young patients who survive their tumor diagnoses may need implants to remain in service for 50 years or more. Moreover, these patients are active and some of them are interested in sports activities. Belzarena et al. [3] were able to show that the latest polyethylene components had a lower hazard ratio for revision in 10 years, and I am interested in how long these components will survive in the still-longer term.

The authors [3] analyzed several polyethylene-related factors such as molding, processing, sterilization, and packaging together, and it was not clear which factor contributed to the longevity of the polyethylene components. Considering the relatively small number of patients with megaprostheses, each polyethylene-related factor can be more-efficiently tested in conventional TKA, although knee kinematics and reasons for failures may be slightly different for distal femoral megaprostheses.

One brand’s distal femoral megaprostheses were specifically evaluated in the study (Biomet, Warsaw, IN, USA), but several companies offer commercially available distal femoral megaprostheses. Orthopaedic oncologists may be interested in revision risks for each megaprosthesis when choosing one for their patients. Analyzing several megaprostheses may be challenging, but the outcome could be robust when a study is successfully designed. For example, using a national registry, Partridge et al. [9] compared conventional and highly cross-linked polyethylene components made by four different companies. Such a study helps elucidate the revision risk because of differences in the specific polyethylene material, regardless of other factors such as the shape, design, and manufacturer of components.

How Do We Get There?

Although we should aspire to randomized controlled trials for research about therapeutic interventions because they mitigate a number of important biases [4], it is not realistic to anticipate that randomized trials will address this set of research questions for several reasons. The needed trial duration—more than 10 years—would make such efforts prohibitively time-consuming and costly. It is likely that other new materials and devices would be introduced even before such a study is finished. Additionally, allocating patients to either an old-fashioned, traditional component or the latest component may not be ethical or feasible in some patients.

Instead, a cohort study using national and regional arthroplasty registries will be a powerful tool to evaluate the revision risk [6, 8]. A study from the United Kingdom took advantage of the National Joint Registry in comparing conventional and highly cross-linked polyethylene in primary TKA [9]. Because the number of patients who underwent megaprosthesis replacement is small compared with the number of patients who undergo regular arthroplasty, national or international collaborative work through the orthopaedic oncology community (such as the Musculoskeletal Tumor Society or International Society of Limb Salvage) to collect data from more patients is needed to address the issue. International work is unlikely to happen for a while amid the coronavirus disease 2019 (COVID-19) pandemic. I think the national registry of each country could be a more realistic tool, especially when national governments are trying to control infections. For example, in Japan, which has a universal healthcare system, starting in 2021, patients will be able to access their prescription drug history using their individual number card with an integrated circuit chip. If information about implants and prostheses is also registered in addition to medication, the system will serve as a flawless national database of prostheses, with less effort. I believe it is likely to happen soon, and if so, it means we took advantage of the COVID-19 crisis.