CORR Insights®: High Early Failure Rate After Cementless Hip Replacement in the Octogenarian

Where Are We Now? Although developed in 1962, the Charnley low-friction arthroplasty is still regarded as the “gold standard” for cemented stem fixation [2, 9]. However, it is clear that surgeon preference for fixation of the femoral stem in THA, and the factors that influence that choice, have changed. In one US community practice joint registry [8], use of cemented femoral stems dropped from more than 80% in 1996 to 3% in 2011. While we await data from the nascent American Joint Replacement Registry (AJRR), epidemiologic estimates suggest that 86% of primary THAs in the United States are cementless on both the femoral and acetabular side, less than 1% are fully cemented, and the remainder are hybrid constructs [3]. Similarly, worldwide registries have reported overall increases in the use of uncemented fixation, with a recent review [7] indicating use of uncemented fixation in primary THA varies from a low prevalence of 15% in Sweden to a high of 82% in Canada. Excellent results with cementless fixation, shorter operative times, and potentially less marrow embolization risk have all undoubtedly influenced this move toward cementless fixation. Yet, controversy persists regarding THA fixation techniques in older patients. Numerous single-institution studies [4, 10] have reported excellent mid-term results with cementless THA designs in octogenarians, but larger registry studies have either showed no difference [8], or outcomes favoring cemented THAs in older (≥ 70 years) patient groups [7]. Mortality associated with fixation method also remains controversial since comorbidities may influence both the fixation selected and mortality itself [5]. Where Do We Need To Go? The present study by Jämsen et al. utilized the PERFormance Effectiveness and Cost of Treatment (PERFECT) episodes database and the Finnish Arthroplasty Register to analyze 4777 THAs performed in octogenarians between the years of 1998 and 2009. The study revealed both the strengths and limitations of registry-based studies. The numbers allowed for more sophisticated sensitivity analyses, but the numbers in the cementless group remained relatively small and at risk of misinterpretation. Revision surgery was a relatively crude endpoint, as the authors acknowledged, but the age group studied lacked information for patients not considered medical candidates, patients who refused further surgical intervention, and patients who had a periprosthetic fracture treated with open reduction and internal fixation rather than revision THA. This missing data further confounded the results. The inclusion of surgeons throughout the country increased the generalizability of the results, but we know little about how the more recent adoption of cementless techniques in Finland, and its relatively low prevalence, may have contributed to the reported findings. In a community registry where nearly twice as many cementless stems compared to cemented stems were performed, researchers found no difference in cumulative revision rate between the fixation modes for patients older than 70 years [8]. Finally, the study contributed relatively little insight into prosthetic factors, despite the fact that one of the more common cementless devices used (Stryker ABG II® [Mahwah, NJ USA]) was identified as having a higher than anticipated revision rate by the Australian Orthopaedic Association [1] and has been the subject of negative reports [6]. Still, this report is another reminder to North American surgeons that a cemented femoral stem of time-tested design implanted with appropriate technique remains a superior and cost-effective alternative in older patients. How Do We Get There? “Which implant is best for my patient?” is the question the surgeon must answer for each THA patient. Surgeons must weigh many patient factors (age, activity level, and comorbidities), implant factors (design, characteristics, and performance), and surgeon factors (experience, comfort level with each device, and surgical approach). Randomized controlled trials seek to minimize these factors. They typically achieve this by doing three things: Including a selected patient population with similar characteristics, involving a known, limited and (usually) expert surgeon base, and generally by permitting only one (or very few) important difference(s) among the implants being tested. However, the difficulty of recruiting surgical patients into randomized controlled trials limits the number enrolled. The length of time required to get meaningful information on small numbers of implants occasionally results in published reports on implants that no longer are marketed. Registries can bludgeon their way to earlier conclusions on the strength of their numbers alone, but often lack granularity of detail that allows for appropriate risk-adjustment. How do we combine these two valid approaches to research on implant performance? Roentgen stereophotogrammetric analysis continues to be an underutilized tool in the introduction of new implants, and its use as part of randomized controlled trials should be encouraged as a way to get better information about implant survival at an earlier stage. But registries do hold the real key here; collaboration between registries as already demonstrated by the Nordic Arthroplasty Register Association, the International Society of Arthroplasty Registries, and the FDA-led International Consortium of Registries allows for vast numbers of procedures to be harmonized under shared definitions and implant catalogs. Registries are looking at better ways to incorporate more risk-adjustment data, patient-reported outcomes data, and radiographic findings into their analysis and reports. The development and ultimate maturation of the AJRR with the attendant US volume of arthroplasty will add immeasurably to this effort. It still will be up to the surgeon, however, to use this information to each patient's advantage.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.