Abstract
Introduction
Cementless fixation in total knee arthroplasty has been proposed to offer advantages long-term once osteointegration has occurred as there is no substrate between the implant and the bone to fail. Radiostereometric analysis (RSA) is a useful tool to study fixation, but typically focused on early migration in the first two post-operative years. Few studies have looked at 10-year RSA migration in cementless fixation and those that have contain small numbers of subjects. The objective of this study was to compare implant migration and inducible displacement between cemented and cementless TKA at 10 years and to compare the 10-year migration to the 2-year data in an effort to validate the predictive modelling of RSA.
Methods
Subjects who had previously participated in RSA migration studies with 2-year follow-up were recruited to return for a long-term follow-up exam, at least 10 years from their surgery. The implants under study included two cemented designs from two manufacturers and one porous metal monoblock cementless design. At the 10-year visit, subjects had supine RSA exams to determine long-term migration as well as a loaded exam (single leg stance) to determine inducible displacement. Differences between cemented and cementless groups were evaluated with the Mann Whitney U test and Spearman's rank correlation coefficients were calculated for early and late migrations. Significance was set at p < 0.05.
Results
Seventy-five subjects were available for long-term follow-up, with average time since surgery of 12 years. This cohort contained 51 women and 24 men with cemented tibial components in 53 cases (37 female) and cementless tibial components in 22 cases (14 female). At the time of surgery, the subjects were 62±7 years old with BMIs of 33±6 m/kg2 (mean±standard deviation). Median migration at the long-term follow-up was 0.6 mm (MTPM; range 0.2–2.8 mm) and was not different between the cemented and cementless groups (p = 0.9, Mann Whitney U Test).
Inducible displacement at 10 years was significantly lower for the cementless implants (p<0.001, Figure 1).
Migration at one, two, and 10 years did not correlate with inducible displacement at 10 years. However, migration at one year and two years did correlate with long-term migration, with the strongest correlation at two years (Spearman's rank correlation coefficient for all components = 0.74, p < 0.001, Figure 2).
Conclusion
Although long-term migration was not different for cemented or cementless (porous metal monoblock) tibial components, inducible displacement at the 10-year visit was significantly lower for these cementless components, suggesting superior fixation. Additionally, long-term migration was strongly correlated to two-year migration, regardless of fixation. These findings support the predictive value of short-term migration in determining long-term fixation.
For any figures or tables, please contact authors directly.
