All-polyethylene tibial components are equal to metal-backed components: systematic review and meta-regression

Abstract

Background: Less than 1% of all primary TKAs are performed with an all-polyethylene tibial component, although recent studies indicate all-polyethylene tibial components are equal to or better than metal-backed ones.

Questions/purposes: We asked whether the metal-backed tibial component was clinically superior to the all-polyethylene tibial component in primary TKAs regarding revision rates and clinical functioning, and which modifying variables affected the revision rate.

Methods: We systematically reviewed the literature for clinical studies comparing all-polyethylene and metal-backed tibial components used in primary TKAs in terms of revision rates, clinical scores, and radiologic parameters including radiostereometric analysis (RSA). Meta-regression techniques were used to explore factors modifying the observed effect. Our search yielded 1557 unique references of which 26 articles were included, comprising more than 12,500 TKAs with 231 revisions for any reason.

Results: Meta-analysis showed no differences between the all-polyethylene and metal-backed components except for higher migration of the metal-backed components. Meta-regression showed strong evidence that the all-polyethylene design has improved with time compared with the metal-backed design.

Conclusions: The all-polyethylene components were equivalent to metal-backed components regarding revision rates and clinical scores. The all-polyethylene components had better fixation (RSA) than the metal-backed components. The belief that metal-backed components are better than all-polyethylene ones seems to be based on studies from earlier TKAs. This might no longer be true for modern TKAs.

Level of evidence: Level II, therapeutic study. See Guidelines for Authors for a complete description of levels of evidence.

Figures

Fig. 1

The risk difference in revision rate for any reason between all-polyethylene and metal backed tibial components is shown. N rev = number revised; RE model = random effects model.

Fig. 2

The risk difference in revision rate for aseptic loosening between all-polyethylene and metal-backed tibial components is shown N rev = number revised; RE model = random effects model.

Fig. 3

The significant influence of the start date of the study on the primary outcome is shown. More recent studies favor the all-polyethylene over the metal-backed components.

Fig. 4

The mean difference in KSS between all-polyethylene and metal-backed tibial components is shown. The mean difference of 0.63 in favor of the all-polyethylene is not significant and not clinically relevant. KSS = The Knee Society Score; RE Model = random effects model.

Fig. 5

The mean difference in the KSS Function between all-polyethylene and metal-backed tibial components is shown. KSS F = The Knee Society Function Score; RE Model = random effects model.

Fig. 6

The mean difference in the Hospital for Special Surgery score between all-polyethylene and metal-backed tibial components is shown. HSS = Hospital for Special Surgery; RE Model = random effects model.

Fig. 7

The forest plot of the MTPM measured using RSA is shown. The MTPM is significantly in favor of the all-polyethylene design. MTPM = maximum total point motion; RSA = radiostereometric analysis.

Fig. 8

A funnel plot of the included studies shows there is publication bias, because there is an appearance of missing studies in the upper left corner.

Fig. 9

The plot shows the ‘trim and fill method’. The solid circles are the original data. The open circles on the upper left side are the ‘missing’ studies, when symmetry is assumed. This figure shows these studies have no effect on the center of the risk difference; therefore, publication bias has little impact on the pooled treatment effect.