RESUMO
PURPOSE AND HYPOTHESIS: The study aim was to assess the outcome of patellofemoral arthroplasty (PFA), paying particular interest to 'revisions for any reason'. The hypothesis was that there is a superior outcome of PFA reported in dependent clinical studies in contrast to independent clinical literature and that there is a superior outcome of 'trochlear-cutting' PFA in comparison with 'first-generation trochlear-resurfacing' implants. METHODS: Studies on PFA from its market introduction in 1955 onwards were systematically reviewed. The revision rate, which was calculated as 'revisions per 100 component years (CY)', was evaluated in 45 studies published in indexed, peer-reviewed international scientific journals. In addition, 'first-generation trochlear-resurfacing' and 'trochlear-cutting' implants as well as dependent and independent clinical literature were analysed. Furthermore, the data of three arthroplasty registers were analysed. RESULTS: A total of 15,306 PFA were included consisting of 2266 cases in worldwide literature data and of 13,040 cases in register data. 2.22 revisions per 100 CY were observed in worldwide literature data, which corresponds to a revision rate of 22.2% after 10 years. Revision rates between 18.9 and 27% after 10 years were shown by the included three national joint registers. In the group analyses no significant differences were detected. CONCLUSIONS: This meta-analysis did not reveal significant differences in the comparison between developer over independent publications and between 'first-generation-resurfacing' over 'trochlear-cutting' implants. In conclusion the data of developer publications do not seem to be biased. 'Trochlear-cutting' devices of PFA had slightly superior outcomes, but that benefit was not statistically significant. Nevertheless, we would recommend 'trochlear-cutting' devices for further use in PFA. LEVEL OF EVIDENCE: Meta-analysis of Level IV case series.