Abstract
The role of unicompartmental knee arthroplasty (UKA) in spontaneous osteonecrosis of the knee (SONK) remains controversial, even though SONK usually involves only medial compartment of the knee joint. We aimed to compare the survival rate and clinical outcomes of UKA in SONK and medial compartment osteoarthritis (MOA) via a meta-analysis of previous studies. MEDLINE database in PubMed, the Embase database, and the Cochrane Library were searched up to January 2018 with keywords related to SONK and UKA. Studies were selected with predetermined inclusion criteria: (1) medial UKA as the primary procedure, (2) reporting implant survival or clinical outcomes of osteonecrosis and osteoarthritis, and (3) follow-up period greater than 1 year. Quality assessment was performed using the risk of bias assessment tool for non-randomised studies (RoBANs). A random effects model was used to estimate the pooled relative risk (RR) and standardised mean difference. The incidence of UKA revision for any reason was significantly higher in SONK than in MOA group (pooled RR = 1.83, p = 0.009). However, the risk of revision due to aseptic loosening and all- cause re-operation was not significantly different between the groups. Moreover, when stratified by the study quality, high quality studies showed similar risk of overall revision in SONK and MOA (p = 0.71). Subgroup analysis revealed worse survival of SONK, mainly related to high failure after uncemented UKA. Clinical outcomes after UKA were similar between SONK and MOA (p = 0.66). Cemented UKA has similar survival and clinical outcomes in SONK and MOA. Prospective studies designed specifically to compare the UKA outcomes in SONK and MOA are necessary.
