Purpose: The purpose of this work was to analyse results of revision procedures for unicompartmental prostheses (UNI) replaced by total knee arthroplasty (TKA) and to detail technical difficulties.

Material and methods: This series included 54 UNI revised with TKA. There were 45 medial and nine lateral UNI. The cause of failure was determined from the clinical history, the postoperative x-rays following UNI implantation and the preoperative x-rays before TKA as well as the operation report. Mean time from failure of UNI to TKA was four years. Clinical results were assessed with IKS criteria. The radiological work-up included an AP and lateral view in single leg stance and goniometry in 44.5% of the cases. Twenty-seven patients were reviewed with standard x-rays, eight patients were lost to follow-up, and 19 were studied from the medical files. One patient died.

Results: Mean follow-up was four years (range 2–12 years). There were seven failures (13%). We noted six cases of phlebitis, two plumonary emoblisms, one secondary infection, and three mobilisations under general anaesthesia. 55.5% of the patients were very satisfied, 36% were satisfied and 8.5% were disappointed. Revision was considered easy or moderately difficult in 82% of the cases. A standard gliding TKA was implanted in 39 cases (72%). Mean knee score was 85 points. Mean flexion was 113°. The mean function score was 62 points. For 91% of the cases, there was no laxity. The femorotibial angle was 90° in 54% of the cases and the mechanical tibial angle was 90° in 46%.

Discussion: Our results are similar to those reported in the literature. They were better than reports of TKA revision TKA. Results of TKA revision of UNI have been good. There is no technical problem related to bone loss, generally at the tibial level (45%). Preoperative planning allows defining the revision modalities (long pivot, metal backing, revision prosthesis). We recommend a long tibial pivot in the event of important tibial bone loss.