High-flexion versus conventional total knee arthroplasty: a 5-year study.

PURPOSE: To compare mid-term outcomes of a high-flexion prosthesis with a conventional prosthesis.

METHODS: Records of 107 consecutive patients who underwent total knee arthroplasty (TKA) for primary osteoarthritis by a single surgeon were reviewed. 21 men and 36 women (mean age, 65 years) used a high-flexion prosthesis (NexGen CR-Flex Mobile, Zimmer), whereas 38 men and 12 women (mean age, 67 years) used a conventional prosthesis (Genesis II, Smith & Nephew) that preserves the posterior cruciate ligament, with mobile-bearing polyethylene inlay. The Knee Society knee and functional scores and the range of motion (especially maximum passive flexion) were assessed. Radiographs were evaluated to identify radiolucent lines at the bone cement interface, patella tracking, tibiofemoral alignment, and implant positioning.

RESULTS: The mean follow-up period was 68 (range, 51- 70) months. In terms of the mean Knee Society scores and range of motion, the difference between groups was not significant (p>0.05), but the improvement in both groups after TKA was significant (p<0.005). Respectively in the high-flexion and conventional TKA groups, 2 and one of the patients developed deep vein thrombosis and were treated with anticoagulants for 3 months. One and 5 of the patients had an unsatisfactory range of motion (<60º) in week 1, which was resolved with mobilisation under general anaesthesia in combination with a peridural catheter for analgesia. No patient developed implant-specific complications such as aseptic loosening (osteolysis, progressive radiolucent lines, implant deviation) or dislocation of the polyethylene insert.

CONCLUSION: The high-flexion prosthesis revealed no significant advantages over the conventional prosthesis in terms of the Knee Society scores and range of motion. Long-term studies are needed to determine whether the high-flexion implant is superior to the conventional implant in terms of polyethylene wear and aseptic loosening.