High tibial osteotomy (HTO) is effective in treating isolated medial osteoarthritis of the knee, but subsequent deterioration is inevitable, and total knee arthroplasty (TKA) is then an option. The present study sought to compare TKA following medial opening-wedge HTO (OW-HTO) versus lateral closing-wedge HTO (CW-HTO) in terms of intraoperative data and clinical results. The study hypothesis was that there is no significant difference in clinical results or complications in TKA following OW-HTO or CW-HTO.

A retrospective multicenter (9 centers) study was conducted for the French Society of Orthopedic Surgery and Traumatology (SoFCOT), including 135 TKAs following HTO (58 OW and 77 CW) at a minimum 5 years’ follow-up. Mean interval between HTO and TKA was 134 months and was longer in case of CW-HTO (P<0.0001). Mean age at TKA was 65.4 years and older in case of CW-HTO (P=0.021). Tibial slope was greater in case of OW-HTO (P=0.024). Prior to TKA, 55.7% of patients could walk without canes, 98.4% found stairs difficult or impossible and only 19.1% could manage a walking distance greater than 1000m. Mean flexion was 110°; 54.2% of patients showed frontal knee stability and 87.8% sagittal stability; 60.1% had a mechanical axis in varus, without difference according to OW- or CW-HTO.

Hardware was almost systematically removed (in 98.5% of cases): in the same step for OW-HTO (P=0.018) or often in 2 steps for CW-HTO. The primary approach was generally re-used (54.2%), but less frequently in the CW-HTO group (P=0.0004). Lateral or medial ligament release was not associated in respectively 78.2% and 79.7% of cases. The TKA implant was usually without stem (87.2%) and was fitted using a conventional technique (74.4%). At a mean 87 months’ follow-up, 78.5% of patients could walk without canes, stairs were still difficult or impossible for 67%, and 74.1% could now walk further than 1000m; mean flexion was 110.5°. Overall, 91.5% of patients showed frontal knee stability and 98.2% sagittal stability, without difference according to OW- or CW-HTO. There were 15 complications within 3 months, more often in the OW-HTO group (12.3%) although not significantly, and with no difference in severity. Late complications comprised loosening (5.5%) and infection (3.6%) and were more frequent in the CW-HTO group (12%) (P<0.05).

The study hypothesis was partially confirmed. The only technical differences concerned hardware removal, often performed in two steps in case of CW-HTO, and TKA approach, which differed from the primary approach in case of CW-HTO. Clinical results were comparable between OW- and CW-HTO, but late complications were more frequent in the CW-HTO group.

III; comparative retrospective study.