Avascular necrosis (AVN) of the medial femoral condyle or, more rarely, of the medial tibial plateau presents anatomical features very similar to those of AMOA (focal loss of bone and cartilage in the medial compartment with the ligaments intact) and therefore is theoretically suitable for OUKA (Fig. 8.1). The aetiology of AVN in the knee is poorly understood although factors such as steroid use, arthroscopy, systemic lupus erythematosus or trauma are sometimes implicated.
Figure 8.1 Avascular necrosis lesions in (a) femoral and (b) tibial condyles.
Early disease can usually be treated conservatively. However, once the disease is established, particularly if there is collapse, arthroplasty is usually required. Preoperative investigation should include MRI in addition to radiographs. Unlike AMOA, varus stressed radiographs of knees with AVN lesions usually do not show full-thickness cartilage loss, as the tibial cartilage is often preserved. The MRI substantially overestimates the extent of the damage because of the surrounding oedema associated with the condition in the acute phase. When assessing the extent of the disease the oedema should be ignored (Figs. 8.2(a) and (b)). MRI is useful for confirming the diagnosis and also ensuring there are no other lesions.
Figure 8.2 Plain AP radiograph and MRI scan showing (a) AVN of the medial femoral condyle and extensive oedema and (b) AVN of the medial tibial plateau and extensive oedema.
The surgical technique is similar to that of AMOA. Apart from the damage to the medial condyle, the rest of the knee is usually completely normal. If there is a very deep femoral defect, care has to be taken to ensure that too much bone is not milled from the femoral condyle. Therefore, the primary milling with the 0 spigot should be stopped when the mill is seen, in the window, to be about 2 mm from the end stop. Even if there is a deep defect, we have found, particularly with the two peg femoral component, that good fixation can be achieved. Traditionally we have used cemented fixation and have filled the defect with cement. We have recently started using cementless fixation but do not as yet have enough data to determine if this is reliable. There is also concern that there may be an increased risk of tibial plateau fracture if a cementless tibia is used.
Marmor (1993) reported generally good results after UKA in 34 knees but with four failures at a mean follow-up of 5.5 years.
Langdown et al. (2005) reported the results of 29 knees with medial compartment osteonecrosis (26 femoral and three tibial) treated by OUKA (Phases 2 and 3). The mean follow-up time was 8.2 years (range 1–13 years). There were no revisions, and the mean postoperative Oxford Knee Score (OKS) (37.8 ± 7.6, where 0 = poor; 48 = excellent) was not significantly different from the score for a similar group of osteoarthritic knees, matched for age, sex, and time since surgery, also treated by OUKA. We recently reviewed 33 cases with SONK treated with OUKA and matched them to 68 cases with OA followed up for 6.2 years (range 1 – 14). The OKS was 40 (SD 8) in SONK and 39 (SD 9) in OA and in SONK the 10-year survival was 97% (95% CI 81 – 100). In an independent series from Yoshida et al. (2103) in Japan, there was a 10 year survival of 100% in 45 SONK cases with mean 6 year (range 1 – 14) follow up and OKS of 41. When the two series are combined, the survival of 99% at 10 years is relatively reliable as there are 78 cases.
We therefore conclude that although rare, SONK is a good indication for the OUKA.