[Patterns of failed internal unicompartmental knee prostheses, allowing persistence of undercorrection]

was to investigate how deliberate slight undercorrection of varus deformities influenced the patterns of failure seen following medial unicompartmental knee replacements.

Between 1980 and 1989, 335 medial unicompartmental knee replacements were performed using the St George Sled prosthesis. The replacements were preformed so that the postoperative mechanical axis, as defined as the line joining the centre of the femoral head to the centre of the talus, passed medial to the centre of the knee joint and through the replaced medial compartment. The mean follow-up for these patients was 6.4 years (range of 2 to 12). By 1992, 33 of these cases had required revision. The radiographs and case notes of cases were available for study.

There were no significant differences in the demographic details between those coming to revision and the successful group. Analysis of the post-operative alignments in the successful group revealed that the desired undercorrection was achieved in 73 per cent of cases. 2 per cent were left grossly undercorrected with their mechanical axis passing medially outside the medial compartment. In 20 per cent of cases the mechanical axis was almost fully restored passing through the centre of the knee. 4 per cent of cases had an overcorrection with the mechanical axis passing through the lateral compartment. Early failures (N = 11), classified as those cases in whom satisfactory result were never obtained, were primarily due to poor patient selection or to gross technical errors at surgery. Late failures (N = 22) were those that required revision for symptomatic failure after an initially successful primary arthroplasty and were due to progressive disease (N = 6), prosthetic failure (N = 22) were those that required revision for symptomatic failure after an initially successful primary arthroplasty and were due to progressive disease (N = 6), prosthetic failure (N = 10) and recurrent medial pain (N = 6).

Patients who otherwise meet the criteria for medial unicompartmental replacement, usually have a mild degree of osteoarthritis in the lateral compartment. Suddenly loading the lateral compartment, by fully correcting a varus deformity, could accelerate the degenerative process in this compartment. Progressive disease in the lateral compartment was seen in only 4 cases in our series. In 6 patients failure was associated with recurrent medial medial joint pain with increasing varus deformity. In these cases the mechanism of failure was probably the result of an excessive medial load from a gross undercorrection of the mechanical axis which passed close to or outside the medial aspect of the knee. The crude revision rate for our series with an average follow-up of 6.4 years was 10 per cent. After excluding the early failures, which resulted largely from the potentially avoidable problems the revision rate was 6.7 per cent.

The results from this series suggest that slight undercorrection of varus deformities does not produce any significant differences in the revision rates for unicompartmental replacements when compared to previous reports. Slight undercorrection would appear, nevertheless, to effect the relative importance of the modes of failure seen. Most notably there are fewer failures from disease in the lateral compartment. This potential advantage, however, is somewhat offset by failures from excessive loading of the medial compartment as technical errors at surgery may result in gross undercorrection of the mechanical axis.