Component malrotation in total knee arthroplasty (TKA) is a reason for early failure and revision. Assessment of possible component malrotation using computed tomography (CT) might be useful when other differentials have been excluded. The aims of our study were to determine the proportion of symptomatic patients with component malrotation on CT, and review the subsequent management of such patients.

A retrospective review of case notes was performed locally for all patients who had a CT scan for a painful TKA. Measurements of the femoral and tibial component rotations were done according to the standard Berger protocol, giving net degrees of either external rotation (ER) or internal rotation (IR). Any subsequent surgery was noted, and patients were followed up as per local practice.

Between 2007 and April 2012, 69 knees in 68 patients had CT scans. There were 25 males and 43 females, and mean age at primary surgery was 65.03 years. The mean femoral component rotation for all knees was 0.1° ER (range 7.0° ER – 6.7° IR), and the mean tibial component rotation for all knees was 19.1° IR (6.6° ER – 37.0° IR). No statistically significant difference was found comparing the mean femoral and tibial component rotations between patients with and without further surgery. Further surgery was performed on 39 (56.5%) knees.

Overall, there were ten cases (14.5%) of isolated femoral malrotation, 26 tibial malrotation (37.7%), and two cases (2.9%) had malrotation of both components. Out of these 38 cases, secondary surgery was performed in 22 knees (57.9%), of which a satisfactory outcome was achieved in fifteen cases (68.1%).

It is impossible to establish component malrotation as the only cause of pain following TKA, however, our study does show that the Berger protocol has its uses when other causes have been excluded.