[Comparison and statistical evaluation of two methods of condylar twist angle measurement in total knee arthroplasty]

The aim of this prospective randomised study was to compare and statistically analyse two methods of condylar twist angle (CTA) measurement in total knee arthroplasty in order to assess their applicability in routine practice.

The study included 238 patients with 256 sites undergoing total knee arthroplasty (TKA) in the period from January 2009 to May 2011. There were 93 men (nine with bilateral TKA) and 145 women (nine with bilateral TKA) with an average age of 69.3 years and a range of 47 to 88 years. The implants NexGen (Zimmer) and ADVANCE® Medial-Pivot Knee (Wright) were used. In each patient, CTA was measured before surgery by the radiologist on a multidetector CT SOMATOM 64 (Siemens) using the Yoshioki method. The other CTA measurement was made intra-operatively by the surgeon using our modification of the Hofmann method which involved the identification of a reference line for optimal rotational alignment of the femoral component. A STATISTICA 9.0 software package was used for statistical analysis. In addition to basic statistical data, selected data were presented in graphical forms as Box and Whisker's plots and histograms. Changes in CTA and differences between the groups were evaluated using the Wilcoxon signed-rank test. Relationships among the variables were studied using Spearman's correlation coefficient.

The statistical analysis showed that the pre-operative CTA value obtained from CT scans was, on the average, higher by 0.5 degrees than the value from intra-operative measurement, as assessed at the level of significance p = 0.001 (signed- rank test). The intra-individual variability was lower than the inter-individual one (14.4% and 30.8%, respectively). This means that both methods are suitable for CTA measurement in the knee joint replacement procedures. Spearman's correlation coefficient was 0.6, which is the value of medium strong correlation. The post-operative CTA assessed on CT scans was in the range of 0 to 2 degrees in 87.5% of the patients. Both the pre-operative and intra-operative CTA values were significantly higher in women than in men (Wilcoxon two-sample test). There was no statistical difference between the left and the right side.

Malrotation of the femoral component is one of the causes leading to patellar subluxation and pain in the front part of the knee. The post-operative CTA value should be zero. Optimal rotational alignment of the femoral component varies with each patient; in our study it was found in the range of 0 to 7 degrees on the basis of CTA values. We do not recommend 7 degrees of external rotation to be exceeded because of the risk associated with balancing the flexion gaps; nor do we recommend to set internal rotation of the femoral component for the risk of patellar complications. The difference of 0.5 degree found in the CTA value between the two measuring methods can be explained by individual differences in the anatomy of the medial epicondylar region, and by the use of only selected whole numbers (0, 3, 5, 7) in intra-operative measurements. This difference does not play any role in routine surgical procedures.

The statistical evaluation of the results of CTA measurement with the two methods showed that both were equally suitable for routine total knee arthroplasty. The results of intra-operative CTA measurements are comparable with those obtained on CT scans; in addition, the intra-operative method is less expensive and eliminates exposure of the patient to radiation. CT-based CTA measurements are useful in the patients with chronic problems after TKA in whom femoral component malrotation needs to be either confirmed or ruled out.