Understanding the bony anatomic landmarks of the knee is imperative in the primary and revision TKA. Current TKA employs a jig at 3˚ of inbuilt external rotation in general referenced to the posterior condylar axis. Being a precision surgery in itself, it is important that one comprehends the variation between the ethnic groups so as to learn the differences and hence aim to deliver long-term success. The important finding in our study was that the clinical or the anatomical epicondylar axis on average is 5.59° externally rotated compared to the posterior condylar line. Hence an added minimum of 2.5° of external rotation in the excess of the inbuilt 3 degrees in the jig would help in the optimal position of the femur component.
Re-creation of the rotational alignment of the distal femur facilitates reduction in implant failure and improving clinical outcomes [3]. In the present scenario, there exists little consensus regarding the optimal rotational axis in the distal femur. The most frequently used ones are Whiteside’s line, Transepicondylar axis, and Posterior condylar axis. Cutting guides by various implant manufacturers generally rely on the bony landmarks and their relationship to each other. Berger et al [7] recommended the use of the surgical epicondylar axis as a reproducible secondary anatomical axis as a reference for rotational orientation of femoral component in TKA when posterior condylar surfaces can’t be used. Yoshioka et al used the clinical epicondylar axis in determining the condylar twist angle [8]. Interobserver variability in his study was 4˚. Arima et al introduced the WL (Whiteside’s line) which is considered superior in the Caucasian population [10, 11]. Victor et al in his systematic review quoted the posterior condylar axis to be on average 3˚ internally rotated relative to surgical epicondylar axis, 4˚ relative to the clinical or anatomical epicondylar axis, and 5˚ relative to the perpendicular to the Whiteside’s anteroposterior axis [12]. The direct visual assessment intra-operatively is often confounded by the difficulty in identifying the medial sulcus. The distal femoral cut takes into account the cartilage thickness. Suter et al [16] noticed the anatomical epicondylar axis to be more reproducible than the surgical axis in CT scans. Studies that employ computed tomography to measure the rotational relationship do not take the cartilage into account [17, 18]. We used MRI to assess the rotational alignment of the distal femur as the cartilage thickness would exert an effect on the angles.
A-TEA to PCL (CTA) averaged 5.59 ± 2˚ in our study. Yoshioka et al first used CTA as the reference for alignment [8]. In our study 97 of the subjects were found to have an angle more than 3˚ if one employs the anatomical epicondylar line as the reference axis. Furthermore, it signifies that 97% of the subjects had more than 3˚of expected internal rotation if the surgeon uses the relationship of PCL with A-EA and employs routine jig. Nagamine et al noted a condylar twist angle of 5.8 ± 2.7˚ in males and 6.2 ± 0.9˚ in females respectively [19]. Yoshioka et al reported 5˚ and 6˚ externally rotated A-TEA in relation to PCL [8]. Mullaji et al investigated the relationship using computed tomography and suggested a CTA of 5˚. The sample he used consisted of 42 men and 8 women [14]. The skewed statistical analysis is debatable and comparison of the same data with other ethnic population-based researches is not meaningful.
Arima et al described the AP axis of the distal femur in the Caucasian population.
It’s found to be a reliable axis in situations where the posterior condyles are dysplastic or in arthritic knees [10]. Tables 1 and 2 describe the variation of various distal femoral rotational angles with respect to gender, sex, and side. A-EA/S-EA was found to be 3.06 ± 0.53° in the younger population (< 50 years) and 3.33 ± 0.57° which was found to be statistically significant (p-value < 0.05). Although the clinical significance of the same is debatable and requires further investigation. Tanavalee et al did a comparative study between A-EA and S-EA for rotational alignment of the femoral component. They found that perpendicular to A-EA is closer to the AP axis and provided adequate external rotation to PCL and hence is more reliable [20]. The mean WL/PCL angle was 5.44 ± 2.88˚. The angle value indicates the component to be kept at 5.44 ± 2.88˚ external rotation to PCL to match the anteroposterior axis. Table 3 shows the comparison of various variables with that of pre-existing literature based on MRI evaluation [13, 15, 21].
In the Indian scenario, most surgeons use the epicondylar prominence for the femoral component position. The 3 degrees of inbuilt external rotation in the conventional jigs will impart about 3 degrees of external rotation if one uses PCL as a reference. This would lead to an improper rotational alignment of the component which can hamper the success of TKA in the Indian population. Mullaji et al found a CTA of 5.8˚ and proposed 2˚ additional external rotation while using posterior condyles to prevent internal rotation [14].
The surgeons should maintain an apt knowledge of the bony landmarks to achieve the precise rotational alignment of the femoral component. Though the surgical epicondylar axis represents the true rotational axis of the knee, intraoperatively assessing the sulcus may be at times a bit difficult.
Our study has got some limitations to be addressed. A larger sample is required to accurately define the relationship of these axes to each other. Preoperative computed tomography is considered the best tool to analyze the bony landmarks for rotational alignment in view of low intra- and inter-observer variability [16]. Evaluation of the cartilage portion of the distal femur is not possible with CT scans. MRI score over the CT scans has been used keeping this aspect in mind. We used MRI of knees to identify the relations of the rotational axis relative to each other. Thirdly, the mechanical axis of the subjects could not be studied which would have been ideal for assessing. Lastly, the age group distribution cut off at 50 years is an arbitrary value taken at the authors' discretion for evaluating the effect of age-related changes in distal femur morphology. An evaluation by taking age as a continuous variable might yield a more robust result.