My preferred method is setting the femoral component external rotation 3 degrees with posterior condylar line. Then, you check if it is consistent with AP axis and epicondylar axis.

Not suitable in cases of hypoplastic lateral femoral condyle which is not frequently seen

I use the method described by Eirik Aunan. I use measured resection technique in CR knees. https://www.ncbi.nlm.nih.gov/m/pubmed/28799446/?i=2&from=aunan%20e

Thank you for sharing your thoughts Ahmed Abdelbadie .

Ove Furnes thanks for sharing the new method described by the authors: CRA.

I believe in gap balancing method to determine the femoral component rotations. However at the same time I look at the bony landmarks (TEA, Whiteside line and PCA) making it a combined usage of measured resection and gap balancing to determine the rotations.

An interesting paper in which a meta analysis has been done comparing MR and GP methods to determine the component rotations in 8 RCTs, published in BJJ in 2017 reveals GB to be superior... I would love to hear more thoughts regarding this.

Article Meta-analysis of gap balancing versus measured resection tec...

The 3 degree ER based on the posterior condyles is ONLY good for Varus knees. Valgus knees need more ER as there is a lateral femoral condylar deficiency. I personally try to align to the epicondylar axis, as I have trouble accurately finding Whiteside's line. Often, there is erosion in the PF joint, and Whitesides line is distorted by that. Patient specific instruments are supposed to be very good at this, but I have my doubts.

@adam Harris: I agree with the increased ER requirement in valgus knees. Which of the transepicondylar axis do you prefer : surgical or anatomical ? I am of the belief that there is a lot of inter and intra surgeon variability in determining the epicondyles and hence marking this reference line. Any suggestions on improving the reproducibility?

PSI definitely helps as one can mark the epicondyles, amongst other anatomical landmarks, more accurately based on pre operative CT or MRI, hence improving the accuracy.

However at times there might be difficulty in accurate sitting of the PSI jigs on the femoral/tibial surface, thus introducing a play, affecting accuracy.

Thoughts ?

Go kinematic. A lot of these questions on component alignment then become redundant. My personal experience is that knees are easier too do using Steve Howlls technique, and they move in a more natural way. So: my answer to your specific question would be: 0 degrees rotation relative to the posterior condylar axis, unless there is posterior cartilage loss, which can be adjusted for by lifting the posterior foot off the bone on the affected side, by the thickness of the saw blade.

To address Dr. Aneja's question, yes, absolute intra-operative determination is only moderately reliable. I've used the anatomic axis.

Rather than "go kinematic", I'd suggest using a truly anatomic design, such as the Journey II.

I do the tibial cut first, then measure the gaps. My plan is for the MCL to be in equal tension throughout a ROM, and the LCL 2-3mm loose in flexion. The improved knee and patient happiness is dramatic. The downside is perhaps a 5% lateral release rate, which I do extra-articular / extra-synovial, with no apparent downside.

For the femoral axial rotation it depends on the implant philosophy. For the Attune and LCS implants, soft tissue flexion gap balancing is most commonly used. However, it is always worthwhile assessing your femoral rotation relative to Insall's (transepicondylar axis) line. This seems to be the most useful anatomical landmark to use considering the posterior condylar axis can very often be misleading secondary to wear and lateral femoral condylar hypoplasia.