The implementation of knee arthrodesis has become synonymous with limb salvage in the presence of chronic sepsis and bone loss around the knee. This can be seen in failed trauma surgery or knee arthroplasty as an alternative to trans-femoral amputation. There is no prior literature assessing which factors affect knee arthrodesis using external fixation devices.

Sixteen consecutive patients (four women and twelve men) made up of eleven infected knee implants, three internal fixations of the tibial following fractured tibial plateau as well as 2 infected native joints were identified. The mean age at initial surgery was 56 years (range 25 to 82 years). All procedures were performed under the direct supervision of the limb reconstruction teams using a standard protocol with either a Taylor spatial frame or Ilizarov frame. The patient records, microbiology results and radiographs of all patients who underwent knee arthrodesis at this institution between 1999 and 2010 were reviewed.

Of the 16 patients in this study knee fusion occurred in eleven patients (69%). The five patients where arthrodesis failed all had significant bone loss on the pre-operative radiographs and confirmed at surgery. We found a relationship between a significant infection of the knee with MRSA and failure to fuse. Three of the five patients had MRSA isolated from inside the knee at some stage during their treatment.

The five patients where fusion failed were on average older (mean age 63 years against 51 years) and had more extensive bone loss. Those who failed to fuse had more co-morbidities. We would conclude that where there is little or no bone loss, arthrodesis of the knee can be reliably achieved with the use of circular frame fixation. A greater number of negative factors also prolongs the amount of time spent in the external fixator. The presence of significant bone loss, infection, increased age and multiple co-morbidities requires careful evaluation and consideration of trans-femoral amputation as an alternative.

Lower limb mal-alignment due to deformity is a significant cause of early degenerative change and dysfunction. Standard techniques are available to determine the centre of rotation of angulation (CORA) and extent of the majority of deformities, however distal femoral deformity is difficult to assess because of the difference between anatomic and mechanical axes. We found the described technique involving constructing a line perpendicular to a line from the tip of the greater trochanter to the centre of the femoral head inaccurate, particularly if the trochanter is abnormal. We devised a novel technique which accurately determines the CORA and extent of distal femoral deformity, allowing accurate correction.

Using standard leg alignment views of the normal femur, the distal femoral metaphysis and joint line are stylized as a block. A line bisecting the axis of the proximal femur is then extended distally to intersect the joint. The angle (Θ) between the joint and the proximal femoral axis and the position (p) where the extended proximal femoral axis intersects the joint line are calculated. These measurements can then be reproduced on the abnormal distal femur in order to calculate the CORA and extent of the deformity, permitting accurate correction.

We examined the utility and reproducibility of the new method using 100 normal femora. Θ = 81 ± sd 2.5. As expected, Θ correlated with femoral length (r=0.74). P (expressed as the percentage of the distance from the lateral edge of the joint block to the intersection) = 61% ± sd 8%. P was not correlated with Θ.

Intra-and inter-observer errors for these measurements are within acceptable limits and observations of 30-paired normal femora demonstrate similar values for Θ and p on the two sides.

We have found this technique to be universally applicable and reliable in a variety of distal femoral deformities.