To the Editor:

Dr. Williams has correctly pointed out that his group previously studied patellofemoral kinematics under weight bearing loads using MRI, and should be acknowledged for this important contribution to the orthopaedic literature. We would like to apologize for the misprint of the negative sign in front of the medial translation numbers in our table. This is clearly inaccurate and should not have happened.

However, the relative changes in patellar position reported in the paper appear accurate. The magnitude of translation was obtained from the table reported by Tennant et al (1). Patellar centralization at 0 degrees is 0.43 while at 30 degrees it is –2.42 giving a total change of 2.87mm. At 55 degrees the value given is –1.31 yielding a change of 1.74. (It could be argued that the numbers could have been calculated from hyperextension to 30 and 55 degrees of flexion; however since most papers have reported values from 0 to maximum flexion, these numbers seemed most consistent. Had we calculated the number from their table based on hyperextension, the translation would have been greater: 5.8 mm at 30 degrees and 4.6 mm at 55 degrees.)

Dr. Williams indicated that our method of simulated weight bearing is not representative of normal patellofemoral loading. Static loading used for MRI studies eliminates dynamic external forces and moments applied to the knee during gait. However, we believe that maintenance of a constant load is important to create a consistent quadriceps contraction force during the time required for MRI scanning, and therefore chose a pulley and weight apparatus. While Dr. Williams’ method permits subjects to be scanned in a more upright position, it appears that the subjects can use their upper extremities during the scan which may decrease quadriceps loads if muscle fatigue occurs. Muscle fatigue may have occurred in the authors study, and may have been manifested as the shaking and leg fatigue the authors refer to on Page 160, not typical in physiological load bearing, as well as the large standard errors reported by the authors.

We also recognize, that fatigue may not be the only factor contributing to the large the standard errors, which could also be depend on the signal to noise ratio, which theoretically might be lower at 0.5 Tesla. We acknowledge that both techniques are not entirely physiological, , but as we stated in the paper, our belief, which was based on the relative importance we ascribed to each of these factors.

Regardless of these differences, we are also pleased that the results in our studies are similar, giving strength to both papers.

Sincerely, Sharmila Majumdar, Ph.D.

1. Tennant S, Williams A, Vedi V, Kinmont C,Gedroyc W, Hunt DM: Patello-femoral tracking in the weight-bearing knee: A study of asymptomatic volunteers using dynamic magnetic resonance imaging: apreliminary report. Knee Surg,Sports Traumatol., Arthrosc 2001; 9: 155- 162.

To The Editor,

The strength of the paper by Patel, et.al, lies in the three-dimensional imaging and the resolution of their images but not, as they claim, in it being the first paper to employ MRI to analyze the patellofemoral joint under weight-bearing conditions.

The paper by Tennant et al[1], of which I am an author, and quoted by Patel et al, was published in 2001. It employed an ‘interventional’ MRI scanner which has a vertical gap between two magnetic coils allowing a subject to stand and have his/her knee imaged during a weight-bearing squat. The subject leans back against a support [10 degrees off vertical] making the ‘squat’ equivalent to a wall-slide. This technique more likely replicates normal patellofemoral loading than when the patient is supine in a conventional MRI scanner as reported in the paper by Patel et al. Yet in the Discussion section of their paper Patel et al list our technique as less physiological than theirs.

It was extremely reassuring that all of our findings have been reproduced in this study [with the exception of minor quantitative variation]. Unfortunately the reader might think that a significant difference existed. In Table 1 of their paper a key finding of ours is misquoted. In that table,the medial translation that we reported has a minus sign against it suggesting we measured lateral translation of the patella during early flexion- we did not. All of our 40 knees [20 subjects] had patellae that started laterally in full extension but moved medially on engaging the trochlea. All were centralised by 30 degrees but most by 10 degrees.

Whilst it is reassuring to read a paper of a good quality that supports our published results it is rather a pity that inaccurate claims and referencing have occurred.

Yours sincerely,

Andy Williams

Consultant Orthopaedic Surgeon And Director of Orthopaedic Research, The Interventional MRI Unit, St. Mary’s Hospital, London W2 U.K.

Reference:

1 Patello-Femoral Tracking in the Weight-Bearing Knee: a study of asymptomatic volunteers utilising dynamic magnetic resonance imaging: a preliminary report. Tennant S, Williams A, Vedi V, Kinmont C, Gedroyc W, Hunt D. Knee Surg, Sports Traumatol., Arthrosc 2001; 9: 155-162.