Can False Profile View Radiographs of the Hip accurately identify CAM lesions? A retrospective radiographic analysis with computed tomography comparison

Abstract

Dunn, Frog Leg Lateral, and AP radiographic views have been verified to accurately correlate with three-dimensional imaging while assessing for CAM lesions of the hip. The false profile radiographic view of the hip has traditionally been used to measure acetabular coverage in the setting of femoroacetabular impingement but some believe it can also be useful in assessing the morphology of the femoral head-neck junction. The objective of this study was to examine the accuracy of using a false profile radiograph of the hip to identify CAM lesions as compared to three-dimensional CT scan findings. An a prior power analysis using the results from a previous study showed we needed 45 patients to have (1-β>0.80. Patients undergoing hip arthroscopy between January 2011 and September 2012 were reviewed. Patients were included if they had a CT scan performed at our institution and if they had adequate screening radiographs in our PACs system(AP pelvis, 90-degree Dunn lateral and false profile). Alpha-angle measurements were performed on radial oblique CT scans and plain radiographs as previously described by Notzli. 2 physicians measured each x-ray and then re-measured 3 months later. An alpha angle > 50° was used as a cut-off for abnormal. SPSS was used to perform all statistics. 45 patients were reviewed, 24 females and 21 males, 32 right hips and 13 left hips. The average age was 35 yo (17-53 yo). 32 hips had a cam lesion on reformatted CT (α-angle > 50°). 31 hips had a cam lesion present about the 2 o'clock position of the femoral head-neck junction and 19 hips had a cam lesion present about the 3 o'clock position. The false profile view was found to be 60% sensitive and 89% specific at identifying cam lesions. When combined with the AP pelvis and 90-degree Dunn lateral views, this series was found to be 86% sensitive and 90% specific. The false profile view had a strong correlation to both the 2 o'clock and 3 o'clock views (r = 0.75 and 0.70 respectively). The False Profile view has a 76% agreement with the 2 o'clock view and an 83% agreement with the 3 o'clock view. Inter and Intra-Observer Correlation Coefficients were 0.81 & 0.88 respectively. This is the first study to suggest that the false profile view can be effectively used to assess the femoral head-neck junction. This has important implications for diagnostic and treatment purposes, as the false profile view has traditionally only been used to only assess for anterior acetabular coverage (pincer-type impingement). We have found that the false profile view effectively characterizes CAM lesions about the anterior portion of the femoral head-neck junction from 2 o'clock to 3 o'clock. Measuring the alpha angle on this view appears to be a reproducible with a significant ICC. In combination with the AP and 90-degree Dunn lateral views the false profile view allows for a physician to fully assess for CAM-lesions about the anterosuperior head neck junction.