Limitations of radiographs in evaluating non-displaced osteochondral lesions of the talus.

Abstract

To evaluate what limitations, if any, radiographs have in detecting and characterizing the morphology of non-displaced OCLTs (size, cystic change, fragmentation, and avascular necrosis [AVN]). Thirty-three OCLTs in 32 patients were reviewed in consensus by a board-certified, fellowship-trained musculoskeletal radiologist and orthopedic surgeon, on radiographs and MRI examinations performed within 15 days of one another. Location, dimensions, and characteristics of the OCLT (fragmentation, bone marrow edema, cystic change, and necrosis) were documented on the radiographs and MRI examinations. There was an average of 7.5 days (range: 1-15 days) between the MRIs and radiographs. Eighteen (55%) medial and 15 (45%) lateral OCLTs were found; none was displaced. OCLTs were categorized into three groups: Group 1-initially diagnosed with radiographs (4/33; 12%), Group 2-diagnosed with radiographs only after MRI correlation (20/33; 61%), and Group 3-not identifiable on radiographs despite MRI correlation (9/33; 27%). The dimensions of all of the lesions in Groups 1 and 2 were underestimated on radiographs. Only 1 of the 4 (25%) lesions in Group 1 and 1 of the 20 (5%) lesions in Group 2 could be measured in the anteroposterior (AP) dimension (using a lateral radiograph). The most common radiographic appearance of non-displaced OCLTs was an "ill-defined" lucency at the talar dome (20/33; 61%). Across all three groups, fragmentation, cystic change, and AVN were radiographically apparent in 3/10 (30%), 4/19 (21%), and 1/6 (17%) cases respectively. Compared with MRI, radiographs are limited in their evaluation of the size (particularly the AP dimension) and characteristics (fragmentation, cystic change, and AVN) of non-displaced OCLTs. The most common appearance of non-displaced OCLTs is an "ill-defined" lucency at the talar dome. When this appearance is also considered, the estimated retrospective sensitivity of radiographs improves considerably.