Computed tomography in the assessment of periacetabular osteolysis

Background: Computed tomography recently has been proposed as an accurate method for diagnosing periacetabular osteolytic lesions. Several investigators have attempted to validate the accuracy of this technique, but they employed cadaveric and animal models, which cannot replicate the adaptive changes that occur over time in vivo. This study was performed to determine the accuracy of computed tomography in identifying and measuring periacetabular osteolytic lesions in hemipelves retrieved at autopsies of individuals with a previously well-functioning total hip prosthesis. Methods: We evaluated nine hemipelves, retrieved at autopsy, that contained a cementless porous-coated acetabular component. The fresh specimens were examined with conventional radiographs and computed tomography and then were embedded and sectioned into 1.5-mm slices for evaluation with slab radiographs. Anteroposterior and iliac oblique plain radiographs as well as axial, coronal, and sagittal computed tomography scans were reviewed to determine the presence and location of any periacetabular osteolytic lesions. These results were then compared with those identified on the slab radiographs. Lesion volume was calculated from computed tomography scans with use of post-processing software. Results: A total of twenty-three periacetabular osteolytic lesions were identified on the slab radiographs of the nine hemipelves. The plain radiographs identified twelve (52%) of the twenty-three lesions, and the computed tomography scans identified twenty (87%) of the twenty-three lesions. Three medial wall perforations were identified on the computed tomography scans but were not detected on the plain radiographs. Computed tomography was accurate in measuring the volume of the osteolytic lesions (r(2) = 0.997) but tended to overestimate the volumes measured on the slab radiographs. Periacetabular osteolytic lesions appeared on the computed tomography scans and slab radiographs as areas devoid of trabecular bone that were delineated by a sclerotic border and communicated with the joint space. Conclusions: In this autopsy model, computed tomography was an accurate method for detecting the location and measuring the volume of periacetabular osteolytic lesions.