Bone support correlation of X-Ray and CT for a new PE-glenoid

The radiographic evaluation of novel cementless anatomic polyethylene (PE) glenoid components featuring a titanium-coated back is still unclear. This study explores potential radiolucent lines (RLL) between the radiopaque titanium layer and sclerotic convex reamed bone in an intermodal comparison analysis with computed tomography (CT) scans.

Eight RM pressfit vitamys glenoids (Mathys^®) were implanted into cadaveric scapulae. In the CT scans, glenoids were quantified by evaluating ideal complete bony support (NO GAP) and gap between bone and titanium coating (GAP). X-rays were in perfect 0-degree projection and tilted in ± 10° and ± 20° mediolateral (ml) and craniocaudal (cc) directions. Radiographs evaluated were graded as NO RLL, RLL (gap > 1 mm) or DL (double line, gap < 1 mm) in an intermodal comparison of CT and X-ray findings.

The inter-rater (Cohen's = 0.643) and intra-rater reliability (Cohen's = 0.714) were good. The overall evaluation showed a significant agreement between (NO) RLL on X-ray and (NO) GAP on CT (p < 0.001). The - 10-degree ml projection showed good agreement between CT and X-ray (Cohen's = 0.628). Adequate agreement was shown at 0 degrees (Cohen's = 0.386), + 10 degrees ml (Cohen's = 0.338), and + 20 degrees cc (Cohen's = 0.327). Compared to the scenario DL = NO RLL, the true a.p. view showed better sensitivity when the DL is classified as RLL. Conversely, the true a.p. view demonstrated both better specificity and significant agreement between the X-ray and CT findings in scenario when DL = No RLL.

Standard true a. p. projections are reliable in ruling out gaps when no RLL or DL is visible and the detection of RLL shows high intermodal agreement. Varying agreement across tilting angles emphasizes the importance of a comprehensive approach in evaluating bone support and CT is indispensable for a scientifically reliable assessment.

Level III Treatment Study.