Plain radiographs are inadequate to identify overlengthening with a radial head prosthesis

PURPOSE: A correctly sized radial head implant helps restore more normal elbow motion and load distribution and allows the collateral ligaments to heal in an anatomic position. No single method of measurement has been agreed on. We hypothesized that plain radiographs could be used to evaluate the appropriate length of a radial head prosthesis when we simulated different patterns of ligament disruption. METHODS: Osteotomies of the medial and lateral condyles were created in 6 cadaver specimens to simulate 4 conditions: ulnar collateral ligament (UCL) and lateral ulnar collateral ligament (LUCL) intact, UCL intact, LUCL intact, and UCL and LUCL disrupted. Radial heads were resected and replaced with correctly or incorrectly sized monopolar radial head implants. Radiographs were taken after simulated ligament injury with a radial head prosthesis that was either 2 mm too short, the appropriate length, or too long by either 2 or 4 mm. Measurements were recorded between radiographic markers to determine ulnohumeral joint displacement. A 4 x 4 x 2 repeated-measures analysis of variance was performed, as well as post hoc analyses using Tukey's honest significant difference test, to determine statistically significant changes in our radiographic measurements. RESULTS: Results showed that disruption of at least 1 ligament alters the ulnohumeral joint separation on the lateral but not on the anteroposterior radiograph. Overlengthening the radial head did not result in a statistically significant change in ulnohumeral space. CONCLUSIONS: Replicating radial length is an important but sometimes difficult step during radial head replacement. Standard radiographic assessment of the ulnohumeral joint will not demonstrate marked changes in the ulnohumeral space when the implant is underlengthened by 2 mm or overlengthened by as much as 4 mm, as long as the joint remains located. Disruption of at least 1 collateral ligament alters the ulnohumeral joint separation on the lateral but not on the anteroposterior radiograph.