Purpose: The glenoid and humeral head curvature radii are mismatched in non-constrained total shoulder prostheses. The purpose of this study was to evaluate the effect of this mismatch on radiographic lucent lines and clinical outcome.

Material and methods: The study population included 319 total shoulder prostheses issuing from a multicentric European cohort. The patients underwent surgery for primary shoulder degeneration. Mean age at surgery was 67 years. Female sex predominated (75%). Partial supraspinatous tears were present in 7% of the shoulders with full-thickness tears also in 7%. A single type of prosthesis was used composed of a humeral pivot with a modular head (seven head sizes) and a cemented polyethylene flat-surface spiked glenoid (three sizes). Variable association of humeral heads and prosthetic glenoids defined the mismatch which varied from 0 to 10 mm (difference in the curvature radii between the head and glenoid). The patients were reviewed clinically and radiologically at a mean follow-up of 53.5 months (24–110 months). Clinical outcome was assessed with the Constant score for pain (15 points), daily activity (20 points), motion (40 points), and force (25 points). The glenoid lucent line was evaluated on the AP view using a 0 to 18 point scale (0=absence, 18=lucent line in 6 zones). Analysis of variance and linear regression were used to assess the effect of mismatch on the glenoid lucent line and clinical outcome.

Results: There was a statistically significant linear relationship between mismatch and glenoid lucent lines. The lucent line score was significantly lower when the mismatch was between 6 and 10mm. Mismatch had no influence on the overall Constant score or the individual scores (pain, motion, force, daily activities) nor on early or late postoperative complications.

Discussion: Based on the results of this study, the first in vivo assessment, the “ideal” gleno-humeral mismatch for total shoulder prostheses would be between 6 and 10 mm, i.e. much greater than is classically recommended (0–5 mm).