LAUSR

BACKGROUND Little is known about the way the newest generation of stems integrate into the proximal humerus and their effect on the surrounding bone. Factors that may influence ingrowth have not been investigated. METHODS A consecutive cohort study was conducted that examined 74 anatomical, reverse, or pyrocarbon hemiprostheses, using a curved modular short stem with a proximal porous coating 2-5 years postoperatively (mean 35 months). X-rays were reviewed by 2 examiners independently. Bone loss was scored with 1 point per zone with partial and 2 points per zone with complete resorption (10 zones). The Constant score was used for clinical correlation. Multiple linear regression was employed to investigate correlations between variables. RESULTS No subsidence or shift of the stems occurred. Two of 74 patients showed 1 zone of periprosthetic lucency of 1 mm. The filling ratio averaged 0.54 (range: 0.36-0.75). Thirty patients (40.5%) displayed bone resorption, first seen at 16.6 months (range: 3-40 months), commonly in zones 1 and 5. A total of 22 patients had ≥1 zone with partial resorption, and 8 (10.8%) developed full thickness resorption after 32 (range: 10-49) months.One new finding was that female sex and older age accounted for 51% of the variation of the filling ratio. A high filling ratio, especially when >0.55, correlated with bone resorption (P < .001). Age, sex, and prosthesis type did not directly predict bone resorption. Bony sclerosis correlated with a high filling ratio (P = .019) and thereby indirectly with resorption. A direct correlation between sclerosis and resorption was narrowly insignificant (P = .058) once correcting for the filling ratio. Reverse shoulder prosthesis had a higher filling ratio than total shoulder prosthesis patients (P < .001), resulting indirectly in more bone resorption. The preoperative diagnosis did not significantly correlate with the filling ratio (P = .59) or the resorption score (P = .69). A varus or valgus alignment did not predict resorption (P = .21) or the formation of sclerotic lines (P = .93). Bone loss did not correlate with clinical results. CONCLUSIONS These short stems are firmly anchored 2-5 years postoperatively. However, significant bone loss, linked to a high filling ratio (>0.55), is observed proximally around these stems. The development of sclerotic lines around the stem indicates oversizing. Other factors were not found to have a significant effect on stem ingrowth. The implantation of stems with a large filling ratio is more common in older females and in patients receiving reverse shoulder prosthesis. Autologous impaction bone grafting could downsize the required stem. If adequate hold is not afforded by a suitably small stem, cementation is advisable.