Pyrocarbon hemiprostheses show little glenoid erosion and good clinical function at 5.5 years of follow-up.

Background: The success of traditional shoulder hemiarthroplasty (HA) with cobalt-chromium heads is limited by painful glenoid erosion with problematic bone loss. Hemiprostheses with pyrolytic carbon (PyC) heads have shown reduced glenoid erosion in experimental laboratory studies. Few in vivo data are available.

Methods: We performed a single-center consecutive cohort study of 31 of 34 patients (91%) who underwent PyC HA between September 2013 and June 2018. In 11 of these patients, concentric glenoid reaming was additionally performed. The mean follow-up period was 5.5 years (range, 3.5-7 years). Standardized radiographs were taken, and clinical function (Constant score) and pain (visual analog scale score) were recorded. Anteroposterior radiographs were analyzed according to an established method by 2 independent observers: A line parallel to the superior and inferior glenoid rim was translated to the most medial point of the glenoid surface. A further parallel line was placed on the spinoglenoid notch. The distance between these 2 lines was measured. Measurements were scaled using the known diameter of the implanted humeral head component. To assess eccentric erosion, anteroposterior and axial images were classified according to Favard and Walch, respectively.

Results: Mean medial glenoid erosion measured 1.4 mm at an average of 5.5 years of follow-up. In the first year, 0.8 mm of erosion was observed, significantly more than the average erosion per year of 0.3 mm (P < .001). Mean erosion per year was 0.4 mm in patients with glenoid reaming vs. 0.2 mm in those without reaming (P = .09). An evolution of glenoid morphology was observed in 6 patients, of whom 4 had a progression of the erosion grade. The prosthesis survival rate was 100%. The Constant score improved from 45.0 preoperatively to 78.0 at 2-3 years postoperatively and 78.8 at latest follow-up (5.5 years postoperatively) (P < .001). The pain score on a visual analog scale decreased from 6.7 (range, 3-9) preoperatively to 2.2 (range, 0-8) at latest follow-up (P < .001). There was a weak correlation (r = 0.37) between erosion and pain improvement (P = .039) and no correlation between erosion and change in Constant score (r = 0.06).

Conclusion: PyC HA caused little glenoid erosion and a sustained improvement in clinical function in our cohort at mid-term follow-up. PyC demonstrates a biphasic development of glenoid erosion, with a reduced rate after the first year. PyC HA should therefore be considered as an alternative to cobalt-chromium HA and to anatomical total shoulder arthroplasty for patients with a high risk of glenoid component complications.