Monday, October 10, 2016

Glenoid erosion after shoulder hemiarthroplasty

Risk factors for glenoid erosion in patients with shoulder hemiarthroplasty: an analysis of 118 cases.

These authors studied the erosion of the glenoid in 118 shoulders (113 patients) at an average of 31 (range 5-86 months) after a humeral hemiarthroplasty performed for primary or secondary osteoarthritis or fractures.

24 cases of hemiarthroplasty were excluded from the study because they had a periprosthetic infection, severe Parkinson disease, or follow-up <1 year (two patients with a follow-up <1 year (5 and 10 months, respectively) were included because of severe erosion within that time).

At the postoperative visits,  a series of standardized radiographs were taken: anteroposterior centered on the glenoid, anteroposterior centered on the humerus in neutral rotation, lateral (Neer view), and axillary view.

Glenoid erosion was graded independently by 2 observers using the method described in this link. Erosion was labeled as none (grade 1), mild (grade 2; erosion into subchondral bone), moderate (grade 3; medialization of subchondral bone with hemispheric deformation), or severe (grade 4; complete deformation/destruction of the glenoid or hemispheric deformation until/beyond the base of the coracoid). With this method the inter-rater reliability was 0.76 (95% CI, 0.67-0.83).

The difference between the average of the two observers for the preoperative and for the postoperative erosion was calculated. Severe erosion was defined as erosion of grade ≥2.5.

An attempt was made to quantify glenoid erosion by drawing a vertical tangent to the lateral edge of the acromion and measure the distance from that line to the most medial point of the prosthetic
head. However, these measurements proved to be unreliable and were excluded from further analysis.

These authors found severe erosion in approximately one-third of their cases within a mean postoperative time of 2.5 years.

Predisposing factors for erosion were
(a) glenoid cysts (odds ratio, 5.4; P < .001, approximately 3 times more frequent in women), 
(b) fatty infiltration of the rotator cuff musculature (R, 0.43; P < .001), and 
(c) rheumatoid arthritis (odds ratio, 3.6; P = .049).

A valgus position of the prosthetic head relative to the glenoid (angle >50°) may have been associated with local destruction of the glenoid cartilage.

Age, the version of the glenoid, and the size of the prosthetic head showed no significant association with glenoid erosion.

Analysis of kinetics in cases of severe erosion showed 2 basic erosion patterns: continuous erosion over time (11 cases) and severe erosion occurring rapidly after implantation (13 cases).

12 patients had revision surgery

Comment: This study combined hemiarthroplasties for glenohumeral arthritis (in which case the glenoid can be assumed to be abnormal) with hemiarthroplasties for acute humeral fractures (in which case the glenoid can be assumed to be normal). Thus, as expected, in the fracture cohort, there was significantly less erosion.

It would have been of interest to see the correlation between the amount of glenoid erosion and the patient self-assessed shoulder comfort and function: did patients with more erosion have worse clinical outcomes?