Background: The orientation and the position of glenoid are two factors which have been pointed out as a cause of notch in case of reverse shoulder arthroplasty (RSA). Our hypothesis is that the notch is mostly depending on the relationship between the humerus and the glenoid.

Material and method: The inclusion criteria of this prospective study were: RSA with a 2 years minimum follow up, pre and post op x-rays of good quality to allow accurate measurements, a minimum of 3 x-rays regularly separated in the first 2 years. On these x-rays, done at rest, we have analyzed: the vertical orientation of the glenoid, the degree of abduction of the humerus, the gleno-metaphyseal angle (GMA) which represents the relationship between glenoid and humerus. Sixty one shoulders in 60 patients (57 females, 7 males) have been included. The mean age was 74,6 y.o (56–82) and the mean body mass index (BMI) was 25,4 (16–36).

Results: The patients with a notch had a significant correlation with a lower BMI (p< 0,001), a more upward pre op orientation of glenoid (p< 0,01), a less downward post op orientation of the glenoid, a lower degree of abduction of the humerus (p< 0,01), a lower GMA (p< 0,001) that means either that the humerus is less abducted or that the glenoid is orientated more upward or both. In addition we found a strongly correlation between the BMI and the degree of abduction of humerus which is lower if the BMI is low (p< 0,001). We studied the evolution of the angles and we noted that there was no difference during the first 3 months. After 3 months, the GMA stayed stable in patients without notch but was decreasing during the first year in those with a notch.

Discussion: The occurrence of a notch is strongly increased in patients with a lower BMI because of a less degree of abduction of the humerus, mostly if the glenoid is orientated upward. So the relationship between humerus and glenoid is more important to analyse than position of glenoid alone. The BMI is another important factor to take into account in case of indication of RSA because the risk of occurrence of a notch is quite higher.

Purpose: The purpose of this study was to analyse outcome of shoulder prostheses after radiotherapy, to define a specific clinicoradiological entity, and evaluate incidence of complications.

Material and methods: Fourteen shoulder prostheses were implanted in 13 women who had been treated for breast cancer with complementary radiotherapy and one man treated for Hodgkin’s lymphoma. The time from radiothearpy to implantation was 16 years, seven months. Two forms were identified on the preoperative x-rays: seven cases with typical avascular osteonecrosis according to the Arlet and Ficat classification, and seven cases with a radiographic presentation of arthritis or degenerative disease. Humeral prostheses were used in five cases and a total shoulder arthroplasty in nine.

Results: Four implants had to be removed, three for sepsis, and five patients required revision surgery. The mean postoperative Constant score for the ten prostheses still in place was 53.1 points with a mean elevation of 111° at three years seven months follow-up. The gain in pain score was 8.5 points with a mean result of 10.9 points. The results were different depending on the initial radiological form, with less favourable outcome observed in typical osteonecrosis.

Discussion: This study demonstrated a particular radio-clinical entity independent of classical osteonecrosis of the humeral head. The surgical procedure was more difficult and the outcome was less satisfactory than in the classical forms with a high rate of complications.