*3.2. Radiologic Outcome*

In 22 patients (88%), RLs around the glenoid component were found. In 21 (84%) patients, the humeral component showed RLs. The overall mean Lazarus grade for the glenoid component was 2.1 points (range: 0–4; SD: 1.1; *p* < 0.001; effect size: 2.0). There were only two patients with an RL thicker than 2 mm. In detail, there were three patients with a Lazarus grade of 0 (12%), three patients with a grade of 1 (12%), eight patients with a grade of 2 (32%), ten patients with a grade of 3 (40%), one patient with a grade of 4 (4%) and none with a grade of 5. The overall mean points given by the classification according to Molé et al. [13] were 3.4 (range: 0–7; SD: 2.17; *p* < 0.001; effect size: 1.6). Most patients had a score of 3, followed by a score of 0 and 6, each with the same number of frequencies. No patient had a score of 1, and only one had a score of 7.

Upward migration of the humeral head was observed in all of the patients. The mean difference between the HHM value of the initial postoperative radiograph and the latest follow-up was 6.4 mm (range: 0.5–13.4; SD: 3.9; *p* < 0.001; effect size: 1.6). In 7 study participants, the humeral head migrated more than 10 mm. On the other hand, only two patients had an HHM of less than 1 mm.

In 23 out of 25 cases, polyethylene wear was detected after a mean follow-up of 62.3 months. For LGHO, the mean difference between the first postoperative radiograph and the last follow-up was 2.6 mm (range: 0–4.0; SD: 1.5; *p* < 0.001; effect size: 1.7). In most patients (9), the inlay wore between 2 and 4 mm. On the other hand, only four patients had an LGHO difference of 1 to 2 mm.

#### *3.3. Complications and Revisions*

Seventeen patients (8 women) were revised, mostly because of polyethylene wear. In patients undergoing revision surgery for any reason, the mean age at implantation of the anatomic prosthesis was 63.3 years (range: 45–80; SD: 10.5), and at revision, the average age was 68.8 years (range: 51–84; SD: 9.6). Among our study group, the probability of prosthesis survival was 32% (17 revisions) after a mean follow-up of 68.2 months (range: 1.8–119.6; SD: 27.9). In 12 cases (71%), PE wear was the most prevalent reason for revision surgery. Three patients had RTC injuries, and one patient had glenoid loosening as the cause of the revision. Only one patient developed a wound infection after surgery, resulting in the need for revision. Figure 3 shows the overall implant survival curve of our study. There, after around 75 months, the median has been reached. Afterward, the revision cases occurred more frequently in less amount of time. Censored were all patients on their last follow-up time who did not undergo revision surgery. In the graph, it can be seen that the first revision occurred quite early, after 1.8 months. The last revision occurred after 119.6 months. The graph shows that the first half of the revisions took 3/4 of the total time span. In contrast, most of the revisions were done after the midpoint of the timeline.

**Figure 3.** Kaplan–Meier plot depicting survival of the aTSA with MBG from implant revision for any reason among the study population.

Altogether, thirteen patients, seven of whom were female, were converted to RSA. Out of the 13 revision cases, there were different reasons for conversion to an RSA. With 85% (11 cases), the most common indication for revision was polyethylene wear. The remaining 2 cases had secondary RTC tears as reasons. In all of them, an explantation of the Eclipse™ implant and switch to Arthrex Reverse TSA was performed. The mean time to revision for conversion to RSA was 80.7 months (range: 40.5–152.6; SD: 30.9), whereas the mean age of the patients at revision was 71.2 years (range: 61–84; SD: 7.9). Figure 4 shows the Kaplan–Meier survival curve from conversion to RSA. The survival rate free of revision for conversion to RSA was 48% at 80.7 months. After approximately 85 months, half of the patients go<sup>t</sup> conversion surgery to RSA. The median of revisions came in later than in the first Kaplan–Meier curve. The first conversion to RSA happened after 40.5 months and the last much later at 152.6 months, as seen in the Kaplan–Meier curve (Figure 4).

The time interval between the first and last revision is described in Figure 4 as considerably higher. Figure 5 shows a radiograph taken right before a revision for conversion to RSA, highlighting both PE wear and humeral implant loosening with varus-tilting.

**Figure 4.** Kaplan–Meier plot depicting survival of the aTSA with MBG from implant revision for conversion to RSA among the study population.

**Figure 5.** The image on the left side shows an X-ray taken shortly before a revision operation for conversion to RSA, demonstrating inlay abrasion with a tilt of the humeral head towards the metalback as a direct sign of inlay abrasion and loosening of the humeral component with varus-tilt due to osteolysis and PE wear. The image on the right side presents an intraoperative situs of the inlay after removal.
