The present database analysis revealed three major findings: First, the majority of conservatively treated isolated GT fractures (68.6%) are nondisplaced, generally justifying conservative treatment. Second, the rate of secondary interventions was 16.2%. Finally, most of the secondary surgical interventions were not related to the secondary displacement of the GT fracture but to concomitant lesions of the rotator cuff and the capsuloligamentous tissue.
According to the criteria originally defined by Neer, a displacement of >10 mm and 45° is an indication for surgery, and all other fractures can be successfully managed with a conservative approach . Later, these criteria were modified to 5 mm and 30° of displacement. However, it is known that as little as 2–5 mm of displacement can cause impingement and requires increased forces for abduction [7,8,9]. In particular, fragment displacement in the posterosuperior direction is associated with impaired function and worse results [2, 27]. Therefore, operative treatment is recommended more aggressively, and numerous articles exist about different techniques and results [1, 3, 5, 6, 13, 17, 23, 26].
By contrast, there are only a few reports on conservative treatment of isolated GT fractures. Platzer et al. reported on the functional results of 135 patients treated conservatively for isolated GT fractures with less than 6 mm displacement . They found good to excellent results in 97% of the cases. This is in accordance with results from other studies of conservative treatment [12, 17, 20, 23, 24]. If the displacement is less than 5 mm, satisfying results can be expected. In addition, the present study shows that patients with a nondisplaced or only slightly displaced fracture (0–3 mm) that did not arise from a shoulder dislocation have a low risk for secondary surgical interventions. Furthermore, fractures with an initial displacement of 3 mm or less are unlikely to have further displacement over time (only 2% in the present study). Patients who required secondary surgery after initial displacement had a primary fracture displacement of >3 mm.
Unfortunately, defining the degree of displacement has also been a matter of debate. When displacement is measured with only plain radiographs, errors of up to 13 mm have been described . A computed tomography (CT) scan may help to minimize these errors. On the other hand, Janssen et al. observed that the imaging modality did not influence the reliability of the fracture assessment or the recommendation for surgical treatment . Mutch et al. suggested using a greater tuberosity ratio (GT ratio) that can be applied to plain radiographs . They found a very strong correlation with computed tomography (CT) scans for superior GT displacement. Furthermore, the GT ratio helped to accurately identify fractures as suitable for conservative or operative treatment or as benefitting from further imaging.
In addition, the amount of initial displacement is relevant to the decision on the treatment modality. However, there is an immediate need to reevaluate patients treated conservatively, since 50–60% of fractures show further displacement over time . Younger patients are at an especially heightened risk. Hebert-Davies et al. found a 5.6-fold higher risk for secondary displacement in patients younger than 70 years compared with patients over 70 years of age . Similarly, in our study, patients with secondary displacement and surgical intervention were younger than the mean age of the cohort (46 vs. 55 years).
Therefore, both aspects, the degree of displacement and how to adequately assess it, must be further investigated.
Another major finding of the present analysis is that concomitant soft tissue lesions lead to a secondary intervention after initial conservative treatment in over three quarters of the cases. In the present analysis, surgical intervention owing to secondary displacement was only performed in four cases (3.8%). Other common interventions include capsuloligamentous and rotator cuff repairs (Table 1). These findings highlight the need for further imaging, particularly MRI scans, to detect any concomitant lesions. Especially in patients with anterior shoulder dislocation and multi-fragmentary GT fracture, concomitant lesions are frequently found and require further operative treatment. Maman et al. reported on 24 arthroscopically treated patients with a GT fracture. Concomitant soft tissue lesions were found in 22 patients (94%) . These findings are supported by Katthagen et al., who found concomitant lesions (i. e., pulley/SLAP and Bankart lesions) in 69% of patients who were treated arthroscopically for a GT fracture . Again, these lesions were found more frequently after shoulder dislocations.
Muhm et al. found concomitant lesions in GT fractures with and without a dislocation . However, in patients with a dislocation, concomitant lesions were more likely to be treated operatively. Interestingly, in the Muhm study, GT fractures with three or more fragments were always associated with anterior shoulder dislocation. In the present analysis, complex fracture patterns were found even in patients without previous shoulder dislocation, although patients were more likely to have a multi-fragmentary fracture when they sustained a shoulder dislocation. The risk for secondary surgical interventions increased with dislocations but not with multi-fragmentary fracture patterns.
Some inherent limitations apply to the present analysis. Only patient records were analyzed, and the final functional and radiographic outcomes remain unclear in most cases. Therefore, we cannot provide proof of whether or not conservative treatment leads to good results in patients without secondary interventions. Furthermore, the decision to apply conservative treatment was not based on a distinct algorithm. There is a clear trend, however, toward conservative treatment in patients with only minimally displaced fractures. In most cases, patients with severely displaced fractures were treated conservatively when there were contraindications for surgery or when patients refused to undergo surgical treatment.
Finally, imaging modalities were not consistent in all the cases since not every patient received a CT and/or MRI scan before the decision to apply conservative treatment was made.