In the present study, investigating patients who underwent surgery due to a malleolar ankle fracture, a high incidence (25.8%) of AITFL avulsion fractures was observed in comparison with the previous literature. Park et al.  reported an incidence of 12.4%. Selection bias could be a possible explanation for this difference in incidence. Were Park and al. only used the standard X-rays to identify AITFL avulsion fractures, this study included only patients with a pre- and postoperative CT-scan, making it easier to identify avulsed fragments. Furthermore, this study reported on patients who were treated in a level-1 trauma center, where generally more “complex” ankle fractures are treated.
Just like in the other studies, no Wagstaffe type 1 avulsion fractures were observed [19, 20]. Historically, only Wagstaffe himself reported the type 1 fracture, an isolated avulsion fracture of the anterior tubercle of the distal fibula . Interestingly, he suspected the existence of this type of avulsion fracture after only physical examination in a time where no X-rays were available. The existence of this fracture type of the distal fibula is questionable, considering the absence of reliable proof and no additional described cases in literature.
On the other hand, we evaluated only operative treated ankles, as CT-scans are mostly not performed in conservative-treated ankles. Wagstaffe type 1 fractures might be underreported, since they could easily be missed on a conventional X-ray.
Park et al. only observed AITFL avulsion fractures in patients with a Weber B type ankle fracture. In contrast, this study also had a proportion of patients with Weber C fractures in the AITFL avulsion fracture group (30.6%). This is the first study which shows avulsion fractures of the AITFL which can occur with the foot in both supinated and pronated position during external rotational forces.
In the previous series, the majority of cases with an AITFL avulsion fracture concerned Wagstaffe type 2 fractures. In this study, we observed a more equal distribution between Wagstaffe type 2 (43.1%) and type 3 fractures (49.2%). Like other studies, a correlation was reported between Wagstaffe type 2 avulsion fractures and Weber B type ankle fractures. However, this study also correlated the Wagstaffe type 3 (Tillaux–Chaput) fracture with Weber C type ankle fractures. With the numbers available, a significant correlation, which was found in this study, could be indicative of a hypothesis that there is more tension on the insertion of the AITFL on the Wagstaffe tubercle with the foot in supinated position, and on the Chaput tubercle during pronated position.
It seems that fragment size has an influence on the way of fixation of the avulsed fragment. Recently, Diallo et al.  presented a case series in which they described screw fixation of ten avulsion fractures, sized 5 mm or larger. This corresponds with the results of this study, in which we observed a significant correlation in fragment fixation (direct fixation) of fragments of 5 mm or larger. This study suggests that the size of an avulsed fragment determines the way of fixation.
Haraguchi et al. described that 35% of the conservatively treated avulsion fractures did not heal, without mentioning fragment size . The correlation between size of the avulsed fragments and type of fixation has not been mentioned before in the literature. Park et al. fixated all the fragments with non-absorbable sutures through anchoring holes made by Kirschner wires. It is conceivable that this technique has its limitations in smaller fragments. In the series of Chung et al.  in 2012, most of the avulsed fragments were fixated with mini screws or Kirschner wires, while the others were repaired with a ligament suture. A variety of types of fixation of the avulsed fractures were mentioned in this study. The fixation with a single screw was most common.
Due to the fact that direct fixation of an avulsed fragment appeared to be dependent on the size of the fragment, we would like to propose a new AITFL avulsion fracture classification system (Fig. 3). We adjusted the classification for this variable into the subtypes: (a) < than 5 mm in diameter and (b) ≥ 5 mm in diameter. Avulsion fractures of subtype b should be taken into consideration for fixation. Furthermore, we added an additional type of avulsion fracture: the isolated avulsion fracture of the anterolateral distal tubercle of the tibia (isolated Tillaux–Chaput fracture). This type of fracture has not been frequently reported in the previous study and was observed twice in this study.
Due a low number of revisions (n = 4) in the AITFL avulsion fracture group, no significance could be observed between the groups with non-direct fixated fragment versus direct fixated fragments. However, all four complications did occur in the non-fixated group. In the three cases which underwent revision surgery, a syndesmotic screw was placed during the initial surgical procedure, which should have stabilized the ankle. It is questionable whether direct fixation of the fragment would have prevented a revision procedure. Other variables, like smoking or osteoporosis, could have affected the healing of the bone as well [29, 30]. In patients with secondary dislocation premature mobilization could not be ruled out. Anterior impingement syndrome is a complication which was also seen in one case in the study by Park et al. .
This study only reports on the incidence of AITFL avulsion fractures in patients who had an indication for surgery for their ankle fracture and underwent a CT-scan. Due to the retrospective nature of this study, the optimal treatment of AITFL avulsion fractures accompanied by ankle fractures can only be suggested.
Finally, this study does not incorporate the results of functional outcome, which could have given a clue into the direction whether it is important to fixate the fragment. However, in retrospective studies, it is not possible to include surveys such as Olerud–Molander Ankle Scores (OMAS) , unless they were part of the standardized postoperative procedure at fixed times.
In conclusion, the current study reports a CT-scan incidence of 25.8% of AITFL avulsion fracture in surgically treated ankle fractures. Significant correlations with Weber B/C type ankle fractures and Wagstaffe type 2/type 3 fractures were found. Most fragments larger than 5 mm were fixated. Based on the size of the avulsed fragments and possibilities of fixation, a new AITFL avulsion fracture classification system has been proposed.