Posterolateral approach for posterior malleolus fixation in ankle fractures: functional and radiological outcome based on Bartonicek classification

Ankle fractures are a relatively common type of limb fracture, while PMFs approximately account for 7–44% of all ankle fractures [1]. However, treatment of PMF remains nowadays a controversial topic. It was once widely accepted that the PM fracture requires further surgical fixation when the size of the posterior marginal fragment is more than 25% of the articular surface accompanied by displacement [2, 3]. Recently, several studies demonstrated that the presence of a PMF regardless of size has a negative influence on the outcome and should be treated properly [4, 5]. Drijfhout et al. found that medium and large sized fragments and postoperative step-off of the PM (≥ 1 mm) might increase the incidence of radiographic osteoarthritis [6]. On the other hand, anatomical reduction and fixation of PM fractures could provide better syndesmotic stability and articular surface congruity, which are associated with short-term clinical outcomes [7, 8].

Direct open reduction and fixation of posterior malleolar fragments from posterior to anterior could provide a better biomechanical stability and a more accurate reduction. Subsequently, the posterolateral approach has been advocated for the treatment of PFM and good results were demonstrated in several researches of the treating PMF by posterior approaches [9‐11].

In recent years, it has become recognized that the size of the posterior malleolar fragment is not the only interfering factor of prognosis. Literatures have shifted the focus towards fracture morphology rather than fragment size [12]. The utilization of CT scan makes it possible to assess the exact shape of the posterior malleolar fragments, leading to a better understanding of the anatomical pattern of PMF. Several classification systems have been developed [13‐15]. The first classification system, based on 3-dimensional (3D) CT reconstruction, was proposed by Bartonicek and Rammelt in 2015 [14]. PMFs were classified as 4 basic types with constant reference to the involvement of the fibular notch. On the basis of these characterizations of the PMF, the treatment recommendations have been proposed [16, 17].

In this study, the functional and radiological results were reviewed in 72 patients with PM ankle fracture treated using a posterolateral approach with screws and/or buttress plates. We further compared the outcomes of patients with different types of Bartonicek classification. We also found superior ankle functional recovery in Bartonicek subtype IIa PMF after surgery, to our knowledge this study was the first to analyze differences in clinical outcome according to Bartonicek type II sub-classification.

This retrospective study was conducted at a level one trauma center. Approval for the study was obtained from our Institutional Review Board.

Basic demographic data, fracture and operative characteristics are reported in Table 1. In 42 out of 72 patients (58.3%), the trauma also caused an ankle fracture-

dislocation. Additional syndesmosis screws (26 cases) or endobutton system (4 cases) was used in 41.7% of the patients after fracture fixation.

Posterior malleolus fractures refer to ankle fractures of the posterior articular surface of the distal tibia, which is believed to participate in the symmetrical coordination of the joint [21] and displaced PMFs often lead to a worse prognosis [22‐25]. In recent years, increased attention has been paid to the importance of anatomical reduction and internal fixation of the posterior malleolus [16].

The posterolateral approach is one of the most commonly used surgical approaches for the treatment of PMFs and several literatures reported good clinical results and high patient satisfaction after surgery [26‐29]. It is especially helpful with relatively small fragments and in the case of additional fragments that can’t be indirectly reduced [30, 31]. In this study, we confirmed that the posterolateral approach has the advantages of adequate exposure and reduction, and precise fixation. 94.5% of the patients reported a good or excellent reduction and no failure of fixation were observed. Besides, this approach could be a useful technique for PMF with satisfactory functional recovery and imperceptible pain, based on the AOFAS and VAS scores at the final follow-up.

The posterior malleolus is an important attachment for the posterior inferior tibiofibular ligament (PITFL). Baumbach et al. found that ORIF of posterior malleolus fragments, independent of their size, significantly reduced the frequency of trans-syndesmotic fixation compared to CRIF or untreated PMF in patients with trimalleolar ankle fractures [32]. Another retrospective review by Mengnai et al. revealed that direct posterior lateral plate fixation of PMFs, in supination external rotation type IV (SER IV) ankle fractures, could effectively stabilize the syndesmosis as well [33].Recently, Caglar et al. suggested that patients may not need additional trans-syndesmotic screw fixation after anatomic reduction of the PM and posterior inferior tibiofibular ligament complex [34]. However, in our study, additional syndesmosis screws or endobutton system was used in 41.7% patients. This phenomenon may be partly attributed to the fact that we used more posterior lag-screws (65.3%) than buttress plate (34.7%) in our PMF fixation. Moreover, traditional estimates of posterior stability may not be adequate to determine the need for posterior malleolar fixation [35]. Finally, low anatomic reduction rate (66.7%) of the posterior malleolus was also another interfering factor.

Posttraumatic osteoarthritis is seen clinically or radiographically in about one-third of patients with a PMF [36]. Several clinical studies identified osteoarthritis as a potential risk factor related to poor long-term prognosis. In a clinical retrospective study,the development of early-onset osteoarthritis contributes to functional limitation in young adults with ankle fractures involving the PM [37]. Verhage et al. found correction of intra-articular step-off of posterior malleolar fragments reduced the risk of developing osteoarthritis and, consequently, worse functional outcome after long-term follow-up [38]. Our present study demonstrated that 65 patients (88.9%) got no or just doubtable post-traumatic osteoarthritis according to Kellgren and Lawrence classification[19], which is likely to benefit from accurate PMF reduction, and subsequently resulted in satisfactory functional recovery at final follow-up.

Bartonicek et al. proposed 4 typical types of PMFs on the basis of an analysis of the 3-dimensional CT scans of 141 patients, representing a scale of increased injury severity and ankle instability [14]. The value of Bartonicek classification system is in its guidance of surgical treatment. Multiple studies have reported clinical results of ankle fractures including PFM in the light of Bartonicek classification. When evaluating 46 patients with PMF conservatively treated, clinical outcomes including OMAS/AOFAS scores worsened with the severity of Bartonicek classification, but early post-traumatic arthritis were not correlated to the injury pattern [39]. Michal et al. demonstrated good mid-term results of type IV Bartonicek classification PMFs after performing direct reduction and fixation via posterolateral or posteromedial approaches [40]. In contrast, Maarten et al. compared the functional outcome and general health status between patients who underwent plate osteosynthesis or not. They suggested that there is no indication for routine plate osteosynthesis of all types of posterior malleolus fractures [41].

To our knowledge, this is the first trial exclusively analyzing the clinical results of patients with PMFs treated by posterolateral approach with respect to Bartonicek classification. In our study, the severity of Bartonicek classification led to lower mean AOFAS scores both at 6 month and at the final follow-up. Bartonicek type II fractures had the highest AOFAS score. However, VAS scores were not statistically different between the various groups at different time points. We also revealed that post-traumatic osteoarthritis was not affected by Bartonicek classification.

Anatomical reduction of the PM fragment to restore the articular surface and the normal anatomy of the inferior tibiofibular joint has been advocated [14]. Until recently, only a few authors had studied the effect of fracture pattern on reduction of the PMFs. Stephen found that more anatomical reduction was achieved as the size of the PM fragment increased [42]. Michal et al. first assessed the reduction accuracy of Bartonicek type IV PMFs with the use of postoperative CT scans. Reduction of the PMF was assessed as anatomical in 14 cases (73.7%) and as satisfactory in five (26.3%) cases [40]. In our study, 58.3% Bartonicek type IV PMFs achieved excellent reduction quality, and 33.3% cases got good reduction. We also observed that there was no significant difference in reduction quality among three Bartonicek types, indicating that fracture morphology might not be the only factor affecting reduction of posterior malleolus.

Bartonicek type II PMF refers to an intraincisural posterolateral fragment involving 1/4–1/3 of the fibular notch and accounting for at least half of all posterior ankle fractures [13‐15]. No study has so far focused on subtypes of Type II PMFs according to the Bartonicek classification. In this study, we retrospectively analyzed the fracture and surgical characteristics, clinical results among different subtypes of Bartonicek type II fractures. First, we found that subtype IIa fractures presented a higher incidence (68.5%) of ankle dislocation. Moreover, a higher proportion of subtype IIa PMFs requires additional syndesmosis fixation, compared with subtype IIb and IIc groups. These discrepancies were not consistent with traditional findings and might be attributed to injury mechanism or sample shortage. While we did not reveal significant differences in reduction quality among different subtypes of type II PMFs, we found that, for osteoarthritis, the results of type IIa were better than those of type IIb and IIc. We attributed these variable results to the fact that, compared with type IIa PMFs, type IIb or IIc fractures usually involved larger size of the articular surface extending to the medial posterior rim or depressed intercalary joint fragments [16]. Finally, the difference in AOFAS among three groups at 6 months was not of statistical significance and at the final follow-up was small. Multiple comparisons found no difference of the AOFAS scores between the various groups. We also noted that the level of pain was not significantly different either. These findings reflect the fact that subtypes of Bartonicek type II PMF may just represent a scale of anatomical increased discrepancy fragments, and have no clear association with functional results.

There are several inherent weaknesses in this study. First, it was a retrospective data collection without patients treated conservatively in our cohort. Second, surgery was conducted by different surgeons varying in their experience. It should be noted that the final outcomes might be influenced by their selection of internal fixation and judgment of syndesmotic fixation necessity. Furthermore, our subtype identification mainly was on the basis of coronal CT scan without sagittal features. Finally, despite an innovative attempt to analyze the differences in prognosis between different subtypes of Bartonicek type II PMFs, our patient cohort was relatively small and the cases in each group varied greatly. All of these may affect the reliability of our clinical data analysis.

This study demonstrates that 72 patients treated for ankle fractures involving PMFs, with different Bartonicek classification, achieved good clinical outcomes through the posterolateral approach. We confirmed the relationship between the severity of PM involvement in ankle fractures on the basis of Bartonicek classification and the worsening of clinical outcomes measured by the AOFAS score. Compared with type IIb and IIc fractures, patients with Bartonicek type IIa fractures got a slightly higher AOFAS score and a remarkably lower incidence of osteoarthritis at the final follow-up. Classification of posterior malleolus fractures according to the Bartonicek classification was reliable. We would recommend all ankle fractures involving the posterior malleolus undergo preoperative CT scanning to guide treatment planning.

Direct reduction and fixation from the posterolateral approach was an effective option in Bartonicek type II, III and IV fractures. Bartonicek subtype IIb, IIc PMFs should pay more attention to achieve satisfactory clinical prognosis.