Identification and treatment of atypical acetabular fractures having an independent acetabular roof fragment: Three-column fractures

Introduction: A unique type of both-column fracture of the acetabulum that also has an independent roof fragment seems to occur relatively often. It is challenging to diagnose, thus our ability to detect it and its frequency are not known. This led us to analyze a database of acetabular fractures to 1) determine the incidence of this type of fracture, 2) describe its radiological characteristics, 3) attempt to set out a specific treatment strategy.

Hypothesis: The need for a dual surgical approach depends on the type and displacement of the independent roof fragment.

Methods: Four surgeons and radiologists independently analyzed a set of acetabular fractures that occurred between 2007 and 2017. The diagnosis was made using two-dimensional (2D) and three-dimensional (3D) CT reconstructions. Once the fractures had been identified, a detailed description was made of all 2D and 3D slices. A retrospective analysis was done of the reduction and fixation of the independent roof fragment relative to the chosen surgical approach.

Results: The study comprised 534 acetabular fractures, of which 96 where both-column fractures. In that subset, 41% (39 fractures) had an independent roof fragment. A detailed analysis resulted in the identification of two subtypes of three-column fracture with independent fragment: type 1 has a posterosuperior fragment (22 cases); type 2 has an independent fragment separated by a juxtatectal fracture line (17 cases). The best radiological outcomes were achieved when the independent fragment was minimally displaced and did not require additional reduction.

Conclusion: The isolated roof fragment is common, as it occurs is more than one-third of both-column fractures. We were able to identify two subtypes that require different surgical strategies. Surgical treatment of these fractures is difficult; thus a combined surgical approach (anterior followed by posterior) may provide the best reduction, especially for subtype 1.

Level of evidence: IV.