Rethinking the Lisfranc injury: Rationale for and development of a new classification system for Lisfranc injuries

Abstract

The Lisfranc joint comprises the articulations between the cuneiforms and bases of the metatarsals and is susceptible to a variety of complex injuries which are frequently missed or diagnosed late. This can result in morbidity with poor outcomes, and renders it challenging to standardise management and allow prognostication. Existing classification systems, including the Myerson-Hardcastle (MHC) and Nunley & Vertullo classifications, do not consider the propagation of force through the midfoot, which is key to identifying structures which may require stabilization. They also demonstrate poor clinical utility due to poor correlation with outcomes. This study aimed to develop a new classification system for Lisfranc injuries using data from a large single-surgeon case series from a tertiary level trauma hospital in Australia. The primary outcome measures for this study were inter- and intra-observer reliability. The Suthersan classification was developed to address shortfalls in current classification systems. Patient records were manually searched for Lisfranc injuries. Direct 'crush' Lisfranc injuries were excluded due to unpredictable injury patterns. Indirect Lisfranc injuries were included in analysis. Two orthopaedics foot and ankle surgeons and one hand surgeon reviewed patient radiographs and classified them using the MHC and Suthersan classification. Inter- and intra-observer reliability for both classification systems was assessed using Cohen's Kappa statistic. 176 indirect Lisfranc injuries with accessible CT scans were identified. Mean patient age was 41.6 y and 45.5% were female sex. Suthersan Grade 1a injuries were most common and grade 4 injuries were least common. 19 CT scans were not classifiable using the MHC. Combined inter-rater reliability was 0.95 (Z = 42.3) for Suthersan classification and 0.65 (Z = 21.7) for MHC. Mean intra-rater agreement was 97.7% for Suthersan classification and 87.9% for MHC. The Suthersan classification demonstrates excellent inter- and intra-rater agreement compared to the MHC. By considering propagation of force, the Suthersan classification may allow for prognostication and the development of management algorithms following prospective clinical observational studies correlating categories with outcomes.