Novel values in the radiographic diagnosis of ligamentous Lisfranc injuries

Abstract

Background: Ligamentous Lisfranc instability is commonly missed on unilateral radiographs. However, measurement protocols for bilateral weightbearing radiographs have not been standardized. The primary aim of this study was to investigate the optimal cut-off values for diagnosing Lisfranc instability by evaluating the side-to-side differences of preoperative bilateral weightbearing radiographs among patients with surgically-confirmed ligamentous Lisfranc instability. A secondary aim was to investigate whether the midfoot measurements for detecting Lisfranc injury could also be used in patients with a pre-existing bilateral Hallux Valgus (HV) deformity by evaluating whether the Lisfranc measurements could be affected by a foot deformity as HV.

Patients and methods: Patients who underwent surgical repair of ligamentous Lisfranc instability, as well as a separate cohort with bilateral hallux valgus deformity, were included in this multicenter retrospective cohort study. A standardized radiographic measurement protocol was used to assess the midfoot and a receiver operator correlation (ROC) analysis was used to identify the optimal cut-off value for measurements. Interclass Correlation (ICC) scores were calculated to assess the interrater reliability of the Lisfranc area measurement.

Results: Forty-seven patients were included in the Lisfranc group with a mean age of 33 (± 15) years and 25 patients were included in the HV group with a mean age of 51 (± 15) years. For the Lisfranc group, measurements that demonstrated a significant side-to-side difference included; increased C1M2 diastasis of 2.4 mm (± 1.4, P<0.001), increased C1M2 surface area of 24 mm² (± 15, P<0.001), C2M2 malignment by 1.7 mm (± 1.2, P<0.001), second tarsometatarsal joint dorsal step-off sign by 0.8 mm (± 0.7, P<0.001), and arch height by 2.5 mm (± 6.4, P<0.048), all greater on the injured side. In the HV group, side-to-side measurements were not significantly different. There was no significant difference comparing the M1M2 measurement in the HV group with the injured (P = 0.16) or uninjured (P = 0.08) foot in the Lisfranc group. The optimal cut-off points were between the injured and uninjured foot in the Lisfranc group were 2.1 mm for C1M2 diastasis, 0.7 mm for the C2M2 alignment, and 30 mm² for the C1M2 surface area. The ICC-score for the second C1M2 area measurement was 0.88.

Conclusion: Bilateral foot weightbearing radiographs can effectively diagnose ligamentous Lisfranc instability using a standardized measurement protocol. Malalignment of the medial aspect of the second metatarsal base ≥0.3 mm relatively to the intermediate cuneiform offers a high sensitivity, and distance ≥2.1 mm between the second metatarsal base and the medial cuneiform has a high specificity. Intermetatarsal distance between the first and second metatarsal base has a low sensitivity and specificity and should not be used in solitary for diagnosis.

Level of evidence: Level III, retrospective comparative study.

Keywords: Diagnosis; Lisfranc; Radiography; Tarsometatarsal; Weight-bearing.