A Novel Classification System for Fifth Metatarsal Base Fractures Is Reliable and Clinically Validated

One goal of classifying fractures at the base of the fifth metatarsal is to reliably identify which fracture patterns have an increased risk of delayed union and nonunion and therefore merit consideration for operative fixation. The most common classification system of proximal fifth metatarsal fractures is the three-part Lawrence and Botte scheme, which has been shown to have poor interobserver and intraobserver reliability and can be ambiguous in terms of clinical treatment and prognosis. In particular, the assignment of the middle classification (Type II) has unclear implications for treatment and prognosis.

We developed a two-part classification system based on prior anatomic and biomechanical studies that we performed, and we asked: (1) What is the interobserver and intraobserver reliability of the system? (2) What association does it have with the risk of delayed union or nonunion?

This novel classification system was based on previous work on the effect of the peroneus brevis insertion on fracture stability. The fractures were divided into two categories, avulsion or indirect, based on their location with regard to the peroneus brevis footprint and with regard to the mechanism of injury. Fractures were considered avulsion if they occurred as the result of a direct avulsive force as ascertained by whether they were located within the peroneus brevis footprint. Fractures were considered indirect if they occurred as a result of a bending moment applied to the fifth metatarsal and were located outside of the peroneus brevis footprint. In that previous study, we used 10 cadaver specimens, and we simulated two loading conditions on the peroneus brevis tendon: (1) fractures within the peroneus brevis insertion (avulsion) were found to be stable and (2) fractures distal to the peroneus brevis footprint (indirect) were found to be unstable. The mean length of the peroneus brevis footprint on those specimens was 15 mm. This corresponded regardless of foot size to anatomic and radiographic features of the base of the fifth metatarsal. The most distal lateral extent of the avulsion fracture was 15 mm from the tuberosity. The most distal medial extent was proximal to the point where the curve of the articular surface of the base of the fifth metatarsal changes direction (inflection point). The most proximal lateral extent of the indirect fracture was distal to 15 mm from the tuberosity. The most proximal medial extent was distal to the inflection point of the curve of the base of the fifth metatarsal. To evaluate the reliability of this classification system, 30 sets of initial presentation radiographs composed of 10 fractures previously clinically diagnosed as representing each of the three zones in the Lawrence and Botte system were used. These radiographs represented a diverse pool of fractures chosen without knowledge of the patients' clinical outcome and without knowledge of the care provided. The 30 sets of radiographs were evaluated and classified by nine orthopaedic surgeons and three orthopaedic foot and ankle fellows to determine interobserver reliability for the proposed two-part system. Four weeks later, the same set of radiographs was reevaluated in random order to determine intraobserver reliability for the two-part classification system. Kappa and intraclass correlation values for interobserver and intraobserver reliability were calculated. For clinical validation, 23 of 30 patients were followed to complete healing as determined by symptomatology and radiography. The groups were too small for us to perform statistics, so we evaluated outcomes qualitatively.

The novel two-part classification of proximal fifth metatarsal fractures demonstrated high interobserver reliability of κ = 0.92 and intraobserver reliability of κ = 0.85 to 1.00. All four indirect fractures treated with closed management went on to nonunion. Two of the remaining 11 indirect fractures, all treated with internal fixation, experienced healing complications, including one nonunion and one refracture, possibly as a result of incomplete healing. Seven of eight avulsion fractures treated with internal fixation healed primarily. One displaced avulsion fracture was treated with excision.

This novel two-part radiographic classification system of proximal fifth metatarsal fractures based on the anatomy and biomechanical effect on fracture stability of the peroneus brevis tendon insertion demonstrated high interobserver and intraobserver reliability. It substantially outperformed the Lawrence and Botte three-part scheme as well as other recently studied two-part classification schemes. This two-part classification system distinguishes fractures resulting from avulsion or indirect injury, providing surgeons with guidance in determining treatment and prognosis. It therefore warrants consideration for use in both clinical and research applications.

Level III, diagnostic study.