Midfoot injuries

Last evidence check March 2011

Navicular fractures may be traumatic or stress fractures. Minor low-energy navicular fractures are relatively common and require little treatment. Some, however, are complex, associated with high energy trauma and difficult to reconstruct or stabilise. Neuropathic fractures can occur, most commonly in diabetics and often as part of a more extensive midfoot (Brodsky type 1) or peritalar (Brodsky type 2) neuroarthropathy.

Traumatic injuries

The commonest traumatic injuries are minor avulsions, usually of the dorsal lip, which represent midtarsal sprains and can be treated functionally. Occasionally the history of the injury or the amount of pain and swelling suggest that the midtarsal joint injury is more severe and 2-3 weeks splintage may be valuable before starting mobilisation.

Avulsions of the navicular tuberosity, usually by the tibialis posterior tendon, may be due to quite significant valgus injuries. An associated impaction or crush fracture of the cuboid suggests a self-reduced midtarsal subluxation or dislocation which would require splintage and possibly even examination of stability under anaesthetic and/or additional imaging by CT.

The most severe injuries are fractures of the body of the navicular. These are principally caused by axial compression forces, with additional valgus or varus. Sangeorzan et al classified these as follows:


Fracture configuration

Foot alignment


Coronal, transverse



Coronal, dorsolateral to plantarmedial





Simple fracture patterns with two or three fragments may be amenable to ORIF with lag screws. Comminuted fractures will need bridging techniques using external fixators, plates or screws into adjacent bones. It is important to minimise periosteal stripping to reduce the risk of osteonecrosis. DiGiovanni's (2004) review article in Foot and Ankle Clinics gives an excellent account from a meticulous surgeon, with many helpful technique tips.

Stress fractures

Stress fractures occur in high-performance athletes, especially runners and basketball players. They occur in the central third of the body. this may be because the central third becomes progressively less vascular with age (Torg 1982), or because it is a region of shear forces between axial forces generated by the first and second metatarsals (Fitch 1989). Torg found that 80% of his patients had metatarsus adductus.

Stress fractures may be merely minor cracks in one cortex, or may spread into the body or to the other cortex. Saxena (2006) found that the time from onset of symptoms to diagnosis was longer in the more extensive fractures, suggesting that the fracture propagates over 6-9 months. Most, but not all, are undisplaced at presentation. 26% of the clinical fractures in Potter's (2006) series were bilateral, but another quarter had healed, asymptomatic stress fractures on CT.

The usual presence is aching pain in a performance athlete, especially in sprinting activities. Tenderness at the "N-spot" (the dorsomedial prominence of the bone) is said to be diagnostic; however this has been questioned in the context of operatively treated fractures.

This is one of the few foot and ankle injuries where non-weightbearing splintage is necessary for healing. In Torg's (1982) series, all ten fractures treated in NWB casts healed, whereas 7/9 patients treated with weightbearing and various degrees of splintage failed to heal or refractured. Khan et al (1992) obtained successful healing and return to sports in 86% of athletes with undisplaced fractures treated in NWB casts compared with 20-69% with various weightbearing regimes. Potter (2006) reported a mean AOFAS midfoot pain score of 51/60 and a mean function score of 25/30 in athletes treated in NWB casts, similar to the results of surgical fixation.

Displaced fractures are uncommon and most authors recommend surgical treatment. The place of surgery in undisplaced fractures is unclear. There are no RCTs comparing surgery with NWB casting - not surprising in a rare fracture. Potter (2006) in a retrospective comparison found no difference in functional results between surgery and casting, but more of the operated patients had persistent navicular tenderness; this was a small study with considerable potential confounding factors. Khan (1992) reported that healing was quicker in surgically treated fractures (3.8m versus 5.6) but the overall success rate was the same. Saxena (2006) found no difference in healing rate or return to sport in a group in which fractures more extensive than single-cortex cracks were fixed, compared with historical controls.

It appears that the first-line treatment of undisplaced navicular stress fractures should be 6-8 weeks NWB in a cast, and this was the conclusion of a recent systematic review (Torg 2010). Where this fails to produce healing surgery should be considered.