Named after Jaques Lisfranc, an eighteenth century surgeon who performed the first foot amputations at the tarsometatarsal (TMT) joint, the Lisfranc joint is an anatomic area where a broad spectrum of injuries from subtle distensions to open fracture dislocations occur [
1,
2]. The incidence of Lisfranc injuries is estimated to be 1/55000/year and they are believed to account for 0.2% of all fractures [
3,
4]. These figures have, however, been challenged as up to 24% of Lisfranc injuries are either misdiagnosed or overlooked during initial evaluation [
5‐
7]. Injuries to the Lisfranc joint occur most often during the third decade of life and men are 2 to 4 times more likely to suffer from these injuries than women [
8]. Lisfranc injuries are caused either by direct or indirect forces to the foot [
9]. Indirect injuries are more common and occur during bending or twisting movements applied to the midfoot [
9]. Injuries caused by direct forces are often induced by a heavy object falling on top of the foot or by crush injuries, such as in motor vehicle accidents [
6,
7]. A wide spectrum of injuries to the TMT and interrelated joints have been recognized, and range from severely dislocated high-energy open injuries to minor midfoot sprains suffered during sports activities [
10‐
12].
An untreated or inadequately treated Lisfranc injury results in multiple late complications, the severity of which depends on the severity of the primary injury [
13]. The most common complications are painful instability of the joint, malformation and arthritis [
5]. All these complications can lead to remarkable dysfunction and foot pain [
5]. Secondary arthrodesis may be used to treat these injuries, but the outcome is poorer the longer the treatment is delayed [
14‐
16]. Therefore, the initial recognition of these injuries is a crucial step in ensuring optimal treatment is provided.
Diagnosis and treatment
Fractures of the Lisfranc joint are known to be rare and are often overlooked [
7,
17‐
19]. Approximately 20 to 24% of these fractures are missed at initial evaluation [
5,
7]. High-energy injuries are often the most obvious due to traumatic history and very apparent clinical findings [
20]. Low-energy injuries, however, are harder to detect because of less traumatic history and less apparent clinical findings [
21]. Typical clinical findings of fracture of the Lisfranc joint are a swollen midfoot, tenderness and pain in the midfoot during passive movements and weight-bearing [
22], plantar ecchymosis [
23] and an extended space between the first and second toe seen in x-ray radiographs that is also known as the ‘gap’ sign [
24].
Although sensitivity is relatively low when compared with CT-imaging, primary diagnosis of Lisfranc injuries is usually based on plain x-ray imaging [
7]. False-negative findings on x-ray radiographs may be the result of weight-bearing not tolerated due to pain [
6]. A typical finding ‘fleck sign’ in plain x-ray radiographs, an avulsion of intra-articular bone, is estimated to be detectable in 90% of cases where the dislocation between the first and second metatarsal is greater than 4 mm [
5]. As the radiographic findings of Lisfranc injuries can be subtle, CT is an important imaging modality in detecting these injuries, and furthermore serves as a useful tool for preoperative planning [
25,
26]. Although the current literature introduces classifications that provide general characteristics for Lisfranc injuries, none of the classifications are useful in predicting treatment or outcome of a Lisfranc injury [
27]. Moreover, the current literature fails to offer a classification based on computed tomography.
Due to the diversity of injuries, there is no single evidence-based policy for treating all Lisfranc injuries in a similar manner [
28]. Nowadays, there is strong consensus that in dislocated injuries it is crucial to achieve exact anatomic reduction and stable internal fixation, which is best obtained with open reduction and screw fixation (ORIF) [
5,
29]. However, even after appropriate treatment with ORIF, up to 40 to 94% of patients will develop post-traumatic arthritis [
5,
13,
30,
31], necessitating conversion to an arthrodesis to relieve pain [
14‐
16]. To prevent the need for secondary operations and the development of post-traumatic arthritis, primary arthrodesis (PA) is suggested [
30,
32‐
34]. The treatment of non-dislocated injuries, in turn, is controversial [
29,
35‐
38]. Some stable injuries might need activity modification only, but surgery is often recommended for even minimally displaced injuries [
5,
29]. There is general agreement, however, that poor functional results are commonly correlated with a delay in diagnosis or the inadequate treatment of unstable or dislocated injuries [
19,
27].
Fixation with screws is the primary fixation technique used to treat dislocated Lisfranc injuries [
13,
31]. K-wire fixation [
5,
12,
39,
40] and screw fixation [
13,
22,
41,
42] are both controversial, but the higher failure rates associated with K-wire fixation have led to an increase in screw fixation [
13,
43,
44]. Another fixation technique, dorsal plate fixation, has been reported to produce similar results as ORIF [
45]. An advantage of dorsal plate fixation is that the plate causes no damage to the articular surface. However, soft-tissue irritation may be more prevalent, and second surgery is often needed to remove the plates [
45].
There is, however, no general agreement on what is the correct nonoperative protocol for treating non-dislocated Lisfranc injuries. In their review, Myerson and Cerrato [
11] concluded that if the foot remains stable in weight-bearing radiographs 2 weeks after the injury, the injury can be treated with immobilization in a boot and weight-bearing is permitted as tolerated until the boot is removed at six to eight weeks. In the study by Nunley & Vertullo [
29], stable injuries were treated nonoperatively. Furthermore, it was suggested that treatment begin with a non-weight-bearing cast for 6 weeks. If the patient is painless at 6 weeks, treatment should continue with a gradual return to normal function with a weight-bearing orthosis for the following 4 weeks.
The commonly used postoperative protocol is nearly identical to nonoperative treatment. In their study, Ly & Coetzee [
30] used a short leg splint for 2 weeks followed by a short leg cast for four to six weeks. The patients advanced to full weight-bearing during the following 4 weeks while wearing a prefabricated fracture boot. In the study by Henning et al. [
33], weight-bearing began at three months with a controlled ankle motion walker.
Interestingly, there are several opinions about postoperative implant removal. Some studies suggest routine screw removal at 8 or 12 weeks [
31,
33,
46,
47], while others prefer routine removal only after the recovery is complete or only if the screws cause irritation or pain [
48‐
50]. Ahmad and Jones [
51] have suggested the use of bioabsorbable screws to remove the need for screw removal. In addition, bioabsorbable screws achieve similar functional results compared with metal screws.
Evaluation of treatment
Most of the previous studies have used Patient Reported Outcome Measures (PROMs) to evaluate treatment. The most common PROM used in Lisfranc injury studies is the American Orthopaedic Foot and Ankle Society Midfoot Score (AOFAS) [
13,
28,
30,
38,
47]. Other commonly used PROMs include Visual-Analogue-Scale Foot and Ankle (VAS-FA) [
52] (also validated in the Finnish language [
53]), Visual-Analogue-Scale (VAS) [
28], Short Form 36 (SF-36) [
28,
33], Baltimore Painful Foot Score (PFS) [
31], Short Musculoskeletal Function Assessment (SMFA) [
33], long-form Musculoskeletal Function Assessment (MFA) [
13], the Maryland foot score [
47] and activities of daily living (ADL) [
47]. In our study, we decided to use AOFAS because it is the most commonly used PROM for Lisfranc injuries and VAS-FA as it is validated in the Finnish language [
53].
Previous studies
The literature does not provide any prospective randomized controlled studies on the nonoperative versus the operative treatment of Lisfranc injuries. Current knowledge is based on a few case-series [
35,
37] and retrospective studies [
5,
22,
38]. Nunley and Vertullo [
29] suggested in their series of midfoot sprains in athletes that only totally non-dislocated sprain injuries should be treated nonoperatively, and that all injuries where the diastasis between the first and second metatarsal is 2 mm or more would benefit from ORIF. Myerson et al. [
5] were the first to study the nonoperative treatment of Lisfranc injury. In their study, only 5 out of a total of 52 patients were treated nonoperatively, and these patients received the treatment unintentionally, due to incorrect diagnosis. Of these five patients, four resulted in a poor result and one resulted in a fair result. Curtis et al. [
22] organized a retrospective study of the treatment of 19 athletes with Lisfranc injuries. Only 14 stable injuries were treated nonoperatively. An excellent functional result was obtained with six patients, a good result with three patients, a fair result in four and a poor result with one patient. An excellent result implied the absence of symptoms and signs; a good result implied minor symptoms or signs; a fair result implied residual signs of symptoms with some disability, and a poor result implied marked symptoms or signs with limitation of function and a request for further treatment, such as arthrodesis. The treatment protocol between patients differed from “none” to “cast for ten weeks”. Crates et al. [
38] studied nonoperative treatment and operative treatment after the failed nonoperative treatment of subtle Lisfranc injuries in 36 patients. The nonoperative protocol consisted of 6 weeks of a short leg walking orthosis and weight-bearing was progressed as tolerated. Progressed weight-bearing in an orthotic was begun after boot removal. Nonoperative treatment was successful in 16 patients, and the treatment failed in 20 patients. The mean AOFAS midfoot score in the successfully treated patients was 62 (49–72) before treatment and 75 (53–100) after treatment.
There have only been two previous prospective randomized studies on ORIF vs PA. Ly and Coetzee [
30] randomly assigned 41 patients with ligamentous Lisfranc injuries to either an ORIF group or a PA group. The PA group had a slightly better functional outcome (AOFAS score 88 vs. 69), a higher return to preinjury activity level (92% vs. 65%), a lower rate of revision surgery and less pain in the final follow-up. Implant removal due to prominent or painful screws was performed on 16 of the 20 patients in the ORIF group and on 4 of the 21 patients in the PA group. The implant removal was only performed due to painful hardware, on average at 6.5 months (range: from five to ten months). Follow-up radiographs showed loss of correction, increasing deformity, and degenerative joint disease in 15 of the 20 patients in the ORIF group and 7 of them required conversion to an arthrodesis. In the PA group, one patient had delayed union at seventeen weeks and one patient required a revision arthrodesis with bone graft. One patient suffered from a post-traumatic intrinsic compartment syndrome that resulted in claw toes. In the study by Henning et al. [
33], 40 patients with acute Lisfranc joint fractures or fracture dislocations were randomized to primary ORIF or PA. A total of 8 patients dropped out before 3-months follow-up. There was a significantly higher rate of secondary surgery in the ORIF group. Statistically significant differences were not found in physical functioning with regard to SF-36 or SMFA scores at any follow-up time interval. In their systematic review and meta-analysis, Smith et al. [
34] concluded that ORIF has a higher risk of implant removal compared with PA (risk ratio 0.23 (0.11–0.45)
p < 0.001), although there were no statistically significant differences in revision surgery, PROMs or non-anatomic alignment. Cochran et al. [
32] organized a retrospective comparative cohort study on PA versus ORIF in young athletic military personnel with low-energy Lisfranc injury. In their study, PA resulted in a faster return to military service, a lower implant removal rate and better fitness scores after 1 year.
In conclusion, PA seems to result in less secondary surgery, less implant removal and a faster return to activity. There is some evidence of a better functional outcome after arthrodesis, but the result is still controversial. Nevertheless, the current overall evidence slightly favors arthrodesis as a primary treatment of dislocated Lisfranc injuries.