Midfoot fractures: Patterns of injury and predictors of stability

Abstract

Background

Lisfranc injuries describe a spectrum of midfoot and tarsometatarsal joint (TMTJ) trauma ranging from purely ligamentous to multiple fracture-dislocations. Lisfranc injuries represent 0.2 % of all fractures and are seen predictably, with mechanisms involving a fall from height, crushing, or torsion. Diagnosis can be challenging, with approximately 20 % of cases being missed, and relies upon clinical acumen and proficient image interpretation. Whilst multiple classification systems have described Lisfranc injuries using a 3-column concept, these add zero prognostic value and are therefore rarely used clinically. Furthermore, existing literature on diagnosis and management is limited to retrospective small series.

Methods

We present a review of 161 midfoot injuries, with the aim of highlighting characteristics of radiological instability and indication for operative management. CT scans and weight-bearing and non-weight-bearing X-rays were reviewed for features of joint instability. These features included metatarsal base, cuneiform and cuboid fractures, tarsometatarsal joint subluxation or dislocation and C1-MT2 diastasis. The subsequent “stable” and “unstable” injury groups were then compared to identify statistically significant indicators for instability.

Results

Avulsion and intra-articular fractures of the medial, middle, or lateral column were all suggestive of instability. Although these appeared in multiple combinations, 95 % involved the middle column. Concomitant inter-cuneiform and cuboid fractures were additional indicators of instability. In cases of uncertain midfoot instability, weight-bearing radiographs were of value with 14.2 % demonstrating a diastasis of C1-MT2 >2 mm.

Conclusion

We propose the need for a new classification of midfoot injuries which emphasises the diagnosis of instability and guides surgical management. We propose that, based on non-weight-bearing X-ray and CT scans, these injuries can be initially classified as “stable”, “unstable”, or “stability uncertain”. Weight-bearing X-rays are a safe and reliable method of detecting instability in the “stability uncertain” group. Fractures of the medial column and cuneiform on initial imaging were suggestive of midfoot instability.

1. Introduction

Midfoot fractures, involving the tarsometatarsal articulations, are reported to have an incidence of 9.2/100,000 person-years.¹ These injuries can occur following a variety of mechanisms, ranging from high-energy direct trauma (e.g. fall from height or crush injuries) to low-energy indirect trauma (axial force through the foot, twisting on a plantarflexed foot and forced foot external rotation) commonly seen in sports injuries.²

The tarsometatarsal joint complex consists of the articulations between the three cuneiform bones (C1-C3), five metatarsal bones (M1-M5), and the cuboid bone.³ The joint complex is supported by dorsal, plantar, and intra-osseous ligaments, including the eponymous Lisfranc ligament running from the base of the second metatarsal to the medial cuneiform.⁴ Of these three ligaments, the plantar and intra-osseous ligaments are stronger.^5,6 The stability of the joint complex is provided by a combination of osseous and ligamentous anatomy with the “keystone” effect of the recessed second tarsometatarsal joint and Lisfranc ligament being of particular importance.⁷

Injury to the midfoot can involve a combination of bone and ligamentous injuries with varying involvement of the tarsometatarsal joint complex itself.⁸ The more severe high-energy injuries (e.g. fracture-dislocations) are often clearly demonstrated on plain film radiographs.⁴ However, many, often more low-energy, injuries are subtler on radiographic imaging and are subsequently often missed.⁴ Diagnosis in the latter group therefore relies on a high index of clinical suspicion, cross-sectional, and weight-bearing imaging. Lau et al., estimates that approximately 20 % of Lisfranc injuries are missed and this can lead to morbidity including progressive midfoot instability, arch collapse, and post-traumatic osteoarthritis.⁹ In light of this, diagnostic accuracy and the early recognition of injuries with midfoot instability are vital to optimise patient outcomes.^2,10

Multiple classifications have been proposed to aid diagnosis of Lisfranc injuries.^4,8 Examples include Quenu and Kuss, later modified by Hardcastle et al. and then by Myerson et al., and the later classification by Nunley and Vertillo.⁴ These classification systems are based on static radiographs and focus on describing the injury in terms of the patterns of displacement between different injured segments dividing injuries into partial (medial or lateral displacement) or complete (homolateral or divergent displacement) injuries.⁴ As a result, these classification systems are primarily descriptive and offer little prognostic value, so provide limited guidance to treatment and injury outcomes.⁸

This study aims to describe the patterns of injury seen in midfoot fractures based on static radiographs, cross-sectional imaging and weight-bearing radiographs and to describe a structured approach to identifying fractures with midfoot instability and thus those injuries with a worse prognosis and likely to require surgical stabilisation. In addition, we aim to identify whether any injury patterns are more predictive of instability of the midfoot and therefore more likely to require surgical stabilisation.

2. Methods

The study was registered with our trust governance team (CARMS-18435), as a retrospective data review. Ethical approval was not required following our trust protocol. We performed an informatics search of all CT reports containing the phrases “midfoot injury or fracture”, “Lisfranc injury or fracture”, “tarsometatarsal dislocation or subluxation or fracture” over 5 years from 2015 to 2020.

Patients included in the study were those with midfoot fractures adjacent to or involving the tarsometatarsal joints with complete radiographs and follow-up until discharge. We excluded those below the age of 16, with non-traumatic aetiology or neuropathy, and those whose imaging records were not complete. All imaging was then reviewed by a consultant orthopaedic surgeon fellowship trained in trauma and foot and ankle surgery. Initial static radiographs were reviewed for evidence of fractures, subluxation (joint displacement with residual joint contact), dislocation (complete loss of joint contact), and diastasis between the medial cuneiform (C1) and base of second metatarsal (MT2) on the anteroposterior radiograph. C1-MT2 diastasis was measured by a Picture Archiving & Communications System (PACS) tool and defined as >2 mm of widening. CT scans were reviewed for evidence of fractures of the metatarsal bases (extra-articular, intra-articular or avulsion type), fractures involving the cuneiform bones, and fractures of the cuboid. The foot was divided into columns consisting of medial (C1 and M1), middle (C2-3 and M2-3), and lateral columns (Cuboid and M4-5).^3,4 In addition, evidence of tarsometatarsal joint subluxation, dislocation, or C1-MT2 diastasis on CT was also recorded. Weight-bearing radiographs were performed when midfoot instability was uncertain following initial imaging. These were then reviewed for evidence of C1-MT2 diastasis suggestive of instability under physiological load or dorsal subluxation of the metatarsal bases. Injuries were considered unstable if any features of joint dislocation, joint subluxation or C1-MT2 diastasis were identified and stable if none of these features were present.

The unstable and stable group of injuries were then compared using the chi-square statistical test. Basic summary statistics were calculated in Microsoft Excel. Statistical analysis beyond this was performed using R 4.3.1 (Ref) in the RStudio IDE 2023.09.0 (Ref). The threshold for statistical significance was set at a type 1 error rate of 0.05.

3. Results

We identified 161 patients with midfoot fractures on CT imaging during the study period. Demographic data, fracture pattern and mechanism of injury are outlined in Table 1.

Table 1.

Demographics, fracture pattern and mechanism of injury of 161 patients with midfoot fractures.

Demographics
Total number of patients	161
Total number of female patients	76
Total number of male patients	85
Mean age of patients	41.2 years (Range: 16–78)
Fracture pattern
	Number of patients
Medial column involvement	107
Middle column involvement	156
Lateral column involvement	93
All three-column involvement	64
Inter-cuneiform involvement	45
Cuboid involvement	24
Mechanism of injury
	Number of patients
Fall from standing (<3 m)	38
Fall from height (>3 cm)	39
Road Traffic Collision	19
Twist injury	49
Crush injury	13
Seizure	2

Based on the initial static plain radiographs and CT scans, 18 fractures were classified as stable (extra-articular fractures) and 68 were classified as unstable due to either tarsometatarsal joint subluxation/dislocation or the presence of C1-MT2 diastasis. Fig. 1 demonstrates an unstable midfoot injury identified on initial non-weight-bearing x-ray. It was not possible to determine the stability of the remaining 75 fractures. These fractures involved the radiological footprint of the tarsometatarsal ligaments but the joints themselves remained anatomically aligned with no subluxation or diastasis. This group was classified as “stability uncertain”.

Fig. 1.

Anteroposterior and lateral non-weight-bearing X-ray of an unstable midfoot injury.

Of the 75 “stability uncertain” fractures, 70 underwent weight-bearing imaging at subsequent review. Of these, a further 10 patients (14.2 %) were found to be unstable due to dorsal subluxation of the tarsometatarsal joints or diastasis of C1-MT2 under physiological load. Fig. 2 demonstrates a patient from this case series with an unstable midfoot injury confirmed on weight-bearing x-ray.

Fig. 2.

Radiographic imaging demonstrating a midfoot injury that had uncertain stability on initial non-weight-bearing X-ray and CT scan, but later confirmed to be unstable by a weight-bearing x-ray. Scan a) is an anteroposterior non-weightbearing x-ray. Scans b) and c) are non-weight-bearing CT scans and Scan d) is an anteroposterior weight-bearing x-ray.

Thus, from the 161 identified midfoot injuries, based on static radiographs, CT scans, and targeted weight-bearing imaging, 83 were diagnosed as stable and 78 were deemed unstable. Fig. 3 summarises the above results.

Fig. 3.

Algorithm of investigation and diagnosis of midfoot fractures according to stability.

The characteristics of the unstable fracture group are outlined in Fig. 4. Of the 78 unstable fractures, 47 underwent surgical stabilisation with the remainder either not suitable for surgery owing to surgical risk factors or declining surgery. Surgical stabilisation procedures included 12 open reduction and internal fixations (ORIFs) with staples, 4 ORIFs with screws, 27 ORIFs with bridging plates, and 2 ORIFs with K wires with the type of surgical stabilisation determined by fracture pattern, soft tissue injury and surgeon preference. 2 patients underwent fusions.

Fig. 4.

Characteristics of unstable Lisfranc Injuries identified using static radiographs, CT imaging and weight-bearing x-rays.

In the stable group, 1 patient underwent surgical stabilisation despite no demonstrated midfoot instability, the remaining patients were managed non-operatively. There were no documented late complications requiring surgical intervention in this group.

A comparison between unstable and stable group injury patterns was then made. The results are displayed in Table 2. From these results, we can establish that all 3-column involvement, medial column involvement, cuneiform fracture, and male gender are all significant risk factors for unstable midfoot injuries.

Table 2.

Comparison between unstable and stable Lisfranc injuries. High energy mechanism of injury has been defined to include fall from height and crush injuries, whilst low energy mechanisms of injury include fall from standing, axial force through the foot, twisting on a plantarflexed foot and forced foot external rotation commonly seen in sports injuries.

	Unstable injury group		Stable injury group		Statistical significance
Sex	30 Females	48 Males	47 Females	38 Males	P = 0.0315
Energy mechanism	36 High energy mechanism	42 Low energy mechanism	33 High energy mechanism	52 Low energy mechanism	P = 0.34
All 3 columns involved	38/78		26/85		P = 0.0179
Medial column involved	63/78		48/85		P = 0.0009
Middle column involved	75/78		83/85		P = 0.58
Lateral column involved	47/78		47/85		P = 0.52
Inter-cuneiform involvement	38/78		7/85		P = 0.0001
Cuboid involvement	14/48		10/56		P = 0.17

4. Discussion

We present one of the largest reported series of midfoot fractures, identified by CT imaging. Based on non-weight-bearing imaging and CT scans we preliminarily classified these injuries into 3 groups: stable (18/161), unstable (68/161) and stability uncertain (75/161). Further weight-bearing imaging in the “stability uncertain” group allowed a final classification into stable (83/161) and unstable (78/161) groups. Injury patterns associated with instability were fractures involving all 3 columns and fractures involving the medial column or cuneiforms. In our series, no stable fracture managed non-operatively developed a late complication.

There has historically been a high incidence of missed Lisfranc injuries based on standard plain film imaging and identifying unstable midfoot injuries that require surgical stabilisation can be challenging. Seow et al. performed a meta-analysis of the radiographic criteria for diagnosing Lisfranc injuries based on x-ray.¹¹ Whilst noting the significant heterogeneity of diagnostic criteria reported, they proposed that C1-MT2 diastasis of >2 mm on anteroposterior views or C2-MT2 subluxation to be the most useful criteria for determining stability. They additionally advocate that further imaging should be considered in cases of high clinical suspicion.¹¹ De Bruijn et al. compared non-weight-bearing and weight-bearing X-rays in patients with subtle Lisfranc injuries.¹² They noted that a larger diastasis was observed on weight-bearing films with a greater interobserver reliability for diagnosis of midfoot instability.¹² Furthermore, in their systematic review of the radiology of Lisfranc injuries, Sripanich et al. concluded that weight bearing X-rays increased the ability to diagnose subtle Lisfranc instability and that CT scans could aid the detection of undisplaced fractures and subtle osseous displacement whereas the role of MRI in detecting instability was unclear.¹³ It is possible that as weight bearing CT scans become more available this will become the investigation of choice for early diagnosis of midfoot injuries with instability, with initial results suggesting this is more sensitive than conventional CT scanning.¹⁴ There has been some suggestion regarding the usefulness of MRI imaging in diagnosing Lisfranc injuries. Whilst MRI scans can be beneficial in identifying purely ligamentous injuries, they are less helpful in identifying Lisfranc or midfoot trauma involving bone fractures. As a result, due to the limited availability of weight-bearing CT scans in all institutions and limited benefit of MRI imaging we consider the use of CT scans with targeted weight bearing x-rays to be the most appropriate method of diagnosing midfoot instability.^15,16

It is recognised that displaced Lisfranc injuries require fixation to achieve the best outcome and therefore displaced or unstable injuries are likely to require surgical stabilisation.¹⁷ In their paper, Myeson et al. identified the quality of reduction as the main determinant of outcome after a midfoot injury and observed a worse outcome in those with a displacement of over 2 mm.¹⁷ There remains ongoing debate as to the optimum method of midfoot stabilisation with options including internal fixation with trans-articular screws, bridge plates, and midfoot fusion. A recent meta-analysis of four observational studies suggested better functional outcomes with bridge plating versus trans-articular screws, whereas meta-analysis comparing primary fusion with fixation of midfoot injuries found no difference in clinical outcome between the groups but a higher rate of metalwork removal in the fixation group.^18,19 Stødle et al. reported on 84 Lisfranc injuries, of which 40 % were unstable.²⁰ Here they found that predictors of instability included female sex, intra-articular fractures of the lateral 2 rays, and a shortened second tarsometatarsal joint height.²⁰ In a later study, Stødle et al. reported on 26 patients with undisplaced Lisfranc injuries, as determined by CT stress testing and weight-bearing x-rays.²¹ They found that excellent outcomes were achieved with non-operative management, with a median follow-up of 55 months and a medial AOFAS midfoot score of 100.^17,21

Our paper is a retrospective series with the inherent limitations of such a study. Patients were identified from CT scan reports, and it is likely that some injuries, in patients who did not have CT scans, were missed. We have not included purely ligamentous injuries of the tarsometatarsal joints in our study as the investigation and treatment of such injuries differ somewhat from midfoot fractures. We have not recorded outcome measures for our patients as the focus of this study was on describing radiological patterns of injury and factors associated with midfoot instability. Whilst we recognise that treatment protocols between units may vary, we suggest that our proposed algorithm of initial non-weight-bearing X-ray (soon after initial injury when meaningful weight-bearing is unlikely to be possible) and cross-sectional imaging in the form of CT scan followed then by selective weight-bearing imaging in a patient with uncertain midfoot stability, reflects the contemporary practice and is validated by our study results.

5. Conclusion

We present a large study of midfoot fractures. We propose that these injuries can be initially classified based on non-weight-bearing X-ray and CT scans into three groups: “stable”, “unstable”, and “stability uncertain”. Weight-bearing X-rays are a safe and reliable method of detecting instability in the “stability uncertain” group of injuries. In our series, involvement of all 3 columns of the midfoot, fractures involving the medial column, and fractures of the cuneiforms on initial imaging were suggestive of midfoot instability.

Author contribution

Dr Serena Patel – Writing – Original Draft (Lead), Writing – Review & Editing (Equal).

Ms Danielle Piper – Conceptualisation (supporting), Data Curation, Investigation, Methodology, Supervision (supporting).

Mr Paul Fenton – Conceptualisation (lead), Data Curation, Investigation, Methodology, Supervision (lead), Writing – Review & Editing (Equal).

Ethical approval and patient consent

Ethical approval and patient consent was not required for this case series under local University Hospitals Birmingham NHS Foundation Trust policy.

Source of funding

We the authors wish to declare that no additional funding was received from a commercial or third party towards the development of this article.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.