Clinical and Radiological Outcomes after Open Reduction and Internal Fixation of Lisfranc Injuries: A Single Centre Experience

Abstract

Background

Lisfranc injuries are uncommon and can be challenging to manage. There is considerable variation in opinion regarding the mode of operative treatment of these injuries, with some studies preferring primary arthrodesis over traditional open reduction and internal fixation (ORIF). We aim to assess the clinical and radiological outcomes of the patients treated with ORIF in our unit.

Materials and Methods

This is a retrospective study, in which all 27 consecutive patients treated with ORIF between June 2013 and October 2018 by one surgeon were included with an average followup of 2.4 years. All patients underwent ORIF with joint-sparing surgery by a dorsal bridging plate (DBP) for the second and third tarsometatarsal (TMT) joint, and the first TMT joint was fixed with transarticular screws. Patients had clinical examination and radiological assessment and completed American Orthopaedic Foot and Ankle Society (AOFAS) midfoot score and Foot Function Index (FFI) questionnaires.

Results

Our early results of 22 patients (5 lost to followup) showed that 16 (72%) patients were pain free, walking normally without aids, and wearing normal shoes and 68% were able to run or play sports. The mean AOFAS midfoot score was 78.1 (63–100) and the average FFI was 19.5 (0.6–34). Radiological assessment confirmed that only three patients had progression to posttraumatic arthritis at the TMT joints though only one of these was clinically symptomatic.

Conclusion

Good clinical and radiological outcomes can be achieved by ORIF in lisfranc injuries with joint-sparing surgery using DBP.

Introduction

Lisfranc fractures are rare injuries [1], with reported incidence of 0.2% of all fractures and 1/55,000 per year incidence in the population. These are reported to occur 2–4 times more commonly in men, and the peak incidence is in the third decade of life [2]. Lisfranc injuries result from disruption of the lisfranc ligament which runs from the lateral aspect of the medial cuneiform to the medial aspect of the base of the second metatarsal on the plantar aspect. It is the strongest ligament of the foot, and injury to this results in disruption of the tarsometatarsal (TMT) articulation.

The spectrum of these injuries varies from low-energy sports injuries to high-energy crush injuries [3]. There is a wide variation in these injuries with pure ligamentous injuries to comminuted forefoot and midfoot fracture dislocations [3]. These injuries are caused by axial loading of hyperplantarflexed foot [4]. These can also be caused by high-velocity trauma. These are increasingly reported as being caused by twisting injuries or minor trips and falls.

Up to 24% of these injuries are missed [5] or frequently diagnosed late. Untreated, delayed treatment, or inadequately treated injuries result in poor functional outcome for the patient [6] in proportion to the severity of the initial injury [7]. This results in substantial pain, midfoot arthritis, pes planus deformity, decreased function, and loss of quality of life [4]. Secondary arthrodesis may be used to treat these injuries, but the outcome is poorer, the longer the treatment is delayed [8–10]. Prompt diagnosis and early treatment is, therefore, essential in ensuring good outcome in these injuries.

Early diagnosis with proper clinical examination and weight-bearing radiographs is important. The key to good functional outcome is anatomical reduction and stable internal fixation. There is good evidence to suggest that quality anatomical reduction of <2 mm is associated with a significant improvement in both radiological and functional outcomes [11, 12].

There is a wide range of opinions regarding the treatment of these injuries. The treatment varies from closed reduction and immobilization, closed reduction and percutaneous K wire fixation, open reduction and internal fixation (ORIF) [13, 14] with transarticular screw (TAS) or extraarticular fixation with joint-sparing surgery by dorsal bridging plate (DBP). A number of studies have also shown good clinical outcomes with primary arthrodesis (PA) in these injuries [15, 16].

The aim of our study is to evaluate the clinical and radiological results of joint-sparing internal fixation of lisfranc injuries. The primary outcome measures were American Orthopaedic Foot and Ankle Society (AOFAS) and Foot Function Index (FFI) functional scores. The secondary outcome measures were rate of reoperation, complication rates, and osteoarthritis (OA).

Materials and Methods

A total of 27 consecutive patients, 15 females and 12 males, who underwent ORIF for lisfranc injuries between June 2013 and October 2018 were included in the study. The age at the time of injury varied from 19 to 72 years, with the average age at the time of injury being 43.3 years. In our series, the median age at injury was 45.5 years.

Of the 27 patients, 2 had other injuries in addition to the lisfranc injury (anterior dislocation of shoulder in one and undisplaced ipsilateral fibular fracture in the other). The remaining 25 patients had an isolated lisfranc injury.

Mechanisms of injury in each of the 27 cases are highlighted in Table 1. Road Traffic Accident (RTA) featured as a cause in only two cases.

Table 1.

Patient demographics and mode of injury (n = 27)

Baseline characteristics	n (%)
Median age at injury (years)	45.5 (19-72)
Mean age at injury (years)	42.9
Gender
Male	12 (44.4)
Female	15 (55.6)
High-energy trauma
Yes	2 (7.4)
No	25 (92.6)
Mechanism of trauma
Motor vehicle	1 (3.7)
Fall >3 m	5 (13)
Fall <3 m	4 (10)
Sports	3 (8)
Motorcycle	1 (3.7)

We used the Myerson Classification for lisfranc injuries. Majority of our patients had Type B (partial lateral or total lateral) injury. Nineteen patients had B2, 6 had C2, and 2 had type A injuries [Table 2].

Table 2.

Results of American Orthopaedic Foot and Ankle Society and Foot Function Index scores with Myerson classification

Myerson classification	Number of patients	Number of columns fixed	Foot Function Index	Average AOFAS score
Type A (total incongruity)
A (medial)	0
A (lateral)	2	5		71.62
Type B (partial incongruity)
B1 (medial)	0
B2 (partial lateral)	13	1		84.21
B2 (total lateral)	6	3		74.36
Type C (divergent)
C1 (partial)	0
C2 (total)	6	5		71.18
Total	27		19.5 (n = 22)	78.11 (n = 22)

AOFAS = American Orthopaedic Foot and Ankle Society

All our patients had anteroposterior and lateral weight-bearing radiographs and computed tomography (CT) scans before surgery. A clear assessment was made of the nature and extent of the injury; extent of comminution; and involvement of the first, second, and lateral rays.

We note that, in our series, a large proportion (13/27) of patients had purely ligamentous injury involving the second TMT joint and hence the high rate of missed or late diagnosis (8 [61.5%] out of 13). These are subtle injuries and difficult to diagnose and have a high propensity to be missed in emergency department (ED).

All patients had ORIF in the next available trauma list by the senior author. The surgery was only delayed in two patients due to severe swelling.

A universal dorsal approach was used, centered between the second and third metatarsal, providing access to the second and third TMT joints Figure 1. Separate dorsal incision was used for the 4/5 TMT joint and a medial incision for the first TMT joint. Adequate care was taken to protect the branches of the superficial and deep peroneal nerve and the dorsalis pedis artery.

Fig. 1.

(a) Per-operative photographs demonstrating the findings at the second tarsometatarsal joint. (b) Dorsal bridging plate applied to the second tarsometatarsal joint

The second TMT joint was reduced first as this was considered the key for the reduction and then the medial TMT and lateral TMT joints were reduced. Initially, K wires were inserted to hold the reduction and then DBP was applied. The remaining TMT joints were then screened for instability. Initially, K wires were inserted to hold the reduction and then DBP was applied Figure 2.

Fig. 2.

Case study with (a) preoperative radiographs and computed tomography scan and (b) operative images and (f-h) postoperative radiographs

DBP was the preferred mode of stabilization for the second and third TMT joints. The Synthes 2.4 mm/2.7 mm variable angle LCP forefoot/midfoot plate was used as the DBP for the second and third TMT joints. First TMT joint was stabilized with 4.0 mm cannulated screws or a plate (1 case). 4 and 5 TMT joints were stabilized by K wires, if needed. One patient had very poor skin and extensive swelling despite 2 weeks of elevation and was treated with open reduction and K wires fixation of all five TMT joints instead of screws or plates. Figure 2 (a)-(h) shows serial radiographs of one patient who required fixation of two columns. Figure 3 (a)-(d) shows serial radiographs of one patient who had purely ligamentous injury.

Fig. 3.

Case study showing purely ligamentous lisfranc injury. (a) Preoperative radiographs and (b) computed tomography scan and (c and d) postoperative radiographs

Postoperatively, all patients were mobilized nonweight bearing in a plaster boot for 6 weeks followed by progressive weight bearing for the next 6 weeks in an aircast boot. All patients were referred for physiotherapy at 6 weeks. Patients were followed up in the clinic at regular intervals with radiographs.

All patients had clinical and radiological assessment and also completed FFI and AOFAS midfoot score questionnaire. All patients were assessed for their pain, mobility, gait, ROM at the ankle and subtalar joints, footwear modification, and functional assessment as well radiological assessment by the first author.

Results

We noted that 8 (29.6%) of 27 patients were not diagnosed initially by the ED and presented more than 3 weeks after the initial injury. One patient was diagnosed 4 months after the initial injury, due to persistent pain. The mean time to surgery for all patients (n = 27) was 21.5 days after injury (range 6 h to 126 days). In the eight patients who had a delay in diagnosis, the average time to surgery after injury was 51.3 days. If the waiting time of these eight patients was excluded, the mean time to surgery for the remaining 19 patients was 8.8 days.

Twenty two patients attended their assessment and completed the AOFAS midfoot and FFI score. Five patients were lost to followup as they had either moved out of the area or were uncontactable or unable to attend.

The followup period varied from 6 months to 65 months, and the average length of followup after surgery was 29 months (2.4 years). Fifteen patients had more than 2 years of followup and 21 patients had at least 1 year of followup at the time of their last review.

The complications encountered are listed in Table 3.

Table 3.

Type of surgical fixation and complications

Treatment characteristics (n = 27)
Implant (number of columns fixed)
TAS for first TMT and plates for second and third TMT (3)	5
TAS for first TMT and plate for second TMT (2)	3
Plates for Second and Third TMT (2)	3
Plate for second TMT only (1)	13
All TMT involved - K wires only (skin poor) (5)	1
All TMT involved - TAS, plates, and K wires (5)	2
Complications (n = 6)
Superficial wound infection	2
Deep infection	0
Loss of reduction (early postoperator)	1
Delayed discharge	1
Implant failure	0
Compromised wound healing	1
Transient numbness	1
Second procedure (n = 5)
Secondary arthrodesis	1
Implant removal	3
TMT joint injection	1
Radiological evidence of posttraumatic OA
Total	3
Clinically symptomatic	1

TMT = Tarsometatarsal, TAS = Transarticular screw, OA = Osteoarthritis

Of 22 patients, 16 reported that they had no pain and 4 had mild intermittent pain. Two patients had moderate pain, of which one patient reported constant pain. No patient had incapacitating pain. Four patients were bothered by dorsal prominence from the metalwork.

Sixteen patients were wearing normal unadapted shoes and no insoles. Four patients had inability to wear normal shoes and two were wearing insoles in normal shoes. All patients had preservation of ankle and subtalar joint movements, and apart from one patient, none had any pain in these joints, though one of them experienced pain in these joints.

Nineteen patients were walking unaided, two used an elbow crutch, and one used a walking stick. One of the patients using the elbow crutch was recovering from the recent surgery (4 weeks prior) of the removal of metalwork. Of 22 patients, 16 had a normal gait pattern, 3 had antalgic but unaided gait, and 3 used walking aids.

All except two patients, who were working before their injury, had been able to return back to their preinjury occupation. Two patients were unable to work as a direct result of their injury. Majority of the working patients were engaged in standing or heavy occupations (14 out of 16).

Of 22 patients, 15 (68%) had been able to return back to sports or undertake running.

The average AOFAS midfoot score of our patients (n = 22) was 78.11. The median score was 81.5, ranging from 63 to 100. The average FFI was 19.5 and the median score was 17.3, ranging from 0.6 to 34. The average AOFAS score was also calculated for different injury patterns as shown in Table 2.

Radiographs were obtained at 2 weeks, 6 weeks, 3 months, and 6 months. The radiographs confirmed maintenance of reduction of TMT joints in all except one patient, in which the reduction of the first TMT joint was lost at 3 months, requiring secondary arthrodesis.

Three patients had radiological features of posttraumatic arthritis affecting the TMT joints with reduction in the joint space. The radiological findings did not translate clinically in two out of three patients. There were no cases of implant breakage or failure at the last followup.

Discussion

Open reduction and anatomic reduction with internal fixation remains the gold standard treatment, and good results are reported by a number of authors [4, 5, 11–14].

Our study also showed good results after ORIF with an average AOFAS score of 78.1 and 68% able to return back to sports or running activities and only 3 (11.5%) of 27 patients requiring metal work removal. Radiological features of secondary OA was seen in 3 (13.6%) of 22 patients, with only one out of these three patients being symptomatic with some midfoot pain.

Our results were comparable to other studies reporting functional outcomes after ORIF. Kuo et al. [4] in 2000 published a retrospective study of 48 patients followed up for an average of 52 months with the mean AOFAS score of 77. Similar results were also reported by Duboise- Dubois- Ferrière et al. [17] in 2016. They reported functional outcomes in 61 out of a total of 128 patients treated surgically with either ORIF or PA at a mean of 10.9 years postoperatively (range 2.4–23.9 years) with an average AOFAS score of 79.0 and FFI of 16.9. They found radiographic evidence of OA in 72.1% of the patients and symptomatic OA in 54.1%.

PA is also the favored option in some series, and a number of authors have shown good results with PA [15, 16]. A systematic review by Smith et al. in 2015 showed that there was no difference in patient-reported outcome scores or in achieving anatomical reduction between ORIF and PA [18]. They reported that there was decreased need for metalwork removal in PA as compared to ORIF. Our study showed that good results can be achieved with ORIF with minimal reoperation rates for metalwork removal or symptomatic secondary OA. Only three of our patients had the removal of metalwork because of dorsal prominence and discomfort between 12 and 20 months after surgery. Three patients had radiological evidence of progression to secondary OA, with only one being symptomatic clinically at the last followup.

There is no consensus regarding the method of fixation. TAS fixation of the midfoot has been the traditional preferred surgical technique up until the use of DBP. The drawback of the TAS is scuffing and damage of the articular surface caused by the screw. It is also associated with screw breakage and implant failure [19, 20]. There can be loss of position on screw removal and potential fracture after screw removal.

Cadaveric studies by Alberta et al. [20] and Gaines et al. [21] showed that using dorsal plating led to less articular surface damage and similar stability compared with TAS. Dekker et al. [22] also demonstrated in a cadaveric biomechanical study that there is a significant increase in articular cartilage injury from TAS fixation with cyclic loading while walking as a result of screw toggle. They showed that this articular damage from the TAS was worst in the second metatarsal and least in the first metatarsal.

We feel that there is a significant advantage of DBP in achieving anatomical reduction in cases with comminution of the metatarsal base where it can bridge fracture site, aiding anatomical reduction. It would be difficult to achieve the same degree of stability with TAS, especially in the presence of intraarticular comminution. In cases of ligamentous injury without fracture, the DBP can maintain the reduction without breaching or damaging the articular surface.

Hu et al. [23] in 2012 described a prospective study of 60 patients with an isolated lisfranc injury treated by dorsal plating or TAS. At 2 years of followup, the mean AOFAS score was approximately 83 and 78 for DBP and TAS groups, respectively. They also reported that chances of achieving anatomic reduction were higher in the DBP group (90.6% vs. 82.1%).

Kirzner et al. [24] in 2018 reported that DBP gave significantly better functional results than TAS or a combination in their series of 108 patients. The patients who had DBP had a mean AOFAS score of 82.5 points, compared with 71.0 for the screw group and 63.3 for the combination group.

In our study, the worse outcomes were noted for Myerson A and C injuries. Dorsal plate fixation alone was associated with much better outcome than a combination. The average AOFAS midfoot score was 84.2 in patients who had only one column involved and had DBP for the second TMT joint only. The average AOFAS score was 71 and 74 when 5 and 3 columns were involved, respectively. These patients also required a combination of TAS and DBP. We also conclude that the lower scores were as a result of the severity of the initial injury and multiple column involvement, rather than a correlation to the method of fixation. This, however, needs to be investigated further in future.

We feel that joint-sparing approach with DBP for the second and third TMT preserves the articular cartilage and avoids iatrogenic damage. It also avoids overcompression or failure to achieve anatomical reduction easily caused by using screws. It also avoids loss of position during screw insertion. The potential drawback of the variable angle LCP midfoot plate used as DBP is the additional exposure needed, requiring soft-tissue stripping and the profile of the implant itself, causing prominence on the dorsum of the foot. It can also cause stiffening of the TMT joint by preventing any movement at this joint.

We highlight that, in our series, there were no implant breakage or failure at the last followup (5.8 years to 6 months). We had to remove metalwork in three patients for dorsal prominence and ongoing discomfort. The healing of ligaments in purely ligamentous injuries is difficult to assess and predict. We also feel that the use of DBP in cases of purely ligamentous injuries is advantageous as it not only adequately holds the joint in the anatomical position but also is an extremely strong construct and does not fail or break even in the absence of any ligamentous healing.

Purely ligamentous lisfranc injuries are specifically difficult to accurately diagnose. These are missed or diagnosed late. We found that 8 (29%) of our patients presented late to us and their injuries were missed in the ED, in keeping with rate of delayed or missed diagnosis of 20%–24% of these injuries [6, 25, 26] reported in different series. There is evidence that delay in diagnosis and treatment is associated with poor outcome [8, 10, 25]. We recommend that weight-bearing anteroposterior and lateral views should be routinely performed. If there is significant swelling, pain, or plantar bruising in the absence of any bony injury or subluxation, magnetic resonance imaging (MRI) scan should be done to accurately diagnose these injuries.

The strengths of this study are that all patients were operated on by one surgeon at one unit and assessed by a single senior assessor. The same treatment protocol was followed for all patients and a large number of patients (22 out of 27) were available for followup assessment. We used both AOFAS and FFI scores to assess our clinical results. Majority of our patients had a good functional and clinical outcome with the mean AOFAS midfoot score of 78.1 and mean FFI of 19.5 comparable to other series.

There are a number of limitations of our study. It is a retrospective study. Some of our patients were seen and treated with a delay after their injury and thereby had considerable swelling and required a further delay till the acute swelling resolved before proceeding to surgery. This is likely to have influenced the outcome and results in our series. The average followup in our series is 2.4 years and a longer followup will provide a better outlook of the long term results of these complex injuries with potential of posttraumatic arthritis and foot dysfunction.

Conclusion

Lisfranc injuries are uncommon injuries with a high propensity to be missed, especially purely ligamentous injuries. High index of suspicion is essential for prompt diagnosis. Weight-bearing radiographs should be undertaken to accurately diagnose these injuries early. There should be low threshold for doing a CT scan if the patient is unable to weight bear or MRI scan when purely ligamentous injury is suspected. Good functional outcomes can be achieved by ORIF with preferably joint-sparing surgery by DBP. Prognosis is predicted by the severity of the initial injury and accuracy of reduction.

Overall good functional outcome can be expected for most patients with this injury. In our study 68% of our patients were able to return back to their preinjury level of function and work. However, a proportion of these injuries will lead to secondary midfoot OA and some residual symptoms.