Comparison of Arthrodesis and Non‐fusion to Treat Lisfranc Injuries

Abstract

Objective

“Lisfranc joint injury” is comprised of a tarsometatarsal joint‐complex injury. The Lisfranc complex injury is always a challenge for orthopedists, and the optimum treatment is still up for debate. Anatomic reduction and stable internal fixation prove to have no satisfactory outcomes. This research aims to compare the clinical curative effects, complications and radiographic features of arthrodesis and non‐fusion of the Lisfranc joint in the follow‐up of the patients who suffered Lisfranc injuries.

Methods

A comparative retrospective study of 25 patients with acute or subacute Lisfranc complex injuries was conducted between September 2013 and March 2015 in the First Affiliated Hospital of Soochow University. All patients were classified by Myerson classification. Eight patients were treated with arthrodesis, while 17 patients received non‐fusion operations. The clinical curative effects, complications and image differences were compared between the two groups. American Orthopaedic Foot and Ankle Society (AOFAS) hindfoot score, Short Form‐36 (SF‐36) and Visual Analogue Scale (VAS) score were evaluated for each patient during the follow‐up. All statistics were analyzed using the SPSS software system.

Results

All fractures healed for both the arthrodesis group and the non‐fusion group. Patients in the arthrodesis group had a higher AOFAS score compared with patients in the non‐fusion group (94.00 vs. 88.58, P = 0.034). Complications occurred in eight patients (8/17, 47%) in the non‐arthrodesis group, including the second and third phalanx abduction (1), talipes cavus (2), eversion deformity of front foot (3), eversion deformity of calcaneus (1), as well as postoperative infection (1). Only two patients (2/8, 25%) in the arthrodesis group suffered complications. One was a limitation of motion of the front foot and pain during walking; the other was an eversion deformity of front foot.

Conclusion

Primary arthrodesis has advantages compared to primary open reduction and internal fixation (ORIF): reduced foot deformity rates, sustained biomechanical morphology of the feet, reduced complications, higher level of function recovery, shorter time of surgical procedures, fewer complications, higher AOFAS score and fewer frequency of complications. According to our research, primary arthrodesis may be a better choice for treating Lisfranc injury.

Introduction

“Lisfranc joint injury” is the term used to describe a tarsometatarsal (TMT) joint complex injury. The term was coined by a French doctor named Jacques Lisfranc in 1815 who amputated the foot of a soldier who suffered a midfoot injury after falling from a horse1. The injury is not a common fracture, and involves ligamentous injury or bony and articular injuries. The injury is often misdiagnosed and treated inappropriately, which can result in unstable TMT joints, pain, and traumatic midfoot osteoarthritis and long‐term disability. It is important to continue to research the most appropriate treatment of Lisfranc injury, so that patients can achieve better clinical efficacy. If timely and definitively diagnosed and treated by a proper surgeon, acceptable results can be achieved.

The Lisfranc complex injury makes up approximately 0.2% of all fractures or dislocations, with an incidence of 1 in 55,000 people every year. It usually occurs after an axial rotation force caused by a plantar flexion of the foot, such as when stepping off a curb or during a high‐energy collision such as high‐speed motor vehicle accident2. In 1986, Myerson proposed a longitudinal three‐column theory. The columnar medialis contains the medial cuneiform bone and the base of the first metatarsal; the intermediolateral column, which is the most rigid, consists of the second and third metatarsals and middle and lateral cuneiforms; the fourth and fifth metatarsals and the cuboid constitute the lateral column, which is the most mobile column. The Lisfranc ligament (oblique interosseus ligament) is the only connection between the medial and middle column; it allows mobility between the medial and middle column. It reaches from the lateral margin of the first metatarsal cuneiform to the distal half of the medial side of the bottom of the second metatarsal, which provides the most strength support of the TMT joint complex3, 4. It holds the base of the second metatarsal and the lateral margin of the medial cuneiform together. The base of the second metatarsal bone, which extends nearly 1 cm proximally to the base of the first metatarsal and approximately 0.5 cm proximally to the base of the third metatarsal bone, plays a highly important role in the steadiness of the Lisfranc joint. Clinical manifestations are pain and tenderness at the TMT joint, pain with the pronation–abduction test of the forefoot, and plantar ecchymosis on the middle of the foot.

The TMT joints make up the bony support structure of the transverse arch of the midfoot5. In the weight‐bearing anteroposterior (AP) radiographies, the inner side of the second metatarsal base should be aligned with the inner side of the middle cuneiform for uninjured feet. For all patients, the typical mechanism of injury involves an axial load forced on a plantar‐flexed foot6. There is still debate over the optimal surgical procedure for patients with Lisfranc injuries. Initial misdiagnosis and missed diagnosis rates of Lisfranc injuries have been reported to be almost 20%7. In recent years, the focus has been on the comparison between primary fusion and non‐fusion Lisfranc injuries, especially with completely ligamentous injuries8. Some authors believe that Lisfranc injuries should be first treated with open reduction and internal fixation. If there are still serious complications, a second‐stage fusion treatment could be an appropriate choice9. However, some authors suggest that primary arthrodesis of Lisfranc injury is the optimal treatment. Many researchers suggest that the first choice is primary fusion of the Lisfranc joint in the case of poor healing of ligament–osseous structure interface10. It is great a challenge for surgeons to treat complete ligamentous disruption of TMT joints. Ignorance of subtle Lisfranc injuries would result in chronic pain and deformity. In these cases, fusion is essential11. Granberry and Lipscomb propose that in any injury type, primary fusion is required, because of high incidence of reoperation rates12.

Open reduction and internal fixation would not lead to a satisfactory outcome, and most patients would suffer traumatic arthritis or other post‐operation complications13. A timely open reduction and internal fixation (ORIF) is the gold standard for treating TMT injuries14. In our research, the primary fusion of Lisfranc joints has had better clinical effect than ORIF15. Thus, our preferred method for the treatment of most Lisfranc fractures/dislocations involves the fusion of medial and middle columns and elastic fixation of the lateral column16. The aims of our approach are to establish a stable joint, to recover the longitudinal arch of the foot, and to form the shape of the plantigrade foot, so that we can minimize subsequent complications17. We compared two groups of patients who suffered Lisfranc complex injuries, and the arthrodesis group turned out to be much better clinical outcomes.

Materials and Methods

Inclusion and Exclusion Criteria

The inclusion criteria are patients who had undergone arthrodesis or non‐fusion procedures of the TMT joint on a fresh Lisfranc complex injury without any other trauma or fractures. Patients with comminuted fracture, multiple injuries, and other basic diseases, such as chronic obstructive pulmonary disease, diabetes mellitus, hypertension, coronary heart disease, and cerebrovascular disease, were excluded.

General Information

According to this criterion, 25 patients with Lisfranc injury who were treated with arthrodesis or open reduction and internal fixation of TMT joints in our hospital by a single surgeon between September 2013 and November 2015 were included in the retrospective comparative research. For all patients, we evaluated the anteroposterior, lateral, and 30° oblique radiograph of the injured foot. All patients were fully instructed that there were two operative methods and they were informed that there was controversy over which method was better. There was no significant difference in patient demographics. There were 8 patients (5 men and 3 women) in the arthrodesis group and 17 patients (12 men and 5 women) in the non‐fusion group. The mean duration of follow‐up was 15 months10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27 in the arthrodesis group. The average age was 37 years (range, 18–65) and the mean duration of follow‐up was 7.5 months6, 7, 8, 9, 40 years (range, 34–52) in the non‐fusion group (Table 1). The clinical manifestations of the patients who suffer from Lisfranc injuries are severe pain, swelling, and deformity of the midfoot. The complications are nerve and blood vessel injuries and compartment syndrome. Plantar ecchymosis can appear in patients with more serious symptoms. The edge of the cuboid and navicular are not collinear with the dorsal and plantar side of the metatarsal in X‐ray. Fleck sign (tiny bone fragments) can be detected in the TMT articular space on a CT scan (Fig. 1A). Lisfranc ligament disruption can be discovered by MRI. All Lisfranc injuries were classified by Myerson classification (Types A, B and C)18. The injuries were classified as homolateral, isolated, and divergent. The most common injury type was B2 (38%). In the non‐fusion group, 7 patients had multiple injuries, and 2 patients had osteoporosis (proved by X‐ray and CT preoperatively, Table 1). In the arthrodesis group, no other fractures or osteoporosis were found in the patients. During the final follow‐up, all patients were evaluated on the basis of the Short Form‐36 (SF‐36), the American Orthopaedic Foot and Ankle Society (AOFAS) hindfoot score, and the Visual Analogue Scale (VAS) score (Table 2)19.

Table 1.

Clinical characteristic of the patients

Groups	Age (mean [range], years)	Gender		Myerson classification			Accompanied osteoporosis
		Male	Female	I	II	III
Arthrodesis group (n = 8)	40 (34–52)	5	3	1	6	1	1
Non‐fusion group (n = 17)	37 (18–65)	12	5	3	10	4	3
P‐value	0.492	0.686		0.725			0.743

Figure 1.

(A) Fleck sign (tiny bone fragments, red arrow) can be detected in the TMT articular space on computed tomography scan of Lisfranc injuries. (B) The first, second, third, and fourth metatarsal fracture. (C) The X‐ray postoperation showed that the TMT joint was anatomically reduced and the cuneiform was collinear with the metatarsal. (D) X‐ray 1 year postoperation showed that bony union was acquired in the TMT joint. (E) After removal of the internal fixation, the patient did not feel any discomfort. (F, G) No postoperative complications or deformities of the foot occurred. During the follow‐up, the patient achieved an acceptable range of motion.

Table 2.

Mean AOFAS hindfoot score, VAS and SF‐36 score at the final follow‐up (mean ± SD)

Groups	AOFAS hindfoot score	VAS score	SF‐36 score
Arthrodesis group (n = 8)	94.00 ± 3.16	0.13 ± 0.35	86.00 ± 3.70
Non‐fusion group (n = 17)	88.58 ± 6.36	1.00 ± 1.70	82.12 ± 5.71
P‐value	0.034	0.056	0.094

AOFAS, American Orthopaedic Foot and Ankle Society; SF‐36, Short Form‐36; VAS, Visual Analogue Scale (VAS).

Surgical Technique

Preoperative Measures

It is better to delay surgery for 1–2 weeks when the swelling has subsided, as indicated by the skin wrinkling20. Unstable joints should be determined under fluoroscopy before surgery. Before surgery, a calf tourniquet is applied and inflated to 250 mm Hg. The operation is performed by making two longitudinal parallel incisions (Fig. 2A, B)21.

Figure 2.

(A) Two parallel incisions: medial incision and lateral incision. (B) Plantar ecchymosis syndrome of the injured foot. (C) Removing all remaining cartilage, sclerosis, and fibrous tissue, the subchondral bone could be exposed. (D) X‐ray post‐operation showed the good reduction of the fractures. The medial and middle columns are steadily fixed with screws, while the lateral column is elastically fixed with K‐wires.

Incision

The medial incision, which is approximately 6 cm, should be in the center of the first and the second metatarsals. The lateral incision, which is nearly 3 cm distal to the first TMT joint, should be made just between the extensor digitorum longus tendon and the extensor hallucis longus tendon. The surgeon must be careful not to make the incision too medial or too lateral. Care must also be taken not to impair the sensory branches of the deep peroneal nerve during the medial approach, which can be avoided by laterally retracting the dorsalis pedis artery. The lateral branches of the superficial peroneal nerve can be damaged by the lateral incision and ought to be isolated and protected. The incision is usually just lateral to the extensor digitorium longus tendon, which is easily elevated to expose the underlying extensor digitorum brevis. It is really important to guarantee an adequate skin bridge between the two incisions. Full‐thickness skin flaps, which contain the dorsalis pedis artery, need to be carefully protected. The Lisfranc ligament and dorsal capsular structures are disrupted and the fracture segments are easily recognized.

Reduction

For arthrodesis, Lisfranc joints must be debrided adequately and free small fragments should be taken off. The articular cartilage is completely removed and little holes are drilled with 2.5‐mm K‐wires. After removing all remaining cartilage, sclerosis, and fibrous tissue, the subchondral bone can be exposed (Fig. 2C). The aim is to make sure all sclerosis, cartilage and fibrous tissue have been removed. Subchondral bone should be exposed with the help of small hinterman distractor curets, rongeurs, and osteotomies. The position of the second metatarsal and the medial cuneiform was maintained by a reduction clamp.

Internal Fixation

After reduction, the alignment was confirmed by fluoroscopy. The TMT joint was temporarily stabilized by K‐wires. One 3‐mm cannulated, headless compression screw was inserted from the medial cuneiform to the base of the second metatarsal in a straight line following the direction of the Lisfranc ligament. The medial column was immobilized with another headless screw from distal to proximal. Finally, another 3‐mm cannulated, headless compression screw was inserted from the medial to the lateral of the metatarsal. Stable fusion was achieved with 3‐mm cancellous compression screws. After that, a large clamp was used to pull the metatarsal base to the “mortise” position. Countersinking all the screw heads to prevent dorsal breakout of the base of the TMT joints is really important. The first metatarsal and the medial cuneiform, and the second metatarsal and middle cuneiform were fixed to each other, respectively. After stabilization of both the medial and median columns, the stability of the lateral column was affirmed. If the laterals turned out to be unstable, they were reduced and immobilized with K‐wires under fluoroscopy (Fig. 2D)22. The procedure is similar to ORIF, with the only difference being that the cartilage is not removed and the cannulated compression screws are fixed in a similar position.

Postoperative Care

All patients were required to not bear weight for at least 8 weeks postoperatively23. Between 8 and 12 weeks, gradual increase to partial weight‐bearing was allowed24. The K‐wires in the lateral column were removed 8 weeks after surgery. After 12 weeks, when the radiographs show bone union, the patient may weight bearing fully walking25. The internal plate and screws were removed between 6 and 12 months postoperatively. Lattermann suggests internal fixations be removed between 12 and 16 weeks after surgery. Some authors suggest that the medial column screws could be preserved unless they caused pain26. Patients were assessed 1, 3, 6, and 12 months post‐surgery, followed by annually outpatient follow‐ups.

Statistical Analysis

Quantitative data from each of the two groups was compared using Student's t‐test; their rates were compared using the χ²‐test with statistical significance defined as P < 0.05. Statistical analysis was performed with SPSS version 19.0 for Windows (SPSS, Chicago, IL, USA).

Results

General Situation

The average operation time was 125.1 (range, 40–235) min in the non‐fusion group and 118.8 min (range, 54–220) in the arthrodesis group. The average blood loss was 182.5 (range, 50–470) mL in the non‐fusion group and 176.5 (range, 50–520) mL in the arthrodesis group. There was no significant statistical difference between these two groups in the average blood loss and average operation time. The mean follow‐up time was 15 (range, 10–27) months in the non‐fusion group and 7.5 (range, 6–9) months in the arthrodesis group.

Clinical Assessment

The mean AOFAS hindfoot score was 87 (range, 71–100) in the ORIF group. This was lower than that for the arthrodesis group, which was 94 (range, 90–100). The mean SF‐36 score was 81 (range, 70–90) in the non‐arthrodesis group, which was lower than that of the arthrodesis group (mean, 87 points; range, 81–92 points); however, no statistically significant differences were found. The mean VAS score was 0 (range, 0–5) in the non‐arthrodesis group and 0 (range, 0–1) in the arthrodesis group (P = 0.056).

Complications

Postoperative pain occurred in 5 (5/17, 29.4%) patients in the non‐arthrodesis group and 1 (1/8, 12.5%) patient in the arthrodesis group. The locations of pain in the non‐arthrodesis group included the location of the steel plate in 1 patient, the planta pedis in 3 patients, and the third metatarsal in another patient. One patient who underwent arthrodesis developed postoperative pain in the planta pedis. Removal of the internal fixation was required in 15 (15/17, 88.2%) patients in the non‐arthrodesis group. In the arthrodesis group, 6 patients (6/8, 75%) needed to have the internal fixation taken out. Complications occurred in 8 (8/17, 47%) patients in the non‐arthrodesis group, including the second and third phalanx abduction (1), talipes cavus (2), eversion deformity of front foot (3), eversion deformity of calcaneus (1), as well as postoperative infection (1), respectively. Limitation of motion of the front foot and pain during walking occurred for 1 patient who underwent arthrodesis, while another patient in the arthrodesis group developed eversion deformity of the front foot. One patient in the arthrodesis group suffered redislocation and underwent re‐revision of the tarsometatarsal joint arthrodesis. After the revision, a radiograph showed that the alignment of the talus and first phalanx was good.

Typical Case

General Information

A 63‐year‐old woman was hospitalized for 1 day because of limited mobility caused by trauma‐induced pain in her left foot. The patient accidentally fell at home and sprained it her foot 1 day prior. She experienced pain, swelling and limited movement of her left foot. She was sent to hospital immediately. The X‐ray showed that it was a first, second, third, and fourth metatarsal fracture. The admission diagnosis was a left‐foot Lisfranc injury. After conservative treatment, the patient felt that the pain had intensified. Physical examination showed swelling of the left foot, good skin condition, plantar ecchymosis, tenderness of the left TMT joint, and activity limitation of the left toe. X‐ray: first, second, third, and fourth metatarsal fracture (Fig. 1B).

Operation Method

After anesthesia, the patient was placed in supine position with a calf and left ankle tourniquet. The first dorsal longitudinal incision was taken between the first and second metatarsals. Skin and subcutaneous fascia were then cut. Special attention was paid to the dorsal neurovascular bundle. The left short extensor tendon was pulled laterally. Severe TMT joint dislocation was obvious after subperiosteal exposure of the medial, middle and lateral cuneiform, as well as the first, second, and third metatarsal. After removal of the intraarticular hematoma, soft tissue, and bone block with forceps, Kirschner wire was used to temporarily stabilize the joint. First, one cannulated screw was inserted from the medial cuneiform to the base of the second metatarsal, parallel to the Lisfranc ligament. Then, two cannulated screws were used to fix the second and third TMT joints longitudinally; the first TMT joint was fixed with plate, and the second TMT joint fusion works by bone grafting. The second dorsal longitudinal incision was taken between the fourth and fifth metatarsal. Skin and subcutaneous fascia were then cut. The fourth base of the metatarsal fracture and dorsal dislocation of the fourth TMT joint can be revealed after subperiosteal exposure of the cuboid. A Kirschner wire spreader can help restore the length of the lateral column. After Kirschner wire elastic fixation of the fourth TMT joint, its tail was bent and buried into soft tissue.

Result

The patient's symptoms were obviously relieved. The X‐ray postoperation showed that the TMT joint was anatomically reduced and the cuneiform was collinear with the metatarsal (Fig. 1C). X‐ray 1‐year postoperation showed that bony union was attained in the TMT joint (Fig. 1D).

The patient was followed up for 8 months. Improvement in pain was remarkable during the follow‐up. The scores for this patient were: 94 on AOFAS, 81.1 on SF‐36, and 1 on VAS. No postoperative complications and deformities of the foot occurred. During the follow‐up, the patient achieved acceptable range of motion. After removal of the internal fixation, the patient did not feel any discomfort (Fig. 1E–G).

Discussion

The Lisfranc injury is not a common fracture and only accounts for 0.2% of all fractures2. Many cases of hhhoLisfranc damage are hidden and subtle. They are easily missed during the clinical diagnosis and treatment process. If not diagnosed early, the majority of patients would suffer TMT joint arthritis in later years. Strong understanding of the anatomy is conducive to the diagnosis of Lisfranc fracture dislocation. The TMT joints have little mobility apart from micro‐sliding. When the middle foot is swollen and causing pain, weight‐bearing and stress radiographs should be used for routine examination of the foot. This is of great significance to reducing the long‐term Lisfranc damage. Closed reduction is unsuccessful in most cases, due to the instability of TMT joints and the inability to restore the translation and remove the soft tissues embedded in the fracture line27. Achieve a plantigrade, painless, and stable foot is our goal28. Anatomic reduction of TMT joints is the gold standard of treatment of Lisfranc injuries29. Only 35% of the patients treated with anatomic reduction developed osteoarthritis compared to 80% of those who underwent non‐anatomic reduction30. Some authors suggest that patients with displaced or unstable Lisfranc injuries require anatomical reduction and internal fixation. Without arthrodesis, primary fusion only occurs in cases of purely ligamentous injuries or severe intra‐articular damage31. However, ORIF of TMT joints does not always achieve excellent clinical effects; according to our research, arthrodesis of the TMT joints attains much better clinical outcomes.

Kuo et al. conducted a retrospective study of 48 patients with TMT joint injuries that were first treated with open reduction and internal fixation. Twenty‐five per cent of the patients developed traumatic arthritis and 50% of them later underwent arthrodesis13. According to another study, the reoperation rate in an open reduction and internal fixation group was more than 75%, while in a fusion group was only 17%–20%32. Sangeorzan et al. found that early arthrodesis after a failed treatment for Lisfranc injury would have a positive effect33. In a randomized prospective study, Coetzee and Ly underwent a study with 41 patients who suffered isolated acute or subacute injuries. The follow‐up results showed that the arthrodesis group reached 92% of the pre‐injury level compared with the ORIF group, which only reached 65% of their pre‐injury level, and the arthrodesis group had higher AOFAS midfoot scores34, 35. A study conducted by Henning showed that 26 patients treated with open reduction and internal fixation (79%) had a higher higher rate compared to the primary fusion group (17%)32. Professional athletes have even been able to return to competitive sports after fusion of TMT joints9.

Open reduction and internal fixation was thought to be the optimal treatment for unstable injuries31. Recently, however, more researchers have supported the use of primary arthrodesis rather than primary ORIF. There would be a high incidence of arthritis even if treated with the most perfect open reduction and internal fixation. The mean AOFAS score was higher in the primary fusion group (Table 2) at a similar injury level. There is no apparent difference in anatomic reduction between primary open reduction and internal fixation and primary arthrodesis groups. The inherent stability of the Lisfranc joints depends on the scar tissue that is formed after ORIF of the Lisfranc joint; however, the scar tissue is not strong enough to support the injured TMT joint, and it has been shown that ligaments do not heal after ORIF and TMT joints deformity formed. Clinical results showed that, compared to the fusion group, the non‐fusion group had more complications. In the open reduction and internal fixation group, 11 patients suffered complications requiring reoperation or conservative treatment. After removal of the hardware, some primary non‐fusion patients had TMT joint pain, talipes cavus, and eversion deformity of front foot (Fig. 3). Talipes cavus deformity was a result of the scar contraction of the middle foot and eversion deformity of the front foot due to unstable structures formed after surgery. Four patients in the ORIF group had postoperative pain, including three with plantar pain and one with internal fixation pain. Changes in the shape of the foot and the loading forces result in plantar pain. One patient had postoperative infection. After several thorough debridements, the patient recovered well. In the fusion group, all patients attained union of TMT joints and only one patient had a little movement limitation of the front foot. Asymptomatic degenerative joint disease developed in 25% of all patients in the fusion group, but did not require surgery9. Patients should not walk bearing weight for at least 6–8 weeks until bony healing. There would be a high risk of malalignment of bones if weight bearing is recommenced early.

Figure 3.

In the follow‐up, we found that some non‐fusion patients had malformation postoperation after removal of internal fixation. (A) Talipes cavus deformity due to the scar contracture of the middle foot. (B) Eversion deformity of the front foot due to unstable structures formed after operation.

Still, there are several weakness in this research. First, the number of patients is small (25 patients). Second, we choose three types evaluation system. AOFAS scale score is the most commonly used system to assess patients. The other two systems have limitations and inherent errors. All evaluation systems were assessed by one doctor to reduce errors. Finally, caution should be taken in the selection of patients. For example, young athletic patients should be excluded36.

A primary fusion of TMT joints could prevent potential risks of foot deformity, reduce plantar pain, and decrease reoperation rates. The majority of the fusion patients had good results and osseous union. As a result, we tended to perform a primary fusion of the medial and middle columns in treating most of our patients who suffered from Lisfranc injuries.

Conclusions

In conclusion, Lisfranc injuries are complex and care must be taken in selecting the appropriate treatment. Primary arthrodesis has advantages compared to primary ORIF, including reduced foot deformity rates, sustained biomechanical morphology of the feet, reduced complications, higher level of function recovery, shorter time of surgical procedures, fewer complications, higher AOFAS score, and fewer frequency of complications. According to our research, primary arthrodesis may be a better choice for treating Lisfranc injuries.