John Joseph Anderson, G. Parker Peresko, Riley J. Rampton
Revisional Surgery of Lisfranc Injuries and Deformities
Indications
Contraindications
Preoperative Considerations
Surgical Techniques
Correcting Surrounding Deformities
Soft Tissue Equinus Deformity
Osseous Malalignment and Deformity
Cavus Foot and Ankle
Valgus Foot and Ankle
Nonunions
Lateral Column Pathology
Retained Hardware
Revisional Surgery of Lisfranc Injuries and Deformities
One of the most debated questions which authors often pose when discussing surgical management of the Lisfranc joint complex is on whether to fixate via open reduction and internal fixation (ORIF) or opt for primary arthrodesis.1–12 A lesser discussed topic is what to do and how to approach the revision surgery after either primary arthrodesis or subsequent arthrodesis post-ORIF fails. The Lisfranc joint complex is the second most common site in the foot that will manifest osteoarthritis, and of the 17% of radiographic osteoarthritis that affects people over 50 years of age,13 10.7% is attributable to these joints.14 This means that arthrodesis is very likely to occur at some point in these patients, with nonunion and subsequent revision surgery being a very possible outcome. Various literature reviews reveal nonunion rates from 0% to 33%,1,4,6–10 with the lower end of the spectrum found in young healthy patients(Clinical Cases 1 and 2). When subjects with Charcot neuroarthropathy (CN) are added to this pool, the nonunion rates can increase considerably.15–17 This is further compounded by the fact that the Lisfranc complex is the most common joint site for CN to occur18 and that surgical success is also dictated by the stability and support of the surrounding joints, which are also prone to collapse. These considerations warrant the lower extremity reconstructive surgeon to exhibit skillful planning, definitive surgical intervention, and precise perioperative management to prevent further complications and successfully address the deformities, and cause of recurrence is imperative regarding this lesser explored, but important surgical circumstance.
This chapter will focus on all the aspects of revision arthrodesis of the tarsometatarsal (TMT) joints, providing detailed surgical techniques and pre- and postoperative management as well as addressing downstream complications stemming from difficult scenarios. Not only will there be an in-depth look at nonunion revision in the “normal” patient but also the challenging endeavor of revisiting the CN Lisfranc complex after failed arthrodesis.
For the approach of TMT pathology, the authors separate this surgical approach into the following categories:
For the planning of the surgical approach, the authors pay close attention at the residual deformity, position, and surrounding joint deformities to find the best approach for deformity correction and pain relief. The authors attempt to salvage any joints if possible in order to allow surrounding joint motion, as we appreciate any restriction on a fused joint will transfer the pressure to surrounding joints. Many times advanced imaging such as magnetic resonance imaging and computed tomography (CT) or isolated injections of anticipated fused joints with a local anesthetic under C-arm fluoroscopic guidance will help making the decision on fusing additional joints but allow for the final decision to be made during surgery based on intraoperative findings.
Indications for revision Lisfranc arthrodesis can include failed primary arthrodesis with delayed union, malunion, or nonunion refractory to conservative measurements. Painful or infected hardware necessitate intervention or lack of stability after explantation of hardware will require revisional surgery. Peripheral neuropathy, regardless of the cause, may represent the number 1 risk factor for failed TMT surgery and arthrodesis in general.
Contraindications for revision Lisfranc arthrodesis include active infection at the revision site, patient health status, or medical comorbidities that would make the risks of surgeries outweigh the potential benefits. Relative contraindications may be poor glycemic control, smoking, inability to stay non−weight-bearing, poor circulation, and noncompliance of patients. Each relative contraindication must be addressed and managed to the best of the patient’s and surgeon’s ability. Certainly any history of bone or joint infection must be evaluated and resolved prior to undergoing surgery. Most revisional surgeries are not one and done, meaning that the surgeon and the patient have to appreciate that long-term additional joints may need arthrodesis or correction and/or hardware removal.
There are numerous factors to recognize and assess prior to surgical reintervention. Socioeconomic factors; addressing nutritional, metabolic, and vitamin abnormalities; detrimental comorbidities such as tobacco use, obesity, and age; operative planning; and what forms of fixation and biologics may be used are just a few to be named that are among the highest priority. Ability to stay non−weight-bearing must be addressed preoperatively as well.
When discussing comorbidities, smoking should be regarded as one with the utmost priority. The discontinuation of smoking long before a surgical incision is made can ultimately be a paramount factor in obtaining a successful outcome in regard to arthrodesis and certainly can modify your surgical approach. Strictly taking nonunion into consideration, it has been shown that smoking can increase the chances of nonunion almost threefold.18,19 Additionally, smoking and tobacco use has also been shown to delay fracture healing as well as decreases in vascularity and flow, imparting an obligation on the surgeon to convey the negative impacts that it could have on wound healing, osseous union, and infection and an array of other detrimental effects to one’s health.19,20 Additionally, one must consider vaping and electronic cigarette form as a subset of smoking and forms of tobacco as being harmful to healing in general. It is known that byproducts of tobacco, nicotine, and other chemicals are harmful to healing and at minimum create a less desirable environment for new bone growth at the osteoblastic and osteoclastic level. A thorough discussion and education on options and methods of smoking cessation with a recommendation to the patient’s primary care physician for further management before operation is paramount for a successful outcome.
Varying literature exists on other comorbidities which can change the outcome of surgery as well. Literature regarding age and obesity has shown findings suggestive of nonunion association.20–24 It should be noted that some of these studies investigated the cause of nonunion after fracture fixation and not arthrodesis surgery, although we believe it is still important to include this information.22,23
Vitamin D plays an integral role in osseous consolidation and ensuring that union is achieved. Normal serum levels are defined as 30 to 100 ng/mL (measured as serum-25 hydroxyvitamin D) while insufficiency falls between the ranges of 21 to 29 ng/mL, and lastly, deficiency is seen as ≤20 ng/mL. The consensus has been established that anything that falls into a “deficient” range is less than ideal for normal skeletal health; this includes an increased incidence of nonunion.25 Other authors, while still advocating for the use of vitamin D, argue that there is not enough concrete evidence to link hypovitaminosis D and nonunion rates.26,27
The authors of this chapter currently advocate for routine testing and subsequent supplementation of vitamin D starting in the preoperative period and continuing as needed, especially in the context of revisional surgery, when not contraindicated. Citing literature such as Moore et al, who revealed that patients with either vitamin D insufficiency or deficiency had a greater risk of 8.1 times that of patients with normal levels to go on to nonunion, we firmly adhere to the practice of routine vitamin D supplementation.28 We routinely supplement patients on 50,000 units D2 weekly and 2 to 6000 units of vitamin D3 daily and a daily multivitamin which may include a portion of vitamin D.
Evaluation of these patients with the added complexity of both diabetes mellitus and CN creates an entirely new and more challenging endeavor when planning for surgery. One must obtain up-to-date laboratory testing and properly manage glucose levels and ensure that the hemoglobin A1C (HbA1C) levels are adequate for surgery. Continuing on the subject of vitamin D levels data, high level studies have shown that diabetic patients, with or without the presence of CN, exhibit higher instances of vitamin D abnormality, reinforcing the need for routine evaluation, as aforementioned.29,30 Other values, such as HbA1C must be closely monitored, as having levels >7% poses an increased risk for delayed union, malunion, and nonunion.31,32 The importance of achieving union can be further stressed as nonunion has been cited as a risk factor associated with amputation. Rationale for controlled glucose levels is found beyond union, extending also into wound healing and infection rates, with each HbA1C percentage increase correlative to an odds of postoperative infection by 1.59 times.33 Neuropathy, alone and in the presence of an elevated HbA1C, has also been indicated as an element of concern when considering surgery in this patient population. Domek and colleagues evaluated 21,854 diabetic patients a year prior to surgery, finding similar results to Humphers et al, with each percentage increase of HbA1c corresponding to a 5% increased likelihood of complication development. Their overall 30-day complication rate was shown to be 3.2% and the average patient experiencing a complication having a HbA1C of 6.29%. In the same study, neuropathy itself posed a 1.78 times risk for sustaining a complication, and when diabetes mellitus presence was coupled with comorbidities, the risk increased to a factor of 3.08 times that of normal.34
Revision itself has been shown to be a pertinent risk factor for nonunion; one such study assessing foot and ankle revision nonunion cases found a 2.8-fold increase in nonunion for every revision attempted, and overall incidence in the single study was observed to be 23%.35 This should be thoroughly considered when implementing the surgical plan and construct you will ultimately apply to the patient as each attempt could pose a more undesirable level of success. The authors additionally try to avoid the use of steroids and nonsteroidal anti-inflammatory drugs in the first phases of bone healing to assist in healing.
Bone graft, biologics, and other supplemental materials allow for additional length, structure, stability, and increased healing potential depending on the type implemented. An abundance of literature demonstrates the additive healing properties of these products, with fresh autograft taken from the donor still held in high regard as the gold standard. Autograft possesses all the inherent healing properties of bone including osteogenesis, osteoconduction, and osteoinduction, thus earning its title as the gold standard. Harvest can be obtained from the distal or proximal tibia approximately at the level of the tibial tuberosity in regard to the latter;36 at the calcaneus, either as a structural graft from the superior lateral aspect or with use of a trephine to harvest through a minimal incision with a lateral approach, thus minimizing donor site morbidity.37–39 Differing types of autograft exhibit varying benefits with cortical graft providing much more structure and stability but taking much longer to incorporate. Conversely, cancellous autograft incorporates readily but cannot reach the levels of structure that its cortical counterpart imparts.37
Allograft is another readily available option, lacking only the osteogenic property that autograft has; however, it is still a high-quality option when other factors may preclude the surgeon from obtaining the patient’s autograft. It is important to note that not all allografts are the same, and multiple studies have shown that freeze-dried allograft causes a loss of osteoinductive property,40,41,42 thus diminishing its overall capacity for healing. Regardless of this, it is still a useful tool in the surgeon’s armamentarium. We routinely mix autograft with other allograft bone and sometimes use other scaffolding sources to assist in osseous union.36,43–45
Preoperative physical therapy evaluation and/or gait training is needed on most patients to assure the importance on non−weight-bearing status after the surgical intervention. Equally important in this discussion is a realistic time period of complete non−weight-bearing which in revisional surgery will be a minimum of 12 weeks, and if more than the TMT joint was fused or extensive autogenous bone graft was utilized, it may be outward of approximately 5 to 6 months. Patients must have this goal preoperatively and need to be educated in a 4- to 12-week transition of gradual weight-bearing toward full weight-bearing status. Postoperative splinting, casting and transitioning into a walking boot, Charcot restraint orthotic walker, ankle-foot orthosis with accommodative high top, and supportive shoes are paramount for the patient’s successful outcome.
Revision surgery is rarely straightforward with one described technique, as such it is challenging to present a technique that caters to this subject. We therefore present a multicomponent technique, incorporating varying aspects, pearls, and tips which may be encountered during these secondary and tertiary revisions.
Equinus deformity should be preliminarily addressed and considered a pillar of the surgical method in diabetic and nondiabetic patients alike, as it has been previously recorded as high as 96.5%.46 To digress momentarily, other benefits of addressing this pathology have been shown to be decreased reulceration and increased incidence of ulcer healing in the diabetic and CN population.47,48 Failure to perform either a gastrocnemius recession or tendo-Achilles lengthening (TAL) at the time of surgery has shown an increased risk of both failure fixation and joint subluxation.49 Additional correction of soft tissue may be achieved through release of the posterior subtalar joint and ankle capsule at the same time as the Achilles or gastrocnemius release. A central foot equinus may involve the need to release a tight abductor hallucis or even a central plantar fascia release. In longer term deformity, evaluation of peroneal longus function should be considered, especially if the medial column will not purchase the ground as seen in peroneal longus rupture or in a residual clubfoot deformity. In patients with peripheral neuropathy and TMT pathology, other tendon balancing and transfer options should be considered.
If the ankle or hindfoot has severe malalignment, a detailed discussion with the patient toward a correction of the surrounding deformity must be addressed to give optimal results. In general, one must think of fusing the surrounding affected joints and all others that have deformity creating unwanted stress across the arthrodesis site must be realigned with osteotomy or arthrodesis. To not address the surrounding deformity will appear easier in the beginning, but in the long-term, the easier path of leaving or ignoring contributing deformity to the arthritis or CN will lend toward a recurrence of the deforming forces across the corrected or arthrodesis sites with subsequent detrimental failures.
When revising a TMT joint in a cavus foot with a heel that stays in varus on attempted heel rise, one must consider both soft tissue balancing such as medial release, peroneal repair, posterior tibial tendon transfer, and addressing any contracture distally. It is known that in traumatic, neuropathic, and long-term equinus deformities, most feet will present with hammertoe deformities, extensor hallucis longus contracture, and lesser metatarsophalangeal joint (MTPJ) contractures that should be addressed. Additional abductor hallucis, distal plantar fascia, lateral or medial first MTPJ capsule, and soft tissue releases must be considered in revisional TMT surgery.
For most patients with cavus deformity, revisional TMT surgery must be combined with a solid structural correction of bone and joint realignment.
In general, the authors in this chapter start with addressing the ankle and heel and work forward with a stable slight valgus position of the hindfoot. If the heel stays in a varus position, unwanted stress across the TMT joint will potentially increase rotational and pivoting forces across the TMT complex. Much more likely is a need to address the lateral column of the TMT and cuboid joints. The calcaneocuboid and lateral TMT joints can rapidly progress into severe arthrosis with malalignment that would be tolerated in a valgus foot-type. When deciding on a cuboid or midfoot osteotomy, one is better served with repositioning the foot into a valgus position. It should be noted that the lateral column is addressed after the medial column of the TMT is fixed in a desired position.
After a rectus or slightly valgus hindfoot is achieved, next the talonavicular and calcaneocuboid joints are addressed. On occasion, the talonavicular joint will need to be fused and even in rare cases is the need to simultaneously perform an arthrodesis of the calcaneocuboid joint to assist in the TMT joint realignment. The talonavicular joint is our exception to working from the hindfoot forward; in other words, many times this joint can be fused and should at minimum be reevaluated after the TMT is fused. One must also appreciate that in a cavus foot, arthrodesis of the talonavicular joint will take more bone and shorten the medial column and therefore further worsening the cavus foot position especially if an arthrodesis of the TMT joint was performed. In a “Z” foot or skewfoot, it is not uncommon to simultaneously perform a talonavicular arthrodesis with a cuboid osteotomy to realign the TMT complex for a successful ideal positioning.
The TMT joint is addressed by evaluating the most deformed joint. If a nonunion is present, this must be addressed. In general, the authors start with the first metatarsocuneiform junction. This alignment is crucial in lining up the remaining lateral part of the TMT complex. Working from the medial to lateral side of the TMT complex is usually the best foundation to build on. In many cases, the authors will resect the intracuneiform joints to create further stability and osseous union. If the bases of the metatarsals have lost a great percentage of their arthrodesis surface, it may also benefit to establish an arthrodesis or fibrosis between the metatarsal bases.
Rather than establishing a standard guide for fixation, the authors generally try to follow sound fixation principles, but with each case, we tailor the fixation to the location of the joint(s) to fuse, bone grafting, and stability needed. We apply the general thought that the maximum amount of bone to bone contact at an arthrodesis site should be combined with the most stable construct.
An abnormal metatarsal parabola and painful lesser MTPJs may be present if all 3 planes of deformity are not appreciated when aligning the TMT complex. The surgeon must plan ahead to address the residual deformities and reduce abnormal metatarsal weight transfer and hammertoes at the end of the case, or with a secondary revision surgery, this is more critical in the rigid cavus and forefoot varus foot types.
In general, a valgus foot type has a much better outcome with positioning and alignment, and it can be accommodated in an orthotic, brace, or walking boot if the position and alignment are not ideal. The ankle and hindfoot must be realigned by focusing on getting the heel under the talus and tibia alignment in proper orientation. Many of these patients will have a concomitant calcaneal valgus deformity and posterior tibial tendon dysfunction. In this case scenario, the repair of the TMT complex in a malaligned hindfoot valgus will not allow for optimal positioning and deformity correction. In many cases, the subtalar joint is simultaneously fused or at minimum a calcaneal osteotomy can move back the heel in line with the tibia axis. A medial calcaneal displacement osteotomy is often useful, and if there is a combined transverse and frontal plane deformity, an Evans type of calcaneal osteotomy can be performed. As with the cavus foot deformity, the authors will many times reserve the talonavicular joint to give additional deformity correction as a final corrective action for the midfoot to forefoot correction and ground purchase.
Next, the naviculocuneiform joint(s) is evaluated. In most cases, a naviculo-first cuneiform arthrodesis is performed if needed for a crush injury or posttraumatic arthrosis, but in revisional cases, an arthrodesis of the navicular and all 3 cuneiforms as a unit is performed. In addition, a resection and bone grafting between the cuneiforms and arthrodesis together for a greater platform to build the TMT arthrodesis is recommended. After the more proximal foundation is created, it is then that the TMT complex is directly addressed.
In a valgus foot, the first TMT joint may be plantarflexed in order to compensate for the usual elevated first ray. This will help off-load any uneven lesser MTPJ weight transfer. The authors try to match the opposite side with relative alignment, but the most important features are a stable TMT complex with an attempt to decrease the usually widening of the foot that occurs in a valgus foot type. In many cases, exostectomies, intercuneiform arthrodesis, hallux abducto valgus, lesser metatarsal or Tailor bunion osteotomies can be added at the end of the surgical procedure to assist in creating a stable foot that actually will fit back into a reasonable shoe gear.
It would be unwise to think that one simply needs to fuse the affected joints to gain a good result. The surgeon must take into account the sagittal, frontal, and transverse planes of the TMT deformity. Width of the foot must be restored; height of the arch and medial column and length or proper parabola of the metatarsals must be appreciated. It is not uncommon to fuse the medial column in an excellent position and alignment, but only when examining the forefoot and lateral aspect of the foot we realize that we need to abduct or rotate the forefoot to achieve optimal purchase with a lateral column procedure, or if isolated to MTPJs, we will perform isolated osteotomies.
For most patients with a valgus foot and ankle deformity, revisional TMT surgery must be combined with a solid structural correction of bone and joint realignment.
In general, nonunions need to be evaluated with a CT scan. It is also very helpful if three-dimensional (3D) CT reconstruction is available. This helps in preoperative planning the amount of bone graft needed and how many or which of the joints need to be realigned or fused. Additional consideration into structural bone grafting and bridge plating must be given. The authors most often will bridge the arthrodesis sites and supplement them with the highest quality of inductive and conductive bone possible. The authors prefer quality and quantity of bone growth and bridging construct as opposed to compression, this is different than primary arthrodesis or open reduction surgery.
The lateral column of the TMT including the cuboid and lateral instability is one of the most difficult areas to address. Our approach to the lateral column pathology is as follows: (1) Ordering advanced imaging making sure the peroneal tendons are intact; (2) Isolated fourth/fifth TMT injection and follow-up evaluation; (3) Consider if any other joint malaligned joints are contributing factors as these joints will also need to be addressed; (4) Orthotic/insole trial will reveal much information until permanent correction is performed when needed; (5) Consider lateral column opening or closing correction and/or medial column shortening/realignment to off-load the lateral column; (6) Consider tendon interposition or lateral column implants which have very mixed results and the authors prefer an aggressive exostectomy/cheilectomy instead with a repair of overriding ligaments included; (7) When this fails, the authors will fuse the lateral column, and this should be taken into consideration after extensive biomechanical/gait examination and discussion with the patient.
In revisional cases, it is best not to leave any retained hardware since the loosening is creating a poor quality fibrotic bone and a focal area of necrotic fibrosis that leads to huge deficits around failed or broken hardware. When intraoperative decisions on hardware are made, it is with the thought of whether the failed hardware/broken hardware are in the way or will it be more destructive to remove the hardware. In some cases of mixed pictures of nonunion versus infection versus CN or neuropathic destruction, the hardware may need to be removed, bone biopsy is performed, and a staged reconstruction may be prudent after negative biopsies or after consolidation of the hardware sites occurs.