Ankle and tibiotalocalcaneal (TTC) arthrodesis procedures are well-established and effective treatments for myriad pathologies.1,2 The goals to be achieved in the short term are pain relief and genesis of a braceable, functional limb. Longer-term goals are preservation of surrounding joints by proper positioning of the talus under the tibia for isolated ankle arthrodesis extended to include the calcaneus when a TTC is performed. More recently, ankle or TTC arthrodesis to preserve structures for potential conversion to total ankle replacement (TAR) when the adjacent joints degenerate has been advocated.3
Nonunion of the ankle and/or subtalar joints result from a variety of factors. Frey et al4 conducted a retrospective review of 78 isolated ankle arthrodesis procedures performed between 1975 and 1990. Factors associated with nonunion included fracture type, especially fracture dislocations of the ankle and talar dislocations, avascular necrosis of the distal tibial plafond and/or talus, open ankle fractures, infection, and major medical comorbidities. Interestingly, factors not associated with ankle nonunion included patient age, prior history of subtalar or triple arthrodesis, or the ankle arthrodesis surgical technique/fixation method. Younger et al5 performed a retrospective review of 224 ankle and 368 subtalar arthrodesis procedures between 2006 and 2010 and determined an arthrodesis nonunion rate of 8.48% for the ankle and 4.35% for the subtalar joints. For both joints the nonunion rate was higher in male gender, tobacco use, elevated body mass index (BMI), diabetes, and advanced age. Chalayon et al6 performed a retrospective review over an 11-year period of straightforward, uncomplicated open ankle arthrodesis in 215 patients. The authors excluded any patient undergoing ankle arthrodesis for (1) Charcot neuroarthropathy, (2) insensate limb, (3) failed TAR, (4) simultaneous arthrodesis of the subtalar joint, or (5) arthrodesis performed within 1-year of injury to salvage failed fixation or painful function due to open fracture, segmental bone loss > 1 cm, infection, or talar body fracture. In their exclusively straightforward, uncomplicated open ankle arthrodesis patient population they identified a reoperation rate of 19% and a nonunion rate of 9%. Ankle arthrodesis nonunion was 3 times higher in patients with prior subtalar arthrodesis and 2 times higher in patients with preoperative varus ankle alignment. Patel et al7 performed a systematic review and meta-analysis of 13 studies (1 case-control, 12 retrospective cohorts) involving 987 patients undergoing isolated ankle arthrodesis (study sample size between 23 and 215) from predominantly the United States (11/13 studies). Ankle arthrodesis nonunion occurred in 21% of patients and ranged between 6.9% and 28%. Meta-analysis identified male gender, tobacco use, and history of operative site infection prior to ankle arthrodesis as strong predictors of nonunion. History of open injury was a moderate risk, and presence of avascular necrosis prior to ankle arthrodesis was determined to be a limited risk for nonunion.
With any revision surgery it is important to understand if the index surgeries were performed for appropriate indications and were technically performed well. To this end, understanding whether the performance of the primary ankle or TTC arthrodesis was via an arthroscopic versus open approach is an important consideration. Arthroscopic ankle or TTC arthrodesis is performed predominantly for (1) primary or secondary “other” (ie, gouty arthropathy, rheumatoid arthritis) causes of end-stage degenerative ankle and/or subtalar joint arthrosis with a well-aligned ankle in the sagittal and transverse planes, (2) flaccid drop-foot deformity correctable to neutral with minimal soft tissue releases, (3) a salvage procedure for multiple failed treatment attempts of a talar osteochondral defect or chronic lateral ankle instability, or (4) when the fixation of choice is 2 or 3 large-diameter compression screws for ankle arthrodesis and retrograde intramedullary nail for TTC arthrodesis.8
With the surgical approach and specific fixation employed being key factors in development of ankle arthrodesis nonunion, Henricson et al10 conducted an analysis of the Swedish Ankle Registry. The Swedish Ankle Registry includes data from 54 departments and involves 96% of all primary ankle arthrodesis and revision ankle arthrodesis procedures performed. Between 2008 and 2015 a total of 1773 primary ankle arthrodesis procedures were performed for a wide array of etiologies including congenital or acquired deformities, septic arthritis, psoriatic arthritis, talar avascular necrosis, unspecified secondary ankle osteoarthritis, instability, and osteochondral lesions. The surgical approach and method of fixation varied and consisted of open screw fixation, retrograde intramedullary nail, arthroscopic screw fixation, plate and screws, external fixation, percutaneous screw fixation, and “other” methods predominantly including missing data and hybrid fixation options. The overall risk of revision ankle arthrodesis was 7.4% at 2.5 years and 7.8% at 9 years. Of interest, the risk of revision ankle arthrodesis was 15% following arthroscopic screw fixation, 8% for open screw fixation, 5% for retrograde intramedullary nail fixation, and 3% for plate-and-screw fixation. Gender did not affect revision rate; however, unspecified secondary ankle osteoarthritis and posttraumatic arthritis were associated with greater risk of revision ankle arthrodesis. Since the procedures performed by department varied widely between 1 and 203, the authors surmised that poor surgical technique and/or surgeon inexperience, as well as poor patient selection, likely contributed to the need for revision ankle arthrodesis. Specifically, they surmised that arthroscopic ankle arthrodesis with screw fixation is more demanding than open approaches, has a longer and more steep learning curve, and requires a higher volume for successful outcomes to be achieved.
Specific to TTC arthrodesis, Pitts et al11 performed a retrospective review of 101 patients undergoing TTC arthrodesis between 2011 and 2019 and determined the following situations resulted in higher nonunion rates: (1) preoperative diagnosis of Charcot neuroarthropathy (44%), (2) preoperative diagnosis of nontraumatic osteoarthritis (39.1%), (3) diabetes (2.22 × increased nonunion risk), and (4) chronic kidney disease (2.37 × increased nonunion risk). Factors associated with increased infection risk were (1) preoperative diagnosis of Charcot neuroarthropathy (29.6%), (2) preoperative diagnosis of nontraumatic osteoarthritis (37%), and (3) age > 60 years (3 times increased infection risk). The use of autogenous bone graft was protective against nonunion. Patel et al12 conducted a systematic review and meta-analysis of 8 retrospective cohort studies from the United States involving 624 patients undergoing isolated TTC (study sample size between 19 and 154). TTC nonunion occurred in 31% of patients and ranged between 5% and 50%. Meta-analysis identified peripheral sensory-motor neuropathic conditions as a strong predictor of nonunion.
Regardless of cause, nonunion of the ankle and/or subtalar joints are poorly tolerated. Krause et al13 performed a post hoc analysis of data previously collected for a prospective randomized clinical trial on recombinant human platelet-derived growth factor-BB of 370 ankle and/or subtalar arthrodesis performed through open surgical approach and internal fixation. A total of 67 (18%) nonunions were identified on blinded computed tomography assessment; however, only 21 (5%) were identified on surgeon clinical examination alone. Nonunion was higher in patients with elevated body mass index, patients with tobacco use, and those who were unemployed. The differences between union and nonunion groups were clinically meaningful for all outcome measures with the nonunion group fairing significantly worse. The concept of an “asymptomatic” nonunion of the ankle or subtalar joint (imaging supporting nonunion but the patient doing well) was not supported. Although not specifically evaluated, malalignment in any plane is a major cause of failure and reason for revision. Ankle and TTC arthrodesis must result in neutral ankle and hindfoot position that will have minimum impact on the rest of the lower limb function.14 An overly plantarflexed talus, as well as anteriorly subluxed talus, will each rapidly wear the hindfoot/midfoot joints and cause a lurching of the body over the joint during walking (Figures 14.1–14.3).15 Similarly, the impact of coronal misalignment will cause early wear of the hindfoot/midfoot joints and pain.14,15
Regardless of cause, nonunion of the ankle and/or subtalar joints is expensive to revise. This is especially true when infection is the proximate cause for the nonunion. Gagne et al16 performed a retrospective review of 24 failed ankle arthrodesis procedures requiring revision for nonunion performed at 2 Canadian academic medical centers. They calculated the total costs (in US dollars) from billing codes, length of operation, period of hospitalization, and postrevision outpatient fees. The additional costs associated with ankle arthrodesis nonunion were $9683 US dollars consisting of $1061 for imaging, $627 for prerevision visits, $3026 for the revision surgery including orthobiologic grafting materials, $3432 for the hospitalization stay that averaged 3 days, and $1754 for the postrevision follow-up. They concluded that the costs associated with revision ankle arthrodesis were both an agonizing and expensive financial burden for the patient and insurance company alike. Similarly, nonunion following TTC arthrodesis can incur lifetime costs of up to $500,000 (in US dollars) for patients depending on the necessary treatment, such as below-the-knee amputation, and is associated with decreases in quality of life.17
Indications and Contraindications
Revision ankle and TTC arthrodesis nonunion surgery involves either a rearthrodesis of the involved joints10 or conversion of the ankle arthrodesis nonunion to a TAR.18 These procedures are indicated in situations where the soft tissues are nonhostile; infection, if present, has been completely eradicated; and the patient is medically healthy enough to survive the revision surgery and accepting of the protracted recovery process, as well as when below-the-knee amputation is not desired. These situations in addition to inadequate vascular supply are considered contraindications.
We employ a standardized approach to the management of each revision ankle and TTC arthrodesis, allowing for deviation and augmentation to this approach on an individual basis as clinically required. A thigh tourniquet is applied but not always inflated and a gel roll is placed under the ipsilateral hip to control physiologic external rotation. The patient receives a musculoskeletal ultrasound-guided popliteal/saphenous nerve blockade by the anesthesiologist and is placed under general anesthesia with endotracheal intubation. Unless the tissues are hostile, we use the same incisions employed during the index surgery, with careful soft tissue dissection to the level of the ankle joint capsule/scar, taking care to protect the adjacent neurovascular bundle(s) and maintain appropriate tissue planes for layered closure. With revision of a failed arthroscopic-assisted ankle arthrodesis, we employ anterior, lateral, or posterior approaches depending on the soft tissue quality and location of internal fixation requiring removal. Regardless of incisional approach, retraction is minimized to only what is necessary to aid in the preservation of the soft tissues. Following entry into the ankle joint space, the periarticular scar tissues are evacuated and sent for histopathology analysis. In addition, multiple deep cultures are procured for fresh frozen analysis and culture with sensitivities owing to the concern for indolent deep infection with revision ankle and TTC arthrodesis procedures. The periosteal tissues are elevated directly off the bone overlying the distal tibia and talus with soft tissue exposure limited to only what is needed to visualize the ankle and/or subtalar joints along with the medial and lateral gutters. Once soft tissue dissection is completed, internal fixation that will cause an impediment to revision arthrodesis is extirpated19,20 after which the ankle and/or subtalar joints are resected until only healthy, bleeding cancellous bone substrate remains. The residual osseous defects can be much more extensive than expected preoperatively. In addition to bone marrow aspirate harvested from the lower extremity,21 liberal use of structural corticocancellous bulk allograft, specifically femoral head allograft, when appropriate the RIA 2 system (RIA 2 System-Reamer Irrigator Aspirator, Synthes USA Products, LLC, West Chester, PA), living cellular bone allografts (ViviGen & ViviGen Formable Cellular Bone Matrix, LifeNet Health, Virginia Beach, VA)22–24 and use of a vascularized pedicle fibula onlay bone graft25 is recommended to enhance arthrodesis with the critical osseous defects encountered during revision ankle and TTC arthrodesis procedures (Figures 14.4 and 14.5). Although fixation methods will depend on the osseous defects encountered, the most common rearthrodesis fixation for ankle arthrodesis nonunion are retrograde intramedullary nail fixation (Figures 14.6–14.8)26,27 and plate-and-screw constructs placed anteriorly,28 laterally29 or posteriorly30 (Figures 14.9–14.11). The use of screw fixation alone (Figure 14.12)31 and condylar blade plate (Figure 14.13)32 fixation should be considered of historical interest only due to the high potential for failure secondary to inadequate rigidity and excessive vascular stripping of the bone, respectively.