Risk Factors for Nonunion in Arthroscopic and Open Ankle Arthrodesis

CATEGORY: Ankle; Ankle Arthritis; Arthroscopy INTRODUCTION/PURPOSE: Ankle arthrodesis remains the most popular surgical treatment option for end-stage ankle arthritis (ESAA) among surgeons in the United States. The primary objective endpoint for judging failure versus success of any arthrodesis is radiographic union versus nonunion. Overall, reported union rates in the last two decades have been excellent; however, there does remain significant variation in results with conflicting evidence regarding both treatment and patient factors that are associated with nonunion. We present a relatively large case series of ankle arthrodeses from a single institution with a high-risk patient population with the goal of further clarifying the patient and treatment factors that lead to nonunion. METHODS: We conducted a retrospective chart review of 118 patients who underwent primary open or arthroscopic ankle arthrodesis at our institution between November 2014 and April 2019. Revision arthrodesis and patients with a history of complex open fracture were excluded. A minimum 6-month postoperative followup was required. The patients were divided into arthroscopic and open arthrodesis cohorts. The primary outcome measure was radiographic union at 6 months. Patient factors including demographics, BMI, medical comorbidities, and smoking status were analyzed as predictors of nonunion. Likewise, treatment factors such as surgical approach, method of fixation, and tourniquet time were analyzed as predictors of nonunion. RESULTS: Of the 43 individuals that underwent arthroscopic ankle arthrodesis seven progressed to nonunion (16.27%). Among those undergoing open ankle arthrodesis 6 patients out of 46 progressed to nonunion (13.04%). In the arthroscopic cohort, individuals with preoperative lower extremity infection had a significantly higher rate of nonunion compared to those without infection (50.00%, p=0.0447). The open group had two significant predictors of nonunion: use of external fixation and low tourniquet time. Individuals who underwent arthrodesis with the use of an external fixator had a 100% nonunion rate compared to 11.11% for those treated with screws and 0% for those treated with plate fixation (p=0.020). Individuals that had a total tourniquet time under 90 minutes had a non-union rate of 66.67% (p=0.0082). CONCLUSION: While it was unsurprising that preoperative infection was a significant risk factor for nonunion, it is interesting that this effect was only shown in the arthroscopic group and not the open group. This could have practice implications and warrants further study. Our findings also add to the body of evidence that external fixation is inferior to modern internal fixation techniques for achieving bony fusion. This result may also reflect the poor preoperative prognosis of those requiring external fixation. There remains little evidence that diabetes, smoking, or BMI are significant risk factors for nonunion in primary ankle arthrodesis.