Nonunion After Ankle Arthrodesis: A Contemporary Review

Introduction

Ankle arthrodesis is frequently performed for ankle arthritis, offering durable pain relief and functional improvement when conservative measures fail. ¹ Despite its long-standing role in orthopaedic practice, nonunion remains a complication, with reported rates in tibiotalar arthrodesis ranging from 1% to 16%. ² Nonunion not only compromises the surgical outcome but also imposes a substantial burden on patient quality of life, often necessitating reoperation and prolonged rehabilitation. ³

Nonunion after ankle arthrodesis is a multifactorial complication influenced by several known risk factors. Risk factors such as smoking, diabetes, obesity, peripheral vascular disease, prior infection, avascular necrosis, and suboptimal mechanical fixation have been consistently reported in the literature.^4,5 Although these factors are well established, the management of high-risk patients remains challenging, and strategies to promote union in compromised biological environments are of increasing interest. Nonunion following ankle arthrodesis may be defined radiographically, clinically, or through a combination of both. Radiographic nonunion typically refers to the absence of progressive osseous bridging across the fusion site on serial imaging, whereas clinical nonunion is characterized by persistent pain, instability, or functional limitation attributable to lack of fusion. Importantly, radiographic findings do not always correlate with clinical outcomes. Lause et al ⁶ demonstrated that patients without early radiographic bridging following ankle arthrodesis may still progress to clinical success over time, suggesting that radiographic nonunion alone should be interpreted cautiously when guiding management decisions. This distinction is critical when evaluating fusion status and determining the need for intervention.

Biologic augmentation has emerged as a potential adjunct to improve fusion rates, particularly in patients with impaired bone healing potential. In foot and ankle arthrodesis, available data are heterogeneous and limited by small cohorts and variable outcome definitions, with most evidence in select orthobiologics (eg, bone marrow aspirate concentrate [BMAC] / demineralized bone matrix [DBM], recombinant human platelet-derived growth factor-BB with beta-tricalcium phosphate [rhPDGF-BB/β-TCP], and rhBMP-2 in higher-risk settings). ⁷

More recently, attention has turned toward next-generation regenerative strategies, including the use of extracellular vesicles (EVs)—nano-sized lipid-bound particles secreted by cells that carry signaling molecules capable of modulating inflammation, angiogenesis, and osteogenesis. Although research on EVs in ankle fusion is currently limited, early preclinical data from spine and long bone models suggest that these vesicles may play a role in orchestrating the bone healing cascade and represent a promising avenue for future therapeutic exploration.^8,9

This review aims to provide a synthesis of current knowledge on nonunion after ankle arthrodesis, with a focus on understanding nonunion along a clinical continuum. Specifically, we will outline established and emerging risk factors for nonunion, evaluate current evidence on biologic augmentation strategies including bone morphogenetic proteins (BMPs), platelet-rich plasma (PRP), and BMAC, and explore the future potential of novel therapies such as extracellular vesicles in enhancing fusion outcomes. By integrating these perspectives, we hope to highlight gaps in the existing literature and identify promising directions for future research to improve patient outcomes.

Background on Ankle Arthrodesis

Ankle arthrodesis, commonly referred to as ankle fusion, is a surgical procedure that aims to relieve pain and restore stability in patients with severe ankle arthritis, deformity, or post-traumatic joint deterioration. There are various techniques for performing an ankle arthrodesis. The traditional method has been an open procedure, which allows for full visualization and precise preparation of the joint surfaces. Fixation is typically achieved with hardware such as large, cannulated screws or plates, or external fixation devices, depending on the surgeon’s preference and the patient’s specific pathology and indications. ¹⁰

An increasingly popular alternative is arthroscopic ankle arthrodesis, which is a minimally invasive approach that uses small portals and an arthroscope to perform the fusion. Some studies have reported that arthroscopic fusion may offer advantages over open fusion, including less postoperative pain, faster recovery, and reduced risk of wound complications, which could be particularly beneficial in diabetic or immunocompromised patients. ¹¹ However, these benefits may vary depending on patient selection and surgical expertise. Moreover, arthroscopic fusion may be technically challenging and less effective in cases with significant deformity or bone loss.

The choice of fixation method is another important consideration in ankle arthrodesis. Internal fixation remains the standard and typically involves the use of cannulated screws placed across the tibiotalar joint.^12,13 In more complex cases, especially those involving bone loss or poor bone quality, anterior and lateral plating systems have been developed and are often used in patients undergoing revision arthrodesis or those with compromised posterior soft tissues. ¹⁴ By providing rigid stability and optimizing load distribution, these fixation constructs can help create a more favorable mechanical environment for fusion, thereby reducing the risk of nonunion. ¹⁵ In patients with severe soft tissue compromise or infection, external fixation is employed.^15,16 External fixation is also advantageous in correcting multiplanar deformities and managing complex cases such as Charcot neuroarthropathy. ¹⁶

In certain clinical scenarios, adjunct techniques are used to enhance the likelihood of successful fusion. Bone grafting is commonly employed in the presence of bone defects or a perceived risk of nonunion. Autografts from the iliac crest are often considered the standard because of their osteogenic potential, whereas allografts or demineralized bone matrix may be used in patients who are poor candidates for autograft harvesting. ¹⁷ Moreover, the use of orthobiologics, such as BMPs and PRP, has been explored in spine surgery to stimulate bone healing, although their routine use remains controversial due to variability in outcomes and cost. ¹⁸

BMPs are osteoinductive growth factors that promote mesenchymal stem cell differentiation into osteoblasts and enhance new bone formation. ¹⁹ They are often used as adjuncts in cases of delayed union or nonunion, particularly when conventional grafting options are limited. However, their efficacy can vary based on patient factors, delivery method, and defect characteristics, whereas concerns remain regarding cost and potential complications. ²⁰ When biologic enhancement alone is insufficient, such as in revision cases or those with substantial bone loss, more advanced reconstructive strategies may be required. In such cases, structural grafts or custom-made 3D-printed implants may be used to restore alignment and provide a scaffold for fusion. ²¹ An overview of graft categories and their biologic properties is shown in Table 1.

Table 1.

Pharmacologic and Adjunctive Stimulation Strategies for Ankle Arthrodesis.

Strategy	Fusion Rates	Indications	Key Components	Advantages	Limitations	References
Teriparatide (PTH analog)	Improved fusion in revision/osteoporotic patients	Osteoporosis, revision nonunion	Recombinant PTH (1-34)	Enhances callus and turnover	Cost; daily injections; off-label	Bolia et al 7
Implantable electrical stimulators	86% consolidation in high-risk patients	High-risk primary/revision	Direct current applied via electrodes	Stimulates osteogenesis; effective in compromised bone	Invasive; hardware complications	Saxena et al 64
PEMF (pulsed electromagnetic field)	Prospective hindfoot: faster union, no nonunions (Dhawan); delayed union: 26% (Saltzman; incl. 7 ankles)	High-risk patients; salvage for delayed healing	Noninvasive electromagnetic coil stimulation	Enhances osteoblast activity + vascularity	Less effective in salvage; ankle-specific outcomes limited	Dhawan et al, 65 Saltzman et al 66
Gene therapy (BMPs, LMP-1, NELL-1 vectors)	Preclinical spine: enhanced union	Investigational, potential ankle use	Viral/nonviral vectors encoding osteogenic proteins	Sustained local growth factor expression	No clinical ankle use; regulatory barriers	Beederman et al 19
Extracellular vesicles (EVs)	Preclinical: improved bone repair (osteogenesis + angiogenesis)	Emerging, future therapy	Nano-vesicles carrying miRNAs, BMP-2, VEGF, TIMPs	Cell-free; modulate immune + vascular pathways	No ankle data yet; production standardization needed	Phinney and Pittenger, 8 Watanabe et al, 72 Davies 83

Abbreviations: BMP, bone morphogenetic protein; miRNA, micro-RNAs; PTH, parathyroid hormone; TIMP, tissue inhibitor of metalloproteinase; VEGF, vascular endothelial growth factor.

Risk Factors for Nonunion

When faced with a potential nonunion, clinicians must first confirm the diagnosis and evaluate for such factors as infection and poor glycemic control. A proposed diagnostic workflow for suspected nonunion after ankle arthrodesis is shown in Figure 1. Nonunion remains a relevant complication following ankle arthrodesis, with 3 systematic reviews reporting rates ranging from 0% to 41%, depending on surgical technique, patient comorbidities, and postoperative care strategies.^22-24 Understanding the multifactorial nature of nonunion is essential for surgeons seeking to optimize outcomes. A variety of patient-related, surgical, and biological factors have been identified that influence the risk of nonunion, and recent literature continues to refine our understanding of these associations. ²⁵

Figure 1.

Proposed diagnostic workflow for suspected nonunion after ankle arthrodesis in patients with persistent pain. Initial history and examination guide imaging with ankle radiographs (AP, lateral, mortise), with CT used as needed to assess osseous bridging. Laboratory tests (eg, ESR, CRP, CBC, HbA1c, vitamin D) help evaluate infection and modifiable host factors, and patient-specific considerations inform treatment planning. AP, anteroposterior; CBC, complete blood count; CRP, C-reactive protein; CT, computed tomography; ESR, erythrocyte sedimentation rate; HBA_1c, glycated hemoglobin.

Patient-Related Factors

Several intrinsic patient characteristics have consistently emerged as critical determinants of nonunion following ankle arthrodesis. Smoking is one of the most well-established risk factors, with numerous studies demonstrating impaired osteogenesis and vascularity at the fusion site in smokers. ²² Nicotine and other tobacco-related substances are known to inhibit osteoblast function and reduce peripheral blood flow, thus compromising bone healing. Reported nonunion rates in smokers undergoing ankle arthrodesis are associated with a substantially increased risk of nonunion. In the most recent systematic review and meta-analysis focused on ankle arthrodesis, smoking demonstrated strong evidence as a predictor of nonunion, with a pooled odds ratio of 2.89 (95% CI: 1.23-6.76). ²² Individual studies within that meta-analysis reported even larger effect sizes, including Cobb et al ²⁶ (OR 16.0, 95% CI 2.16-118.27) and Fragomen et al ²⁷ (OR 10.21, 95% CI 2.75-37.95).

Nonunion risk may also be linked to obesity.^28,29 Increased mechanical loading, a poor soft tissue envelope, and impaired metaphyseal cancellous bone quality in obese patients can impair the healing environment. ³⁰ In a large propensity score–matched Medicare database analysis of 7756 patients undergoing ankle arthrodesis, Kamalapathy et al ³¹ found that neither obesity (BMI 30-40) nor morbid obesity (BMI > 40) was independently associated with increased nonunion. Reported nonunion rates were comparable across BMI cohorts and ranged from 5.5% in normal-weight patients to 6.7% in morbidly obese patients (P > .05). ³¹ However, a study by Shah et al ³² found that beyond smoking and diabetes, obesity and cardiovascular disease were not significantly associated with increased nonunion risk. In their cohort of 87 high-risk patients undergoing primary ankle arthrodesis, nonunion rates were similar between the open and arthroscopic groups (11% vs 12%, respectively). They found that obesity was common in both groups, but after analysis obesity was not statistically associated with nonunion, consistent with their conclusion that smoking and diabetes remain more reliable predictors.

Diabetes mellitus is a particularly concerning comorbidity in foot and ankle surgery. In a recent population-based case-control study, Chiang et al ³³ demonstrated a significantly increased risk of nonunion in diabetic patients undergoing arthrodesis of the ankle, tarsometatarsal, metatarsophalangeal, toe phalangeal, and tarsal joints, with diabetes associated with a 1.71-fold increase in nonunion risk. The authors attributed this to impaired microcirculation, delayed inflammatory responses, and altered collagen synthesis in diabetic tissue. The pro-inflammatory state associated with diabetes can prolong the inflammatory phase of healing and delay the transition to the reparative phase. ³⁴ Additionally, diabetic microvascular disease limits oxygen and nutrient delivery to the fusion site, further compromising bone regeneration. ³⁵ Peripheral neuropathy, common in long-standing diabetes, may lead to abnormal gait mechanics and uneven load distribution across the arthrodesis site, increasing mechanical strain on developing fusion tissue. ³⁶ Poor glycemic control has also been correlated with higher rates of surgical site infection and slower radiographic evidence of union in orthopaedic procedures, compounding the challenge of achieving a solid ankle fusion.^37-39

These challenges are particularly pronounced in patients with diabetes-related Charcot neuroarthropathy (CN). In a multicenter review, patients with CN who underwent ankle arthrodesis had higher postoperative complication rates compared with non-Charcot patients, including surgical site infection, hardware removal, and early amputation. ⁴⁰ Within the diabetic population, CN was associated with a greater than 3-fold increase in the odds of amputation (OR 3.43) compared with diabetes without CN, whereas diabetes without CN was associated with increased odds of amputation compared with patients without diabetes (OR 2.26). ⁴⁰

Shibuya et al ⁴¹ conducted a retrospective cohort study of 165 diabetic patients undergoing foot or ankle arthrodesis, osteotomy, or fracture fixation. They found that a history of foot ulcer, peripheral neuropathy, and prolonged surgical duration were significantly associated with postoperative bone healing complications on bivariate analysis. After multivariate adjustment, only peripheral neuropathy, surgical duration, and hemoglobin A1c >7% remained independent predictors. Peripheral neuropathy showed the strongest association, likely due to the combined effects of altered biomechanics and impaired microvascular supply on bone healing.

Poor bone quality, either due to osteoporosis or osteopenia, represents an important patient-specific risk factor that can hinder fusion success. Shah et al ³² emphasized this in their review of high-risk cohorts undergoing ankle arthrodesis, noting that reduced bone stock may compromise optimal joint surface contact. They also observed that, despite attempts to augment fusion with adjuncts such as DBM or PRP, outcomes remained suboptimal in this subgroup.

Finally, advanced age has been variably reported as a risk factor for nonunion. A recent survey of orthopaedic surgeons indicated that patient age is generally perceived as a relatively minor risk factor for nonunion after foot and ankle arthrodesis, despite biologic plausibility for age-related declines in autologous graft quality. However, the results from a large multicenter randomized clinical trial challenge this perception. ⁴² This trial analyzed 397 patients (597 joints) undergoing hindfoot or ankle arthrodesis supplemented with either autograft or an osteoinductive autograft alternative, rhPDGF-BB/β-TCP. In the autograft cohort, patients younger than 60 or 65 years had more than twice the odds of achieving successful fusion compared with older patients, whereas no significant differences were observed at the 55-year threshold. In contrast, rhPDGF-BB/β-TCP demonstrated consistent fusion success across all age thresholds, with approximately double the odds of union compared with autograft in patients older than 60 years. ⁴³ This aligns with earlier orthopaedic literature, which has suggested that bone healing capacity diminishes with age because of lower mesenchymal stem cell availability and impaired angiogenesis. ⁴⁴

Technical Factors

Surgical technique and the precision of execution play pivotal roles in achieving successful arthrodesis. In particular, the quality of joint surface preparation and the adequacy of compression across the fusion site are fundamental to success. ⁴⁵ A recent multicenter retrospective cohort study by Takahashi et al ³ highlighted the importance of radiographic alignment and bony apposition. Their study of 154 patients undergoing arthroscopic ankle arthrodesis revealed that a larger postoperative tibial bony gap was significantly associated with nonunion, with mean bony gap measurements ranging from 1.65 to 3.01 mm in the nonunion group compared with 1.03 to 2.03 mm in the union group, and an overall nonunion rate of 9.0% (P < .05).In addition to joint preparation, implant construct characteristics such as the use of anterior plate augmentation to enhance fixation rigidity and load distribution may influence fusion outcomes, with comparative clinical data demonstrating a trend toward lower nonunion and revision rates compared with compression screw–only constructs. ⁴⁶ Thus, constructs that provide rigid compression and rotational stability may optimize the mechanical environment for fusion, whereas inadequate fixation or suboptimal implant configuration can contribute to micromotion and increase the risk of nonunion. ¹⁴

The surgical approach can also influence fusion outcomes. Although both open and arthroscopic techniques are used for ankle arthrodesis, the relative impact of these techniques on nonunion remains debated. Some studies suggest that arthroscopic approaches may yield higher fusion rates because of less soft tissue disruption and better preservation of blood supply. A systematic review and meta-analysis that compared open and arthroscopic ankle arthrodesis demonstrated that arthroscopic fusion was associated with significantly higher union rates, whereas open techniques showed substantially lower odds of fusion (OR 0.26, 95% CI 0.13-0.52; P = .0002). ⁴⁷ This may suggest that arthroscopic approaches possibly confer 3- to 4-fold higher odds of achieving successful fusion compared with open surgery. However, complications can still arise when joint surfaces are inadequately prepared or insufficient compression is achieved. In their retrospective analysis, Collman et al ⁴⁸ found no significant difference in nonunion rates when DBM was used in arthroscopic ankle arthrodesis, further complicating the question of biologic augmentation efficacy in technical cases where fusion is mechanically compromised.

Postoperative compliance and weight-bearing protocols are also important considerations following ankle arthrodesis. Mechanical loading plays a complex role in bone healing, as moderate, controlled loading has been shown to stimulate osteogenesis, whereas excessive or delayed loading may impair fusion. Experimental and clinical data suggest that early, protected weight-bearing can promote callus formation and improve healing biology. A systematic review by Potter and Freeman ⁴⁹ found no clear increase in nonunion rates with early weight-bearing following ankle arthrodesis in appropriately selected patients, although protocols varied widely across studies. This speaks to the importance of postoperative rehabilitation strategies that balance mechanical stimulation with construct stability and patient-specific risk factors.

Prior Infection, Avascular Necrosis, and Open Injury

A history of prior infection or open injury presents another layer of complexity in ankle arthrodesis procedures. Patel et al, ²² in their systematic review and meta-analysis, identified prior operative site infection and preoperative avascular necrosis as moderate to strong predictors of nonunion. These factors contribute to poor local vascularity, persistent inflammation, and a challenging biological environment for healing. Open injuries, in particular, are known to disrupt periosteal integrity and are often associated with soft tissue compromise, further hampering the regenerative process.

The challenges presented by avascular necrosis are not limited to impaired vascularity but also include decreased bone turnover and remodeling capacity. These factors can impair graft incorporation and fusion, necessitating more aggressive surgical debridement or structural grafting strategies in such cases.^22,50 Moreover, patients with a history of infection may also be at higher risk for biofilm-related complications and require longer durations of antibiotic therapy, which can indirectly affect bone healing dynamics.

Augmentation Strategies

Contemporary strategies have evolved to include a broad array of orthobiologic materials, synthetic graft substitutes, pharmacologic agents, and mechanical augmentation techniques. The optimal approach is multifactorial and should be tailored to patient-specific variables, including comorbidities, bone quality, and surgical complexity. It is important to distinguish outcomes reported for primary ankle arthrodesis from those achieved in revision or salvage settings. Fusion rates following revision arthrodesis are generally lower and may reflect compromised local biology, altered biomechanics, infection history, and prior hardware failure that may have contributed to the initial nonunion. When interpreting the reported efficacy of biologic augmentation strategies, clinicians should recognize that favorable outcomes observed in primary fusion cohorts may not directly translate to revision cases, where host and mechanical factors often limit healing potential.

Autologous Bone Graft

Autologous bone graft remains the historical standard because of its comprehensive osteogenic, osteoinductive, and osteoconductive properties. Bone graft and biologic options used in ankle arthrodesis can be broadly categorized along a spectrum from highly osteogenic constructs to scaffold-only materials (Figure 2). Typically harvested from the iliac crest, tibia, or calcaneus, an autograft continues to demonstrate high union rates ranging from 75% to 100%, with the majority of these data derived from primary ankle arthrodesis cohorts. ⁵¹ However, iliac crest autograft harvesting is associated with several risks, including increased surgical operation time, the need for a separate incision site, and additional donor site morbidity such as pain and risk of infection.^52,53 Logistic regression analyses of more than 5000 cases confirm that structural autografts and cancellous autografts offer among the highest probabilities of union, approximately 94% and 93.7%, respectively, again reflecting predominantly primary fusion populations. ⁵⁴ Interestingly, local autografts harvested from the surgical site have shown outcomes comparable to iliac crest grafts, minimizing complications without sacrificing efficacy. ⁷ Importantly, the osteogenic quantity and biologic activity of autograft vary by harvest site. Axial sources such as the iliac crest demonstrate higher concentrations of osteoprogenitor cells and colony-forming units compared with peripheral harvest sites, including the tibia and calcaneus, which may exhibit reduced cellular activity. ⁵⁵ This variability may be particularly relevant in revision arthrodesis, where compromised host biology and prior surgical insult may necessitate greater osteogenic potency than is typically required in primary fusion.

Figure 2.

Overview of available bone graft methods for arthrodesis. Multiple donor, recombinant, or synthetic materials are available, which vary in osteoinductive and osteoconductive properties. Autologous bone graft, whether local or harvested, has long been the gold standard in arthrodesis surgery. Given limited local autograft and morbidity of graft harvesting, growth factors such as rhBMP-2 are popular osteoinductive options, whereas viable cell allografts and DBM with synthetic additives are often choices that optimize osteogenic or osteoconductive metrics, respectively. Donor allograft is a widely available graft material notable for its osteoconductive properties but lacking osteoinductive potential.

Allografts and Mesenchymal Stem Cell–Based Grafts

Allografts, both structural and particulate, offer a readily available alternative but lack intrinsic osteogenic potential. When combined with autograft or bone marrow aspirate, allografts have demonstrated union rates of 59% to 100% in foot and ankle arthrodesis, depending on formulation and surgical indication. ⁵¹ Viable cellular allografts, which contain mesenchymal stem cells (MSCs) embedded in a matrix, offer enhanced biologic potential and have demonstrated promising results. For instance, Dekker et al ⁵⁶ reported an 83% fusion rate in a high-risk cohort using cellular allografts, whereas Hollawell ⁵⁷ observed a 100% fusion rate at 6 months in patients with diabetes and smoking history using a mesenchymal stem cell–based product. Given the increasing use of biologics and cellular therapies in orthopaedic procedures, similar MSC-based strategies could offer valuable adjuncts to promote union in ankle arthrodesis.

Demineralized Bone Matrix and Bone Marrow Aspirate Concentrate

DBM, composed of collagen, noncollagenous proteins, and growth factors, is frequently used as a graft extender or substitute. Fusion outcomes vary widely depending on the formulation, with reported union rates ranging from 56% to 100%. ⁵¹ DBM appears to be most effective when combined with osteogenic agents, such as BMAC. BMAC, which contains osteoprogenitor cells and a milieu of growth factors, has been shown to improve fusion rates when paired with DBM or synthetic scaffolds. ⁵⁸ In particular, DBM combined with BMAC demonstrated fusion rates as high as 97% in small case series, reflecting synergistic osteoinductive and osteogenic properties. ⁵¹

Synthetic Graft Substitutes and Recombinant Growth Factors

Among synthetic graft alternatives, β-tricalcium phosphate (β-TCP) and calcium sulfate-hydroxyapatite composites have garnered attention for their structural integrity and osteoconductivity. One of the most promising developments in this category is the use of recombinant human platelet-derived growth factor-BB (rhPDGF-BB) in conjunction with β-TCP.^43,51 This combination has demonstrated clinical efficacy in multiple randomized trials. In a large prospective study by Daniels et al, ⁵⁹ patients treated with rhPDGF-BB/β-TCP achieved an 84% fusion rate at 24 weeks, compared with 65% in the autograft control group, with fewer donor site complications.

Bone morphogenetic proteins, especially rhBMP-2, continue to play a role in challenging cases such as revisions or large defects. Although fusion rates as high as 100% have been reported, concerns remain regarding heterotopic ossification, inflammatory responses, and off-label use in the foot and ankle. ⁶⁰ Nevertheless, retrospective studies and limited comparative data suggest rhBMP-2 is effective in enhancing union and may shorten recovery time, particularly in complex reconstructions.^7,61

Emerging Biologics

Additional biologic options under investigation include the osteoinductive peptide B2A and autologous conditioned PRP. Early trials of B2A in foot and ankle fusion demonstrated similar outcomes to autograft, although sample sizes were small and further validation is needed. ⁷ Further research is also needed to determine the efficacy of B2A in ankle arthrodesis specifically. Another biologic of interest is PRP, which is rich in growth factors such as PDGF and transforming growth factor β (TGF-β), and has shown potential to enhance bone healing when used adjunctively. However, primary clinical data supporting the use of PRP in foot and ankle arthrodesis are limited, with only isolated studies evaluating its role as an adjunct in fusion procedures. While definitive clinical evidence remains sparse, early and heterogeneous reports suggest that PRP may offer a modest adjunctive benefit in select high-risk populations when used alongside structural grafts rather than as a standalone biologic strategy. ⁶²

A 2024 systematic review and meta-analysis including 10 studies, 6 of which were randomized controlled trials, evaluated PRP for ankle osteoarthritis and talar osteochondral lesions. ⁶³ Across randomized trials, PRP did not demonstrate a statistically significant overall improvement in AOFAS scores compared with controls (mean difference 4.14 points, 95% CI −0.60 to 8.87; I² = 86%). ⁶³ Subgroup analysis showed greater benefit in talar osteochondral lesions with an AOFAS mean difference of 8.66 points (95% CI 6.61-10.71; I² = 0%). ⁶³ Pain reduction was observed overall with a standardized mean difference of −0.62 (95% CI −1.13 to −0.10; I² = 77%), and the effect was larger in talar cartilage injury with a standardized mean difference of −1.24 (95% CI −1.68 to −0.81; I² = 0%). Importantly, these data reflect symptomatic and functional outcomes rather than fusion or nonunion rates, and significant variability in PRP formulation and dosing was reported. Dedicated studies evaluating PRP in ankle arthrodesis are still required to determine its effect on union.

Pharmacologic agents have also been evaluated for their ability to improve bone healing. Teriparatide, a recombinant parathyroid hormone, has demonstrated success in accelerating fusion in patients with osteoporosis or previous nonunion. Several studies have reported significantly improved union rates in patients treated with teriparatide postoperatively, particularly in the context of revision arthrodesis. ⁷

Pulsed Electromagnetic Field Stimulation and Electrical Bone Stimulation

Mechanical stimulation techniques, such as implantable electrical bone stimulators, offer another dimension to nonunion prevention. These devices deliver low-level electrical currents to the fusion site, stimulating osteogenesis through cellular pathways including calcium ion flux and upregulation of osteogenic genes. In a multicenter study of high-risk patients undergoing arthrodesis, Saxena et al ⁶⁴ demonstrated an 86% radiographic consolidation rate with electrical bone stimulation, even in patients with diabetes, obesity, or prior nonunions. Although complications like cable breakage occurred in a minority of patients, the overall efficacy suggests that electrical stimulation is a viable adjunct for patients with impaired healing potential.

Pulsed electromagnetic field (PEMF) stimulation is another well-studied postoperative treatment strategy aimed at promoting bone healing through the delivery of electromagnetic energy. Although initially approved by the FDA for long bone nonunions, emerging data have supported its utility in elective arthrodesis procedures. In a prospective randomized study of 64 patients undergoing hindfoot arthrodesis, Dhawan et al ⁶⁵ demonstrated that PEMF significantly reduced time to radiographic union across several joints. Specifically, talonavicular and calcaneocuboid fusion times were significantly shorter in the PEMF-treated group compared with controls, and no nonunions were observed in the PEMF cohort. In a retrospective review of 19 cases of delayed union following foot and ankle arthrodesis, Saltzman et al ⁶⁶ reported a lower success rate of 26% achieving union with PEMF. The series included 7 ankle arthrodeses, although the authors did not report outcomes separately by joint, limiting ankle-specific conclusions. The majority required revision surgery with autograft and continued PEMF to achieve fusion. Taken together, these data imply that PEMF may be more effective in the primary fusion setting than as a salvage strategy after delayed union. Pharmacologic and adjunctive strategies, including teriparatide, electrical stimulation, and PEMF, are summarized in Table 2.

Table 2.

Bone Graft and Biologic Augmentation Options for Ankle Arthrodesis.

Strategy	Reported Fusion Rates	Indications	Key Components	Advantages	Limitations	References
Autograft (iliac crest, tibia, calcaneus, local bone)	75%- 100%; logistic regression analysis ~94% structural, 93.7% cancellous	Gold standard, primary and revision cases, bone defects, high-risk nonunion	Living osteogenic cells, growth factors, natural scaffold	Osteogenic + osteoinductive + osteoconductive; proven long-term efficacy	Donor site morbidity, limited volume, increased operative time	Lareau et al, 54 Greer et al 60
Allograft (structural/particulate)	59%-100% when combined with autograft or BMAC	When autograft not feasible; supplement to structural grafting	Mineralized scaffold (osteoconductive only)	Readily available, no donor site morbidity	Lacks intrinsic osteogenic potential; risk of immune reaction	Greer et al 60
Viable cellular allograft (MSC-based)	83% (Dekker); up to 100% in high-risk patients (Hollawell)	High-risk populations (smokers, diabetes), revision fusion	Scaffold + viable mesenchymal stem cells	Provides osteogenic cells in addition to scaffold; avoids donor site morbidity	Higher cost, variable radiographic union times	Dekker et al 56 Hollawell 57
Demineralized bone matrix (DBM)	56%-100%; up to 97% when combined with BMAC	Graft extender; revision or compromised bone quality	Collagen, noncollagenous proteins, residual growth factors	Contains growth factors; synergistic with BMAC	Fusion highly formulation-dependent; weak alone	Greer et al, 60 Harford et al 58
Bone marrow aspirate concentrate (BMAC)	Up to 97% when paired with DBM or scaffolds	High-risk patients, revision settings, compromised biology	Osteoprogenitor cells, growth factors (VEGF, TGF-β, BMPs)	Provides osteoprogenitors + growth factors; minimally invasive harvest	Variable cell yield; requires adjunct scaffold	Harford et al 58
Synthetic scaffolds (β-TCP, Ca sulfate-HA)	Variable alone; up to 100% when combined with BMAC or rhPDGF-BB	Structural support with biologic augmentation	Calcium-based osteoconductive ceramics	Osteoconductive; unlimited supply	Weak union potential if used alone	Loveland, 43 Daniels et al 59
rhPDGF-BB + β-TCP	84% at 24 wk vs 65% autograft; especially effective in older patients	Primary or revision arthrodesis, older/osteoporotic patients	Recombinant PDGF-BB + β-TCP matrix	Angiogenic + osteoinductive; avoids donor morbidity	Cost; not yet widely available in foot/ankle	Daniels et al, 59 Berlet et al 42
rhBMP-2	Up to 100% in select series	Large defects, revision or salvage cases	Recombinant BMP-2 protein + carrier (collagen sponge)	Strong osteoinductive effect; may shorten time to union	Off-label in ankle, risk of HO, inflammatory response, cost	DeVries and Scharer, 61 Greer et al 60
B2A osteoinductive peptide	Early foot/ankle trials show similar to autograft	Investigational; adjunct for fusion	Synthetic peptide upregulating endogenous BMP-2	Osteoinductive potential; synthetic and scalable	Small studies; efficacy in ankle needs confirmation	Bolia et al 7
Platelet-rich plasma (PRP)	Limited data; adjunctive use suggested	High-risk patients, adjunct to structural graft	Autologous plasma enriched in PDGF, TGF-β, VEGF	Growth factor-rich; autologous	Inconsistent protocols; weak evidence in ankle	Bibbo and Hatfield 62

Abbreviations: BMP-2, bone morphogenetic protein-2; MSC, mesenchymal stem cell; PDGF, platelet-derived growth factor; PDGF-BB, platelet-derived growth factor-BB; PRP, platelet-rich plasma; TGF-β, transforming growth factor β; VEGF, vascular endothelial growth factor.

Future Directions: EVs in Bone Regeneration

EVs are nanoscale, membrane-bound particles secreted by MSCs and other skeletal cells that carry proteins, lipids, growth factors, and miRNAs with the capacity to influence inflammation, angiogenesis, and osteogenesis. ⁶⁷ Their cell-free therapeutic profile makes them an emerging option for addressing nonunion and impaired fracture healing.

EVs potentially promote bone repair through multiple mechanisms. They shift macrophages from a pro-inflammatory M1 phenotype toward the pro-healing M2 phenotype, improving the local environment for osteogenesis and angiogenesis.^67,68 They stabilize the extracellular matrix by delivering tissue inhibitors of metalloproteinases (TIMPs), counteracting collagen degradation, and maintaining osteoid formation. ⁶⁷ They also reduce osteoclast activity by carrying soluble TNF receptor I (sTNF-RI), which neutralizes TNF-α and supports osteoblast survival. ⁶⁷ In addition, angiogenesis is enhanced through the delivery of VEGF, fibroblast growth factors (FGFs), and hepatocyte growth factor (HGF), which stimulate neovascularization and limit fibrotic nonunion. ⁶⁷

MSC-derived EVs directly promote osteogenesis by transferring pro-osteogenic miRNAs (eg, miR-29a, miR-335) that activate PI3K/Akt and mitogen-activated protein kinase (MAPK) pathways.^69-73 The regenerative potential of EVs is influenced by donor age and tissue of origin, with bone marrow MSCs and younger donors showing greater osteogenic activity.^74,75 Beyond MSCs, osteoblast-derived EVs stimulate osteogenic differentiation but may also promote osteoclastogenesis via RANKL and miR-143,^74-77 whereas osteoclast-derived EVs exert both inhibitory and anabolic effects depending on their cargo, including SPP1-mediated activation of TGFβ1/Smad3 signaling to enhance MSC osteogenic differentiation.^78-80 Preclinical murine fracture models confirm these effects, with bone marrow MSC-EVs accelerating healing via miR-335 regulation of VapB. ⁸¹ EV cargo has also been shown to engage Wnt/β-catenin, BMP/Smad, and PI3K/Akt pathways, reinforcing their role in osteogenesis.^82,83 Engineering strategies to enhance efficacy include osteogenic preconditioning of MSCs to enrich pro-osteogenic miRNAs (miR-146a-5p, miR-503-5p) and surface modifications such as SDSSD peptide conjugation, which improved bone regeneration in murine tibial defect models.^84,85 Biomaterial delivery systems, including hydrogels and scaffolds, provide localized and sustained release.

Despite these advances, substantial barriers remain, including standardization of EV isolation, variability in biologic potency, immune safety, and cost-effectiveness. Nevertheless, EVs represent a promising biological strategy with potential application in ankle arthrodesis, particularly in patients at high risk of nonunion.^67,83,85

Conclusion

Nonunion following ankle arthrodesis represents a complex, multifactorial challenge that demands a personalized and evidence-based approach. A wide array of patient-related factors, including smoking, obesity, diabetes, and advanced age, interact with surgical technique, joint-specific biomechanics, and the biological milieu to influence fusion outcomes. Although traditional strategies such as autografts remain central to treatment, contemporary advancements in orthobiologics, synthetic scaffolds, and mechanical stimulation have substantially expanded available therapeutic options. With continued refinement and clinical validation, MSC-derived EVs may emerge as a cell-free biologic adjunct to improve fusion outcomes in complex ankle arthrodesis.

Supplemental Material

sj-pdf-1-fai-10.1177_10711007261438446 – Supplemental material for Nonunion After Ankle Arthrodesis: A Contemporary Review

Supplemental material, sj-pdf-1-fai-10.1177_10711007261438446 for Nonunion After Ankle Arthrodesis: A Contemporary Review by Ryan C. Rizk, Kevin Liebmann, Cameron A. Rivera and Solangel Rodriguez-Materon in Foot & Ankle International