Open fracture
Age
Smoking
Alcohol
NSAIDs
Nutritional deficiency
Prior radiation
Endocrinopathies
Infection
13.2.1 Mechanical
Instability may cause excessive motion at the fracture site, which encourages stem cells to differentiate into fibroblasts, resulting in the formation of fibrous tissue formation and delayed union or nonunion. Factors that promote instability at the fracture site include bony comminution, inadequate or poor plate fixation, small diameter nails, poorly constructed external fixation constructs, and inadequate bony contact.
Comminuted tibial shaft fractures are typically treated with relative stability techniques including intramedullary nailing, bridge plating, and external fixation. When treating these fractures, it is important for the surgeon to choose the construct based on appropriate stiffness. This can be a difficult task and requires much experience with tibial nonunions and the fixation constructs available. In general, the use of larger diameter nails, stiffer plates, and multi-planer external fixation constructs can decrease the risk of nonunion in comminuted tibial shaft fractures.
Bony contact is also an important factor in providing stability to the fracture. As bony contact decreases, fracture stability will also decrease, predisposing to hypertrophic nonunion.
13.2.2 Biological
There are many factors that contribute to “biologic” nonunions. A common cause includes poor blood supply to the fracture often secondary to the soft tissue injury, surgical technique, or a combination of the two. Bishop et al. [17] described patient-related contributors of nonunions and included medical comorbidities, advancing age, smoking, alcohol abuse, nonsteroidal anti-inflammatories, nutritional deficiency, prior radiation treatment, genetic disorders, and various metabolic diseases.
13.2.2.1 Open Fracture
The tibial shaft maintains a subcutaneous anatomical location for a substantial portion of its length especially along its medial border. Poor soft tissue coverage of the tibia has long been associated with a higher incidence of open fractures, which have a higher likelihood of nonunion. Rosenthal et al. retrospectively reviewed 104 open tibial fractures for the relationship between initial wound presentation and potential for healing. Records were analyzed for some 71 patients: all Gustilo type I fractures united, two patients in type II continued to nonunion, and 13 patients in the type III fracture classification went on to nonunion. The authors concluded that there was a strong association between fractures that suffered nonunions and extensive soft tissue loss [18].
13.2.2.2 Smoking
While surgeons cannot always choose their patients in trauma, care should be taken in selecting patients for elective nonunion surgery. Cigarette smoking and nicotine have been implicated in inhibiting fracture healing and increasing the risk of delayed union or nonunion [19, 20]. The effects of smoking are related to its inhibitory effects on the formation of fibroblast-rich granulation tissue leading to impaired healing [21]. Smoke inhalation leads to low concentration of antioxidant vitamins and reactive oxygen species that cause cellular damage, particularly in osteoblasts, fibroblasts, and macrophages. Nicotine has been shown to increase platelet aggregation, to inhibit fibroblast function, and to decrease blood flow to extremities due to increased peripheral vasoconstriction [22].
A large number of studies document the effects of smoking and nicotine in various animal models. In the rabbit model, Donigan et al. studied the effects of transdermal nicotine on fracture healing in 22 mid-shaft tibial osteotomies treated with plate fixation. They noted that, although the nicotine-treated rabbits had similar areas of periosteal callus formation, these rabbits had significantly less torsional resistance and stiffness at 21 days postoperatively and three rabbits had gross nonunions [23]. Similar results reported by Raikin et al. [24] showed that nicotine-exposed rabbits had tibial healing that was 26% weaker resistant to three-point bending than those not exposed. In humans, the majority of research confirmed similar associations as that of the animal models; however, this has been mostly retrospective reviews rather than the understandably difficult prospective, randomized study. Castillo et al. as part of their prospective lower extremity assessment project of 268 tibial fracture patients revealed that current smokers had a higher incidence of nonunion at 24 months after injury compared to nonsmokers (24.1 vs. 9.9%, respectively). Smokers were also more than twice as likely to develop infection and 3.7 times as likely to develop osteomyelitis [25]. In a retrospective study, Adams et al. [26] compared 140 smoking and 133 nonsmoking patients. Mean time to union was 32 weeks compared to 28 weeks, respectively. Clearly, there is an association with smoking and delayed fracture healing, but further research is necessary to identify the exact molecular pathway and possible therapeutics to counteract its effects. Prior to any surgical intervention, smoking cessation should be emphasized to enhance the likelihood of healing. Urine and/or blood screening for nicotine and cotinine can be used to confirm patient’s smoking status. Clinical experience has shown that blood levels of nicotine will return to normal within 2 weeks of cessation while urine will be positive for several weeks.
13.2.2.3 Medications
Nonsteroidal anti-inflammatory drugs (NSAIDs) are commonly cited in the literature as being associated with delayed unions and nonunions, while controversy remains regarding their effect on fracture repair [27]. The exact biochemical pathway is an area of further research, but many authors have theorized that these medications inhibit cyclooxygenase leading to less prostaglandin E2, which leads to less bone formation by osteoblasts [28, 29]. Zhang et al. [30] proposed a schematic model for cyclooxygenase-2, (COX-2) effects on bone repair after fracture using COX knock out mice, whereby decreased levels lead to decreased production of prostaglandin E2, which may lead to low levels of (bone morphogenic) protein. Simon and O’Connor expanded on this murine model and administered various doses of celecoxib, a selective COX-2 inhibitor, to explore the dose-dependent and time-dependent effects of this NSAID. The authors found impaired healing with increasing dosage radiographically, in torsional stability, and overall increased formation of nonunion [31]. Giannoudis et al. [32] retrospectively reviewed the effects of NSAIDs on femoral nonunions in 32 patients and noted a strong correlation. While this association has not been proven definitively in humans with a prospective randomized control trial, caution should be used when prescribing NSAIDs in the setting of tibial fractures, especially in those patients with impaired healing, e.g., smokers, diabetics, etc.
13.2.2.4 Endocrinopathies
Patients who present with a tibial nonunion without an obvious cause should be worked up for an endocrinopathy. Vitamin D, vitamin C, calcium, thyroid hormone, and parathyroid hormone abnormalities have all been implicated in the formation of nonunions. Brinker et al. [33] analyzed 37 prescreened nonunion patients with the hypothesis that these idiopathic nonunions identified actually had underlying endocrine and metabolic abnormalities. They found that 83.8% of the 37 patients had some type of endocrinopathy with the most common being vitamin D deficiency. These authors proposed a diagnostic algorithm for identifying these patients for further workup by an endocrinologist as part of their study. Additional research may further elucidate the causal nature of various endocrinopathies and metabolic disorders and their relationship to nonunions, as well as potential medical treatments.
13.2.2.5 Infection
Infected tibial nonunions pose a complex clinical problem for surgeons and can lead to significant morbidity. In the setting of tibial fractures, infections are propagated from open wounds or introduced during surgical management. Staphylococcus aureus is the most commonly implicated organism and has been found in 65–70% of patients with long bone infections [34]. On the microscopic level, bacteria will form a biofilm or glycocalyx that significantly inhibits ability of the immune system to clear the infection. This leads to involucrum formation, which is reactive bone, as the body attempts to limit the spread of the infection. Shortly following is sequestrum formation, or necrotic bone, indicating a chronic infection with little ability to heal without intervention.
13.3 Evaluation and Diagnosis
13.3.1 History
Of the utmost importance in defining the scope of the problem is the history of the tibial fracture and prior treatment modalities that have failed to obtain fracture union. This includes mechanism of injury, prior open wounds, pain with weight bearing, feelings of instability, and any delayed wound healing. Patients who present with tibial nonunions have often had an extensive treatment history at multiple institutions. Previous records, including operative notes, injury and postoperative radiographs, and any pertinent laboratory values, should be obtained from all previously treating physicians. Questions specific to infectious etiology are particularly important, covering wound drainage, prior cellulitis, constitutional symptoms, pertinent cultures/sensitivities, and previous antibiotic treatment regimens. A complete account of the patient’s chronic illnesses is also important and will help to guide treatment algorithms. This should include the patient’s nutritional status, smoking history, constitutional symptoms, and prior history of nonunion.
13.3.2 Physical Exam
The physical examination of all tibial nonunions begins with observation of the lower extremity for prior wounds, surgical incisions, erythema, gross deformity, and the general state of the surrounding soft tissue. Tenderness to palpation about the nonunion site should be noted and gross motion may be found as well. The surgeon should document the limb vascularity, limb lengths, and range of motion of the knee and ankle joints, as contractures may have occurred.
13.3.3 Laboratories
Important laboratory markers in the evaluation of tibial nonunions that can help guide the surgeon’s treatment include erythrocyte sedimentation rate (ESR) , C-reactive protein (CRP) , and white blood cell (WBC) count . Unfortunately, many authors note that negative laboratory markers do not completely rule out indolent infection. In nonunions with reasonable stabilization, laboratory evaluation for metabolic and endocrine disorders should be obtained in consultation with an endocrinologist, as previously discussed. These markers include serum calcium, serum 25-hydroxy-vitamin D, thyroid-stimulating hormone, phosphorus, and alkaline phosphatase levels.
13.3.4 Radiographs
Radiographs on initial presentation should include the standard anteroposterior and lateral of the tibia/fibula to document the characteristics of the nonunion. Forty-degree internal and external oblique views and stress views may also be useful to better characterize the nonunion.
13.3.5 CT/MRI
Computed tomography (CT) and magnetic resonance imaging (MRI) of nonunions are important tools for defining the three dimensional extent of tibial pathology. CT scans can provide useful information regarding the number of cortices that have healed across a tibial fracture site with bridging callus formation. One study of CT scans for the presence of tibial nonunion found 100% sensitivity and 62% specificity [35]. MRI of tibial nonunions may delineate soft tissue infections from underlying osteomyelitis. Osteomyelitis appears as a low signal intensity of T1-weighted images and high signal intensity of T2-weighted sequences (Fig. 13.1a, b).
Fig. 13.1
Demonstration of osteomyelitis on magnetic resonance imaging . T1-weighted image are low signal and appear dark (a) while T2-weighted images show high signal (b). Images courtesy of Animesh Agarwal, MD Animesh Agarwal, MD, Department of Orthopedics, University of Texas Health Science Center, San Antonio, Texas, USA
13.3.6 Nuclear Imaging
Various modalities are available and are used to evaluate infection as an etiology of the nonunion. The most commonly used nuclear medicine scans of nonunions include technetium-99 m, gallium-67 citrate, and indium-111-labeled leukocyte. Madsen recently reported a case report on the use of bone SPECT/CT imaging to detect sequestrum formation in a chronically infected tibial nonunion [36]. Further research is required to determine the clinical applicability of using SPECT/CT in the setting of tibial nonunions.
13.4 Treatment
The most important aspect of treating tibial nonunions is identifying and correcting the underlying cause of the nonunion. This may be a systemic issue, such as an endocrinopathy, or a localized pathology, for example an infection. Once the etiology of the nonunion has been addressed, the surgeon can continue his or her plan to repair the nonunion. Basic treatment modalities include improving fracture mechanics, restoring the local biology of the fracture, providing electrical or ultrasonic stimulation, and various combinations of these modalities. The surgeon should also consider factors such as the patient’s functional level, occupation, and expectations when developing a treatment plan to ensure its eventual success for both the surgeon and the patient. Various treatment modalities will be discussed and the surgeon should select one based on their training, comfort level, and prior experience.
13.4.1 Based on Nonunion Type
13.4.1.1 Hypertrophic
In most cases of hypertrophic nonunion, fracture stabilization is the fundamental management concept and bone grafting is generally not necessary. Hypertrophic tibial nonunions present with callus formation about the fracture ends, leading to the very characteristic flared ends as the bone attempts to unite (Fig. 13.2). These nonunions tend to be well vascularized but are thought to lack the requisite mechanical stability for bone formation. Once bony stability has been restored, motion at the fracture site is decreased and allows for capillary ingrowth with eventual enchondral ossification. Often this case presents as a tibial shaft that had been previously treated with an intramedullary nail. In such a case, exchange nailing with a larger nail can provide the necessary stability to promote bone healing. In cases where the nonunion had been treated nonoperatively or with an unstable external fixator, intramedullary nailing or compression plating is all that is required to obtain fracture healing. Figure 13.3 demonstrates a hypertrophic nonunion due to a lack of mechanical stability. A typical large callus has formed in the attempt to increase fracture stability. The nonunion was treated with exchange intramedullary nailing with a larger diameter nail. No bone grafting or other biological adjunct was used to achieve union.
Fig. 13.2
Hypertrophic nonunion . Broken distal interlock (red arrow) consistent with excessive motion at the fracture site. Note the abundant callus formation which is a hallmark of a hypertrophic nonunion
Fig. 13.3
32-year-old male with a closed tibial shaft fracture (a). Closed intramedullary nailing with a 10 mm nail resulted in a hypertrophic nonunion (b). Exchange nailing was successfully performed using a 11.5 mm nail (c, d)
13.4.1.2 Atrophic
Unlike hypertrophic nonunions, atrophic tibial nonunions present with poor callus formation, indicating little to no attempt at fracture healing (Figs. 13.4 and 13.5). Classically, these nonunions are thought to be poorly vascularized, but recent research has elucidated a more complicated understanding of atrophic nonunions. Matuszewski and Mehta recently described a case report of a 30-year-old patient who sustained a type IIIC tibial shaft fracture initially treated with vascular repair, soft tissue coverage and plating and yet unfortunately progressed to an aseptic, atrophic nonunion. The treating team noticed pallor of the lower extremity and angiogram revealed stenosis of both the anastomosis sites. After angioplasty, the patient planned on further intervention but was delayed secondary to pregnancy. When she returned to clinic 5 months later, 15 months after initial treatment, the fracture site was radiographically healed without further intervention, implying the importance of vascular supply in the setting of atrophic nonunions [37]. Brownlow et al. analyzed 16 rabbits with atrophic nonunions at various time points to document the vascularity at the fracture site compared with controls. The authors found that at 1 week the control fracture sites were vascularized and the experimental fracture sites were nonvascularized, but this difference resolved by 8 and 16 weeks [38].
Fig. 13.4
Classic atrophic nonunion . Notice there is no evidence of bone healing. These types of nonunions typically require a biological stimulus to promote healing
Fig. 13.5
22-year-old male with a type IIIC open tibia (a). Free flap with intramedullary nailing and cement spacer (b) that was eventually bone grafted using a reamer-irrigator/aspirator technique. Nonunion eventually healed with small anteromedial defect (c, d)
Treatment goals should be focused on the underlying etiology of the nonunion, stimulating a healing response, and providing stable fixation if needed. Bone grafting with autogenous graft remains the gold standard but adjuncts such as bone morphogenic protein-2 and parathyroid hormone (PTH) can also be useful. Atrophic nonunions are most likely multifactorial and present an area for further research.
13.4.1.3 Oligotrophic
Oligotrophic nonunions are those that have characteristics of both atrophic and hypertrophic nonunions, as previously discussed. Management options follow those for hypertrophic nonunions as well as examination for possible causes of a biological lack of bony healing.
13.4.1.4 Infected
Infected nonunions of the tibia can be a challenging problem. Multiple surgeries are usually required to get adequate debridement and eventually restore the function of the limb. Patzakis and Zalavras [39] summarized the basic tenets of care, which includes surgical debridement, antibiotics, fracture stabilization , adequate soft tissue coverage, and eventual restoration of bone defects.
Cierny et al. [40] described the basic classification schema of osteomyelitis based on anatomic types, e.g., medullary, superficial, localized, and diffuse, and patient characteristics based on underlying comorbidities (Fig. 13.6). Clinically, these patients present with ongoing pain, erythema, swelling, and possibly a draining sinus. Laboratory markers include ESR, CRP levels, and a WBC count may be elevated and can be used to diagnose and demonstrate clinical improvement after treatment. Computed tomography or MRI are more useful than plain radiographs in identifying affected areas, periosteal reactions, or abscess formations in the preoperative setting. Erdman et al. [41] demonstrated that MRI of patients suspected of having osteomyelitis had a sensitivity of 98% and a specificity of 75%.
Fig. 13.6
Basic classification schema of osteomyelitis based on anatomic types. Modified from Cierny et al. [40], with permission
Intra-operative cultures from the sinus tract, purulent discharge, soft tissue, and curetted/debrided bone are imperative, as they can help determine a proper antibiotic regimen. Perry et al. [42] noted that superficial wound cultures and needle aspirations were insufficient to rule out infection perioperatively.
For the medullary and superficial cases of osteomyelitis (Cierny-Mader type I and II), general consensus treatment includes removal of metal implants and radical debridement of all involved bone and soft tissue. It is important that the surgeon does not sacrifice a thorough debridement for the hope of an easier reconstruction. While reconstructing large soft tissue and bony defects can be difficult, an inadequate debridement will be doomed to failure. For the localized and diffuse cases (Cierny-Mader type III and IV), nonviable bone must be debrided fully and may require reconstruction at a later date. The complicated diffuse cases with extensive bony and soft tissue defects may result in amputation as the only viable treatment option, especially when presenting in patients with severe comorbidities. Amputation versus limb salvage is a clinical judgment based on patient comorbidities, soft tissue defects, bony involvement, neurovascular assessment, and the desires of the patient. In either case, a good support system for the patient is imperative for a successful outcome.
Fracture stability is of the upmost importance in treating infected tibial nonunions . The senior author was taught as a resident, and this still stands true that “an infected STABLE nonunion is better than an infected UNSTABLE nonunion.” Patzakis and Zalavras [39] have similarly recommended retaining the implants in infected nonunions in certain clinical situations, e.g., early diagnosis, known bacterial species, antibiotic sensitive species, etc. Implants that may be colonized or have failed should be replaced with either external fixation or intramedullary nailing; however, plate fixation may also be reasonable in certain settings. Megas et al. [43] treated nine patients with infected tibial nonunions and bone defects of 2–12 cm after intramedullary nailing with Ilizarov external fixation and reported a 100% union rate with a mean external fixation time of 187.4 days. Consideration should also be given to the placement of poly(methyl methacrylate) (PMMA) beads impregnated with heat-stable antibiotics such as tobramycin and vancomycin. Holtom and Patzakis [44] recommended approximately 2.4–4.8 g of tobramycin, or vancomycin, 2–4 g, per 40 g of PMMA cement to achieve local bactericidal conditions.
Adequate soft tissue coverage should also be obtained during the wound debridement of infected tibial nonunions. This is usually accomplished with a rotational muscle flap or a free muscle transfer, depending on the integrity of the local tissues and the size and location of the soft tissue defect. Muscle transfers are particularly useful for providing a new influx of vascular supply, which improves antibiotic dispersal and host immune system, preventing further microbial seeding. Figure 13.7 demonstrates the usefulness of muscle transfers to help eradicate osteomyelitis and heal a nonunion. A thorough debridement of the infected area is vital to a good outcome. Once the surgical debridement is complete, an antibiotic cement depot is inserted and the patient receives intravenous antibiotics tailored to the specific bacteria that is cultured. After 6 weeks of antibiotics, the flap is elevated, the cement spacer is removed, and autogenous bone graft is placed in the defect.
Fig. 13.7
28-year-old female with a Gustilo-type IIIB open tibia and a segmental bone defect. Intramedullary nailing was performed with insertion of an antibiotic impregnated cement spacer (red arrow, a, b). Harvesting of a free gracilis muscle (c). Final coverage with split thickness skin graft (d). The patient would later go on to have removal of the spacer with autogenous bone grafting 6 weeks later. (Images courtesy of Garrett A. Wirth, MD, University of California–Irvine, Orange, California, USA)
13.4.1.5 Nonunion Location
Peri-articular nonunions are a relatively uncommon occurrence but they can be difficult clinical problems to treat. Metaphyseal bone of the peri-articular region is well vascularized which provides the basic components for fracture healing. Nevertheless, nonunions do occur in this area and the soft metaphyseal bone may not provide the best implant fixation. Treatment regimens include fixed-angle plating, external fixation, and at times intramedullary nailing. Harvey et al. [45] reported on 17 proximal and 13 distal tibial nonunions using customized blade plate fixation, with 29 unions and five persistent nonunions after initial blade plate fixation. These authors found that blade plate fixation was able to achieve eventual union in 97.2% of peri-articular fractures. Gardner et al. performed a retrospective clinical study on 16 patients with proximal tibial nonunions treated with deformity correction, bone grafting, and lateral plating. The authors found that all nonunions healed at an average of 4 months, Knee Society function scores improved significantly, and 88% were able to return to their prior activities [46]. Reed and Mormino reviewed functional outcomes after distal tibial metaphyseal nonunion fixation with blade plates. The authors found all 11 patients had healed and AOFAS scores improved from average scores of 29–89. Pain scores also improved from an average preoperative score of 14–36 postoperatively [47]. Alternatively, Richmond et al. [48] reported on 32 patients with distal tibial nonunions treated with intramedullary nailing and noted 91% union rate at an average of 3.5 months as long as there was enough room distally for two interlocking screws.
13.4.1.6 Segmental Defects
Segmental tibial nonunions are a clinically challenging problem to manage, as the body’s natural ability to fill in bony defects is fairly limited [49]. Surgical options include acute shortening with possible future bone lengthening, autologous cancellous bone grafting, vascularized fibula cortical bone graft, and bone transport with an Ilizarov frame or over an intramedullary nail. Although there is no formal consensus on treating segmental tibial nonunions, many surgeons approach them with treatment guided by the size of the defect. Bone loss of less than 2 cm can be effectively treated with autologous bone graft ing. Defects between 2 and 6 cm may be treated with large autologous bone grafting, such as the Masquelet technique, or bone transport. The Masquelet technique involves a two-stage procedure starting with radical debridement and cement spacer placement, which induces an osteoinductive membrane, followed by autologous bone grafting after removal of the cement spacer [50]. This technique can be quite powerful. Our institution has had success with defects up to 9 cm in the tibia and even larger in the femur. Bone defects larger than 6 cm are often treated with bone transport, free vascularized fibular transfer or amputation. The Ilizarov bone transport technique is a useful tool and allows for bifocal or trifocal correction of large segmental defects. Sala et al. [51] reviewed results from 12 patients with post-infectious segmental tibial nonunions treated with Ilizarov bone transport (Taylor Spatial Frame) in a bifocal or trifocal method and noted 100% union rate in an average external fixation time of 418 days (range 300–600 days).
Stafford et al. [49] retrospectively reviewed 19 segmental tibial nonunions treated with the reamer-irrigator/aspirator (RIA) system (Synthes, Paoli, PA, USA) and a two-stage Masquelet technique for bone defects from 1 to 25 cm in length. At the final clinical follow-up at approximately 1 year postoperatively, 17 of the nonunions had achieved clinical union. Kundu et al. reported results of the Huntington’s procedure, a tibialization of the fibula, for bone defects over 6 cm in size in 22 patients. The authors described clinical unions in 21 of the 22 patients with full unprotected weight bearing at an average of 16 months [52]. Figure 13.8 demonstrates the Masquelet technique in a 38-year-old female who presented from an outside institution 6 months out with a draining sinus over her anterior tibia. She underwent a staged procedure with irrigation and removal of all dead and infected bone, placement of an antibiotic spacer, and a free gracilis muscle transfer. She was placed on intravenous antibiotics for 6 weeks. After a 2 week antibiotic holiday, infectious laboratory markers were drawn and were normal. She was taken back to the operating room for removal of the antibiotic spacer and autologous bone grafting using RIA on her ipsalateral femur. A robust pseudomembrane was generated and preserved (see Fig. 13.8d).
