Nonunion of a first MTPJ arthrodesis is a potential complication encountered postoperatively. While rates of nonunion have been documented between 0 and 23% [31–33, 60–62], the most commonly reported and accepted nonunion rate is approximately 10% [14, 19, 20, 29, 30, 46, 54, 58, 63–72]. With advancements in surgical technique and fixation devises, the rate of nonunion has decreased [14]. In a systematic review performed by Roukis [14], 2818 first MTPJ arthrodesis procedures were evaluated to determine if nonunion rates were different than the historical value of 10%. Inclusion criteria included studies utilizing modern osteosynthesis techniques that had a minimum of 30 ft. Exclusion criteria involved studies involving only patients with rheumatoid arthritis or use of structural bone graft [14]. He found lower incidence of nonunion at 5.4% when utilizing modern osteosynthesis techniques, concluding that the historical rate of 10% is inaccurate.

Even with advancements in surgical fixation devises, higher incidence of nonunion can occur without proper joint preparation. Numerous joint preparation techniques such as conical reamers, high-speed burrs, flat cuts, and curettage have been described (Fig. 13.9) [58, 73, 74]. It has been hypothesized that increased bone temperature caused by high levels of friction created by saws, high-speed burrs, and conical reamers may lead to aseptic thermal necrosis and decrease the odds of successful fusion (A16) [75–78]. The use of saline water irrigation, reamer-irrigator-aspirator systems, and even subaquatic reaming has all been shown to dissipate heat to prevent thermal necrosis of bone [75, 79, 80]. Iatrogenic shortening of the first ray from overzealous joint resection can also be a consequence of powered instrumentation. The author’s preferred techniques include manual curettage or cortical reaming without power to a level of raw bleeding bone surface. Further preparation by fish-scaling and drilling of the bone surfaces with a 1.5 mm drill is suggested (Fig. 13.10). While a K-wire is commonly used for fenestration of the joint surface, we believe that thermal necrosis may be induced utilizing this method [81–83]. Likewise, the act of K-wire fenestration may produce a sterile subchondral plug preventing the intent of active bleeding. By using a drill, the fluted tip allows churning of the cancellous bone below the subchondral plate allowing for release of cancellous bone and pluripotent stem cells into the fusion site, believed to aid in the arthrodesis process.

During the process of joint preparation, attention to the position and abnormal anatomy of the sesamoids should be considered. Quite often, the sesamoids are degenerative, hypertrophic, and at times adhered to the first metatarsal head. This pathological condition can prevent adequate bone apposition plantarly, leading to a nonunion of the plantar one-third of the joint or a dorsiflexed malunion of the hallux. Debridement, mobilization, planning, and even removal of the sesamoid may be needed in this situation.

Fixation failure is another important factor when understanding nonunions. Often, we may ask ourselves if the nonunion was caused by hardware failure or hardware failure from the nonunion. To this, we suggest that hardware does not fail on its own, but must have deforming forces driving the process. Motion at the arthrodesis site is the most plausible force causing hardware deformation and ultimately leading to its failure (Figs. 13.11 and 13.12). Therefore, the hardware used to strengthen and protect the surgical site must be able to withstand these forces.

Mini-external fixation has long been accepted as a possible fixation construct for first MTPJ arthrodesis. Caution should be exercised when using this approach due to the high incidence of malunion and nonunion. In the author’s opinion, proper positioning of the half-pins is paramount for a successful outcome. Due to the complex nature of the joint anatomy, compression at the fusion site can lead to abnormal anatomical alignment of the hallux if perfect positioning of the half-pins has not been achieved. The technical difficulty of placing half-pins correctly, while at the same time visualizing the three dimensional spacial orientation of the first MTPJ during compression through the fixator is what we believe leads to a high malunion and nonunion rate (Fig. 13.13).

Certain plate constructs such as a non-conformed plate are not anatomic and require bending prior to placement. When bending of the plate is needed, the required location of this bend is most commonly found at the fusion interface. This is the site where motion is most likely to occur postoperatively. Likewise, this area is undergoing bone remodeling and revascularization and is most vulnerable to deforming forces. The authors have seen this fixation type fail most commonly at the bend site and it has even been reported in the literature (Fig. 13.14) [70]. Pre-conformed anatomical plates allow for a rigid construct that can disperse these deforming forces better than other forms of fixation.

As the trend for early to immediate weight bearing continues, a solid fixation construct , such as pre-conformed anatomical plates, is an important consideration in avoiding complications [68]. The ideal plate construct should provide the following: (1) mechanically generated cross joint compression through the plate, (2) lobe contour and screw positioning that resists rotation in all four planes, (3) the ability for a locking component, and (4) length sufficient to dissipate the weight bearing load off of the fusion site to more proximal and distal uninjured bone (Fig. 13.15). Plating that incorporates these four components will provide a fixation construct stable enough to protect the fusion site during the initial phases of bone healing while allowing immediate postoperative weight bearing. We believe that immediate weight bearing decreases the risk of deep venous thrombosis and increases micro-motion at the fusion site promoting the neovascularization necessary for bony union. In patients where the bone quality is poor, osteoporotic, or incapable of accepting screw and plate fixation such as in a patient with rheumatoid arthritis, a 3 crossing K-wire or Steinman pin fixation construct may be a necessary and acceptable fixation method [46, 53, 55, 84, 85].

Other co-morbidities such as diabetes, deformity type, nicotine use, endocrine disorders, vitaminosis D, and other nutritional deficiencies may lead to nonunion. In a study evaluating 76 diabetic patients who underwent first MTPJ arthrodesis by Anderson et al. [86], those with peripheral neuropathy were found to be at higher risk for complication. In total, 80% of neuropathic patients resulted in nonunion and concluded that although first MTPJ arthrodesis is an effective and beneficial procedure in patients with diabetes, caution must be exercised in this population. Shibuya and colleagues [87] evaluated factors in diabetic patients leading to nonunion, delayed union, and malunion and reported that surgical duration, peripheral neuropathy, and hemoglobin A1c levels greater than seven were all contributors. They noted that peripheral neuropathy had the strongest association with bone healing complications.

Metabolic and endocrine abnormalities should be considered when evaluating patients with a nonunion. Vitamin D has been shown to play an essential role in fracture healing, specifically mineralization, which is a key component during hard callus formation and bone remodeling. Although not specific to nonunions of first MTPJ arthrodesis, vitaminosis D has been associated with bony nonunions [88–92].

Brinker et al. [92] evaluated a series of patients who experienced an unexplained nonunion without technical error, poor reduction , or other obvious etiology. They found that of the 37 patients that met their criteria, 31 of them had one or more newly diagnosed metabolic or endocrine abnormalities. Vitaminosis D was present in 68% of these patients. Other diagnosed abnormalities included central hypogonadism, thyroid disorders, calcium imbalances, and parathyroid hormone disorders. Of these patients, eight went on to achieve bony union with medical treatment alone. We recommend routine screening and coordinated care with an endocrinologist in this patient population. We have also observed that regional factors such as areas with less daylight may be a contributing factor in vitaminosis D.

The effects of smoking on musculoskeletal healing are well documented. Postoperative complications including wound healing, increased infections, delayed unions and nonunions have been shown to be associated with smoking [93–97]. Smoking cessation of 3–4 weeks prior to surgery has been shown to effectively reduce the overall postoperative complication rate substantially [96, 97].

Effect of underlying first MTPJ pathology has also been discussed as a possible influence on union rates. Korim and Allen [67] observed an 8.2% nonunion rate in 134 patients. 100% of the nonunions reported were found in patients with hallux valgus. They suggested that fixation may have difficulty resisting the deforming forces of the adductor hallucis, flexor hallucis brevis, and extensor hallucis longus. They also noted that bone in patients with hallux rigidus is often more sclerotic and may offer a stronger fixation construct and screw purchase. Grimes and Coughlin [30] reported a 12% nonunion rate for first MTPJ arthrodesis for failed hallux valgus surgery. Of the four nonunions, only one was symptomatic and underwent successful revision.

Severity of the hallux valgus deformity provides for difficult rectus positioning with increased strain on the fixation due to soft tissue adaptation. Significant deformity often requires resection of bone versus curettage. By doing so, stable fixation becomes more difficult due to the amount of hard subchondral bone resected which is required for adequate screw purchase. This may be less of a concern when plating is utilized as opposed to crossed screw fixation.

Ruling out underlying osteomyelitis is important before considering revision (Fig. 13.16). There should always be a high suspicion of such an underlying process. Radiographic evidence of lucency and erosions at the nonunion site coupled with clinical suspicion requires appropriate lab work and ultimately a bone biopsy. Computer assisted tomography and white blood cell labeled bone scans can assist in the diagnosis. If osteomyelitis is found, antibiotic therapy, hyperbaric oxygen therapy, resection of infected bone, or amputation may be indicated.

Nonunion of the first MTPJ can either present as being symptomatic or asymptomatic. The authors encounter asymptomatic nonunions more often than symptomatic. This clinical observation is also documented by others in the literature. In a systematic review, 50 out of 153 (32.7%) nonunions were reported as symptomatic [14]. A pseudoarthrosis formation of the first MTPJ and breakage of internal fixation is sometimes well tolerated without further intervention needed (Fig. 13.17) [46, 70]. When symptomatic, conservative measures such as prolonged immobilization and the use of a bone stimulator are limited. These patients ultimately will require revisional surgery. The patient’s activity level, job demands, and expectations are all factors that need careful consideration when choosing a procedure. One of the most difficult decisions for a surgeon in managing a nonunion is determining when to surgically revise. Sometimes just a little time and patient education is all it may take for bony union to occur.

Identifying the underlying factors which led to a nonunion of the first MTPJ is imperative when determining revision surgical options for these patients. Identifiable co-morbidities such as a history of smoking, uncontrolled diabetes, peripheral neuropathy, nutritional deficiencies, and chronic steroid use have an increased likelihood to fail a revisional arthrodesis. This population may benefit more from removal of hardware and joint debridement with or without implant arthroplasty. Hope and colleagues [98] reviewed their treatment of 12 symptomatic nonunions. Eleven of twelve patients elected to undergo removal of hardware and joint debridement. Four of these patients went onto painful pseudoarthrosis and required either revision fusion or first MTPJ replacement. They concluded that removal of hardware and joint debridement alone is a reasonable option to offer as a relatively minor procedure following failed arthrodesis. It is important to discuss preoperatively with the patient that this surgical procedure often leaves a short, non-functional, floppy toe, with the primary goal being relief of pain (Fig. 13.18).

In a retrospective chart review performed by Thun and colleagues [99], 77 consecutive first MTPJ arthrodesis procedures were divided into two groups consisting of 34 K-wire and 43 modern osteosynthesis fixation techniques with a minimum follow-up of 10 months. Modern osteosynthesis was defined as compression screws, dorsal plating, or dorsal plate with compression screw fixation. All other fixation methods were excluded from this study. They hypothesized that the nonunion rate would be lower in fusions utilizing modern osteosynthesis techniques. The overall nonunion rate was 6.5%. The K-wire group nonunion and delayed union rate was 11.76% and 5.88%, respectively. Modern osteosynthesis nonunion and delayed union rates were both 2.33%. Co-morbidities identified prior to surgery were diabetes mellitus (14.2%), smoking (10.4%), and previous surgery (13%). Of the five nonunions, two (40%) were smokers and three (60%) were diabetic with peripheral neuropathy. One of the patients was both a smoker and diabetic. Fusion was achieved in all patients with a previous surgery. Two of the five (40%) nonunions were symptomatic requiring a second surgery (one revisional first MTPJ fusion with autogenic bone graft and one hardware removal with debridement) resulting in resolution of pain. Of interesting note, 75% of the nonunions found in the K-wire group were asymptomatic (Fig. 13.19). This may be in part due to the fact that these patients did not have any retained hardware postoperatively. As a failed arthrodesis results in pseudoarthrosis, motion at the first MTPJ occurs. These patients often experience irritation provoked by the retained hardware. Similar to Hope’s [98] findings, it should come as no surprise that removal of hardware and debridement alone may be a reasonable solution for some patients where relief of pain is the primary goal. There was only one nonunion within the modern osteosynthesis group. This nonunion was painful and required removal of hardware and debridement. Resolution of pain was achieved postoperatively (Fig. 13.20).