Secondary malalignment refers to a change in fracture reduction and alignment at some point during the postoperative period. This can usually be attributed to a loss of fixation in poor quality bone, inadequate fixation in combination with an unstable fracture pattern, premature dynamization or noncompliance with appropriate weight-bearing restrictions. Brumback et al. demonstrated in their work on femoral shaft fractures that cases which did not have static interlocking had a 10.5 % rate of major unanticipated malalignment postoperatively [26]. This occurred even in seemingly length stable fracture patterns with >50 % cortical contact. Furthermore, data from statically locked femoral nailing has shown that this technique does not necessarily retard the bony healing process, with a published union rate of 98 % [27]. Lastly, early weight bearing for femoral shaft fractures can be allowed when treated with statically locked nails, provided they are of sufficiently large nail and screw diameter [28, 29].

There is some published data on secondary malalignment seen after tibial intramedullary nailing (Fig. 8.3). Overall this occurs at a low rate. Ricci et al. reported 1 case out of 11 proximal tibia fractures that displaced from 6° of valgus immediately postoperatively to 10° of valgus (9 %) [10]. Lower rates of secondary malalignment have been reported by Josten (5 %) and Nork (0 %), both dealing with closed nailing of proximal tibia fractures [6, 24]. Vallier reported an 11 % rate of secondary malalignment in a study of only distal tibia fractures [11]. The majority of these cases displaced into valgus. At least two distal interlocking screws should be used, as a high rate of failure and loss of reduction has been reported with the use of a single screw [30].

In conclusion, secondary malalignment of fractures treated with intramedullary nailing can be minimized through the use of static interlocking of sufficiently sized nails. This is particularly true in the femur. The use of multiple interlocking screws is advisable as well. Lastly, if bone quality is poor or the fracture pattern has minimal inherent stability, weight bearing restriction or adjunct fixation is an option as well to help prevent loss of reduction and secondary malalignment.

In the setting of an appropriately reduced and fixed fracture that heals, hardware failure is rare [6, 28, 31, 32]. Hardware failure is seen when the implant sustains more load than intended, either due to early excessive weight bearing or prolonged cyclical loading, as in a non-united fracture. Logically, failure will occur at the weakest portion of the construct, which is the interface between the interlocking screws and the nail (Fig. 8.4), although breakage of the nail itself has certainly been observed in rare instances (see also Fig. 6.​5). Biomechanical data has clearly demonstrated that the fatigue strength of an intramedullary nail construct is proportional to the size of the interlocking screws. Griffin et al. performed cyclical loading of tibial intramedullary nails with various sized interlocking screws; they found that the average fatigue life of a single 4.0 mm screw was 1,200 cycles [33]. This corresponds clinically to less than a full day of weight bearing. They found that by increasing the screw size, a 4.5 mm screw had a 50 % probability of withstanding a week of weight bearing and a single 5.0 mm screw had a 90 % probability of withstanding a week of weight bearing. Increasing the number of interlocking screws also correlated with increased load to failure.

Clinical data supports the notion that hardware failure is relatively rare. Small case series have reported no cases of hardware failure for tibial and femoral nailing [6, 32], while others report low rates of mechanical failure at 0.8 % with cephalomedullary nails and 0.3 % in the SPRINT trial [31, 34]. A systematic review published by the Cochrane Database suggested an association between failure of hardware and the use of unreamed nails for tibial fixation (35/789, 4.4 % versus 79/756, 10.4 %, RR 0.42) [35]. This difference may be attributable to a higher rate of delayed/nonunion with the use of unreamed nails, thus placing them at an increased risk of failure. Additionally, they reported that there was a significantly higher risk of implant failure with the use of one distal screw versus the use of 2 distal screws (RR 11.82). This original research was first reported by Kneifel and Buckley [30]. Tibia fractures with 1 distal interlocking screw in their cohort failed at a rate of 59.5 % (13/22) versus 5 % (1/20) in fractures secured with two distal interlocking screws. A similarly high rate of failure (25 %) was reported by Lang et al. with the use of a single proximal interlocking screw for intramedullary nailing of proximal tibia fractures [7].

While early weight bearing with statically locked intramedullary nails is supported in the literature [28, 32], even in the presence of unstable fracture patterns, surgeons must take precautions and exercise good judgment to prevent hardware failure. The abovementioned studies were performed in a population of femur fractures, using large size intramedullary nails (at least 12 mm diameter) and interlocking screws (greater than 5 mm). This caliber of fixation cannot be routinely applied to the tibia, thus they must be considered separately. However, increasing the number, size and type of interlocking fixation can help strengthen the construct and in turn hopefully avoid hardware failure (Fig. 8.5). Biomechanical studies have demonstrated that increasing the interlocking screw diameter as well as the number of interlocking screws will reliably increase the fatigue strength of the construct [33, 36]. The development of angular stable interlocking fixation has also proven to have increased fatigue performance [36, 37]. In the setting delayed/nonunion (Fig. 8.4) or limited healing potential, such as in pathologic fracture fixation, the surgeon must anticipate that the hardware will eventually fail without further intervention [38]. If the bone does not eventually accept the weight bearing demands of the limb, the hardware is bound to fail. This is a race between osseous union and mechanical failure of the implant. Towards that end, if delayed/nonunion is suspected, then steps must be taken to address the underlying problem prior to implant failure.

Infection is a major complication of intramedullary fracture fixation. It can result in secondary operations, increased risk of secondary complications, delayed/nonunion and worse clinical outcome for the patient. Fortunately, infection in conjunction with intramedullary nailing occurs in the minority of cases. Rates of infection for tibial nails are reported between 1.1 and 6.9 % without stratifying by fracture type or patient comorbidities [11, 18, 24, 31, 39, 40]. Rates in the femur are reported at slightly lower rates (1.5–3.2 %) largely because of the more generous soft tissue envelope and decreased risk of open fracture [18, 19, 40, 41]. The rate of infection following intramedullary nailing of humerus fractures is similarly low [40].

There are several well accepted risk factors for infection following intramedullary nailing. The presence of an open fracture overwhelmingly predisposes patients towards the development of deep infection [11, 31, 39, 42]. This has been demonstrated in several large clinical studies, with infection rates reported as high as 22.6 % in Grade III open injuries by Yokoyama et al. [43]. Single or serial thorough debridements are of utmost importance for prevention of deep infection. Delay or lack of peri-operative prophylactic antibiotics also increases the risk of postoperative infection by up to 29 % [40]. Other risk factors for infection following intramedullary fixation include lower socioeconomic status [40], delay in soft tissue coverage/wound closure [43] and duration of the surgical procedure [18, 42]. Prior stabilization with external fixation has also been shown to increase the risk of postoperative infection after definitive intramedullary nailing [18, 44, 45]. While this data may be confounded by the increased use of external fixation in the face of open fracture, the potential for contamination of the intramedullary canal by means of pin tracts is real. In order to minimize this complication, it is recommended that conversion to intramedullary nail takes place prior to 14 days. This is based upon data by Harwood et al., who showed that the rate of infection dramatically increased in those patients who received definitive intramedullary nailing greater than 2 weeks after initial external fixation had been applied [45]. Clearly, factors intrinsic to each individual patient, including metabolic derangements, immunodeficiency and general health status can increase the risk of infection and will not be discussed in detail.

Treatment of infection following intramedullary nailing requires a comprehensive approach. Presence of deep infection should mandate serial debridement with the intent of decreasing the bacterial load and obtaining accurate cultures. Culture specific antibiotic treatment, preferably in conjunction with an infectious disease specialist is the rule, with duration of treatment sufficient to allow bony healing. Soft tissue coverage should be performed, should this be an issue. In the acute postoperative period, a strategy of irrigation and debridement, systemic antibiotic treatment and suppression and hardware retention is preferable. This allows for bony stability that is in-fact a prerequisite for successful eradication of infection. However, this strategy may not succeed. In fact, data suggests that the presence of an intramedullary nail (rather than plate and screw fixation) places the patient at a higher risk of failure of this strategy, thus requiring nail removal [46]. Should nail removal be necessary, a two stages or single stage treatment protocol can be persued. Hardware removal followed by temporary local antibiotic therapy in the form of beads, followed by delayed definitive fixation has been reported with good success [47, 48]. However, a more aggressive single stage protocol, consisting of hardware removal, thorough debridement and fixation with an antibiotic coated intramedullary nail has been described as well [49, 50]. The nail provides stability and the antibiotic cement coating supplies local antibiotic therapy (in conjunction with systemic treatment) which can theoretically create an environment suitable for infection suppression as well as ultimately osseous union without the need for additional operations.

Because of the biologically friendly nature of the procedure, nonunion following intramedullary nailing occurs at a predictably low rate. The ability to reduce and stabilize the fracture in a closed fashion and create minimal disturbance to the blood supply and soft tissue is advantageous towards expeditious fracture healing. However, delayed union and nonunion can be expected to occur in a small percentage of cases. Different authors have varying definitions of what constitutes delayed union and nonunion. Delayed union is present when a fracture, while showing clinical or radiological signs of ongoing healing, fails to unite within the anticipated healing time for that fracture. Non-union can be defined as when the normal biological healing process of bone has ceased, without union occurring [51]. Most data on the incidence of delayed union after intramedullary fixation focuses on tibia fractures, with a rate between 8 and 12 % reported [5, 6, 24, 39]. Reported rates of delayed union in the femur are slightly lower, around 5 % [52]. Rates of nonunion in large series of femoral nails range between 0.9 and 6 %, with no observed differences between antegrade and retrograde techniques [9, 52, 53]. Nonunion occurs at higher rate for intramedullary tibia fracture fixation, likely explained by an increased prevalence of open fractures, inferior soft tissue envelope and blood supply. Rates of nonunion for these cases are reported between 2.6 and 16 % [5, 11, 24, 31, 39, 54]. Humeral nailing nonunion rates are also historically higher than those of the femur. Rates anywhere between 3 and 13 % have been observed [55, 56]. This may be secondary to the non-weight bearing nature of the bone, as well as high torsional forces seen in the upper extremity, in comparison to the femur or tibia. However, meta-analysis would suggest that this nonunion rate is no different than humeral fractures treated with plate and screw constructs [57].

There are several known risk factors for impaired osseous union following intramedullary fixation. Knowledge of these factors is critical in order to help ameliorate their effects, as well as appropriately counsel patients. Open fractures are at an increased risk of delayed/nonunion [5, 6, 31]. This can be attributed to disruption of the soft tissue envelope and compromise of blood supply, as well as increasing the risk of infection. Tobacco use has also shown a detrimental effect on bone healing following intramedullary fixation according to studies by Hak and Josten [6, 58]. Reaming has been shown to have a beneficial effect on bone healing after intramedullary nailing, despite concerns of endosteal blood supply disruption. As such, unreamed nailing is associated with an increased rate of delayed and nonunion [31, 35, 54]. Attal et al. found that tibia fractures within the mid-diaphysis, followed by the distal third had a higher rate of delayed union versus proximal third fractures (16.7 & 10.5 % versus 5.9 %) [5]. Again, this would logically seem to be related toward the issue of available soft tissue coverage and blood supply. Also, residual fracture gaps and lack of fracture site compression have been associated with a higher risk of delayed/nonunion [39, 59].

Management of delayed and nonunion following intramedullary nailing must be considered on a case by case basis. First, any underlying metabolic and endocrine abnormalities should be addressed. Appropriate bony healing cannot be expected to reliably occur in the presence of vitamin/mineral deficiencies or other inhibiting medical problems. Brinker et al. found that amongst femoral nonunions, 84 % of all cases had an underlying metabolic and/or endocrine abnormality [60]. The most common abnormality observed was vitamin D deficiency, as well as calcium imbalances, hypogonadism, thyroid disorders and parathyroid hormone abnormalities. Clearly any suspicion of metabolic bone pathology should be actively pursued, preferably with the assistance of an endocrinologist or other professional with experience in these disorders. The management of delayed unions through the use of dynamization is controversial. Dynamization is a technique that is based upon the principle of increasing load and compression across the fracture site to promote and enhance fracture union. Dynamization was routinely employed in the past for this purpose [29, 61]. This has unfortunately not proven to be a panacea for bone healing following nail fixation as it does not guarantee healing and has risks. Wu et al. reported only a 58 % success following femoral nail dynamization at 4 months post-op; additionally, 21 % of cases experienced 2 cm or more of shortening after removal of statically locked fixation [62]. Similar results have been observed with tibia fracture dynamization [63]. One study comparing dynamization to continued static interlocked fixation did find a minimal decrease in mean time to union (dynamized 13–28 weeks, mean 19; 1 nonunion) versus not dynamized (16–30 weeks, mean 23.5, 2 nonunions), but this is of questionable clinical significance [64]. Overall, dynamization is a widely practiced technique, but expectations should be tempered and patients and providers should understand the legitimate risk of secondary malalignment and need for further operative intervention. It is the authors’ recommendation that dynamization should not be performed as a solitary procedure intended to avoid or treat a delayed or nonunion.

In the presence of an established nonunion following intramedullary nailing, reaming and exchange nailing has proven to be a viable technique. Usually, the second nail is thicker than the first one (Fig. 8.5). Some revision nails offer the option of creating interfragmentary compression. We refer to the specific chapter on nonunion. The best results using exchange nailing have been reported by Shroeder et al. In a cohort of 42 femoral shaft fractures treated initially with intramedullary nailing, all underwent closed exchange nailing and were met with an 86 % union rate (36/42) without any additional procedures [65]. At the other end of the spectrum, Waresh et al. found that reamed exchange nailing resulted in 53 % union (10/19); 8 of the 9 fractures that did not heal went on to heal with an additional revision operation following the exchange nailing (including bone grafting and either compression plating or revision intramedullary fixation) [66]. Hak et al. reported a 78 % union rate following reamed exchange nailing in femoral shaft nonunions [58]. Interestingly, all nonsmokers in this cohort of 23 patients had a successful outcome with exchange nailing alone, whereas only 2/3 of smokers obtained union. In conclusion, reamed exchange nailing is a viable option in the presence of nonunion following prior intramedullary fixation. This technique should reliably result in union in the majority of patients. Failure of exchange nailing should prompt additional operative intervention such as open debridement, bone grafting and internal fixation. It is the authors’ recommendation that nonunions be treated on an individual basis, with fixation and other adjunct procedures (including nonunion debridement, intraoperative cultures and bone grafting) chosen based upon the condition of the patient and type of nonunion.