Orthopedic surgeons continue to encounter fracture healing complications, including malunions, delayed unions, and nonunions. Even in the presence of rigid fixation, such as locking plate constructs, failure rates as high as 20% have been reported in difficult to treat areas such as the distal femur.1 Therefore, optimizing the speed and strength of fracture healing continues to be a research emphasis. In the early period of the modern fracture treatment era, there was an emphasis on anatomic fracture reduction with rigid internal fixation. This led to stable fixation constructs requiring extensive soft tissue dissection and bony devascularization, compromising healing.2 A subsequent shift toward indirect reduction techniques and the development of limited contact plates helped preserve the bony vasculature. The introduction of locked plating provided fixed-angle fixation with minimal insult to the biologic environment, allowing surgeons to tailor implant strength, stiffness, and size to specific fracture requirements.3
Advances in the understanding of the biology of bone healing suggest that the optimal biomechanical conditions at the fracture site change throughout the stages of healing. Implant dynamization has been used with intramedullary nails (IMNs) and external fixators to optimize the healing environment. The biomechanical properties of current plating systems remain relatively static after insertion. There may be a role for dynamic plating, but options are currently limited.4
Despite reports of clinical success, controversy remains regarding the optimal techniques, timing, and direction of dynamization. This article explores the biology and stages of fracture healing, reviewing laboratory, animal, and clinical studies investigating the impact of dynamization on bone healing.
Biology of Fracture Healing
Fractures heal through primary or secondary processes. Primary, or direct, bone healing occurs when mechanical strain, defined as the movement at the fracture site divided by the size of the original fracture gap, is less than 2%.5 Intramembranous healing occurs via Haversian remodeling, leading to direct remodeling of the fracture site with bone without a cartilage scaffold. Rigid fixation necessary for primary bone healing requires stable fracture reduction, producing stress shielding at the fracture site and cortical necrosis via periosteal damage.6 The transmission of large loads across the fracture gap can disrupt bony healing, leading to plate failures and high nonunion rates. Non-gliding implants such as lag screws and compression plates promote primary bone healing.
Secondary, or indirect, bone healing is stimulated by interfragmentary motion (IFM) at the fracture site when mechanical strain is between 2% and 10%.5 Less rigid fixation may allow micromotion at the near and far cortex, enhancing secondary bone healing.7 The process of bone healing has 4 overlapping stages: inflammation, proliferation, consolidation, and remodeling.8 Inflammation and hematoma formation occur within a few hours of fracture, stabilizing the fracture site and catalyzing the osteogenic response. Fibroblasts and mesenchymal cells then migrate to the fracture site, where osteoblasts and fibroblasts proliferate. Within 3 to 4 weeks, a soft callus is formed via intramembranous ossification. This fibrous, temporary callus is replaced by hard callus, or primary bone callus, via endochondral ossification. This consolidation can start 3 to 4 weeks after injury and lasts for 2 to 3 months, until bony union. The final remodeling phase involves woven bone being replaced with lamellar bone. Secondary healing is generally faster and stronger than primary bone healing, but overall healing time is dependent on many factors and can be more than 1 year. Fixation techniques that promote secondary healing include external fixators, dynamized IMNs, and bridge plates.
The mechanical environment also affects angiogenesis at the fracture site, which is vital for both osteogenesis and mineralization of soft callus.9–11 In primary healing with more stable fixation, intramedullary blood vessel bridging occurs, leading to endosteal callus. This contrasts with secondary bone healing, where micromotion at the fracture site stimulates angiogenesis circumferentially around the fracture, leading to soft callus.10
Primary Versus Secondary Healing
Fracture pattern and location affect the optimal bone healing process and, likewise, the surgeon's implant choice. For example, articular fractures are best treated with anatomic reduction and primary bone healing, while comminuted diaphyseal fractures respond favorably to secondary healing via bridge plating or IMN.6
Although bony union occurs through either process, animal studies have demonstrated significant improvements in healing time, callus size, and symmetry when secondary bony healing is coupled with controlled IFM.7,12–16 Flexible fixation also decreases stress shielding seen with rigid fixation, leading to a lower nonunion rate.17 However, there is an upper limit to movement at the fracture site; excessive interfragmentary displacement is correlated with delayed healing and an increased risk of malreduction, refracture, and nonunion.5,6 Additionally, flexibility often comes at the price of plate strength. Plates must maintain sufficient strength to carry repetitive loading until bony union. Optimal plate fixation must achieve a delicate balance, allowing adequate IFM to promote callus formation and sufficient strength to maintain fracture reduction. As healing progresses, callus formation provides greater fracture site stability, changing the ideal biomechanical properties of the fixation construct and optimal IFM.
Traditionally, fracture fixation has been static, with the biomechanical properties of the construct remaining relatively unchanged over the life of the implant. However, current understanding of bone biology suggests that optimal conditions for bony healing, specifically the amount of fracture site IFM, may vary throughout the stages of healing.15 Construct dynamization—the transition between rigid and flexible fixation—offers a means for addressing this need.6,14,16,18–20
The use of both primary and secondary processes facilitates faster bone healing, where mechanical strength is closest to intact bone.21 Dynamic constructs may allow surgeons to tailor treatment by fracture pattern. For example, a diaphy-seal femoral shaft fracture treated with an IMN and a distal locking screw can be dynamized a few weeks after implantation to increase IFM. After dynamization, this increased IFM stimulates secondary healing, while the callus developed during rigid fixation guards against excess mobility.22
Inverse dynamization, where a construct is initially flexible and then becomes rigid over time, has also been described. However, given its limited success in biomechanical and clinical studies, it is not the focus of this review.23–26
Studies on early dynamization include implants flexible from the time of implantation and those that change from rigid to flexible within 1 to 2 weeks of implantation. Theoretically, early, high-magnitude mechanical stimulation via IFM affects stem cell proliferation, promoting differentiation into a cartilage phenotype and subsequent bone healing.27 While the results of animal studies have been mixed, several studies suggest that early dynamization stimulates greater callus formation compared with rigid controls.28,29 Conversely, other reports suggest that strain on the cartilage during its differentiation phase may actually delay healing.14,15 When comparing studies, it is important to differentiate the fixation method and whether axial or elastic dynamization is being studied. Additionally, the animal model is also important, as normal healing times vary between species.30Table 1 summarizes the early dynamization studies reviewed below, categorizing them by elastic vs axial dynamization, animal model and bone, type of fixation, and timing of dynamization.
Summary of Early Dynamization Studies
Early Elastic Dynamization—Animal Studies. Elastic dynamization involves destabilization or deformation of the fixation construct under physiologic loading, with fracture gap recovery during unloading. No plane of motion is necessarily favored. Traditional external fixators are elastically dynamized through pin removal. This makes the construct less rigid, which allows for increasing IFM. However, limited data exist regarding the benefits of early elastic dynamization for fracture healing.
Claes et al29 externally fixed rat femoral osteotomies, removing the inner fixator bar at 1 week. At 5 weeks, no differences were seen in flexural rigidity, callus volume, osteotomy volume, tissue mineral density, or bone mineral density between early dynamized and constantly flexible constructs (P>.05). Compared with constantly rigid constructs, the dynamized group did have greater fracture callus bone volume (P=.005) and total callus volume (P=.004). However, the authors interpreted increased fracture volume at 5 weeks as a delay in bone healing, representative of late consolidation.
Similarly, Willie et al15 investigated unilateral external fixators in a rat femoral model dynamized at either 1, 3, or 4 weeks by removal of the inner fixator bar. After 5 weeks, the group dynamized at 1 week demonstrated greater cartilage compared with the rigid group (P=.020). However, fewer samples dynamized at 1 week demonstrated bony bridging compared with rigid specimens and specimens dynamized at 3 and 4 weeks. Rats dynamized at 4 weeks were the only group with complete bridging of all periosteal and intracortical regions, both medially and laterally. The authors interpreted these data to suggest that early elastic dynamization delays healing compared with rigid fixation or delayed dynamization.
Early Elastic Dynamization—Clinical Studies. Kulshrestha et al31 performed a randomized, controlled study on 60 displaced, unstable comminuted distal radius fractures that were treated with dynamic or static external fixators. In the dynamic group, the fixator was manipulated at 3 weeks to allow 30° range of motion, with full mobilization at 5 weeks. The dynamic group had improved early functional and anatomical outcomes compared with the static control group, including restoration of palmar tilt (P<.001), reduced loss of ulnar tilt (P=.05) and radial height (P=.04), and improved Gartland and Werley functional outcome scores at 3 and 6 months. However, 2-year follow-up showed no significant difference in functional outcome measured by Disabilities of the Arm, Shoulder and Hand scores (P=.14).
Early Axial Dynamization—Animal Studies. Most axial dynamization studies use modified external fixators to allow uniplanar, axial motion. Loading collapses the fracture gap, often without recovery. Larsson et al28 investigated canine tibia fractures treated in this manner using a roller-bearing, telescoping sleeve, axially compressing the osteotomy under functional loads. External fixators were applied to 2-cm osteotomy gaps in both hind leg tibiae. The control leg was rigidly locked. At 1 week, the experimental side was unlocked, allowing free axial movement. Dynamized legs showed increased density of endosteal bone and greater periosteal callus area at 4 weeks. Conversely, the control side demonstrated greater callus at 9 weeks and increased density at 12 weeks. Torsional stiffness of the dynamized side was significantly greater than the control side at 6 weeks. No statistically significant differences in mechanical properties were reported before or after this 6-week period. No significant differences in periosteal or endosteal callus tissue composition were reported at any time.
A similar canine model by Egger et al32 used bilateral transverse diaphyseal tibia fractures with a 2-mm gap. The control leg used a rigid fixator, while the study limb employed a modified fixator with an axial telescoping mechanism dynamized at 1 week. Dynamization led to physiologic loading of the osteotomy site, closing the fracture gap and increasing functional weight bearing on the dynamized leg compared with controls. Mechanical testing after the animals were killed demonstrated a statistically significant increase in torsional strength for the dynamic side. No differences in periosteal or endosteal callus formation were seen. The authors hypothesized that improved fracture healing in the dynamized group was due to fracture gap closure rather than an effect on the bony healing process.
While physiologic loading via weight bearing is dependent on patient mobility, mechanical stimulation was theorized as a more precise method to evaluate fracture healing. Goodship and Kenwright7 investigated an ovine model with a 3-mm tibial fracture gap. The control group had rigid fixation throughout the 12-week study. The experimental group was exposed to mechanical axial stimuli for 17 minutes (500 cycles at 0.5 Hz, 1-mm displacement) each day, starting postoperative week 1. Prior to invoking the daily stimuli, the study fixator was unlocked and subsequently re-locked once completed. Fracture stiffness was measured every 2 weeks. During the first 4 weeks, the two groups demonstrated similar increases in stiffness, judged by a strain gauge transducer attached to the fixator bar. At weeks 8 and 10, the stimulated group was stiffer (P<.05). Postmortem assessment at 12 weeks demonstrated higher torsional stiffness in the stimulated group (83%±3.5% vs 54%±10%, P<.01), with more complete bridging of the fracture callus as judged by radiographic interpretation.
Bottlang and Feist33 described a far cortical locking screw (FCL) that has an unthreaded, reduced-diameter mid-shaft that is postoperatively dynamized by physiologic loading. In an ovine tibial osteotomy model, 3-mm fracture gaps were randomized to FCL plates or traditional locking plates. Far cortical locking screw plates reduced construct stiffness by 84%, allowing micromotion at both near and far cortices. After the animals were killed at week 9, FCL constructs had 36% more callus volume (P=.03). Far cortical locking screw constructs demonstrated statistically similar bone mineral content in both the near and far cortices (P=.91), in contrast to traditional locking plates. Far cortical locking screw constructs also exhibited greater torsional strength (P=.023), with greater energy to failure in torsion (P<.001).
Use of FCL screws in small-diameter bone is theoretically limited because they rely on long screw shafts for elastic flexion. Tsai et al34 addressed this limitation by inserting silicone spacers in locking screw holes, thereby elastically suspending screws within the plate. This was named “active plating” because it provided for approximately 1 mm of axial translation under compressive loads. Using a biomechanical synthetic bone model, dynamized constructs had greater IFM at the near and far cortices than traditional locking plates, with decreased axial and torsional stiffness. Bending stiffness was similar. Active plating was further compared with standard locked plating using an ovine model with tibial osteotomies, demonstrating greater callus area on repeat postoperative radiographs (P<.001), along with increased failure energy.
Comparing early axial vs elastic dynamization, Arazi et al35 used external fixators in a canine model. Tibial osteotomies were performed on both hind legs and rigid fixators were applied. At 2 weeks, fixators on the left leg were axially dynamized, while the right was made progressively less stiff by removal of fixator pins. At 2 months, maximum load to failure and maximum deflection were not significantly greater (P=.20) in the elastically dynamized group, implying no difference in fracture strength or stiffness.
Early Axial Dynamization—Clinical Studies. Bottlang et al36 conducted a prospective observational study of 32 human distal femur fractures fixated with FCL screws. These authors found that 30 of 31 fractures healed at 15.6±6.2 weeks, with no implant or fixation failure after 17±4 months, suggesting safe and effective fixation.
Linn et al19 investigated supracondylar femur fractures treated with traditional locking plates that were axially dynamized secondary to overdrilling the near cortex screw hole. Compared with traditional techniques, the dynamic group had statistically greater callus formation after 6 weeks (P=.048) with decreased nonunion rates (P=.9).
Delayed, or late, dynamization typically describes a change in construct stiffness from rigid to flexible. From a biologic perspective, excessive early IFM, especially in a comminuted or unstable fracture, would hinder fracture stabilization.3 Once the fracture is in the remodeling stage, with the callus widely calcified and close to bony bridging, the fracture would be capable of distributing increased interfragmentary strain.15 At this time, a more flexible construct would be desirable to both provide the IFM necessary for secondary bone healing and avoid the stress shielding associated with rigid constructs.36,37
There is limited consensus between studies on the optimal timing for delayed dynamization, likely due to differences in experimental design and fracture location. The time at which motion is introduced is generally reported as between 4 and 12 weeks after implantation, with a distinction made between delayed elastic and axial dynamization. Table 2 summarizes the late dynamization studies reviewed below, categorizing them by elastic vs axial dynamization, animal model and bone, type of fixation, and timing of dynamization.
Summary of Delayed Dynamization Studies
Delayed Elastic Dynamization—Animal Studies. Claes et al14 investigated delayed elastic dynamization with external fixators in diaphyseal rat femur fractures. The fixator bar was dynamized at either 3 or 4 weeks. Fracture healing was evaluated at 5 weeks, with enhanced healing demonstrated in both groups. Both had statistically greater flexural rigidity than the constantly flexible controls (P<.05). The elastic modulus in the late dynamized group was 88% greater than the rigid control (P=.031) and 282% greater than the flexible control (P=.002). There was greater callus density in the late dynamized group compared with controls, with significantly less volume (30% less than the rigid control, P=.016; 37% less than the flexible group, P=.026), implying advanced bone remodeling. This study complements the study by Willie et al15 described previously with external fixators in rat femoral fractures showing increased bony bridging in fracture sites dynamized at 3 and 4 weeks vs those dynamized at 1 week.
Delayed Axial Dynamization—Animal Studies. Dynamization of IMNs is the most common example of delayed axial dynamization. Typically, distal locking screws are removed 3 to 12 weeks postoperatively. Although IMN dynamization is a widely accepted practice, limited clinical data support its use. There is also debate about whether dynamization should be routine or reserved for delayed unions.38
Georgiadis et al12 studied dynamic IMN in a canine model using transverse tibial osteotomies. After 20 weeks, fractures dynamized at 8 weeks demonstrated greater stiffness at the fracture site (P<.05) and more dense trabecular callus patterns than statically locked controls. However, the rate of bony union was not significantly different.
Utvåg et al13 used IMNs in transverse diaphyseal rat femur fractures to compare unlocked, locked, and those with locking screws dynamized after 20 days. At 6 and 12 weeks, 10 rats from each group were killed for mechanical testing. At 12 weeks, the dynamized group had a larger callus area than the locked control (P=.01). Compared with the unlocked control, the dynamized group had greater maximum bending load at 6 weeks (P=.03) and 12 weeks (P=.04), and higher bending rigidity at 12 weeks (P=.01). Dynamized fracture sites had significantly increased fracture energy at 6 weeks compared with the locked group (P=.01). There were no nonunions in any group. The Georgiadis et al12 and the Utvåg et al13 studies both suggest that routine IMN dynamization may positively impact early bone healing but has no effect on the rate of bony union.
Delayed axial dynamization has also been studied with modified external fixators. Goodship and Kenwright7 had mixed results investigating diaphyseal tibial shaft fractures in sheep treated with dynamized external fixators. Starting 1 week after fixation, fractures were mechanically dynamized with axial displacement rates of 2 mm/sec, 40 mm/sec, or 400 mm/sec of applied cyclic micro-movement for 500 cycles. This occurred 5 days per week for 11 weeks. A second group of 6 sheep were dynamized at 6 weeks postoperatively, with a displacement rate of 400 mm/sec for 500 cycles. Contrary to more recent findings on delayed dynamization, the delayed group had a lower rate of bone mineral content (P<.001), decreased torsional strength (P=.03), and decreased stiffness (P=.17) when killed at 12 weeks. Less callus and consolidation were also noted in this group. Within the group dynamized at 1 week, the 400-mm/sec group had a greater walking stiffness index compared with the other groups (P<.001) and the 40-mm/sec group had significantly higher values of bone mineral content, torsional stiffness, and strength (P<.05).
Delayed Axial Dynamization—Clinical Studies. Dynamization of tibial and femoral IMNs has been thoroughly investigated. General consensus indicates that routine dynamization is unnecessary.38,39 However, it may be a good first option in cases of delayed union or nonunion. Reported rates of healing after dynamization vary greatly, ranging from 19% to 82%.40–42
Basumallick and Bandopadhyay39 conducted a randomized, prospective trial dynamizing interlocking nails in femoral fractures that required open reduction. Nails dynamized 2 to 3 months after fixation showed faster healing compared with static fixation (19.2 weeks vs 23.5 weeks, P<.05), with no difference in union rate.
Vaughn et al40 retrospectively looked at 35 tibial and femoral shaft fractures treated with IMNs with subsequent delayed unions or nonunions that were dynamized by removing the interlocking screws an average of 6 months postoperatively. There was a 54% union rate. In a similar, larger study, Litrenta et al43 retrospectively looked at 194 tibia fractures that underwent dynamization or exchange nailing for delayed union/nonunion. Fractures achieved union in 83% of dynamized and 90% of exchange nails. There was no difference in their radiographic outcome measures if the intervention took place before or after 6 months.
External fixators have also shown promise with delayed dynamization in clinical practice. In a prospective, randomized trial, Noordeen et al44 investigated 56 tibial shaft fractures treated with external fixators using immediate elastic, delayed elastic, or delayed axial dynamization. The elastic groups had fixators modified with a silicone cushion to allow cyclic micromotion up to 2 mm under weight bearing. The axially dynamized group used modified fixators allowing for free axial compression. Both delayed groups were dynamized between 4 and 6 weeks postoperatively. The mean time to fracture healing was 14.1 weeks for delayed axial, 15.9 weeks for delayed elastic, and 19.3 weeks for immediate elastic dynamization, respectively (P=.004, F2,33=6.58). The study authors endorsed late dynamization, while supporting pure axial dynamization as superior to elastic flexibility. However, the elastic external fixators allowed for IFM up to 2 mm, doubling the optimal IFM of 0.2 to 1 mm that Good-ship and Kenwright7 recommend. Data were only analyzed for 36 of the original 56 patients due to attrition and exclusions.
Success with delayed dynamization in external fixators and IMNs has led to attempts to dynamize plating constructs, specifically in areas such as the distal femur where nonunions are common. Described techniques include overdrilling the near cortex, removing screws close to the fracture site to increase the working length of the plate, exchanging locking screws with non-lockers or bicortical screws with unicortical screws near the fractures, or removing lag screws.4,19,45 However, none of these options can dynamize over time in vivo without intervention. Schultz46 described delayed dynamization in a plate construct having a locking screw with a threaded degradable polymer locking mechanism. At insertion, the system functions in a manner similar to locked plating. Over time, the polymer locking mechanism resorbs, creating an avenue for delayed dynamization.
Fracture healing complications, including nonunions and delayed unions, can often be attributed to implants that are either too flexible, leading to loss of reduction or hypertrophic nonunion, or too stiff, leading to stress shielding and atrophic nonunion.3 Mavcic and Antolic37 described an ideal construct that would provide rigid compression early while allowing larger amplitudes of load bearing (dynamization) later in the healing process. Construct dynamization offers a potential solution to this problem, but as this review demonstrates, there is still no clear consensus on the optimal timing and type of dynamization. Part of this is due to the imperfect understanding of fracture healing in humans. Generalizing small animal models to humans has its limitations, but as a large percentage of animal studies on fracture healing and dynamization are done in mice and rats, these data are still relevant.30 Another challenge is the variability of experimental design across studies. Differences in fracture type, fixation construct, dynamization timing, and histologic and biomechanical data collection and interpretation confound direct comparisons and make a clear consensus difficult.
For example, several axial dynamization studies involved telescoping modifications that lead to closure of the fracture gap.7,28,32 As smaller osteotomy gaps tend to heal faster, it becomes difficult to differentiate whether accelerated bone healing is due to increased IFM and cyclic load bearing or simply the result of a smaller fracture gap.47 Claes et al29 found no increased healing compared with rigid controls in early dynamization of rat femurs using fixators that maintained fracture gap size after dynamization, while Egger et al32 reported statistically greater torsional strength in canine tibial fractures dynamized with a telescoping modified external fixator leading to immediate fracture gap closure.
While most studies described a positive correlation between bone volume and fracture healing, Claes et al29 suggested that increased bone volume at later stages of fracture healing actually represents delayed healing. As the consolidation phase of fracture healing can begin as early as 3 to 4 weeks, increased callus volume after 5 weeks may represent a delay in the consolidation phase where soft callus is replaced by hard callus. This is consistent with tibial osteotomies in sheep and dogs that have shown maximum callus formation approximately 6 weeks after implantation, before remodeling occurs. Therefore, increased callus volume later in bone healing may actually be a sign of delayed healing.48
Additional controversy exists regarding the correlation between callus formation and biomechanical stability. Claes et al49 found no correlation between the amount of periosteal callus formation and bending stiffness.
Overall in animal models, it appears that controlled dynamization, regardless of timing or type, does improve histologic and biomechanical properties compared with controls that are either statically rigid or flexible. In limited studies that compared early with delayed dynamization directly, delayed dynamization at 3 to 4 weeks had improved histologic results.14,15 There were no obvious differences in outcomes across animal species.
Clinically, the most common uses of dynamization currently are with IMN and with plating. While the benefits of routine IMN dynamization remain unknown, evidence supports that delayed axial dynamization can accelerate fracture healing in the presence of delayed unions and nonunions.40,43 This should be strongly considered as a quick and minimally invasive clinical practice, especially in fractures that are notoriously difficult to treat, such as the distal femur. The effect of “plate dynamization,” either using an overdrilling technique or with a dedicated implant, appears to be effective in stimulating callus formation.19,21 In many situations, a blunted callus response occurs after bridge plate fixation of distal femur fractures.50 Although the cause of this cannot be fully attributed to lack of fracture site motion, it is becoming clearer in the literature that callus formation is enhanced with a dynamic plating situation. Therefore, the authors recommend using a dynamic plating technique or implant in situations where the patient is at high risk for nonunion, such as systemic biological impairments (eg, diabetes), high-risk injury factors (eg, open fracture), or in osteoporotic bone.
As discussed above, this review's main limitations are related to variability of experimental design, including animal model, type and timing of dynamization, and interpretation of results, making direct comparisons and a clear consensus difficult.