Introduction

The repair process after bone injury requires the participation of several cell and tissue types to achieve a successful healing outcome. Bony tissue represents an impressive biomaterial due to its ability to completely regenerate under normal healing conditions. Despite this efficacy, still up to 10–15% of the fracture patients show an impaired healing process, leading to a delayed healing outcome or even to a non-union [1]. This is not only a burden for the patient’s life due to additional surgical treatment and hospitalization time but also for the socio-economic and health care systems [2]. Hak and colleagues reported in 2014 that a treatment of an established non-union of long bones costs over $10,000 (average costs: Canada $11,800, USA $11,333, and UK £29,204) [3].

The process of bone fracture healing can be divided into five distinct but overlapping phases: hematoma formation with an accompanying inflammation (which is separated into a pro-inflammation [1] and anti-inflammation [2]), soft callus formation [3], hard callus formation [4], and remodeling [5]. A hematoma is formed in the fracture area due to the blood influx after vessel disruption [4, 5]. Cells of the innate immune system are one of the first cells infiltrating the fracture area [6,7,8]. Based on their secretion profile, they create a pro-inflammatory state, which induces the recruitment of cells of the adaptive immunity and mesenchymal stromal cells (MSCs). For the progression of the repair process, the pro-inflammation has to switch to an anti-inflammation. The switch to the anti-inflammatory state initiates the revascularization of the fracture area, which is another prerequisite for a successful healing outcome [9••]. In the next step, fibrocartilage tissue refills the fracture area and a soft callus is formed leading to a first stabilization of the injury site. The cartilaginous tissue matures, becomes hypertrophic, and starts to mineralize. An external hard callus is build composed of newly formed woven bone, which replaces the hypertrophic chondrocytes. The last step is the remodeling of the fracture area [10]. In humans, this remodeling process can last up to several years, depending on the general condition of the patient and on the fracture type/location.

The question is: When is an impaired healing fracture called a non-union? There is still no standard definition how to define in general a non-union. Among others, it depends on the site of injury. Based on the definition of the US Food and Drug Administration (FDA) for long bones like the femur, a non-union is present when the fracture is not healed within the first 9 months after injury and showed no signs of healing progression for at least 3 months [11].

The definition of non-union is not simple or satisfying implying that also the diagnosis of a non-union is difficult. This is due to the fact that not two non-union cases are alike as causes are multi-factorial. The affected bone has to be considered, the type of injury, the bone quality, the soft tissue cover, the patient with its habits and comorbidities, and the environmental factors. This not only hinders clear treatment guidelines but also research into non-union to gain a better understanding. To overcome these difficulties, a non-union scoring system (NUSS) has evolved similar to the Injury Severity Score or the scoring system to grade joint disease. Early scoring systems relied solely on radiographic assessments; however, this was proven to be insufficiently reliable—being compared with tossing a coin [12]. The NUSS exceeds the radiographic evaluation to also consider the bone quality, the history of the fracture as being open or closed, the clinical interventions up to the current point, the soft tissue state, and uses the American Society of Anesthesiologists (ASA)-grade system for patients to include the individual patient constitution. The score for the individual patient situation would provide the surgeon with a treatment guideline. Low scores would receive a standard treatment of refreshment and fixation revision, medium low scores would require a more specialized treatment including a supplementation with osteoinductive factors (autologous spongiosa or growth factors) while fixation is revised. The higher score would indicate that a specialized treatment in form of a viable bone graft or segment transport is necessary in addition to applying osteoinductive factors (autologous spongiosa or growth factors) with a specialized fixation. The highest score would indicate that an amputation has to be considered a final solution. The NUSS (Table 1) has been proposed in 2008 [13], giving a guideline for the classification of non-union fractures and indicators for the clinical treatment but lacking in validation at that time. Bastenberg et al. recently evaluated the scoring system, confirming that the NUSS score led to a high agreement in classifying non-union fractures between observers [14•]. Earlier performed evaluation studies in 2011 [15] and 2014 [16] also support the validity of the NUSS system to evaluate non-union fractures.

Table 1 Non-union scoring system (NUSS) is first proposed by Calori et al. in 2008
Full size table

Reasons for Non-union

The underlying causes for the occurrence of fracture non-unions are various and depend among others on the mechanics (e.g., site of injury), associated concomitants (e.g., infection), patient-dependent factors (e.g., age, lifestyle, chronic diseases), and the type and severity of the fracture itself.

Treatment in cases of a non-union often includes an optimization of the fracture stabilization (Fig. 1). Movement of the bone fragments exceeding a certain window will lead to an arrest of the healing (Fig. 2). However, a too rigid fixation will also affect the healing negatively. The mechanical stimulus to form bone and to repair a fracture would be missing and the bone-forming process would cease. This phenomenon led to an intensive study of the optimal fixation stiffness and technique to treat broken bones [17,18,19,20,21,22,23,24,25]. In addition, the concept of using a mobilization of the fractured bone during the healing cascade has been extensively investigated [26,27,28,29]. So far, no defined mobilization treatment strategy has reached the clinic, however, and early weight bearing is favored more and more in lieu of prolonged bed rest and immobilization of the affected limb.

Fig. 1
figure 1

Non-critical osteotomy gap in the femur of BL6 female 12-week-old mice were fixated with an external fixation (RISystem) with a rigid (a) and a semi-rigid (b) stabilities. Twenty-one days after osteotomy, the healing outcome was measured histologically (Movat Pentachrome staining: bone-yellow, cartilage-green, bone marrow-dark red, muscle-orange). The optimal fixation stability (a) allowed complete bridging and progression of remodeling while the unstable fixation (b) led to a larger callus formation to compensate for the missing stability thus leading to a delayed healing were at the 21-day time point bridging has occurred; however, the bone marrow cavity is still closed and remodeling has not yet succeeded. While the lack of stability in this case leads to a delayed healing, the bone has been able to overcome the lack by forming a larger callus and compensate for the non-optimized stabilization

Full size image
Fig. 2
figure 2

In a sheep tibia osteotomy model, a 3-mm gap was stabilized with a rigid external fixator (left) or with a rotationally instable fixator (right) (Movat Pentachrome histology). In this case, the mechanical instability was so high that healing was not possible (visible in the displaced bone ends on the right upper image). While under stable fixation at day 42 post-surgery, a woven bone callus bridged the gap (left lower panel), rotational instability caused the formation of a pseudo joint (right lower panel)—thus a non-union ensued

Full size image

Non-unions, arrests in the fracture repair process, can be classified as:

  • Septic non-unions (bacterial infection impeding the healing)

  • Pseudoarthrosis (atrophic, non-viable bone ends forming an artificial joint capsule)

  • Hypertrophic non-union (viable bone ends, indicating problems with the fixation rather than the biology)

  • Atrophic non-union (dysvascular bone ends, indicating problems with the biology)

  • Oligotrophic non-union (an intermediate of the above)

While in most of these cases the first step is an optimization of the fixation, the most critical distinction is the viability of the bony ends of the fracture. If the bone ends are still viable then the biology to heal the fracture is still available to aid the healing process and once stabilization is regained, healing will proceed. However, if the bone ends are no longer viable, changing fixation will not gain a progression in healing—in these cases, the bone ends have to be removed. The best prognosis can be given for those cases, where the bone has closed the bone marrow cavity in an attempt to “heal” the bone (Fig. 3). The bone has an active healing capacity but has not been able to overcome the gap between the bone fragments—narrowing the gap between the bone ends after reopening the bone marrow cavity will enable a bridging.

Fig. 3
figure 3

Healing in a critical-sized defect (5 mm) in a rat femur osteotomy model stabilized with an external fixation (custom made) has been analyzed histologically (Movat Pentachrome staining) after 3 and 6 weeks, respectively. Already at the 3-week time point, the closing of the bone marrow cavity at the bone ends is detectable. While the attempt to close the gap is visible in the cone formation at the left bone end in the 6-week sample, healing clearly has stopped and a non-union has formed. To enable healing, the bone cavities would have to be reopened and the bone fragments moved closer together to allow contact

Full size image

The most common reason for an insufficient biology in case of occurring non-union in bone is the lack in the revascularization and thus of angiogenesis in the fracture area. During the fracture healing cascade, this process occurs twice: first at the very beginning of the healing, after the injury disrupts the blood vessels and a hematoma is formed, the hematoma matures to an organized granulation tissue were the newly formed blood vessels reestablish the adequate supply of the fracture area. Angiogenic signaling during this phase is closely coupled with the inflammatory reaction, and both processes are interlinked [30]. An upregulation of the angiogenic signaling cascade occurs during the undisturbed healing when the first initial pro-inflammatory reaction abates. Therefore, the timely termination of the pro-inflammatory cascade is a key element of successful bone healing [9••, 31]. A second revascularization step is needed upon the transformation of the cartilage-dominated avascular callus towards the mineralized woven bone callus [32, 33] (Fig. 4). Through these newly formed vessels, cells important for the remodeling of the callus infiltrate.

Fig. 4
figure 4

Endochondral ossification in a mouse fracture model was evaluated histologically 14 days after three point bending fracture and internal fixation with an intramedullary nail. Hyaline cartilage transformed to hypertrophic cartilage (a) occurring blue green in the Movats Pentachrome staining. Cartilage consist of one cell type only, chondrocytes, and there are no blood vessels apparent. Consecutive formation towards woven bone proceeds with a revascularization step. Woven bone is highly vascularized (b). While bone appears yellow, bright red erythrocytes filling the newly formed vessel structure are clearly visible within the bone marrow interspersing the bony columns. Immunohistological staining of the vessels (c) (laminin staining, green) and T cells (CD3, red) shows that immune cells infiltrate the callus via the newly formed vessels at the border of cartilage and woven bone (cell nuclei, white)

Full size image

Non-unions with a lack in the biology, showing a disturbed revascularization require additional treatment to enhance the biological healing capacity of the bone to achieve a bony bridging and thus a functional bone structure. To date, several treatment options are available.

Treatment Strategies

A bone injury prone for non-union problems is the fracture of the lower extremity. For Germany, over 220,000 cases were reported for the year 2015 (Statistisches Bundesamt 19 Nov 2017) with an about equal distribution of male and female patients. About 90,000 patients were released from the clinic within 3 days. The German statistic Department listed between 13,400 and 14,800 cases of non-unions of fracture ends per year between 2010 and 2015 (Statistisches Bundesamt 19 Nov 2017) (Fig. 5).

Fig. 5
figure 5

ICA classification M84.1 lists cases of non-union of fracture ends; numbers refer to patients treated in German hospitals from 2010 to 2015. These numbers were released by the Statistisches Bundesamt, 19 Nov 2017

Full size image

Current clinical treatment options for a non-union that affects the lower extremity are reported here with an example of a patient who suffered from a III° open fracture of the lower leg with a resulting bone defect (Fig. 6a). The bone was stabilized with an external fixator executing the Ilizarov principle that allows to transport a bone segment (Fig. 6b). To bridge the critical bone defect at the distal end of the tibia shaft, an osteotomy was performed mid-tibial. The bone fragment thus created was then translocated distally. To enhance the healing, the ensuing gap was filled with autologous bone graft supplemented with bone morphogenetic protein 7. To account for the missing soft tissue coverage of the fractured bone, a muscle flap was performed. The gap constructed by the segment transport mineralized successfully; however, the docking site developed an infection and in consequence a non-union (Fig. 6c). During surgery, the infected bone was removed and a new gap was thus created. This gap was filled with a cement spacer loaded with antibiotics to eradicate the infection (Fig. 6d). While treating the infection, this method also created a fibrous sleeve resembling an artificial periosteum spanning the gap section. The Masquelet technique was therefore applied when the cement spacer was removed during the next surgical procedure. Due to the size of the defect, a vascularized fibular bone segment was implanted within the bone defect and the defect was again treated with autologous spongiosa that was inserted into the Masquelet membrane (Fig. 6e). At this stage, the Ilizarov fixator was removed and exchanged against a plate osteosynthesis.

Fig. 6
figure 6

A critical-sized bone defect in the lower leg ensued in a non-union. Upon treatment of the defect, a several techniques were used to reach a satisfactory healing outcome. This included a segmental transport (b), augmentation with autologous spongiosa and growth factor bone morphogenetic protein 7 (c), debridement after infection and a cement spacer application (d), and the transplantation of the vascularized fibular segment (e)

Full size image

This case combines a multitude of available treatment options due to the complication that occurred over the healing period. The critical-sized bone defect offered additional problems that allow the presentation of the available treatment strategies within one case report. The critical-sized defect however occurs during non-union treatment due to the debridement necessary in the case of avital or infected bone ends and therefore should be considered within this context.

This case shows the necessity of a change in the fixation of a non-union patient, albeit at a late time point. The external fixation system is exchanged for a plate osteosynthesis method. Changing the fixation is often necessary if difficulties in the bone healing process occur and is often the first step to counter healing difficulties.

The application of osteoinductive autologous spongiosa is still considered the gold standard to treat non-unions, gaining the best results. In this case the spongiosa was additionally enhanced by adding a growth factor, bone morphogentic protein 7 (BMP7). BMP2 and BMP7 are the only growth factors that gained approval for very specific non-union treatments. As they proved very successful in enhancing the bone formation, off-label use often occurs within the clinic in problematic bone healing cases [34, 35].

Tibial fractures often result in a lack of soft tissue coverage of the fracture area. In cases of a missing soft tissue coverage healing is delayed, mostly because of an impaired revascularization step [36,37,38]. Even today, the optimal time point for the muscle flap application in open fractures is controversial [39] mostly because of the high probability of infection due to the open-fracture scenario.

In the here-presented case, the non-union occurred due to an infection at the docking side after the successful bone segment transfer. In case of an infection, a thorough debridement is a necessity [40, 41] together with a thorough antibiotic treatment. Such an antibiotic treatment is often applied with a cement spacer in the orthopedic setting [42]. Ceramic biocomposites might offer an alternative that is biodegradable. Newly formed bone following the degradation after eradication of the infection could obliterate the second surgical intervention needed to remove the cement spacer [43].

In this case, however, the cement spacer was used to produce an artificial periosteum as described for the Masquelet technique [44]. The Masquelet technique represents a two-stage surgical treatment often used to treat patients with a severe open fracture [45]. It is independent of the size of the bone defect which is filled with an antibiotic impregnated cement spacer to induce a so-called periosteal membrane. Even though this membrane is equivalent to a fibrous capsule, the positive healing results support the reference to a periosteal membrane. During the second surgical intervention, a longitudinal incision is made through the periosteal membrane and the spacer is removed. After freshening the bone ends, the hollowed periosteal cavity is filled with autologous bone graft.

In the case of the here-presented patient, the filling of the periosteal cavity was further enhanced by transplanting a vascularized fibular bone segment into the gap [46]. A technique first reported in 1975, reported to have a 95% success rate, and used to enhance bone healing in bones from head to toe [47].

This clinical case demonstrated the multitude of treatment options currently available to treat non-unions; however, they all are time intensive and require a high compliance from the patient and a surgical specialist. Therefore, further improvement of the non-union treatment is desirable and several research approached are being investigated to date Table 2.

Table 2 Summary of the diagnostic, classification, risk factors, clinical treatment options, and future innovative treatment approaches of impaired or even non-healing fractures
Full size table

Treatment Strategies: Research Approaches

A multitude of studies was already conducted in order to elucidate a biological biomarker that determines a non-healing fracture when the injured patient is coming into the clinic. The finding of such a suitable biomarker would enable the development of appropriate treatment strategies for non-healing conditions at the time of the initial fracture stabilization procedure. Due to the complexity of the patient-related and non-related risk factors for a non-union, several research studies using biological approaches have been initiated and evaluated for their potential to improve bone healing. These will be considered in the following paragraphs.

Mesenchymal Stromal Cells

The best approach would be the use of the patient’s own biological material to foster bone regeneration. MSCs are the precursors of bone-forming osteoblasts and thereby represent a potential cell population to improve bone regeneration under compromised healing conditions. Results obtained from several animal studies already confirmed the potential use of MSCs to enhance bone formation in general [48,49,50,51,52]. Intravenously injected MSCs were able to reach the site of injury, already implicating an attraction of bone-forming precursor cells to the fracture gap. Dreger et al. evaluated the competence of CD127-MSCs in a murine unilateral closed femur fracture model [49]. They showed that the time point of MSC application is critical for the success of the treatment outcome. One study analyzed the impact of MSCs in a non-union mouse model [50]. Expanded murine MSCs were used and an intravenous injection 24 h post-fracture led to improved bone formation thus restoring the impaired healing characteristics. In the human situation, the question arises where to get the MSCs from (bone marrow vs. adipose tissue derived MSCs) and how many MSCs would be needed to overcome impaired healing. Hernigou and colleagues analyzed the needed number of progenitor cells in human autologous bone marrow grafts for successful treatment of non-unions [53]. Bone marrow aspirates with > 1500 progenitors/cm3, corresponding to a total average number of 60,000 progenitor cells, were needed in order to overcome non-unions. However, in this study, not only osteoblast progenitor cells were present in the aspirates but also other, mononuclear cells, further supporting bone regeneration.

Bone Morphogenetic Protein

Bone morphogenetic proteins (BMPs) are a group of growth factors belonging to the TGF-β superfamily. BMPs are characterized by the stimulation of bone and cartilage tissue formation [54, 55]. Thus, BMPs are potent agents to foster bone regeneration under compromised conditions. In 2001, the FDA approved BMP7 for the treatment of long bone non-unions. One year later, in 2002, the FDA approved BMP2 for the use in tibia shaft fractures. Until today, there is no approval of BMP2 to treat non-unions. Both, BMP2 and BMP7, are the most reported BMP members for the treatment of bone injuries.

Govender et al. reported the positive impact of BMP2 in a prospective study with 420 patients having an open tibial fracture [56]. The patient group receiving the higher BMP2 dosage (total dosage of 12 mg) at the time point of trauma management showed significantly more healed fractures 12 months post-operative as well as less infection and hardware failure with regard to the control group. In contrast, Aro et al. published a higher infection rate in BMP2-treated patients with an open tibial fracture in comparison with the control group [57]. In addition, the authors did not find a significantly increased healing rate with BMP2. In the context of non-unions, Friedlaender et al. compared the outcome of BMP7 usage in a collagen type I carrier to the gold standard usage of autologous bone graft [58]. Both treatment strategies showed comparable healing rates (81% (BMP7) vs. 85% (bone graft)). This is in accordance with a report published by Desmyter and colleagues in 2008, reporting the healing rate of non-unions in Belgium after the use of BMP7 [59]. Giannoudis reported in 2009 a 100% healing rate of 45 patients with aseptic atrophic non-unions after the treatment of BMP7 together with bone autograft [60]. A case study including 175 patients compared the usage of BMP2 vs. BMP7 in 214 limb segment non-unions [61]. At an overall result, the BMP2-treated group displayed a better healing outcome (higher rate of radiographic healing and faster weight bearing) with regard to the BMP7-treated group. The BMP2 group showed a lower complication rate, although this was not statistically significant. Although a multitude of studies showed a positive impact of BMP2 and BMP7 in treating bone fractures and non-unions, also several unwanted side effects are reported for BMP2 in non-spinal bone applications [62]. This included local infections, wound complications, and heterotopic bone formation. However, the authors also stated that the reported observed side effects of BMP2 usage vary depending on the type and location of the fracture and the performed surgical procedure. The nature of fracture non-unions is very diverse and complex. The function and success after BMP treatment in bone healing is still under debate due to the consistently seen and reported site effects. Thus, using BMP for the treatment of bony non-unions is a risky option and can lead to more unwanted complications in the fracture treatment and thus to even higher costs and more pain and hospitalization time for the patient. In addition, BMP7 is no longer available for a clinical application has it has been withdrawn from the market [63].

Vascular Endothelial Growth Factor

Another possibility to foster bone regeneration is to stimulate angiogenesis, thus the rebuilding of the vessel network in the fracture area needed, i.e., sufficient nutrition supply [64]. The vascular endothelial growth factor (VEGF) is one of the main proteins stimulating vessel formation.

Animal studies already reported a positive effect of VEGF application in the treatment of bony non-unions [65, 66]. Although the experimental setup was different (VEGF application directly during setting the fracture [65] vs. in an already established non-union [66]), both studies showed improved healing outcome in comparison with the non-treated animal groups. Garcia and colleagues analyzed vascularization pattern in a non-union rat model and observed an even higher revascularization in the fracture zone of non-union animals in comparison with the normal healing group [67]. However, they found a decreased expression of pro-osteogenetic factors BMP2 and BMP4. Thus, the authors concluded that the ratio of expressed pro-angiogenetic and pro-osteogenetic factors could determin non-union formation. One study compared the application of BMP2, VEGF, and platelet-derived growth factor (PDGF) in an atrophic non-union model in rats [68]. Kaipel et al. used a silicon spacer for 4 weeks in order to impair the revascularization of the fracture area. After removing the spacer, growth factors were applied and included into a fibrin clot. The control group only received a fibrin clot. The PDGF and VEGF groups failed to increase bone formation over the observation duration. Whereas, the BMP group displayed increased bone regeneration. The time point of application could explain the non-functional treatment of the pro-angiogenetic factors. Both, PDGF and VEGF, mainly interact in the acute early fracture healing phase by promoting angiogenesis which is a pre-requisite for bone formation. However, Kaipel and colleagues applied these factors 4 weeks after the setting of the fracture and further impaired the healing cascade due to the silicon spacer. Therefore, the administration of PDGF and VEGF could have at least partial anti-osteogenetic effects when used in the later healing phases. This result is even more important for a potential clinical application.

Platelet-Rich Plasma

Another biological material stimulating angiogenesis is platelet-rich plasma. Activated platelets are known for the secretion of growth factors stimulating bone regeneration [69]. One of these growth factors is PDGF. For PDGF, it was already shown that it acts on human osteoblasts and stimulates their proliferation [70]. In the context of non-unions, Labibzadeh et al. and Centeno et al., respectively, reported two small case studies with human fracture patients with long bone non-unions [71]. In these studies, the combination of platelet lysate and autologous MSC application for non-union treatment was analyzed. At least for the study reported by Labibzadeh, all analyzed patients tolerated the treatment approach of the cell-lysate complex well. In four out of seven (Labibzadeh) and four out of six (Centeno), respectively, it even led to a bridging of the non-union gap after 12/6+ months (Labibzadeh/Centeno). For both studies, one explanation for the non-responder could be the long duration of the non-union from 16 months to several years. Although both studies used different platelet lysates (autologous (Labibzadeh) vs. allogenic (Centeno)) and different expansion time and implanted passage of cultured MSCs, both studies demonstrated the potential and feasibility of using both, platelet lysate and autologous, culture expanded MSCs, in combination to treat non-unions. However, the used MSCs were isolated out of bone marrow aspirate obtained from the iliac crest. This means a second surgical intervention with a potential risk of an infection and further pain for the patient. In 2016, a one-patient study reported the sole use of autologous platelet lysate to treat a non-healing tibia and fibula fracture that failed the initial stabilization by an internal metal plate [72]. Injection of autologous platelet lysate led to bony union after 8 months post-injection without any signs of discomfort for the patient or infections or other complications. Based on the reported findings in the literature, autologous platelet lysate represents a potential option treating bone defects without any reported negative side effects so far.

Low-Intensity Pulsed Ultrasound

Low-intensity pulsed ultrasound (LIPUS) is reported as a possible non-invasive technique to stimulate bone formation (company site: http://www.exogen.com/). This method is based on the emission of ultrasound waves through the skin to the fracture site. These waves will then stimulate the activity of cells and thus foster bone regeneration. In vitro analysis already showed that the application of low-intensity ultrasound stimulates periosteal cells as well as MSCs to proliferate and to differentiate into the osteogenetic lineage [73, 74]. Furthermore, it was already shown that LIPUS enhances the migration of MSCs to the fracture site. However, this was not significant in comparison with the control group. Zura et al. reported a case study including 767 patients with established non-unions from more than 1 to 10 years. LIPUS-treated patients showed a healing rate of 86.2%. This is in accordance with published findings from other studies, stating an increased healing rate of delayed healing or non-union fractures [75,76,77,78]. Although there exist a multitude of reported cases showing the positive outcome after LIPUS treatment, one important disadvantage becomes obvious after having a closer look into the patient selection of the studies. Often, complicated delayed or non-healing cases were not included into the study design, thus falsifying the real existing patient cohort suffering from impaired bone healing. Furthermore, a control group receiving standard surgical procedures does not accompany the follow-up of LIPUS-treated patients. Thus, the outcome-healing rate after LIPUS application cannot be compared relative to standard treatment approaches for non-unions. However, the National Institute for Health and Care Excellence (NICE) in the UK published a cost saving of around $1726 per patient when using LIPUS instead of surgical intervention to treat non-union [79]. LIPUS seems to be a promising and potential alternative to treat non-invasively impaired healing fractures. However, clinical trials including a broader patient cohort and well-defined control groups are still missing and have to be initiated and done in order to correctly judge and interpret the observed putative beneficial results after LIPUS treatment.

Future Innovative Treatment Approaches Conducted in our Institute

Immunotherapy

The immune system plays a key role in bone regeneration [33]. The interdisciplinary research field “osteoimmunology” combines both research areas, the bone and the immune system [80]. The bone and the immune system share a multitude of factors and molecules and thus are interdependent on each other, meaning, intervention in the one system will also influence the other one. Therefore, immune therapy represents another promising research approach to overcome impaired healing. Terminally differentiated CD8+ effector memory T cells (TEMRA), a subpopulation of the T cell compartment of the adaptive immunity, could be a promising biomarker determining impaired healing fractures. We already showed a positive correlation between a higher amount of pro-inflammatory CD8+ TEMRA with a poorer healing outcome after closed tibia head fracture [81••]. This has been accounted to a prolonged pro-inflammatory state in the fracture area due to the higher CD8+ TEMRA population. CD4+ regulatory T cells (Treg), another subset of the adaptive immunity, could be one possible counterpart to CD8+ TEMRA cells. An in vivo study in a murine caldaria defect model demonstrated the pro-osteogenetic effect of Treg application together with bone marrow MSCs. We also analyzed the effect of Treg in the treatment of a fracture in long bones. Our results support the positive modulating impact of Treg in bone regeneration, whereas this was dependent on the status of the adaptive immune system of the recipient (manuscript in preparation). Our observation further highlights the sensitive and complex interplay between the bone and the immune system in the context of immune therapy. Besides the adaptive immunity, also macrophages as part of the innate immunity play a crucial role in bone regeneration. Stimulating macrophages in the “alternative” M2 lineage at the time point of fracture by the administration of IL-4 and IL-13 led to improved healing outcome with regard to the control group in a murine osteotomy model [82].

Immune therapy in bone-related diseases comes more and more into the clinical focus. However, we are still at the beginning to understand the mechanism how immune cells interact in the bone systems and vice versa, patient-based immunotherapeutical approaches are a very promising treatment approach to determine and overcome impaired healing outcome.

Mechanotherapy

Another direction seeing reasonable considerations are approaches using mechanobiological cues to induce regeneration and considered to enable “mechanotherapy” in musculoskeletal healing [83]. With the help of patient data, pre-clinical animal models and computational models, patient-specific mechanical constrains can be modulated at bone defects, fracture zones, or joints [84]. Such mechanobiologically optimized conditions or the active mechanical stimulation allow to enable endogenous regenerative cascades to overcome impaired bone regeneration.

Conclusion

Looking for actually registered clinical trials corresponding to the keyword “non-union” in the US and EU clinical trial registers (www.clinicaltrials.gov and www.clinicaltrialsregister.eu) revealed 124 (US) and 8 (EU) hits, respectively. Among the listed clinical trials, 15 (US) and 4 (EU), respectively, evaluate the impact of autologous stem cells, mostly MSCs, in bony non-unions. A few studies apply for low-intensity ultrasound or electromagnetic field stimulation to foster bone healing (7/124). Only six trials use BMP2 (2/124), BMP7 (1/124; 1/8), or platelet lysates (2/124). The other listed clinical trials deal with (autologous) bone grafts, different fixation, and screw types, the effect of vitamin supply and fluid lavage or reaming-irrigator aspirates (RIA). The wide and diverse range of the at the-moment registered clinical trials concerning non-unions highlights the complexity of the nature of a non-union and therefore the multitude of treatment options. It further shows even more the importance of ongoing research in order to (1) understand the underlying pathomechanisms leading to the formation of a non-union and to (2) develop appropriate treatment strategies to improve the patient’s life and simultaneously to decrease the burden for the socio-economic system.