An Overview of Drug Effects on Bone Healing on Animal Research Models

Bone fracture is a common health problem in humans and animals, and the healing of the bone fracture is a complicated process. Several drugs may be used concurrently with the treatment of fractures, but they may interfere with the healing process of the bone. The present research reviewed previously published studies with the objective to enhance the understandings of the effects of different drugs on bone healing. There is clear evidence that antibiotics, corticosteroids, nonsteroidal inflammatory drugs, and chemotherapeutic drugs all affect bone healing. By contrast, the effect of anticoagulants on bone healing is controversial, so more research is needed to determine its efficacy. In addition, there is no direct evidence to approve the effect of anesthetics on bone healing, so this is another area in need of further research.


Introduction
Bone fracture causes harmful effects to patients' health, degrades their quality of life, and is responsible for costly treatments. Annually, millions of musculoskeletal procedures are performed worldwide (Jahangir et al., 2008). Each year, billions of dollars have been spent for the treatments of bone fracture and hip replacement, and the number has been increasing over time. In 2005, more than USD 20 billion was spent on medical care for over 300,000 spinal fusions (Porter et al., 2009). Pathogenic bone defects caused by cancer were responsible for more than 3,000 pediatric hospitalizations, and cost over USD 70 million (Porter et al., 2009).
In animals, about 800,000 pets are involved in vehicle crashes annually in USA (Strickland, 2014). High rise syndrome may cause limb fracture in 46.2% of fallen cats, of which 38.5% of fractures are in forelimbs and 61.5% were in the hind limbs are. Tibiae are most commonly fractured followed by the femur (Vnuk et al., 2004). The treatment cost of bone fractures in companion animals can procure several million Vietnam Dong per case.
Antibiotics, anti-inflammatory drugs and anesthetics are usually used during treatment of bone fracture. Sometimes, anticoagulants or chemotherapy drugs may be applied in orthopaedic animals treatments. All of these drugs may have some effects on the course of bone healing. The pharmacological factors associated with bone healing in humans and animals are similar in many aspects; therefore, understanding these factors is useful and can enhance the successful treatment. This narrative review focuses on the effects of drugs used on the bone healing process based on the information retrieved from an extensive literature review.

Effects of Drugs on Bone Healing
Prevention and/or treatment of diseases may involve many types of drugs. They can be generally divided into 6 groups including antibiotics, corticosteroids, non-steroidal antiinflammatory drugs (NSAIDs), anticoagulants, chemotherapeutic agents and anesthetics. When used in animals with bone fractures, some of these drugs may interfere with the bone healing process. Tables 1-5 present the effects of such drugs on fracture healing in animal models.

Effects of Antibiotics on Bone Healing
Many antibiotics have been reported to affect bone healing including quinolones (Gough et al., 1996), ciprofloxacin (Huddleston et al., 2000), levofloxacin and trovafloxacin (Perry et al., 2003), tetracycline (Kim et al., 2004) and cefuroxime (Natividad-Pedreño et al., 2016). Effect of gentamicin on bone healing is controversial. Kim et al. (2004) reported that it decreased bone formation, however, Fassbender et al. (2013) and Haleem et al. (2004) did not find similar results. Cefazolin had no effect on fracture healing parameters including callus mechanical resistance and histological scores (Natividad-Pedreno et al., 2016). Also, vancomycin did not influence biomechanical and radiographic scores of healing bone (Haleem et al., 2004). Interestingly, doxycycline inhibited osteolysis, and is suggested for prevention and treatment of wear particle-induced osteolysis and aseptic loosening (Zhang et al., 2007).
The negative effects of antibiotics on bone healing may be due to their high doses such as 100 mg/kg/day for cefuroxime (Natividad-Pedreno et al., 2016) and ciprofloxacin (Huddleston et al., 2000), 70 mg/kg/day for trovafloxacin and 50 mg/kg/day for levofloxacin (Perry et al., 2003), and due to long treatment durations, such as 3-4 weeks (Huddleston et al., 2000;Perry et al., 2003;Natividad-Pedreño et al., 2016). Antibiotics play a pivotal role in prevention and treatment of bacterial infection in many diseases and procedures, especially, orthopedic cases. The use of antibiotics is inevitable as long as their doses and durations are carefully calculated and followed.

Effects of corticosteroids on Bone Healing
In vivo studies have demonstrated that many corticosteroids including cortisone (Sissons & Hadfield, 1955), prednisolone (Luppen et al., 2002;Yaghini et al., 2017), prednisone Waters et al., 2000), methylprednisolone (Xie et al., 2011) and dexamethasone (Sawin et al., 2001) affected the bone healing process. The common aspect of these studies wa that the duration of treatments was prolonged, i.e., 42 days for dexamethasone (Sawin et al., 2001), 56-78 days for prednisolone (Luppen et al., 2002;Yaghini et al., 2017) and 96 days for prednisone Waters et al., 2000). On the other hand, when used for a shorter duration, i.e., 14-21 days, corticosteroid doses were very high, i.e., 20mg/kg/day for methylprednisolone (Xie et al., 2011), and 10-20 mg/kg/day for cortisone (Sissons & Hadfield, 1955). With a similar duration of treatment (21 days) but several times lower doses (0.5 mg/kg/day), no significant effect of prednisolone on fracture healing was observed although the numeric values of the mechanical parameters were lower in the treated group . Being used for 4 consecutive days, methylprednisolone at 2 mg/kg/day (Hogevold et al., 1992) and prednisone at 0.02 mg/kg/day (Aslan et al., 2005) had no effect on bone healing. Rat calvaria Antibiotics had no effects on biomechanical and radiographic scores.

Vietnam Journal of Agricultural Sciences
Haleem et al.
Levofloxacin: 25 mg/kg, trovafloxacin: 35 mg/kg. Both drugs were used twice a day for 3 weeks, started 1 week post femoral fracture.
Rat femur Antibiotic induced less woven bone and more cartilage in the fracture sites. Perry et al. (2003) Ciprofloxacin: 50 mg/kg, twice a day for 3 weeks, 1 week post femoral fracture.

Rat femur
Decreased radiographic results, lowered torsional strength, and abnormal cartilage morphology were observed in the treatment group.
Huddleston et al.
Rabbit joints: shoulder, hip, knee Drug caused degenerated or hypertrophic chondrocytes, loss of collagen and proteoglycan. Effects were not clearly doserelated. Gough et al. (1996) The negative effect of corticosteroids on bone healing was consistent in the literature, especially if the high doses and/or long treatments were applied. Nonetheless, there is also some evidence that corticosteroids can be used for a short time with appropriate doses if their benefits outweigh their potential side effects.

Rat femur
No significant effect of prednisolone on bone healing was evident although the absolute values of mechanical parameters in prednisolone group were lower in comparison with the control.
Bissinger et al.
Rabbit femur Methylprednisolone induced osteonecrosis, impaired healing and maturation of bone tissue. Xie et al. (2011) Prednisone: 0.02 mg/kg, 4 consecutive days started prior to surgery.

Rat femur
Prednisone had no effect on bone healing. Aslan et al.
Prednisolone: 0.35 mg/kg per day, three times a week for 6 weeks before surgery.

Rabbit ulna
Prednisolone inhibited bone healing characterized by a small callus area, low torsional strength.
Dexamethasone: 0.05 mg/kg per time, twice every day for 42 days since surgery.
Rabbit spine Dexamethasone decreased the rate of bone graft union.
Sawin et al.

Rabbit ulna
Prednisone induced bone loss. Bostrom et al.
Prednisone: 0.15 mg/kg per day, 60 days before osteotomy to 6 weeks post-surgery.

Rabbit ulna
Prednisone caused lower results in callus size, radiographic density, bone mineral content and mechanical strength.
Waters et al.

Rabbit femur and tibia
Longitudinal bone growth instantly stopped after the commencement of cortisone administration. The cartilage was thinned as soon as day 6. By day 24 extensive destruction of metaphyseal trabeculae was apparent. Sissons & Hadfield (1955) Cortisone: 20 mg/kg per day for 21 days.

Rat femur and tibia
The harmful effect of cortisone on bone growth in rats was less severe than in rabbits.
Although there are a few studies supporting the use of some NSAIDs, such as ketorolac (Cappello et al., 2013;Fracon et al., 2010), paracetamol, etoricoxib (Fracon et al., 2010) during bone healing treatment, there are far more studies reporting the newgative impact of bone healing. Therefore, use of NSAIDs during fracture healing should only be considered when the drugs' benefits over-ride their negative impacts ( Table 3).

Effects of anticoagulants on Bone Healing
The use of anticoagulants for prevention of thrombosis and pulmonary embolism in traumatic and orthopedic cases is common . Anticoagulants may elicit inhibitory effects on osteoblast formation, and may intensify bone resorption (Kapetanakis et al., 2015). Enoxaparin has been found to exert a negative effect on fracture healing in rabbit ribs (Street et al., 2000, whereas this drug does not influence bone healing of rat femur (Curcelli et al., 2005;Demirtas et al., 2013;Say et al., 2013). Bone fracture healing is independent of many anticoagulants, including heparin (Curcelli et al., 2005;Erli et al., 2006), dalteparin (Erli et al., 2006;Hak et al., 2006;Say et al., 2013), certoparin (Erli et al., 2006), nadroparin (Say et al., 2013), and rivaroxaban (Demirtas et al., 2013). Rivaroxaban increased the callus volume but decreased bone mineral density resulting in unchanged mechanical parameters (Kluter et al., 2015;Prodinger et al., 2016). It appears that negative effects of anticoagulants on bone fracture healing is still minimal. A definitive conclusion of the effect of anticoagulants on bone healing is impossible discern due to the differences in drugs used, time of drug exposure, drug doses, animal models, fractured bones, time of evaluation and means of evaluation. Therefore, these drugs are still a gold standard for prevention of thrombosis and pulmonary embolism in animals with bone fracture(s). Nevertheless, more well-designed studies are required to determine the true effects of anticoagulants on bone fracture healing ( Table 4).

Effects of chemotherapeutic agents on Bone Healing
Chemotherapeutic agents are commonly prescribed for treatment of cancers and chronic inflammation. High doses are recommended for oncological treatment, and low doses are for chronic inflammation (Cavalcanti et al., 2014). Results show that these drugs are also significantly involved in the bone healing process. Fracture healing was completely prevented by angiogenesis inhibitor TNP-470 via suppression of both intramembranous and endochondral ossifications (Hausman et al., 2001).
Histological and radiographical Incidence of nonunion in rofecoxib and ibuprofen groups was higher than that in the control group. Leonelli et al. (2006) Indomethacin: 1 mg/kg per day for 4,8,12 weeks. Celecoxib: 3 mg/kg per day for 4,8,12 weeks.

Rat femur
Both indomethacin and celecoxib delayed bone healing at 4 and 8 weeks post operation.
Brown et al.
Ketorolac: 4 mg/kg per day, for either 21 or 35 days. Parecoxib: 0.3 and 1.5 mg/kg per day for either 21 or 35 days.

Rat femur
Both parecoxib and ketorolac impaired bone healing, but ketorolac expressed more harmful effects. Gerstenfeld et al. (2003) Etodolac: 20 mg/kg per day for 3 weeks started from surgery until day 21, or for 1 week started from surgery until day 7, or for 1 week started from day 14 until day 21.

Rat femur
In all groups of treatment, radiographic and mechanical testing showed lower scores which predisposed fractures to delayed healing.
Endo et al.

Rat femur
Indomethacin impaired bone healing. This effect was more severe when fracture healed under unstable condition. Hogevold et al. (1992) assessments revealed that doxorubicin and cisplatin affect both quantity and quality of the healing bone of rabbit segmental radial diaphyseal fractures (Morcuende et al., 2004). Even a single dose of doxorubicin at the time of surgery significantly inhibited the process of spinal fusion (Tortolani et al., 2004). Similarly, a single dose of methotrexate at either 30 or 250mg at surgery decreases bone formation resulting in impaired bone healing (Cavalcanti et al., 2014;Satoh et al., 2011). By contrast, methotrexate at a low dose (3mg every week, for 1-4 weeks) did not delay fracture healing process (Cavalcanti et al., 2014;Satoh et al., 2011). The effect of chemotherapeutic agents on bone healing may depend on drug types, and the doses of drugs. Although the effects of chemotherapeutic agents on bone healing are apparent, their use in some cases, including cancer, can not be avoided. In such cases, adjunctive treatment such as growth factors may be considered to reduce the drugs' negative effects and normalize bone healing process ( Table 5). Rivaroxaban increased callus volume and induced a marginal increase in bone mineral density, but torsional rigidity was not affected by the treatment.

Rabbit femur
Histological examination revealed that there was no significant effect of anticoagulants on bone healing. Erli et al. (2006) Dalteparin: 70 unit/kg per day for 14 days. Rat femur Dalteparin did not alter radiographic, histological and, mechanical results. Hak et al. (2006) Heparin: 400 IU/kg per day, enoxaparin: 2 mg/kg per day for 28 days.
Rat femur Heparin and enoxaparin had no effect on histological and mechanical grades. Curcelli et al. (2005) Enoxaparin: 2 mg/kg per day for either 3, 7, 14, or 21 days. Rabbit rib Inhibitory effect of enoxaparin on bone healing was detected at all times of evaluation.
Street et al.

Effects of Anesthetics on Bone Healing
Anesthetics are routinely utilized in many surgical procedures, particularly in orthopedics. Besides their anesthetic properties, they also have anti-inflammatory characteristics. Isoflurane reduces the systemic release of TNFα and IL-1β (Flondor et al., 2008). TNF-α, IL-6 and IL-8 production from human whole blood culture was also suppressed by ketamine (Kawasaki et al., 1999). Furthermore, ketamine lowers TNF-α activity and mortality of carrageenan-sensitized endotoxemic rats (Koga et al., 1994), and decreases TNF-α production in LPS-induced endotoxemia in mice (Taniguchi et al., 2001). Interestingly, ketamine did not exert those effects if entotoxins were absent. The mechanism of ketamine-induced antiinflammation is believed to be partially due to its effect on cyclooxygenase activity (Suliburk et al., 2005). Furthermore, in rats, the antiinflammatory effect of ketamine was seen only at sub-anesthetic dosages (0.5-5 mg/kg), whereas at high dosage (50 mg/kg) it enhanced the TNF-α production (Sun et al., 2004). Although there are no direct results showing that anesthetics influence bone fracture healing, their effects on inflammatory cytokines are apparent. The bone healing process is, on the other hand, involved in the inflammatory mechanism. Therefore, it may be hypothesized that some anesthetics such as isoflurane and ketamine may influence bone fracture healing (Histing et al., 2011). Because the true effect of anesthetics on bone healing has not been determined to date, their use still contributes to the success of millions of orthopedics procedures worldwide.

Conclusions
This overview showed that the course of bone healing may be influenced by several drugs and treatment regimes. However, the negative effects of several antibiotics and steroids on bone healing may be due to their high doses and/or long treatments because such effects were absent when normal doses and treatment periods were applied. The use of such drugs in partients with bone fracture(s) is reasonable if the normal doses and durations of treatments are followed. In contrast, many non-steroidal inflammatory drugs may affect bone healing even at their standard treatment usage. Hence, these drugs should not be prescribed for orthopedic animals treatments. There is a large body of research demonstrating negative effects of chemotherapy agents on bone healing. However, a low dose of methotrexate may be appropriate for animals that have cancer or chronic inflammation and acquire bone fracture(s). More well-designed studies are needed before a conclusion on the effect of anticoagulants and anesthetics on bone healing process can be ascertained. The use of some drugs is inevitable during the treatment of bone fractures, therefore, understandings of their effects on bone healing are imperative to increase the chances of successful treatments.