THE EFFECTS OF CLODRONATE DISODIUM ON EQUINE JOINT TISSUES AND OSSEOUS METABOLISM: FROM IN VITRO ANALYSIS TO A PRE-CLINICAL, OVINE MODEL UNDER EXERCISE By Fernando Benjamin Vergara Hernandez A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Animal Science – Doctor of Philosophy 2023 ABSTRACT Bisphosphonates (BPs) are commonly used drugs for managing bone loss or bone resorption in skeletal diseases, such as post-menopausal osteoporosis, Paget’s disease, and bone cancer. In 2014, the FDA approved clodronate disodium (CLO) and tiludronate disodium for treating navicular syndrome in horses over 4 years old. However, concerns have arisen regarding the extra-label use of CLO, particularly in juvenile individuals subjected to exercise, such as racehorses, where BPs may impact bone metabolism by affecting bone modeling/remodeling. The overall objective of this dissertation was to determine the effects of CLO in equine joint tissues in vitro and CLO effects in vivo in juvenile, exercising sheep as a model for juvenile horses. A prior publication found that CLO reaches the synovial fluid when administered intramuscularly. Therefore, an initial in vitro study exposed equine cartilage explants, chondrocytes, and synoviocytes to recombinant equine interleukin-1β (reqIL-1β) to determine the effects of CLO in joint tissues. The results confirmed that reqIL-1β increased the release of inflammatory markers from joint tissues (e.g., GAG, IL-6, PGE2, NO, P < 0.05), yet CLO did not reduce the inflammatory effects of reqIL-1β. Hence, this in vitro study established that CLO was neither cytotoxic nor cytoprotective to joint tissues. Sheep have served as models for assessing the impact of BPs in human medicine and have also been used as a suitable model for studying exercising horses. Before using sheep as a model for intramuscular CLO administration, we determined the pharmacokinetics (PK) and plasma protein binding (PPB) of CLO in sheep and compared them to horses. Sheep PK parameters were similar to those previously published for horses, particularly Cmax and AUCall. Unbound fractions of CLO differed by less than 1.4-fold between sheep and horses. Having established a dose that resulted in PK values similar to those in horses, we subjected the sheep to a novel exercise protocol, using a high-speed exerciser. The sheep adapted quickly to the exercise protocol and tolerated it well with no evidence of significant lameness. Finally, the juvenile sheep were divided into four groups: CLO on day 0, CLO on day 84, CLO on days 0 and 84, and a control group (saline). Physical examinations and lameness evaluations were measured every 14 days and blood was collected every 28 days for the measurement of serum bone biomarkers. Bone was harvested from the tuber coxae halfway through the study and again at euthanasia. In addition, samples of lumbar vertebrae and fused metacarpal III+IV were obtained at euthanasia for bone microstructure analysis and biomechanical testing. No measurable effects of CLO on the sheep skeletons were detected. Serum bone formation markers, bone-specific alkaline phosphatase (BALP) and procollagen type I amino-terminal propeptide, increased over time. Serum bone resorption marker, carboxy-telopeptide of type I collagen cross-links (CTX-I), decreased at several time points consistent with the exercise stimulus. Male sheep had decreases in compressional stress and increases in modulus of elasticity of the fourth lumbar spine in comparison to female sheep, likely due the low levels of sex hormones in the castrated males. The relatively low dose of CLO used in large animals compared to humans may explain the lack of skeletal effects. Previous BPs studies on horses have reported improvements in lameness without evidence of reduced bone resorption, suggesting that analgesic effects may occur without significant changes in bone microstructure or serum markers. Future research should focus on the potential analgesic effects of CLO at low doses which could provide palliative care without significant effects on bone metabolism. Copyright by FERNANDO BENJAMIN VERGARA HERNANDEZ 2023 This dissertation work is dedicated to David’s Shepherd. v ACKNOWLEDGEMENTS First, I would like to thank the Fulbright Foreign Student Program and the Chilean National Agency for Research and Development (ANID) [56150020] for their funding and orientation prior to and during my stay in the U.S. Without their support, I would not have been able to imagine studying abroad. I am also grateful to the National Institute of Food and Agriculture – United States Department of Agriculture [2021-67015-34079] and the American College of Veterinary Surgeons for funding the research presented in this dissertation work. I extend my sincere gratitude to Dr. Brian Nielsen for his constant support, always accompanied by a warm smile, and for challenging me throughout my studies. I am also grateful for the invaluable feedback, scientific support, and engaging discussions provided by Dr. Aimee Colbath. Both of your contributions have played a significant role in shaping me into a better scientist. I would like to express my appreciation to my committee members: Dr. Daniel Buskirk, especially for your emotional support during my program; Dr. Richard Ehrhardt, for your hands- on assistance and technical expertise with sheep; and Dr. Jessica Leatherwood, for your kindness and tireless efforts in organizing and leading the entire USDA team, which secured funding for a big portion of this research work. Additionally, I extend my gratitude to my “honorary co- advisors”: Dr. Jack Kottwitz, for your deep understanding and continual assistance in navigating the world of pharmacokinetics; and Dr. John Popovich Jr., thank you for training me and for your full support with the micro-CT image analysis. Your dedication and guidance have greatly contributed to my growth as a scientist. I am also immensely grateful for Dr. Cara Robison. Thank you for your unwavering support, guidance, and the laughter shared over language jokes. Without your help, I might not have been able to continue in this program. Additionally, I want vi to express my gratitude to Char Panek for the countless hours of lab work and your enduring patience in trying to understand me despite my broken English. Furthermore, I would like to acknowledge the support of the Department of Animal Science, in particular, I am grateful to Dr. Ernst, Dr. Siegford, and Karla Macelli. I deeply appreciate the assistance of several undergraduate students who played a significant role in data collection during this dissertation work. My thanks go to Rachel Postello, Julia Baker, Taylor Collier, Faith Kurtz, Jenna Darling, Nick Peters, Bailey Curtis, Alondra Gallego, Nicole Hamlin, Gretel Keller, and Rebekah Agnew. I would also like to thank my family in Chile, especially my niece and nephew, Florencia and Julian, my sister Marcela, and my mom, Gloria. Your unconditional support and love have allowed me to pursue further studies by leaving Chile. I want to express my gratitude to my family and friends that I have made here in the U.S. A special thanks to my ΧΑ Christian fellowship, particularly Kate, Janet, Courtney. To Doug and Kathie, who became my parents in a foreign land; thank you for your unwavering love, support, and constant prayers. I also extend my love to my brothers Dan and Brian and their families. To Grandpa and Grandma Kruithof, thank you for your abounding generosity in prayers and otherwise. Lastly, I reserve this paragraph for my beloved wife, Keri. Our journey of getting to know each other and becoming husband and wife has been incredible. I could never thank you enough for your direct and indirect support, including reading and proofreading my drafts, standing by my side for countless hours, and sacrificing so much of your time for me. Meeting you is a blessing I cannot fully express. You have played an integral part in this work, as you are an essential part of my life and my heart. Te amo! vii TABLE OF CONTENTS CHAPTER 1: Is the use of bisphosphonates putting horses at risk? An osteoclast perspective .... 1 CHAPTER 2: Clodronate disodium is neither cytotoxic nor cytoprotective to normal and recombinant equine interleukin-1β-treated joint tissues in vitro .................................................. 16 CHAPTER 3: Pharmacokinetics and plasma protein binding of a single dose of clodronate disodium are similar for juvenile sheep and horses ...................................................................... 18 CHAPTER 4: Exercising sheep as a novel pre-clinical model for musculoskeletal research ...... 33 CHAPTER 5: Clodronate disodium does not produce measurable effects on bone metabolism in an exercising, juvenile, large animal model .................................................................................. 49 CHAPTER 6: Conclusions ........................................................................................................... 80 LITERATURE CITED ................................................................................................................. 83 APPENDIX ................................................................................................................................. 105 viii CHAPTER 1: Is the use of bisphosphonates putting horses at risk? An osteoclast perspective F.B. Vergara-Hernandez, B.D. Nielsen, A.C. Colbath, Is the use of bisphosphonates putting horses at risk? An osteoclast perspective, Animals. 12 (2022) 12131722. https://doi.org/10.3390/ani12131722. SIMPLE SUMMARY Bisphosphonates are a group of drugs that intervene in the bone resorption process, producing cellular death of osteoclasts. These drugs are used for skeletal conditions, such as osteoporosis in humans and are available for veterinary medical use. Clodronate and tiludronate are bisphosphonates approved for the treatment of navicular syndrome in horses greater than four years old. However, these drugs are sometimes used in juvenile animals under exercise, where osteoclast activity is higher. Bisphosphonate use in juvenile and/or exercising animals could have adverse effects, including maladaptation to exercise or accumulation of microdamage. Furthermore, bisphosphonates can be bound to the skeleton for several years, resulting in a prolonged effect with no pharmaceutical reversal available. This review presents an overview of osteoclast function and a review of bisphosphonate characteristics, mechanisms of action, and side effects in order to con-textualize the potential for adverse/side effects in young or exercising animals. ABSTRACT Osteoclasts are unique and vital bone cells involved in bone turnover. These cells are active throughout the individual's life and play an intricate role in growth and remodeling. However, extra-label bisphosphonate use may impair osteoclast function which could result in skeletal microdamage and impaired healing without commonly associated pain, affecting bone remodeling, fracture healing, and growth. These effects could be heightened when administered 1 to growing and exercising animals. Bisphosphonates (BPs) are unevenly distributed in the skeleton; blood supply and bone turnover rate determine BPs uptake in bone. Currently, there is a critical gap in scientific knowledge surrounding the biological impacts of BP use in exercising animals under two years old. This may have significant welfare ramifications for growing and exercising equids. Therefore, future research should investigate the effects of these drugs in skeletally immature horses. INTRODUCTION Since their discovery in 1873, osteoclasts have been recognized for their bone resorption ability [1]. Osteoclasts’ resorption ability makes them a key cell for musculoskeletal development, bone metabolism, and bone repair throughout life [2]. Due to their intricate involvement in physiological processes, substantial pathologies are associated with overactive or impaired osteoclast function. In conditions where osteoclast resorption overcomes bone formation, it can be useful to decrease osteoclast activity through pharmaceutical interventions. Bisphosphonates (BPs), a class of drugs known to impair osteoclast function, have been available for human skeletal conditions for over 50 years and have been available in veterinary medicine for about 25 years [3]. In 2014, two bisphosphonates (Osphos® and Tildren®) were approved by the United States Food and Drug Administration (FDA) to treat navicular syndrome in horses over four years of age [4]. However, it is unknown how these drugs affect juvenile animals under exercise, where skeletal adaptations depend upon normal bone metabolism and normal osteoclast function [5]. The objective of this review is to explore the available scientific literature regarding the origin of osteoclasts and their functions, how BPs affect these cells, the current use and side effects of BPs in humans and other animal models, and the potential negative effects of BPs use 2 in the juvenile horses under exercise. ORIGIN OF OSTEOCLASTS Osteoclasts precursors are found in the bone marrow and originate from the myeloid lineage [6]. Key molecules involved in osteoclastogenesis include members of the tumor necrosis factor family α (TNFα) such as the receptor activator of nuclear factor (NF)-ĸB (RANK), RANK ligand (RANKL), osteoprotegerin (OPG), and the macrophage-colony stimulating factor (M-CSF) [7]. In the presence of M-CSF and other growth factors such as interleukin-3 (IL-3), osteoclast precursors proliferate and become preosteoclasts. These preosteoclasts fuse and generate mature multinucleated osteoclasts [8]. OPG serves as a negative feedback molecule for osteoclastogenesis, decreasing the RANKL function and osteoclast differentiation [6]. Osteoclastogenesis and subsequent osteoclast-mediated resorption are necessary for skeletal development, bone remodeling, and bone repair. Disruptions in osteoclastogenesis can lead to disease processes ranging from osteoporosis to osteopetrosis [8– 10] A summary of the main molecules involved in osteoclastogenesis are presented in Table 1. Table 1. Summary of osteoclastogenesis molecules, origin, and functions. Molecules essential for osteoclastogenesis include: receptor activator of nuclear factor (NF)-ĸB ligand (RANKL), osteoprotegerin (OPG), macrophage-colony stimulating factor (M-CSF). Molecules Origin RANKL Bone marrow-derived stem cells, osteoblasts, osteocytes OPG Osteoblast and osteocytes M-CSF Bone marrow-derived stem cells, osteoblasts, osteocytes Function Primary differentiation factor controlling gene expression binding to RANK [11,12] Decoy receptor for RANKL competing with RANK. Blocks RANKL-RANK interaction [11] Activates pathways stimulating proliferation and survival by binding to macrophage colony-stimulating factor 1 receptor (CSF-1 R/c-Fms)[11,12] OSTEOCLASTS ARE CRITICAL FOR BONE MODELING AND REMODELING Osteoclast morphology plays an important role in their function; osteoclasts are multinucleated cells with an active “ruffled membrane” where resorption occurs [13]. When adjacent to the bone, the finger-like extensions of the cytoplasm adjacent to the bone produce a 3 microenvironment through a proton pump, acidifying and demineralizing the bone matrix [14]. Alterations in osteoclast function can lead to critical bone diseases including osteopetrosis [15] but may also have more subtle effects on bone modeling and remodeling. Bone is a connective tissue with multiple roles, including mechanical support, protection, locomotion, mineral homeostasis, and endocrine functions [12,16,17]. Bones can be erroneously perceived as static tissue when, in fact, bones are constantly adapting to strain [14,18]. Bone adaptation is driven by two processes: bone modeling and remodeling [2]. Bone modeling is the process of mineral uptake and removal in growing organisms, leading to bone maturation. Meanwhile, bone remodeling is a process that takes place throughout the life of organisms as the bone adapts to new mechanical loads and repairs microdamage, allowing the bone to reach a proper geometry [19]. In bone remodeling, bone resorption and bone formation are tightly coupled. Osteoclasts are recruited and activated, resulting in bone resorption and then undergoing apoptosis. Then, osteoblasts produce a new organic bone matrix, followed by mineralization [19]. OSTEOCLAST-MODIFYING DRUGS: BISPHOSPHONATES Bone modeling and remodeling can be affected by pharmacological interventions including bisphosphonates. Bisphosphonates (BPs) vary in their chemical structure, complexity, and binding capacity to bone tissue.[20] BPs are chemically stable analogs of inorganic pyrophosphates (PPi). BPs are not subject to enzymatic hydrolysis, are resistant to high temperatures, are not biodegradable [21,22], and have a high affinity to bone hydroxyapatite (HAP) [20] (Figure 1). 4 Figure 1. Comparison of pyrophosphate and basic bisphosphonate structures. Bisphosphonates differ from pyrophosphates primarily by the change of oxygen from their central atom to carbon, providing resistance to biological degradation. Bisphosphonates have been classified into “generations” and mechanisms of action: first- generation BPs (i.e., clodronate), second-generation BPs (i.e., tiludronate and alendronate), and third-generation BPs (i.e., ibandronate, and zoledronate). Newer generations have greater antiresorptive capabilities, which allows for a lower dose administration [22]. BPs are also classified according to their mechanism of action: simple or non-nitrogen-containing BPs (sBPs) and nitrogen-containing BPs (nBPs). Both BP types interfere with the osteoclast activity but with different relative potency [23]. The R2 side chain of BP structure determines the biological activity (Figure 1), and the presence of nitrogen atoms in the R2 side chain has shown a larger antiresorptive effect (Figure 2) [24]. Simple BPs (i.e., clodronate and etidronate) are metabolized in the cytoplasm and generate cytotoxic ATP analogs (5- [β, γ-dichloromethylene] triphosphate) resulting in apoptosis due to a lack of free/functional ATP for cellular enzymatic function [20,23]. On the other hand, nBPs affect intracellular signaling, through farnesyl diphosphate synthase (FPPS) inhibition [20,23]. This interferes with the prenylation of GTPase proteins, vital for functions such as the formation of the ruffled membrane, vesicular transportation, or apoptosis [24]. FPPS impairment also produces an accumulation of isoprenyl pyrophosphate (IPP), producing another ATP analog (1-adenosine-5’-yl ester 3-[3-methylbut-3-enyl] ester), causing a similar effect to sBPs [20,22,23,25]. Bone resorption, in turn, affects the bone 5 formation phase due to the complex and coupled process of bone modeling and remodeling [26,27]. Figure 2. Chemical structure comparison between a third-generation nitrogen-containing BP (zoledronate) and a first-generation non-nitrogen-containing BP (clodronate). The pharmacokinetics (PK) of BPs depends on the route of administration. BPs are poorly absorbed orally, attributed likely to their low lipophilicity [28]. Oral bioavailability in humans has been estimated between 0.3% for pamidronate [29], 0.7% for alendronate [30], and 1-2% for clodronate [31]. Parenteral routes provide better and nearly complete absorption of BPs [28]. Initial tissue distribution depends on the protein-biding properties and will vary depending on blood pH, serum calcium, drug dosage, and species [28]. Though there is evidence of non- calcified tissue retention of BPs at high dosages [32,33], the probability of this occurring at therapeutic dosages is minimal [28]. BP distribution is not homogeneous in the skeleton and may be affected by sex and age. BPs tend to bind to the trabecular bone, because of larger amounts of bone turnover and blood supply in comparison to cortical bone [34–36]. Young animals may have an increased absorption rate compared to adults, and females may absorb less BP than males [33]. Once in the bone, BPs are absorbed by osteoclasts via endocytosis.[23,25] As bone resorption continues, BPs reactivate, resulting in prolonged osteoclastic inhibition over time.[20] For this reason, BPs’ half-life is difficult to define and may be up to 10 years depending on species, age, and the specific BP [4,28]. 6 Clodronate and etidronate (sBPs) can undergo intracellular metabolization [37]; on the other hand, there is no evidence of metabolization of nBPs [36]. BPs not taken up by the bone are largely excreted unaltered by the kidneys [28,36]. In human and rat studies, BPs half-life ranges between 1 to 2 hours, a quick bloodstream elimination that depends on kidney excretion (renal clearance) and bone uptake (nonrenal clearance); this ratio varies among BPs. In humans, the clodronate renal/nonrenal clearance ratio is between 1.8 to 3 in comparison to pamidronate which is 0.18 [28]. Additionally, the same class of BPs may differ in some PK parameters. For example, clodronate and tiludronate are sBPs with similar potency [22], yet the half-life of clodronate is reported to be between 2 and 3 hours [38,39], and tiludronate’s half-life is much longer (51 hours) [40]. Therefore, it is not accurate to extrapolate drug properties, even if BPs belong to the same class [41]. Because of this, studies must evaluate each BP and its individual effects on bone modeling and remodeling. THERAPEUTIC EFFECTS OF BISPHOSPHONATES Human conditions treated with BPs include postmenopausal osteoporosis, Paget's disease, osteogenesis imperfecta, and bone cancer/metastasis [42,43]. Bisphosphates are used to decrease bone resorption and increase bone mineral density (BMD) by decreasing bone catabolism. Besides their main antiresorptive properties, BPs are believed to have anti-inflammatory and analgesic effects [4,44–48], which make them an attractive potential treatment for multiple diseases including osteoarthritis (OA). Meta-analyses and systematic reviews have concluded that BP studies have controversial results and BP effects may be more related to pain relief than disease modification [48–50]. These pain-relieving effects could be beneficial for some individuals, but in human or animal athletes, masking pain could also be dangerous and lead to 7 further deterioration of joint conditions. BPs anti-inflammatory and pain-relieving effects may be due to a reduction in inflammatory mediators such as Prostaglandin E2 (PGE2). PGE2 is considered one of the key inflammatory mediators, generating pain in OA [51], and is a critical outcome measure in equine OA studies [52,53]. There is evidence that people treated with neridronate, a nBPs, for osteogenesis imperfecta have reduced serum concentrations of PGE2 and CTX-I/creatinine ratio [44]. Equine studies have demonstrated pain relief from BP administration in back OA [54], lower hock osteoarthritis [55], and navicular syndrome [56]. However, it is still unclear how BPs decrease pain in horses [53]. Additional pain-relieving mechanisms may be involved. Recently, clodronate was identified as a selective and potent inhibitor of vesicular nucleotide transport (VNUT) [57]. Typically, this transporter is responsible for storing ATP in neurons. Research suggests clodronate may inhibit the release of ATP acting as a presynaptic blocker attenuating chronic neuropathic pain [57]. This activity may explain why in an equine study, clodronate did not change serum bone resorption markers but did significantly improve lameness [56]. Further investigation is warranted as masking pain could lead to adverse outcomes especially in exercising animals. This is especially important as bisphosphonates may remain active in the bone for months to years following administration. BISPHOSPHONATES’ SIDE EFFECTS: HUMANS AND ANIMALS Bisphosphonates have short and long-term side effects. In humans, the kidneys excrete about 50 to 60% of BPs without major biotransformation [21,22,25], and a rapid intravenous infusion can produce focal glomerulosclerosis [58]. Humans treated with BP can experience short-term adverse effects including fever, muscle aches, vomiting, and transient hypocalcemia [59]. Long-term exposure in humans can result in serious side effects including osteonecrosis of 8 the jaw (ONJ) [60] and atypical femur fracture (AFF) [61]. In horses, short-term side effects may include renal toxicity, especially when the animal has a history of renal disease or has been treated with nonsteroidal anti-inflammatory drugs [47]. Further, transient colic-like symptoms have been documented following intravenous infusion [22,38,40]. Other adverse side effects have not been well documented in horses, but a lack of documentation may be related to the scarcity of long-term studies currently available. Between one and twelve percent of human oncologic patients experience ONJ after three years of IV treatment with BPs [62]. This condition has been also reported in several animal models, including sheep [63,64], mini-pigs [65], and dogs [66]. Recently, clinical reports showed symptoms compatible with ONJ in cats treated with long-term BPs for idiopathic hypercalcemia [67,68]. An additional serious adverse side effect, AFF, accounts for 1.1% of all femur fracture cases in humans [69]. Likewise, a bilateral patellar fracture has been reported in a cat after 8 years of alendronate treatment [70]. Changes in the femoral neck with increased bone brittleness have been found in mice using ibandronate [71]. Even though there is not currently a clear connection between equine stress fractures or catastrophic injuries and BPs, these drugs have shown the potential to produce severe adverse effects in multiple animal models and humans. In horses, clodronate has been isolated from bone samples 18 months following a single administration [72], and tiludronate has been found in low concentrations in plasma (0.05 – 1.0 ng/ml) and urine samples (0.03 – 1.5 ng/ml) after three years following administration [73]. Fluctuations in plasma and urine concentrations over time may have been influenced by activity level, health status, growth, and animal to animal variation [73]. BPs can be present in the skeleton of horses for long periods of time, potentially masking pain, and are documented to cause adverse bone effects in multiple species. Consequently, further investigation into the 9 relationship between BPs and bone injuries in horses is crucial for equine health. BISPHOSPHONATE IN ADULT HORSES Two BPs were approved for use in the horse by the FDA in 2014 for the treatment of navicular syndrome in horses over 4 years old [4]. Clodronate and tiludronate dosage characterizations are detailed in their respective FDA Freedom of Information Summary [38,74]. The lowest effective clodronate dosage found to decrease one grade in navicular syndrome- associated lameness was 1.8 mg/kg or 900 mg per horse [73]. For tiludronate, 1 mg/kg was found to alleviate symptoms associated with navicular syndrome [38,75]. BPs have resulted in reduced pain and lameness in other skeletal conditions, such as back pain, lower hock OA, and fetlock OA [54,55,76]. Additionally, BPs have been used to treat bone fragility disorder, an osteoclast-mediate osteoporosis [77,78]. Multiple publications have focused on short-term benefits of bisphosphonate use in horses. However, long term studies investigating potential long-term adverse effects are lacking. During the 2019 American Association of Equine Practitioners convention, a roundtable discussion covered the extra-label use of BPs by equine practitioners. Participants indicated BPs were being used for various conditions with radiographic or nuclear scintigraphic abnormalities of the sacroiliac area, pelvis, or limb [79]. Participants described frequent BP administration (ex., three full doses in a month); despite the manufacturer’s recommendation of a six-month separation between doses [38]. Researchers have raised concern about the extra-label use of BPs [5], especially in younger horses where bone turnover is significantly higher in individuals under 24 months of age [80]. 10 THE USE OF BISPHOSPHONATES IN YOUNG/EXERCISING ANIMALS Racehorses often start training and racing at 2 years of age. There is evidence that improper training and management is more of a factor in skeletal injury than age [81], as high- performance exercise may result in progressive microdamage accumulation, potentially leading to stress fractures (SF) [82]. Stress fractures have been associated with a high remodeling rate, leading to bone weakness and accumulation of microdamage over time [83]. It is believed BPs may be useful in preventing athletic SF due to their antiresorptive properties [84]. However, there is no conclusive evidence indicating SF healing by BPs [85], and their use in this condition is not recommended [86]. In truth, bone modeling and remodeling are complex processes especially when growth and exercise intersect. SF have been associated with normal remodeling and high strains, or normal strains with decreased remodeling [87]. Even though it is not clear what pathophysiological mechanism prevails in racehorses, any interruption in normal osteoclast resorption could be harmful and lead to damage accumulation over time. In horses, common skeletal conditions, such as dorsal metacarpal disease, commonly known as bucked shins, and sesamoiditis, have been treated with BPs for their perceived skeletal and analgesic effects [4]. Dorsal metacarpal disease results from microfractures on the metacarpal cortical area, and sesamoiditis is the result of disease or osteolysis of the vascular channels of the proximal sesamoid bones [5]. It is believed that BPs may prevent pain and radiographic evidence of these pathologies [4,5]. However, the resolution of radiographic evidence of disease may be accompanied by detrimental effects in juvenile horses, where increasing bone density may not equate to increased bone strength. Further, impairing osteoclast function may harm normal bone remodeling and healing necessary for juvenile horses under high-performance exercise [5]. 11 Bone turnover can be affected by exercise, and BPs can influence physiological adaptation to exercise. Bone resorption increases in response to acute exercise [88]. In long-term exercise, there is a BMD increase, indicating that prolonged exercise can be an osteogenic stimulus [88]. The increased mechanical load under exercising conditions induces osteoclast activation that can result in increasing serum markers of bone remodeling, such as CTX-I [88]. However, serum bone remodeling markers are not strong predictors of bone formation and/or resorption in human subjects [88]. For example, calves subjected to sprints 1 to 5 times per week have increased fracture force and dorsal width of their fused metacarpus compared to a non- exercise group, but no differences in CTX-I were detected between groups [89]. On the other hand, procollagen type II C-propeptide (CPII) and CTX-I increase as a response to exercise and bone turnover in foals [90]. In humans, BPs may reduce serum bone markers over time [91–93]. In horses, conflicting reports exist regarding the effect of BPs on CTX-I [40,56,76,94]. In conclusion, BPs may alter the normal skeletal adaptation to exercise, and assessment of the antiresorptive effects of BPs through serum bone markers is likely insufficient if performed alone. Future studies should consider new, comprehensive approaches to evaluate BP effects including measuring bone mineral density, fracture healing, and biomechanical testing whilst simultaneously determining BP concentration within the bone. In addition, advanced imaging such as micro-computed tomography (µCT) and Positron Emission Tomography (PET) CT may be warranted. The use of BPs may have a greater impact on young horses due to their active growth, where osteoclasts play a significant role in the endochondral ossification process [13,19]. Osteoclasts are abundantly present in growth epiphyseal plates up to 2 years old [80]. The extra- label use of BPs in young animals could impair physiological bone development in this 12 population [80,95]. This has been demonstrated in a rabbit model where BP administration caused a 3% decrease in the length of the tibia [96]. Hence, BPs use in young animals could pose a significant risk to skeletal growth and/or adaptation to exercise, resulting in microdamage accumulation in juvenile horses without degenerative bone disorders. BISPHOSPHONATES AND FUTURE STUDIES Multiple animal models have already been used to investigate BP including mice, rabbits, mini-pigs, dogs, and sheep [63–66,71,96]. The authors recognize the ethical concerns around using animals for research purposes. However, some animal models may be particularly useful depending on the research goals and prior studies available. In particular, the sheep model has proven to be a reliable orthopedic model for human BP use. Sheep have a similar body weight and skeletal size to humans, procedures such as bone biopsies and blood sampling are simple, they are easy to handle, and large numbers of animals are usually available [97–100]. Furthermore, sheep can be trained to undergo forced exercise [101], making sheep a suitable animal model for investigating potential BP-associated bone changes under different exercise regimens. Although animal models have been used to investigate long and short-term BP effects with a focus on human health, few studies are available to guide equine use especially in juvenile and exercising populations. Future studies may include experimental large animal models of BP use which incorporate exercise to mimic athletic training. Specifically, terminal ovine models may allow for mechanical testing, advanced imaging, and analysis of long-term BP retention in bone and other organs. These studies, coupled with focused equine experimental trials, prospective and retrospective studies would provide a more comprehensive explanation of the benefits and risks of BP use in the horse. To date, a single, large, retrospective study has evaluated the efficacy and safety of 13 tiludronate in 1,804 horses; 343 horses were followed for greater than 1 year [102]. The study revealed a low incidence of short-term adverse effects (1.3%), with colic-like symptoms being the most frequent. Less than 20% of horses were treated for navicular syndrome, confirming the extra-label use of BPs. Between one and nine doses of tiludronate were administered to horses included in the study [102]. Treated horses ranged in age from 2 years old to 26 years old. Future retrospective studies would ideally report diagnosis, age at administration, number and frequency of doses, long-term follow-up, concurrent treatments and evidence of disease progression. Future prospective studies will ideally look beyond serum biomarkers and report multiple clinical and experimental parameters. These could include physical and lameness examinations coupled with bone biopsies, synovial fluid analysis, advanced imaging and biomechanical testing. Veterinarians, owners and researchers alike would benefit from a better understanding of the half-life of BPs within the skeleton and the physiologic factors such as age and exercise which may change the half-life of BPs. Tiludronate has been previously measured in tuber coxae biopsies, a relatively non-invasive location for bone biopsy [103]. Tiludronate can be detected with ultra-high-performance liquid chromatography-high-resolution mass spectrometry for up to three years in plasma and urine samples and clodronate was detected in bone 18 months following administration in a single horse in a single study [72,73]. However, long-term presence of BPs in bone in a large clinical population is currently unreported [72]. Little information is currently available to guide frequency of dosing to ensure clinical efficacy and safety. The pain-relieving effects of BPs are still being investigated. Although pain relief may be a clinical benefit, it could also result in further injury especially in high performance athletes. BPs have been detected in the synovial fluid after systemic administration [39]. Further investigation is necessary to understand the potential anti-inflammatory effects of BPs 14 systemically and within in the joint environment. This can be accomplished through in vitro studies, in vivo animal models of pain and inflammation, and clinical studies [104]. CONCLUSION Bisphosphonates are well-known for their antiresorptive properties, impairing osteoclast functionality. In 2014, two bisphosphonates (clodronate and tiludronate) were approved by the FDA to treat navicular syndrome in horses over four years of age. Several in vitro, animal models, and human studies indicate that bisphosphonates may have anti-inflammatory and pain- relieving effects, which has led to extra-label use of these drugs for other conditions and in juvenile horses. Although there may be therapeutic effects, there are concerns regarding impairment of normal physiological functions (growth, bone repair, and bone remodeling) especially in juvenile and exercising animals. Additional research must focus on the identifying the short-term and long-term effects of bisphosphonates in young and exercising animals to ensure the efficacious and judicious use of this powerful, long-lasting group of drugs. 15 CHAPTER 2: Clodronate disodium is neither cytotoxic nor cytoprotective to normal and recombinant equine interleukin-1β-treated joint tissues in vitro This chapter has been published in Veterinary Surgery and is available at the following citation: F.B. Vergara-Hernandez, C.L. Panek, B.D. Nielsen, C.I. Robison, A.C. Colbath, Clodronate disodium is neither cytotoxic nor cytoprotective to normal and recombinant equine interleukin- 1β-treated joint tissues in vitro, Vet Surg. 52 (2022) 146–156. https://doi.org/10.1111/vsu.13898. ABSTRACT Objective: To determine the effects of clodronate disodium (CLO) on control and recombinant equine interleukin-1β (IL-1β)-treated equine joint tissues. Study design: In vitro experimental study. Sample population: Cartilage explants, chondrocytes, and synoviocytes (n = 3 horses). Methods: Monolayer cultures of chondrocytes and synoviocytes from three horses were subjected to: control media (CON), 5 ng/ml CLO (C/low), 50 ng/ml CLO (C/med), 100 ng/ml CLO (C/high), with and without IL-1β, and 10 ng/ml IL-1β (IL) alone for 72 hours. Cartilage explants from three horses were subjected to CON, IL, C/low, and C/med with and without IL- 1β for 72 hours. Culture media was analyzed for prostaglandin-E2 (PGE2), interleukin-6 (IL-6), and nitric oxide (NO). Explant media was analyzed for glycosaminoglycan (GAG) content and NO. At 72 hours, explant and monolayer culture viability were assessed, and explant GAG content was measured. Results: IL-1β treatment resulted in higher media concentrations of GAG, NO, PGE2, and IL-6 compared to the CON treatment (P < 0.05), demonstrating a catabolic effect of IL-1β on explants and monolayer cultures. CLO treatments did not increase media concentrations of GAG, NO, PGE2, or IL-6 compared to CON, indicating no cytotoxic effect. Nevertheless, CLO 16 treatments administered to IL-1β-treated monolayer cultures and explants did not significantly reduce the inflammatory response regardless of concentration. Conclusion: CLO did not demonstrate cytotoxic nor cytoprotective effects in normal and IL-1β-stimulated chondrocytes, synoviocytes or explants in culture. Clinical significance: This study does not support the use of CLO as an anti- inflammatory treatment. Further research is necessary to confirm any anti-inflammatory effects of CLO on joint tissues. 17 CHAPTER 3: Pharmacokinetics and plasma protein binding of a single dose of clodronate disodium are similar for juvenile sheep and horses F.B. Vergara-Hernandez, B.D. Nielsen, J.J. Kottwitz, C.L. Panek, C.I. Robison, B.L. Paris, T.H. Welsh Jr., A.N. Bradbery, J.L. Leatherwood, A.C. Colbath, Pharmacokinetics and plasma protein binding of a single dose of clodronate disodium are similar for juvenile sheep and horses, Amer J Vet Res. 84 (2023) 1–7. https://doi.org/10.2460/ajvr.23.03.0051. ABSTRACT Objective: To determine the single-dose pharmacokinetics of clodronate disodium (CLO) in juvenile sheep and the plasma protein binding (PPB) of CLO in juvenile sheep and horses. Animals: Eleven juvenile crossbred sheep (252 ± 6 days) for the pharmacokinetic study. Three juvenile crossbred sheep (281 ± 4 days) and three juvenile Quarter Horses (599 ± 25 days) for PPB analysis. Methods: CLO concentrations were determined using liquid chromatography-mass spectrometry. Pharmacokinetic parameters were calculated by noncompartmental analysis from plasma samples obtained at 0, 0.5, 1, 3, 6, 12, 24, 48, and 72 hours after CLO administered IM at 0.6 mg/kg. PPB was determined using equine and ovine plasma in a single-use rapid equilibrium dialysis system. Results: The mean and range for maximum plasma concentration (Cmax: 5,596; 2,396– 8,613 ng/mL), time of maximal concentration (Tmax: 0.5; 0.5–1.0 h), and area under the curve (AUCall: 12,831; 7,590–17,593 h × ng/mL) were similar to those previously reported in horses. PPB in sheep and horses was moderate to high, with unbound fractions of 26.1 ± 5.1% in sheep and 18.7 ± 7.5% in horses, showing less than a 1.4-fold difference. Clinical relevance: The pharmacokinetic parameters and PPB of CLO in juvenile sheep 18 were similar to those previously reported in horses. The results suggest sheep may be an appropriate animal model for studying the potential risks and/or benefits of bisphosphonate use in juvenile horses. INTRODUCTION Bisphosphonates (BPs) have been used for over 40 years in human medicine and their use in veterinary medicine has gained interest in the last 10–20 years due to their effects on osteoclast inhibition [5]. In 2014, two BPs, tiludronate disodium (C7H9ClO6P2S) and clodronate disodium (CH2Cl2Na2O6P2), were approved by the Food and Drug Administration under the New Animal Drug Application (FDA NADA) 141-420 (tiludronate disodium) and 141-427 (clodronate disodium [CLO]) for the treatment of navicular syndrome in horses four years of age and older [38,74]. BPs’ ability to impair osteoclast function [24] with possible concomitant pain- relieving effects [57] make them potentially useful for a myriad of musculoskeletal conditions, including fetlock and distal tarsal osteoarthritis [55,76] and thoracolumbar vertebral disease [54]. After administering BPs, the drug that is not absorbed by the skeleton is excreted unchanged by the kidneys [105]. In veterinary medicine, clinicians have raised concern over adverse effects from extra-label use in young and exercising horses including impediment of growth, masking of pain, and fracture predisposition, in addition to known potentially adverse effects such as renal toxicity and gastrointestinal discomfort [4,5,106]. Sheep have been used to assess the efficacy of BPs for osteoporosis and to investigate adverse effects experienced by humans such as necrosis of the jaw and atypical femoral fractures [64,107–110]. In addition, sheep have been used in exercise studies as a model for horses [101]. Therefore, juvenile sheep may be suitable for investigating the effects of BPs under exercise, providing a model to assess the effects of BPs in juvenile, exercising horses. 19 To use sheep as a model for juvenile, exercising horses, an appropriate dose of CLO must be determined. The ideal dose would result in pharmacokinetic and plasma protein binding profiles similar to those observed in horses. Pharmacokinetics (PK) focus on how drugs enter the body, travel to their site of action, and are removed from the organism [111]. Drug accumulation and elimination rates from an organism are determined by the absorption, distribution, metabolism, and excretion of a specific drug over time [112]. The quantification of these parameters help establish a safe and effective dosage of a drug [112]. Plasma protein binding (PPB) determines the fractions of drug bound (fb) and unbound (fu) to plasma proteins in the blood, which influence the volumes of distribution (Vd) [113] and clearance (CL) of drugs [114– 116]. Not evaluating PPB leads to misinterpretations of the total drug plasma concentration, as fu is directly associated with the CL of the total drug plasma concentration [117]. Hence, differences in PPB values between species may affect the safety margin of drugs, leading to potential incompatibility between species [118]. The PK parameters of a commercially-available, intramuscular (IM) formulation of CLO have been published for horses [38,39,119] but not sheep, and the authors are unaware of studies that describe the PPB of CLO in sheep or horses. The current study compared the PK of a single dose of CLO (0.6 mg/kg IM) in juvenile sheep and determined the PPB of CLO in both sheep and horses. The CLO dose of 0.6 mg/kg IM was based on a preliminary study [120]. We hypothesized that a single 0.6 mg/kg dose of CLO administered IM in sheep would lead to similar PK parameters as a single dose of CLO administered at 1.8 mg/kg IM in adult horses. Furthermore, we hypothesized that there would be less than a 5-fold difference between the fu of CLO in sheep and horses. 20 METHODS Sheep The animal use experimental protocols were approved by the Michigan State University (MSU) Institutional Animal Care and Use Committee (202000264). All sheep were obtained from an established flock at the MSU Sheep Teaching and Research Center (Lansing, MI). Eleven juvenile crossbred (Dorset × Polypay, 76 ± 8 kg, 252 ± 6 days of age, five castrated male and six female) sheep were used in the PK study. A juvenile sheep was defined as an animal that had reached approximately 80% of its adult weight.29,30 All sheep had a physical examination, including heart rate (HR), respiratory rate (RR), and rectal temperature (RT), prior to sample collection. Three additional sheep (Dorset × Polypay, 68 ± 11 kg, 281 ± 4 days of age, one castrated male and two females) were used as PPB study subjects; these sheep had never received CLO and had a physical examination prior to the sample collection. Three additional sheep (Dorset × Polypay, 68 ± 11 kg, 281 ± 4 days of age, one castrated male and two females) were used as PPB study subjects; these sheep had never received CLO and had a physical examination prior to the sample collection. All sheep were housed in a 21.6 m2 pen in an indoor facility at the MSU Bennett Road Farm. Sheep underwent a 2-week acclimation period prior to sampling. All sheep had free access to a 90% dry matter (DM) total mixed ration that contained chopped hay (82%), a corn/soybean blend (16.5%), and 1.5% of a mineral blend (Caledonia Farmers Elevator). The sheep ingested approximately 2.0-2.3% of their body weight (BW) in DM per day and had ad libitum access to water. 21 Horses The animal use experimental protocols were approved by the Texas A&M Institutional Animal Care and Use Committee (2019-0325). All horses were obtained from Texas A&M Dick Freeman Arena (College Station, TX). Horses were group housed in dry lots at the time of sample collection for PPB. Three juvenile horses (Quarter Horses, 411 ± 18 kg, 599 ± 25 days of age, one castrated male and two females) were used for the PPB. A juvenile horse was defined as an animal that had reached approximately 80% of its adult weight [121]. Horses were fed coastal Bermuda grass hay ad libitum and supplemented twice daily with 1.4 kg of a 12% protein, 8% fat commercially available concentrate. All animals were used as part of an equine behavior and training course and health status was monitored daily. Pharmacokinetic study dose selection The CLO dose was determined by a preliminary study in which twelve adult sheep were administered three different single doses of CLO (n = 4/treatment group: 0.6, 1.8, 3.0 mg/kg IM respectively) [120]. Preliminary study findings suggested a single dose of 0.6 mg/kg IM resulted in similar plasma concentrations of CLO as reported in mature horses for 48 hours following administration [38,120]. Pharmacokinetic study design Sheep were moved through a chute system in a randomized order and restrained in a crate for sampling. Ten milliliters of blood were collected from the right jugular vein using an 18- gauge needle and vacutainer prior to single administration (hour 0) of CLO (OSPHOS®, Dechra Veterinary Products) at 0.6 mg/kg IM in the right side of the neck. An additional 10 mL of blood was collected from the right jugular vein at 0.5, 1, 3, 6, 12, 24, 48, and 72 h in the same 22 collection order established at hour 0. Blood was collected in a K2 EDTA blood collection plastic tube and immediately placed on ice for transport followed by sample centrifugation at 2,000 × g for 10 min. Plasma was aliquoted into 2.0 mL microcentrifuge tubes and frozen at -80 ºC until analysis. All samples were analyzed within four months of collection. Animals were monitored hourly during the first six hours and then daily following drug administration for acute adverse effects (e.g., syncope and sudden death), side effects described in other species (e.g., gastrointestinal discomfort and/or swelling at the injection site), and/or changes in normal behavior (e.g., agitation and depression) [22,38,59]. Bioanalytical methods The liquid chromatography-mass spectrometry (LC-MS/MS) analysis was based on the methods of Hasan and colleagues.[122] Samples were removed from -80 ºC and thawed to room temperature. An internal standard of 20 µL (10 ng/µL of etidronate solution, Sigma-Aldrich) was added to 1.0 mL of plasma. Then, 200 µL of perchloric acid (10%) was added to precipitate proteins. The solution was thoroughly mixed using a vortex mixer and centrifuged at 17,000 × g for 10 min. The supernatant was evaporated until visibly dry under a stream of nitrogen at ~65 ºC, then dissolved in 150 µL glacial acetic acid and mixed with 500 µL trimethyl orthoacetate. The solution was incubated at 100 ºC for 30 min for derivatization. Samples were allowed to cool to room temperature and then 300 µL formic acid and 500 µL DI water were added. The solution was then transferred to a screw-top tube, followed by liquid-liquid extraction with methyl tert-butyl ether. The tube was capped, placed in a rotorack for 10 min, and centrifuged at 12,000 × g for 5 min. The bottom layer was removed by aspiration. The supernatant was transferred to a new glass tube and evaporated until visibly dry under a stream of nitrogen at ~65 23 ºC. The residue was reconstituted in 80 µL of a 1:1 methanol-water mixture and transferred to an autosampler vial (Zorian, Inc.) for the LC-MS/MS analysis. Plasma clodronate concentrations were quantified by LC-MS/MS analysis conducted in a Thermo TSQ Altis™ (ThermoFisher Scientific) triple quadrupole mass spectrometer with an electrospray ionization source, with a lower limit of detection (LOD) of 10 ng/mL for CLO. Pharmacokinetics data analysis Non-compartmental PK parameters were determined using commercially available software (Phoenix WinNonLin 8.3). The PK parameters included the time of maximum concentration (Tmax), maximum concentration (Cmax), the slope of the second phase (slow distribution phase) of the drug’s concentration curve (λz), terminal half-life (t1/2λ), area under the curve estimated to the last observation (AUCall), area under the curve extrapolated to infinity (AUC0-inf), and mean residence time (MRT). Plasma protein binding assay Twenty milliliters of blood were collected from the jugular vein of sheep and horses using an 18-gauge with a hub. Blood was immediately placed in a K2 EDTA blood collection plastic tube and placed on ice for transport. Plasma was harvested and stored as previously described for no longer than three months. Plasma samples were thawed at room temperature and CLO (pharmaceutical grade, Millipore Sigma) was added to 2 mL of plasma to obtain the desired concentrations for sheep (0, 100, 1,000, 10,000, 20,000, 40,000 ng/mL) and horses (0, 100, 1,000, 7,500, 10,000, 15,000 ng/mL). These concentrations were selected based on our preliminary sheep study (Cmax 11,605 ng/mL) [120] and a previously disclosed equine study (Cmax 7,460 ng/mL) [38]. A single-use rapid equilibrium dialysis (RED®, ThermoFisher Scientific) system was used according to the manufacturer’s recommendations. 24 In brief, 500 µL of the spiked samples were added to the plasma chamber and 750 µL of 1X PBS were added to the buffer chamber. The kit was covered with a sealing plate and incubated at 37 ºC for 6 hours on an orbital shaker at 250 rpm. The content of the plasma chamber and buffer chamber were then pipetted into separate microcentrifuge tubes and stored at -80 ºC. Samples were diluted with distilled water to increase the volume for laboratory analysis. Each biological replicate was measured in duplicate as a technical replicate. The CLO concentrations within both chambers were calculated using LC-MS/MS as described. The technical replicates for each biological replicate were averaged to create an individual data point for the analysis. The percentage of the unbound drug was calculated as follows: %free drug = (drug concentration buffer chamber) / (drug concentration plasma chamber) × 100. Statistical analysis Normality of the PK data was assessed by a Shapiro-Wilk test using Phoenix WinNonLin 8.3. Normally distributed data including physical examination, PPB, and PK parameters, except for Tmax, were reported as means and ranges. Non-normally distributed data (Tmax) were reported as median and range. RESULTS Pharmacokinetics of clodronate in sheep Sheep were determined to be bright and alert prior to PK sampling. Physical examination parameters reflected mild stress associated with temporary restraint: HR (138 [102-160] beats per min), RR (97 [60-132] breaths per min), and RT (39.5 [39.1-39.8] ºC]). No animals or data were excluded from the analyses. No adverse effects were detected following the administration of CLO to juvenile sheep (n = 11). Mean values of CLO plasma concentration in sheep over time are presented (Figure 3). 25 Following IM administration of CLO, the Tmax was reached at 0.5 (0.5-1.0) h post- administration, with a Cmax of 5,596 (2,396-8,613) ng/mL. The λz was 0.034 (0.023-0.042) 1/h, t1/2λ reached 21.2 (16.4-30.3) h with an AUCall of 12,831 (7,590-17,593) × ng/mL, and AUC0-inf of 13,334 (7,947-17,973) × ng/mL (Table 2). Figure 3. Plasma clodronate disodium concentration (ng/mL) time curves over time (72 h) after a single intramuscular administration of 0.6 mg/kg (OSPHOS®) in 11 juvenile sheep. Table 2. Plasmatic pharmacokinetics (PK) parameters (mean or median [Tmax] and range) for clodronate disodium (OSPHOS®) following single-intramuscular administration in juvenile sheep (0.6 mg/kg) determined through liquid chromatography-mass spectrometry and non- compartmental analysis compared to a previously published single dose (1.8 mg/kg) PK study in mature horses [119]. Mean Range Parameter Unit Sheep (n = 11) 0.5 5,596 0.034 21.2 12,831 13,334 12.1 h ng/mL 1/h h h × ng/mL h × ng/mL H Tmax Cmax λz † t1/2λ AUCall AUC0-inf MRT0-inf Abbreviations: AUC0-inf = area under the curve extrapolated to infinity. AUCall = area under the curve to the last observation. Cmax = maximal concentration. λz = slope of the distribution phase. MRT0-inf = mean resident time from 0 to infinity. t1/2λ = terminal half-life. Tmax = Time of maximal concentration. †Harmonic mean Sheep (n = 11) 0.5-1.0 2,396-8,613 0.023-0.042 16.4-30.3 7,590-17,593 7,947-17,973 8.1-22.3 Horse (n = 7) [119] 0.25-0.75 1,969-5,541 0.0001-0.016 43.1-6,814.8 6,393-17,026 6,507–17,606 - Horse (n = 7) [119] 0.5 4,155 0.002 141.7 11,564 11,912 - 26 Plasma protein binding The mean percentage of unbound CLO in sheep plasma was 26.1 ± 5.1% (n = 3 for each of 5 concentrations) from 100 to 40,000 ng/mL of CLO (Table 3). The mean percentage of unbound CLO in horse plasma was 18.7 ± 7.5% (n = 3 for each of 5 concentrations) from 100 to 15,000 ng/mL of CLO (Table 3). Table 3. In vitro unbound fraction (fu, mean ± SD) of clodronate disodium (ng/mL) in juvenile sheep and horses. The table indicates the unbound clodronate disodium fractions in sheep plasma (0, 100, 1,000, 10,000, 20,000, 40,000 ng/mL, n = 3) and horse plasma (0, 100, 1,000, 7,500, 15,000 ng/mL, n = 3), determined using single-use rapid equilibrium dialysis and liquid chromatography-mass spectrometry analysis. (ND, not determined,