Population pharmacokinetic analysis of blood and joint synovial fluid concentrations of robenacoxib from healthy dogs and dogs with osteoarthritis

The history and future of population pharmacokinetic analysis in drug development

1. The analysis of pharmacokinetic data has been in a constant state of evolution since the introduction of the term pharmacokinetics. Early work focused on mechanistic understanding of the absorption, distribution, metabolism and excretion of drug products.2. The introduction of non-linear mixed effects models to perform population pharmacokinetic analysis initiated a paradigm shift. The application of these models represented a major shift in evaluating variability in pharmacokinetic parameters across a population of subjects.3. While technological advancements in computing power have feuled the growth of population pharmacokinetics in drug development efforts, there remain many challenges in reducing the time required to incorporate these learnings into a model-informed development process. These challenges exist because of expanding datasets, increased number of diagnostics, and more complex mathematical models.4. New machine learning tools may be potential solutions for these challenges. These new methodologies include genetic algorithms for model selection, machine learning algorithms for covariate selection, and deep learning models for pharmacokinetic and pharmacodynamic data. These new methods promise the potential for less bias, faster analysis times, and the ability to integrate more data.5. While questions remain regarding the ability of these models to extrapolate accurately, continued research in this area is expected to address these questions.

Phenylbutazone concentrations in synovial fluid following administration via intravenous regional limb perfusion in the forelimbs of six adult horses

Background

Pain management is critical to equine welfare with non-steroidal anti-inflammatory drugs (NSAID) commonly used in horses. However, systemic NSAID use is limited by harmful gastrointestinal and renal side effects. Intravenous regional limb perfusion (IVRLP) is a technique used in horses that produces high, local antibiotic concentrations while limiting systemic circulation. NSAID-IVRLP would be a novel method of local pain management while limiting systemic NSAID side effects. To date, NSAID-IVRLP administration has not been reported in horses. This study aimed to identify the pharmacokinetics and local complications associated with using the NSAID phenylbutazone (PBZ) for IVRLP in six standing adult horses.

Methods

PBZ-IVRLP, at a dose of 2.2 mg/kg PBZ, was performed in a randomly assigned forelimb cephalic vein in 6 standing, healthy adult horses. A placebo-IVRLP was performed in the contralateral forelimb for comparison. Systemic serum and radiocarpal joint synovial fluid PBZ concentrations were identified at various timepoints (before IVRLP T-0 h, just after tourniquet removal T-0.5, 1.5, 3, 5, 12, 16, and 24 h post IVRLP) for non-compartmental pharmacokinetic analysis and concentration over time curves. Local complications associated with PBZ-IVRP were evaluated for up to 7 days following PBZ-IVRLP using physical and ultrasonographic assessment. On day 7 horses were humanely euthanized with histology performed on both forelimbs at PBZ-IVRLP and placebo-IVRLP administration sites.

Results

Non-compartmental pharmacokinetics for PBZ, and its major metabolite oxyphenbutazone (OBZ), were determined for serum and synovial fluid. Synovial PBZ concentrations (mean ± SD; 1.9 ± 2.1 μg/mL) were significantly lower (p = 0.03; CI,0.46-7.36) than serum PBZ concentrations (5.8 ± 5.1 μg/mL) at any time point. Physical and ultrasonographic measurements were not significantly different between PBZ- and placebo-IVRLP forelimbs. The most common histologic findings included focal deep dermal/subcutaneous hemorrhage and edema. Two horses showed perivasculitis and one horse showed a resolving thrombus in the cephalic vein of the PBZ-IVRLP limb. This horse also had severe perivasculitis and fibrinosuppurative dermatitis/panniculitis in the placebo-IVRLP limb.

Conclusion

PBZ-IVRLP pharmacokinetics at a 2.2 mg/kg dose showed no benefit compared to systemic PBZ administration in standing adult horses. Local complications associated with PBZ-IVRLP were similar to placebo-IVRLP on physical and ultrasonographic evaluation.

Pharmacology, safety, efficacy and clinical uses of the COX-2 inhibitor robenacoxib

Robenacoxib is a veterinary‐approved non‐steroidal anti‐inflammatory drug (NSAID) of the coxib group. It possesses anti‐hyperalgesic, anti‐inflammatory and anti‐pyretic properties. Robenacoxib inhibits the cyclooxygenase (COX)‐2 isoform of COX selectively (in vitro IC₅₀ ratios COX‐1:COX‐2, 129:1 in dogs, 32:1 in cats). At registered dosages (2 mg/kg subcutaneously in dogs and cats, 1–4 mg/kg orally in dogs and 1–2.4 mg/kg orally in cats), robenacoxib produces significant inhibition of COX‐2 whilst sparing COX‐1. The pharmacokinetic (PK) profile of robenacoxib is characterized by a high degree of binding to plasma proteins (>98%) and moderate volume of distribution (at steady state, 240 ml/kg in dogs and 190 ml/kg in cats). In consequence, the terminal half‐life in blood (<2 h) is short, despite moderate body clearance (0.81 L/kg/h) in dogs and low clearance (0.44 L/kg/h) in cats. Excretion is principally in the bile (65% in dogs and 72% in cats). Robenacoxib concentrates in inflamed tissues, and clinical efficacy is achieved with once‐daily dosing, despite the short blood terminal half‐life. In dogs, no relevant breed differences in robenacoxib PK have been detected. Robenacoxib has a wide safety margin; in healthy laboratory animals daily oral doses 20‐fold (dog, 1 month), eight‐fold (cat, 6 weeks) and five‐fold (dog, 6 months) higher than recommended clinical doses were well tolerated. Clinical efficacy and safety have been demonstrated in orthopaedic and soft tissue surgery, and in musculoskeletal disorders in dogs and cats.

Observations on the use of a pain numbing device for repetitive percutaneous sampling in sheep

AIMS

To evaluate the success of a commercially available analgesic device (CoolSense; Coolsense Ltd, Tel Aviv, Israel) in ameliorating pain while sampling from subcutaneous tissue cages in sheep.

METHODS

The CoolSense device was used as part of a major parent study involving repetitive percutaneous sampling of subcutaneous tissue cages in seven sheep. Sampling was performed by passing a hypodermic needle through the skin and withdrawing fluid from the tissue cage. Each sheep had 10 tissue cages that were individually sampled 14 times over 74 h. The device was placed on the skin of the sampling site immediately before sampling cooling and numbing the skin. The reaction of the sheep was observed by the operators, flinching or jumping as the needle was passed through the skin was deemed to be a failure. We recorded the success or failure of the device for each needle stick. This was opportunistic data collection as part of a pharmacokinetic trial, therefore no controls were included.

RESULTS

A total of 1655 observations were recorded and then analysed using a generalised linear mixed model. Overall, 1380 of 1655 (83.4%) observations were recorded as successfully providing analgesia. Marked inter-occasion variability was noted with success ranging from 61.42% to 92.86% across sheep:period (approximately 140 observations each). As no controls were available, the effect of treatment could not be evaluated.

CONCLUSIONS AND CLINICAL RELEVANCE

The CoolSense device is a viable option for veterinary research and clinical applications.

Evaluation of the Effects of Undenatured Type II Collagen (UC-II) as Compared to Robenacoxib on the Mobility Impairment Induced by Osteoarthritis in Dogs

Osteoarthritis (OA) is a chronic disease that requires a multimodal therapeutic approach. The aim of this study was to evaluate the effects of undenatured type II collagen (UC-II) as compared to robenacoxib in dogs affected by OA. Our hypothesis was that the two compounds would be similar (non-inferiority) in improving mobility. To test this hypothesis, a complete orthopedic examination, x-ray and the Liverpool Osteoarthritis in Dogs (LOAD) survey were performed in dogs affected by OA before and after the treatments. The study was designed as a clinical, randomized, controlled and prospective study. Sixty client-owned dogs were randomized in the R group (n = 30, robenacoxib 1 mg/kg/day for 30 days) and in the UC-II group (n = 30, UC-II 1 tablet/day for 30 days). Thirty days after the beginning of the treatment (T30), the dogs were reassessed for the LOAD, MOBILITY and CLINICAL scores. Based on the data obtained from the study, a significant reduction in LOAD and MOBILITY scores was recorded between T0 and T30 with a similar magnitude among the two groups (R = 31.5%, p < 0.001; UC-II = 32.7%, p = 0.013). The results of this study showed that UC-II and robenacoxib were able to similarly improve mobility of dogs affected by OA.

Meloxicam vs robenacoxib for postoperative pain management in dogs undergoing combined laparoscopic ovariectomy and laparoscopic-assisted gastropexy

OBJECTIVE

To compare meloxicam and robenacoxib for short-term postoperative pain management after combined laparoscopic ovariectomy and laparoscopic-assisted gastropexy.

STUDY DESIGN

Double-blind, prospective, randomised clinical trial.

ANIMALS

Twenty-six client-owned female dogs.

METHODS

Dogs undergoing combined laparoscopic ovariectomy and laparoscopic-assisted gastropexy were randomly divided into 2 groups. Before induction of anesthesia, 13 dogs received meloxicam (0.2 mg/kg subcutaneously), and 13 dogs received robenacoxib (2 mg/kg subcutaneously). Pain was scored with the Glasgow Composite Pain Scale (short form) before surgery and at 1, 6, 12, 18, and 24 hours after extubation. Rescue analgesia (tramadol, 3 mg/kg) was provided to dogs with a Glasgow pain score (GPS) ≥5. Glasgow pain scores were analyzed by ANOVA with treatment, age, and surgical time as fixed factors.

RESULTS

Glasgow pain scores were higher at 24 hours postsurgery in dogs treated with robenacoxib (2.18 ± 0.29) compared with those treated with meloxicam (0.68 ± 0.41, P = .04). Two dogs treated with meloxicam and 7 dogs treated with robenacoxib required rescue analgesia. Regardless of the treatment, the overall GPS was lower at 18 and 24 hours postsurgery when the surgical time was >40 minutes compared with surgical times ≤40 minutes, but surgical site inflammation was likely a confounding factor in this finding. Glasgow pain score was not affected by patient age.

CONCLUSION

Meloxicam was more effective than robenacoxib at controlling pain in the population of dogs reported here.

CLINICAL SIGNIFICANCE

Preoperative administration of meloxicam effectively controls pain for 24 hours after combined laparoscopic ovariectomy and laparoscopic-assisted gastropexy, but rescue analgesia may be required.

Influence of sevoflurane anesthesia with mechanical ventilation and fluid-therapy on distribution of subcutaneously administered robenacoxib in dogs

Robenacoxib is a novel nonsteroidal anti-inflammatory drug approved for dogs. The present study aimed to evaluate influences of sevoflurane anesthesia on the distribution of robenacoxib in dogs. Ten healthy beagle dogs (1 to 11 years old, 9.3 to 14.3 kg body weight, 6 males and 4 females) were subcutaneously administered robenacoxib (2 mg/kg) under conscious condition or sevoflurane anesthesia inhaled a 1.3-fold predetermined individual minimum alveolar concentration of sevoflurane at a 28-day interval. The dogs under sevoflurane anesthesia were also mechanically ventilated and received fluid-therapy. On each occasion, serum samples were collected from the dogs before and at 5, 15, 30, 60, 120, 180, and 240 min after the robenacoxib administration. Serum robenacoxib concentration was measured by a liquid chromatography-tandem mass spectrometry. Maximum serum concentration of robenacoxib (C_max) was 2.2 µg/ml [range: 1.2–4.6] (median [range: minimum-maximum]) and time of C_max (T_max) was 90 min [range: 60–120] in the conscious dogs. In the sevoflurane-anesthetized dogs, the C_max significantly declined (1.3 µg/ml [range: 0.8–1.4], P=0.008) and T_max was delayed (120 min [range: 120–240], P=0.018) compared with those in the conscious dogs. The serum robenacoxib concentration at 240 min (C₂₄₀) decreased to 0.5 µg/ml [range: 0.2–0.9] in the conscious dogs, while it remained higher in the sevoflurane-anesthetized dogs (1.0 µg/ml [range: 0.3–1.4], P=0.011). In conclusion, the anesthetic procedure with sevoflurane, mechanically ventilated, and received fluid-therapy might affect the pharmacokinetics of subcutaneously administered robenacoxib in dogs.

Utilization of data below the analytical limit of quantitation in pharmacokinetic analysis and modeling: promoting interdisciplinary debate

Traditionally, bioanalytical laboratories do not report actual concentrations for samples with results below the LOQ (BLQ) in pharmacokinetic studies. BLQ values are outside the method calibration range established during validation and no data are available to support the reliability of these values. However, ignoring BLQ data can contribute to bias and imprecision in model-based pharmacokinetic analyses. From this perspective, routine use of BLQ data would be advantageous. We would like to initiate an interdisciplinary debate on this important topic by summarizing the current concepts and use of BLQ data by regulators, pharmacometricians and bioanalysts. Through introducing the limit of detection and evaluating its variability, BLQ data could be released and utilized appropriately for pharmacokinetic research.

Population variability in animal health: Influence on dose-exposure-response relationships: Part II: Modelling and simulation

During the 2017 Biennial meeting, the American Academy of Veterinary Pharmacology and Therapeutics hosted a 1-day session on the influence of population variability on dose-exposure-response relationships. In Part I, we highlighted some of the sources of population variability. Part II provides a summary of discussions on modelling and simulation tools that utilize existing pharmacokinetic data, can integrate drug physicochemical characteristics with species physiological characteristics and dosing information or that combine observed with predicted and in vitro information to explore and describe sources of variability that may influence the safe and effective use of veterinary pharmaceuticals.

Mathematical modeling and simulation in animal health. Part III: Using nonlinear mixed-effects to characterize and quantify variability in drug pharmacokinetics

A common feature of human and veterinary pharmacokinetics is the importance of identifying and quantifying the key determinants of between-patient variability in drug disposition and effects. Some of these attributes are already well known to the field of human pharmacology such as bodyweight, age, or sex, while others are more specific to veterinary medicine, such as species, breed, and social behavior. Identification of these attributes has the potential to allow a better and more tailored use of therapeutic drugs both in companion and food-producing animals. Nonlinear mixed effects (NLME) have been purposely designed to characterize the sources of variability in drug disposition and response. The NLME approach can be used to explore the impact of population-associated variables on the relationship between drug administration, systemic exposure, and the levels of drug residues in tissues. The latter, while different from the method used by the US Food and Drug Administration for setting official withdrawal times (WT) can also be beneficial for estimating WT of approved animal drug products when used in an extralabel manner. Finally, NLME can also prove useful to optimize dosing schedules, or to analyze sparse data collected in situations where intensive blood collection is technically challenging, as in small animal species presenting limited blood volume such as poultry and fish.

Modeling of Large Pharmacokinetic Data Using Nonlinear Mixed-Effects: A Paradigm Shift in Veterinary Pharmacology. A Case Study With Robenacoxib in Cats

The objective of this study was to model the pharmacokinetics (PKs) of robenacoxib in cats using a nonlinear mixed-effects (NLME) approach, leveraging all available information collected from cats receiving robenacoxib s.c. and/or i.v.: 47 densely sampled laboratory cats and 36 clinical cats sparsely sampled preoperatively. Data from both routes were modeled sequentially using Monolix 4.3.2. Influence of parameter correlations and available covariates (age, gender, bodyweight, and anesthesia) on population parameter estimates were evaluated by using multiple samples from the posterior distribution of the random effects. A bicompartmental disposition model with simultaneous zero and first-order absorption best described robenacoxib PKs in blood. Clearance was 0.502 L/kg/h and the bioavailability was high (78%). The absorption constant point estimate (Ka  = 0.68 h-1 ) was lower than beta (median, 1.08 h-1 ), unveiling flip-flop kinetics. No dosing adjustment based on available covariates information is advocated. This modeling work constitutes the first application of NLME in a large feline population.

Pharmacokinetic-pharmacodynamic analysis of spiroindolone analogs and KAE609 in a murine malaria model

Limited information is available on the pharmacokinetic (PK) and pharmacodynamic (PD) parameters driving the efficacy of antimalarial drugs. Our objective in this study was to determine dose-response relationships of a panel of related spiroindolone analogs and identify the PK-PD index that correlates best with the efficacy of KAE609, a selected class representative. The dose-response efficacy studies were conducted in the Plasmodium berghei murine malaria model, and the relationship between dose and efficacy (i.e., reduction in parasitemia) was examined. All spiroindolone analogs studied displayed a maximum reduction in parasitemia, with 90% effective dose (ED₉₀) values ranging between 6 and 38 mg/kg of body weight. Further, dose fractionation studies were conducted for KAE609, and the relationship between PK-PD indices and efficacy was analyzed. The PK-PD indices were calculated using the in vitro potency against P. berghei (2× the 99% inhibitory concentration [IC₉₉]) as a threshold (TRE). The percentage of the time in which KAE609 plasma concentrations remained at >2× the IC₉₉ within 48 h (%T_>TRE) and the area under the concentration-time curve from 0 to 48 h (AUC_0–48)/TRE ratio correlated well with parasite reduction (R² = 0.97 and 0.95, respectively) but less so for the maximum concentration of drug in serum (C_max)/TRE ratio (R² = 0.88). The present results suggest that for KAE609 and, supposedly, for its analogs, the dosing regimens covering a T_>TRE of 100%, AUC_0–48/TRE ratio of 587, and a C_max/TRE ratio of 30 are likely to result in the maximum reduction in parasitemia in the P. berghei malaria mouse model. This information could be used to prioritize analogs within the same class of compounds and contribute to the design of efficacy studies, thereby facilitating early drug discovery and lead optimization programs.

Pharmacokinetic/pharmacodynamic modeling for the determination of a cimicoxib dosing regimen in the dog

BACKGROUND

Cimicoxib is a new coxib anti-inflammatory drug for use in the dog. To determine a preclinical dosage regimen for cimicoxib in dog, a reversible model of kaolin-induced paw inflammation was used. Dosage regimens were established using pharmacokinetic/pharmacodynamic (PK/PD) modeling approach (indirect response model).

RESULTS

Analgesic, anti-inflammatory and antipyretic endpoints investigated with the inflammation model established the efficacy of cimicoxib at a dose of 2 mg/kg administered orally (single dose) in 12 beagle dogs.For both the oral and IV route of administration two groups of dogs to be identified namely Poor Metabolizers (PM) and Extensive Metabolizers (EM).The terminal half-life after oral administration was 8.0 ± 0.6 h for the PM and 4.6 ± 2.6 h for the EM groups, with the corresponding values after the IV route being 5.6 ± 1.7 h and 2.7 ± 0.9 h (mean ± SD).The main pharmacodynamic parameters (potency, efficacy, and sensitivity) were estimated for four endpoints (body temperature, creeping speed, ground vertical reaction force and clinical lameness score). The plasma concentration corresponding to half the maximum of the indirect effect were 239 μg/L for creeping speed, 284 μg/L for the lameness score, 161 μg/L for the ground reaction vertical force and 193 μg/L for the body temperature.To document possible polymorphism of the cimicoxib disposition in the target dog population, cimicoxib was administered by the intravenous route to 40 dogs (four different sized breeds). The cimicoxib half-lives in these 40 dogs were of same order of the magnitude as those of the EM beagle dogs. Thus pharmacokinetic and pharmacodynamic parameters obtained from the EM beagle dogs were selected to simulate the dose-effect relationship of cimicoxib after an oral administration allowing a dosage regimen to be selected for confirmation by a clinical trial.

CONCLUSIONS

Cimicoxib was an efficacious anti-inflammatory, antipyretic and analgesic drug and a dosage regimen of 2 mg/kg daily was determined for confirmatory clinical trials.

Application of population pharmacokinetics for preclinical safety and efficacy studies

From the beginning of the 1980s, population PK has been primarily used in clinical development and only in the last decade has it been convincingly applied in a preclinical setting. Sparse sampling and covariate analyses are key features of preclinical popPK, useful for toxicology and efficacy studies in animals to assemble data obtained from different studies; for describing individual PK and PD; for building mechanistic models; and for performing interspecies scaling-up of disposition and efficacy. Application in disease models, mainly in behavioral and neurological models, allows the quantitative description of PK and PD without frequent blood sampling and recurrent physiological measurements, which are the critical and compromising perturbations of experimental systems. A preclinical population approach to PK and PD, by its versatility and possibility of simulating ‘what if’ scenarios, offers a unique and potent tool in the development of new drugs, in particular biologics.

The effect of robenacoxib on the concentration of C-reactive protein in synovial fluid from dogs with osteoarthritis

Background

Robenacoxib is a novel and highly selective inhibitor of COX-2 in dogs and cats and because of its acidic nature is regarded as being tissue-selective. Thirty four dogs with stifle osteoarthritis secondary to failure of the cranial cruciate ligament were recruited into this study. Lameness, radiographic features, synovial cytology and C-reactive protein concentrations in serum and synovial fluid were assessed before and 28 days after commencing a course of Robenacoxib at a dose of 1 mg/kg SID.

Results

There was a significant reduction in the lameness score (P < 0.01) and an increase in the radiographic score (P < 0.05) between pre- and post-treatment assessments. There was no difference between pre- (median 1.49 mg/l; Q1-Q3 0.56-4.24 mg/L) and post – (1.10 mg/L; 0.31-1.78 mg/L) treatment serum C-reactive protein levels although synovial fluid levels were significantly reduced (pre- : 0.44 mg/L; 0.23-1.62 mg/L; post- : 0.17 mg/L; 0.05-0.49 mg/L) (P < 0.05). There was no correlation between C-reactive protein concentrations in serum and matched synovial fluid samples.

Conclusions

Robenacoxib proved effective in reducing lameness in dogs with failure of the cranial cruciate ligament and osteoarthritis of the stifle joint. The drug also reduced levels of C-reactive protein in the synovial fluid taken from the affected stifle joint. Robenacoxib appears to reduce articular inflammation as assessed by C-reactive protein which supports the concept that Robenacoxib is a tissue-selective non-steroidal anti-inflammatory drug.