Carprofen Attenuates Postoperative Mechanical and Thermal Hypersensitivity after Plantar Incision in Immunodeficient NSG Mice

Abstract

Immunodeficient NSG mice are reported to be less responsive to buprenorphine analgesia. Here, we used NSG mice to compare the efficacy of the commonly used dose of carprofen (5 mg/kg) with 5 and 10 times that dose (25 and 50 mg/kg) for attenuating postoperative mechanical and thermal hypersensitivity following an incisional pain model. Male and female NSG mice (n = 45) were randomly assigned to one of 4 groups and received daily subcutaneous injections for 3 d: saline (5 mL/kg), 5 mg/kg carprofen (Carp5), 25 mg/kg carprofen (Carp25), and 50 mg/kg carprofen (Carp50). Mechanical and thermal hypersensitivity were assessed 24 h before and at 4, 24, and 48 h after surgery. Plasma carprofen concentrations were measured in a separate group of mice (n = 56) on days 0 (at 2, 4, 12, and 23 h), 1, and 2 after the first, second, and third doses, respectively. Toxicity was assessed through daily fecal occult blood testing (n = 27) as well as gross and histopathologic evaluation (n = 15). Our results indicated that the saline group showed both mechanical and thermal hypersensitivity throughout the study. Carp5 did not attenuate mechanical or thermal hypersensitivity at any time point. Carp25 attenuated mechanical and thermal (except for the 4-h time point) hypersensitivity. Carp50 attenuated only thermal hypersensitivity at 24 h. Fecal occult blood was detected in 1 of 8 Carp25-treated mice at 48 and 72 h. Histopathologic abnormalities (gastric ulceration, ulcerative enteritis, and renal lesions) were observed in some Carp50-treated mice. Plasma carprofen concentrations were dose and time dependent. Our results indicate that Carp25 attenuated postoperative mechanical and thermal hypersensitivity more effectively than Carp5 or Carp50 in NSG mice with incisional pain. Therefore, we recommend providing carprofen at 25 mg/kg SID for incisional pain procedures using immunodeficient NSG mouse.

Introduction

Providing appropriate analgesia to animals undergoing painful procedures is an ethical and moral obligation. The most common analgesic classes used in mice are opioids (for example, buprenorphine), NSAIDs (for example, carprofen and meloxicam), and local anesthetics (for example, lidocaine and bupivacaine).^{1,14,17,18,21} NSAIDs are commonly used to treat mild to moderate pain related to inflammation, either alone or in combination with buprenorphine or local anesthetics to provide multimodal analgesia to mice.^1,45,46 NSAIDs inhibit cyclooxygenase (COX), the enzyme that converts arachidonic acid to its metabolites, including prostaglandins, prostacyclin, and thromboxanes.^33,63 The analgesic properties of NSAID are mainly attributed to the inhibition of eicosanoids that elicit inflammation, pain, and fever due to tissue damage or infection.^33,63 Carprofen, a COX-2-selective NSAID, has low affinity for the COX-1 isoform and thereby can better control inflammation and pain and has less toxicity.⁴⁹ Carprofen is available in parenteral formulations and orally as a flavored chewable tablet or additive to drinking water.^8,15,25,46

Many factors impact analgesia efficacy, including drug formulation, dose, and the strain, sex, and age of the animals.^{32,52,53,58,60,61} Currently, commonly recommended mouse carprofen dosing regimens are 2 to 5 mg/kg every 12 to 24 h;^19,20 however, limited literature is available that specifically evaluates efficacious carprofen doses for mice. Previous studies have reported that the effective dose is higher than the frequently recommended dose.^26,39 For example, a study using CD-1 mice found that carprofen can prevent spontaneous pain after laparotomy but only at doses 2 to 4 times (29 mg/kg) higher than the currently recommended mouse dose (2 to 5 mg/kg).³⁹ A sham mouse embryo transfer surgery study assessing different strains of mice (C57BL/6J, DBA/2J, and B6D2-Tg [Pr-mSMαActin)] V5rCLR-25) found that only a high carprofen dose (50 mg/kg) improved the nest complexity score.²⁶ Higher NSAID doses are potentially risky, as they may cause gastric and intestinal ulcerations,^5,22,37,65 altered platelet function,^37,40,52 and nephrotoxicity.⁴⁰ In addition, subcutaneous administration of carprofen at 5-, 2.5-, and 1.25-mg/kg doses was reported to interfere with early phase wound healing in male Wistar rats.⁶¹

Little information is available on efficacious analgesic dosing regimens for immunodeficient mouse strains including the commonly used NOD.Cg-Prkdc^scid Il2rg^tm1Wjl/SzJ (NSG) mice. The NSG strain was generated by knocking out the Il2rg gene required for natural killer cell development and the NOD-scid mutation eliminates immunologically active B and T cells.^56,57 NSG mice are used in numerous biomedical research fields including oncology, immunology, regenerative medicine, drug discovery and infectious disease.^6,55,56 Although the unique immunology of NSG mice makes them a useful research model, the immune impairments also affect analgesic efficacy, particularly with regard to opioids.⁵⁰ A recent study of NSG mice found that buprenorphine did not adequately attenuate mechanical and thermal hypersensitivity in a plantar incision model.⁴ Carprofen may be an efficacious alternative analgesic for NSG mice, but to our knowledge, its dosing regimen and adverse effects have not been previously evaluated in mice in general or in specific immunodeficient mouse strains.

The goals of this study were to 1) evaluate carprofen analgesic efficacy at 3 different doses (5, 25, and 50 mg/kg SID, SC) by using a plantar incisional pain model and mechanical and thermal hypersensitivity testing; 2) characterize the pharmacokinetic profiles of these 3 dosing regimens (5, 25, and 50 mg/kg of carprofen) over a 24-h time period before drug redosing and after repeated dosing at 24 and 48 h; and 3) perform a selective toxicological evaluation of the 3 different dosages by performing fecal occult blood testing and histopathologic examination. We hypothesized the higher carprofen doses (25 and 50 mg/kg) would attenuate mechanical and thermal hypersensitivity and maintain therapeutic blood plasma levels more effectively than a low dose of carprofen (5 mg/kg) in NSG mice. We also hypothesized the highest dose (50 mg/kg) would increase the incidence of positive fecal occult blood results and cause more histopathologic evidence of gastrointestinal ulceration than would the lower doses (5 and 25 mg/kg).

Materials and Methods

Mice.

Adult male and female NOD.Cg-Prkdc^scid Il2rgt^m1Wjl/SzJ (NSG) mice (mean weight 28 g [range, 19 to 35 g]; average age 12 wk [range, 8 to 30 wk]) were used to test hypersensitivity (n = 45), measure carprofen plasma concentrations (n = 56), and evaluate fecal occult blood (n = 27) and histopathology (n = 15).

Mice were bred in-house in a barrier facility using mice obtained from The Jackson Laboratory (Bar Harbor, ME). Breeder mice were not regularly replaced so mice could be viewed as a substrain of the initially purchased mice. The mice were housed in same-sex groups in the barrier facility until the start of the study in disposable, irradiated IVC cages (Innovive, San Diego, CA) that had been prefilled with corncob bedding and included Enviro-dri (Lab Supply, Fort Worth, TX) for enrichment. Sentinel mice were free of mouse hepatitis virus, mouse parvovirus, mouse rotavirus (EDIM), minute virus of mice, Sendai virus, Theiler murine encephalomyelitis virus, ectromelia virus, murine adenovirus 1 and 2, pneumonia virus of mice, lymphocytic choriomeningitis virus reovirus, murine norovirus, Helicobacter spp., Mycoplasma pulmonis, Rodentibacter pneumotropicus (Pasteurella pneumotropica), pinworms, and endo- and ectoparasites.

At least one week before the experiment started, mice were transferred and acclimated to a holding room under the same type of housing conditions in a conventional facility that contained the equipment required for daily hypersensitivity testing. Rooms were maintained on a 12:12-h dark:light cycle (lights on at 0700, lights off at 1900, and fluorescent lighting, 100 to 400 lx), at 68 to 79 °F (20 to 26 °C), and 30% to 70% relative humidity. Autoclaved commercial rodent diet (Teklad Global; 18% Protein Rodent Diet 2018SX, Envigo, Indianapolis, IN) and chlorinated water (Aquavive; Innovive, San Diego, CA) were provided ad libitum. All experiments in this study were conducted with approval from the Stanford University IACUC (Administrative Panel for Laboratory Animal Care), and all animals were handled and treated according to the Guide for the Care and Use of Laboratory Animals.²⁴

Hypersensitivity testing.

Study design.

Mice (n = 45) were acclimated to the housing and experimental room for at least 72 h before baseline testing. Male (n = 21) and female (n = 24) mice were randomly grouped into one of 4 treatment groups. One day before surgery (day 1 [D1]), baseline body weights and mechanical and thermal hypersensitivity data were collected. On the day of the surgical procedure (day 0 [D0]), 1-mL tuberculin syringes with 25-G needles were used to inject mice subcutaneously in the interscapular region with saline (saline; 5 mL/kg; n = 12 [6 males, 6 females], 0.9% sodium chloride; Hospira, Lake Forest, IL]; carprofen 5 mg/kg (Carp5, n = 11 [5 males, 6 females]; Norbrook Laboratories, Rossmore Industrial Estate, Ireland); carprofen 25 mg/kg (Carp25, n = 11 [5 males, 6 females], ibid) or carprofen 50 mg/kg (Carp50, n = 11 [5 males, 6 females], ibid). The 50-mg/mL stock solution of carprofen was diluted with sterile water to make 5-mg/mL concentrations for Carp25 and Carp50 groups and a 0.5-mg/mL concentration for Carp5 group.

All injections were administered 2 to 4 min before the incisional paw surgery was performed. The treatments were repeated every 24 h on the next 2 consecutive days (D1 and D2) by manually scruffing the mice. Mice were tested for both mechanical and thermal hypersensitivity 4 h after drug administration on D0 and an hour after drug administration on D1 and D2 by the same experimenter, who was blind to the treatment. Mechanical hypersensitivity testing always preceded thermal hypersensitivity testing. Upon completion of the final hypersensitivity assessment on D2, mice were euthanized by carbon dioxide asphyxiation and a postmortem examination was performed to assess gross pathology.

Mechanical hypersensitivity testing.

Mechanical stimuli responses were evaluated using the von Frey monofilament nociceptive assay. Mice were transferred to bottomless clear plastic chambers (10.1 × 10.1 × 12.5 cm) on an elevated wire mesh platform (Electronic von Frey Mesh Stand; IITC Life Science, Woodland Hills, CA). After a 15-min acclimation period, a calibrated Semmes-Weinstein von Frey filament (0.4 g) was applied 10 times to the plantar surface of the contralateral (control) and ipsilateral (test) hind paws at random locations excluding the heel, pads, and toes. For every trial, filaments were applied for 1 to 2 s with an interstimulus interval of approximately 5 s. Any nocifensive behavior (withdrawal, shaking, licking of the stimulated paw) was considered a positive response; absence of the behavior was considered a negative response. Mechanical hypersensitivity was defined as a significant increase in the frequency of paw withdrawals after the mechanical stimulus as compared with baseline (the value obtained 24 h before the incision). Mechanical hypersensitivity testing always preceded thermal hypersensitivity testing.

Thermal hypersensitivity testing.

Thermal stimuli responses were evaluated using the Hargreaves nociceptive assay. Mice were transferred into bottomless clear plastic chambers (10.1 × 10.1 × 12.5 cm) and acclimated for 15 min before testing on an elevated glass platform preheated to 29 °C. Once the mice were acclimated, a focal (4 × 4 mm) thermal stimulus with a 25% beam intensity produced by radiant heat generated from a 50-W light bulb (Plantar Analgesia Meter; IITC Life Science, Woodland Hills, CA) was directed to the plantar surface of each hind paw. A cutoff time of 20 s was used as a maximum to prevent tissue injury. Each paw was tested 4 times, with a minimum of a 4 min interval between each trial. Each mouse’s right (contralateral) hind paw served as a control for the left (ipsilateral) hind paw. Mean withdrawal latency was calculated by taking the average of the last 3 trials, omitting the first trial in all cases. The criteria used for positive and negative responses were the same as for the mechanical hypersensitivity assay. Thermal hypersensitivity was defined as a significant decrease in paw withdrawal latency in response to the onset of focal thermal stimuli as compared with baseline (the value obtained 24 h before the incision).

Surgery.

Mice were placed singly in an induction chamber and inhalant anesthesia was induced using 4% to 5% isoflurane in 100% oxygen at a delivery rate of 2 L/min. Anesthesia was then maintained using a nose cone with 2% to 2.5% isoflurane in 100% oxygen at a delivery rate of 1 L/min. Sterile ophthalmic ointment was applied to both eyes, and injections were administered subcutaneously in the interscapular region 2 to 4 min before the incision. A single dose of Cefazolin (30 mg/kg SC; GlaxoSmithKline, Research Triangle Park, NC) and prewarmed 0.9% saline (5 mL/kg SC) were administered subcutaneously over the right shoulder. The mice were kept on a warm water circulating blanket for the entire duration of the surgery. Mice were placed in sternal recumbency, and the left hind paw was gently retracted and taped to stabilize it for the surgery. The contralateral (right) hindpaw was left unincised and served as a control for each mouse. The plantar surface of the left paw was aseptically prepared with 3 alternating povidone-iodine swabs (Povidone-Iodine Swabsticks; PDI, Orangeburg, NY) and 3 alcohol wipes. The paw was covered with a sterile drape, and the paw withdrawal reflex was checked by pinching the right paw. After a surgical plane of anesthesia was confirmed by the lack of paw withdrawal, a No.15 blade was used to make a 0.5-cm skin incision along the plantar aspect of the paw beginning 0.3 cm from the tibiotarsal joint and extending distally. The plantaris muscle was then separated from the surrounding connective tissue and raised using curved iris tissue forceps. A stab incision was made through the muscle’s center without disturbing its attachment, and a second curved iris tissue forcep was inserted into the incision to apply gentle lateral traction to the muscle for 10 s. Saline was applied to the surgical area and absorbed using a sterile cotton-tipped swab. The incision was closed with a single horizontal mattress suture using 4-0 silk. The total surgery time for each mouse was less than 10 min. Mice were transferred to the recovery cage with thermal support and monitored until fully ambulatory and then returned to their home cage.

Clinical observations.

All mice were closely monitored daily for any abnormal behaviors or clinical signs (for example, behavioral pain indicators, altered activity levels, posture, mobility, hydration, and feces). Mice were weighed before surgery for drug dosing and daily thereafter at the conclusion of hypersensitivity testing. A single individual who was blind to the experimental groups made all evaluations.

Statistical analysis.

Data sets were analyzed by using 2-way repeated measures ANOVA with Bonferroni correction for multiple comparisons (R Development Core Team, 2015) to identify significant differences within and between groups and over time. Data were expressed as means ± SEM. A P value of less than 0.05 was considered significant.

Group size determinations.

Sample size was calculated (0.25 effect size, 0.4 correlation of repeated measurements, and 0.05 significance level) and indicated that at least 11 mice per treatment would provide power of approximately 84%. Power analysis was not performed for plasma concentration. Randomization of mice was performed by selection of lottery tickets for the 4 treatments. Two mice that initially appeared unhealthy when examined by the veterinarian were excluded.

Plasma drug concentration analysis.

Study design.

Male (n = 27) and female (n = 29) mice were randomly assigned to one of the same 4 treatment groups as for hypersensitivity assessment (Figure 1). To measure plasma concentration over 24 h after the initial dose, blood was collected at 2, 4, 12, and 23 h after administration of the first dose administration. To determine the concentration after repeated dosages, blood was also collected 1 h after administration of the second and third doses at 24 and 48 h, respectively. Mice were manually restrained by scruffing them around the neck and between the shoulders and holding the tail to allow subcutaneous injection of saline (n = 2 females), Carp5 (n = 18), Carp25 (n = 18), and Carp50 (n = 18) in the interscapular region with a 1-mL tuberculin syringe and a 25-gauge needle. Three mice per group per time point underwent terminal blood collection except for the saline group, from which blood sample was collected from 2 female control mice at the 2-h time point. Both sexes were represented in each carprofen-treated group.

Figure 1.

Workflow of experimental procedures. (A) Hypersensitivity testing (saline, n = 12; Carp5, n = 11; Carp25, n = 11; and Carp50, n = 11). (B) Plasma drug determination (saline, n = 2; Carp5, n = 18; Carp25, n = 18; and Carp50, n = 18). (C) Fecal occult blood testing (saline, n = 3; Carp5, n = 8; Carp25, n = 8; and Carp50, n = 8), and gross necropsy and histopathologic evaluation (saline, n = 3; Carp5, n = 4; Carp25, n = 4; and Carp50, n = 4).

Plasma collection.

Mice were anesthetized in an induction chamber by delivering 4% to 5% isoflurane in 100% O₂ at a 2 L/min flow rate. As soon as mice were removed from the induction chamber and a surgical anesthetic plane confirmed via toe pinch, a retroorbital technique was used to collect 0.6 to 1 mL of whole blood into a nonheparinized capillary tube; the blood was then transferred into a 1-mL lithium heparin microtainer. Euthanasia was performed by exsanguination of mice followed by cervical dislocation as a secondary euthanasia method. The blood samples were spun in a microcentrifuged at 3,451 × g for 20 min. Plasma was removed, transferred to 1-mL cryogenic tubes, and stored at −80 °C until analysis.

Plasma carprofen concentration analysis.

Plasma carprofen concentration was measured by the Pharmaceutical Sciences Research Institute at the McWhorter School of Pharmacy (Samford University, Birmingham, AL) via liquid chromatography-tandem mass spectrometry (HPLC MS/MS). Individual samples had a volume of 0.2 mL and were shipped overnight on dry ice. Calibration standards, blanks, and quality controls were prepared by spiking mouse plasma (20 µL) with the appropriate carprofen amount to achieve plasma concentrations that ranged between 1 and 500 µg/mL. Standards, blanks, quality controls, and samples were spiked with internal standard (IS; 10 µL of 100 µg/mL ketoprofen in acetonitrile), and 200 µL of acetonitrile containing 0.1% formic acid was added to precipitate the proteins. The samples were vortexed and centrifugated for 5 min at 21,130 × g. The supernatant was transferred to 0.5-mL autosampler vials and analyzed by HPLC with UV detection at 254 nm.

Chromatographic separation of the compounds was achieved using a Shimadzu 2010 HPLC system (Shimadzu Scientific Instruments, Columbia, MD) with a 150 × 4.6 mm Luna C18 reverse phase column (Phenomenex, Torrance, CA) at ambient temperature. Mobile phase A consisted of deionized water containing 0.1% formic acid and mobile phase B consisted of Acetonitrile. The carprofen and the internal standard (ketoprofen) were analyzed using a gradient elution profile in which mobile phase B was held at 50% for 1 min, then increased to 90% over 4 min, held at 90% for 1 min, returned to 55% over 0.5 min, and equilibrated for 3.5 min. The flow rate was set at 1 mL/min and 10 µL was injected onto the column. Carprofen was eluted at 5.4 min and ketoprofen (IS) at 4.4 min.

Toxicity assessment.

Study design.

Carprofen was administered to separate groups of mice as described for the plasma drug concentration analysis (Figure 1). Mice were randomly assigned to groups and injected subcutaneously once on D0 using 1-mL tuberculin syringes with 25-gauge needle as follows: saline, n = 3, 2 males and 1 female; Carp5, n = 8, 2 males and 6 females; Carp25, n = 8, 2 males and 6 females; or Carp50, n = 8, 2 males and 6 females. Injections were repeated on the following 2 consecutive days (D1 and D2).

Fecal occult blood test.

Fecal occult blood tests were performed immediately after fecal sample collection at 0 h (D0) immediately before any drug administration, and at 4 (D0), 24 (D1), 48 (D2), and 72 (D3) h after drug administrations. To collect fecal samples, mice were transferred to a clear plastic chamber on top of a clean absorbent pad until they produced a fecal sample. One to two fecal pellets were collected from each mouse and transferred to a 1.5-mL microcentrifuge tube. One drop of sterile water was added to the tube, and the fecal sample was crushed using a wooden applicator stick. Fecal occult blood tests were performed immediately after collection (Hemoccult, Beckman Coulter, Brea, CA). The front flap of the slide was opened, and the fecal material was smeared with the applicator strip into two different spots of the testing slide labeled with box A and box B. The flap was closed, and the test was kept at room temperature for 3 to 5 min. The back flap of the slide was opened, and 2 drops of Hemoccult Developer were directly added over each smear. Tests were read after 60 s. A positive fecal occult blood was indicated by any trace of blue on or at the edge of the smear. Negative results were recorded if blue was not seen on or at the edge of the smear.

Gross and histopathologic evaluation.

On D3 after fecal sample collection and fecal occult blood testing, mice were euthanized by carbon dioxide asphyxiation, and gross postmortem evaluations were performed. Tissues from control (n = 3) and treated mice (n = 4/group), including liver, spleen, kidney, heart, lung, stomach, duodenum, ileum, cecum, and colon, were immersion fixed in 10% neutral buffered formalin for 72 h. Formalin-fixed tissues underwent routine processing and were embedded in paraffin, sectioned at 5 µm, and stained with hematoxylin and eosin. Tissues were evaluated by a board-certified veterinary pathologist who was blind to the treatment. Photomicrographs were taken on an Olympus BX46 microscope equipped with an Olympus DP27 camera using Olympus cellSens Standard imaging software.

Results

Hypersensitivity testing.

To determine the effect of sex, it was included in the analysis and considered as a block. No sex-dependent differences were observed; therefore, data from male and female mice were combined for further analysis to increase the sample size for each drug group and increase the power of the test in the analysis.

Mechanical hypersensitivity.

Baseline (D-1) sensitivities of ipsilateral (left, incised) and contralateral (right, control) hind paws to von Frey filaments were not significantly different. Baseline responses were not different among treatment groups.

Incised hind paw (Figure 2A and Table 1): In the saline, Carp5, and Carp50 groups, mechanical hypersensitivity was significantly higher than baseline (D-1) on D0, D1, and D2 (P < 0.0004). The Carp25 group showed attenuation of mechanical hypersensitivity at all study time points, and the values were not significantly different from baseline (D-1). On D0, the saline group showed significantly higher mechanical hypersensitivity when compared with all the carprofen groups (Carp5; P = 0.003, Carp25; P < 0.0001, and Carp50; P = 0.001). On D1 and D2, hypersensitivity of the saline group was significantly greater than that of the Carp25 (D1; P = 0.004, D2; P = 0.003) and Carp50 groups (D1; P = 0.048, D2; P = 0.033). The carprofen groups showed no significant differences in hypersensitivity attenuation throughout the study (D0, D1, and D2).

Figure 2.

(A) Mechanical hypersensitivity: measured as the number of paw withdrawals (mean ± SEM) of ipsilateral (left) hind paw. (B) Thermal hypersensitivity: thermal latency measured in seconds (mean ± SEM) was defined as the response time of the ipsilateral (left) hind paw to thermal stimulus. Arrow indicates surgery on D0. *, P < 0.05, value is significantly different when compared with baseline (D-1) value of the same treatment group. #, P < 0.05, value is significantly different when compared with the saline group at a specific time point.

Table 1.

Mechanical and thermal sensitivity of the incised hind paw in NSG mice

Mechanical hypersensitivity (number of withdrawals of ipsilateral hind paw) (mean ± SEM)
Time points	Saline (n = 12)	Carp5 (n = 11)	Carp25 (n = 11)	Carp50 (n = 11)
D-1	1.7 ± 0.2	1.5 ± 0.3	2.4 ± 0.2	1.7 ± 0.2
D0	6.7 ± 0.5*	4.7 ± 0.4*	3.5 ± 0.4	4.5 ± 0.5*
D1	5.8 ± 0.4*	4.5 ± 0.5*	3.7 ± 0.4	4.2 ± 0.5*
D2	5.9 ± 0.5*	4.4 ± 0.5*	3.8 ± 0.4	4.3 ± 0.6*
Thermal hypersensitivity (thermal latency of ipsilateral hind paw measured in seconds) (mean ± SEM)
D-1	19.1 ± 0.6	19.1 ± 0.4	18.4 ± 0.8	19.1 ± 0.6
D0	4.9 ± 1.4*	9.7 ± 1.8*	13.0 ± 1.5*	12.0 ± 1.9*
D1	6.2 ± 1.1*	11.7 ± 1.7*	14.9 ± 1.1	14.4 ± 1.7
D2	5.3 ± 0.8*	12.13 ± 1.7*	15.2 ± 1.5	11.6 ± 2.1*

^* P < 0.05, value is significantly different when compared with baseline (D-1) value of the same treatment group.

Intact hind paw: The paw withdrawal responses to mechanical stimuli did not differ significantly between treatment groups at any time point throughout the study (data not shown).

Thermal hypersensitivity.

The baseline (D-1) responses to thermal stimuli responses did not differ significantly between the ipsilateral (left, incised) and contralateral (right, control) hind paws, and baseline responses were not different among groups.

Incised hind paw (Figure 2B and Table 1): The thermal latency of saline and Carp5 groups was significantly lower throughout the study time as compared with their respective baseline values (P < 0.003). Thermal latency of Carp25 group was significantly lower on D0 (P = 0.043) as compared with D-1; however, on D1 and D2 thermal hypersensitivity was attenuated, and the thermal latency was similar to the baseline. Carp50 group thermal hypersensitivity was attenuated only on D1 and thermal latency was significantly lower on D0 and D2 (P < 0.002) as compared with D-1. On D0, thermal latency of the saline group was significantly lower than those of the Carp25 and Carp50 groups (P < 002). On D1 and D2, thermal latency of the saline group was significantly lower than that of all the carprofen groups (P < 0.03). No significant differences in attenuation of thermal hypersensitivity were detected among the carprofen groups throughout the study (D0, D1, and D2).

Intact hind paw: Responses to thermal stimuli did not differ significantly between treatment groups at any time point throughout the study (data not shown).

Body weight.

Baseline body weight did not differ between treatment groups. Males weighed significantly more than females (P < 0.05). Mouse body weights did not significantly differ between treatment groups or at any study time point (data not shown).

Clinical observations.

Limping and/or licking of the incised hind paw was observed in 73% (8 of 11) mice in the saline group, 55% (6 of 11) Carp5 mice, 45% (4 of 11) Carp25 mice, and 73% (8 of 11) Carp50 mice during hypersensitivity testing at all study time points. No other abnormal behaviors were observed in saline, Carp5, and Carp25 groups throughout the study; however, the Carp50 mice had subjectively more locomotor activity throughout the study.

Gross pathology.

After the last hypersensitivity assessment (D2), mice were euthanized, and a postmortem examination was performed for all the animal tissues including the incised paws by an individual who was blind to the animal’s groups. No gross abnormalities were observed in any treatment group.

Plasma carprofen concentration analyses.

Two saline-injected mice were used as negative controls at the 2-h time point (data not shown). To measure plasma concentration over 24 h, samples were collected at 2, 4, 12, and 23 h after the first carprofen dose. To determine the concentrations after repeated dosages, samples were also collected at 24 and 48 h, one hour after the second and third carprofen doses at 23 and 47 h, respectively. The time courses of plasma carprofen concentrations are shown in Figure 3. In the Carp5 group, the plasma concentration at the 2-h time point was not significantly lower or higher than concentrations measured at 4, 12, 23, 24, or 48 h. In the Carp25 group, plasma concentrations were significantly lower at 12- and 23-h time points when compared with the 2-h time point of the same group (P < 0.001). However, the concentration at the 24-h time point was significantly higher than the 2-h time point of the same group (P < 0.001); concentrations at the 4- and 48-h time points were not significantly lower or higher than the 2-h time point values. In the Carp50 group, concentrations were significantly lower from 2-h values at 12, 23, and 48 h after drug administration (P < 0.048); however, values at 4 and 24 h were not significantly lower or higher from 2-h values. Concentrations at 2-, 4-, 24-, and 48-h time points differed significantly between all treatment groups (P < 0.003). At the 12-h time point, only the Carp50 group’s concentration was significantly higher than the Carp5 group’s concentration at the same time point (P = 0.001). No significant differences were detected among treatment groups at 23 h.

Figure 3.

Plasma Carprofen concentrations (μg/mL, mean ± SEM) in mice (n = 3/group/time point) treated with either Carp5, Carp25, or Carp50. Plasma drug concentrations were measured at 2, 4, 12, 23, 24, and 48 h. Second and third doses were administered at 23 and 47 h, respectively.

Toxicity assessment.

Fecal occult blood test.

Fecal occult blood tests were performed immediately after fecal sample collection at 0 h (D0) immediately before any drug administration and at 4 (D0), 24 (D1), 48 (D2), and 72 (D3) h after drug administration. Fecal occult blood was not detected in saline-treated mice at any time point. Positive fecal occult blood was detected in 12.5% (1 of 8) of mice treated with Carp25 at 48 (D2) and 72 (D3) h (the same mouse tested positive at both time points). Fecal occult blood was not detected in the Carp5 and Carp50 groups at any time points.

Histopathologic findings.

A summary of histologic findings is shown in Table 2. Gastric ulceration occurred in 50% of mice treated with Carp50 (Figure 4) but was not present in any other treatment groups or controls. When present, gastric ulceration was located at the gastric cardia at the junction of squamous and glandular epithelium. Lesions were characterized by loss of overlying squamous or glandular epithelium with abundant mucosal karyorrhectic debris and fibrin (Figure 4, asterisks). Ulcerative enteritis with villus blunting and fusion was identified in 25% of mice treated with Carp50. Renal lesions were limited to those mice receiving Carp50 and consisted of mild neutrophilic tubulitis (50% of mice) and mild interstitial fibrosis and tubular regeneration (75% of mice). No histologic abnormalities were observed in the heart, liver, spleen, duodenum, jejunum, ileum, cecum, or colon.

Table 2.

Summary of histopathologic findings

	Gastrointestinal		Kidney
Carprofen dose	Gastric ulceration	Ulcerative enteritis	Neutrophilic tubulitis	Interstitial fibrosis and tubular regeneration
0 (saline)	0 of 3	0 of 3	0 of 3	0 of 3
5 mg/kg	0 of 4	0 of 4	0 of 4	0 of 4
20 mg/kg	0 of 4	0 of 4	0 of 4	0 of 4
50 mg/kg	2 of 4	1 of 4	2 of 4	3 of 4

Figure 4.

Histopathologic findings. (A, B) Gastric ulceration was identified in 2 of 4 mice treated with Carp50. Lesions consisted of abundant mucosal karyorrhectic debris and fibrin (asterisks). (C) Normal gastric mucosa from a control mouse. Hematoxylin and eosin. Scale bar = 50 µm.

Discussion

This study is the first to investigate the efficacy of carprofen for attenuating hypersensitivity in immunodeficient (NSG) mice after the creation of incisional pain. The aims of this study were to 1) evaluate carprofen’s analgesic efficacy at 3 different doses (5, 25, and 50 mg/kg SID, SC) using mechanical and thermal hypersensitivity testing after creating incisional pain; 2) characterize the pharmacokinetic profile of these 3 dosing regimens over 48 h; and 3) perform a toxicological evaluation using fecal occult blood testing and histopathologic examination. Our results indicated carprofen at 25 mg/kg effectively attenuated mechanical (on D0, D1, and D2) and thermal (on D1 and D2) hypersensitivity as compared with baseline in NSG mice after a plantar paw incision. Carp25 attenuated both mechanical and thermal hypersensitivity at all time points tested as compared with the saline group. The carprofen dosages of 5 and 50 mg/kg did not attenuate mechanical hypersensitivity at any time point as compared with their respective D-1 values. However, Carp5 (on D0) and Carp50 attenuated mechanical hypersensitivity on D0 and on D0, D1, and D2, respectively, as compared with saline. As compared with their D-1 values, Carp5 did not attenuate thermal hypersensitivity, and Carp50 attenuated thermal hypersensitivity only on D1. As compared with saline, Carp5 and Carp50 attenuated thermal hypersensitivity on D1 and D2 and on D0, D1, and D2, respectively. The plasma carprofen concentration peaked during the first 4 h after administration of the first dose and then gradually fell until administration of the second dose. The plasma carprofen concentration in the Carp25 group was significantly higher than that of the Carp5 group until the 12-h time point.

Paw incisional pain provides a reliable model for animal pain research because it produces sufficient and measurable hypersensitivity in mice and rats^{4,7,9,11,13,27,40,44,48,54,66} Mechanical hypersensitivity was assessed using the von Frey monofilament test and thermal hypersensitivity was assessed with the Hargreaves test. We used these testing modalities as our group has extensive prior experience using them in surgical models of pain. Hypersensitivity to mechanical and thermal stimuli in this model can vary as a function of animal species,^7,48 strain^42,58 and sex.⁵⁷ For example, a study using C57BL/6 mice reported that mechanical and thermal hypersensitivity lasted for 2 d, whereas hypersensitivity lasted for 2 to 7 d in a study using C3H/He mice.^44,48 A previous study that used NSG mice reported that mechanical and thermal hypersensitivity lasted for 2 to 3 d.⁴ In our current study, mechanical and thermal hypersensitivity was detected in saline-treated mice as soon as 4 h after surgery and lasted throughout the 48-h duration of the study.

Our recent work using the paw incisional pain model in NSG mice indicated that the most commonly used buprenorphine formulations (buprenorphine hydrochloride and 2 different extended-release formulations) failed to attenuate mechanical and thermal hypersensitivity.⁴ This study tested whether 3 different carprofen dosages would attenuate mechanical and thermal hypersensitivity. Carprofen is a nonsteroidal antiinflammatory analgesic commonly used to reduce inflammation and attenuate mild postoperative pain.^16,21,31,64 The currently recommended carprofen dosing regimen for mice and rats is 5 to 10 mg/kg given every 12 to 24 h.^{25,30,35,39,64} Although the efficacy of carprofen is affected by the route of administration, dose, strain, and sex of animals;^8,39,46 the dosing regimen has not been extensively evaluated in mice in general or in immunodeficient mice such as the NSG mouse.

Previous research indicated the need to evaluate the carprofen dosing regimens more extensively in mice. For example, in one study, the standard carprofen dosing regimen (5 to 10 mg/kg) did not attenuate pain after laparotomy in CD1 mice; however, at higher doses (20 and 25 mg/kg), it reduced the mean mouse grimace score.³⁹ In another study, carprofen at 5 mg/kg did not alleviate pain for female FVB mice after surgical removal of the mammary fat pad.² In a third study, carprofen administered at 30 mg/kg via drinking water to C57BL/6 and CD1 mice for 72 h did not attenuate pain after laparotomy.⁴⁶ In contrast, a study conducted on C57BL/6 and CD1 mice reported that carprofen administered subcutaneously or orally at 10 and 25 mg/kg, respectively, reduced mouse grimace scale scores after craniotomy.⁸ These studies indicate selecting a dosing regimen for carprofen requires consideration of multiple factors including strain, surgical procedure, dose, strain, and route of administration.

Our current study evaluated the efficacy of carprofen at the standard dose (5 mg/kg) and at doses that were 5 or 10 times higher (25 and 50 mg/kg). Our results indicated that Carp5 did not effectively attenuate mechanical and thermal hypersensitivity in the NSG mice with incisional pain. Carp25 attenuated both mechanical (on D0, D1, and D2) and thermal (on D1 and D2) hypersensitivity. Carp50 was no better than Carp5 or Carp25. We attribute this lack of attenuation in Carp50 mice to be related to possible toxicity at this higher dose.^12,34,38,65

The clinically effective plasma carprofen concentration for mice is unknown, and to our knowledge, this is the first study to evaluate the effect of different carprofen dosing regimens on the carprofen plasma concentration in NSG mice. The putative therapeutic plasma level of carprofen in other species is reported to be 20 to 24 μg/mL.³³ Moreover, the oral LD50 is reported as 282 and 149 mg/kg in mice and rats, respectively.¹⁰ In our study, we measured plasma carprofen concentration after 3 different dosing regimens (Carp5, Carp25, and Carp50) at time points (2, 4, 23, 24, and 48 h) that approximate the time points evaluated during hypersensitivity testing. After administration of the first carprofen dose, the plasma carprofen concentration increased in a dose-dependent manner for all 3 doses between 0 and 4 h; concentrations then fell between the time points of 4 and 23 h. The highest carprofen plasma concentration for all treatment groups was detected 1 h after administration of the second dose (at 24 h). These data suggest that carprofen plasma levels rise immediately after administration and fall below therapeutic levels at 12 h. These findings are consistent with previous pharmacokinetic studies of carprofen in mice.^25,30 In a study with C57BL/6 mice, peak plasma carprofen concentration (20 μg/mL) occurred 2 h after administration of 10 mg/kg by oral gavage.²⁵ Similarly, another study with female CD1 mice found a peak plasma concentration at 2 h after subcutaneous administration (5 mg/kg).³⁰ In our current study, the concentration in the Carp25 group was significantly higher than that of the Carp5 group at the 2- and 4-h time points, suggesting that the Carp25 plasma carprofen concentration at 4 h was sufficient to attenuate mechanical hypersensitivity, although it was not effective in attenuating thermal hypersensitivity for this model. This difference could indicate that a higher dose is necessary to produce analgesia that will attenuate thermal hypersensitivity. The dose of analgesics needed to attenuate thermal hypersensitivity is known to be different from the dose required for attenuation of mechanical hypersensitivity.^3,18,44 Moreover, in a study with male Sprague–Dawley rats, carprofen gel (5 mg/kg) did not attenuate thermal hypersensitivity but did attenuate mechanical hypersensitivity over 4 d.⁵⁴ In the present study, the Carp25 plasma concentrations in the Carp25 and Carp5 groups were not significantly different at 12 h. These results are consistent with previous studies that recommend administering carprofen to CD1 mice every 8 to 12 h³⁰ and found that carprofen (10 mg/kg PO) was not detectable in plasma of male C57BL/6 mice at 24 h after administration.²⁵ These data support the need to redose carprofen at or before 12 h after the initial administration to maintain analgesic efficacy in NSG mice.

All mice in this study were closely evaluated for any abnormal behaviors or clinical signs. No abnormal behaviors other than lameness and/or licking of the ipsilateral hind paw were observed in saline, Carp5, and Carp25 groups. Carp50-treated mice exhibited hyperactivity and subjective increase in locomotion in addition to lameness and/or licking of the ipsilateral hind paws throughout the study. These mice were very reactive to any movement by the experimenter’s hand during testing. This increased activity and restlessness could be due to potential toxicity from the high dose. Carprofen toxicity has been associated with neurologic signs in dogs at doses greater than or equal to 281 mg/kg.⁴¹

Carprofen is a COX-2 selective inhibitor that has antipyretic, analgesic, and anti-inflammatory effects.^12,45 Concurrent inhibition of COX-1 and COX-2 can have adverse effects including gastric ulceration, renal toxicity, and platelet function inhibition.^{12,34,38,62,65} Short-term high dose or chronic exposure to NSAIDs is associated with gastrointestinal ulceration and the frank blood in fecal samples.^{5,29,36,37,47,59,65} In this study, we used a fecal occult blood test^23,28,29,43 to evaluate fecal samples for blood. A positive fecal occult blood test occurs when hemoglobin and its iron-containing degradation products oxidize α-guaiaconic acid to produce a blue color (positive result).⁵¹ One of the 8 of mice treated with Carp25 had a positive fecal occult blood test on D2 and D3. This mouse was clinically normal (no behavior changes, postural changes, alterations in body condition or weight) and had no abnormalities on histopathologic examination. None of the other fecal samples tested positive for occult blood at any time point. Positive fecal occult blood has been reported previously in carprofen-treated mice and may indicate either gastritis or a false positive result.²⁹ Histopathologic examination was performed to check for evidence of carprofen toxicity in organ systems including the gastrointestinal, kidney, liver, heart, and spleen. No apparent histologic changes were observed in the Carp5 and Carp25 groups. In the Carp50 group, the histopathologic evaluation revealed gastric ulceration in 2 of 4 mice, ulcerative enteritis in one of 4 mice, and renal lesions in 3 of 4 mice. Thus, gastrointestinal and renal toxicity is evident in mice in the Carp50 group.

One of the limitations of our current study is that we did not test opioid medications or other NSAIDs in NSG mice. The second limitation is that we used only behavioral readouts for the von Frey and Hargreaves nociceptive tests. We wanted to retain the same experimental setup that we used previously to evaluate the efficacy of commonly used opioid medications in NSG mice.⁴ The results of the current and previous studies can thus provide a better comparison of the efficacy of carprofen and opioid medications in the same strain of mice. A third limitation is that we administered the carprofen shortly before the surgical procedure; the onset of analgesic effect of carprofen for rodent surgical procedures is an area of interest for future studies. The fourth limitation of this study is the lack of a power analysis to determine if adequate numbers of mice were used in the plasma concentration studies. Another limitation of this study is that we only tested 3 doses, 1 dosing interval, and 1 route of administration.

Our results indicate the need to closely evaluate the NSAID dosing regimens with regard to the species, strain, and sex of the animal, as all can affect the antinociceptive efficacy of carprofen. Our findings indicate a carprofen dose of 5 mg/kg is insufficient to attenuate mechanical or thermal hypersensitivity after paw incision in NSG mice. In NSG mice with paw incisions, a dose of 25 mg/kg SC provided greater attenuation of postoperative mechanical and thermal hypersensitivity than did the commonly recommended dose (5 mg/kg). The highest carprofen dose studied (50 mg/kg) showed evidence of gastric and renal toxicity, and this dose is therefore not recommended for NSG mice. Based on these results, we recommend using carprofen at a dose of 25 mg/kg SC SID for incisional pain in NSG mice. The analgesic efficacy of NSAIDs in mice should be further evaluated with other pain models, strains, dose regimens, and NSAID classes.

Conflict of Interest

The authors have no competing interest to declare.

Funding

This study was supported by the ACLAM Foundation.