Evaluation of Extended- and Immediate-Release Buprenorphine Formulations on Pharmacokinetics and Injection Site Lesions in Different Strains and Sex of Mice

Ethical review.

AbbVie is committed to ensuring the humane care and use of laboratory animals in the company’s research and development programs. Our programs exceed regulatory agency standards, and we are committed to the internationally accepted principles of the 3Rs (refinement, reduction, replacement). All animal studies were reviewed and approved by AbbVie’s IACUC (in accordance with national regulations). Animal studies were conducted under an AAALAC-accredited program, where veterinary care and oversight were provided to ensure optimal animal care.

Animals, husbandry, welfare.

Mice were obtained from Charles River Laboratories (Wilmington, MA) and the C57BL/6J strain was from The Jackson Laboratory (Bar Harbor, ME), consisting of both sexes (8 female, 8 male) from 4 different strains (3 inbred: C57BL/6J, CB17/Icr-Prkdc^scid/IcrIcoCrl, and BALB/cAnNCrl and 1 outbred: Crl:CD1(ICR)). All mice were 7 to 8 weeks of age on arrival and free of pathogens of concern, including viruses, bacteria, and external and internal parasites. Mice were group housed (4 per cage per sex/strain) in polycarbonate microisolator caging ( 25 × 48.2 × 15.2 cm; 19 x 19 x 6 in.), on ventilated racks (Allentown, Allentown, NJ). Individually ventilated cages received Sani-Chips bedding (PJ Murphy Forest Products, Montville, NJ) and 8 g of paper nesting material for enrichment (Bed-r’Nest; The Andersons, Maumee, OH). Mice were maintained on a 12-h light/12-h dark cycle. Food (Teklad 2014 rodent diet; Inotiv, West Lafayette, IN) and water (reverse osmosis/hyperchlorinated) were provided ad libitum. All caging components, including bedding, water, and feed (2018SX, Teklad, Greenfield, IN), were autoclaved for CB17 SCID mice, and manipulations with CB17 SCID mice were performed in a biological safety cabinet.

Compounds, formulation, and administration.

Mice received either Bup-ER (1 mg/mL, ZooPharm, Fort Collins, CO) at a dose of 1 mg/kg or Bup-IR (0.3 mg/mL, Reckitt Benckiser Pharmaceuticals, Richmond, VA) at 0.1 mg/kg via subcutaneous injection into ‘tented’ skin over the neck and between the shoulder blades as per the manufacturers’ recommendations. Physiologic saline was also administered subcutaneously into tented skin on the caudal dorsum of each mouse as a vehicle control. Doses for both formulations were based on published doses and an internal formulary using a published reference.²⁷

Mouse preparation and inoculation.

Body weights were collected for all mice on days 1 and 4. Mice were randomly assigned to the treatment groups using a computer-based random number generator to assign each cage a number. Prior to inoculation with buprenorphine, mice were lightly anesthetized using isoflurane (Patterson Veterinary, Loveland, CO) at a rate of 2% to 3% per liter of O₂ until loss of righting reflex, then briefly maintained on isoflurane at a rate of 2% to 3% per liter of O₂. Upon loss of righting reflex, the caudal neck and rear dorsum areas were shaved for administration of buprenorphine or vehicle; lubricating ophthalmic ointment was applied to both eyes for protection. Both buprenorphine and saline were administered to the mice by the same person. A 22-gauge needle was used to administer the Bup-ER and a 27-gauge needle for the Bup-IR stock formulations; both formulations were warmed to room temperature per the manufacturers’ instructions. The skin was pinched off following slow withdrawal of the needle based on the manufacturers’ instructions. Misdosing was noted for 2 animals (no. 57 CD-1 female and no. 63 BALB/c male, Table 1) due to an error in reading the syringe, with both mice likely receiving more material per dose than intended. Both of these animals were excluded from PK determinations but were included in all other data collections. Recovery was quick and unremarkable following cessation of isoflurane anesthesia.

Skin scoring, histopathology, and PK processing.

A scoring system (Table 2) was devised to assess edema, erythema, skin lesions in the form of papules or scabbing, and location of lesion in proximity to the site of injection (center of shaved area) or spreading to edges and away from injection site. A scale of 0 to 3 was used to rate the severity of response to injection, with 0 being no visible lesion and 3 being a severe reaction as defined in Table 2. Two different technicians performed the scoring, and although blinded to treatment, they were aware that the vehicle was injected into the caudal dorsum area. Mice were euthanized on day 8 via CO₂ asphyxiation and cervical dislocation to confirm death, within 24 hours of the last observation. Skin was collected from the neck and caudal dorsum areas from animals that had a cumulative score >0 at any point during the study. Skin sections were placed in 10% buffered formalin for a minimum of 48 hours. Tissues were infiltrated to paraffin on an automated histology processor using a standard ethanol/xylene/paraffin program, embedded, and sectioned at nominal 6 µm thickness and stained with hematoxylin and eosin on an automated staining platform. Two slides from each mouse were labeled with the mouse number and site number, corresponding to the saline and buprenorphine injection sites. The pathologist evaluating the tissue sections was blinded to treatment information as well as the sex and strain of the animals for each slide, and this information was not revealed until after pathologist (W.R.B.) scoring. Histopathology scoring was performed using a subjective categorical scoring system for 5 characteristics. (1) Epidermis was scored for the presence of epidermal hyperplasia (score of 1) or crusting with or without hyperplasia (score of 2). (2) Follicular dropout (score of 1) or follicular anagen (score of 2) was noted when a localized region was so affected. (3) Dermal fibrosis was scored if present (score of 1). (4) Subcutaneous cavitation with a thin wall of reaction was scored as 1, whereas abundant cellular or amorphous debris with a heavy cellular infiltration of the wall was scored as 2. (5) Subcutaneous fibrosis was scored 1 if mild and 2 if moderate or more severe. Similarly, blood submitted for PK analysis was blinded for strain and formulation. In alignment with 3Rs principles, we used microsampling as a refinement, collecting all time points from the same mice. Blood (40 μL per time point) was collected at 1, 4, 24, 48, 72, and 96 hours after injection of the Bup-ER and 0.25, 0.5, 1, 2, 3, 6, 9, 12, and 24 hours following Bup-IR from the lateral saphenous vein for each time point.²⁸ Once histopathologic and PK analyses were complete, the data were unblinded and results organized by strain and formulation type.

Liquid chromatography–tandem MS analysis of buprenorphine levels.

Chromatographic elution of buprenorphine was achieved using a linear gradient on an Xbridge C18 column (30 × 2.1 mm, 5 µm) (Waters Corporation, Milford, MA) at a flow rate of 1.5 mL/min, and with a total run time of 1 min. The mobile phase consisted of 0.1% formic acid in water (mobile phase A) and acetonitrile (ACN, mobile phase B). Samples were analyzed using a 6500+ triple quadrupole mass spectrometer (AB Sciex, Framingham, MA) coupled with an Agilent 1290 UPLC system (Agilent Technologies, Santa Clara, CA) and PAL autosampler (CTC Analytics, Zwingen, Switzerland). Buprenorphine and the internal standard d4-buprenorphine were detected using electrospray ionization with multiple reaction monitoring in positive mode. The multiple reaction monitoring transitions and collision energy were 468.3/396.3 at collision energy of 51 and 472.3/400.3 at collision energy of 55 for buprenorphine and d4-buprenorphine, respectively. The ion spray voltage and source temperature were kept at 5.5 kV and 500 °C.

Standard solutions and sample preparation.

For quantitative analysis, stock solutions of buprenorphine and d4-buprenorphine were prepared in DMSO at 1 mg/mL and stored at −20 °C. On the day of analysis, the calibration spiking solutions from 0.002 to 20 ng/mL were prepared using a Tecan D300e digital dispenser (Tecan Group, Männedorf, Switzerland) in ACN/H₂O 50:50 (v/v). Internal standard working solution was prepared in ACN/methanol 50:50 (v/v) at 5 ng/mL. K2-EDTA CD-1 mouse plasma was used as the blank matrix to construct the calibration curve. Sample preparation was achieved by liquid–liquid extraction on a Hamilton workstation (Hamilton Robotics, Reno, NV) to reach the sensitivity required for the limited sample volume. Briefly, to 10 µL of sample or blank plasma in a 96-well plate, 10 µL of ACN/H₂O 50:50 (v/v) or calibration spiking solution was added, followed by 25 µL of internal standard working solution and 50 µL of 100 mM ammonium hydroxide solution. After mixing the plate on a plate mixer (Eppendorf, Enfield, CV) at 1600 rpm for 1 min, 700 µL of t-butyl methyl ether was added to each well, and the entire plate was sealed, vortexed at 1100 rpm for 45 min, then centrifuged. Then, 450 µL of the top layer was transferred to a new plate and dried down under gentle nitrogen flow. The residue was reconstituted into 100 µL of can and then subjected to liquid chromatography–tandem MS analysis. The standard curves were fitted using linear regression with a weighting factor of 1/x² and met a minimum r value of 0.99. The lower limit of quantitation was 0.01 ng/mL. Sample concentrations were calculated using the equation derived from the regression analysis of the peak area ratio (analyte/internal standard) of the standards compared with concentrations.

PK analysis for buprenorphine levels.

PK analysis of plasma concentration data was performed by using validated WinNonlin, version 5.3 (Pharsight, Cary, NC). Plasma concentration–time data were obtained after analysis of the plasma samples collected from the mice and were analyzed by using a noncompartmental PK model. C_max and T_max were determined directly from the plasma concentration data for each animal. The plasma AUC from 0 to t hours after dosing (AUC_0-t, where t indicates time of the last measurable plasma concentration) was calculated using the linear trapezoidal rule. The residual area extrapolated to infinity, determined as the final measured plasma concentration (C_t) divided by the terminal plasma elimination rate constant (β), was added to the AUC_0-t to produce the total AUC (AUC_0-∞). The apparent total plasma clearance (CL_p) was calculated by dividing the administered dose by the AUC_0-∞. The initial volume of distribution (V_c) was calculated as the dose divided by the extrapolated concentration at time = 0 (C₀). The volume of distribution at steady state, V_ss, was estimated as a product of the total plasma clearance (CL_p) and the mean residence time (MRT); the terminal-phase volume of distribution, V_β, was derived from the total plasma clearance value (CL_p) divided by the terminal plasma elimination rate constant (β).

Mouse cytokine analyses.

For the separate cytokine analysis study, 28 male C57BL/6J mice, 7 to 8 weeks of age, were single housed on arrival in caging and under parameters as described above. Mice were randomly assigned to treatment group and had skin shaved and were anesthetized and inoculated with either Bup-ER, polymer, or saline; 12 mice received Bup-ER and 8 mice received polymer or saline. The polymer (poly(d,l-lactide–co-caprolactone): 40% d,l-lactide and 60% caprolactone; Sigma-Aldrich, St. Louis, MO) was dissolved in 1-methyl-2-pyrrolidone at 20% w/v as described.²⁹ Our rationale in choosing this polymer was based on information contained within the relevant patent.³⁰

Half of the mice (n = 14) had blood and skin near the injection site collected at 24 hours and the other 14 mice remained on study for 7 days. Six mice inoculated with Bup-ER had blood collections occurring at 1, 4, 24, 48, 72, 96, and 168 hours for PK analyses as described above (Table 3). Additional blood was collected at 24 and 168 hours as terminal collections during euthanasia for cytokine analyses. Skin at the injection site was scored as described with personnel scoring the skin blinded to treatment, and skin was collected at 24 and 168 hours postinjection for histopathologic analyses. Histopathology scoring was performed using a subjective categorical scoring system for 5 characteristics as described. The pathologist evaluating the tissue sections was blinded to treatment until after tissues were scored. Cytokines in mouse blood serum collected from mice at 24 and 168 hours were assayed in duplicate using the Milliplex mouse cytokine/chemokine panel 32-plex kit (Millipore, Billerica, MA) following the manufacturer’s directions on the Bio-Rad Bio-Plex system (Bio-Rad, Hercules, CA).

MS imaging methods.

Skin tissues were collected from 4 male C57BL/6J mice at 24 or 168 hours after dosing with Bup-ER and stored at −80 °C. Frozen skin samples were sectioned on a Thermo HM550 cryostat at −16 °C using CryoJane tape transfer (Electron Microscopy Sciences, Hatfield, PA) at 12 µm and transferred onto indium tin oxide–coated microscope slides (CG-81IN-S115; Delta Technologies, Loveland, CO). Slides were dried overnight, then stored under vacuum for <8 days before applying DHA solution (100 mg of dihydroxyacetophenone, 9 mL of ACN, 1 mL of water, 150 μL of trifluoroacetic acid) with an HTX TM sprayer M3+ using consistent parameters (4 passes, 0.1 mL/min flow rate, 1100 mm/min nozzle moving speed, 2.0 mm/track, 40 mm nozzle height, 70 °C nozzle temperature, 10 pounds per square inch nitrogen pressure, and criss-cross spraying mode). Images were collected with a Bruker timsTOF fleX mass spectrometer (Bruker Corporation, Billerica, MA) in positive ion mode with laser settings of 200 shots/pixel at 1000 Hz with 10-ms trigger delay and 20-µm spatial resolution using the smart beam profile, trapped ion mobility, and the MALDI-2 laser. Analysis of m/z 468.30 (protonated buprenorphine) intensity data was processed with a Bruker SCiLS Lab MVS 2024a Pro. After scanning, slides were stained with hematoxylin and eosin beginning with 95% ethanol for 2 min to remove the matrix coating. Stained slides were scanned on a MoticEasyScan Pro 1 scanner at 20×, and images were exported using QuPath.³¹ Eosin-Y solution was obtained from Fisher Scientific (Hampton, NH), and all other chemicals were obtained from Sigma-Aldrich (St. Louis, MO).

Statistical methods and data analysis.

Group sizes were based on historical data from PK studies conducted inhouse where 3 to 6 mice per dose group provide acceptable SEM and SD. Statistical analyses on histopathologic scores, visual skin scores, body weight changes, and PK parameters were conducted to evaluate the potential difference caused by formulation and other factors. More specifically, for histopathologic scores, mixed effect logistic regression on total score was conducted, with formulation (Bup-ER compared with Bup-IR compared with saline), strain, and sex as fixed effects, and individual animal as a random effect. The histopathologic total score was transformed to 0 or 1 based on whether or not it is 0. For visual skin scores, a logistic regression model on averaged total score of 7 days was conducted, with formulation (ER compared with IR), strain, and sex as covariates and visual averaged total skin score transformed to 0 or 1 based on whether or not it is 0. For body weight changes and PK parameters, linear regression was conducted, with formulation, strain, and sex as covariates. After model construction, ANOVA was conducted to evaluate the impact of each factor. For histopathologic scores, the contrast between Bup-ER and Bup-IR was subsequently evaluated because the formulation factor itself also includes saline.

Statistical evaluation of cytokine levels was performed using a Wilcoxon rank sum test comparing time points and Bup-ER to polymer or saline vehicle. All cytokine data were log₂ transformed for better normality. P values of the Wilcoxon rank sum test for significant comparisons are displayed in the results.

All P values reported herein are nominal, without multiplicity adjustments.

Scoring of skin lesions and corresponding histopathology.

Personnel performing skin scoring and histopathology assessments were blinded with respect to treatement group. In addition, the pathologist evaluating the histopathology was blinded to control site, inoculation site, and strain/sex of the animals for the tissues being evaluated. Scoring of the skin was performed on all mice for days 1 to 7 postinjection (Tables 1, 2). All 32 animals that received Bup-ER had at least 1 day with a score greater than 0 for edema or erythema. Only 16 of the 32 animals that received Bup-IR had at least one day with a score greater than 0. For all mice, regardless of treatment, none had a score in any category >2, and all had scores of 0 for injection site location. Only 5 mice developed papules, and all had been injected with Bup-ER formulation: 1 male and 2 female C57BL/6 mice had a score of 1 for papules; 1 male and 1 female BALB/c mouse had a score of 2 for papules. We suspected barbering issues for the female BALB/c and C57BL/6 mice as the cause for papules. Shaving site irritation was also considered as a cause, but since so few mice had reactions, we interpret this to result from behavior of the affected strains.

In the group of male CD-1 mice injected with Bup-ER, 3 of 4 mice had scores (mice nos. 9, 10, and 11; Table 1) for edema and erythema near the injection site, whereas mouse no. 12 had almost no issues (Table 1). We identified fighting issues with this group of male mice and believe this to be the cause of increased edema and erythema; however, the fight wounds were mild, consisting of tooth marks on tails and an inflamed area at the back of the neck. Extra enrichment was added to this cage with no escalation of issues noted during the remainder of the study.

For the separate cytokine study, skin scores revealed more significant lesions for both the polymer- and Bup-ER–inoculated mice that were evaluated over the 7-day course (Tables 2, 3). Mice nos. 14 and 15 receiving the polymer vehicle had scores of 2 and 1, respectively, for edema, erythema, and site location, and mouse no. 16 had scores of 2 for edema and erythema (Table 3). Two mice receiving Bup-ER had edema, erythema, and site location scores (mice nos. 23 and 25; Table 3). In addition, mouse no. 26 had scores of 1 for edema, erythema, and site location along with mild body weight loss (Tables 2, 3).

In summary, most of the injection site skin lesions and scoring were seen in the Bup-ER–injected mice, with generally mild lesions limited to erythema and edema. Statistical modeling on the averaged total visual skin score showed that both the formulation (ER compared with IR) and the strain are significant factors (both P < 0.001), and ER formulation is significantly more likely to have a score >0.

Histopathologic cumulative skin score totals for each day were used for statistical comparison (Tables 4, 5). Mice without a skin lesion score were excluded from histopathologic analysis and were all from the Bup-IR–treated group. In addition, there were only 2 groups of Bup-IR–injected mice with skin lesion scores of ≥1, and these groups had injection site skin submitted for histopathology; the saline-injected skin was not included. Daily score tallies included any score for any issue during that day. Daily cumulative scores did not exceed a score of 4 in mice that received Bup-ER. In mice that received Bup-IR, only 16 had scorable lesions, with daily cumulative scores not exceeding a score of 2.

Comparison of histopathology between Bup-IR and Bup-ER across 4 strains of mice found that cavitation and mild to moderate inflammation of the subcutis was restricted to mice dosed with Bup-ER (Figure 1). The incidence of mild to moderate inflammation was 1 of 8 BALB/c, 2 of 8 CB17 SCID, 2 of 8 C57BL/6J, and 3 of 8 CD-1 mice administered the ER formulation. Inflammation was often associated with cavitation, which is the expected histologic appearance of the in situ polymer after tissue processing. In the separate cytokine study, MS imaging confirmed the presence of buprenorphine within these areas (Figure 2). The degree of inflammation is consistent with cell-mediated clearance of the polymer through phagocytosis and lysosomal digestion. The character and severity of inflammation was variable without a clear correlation to mouse strain. CB17 SCID mice are immune deficient but had similar inflammation, suggesting that innate immunity drove the response. When comparing the proprietary Bup-ER formulation with a common long-acting injectable polymer formulation or saline, the long-acting formulations were similar with respect to incidence of mild to moderate inflammation at 24 hours, but mice receiving the polymer alone had a lower incidence of inflammation at 168 hours. This indicates that the polymer formulation provokes a modest degree of inflammation, not seen with saline, and which may recover more rapidly in the absence of buprenorphine.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Statistical modeling on the total histopathologic score showed that formulation (Bup-ER compared with Bup-IR compared with saline) is a significant factor (P < 0.001), and the contrast between Bup-ER and Bup-IR showed that Bup-ER is significantly more likely to have a score >0 (P = 0.003). Given these findings, and the potential use of Bup-ER in mouse models of immunity, we sought to determine whether the innate response to the polymer would significantly impact cytokines commonly monitored as indicators of immune function.

Body weights.

While we were not expecting to see any body weight loss since mice had no surgical manipulations performed, we found body weight loss of ≥5% in some of the mice. This appeared to be sex- and formulation-dependent (Table 1). There was greater body weight loss found in the Bup-ER–dosed group (17/32 mice) than in the Bup-IR–dosed group (9/32 mice), with most of the body weight loss being <5% between days 1 and 4 postinoculation. What is interesting is the strain effect, with 6 C57BL/6 mice in the Bup-ER group (4 males and 2 females) having mild decreases in body weights (2% to 4% weight loss). BALB/c and CD1 mice had 4 mice each from the Bup-ER groups with body weight loss and 4 mice from the CB17 SCID mice with body weight loss (Table 1, individual body weights not shown). The group mean body weights shown in Table 1 are consistent with the above findings revealing body weight loss in both sexes of C57BL6/J mice, female CD-1 mice, male BALB/C mice, and both sexes of CB17 SCID mice (individual body weights not shown). Body weight loss for the Bup-ER groups was >5% in one female CD1 mouse with 9% and one male CB17 SCID with 6.9%. In contrast, for the Bup-IR–dosed groups there were few occurrences of body weight loss (9/32 mice), with no body weight loss found in either sex of C57BL/6 mice. None of the male CD1 mice in the Bup-IR group had body weight loss, with only 2 female CD1 mice losing 1.5% and 5.1%, respectively. Only 1 of the 4 male and female BALB/c mice in the Bup-IR group had body weight loss of 1% and 2.5%, respectively. No male CB17 SCID mice had body weight loss in the Bup-IR group, whereas all 4 female CB17 SCID mice had weight loss (3 female mice had weight loss between 6% and 7%, and 1 female had 3.9% body weight loss). These findings are consistent with the group mean body weights shown in Table 1 (individual body weights not shown). In addition, in a separate cytokine study, 2 of the male C57BL/6J mice (nos. 24 and 26) also displayed minor weight loss (<5%) a few days after dosing of the Bup-ER (Table 3, individual body weights not shown). In summary for both studies, none of the body weight loss was considered to be marked, as it was not ≥10%, but differences seen in sex and between the 2 formulations proved interesting. Statistical modeling on the change in body weights showed that formulation (ER compared with IR) is a significant factor (P = 0.014), with the ER formulation more likely to lose body weight compared with the IR formulation. Sex is also significant (P = 0.016), with females being more likely to lose body weight. However, the strain of the mouse was found to not significantly impact body weight loss.

PK evaluations.

It was striking how quickly the Bup-IR plasma levels dropped following the 6-hour time point, and levels were not detectable in most mice following 9 hours, with only CB17 SCID mice (Figure 3C) having detectable levels at 9 h. Indeed, Bup-IR was not detectable following 9 hours for all mouse strains except for the CD-1 strain (Figure 3B), where one male and one female had detectable levels at 9 hours that were <0.35 ng/mL. Male CB17 SCID mice, along with 1 female mouse of the same strain, had no detectable levels following 6 hours, whereas 3 of the 4 CB17 SCID females had levels at 6 hours of ≥0.8 ng/mL, and 1 of these 3 mice had a level of 0.24 ng/mL at 9 hours. For the BALB/c mice (Figure 3D) treated with Bup-IR, one male had no detectable levels following 3 hours and one female had no detectable levels following 6 hours. There were few differences for Bup-IR levels between sexes of the various strains of mice (Figure 3, Table 6).

View Figure

• Download as Powerpoint
• Download Figure

The Bup-ER levels from the study with 4 different strains and both sexes were still detectable at 96 hours, although low, <1.6 ng/mL, except in one male C57BL/6J mouse (Figure 3, Table 6). This male C57BL/6J mouse had the highest buprenorphine plasma level of all mice, peaking at 65 ng/mL at the 4-hour time point. All mice received the same dose of Bup-ER, and it is uncertain why this one mouse had higher levels than the other mice. Bup-ER levels rose between 1 and 4 hours after injection for most mice, with one male and one female C57BL/6 mouse that had levels below detection at 72 and 48 hours, respectively. One female and 2 male CD1 mice, and 1 male BALB/c mouse, had levels that declined between 1 and 4 hours postinjection, but their levels remained detectable out to 96 hours. In addition, one female and one male CB17 SCID mice had no detectable levels at 72 and 96 hours, respectively. As noted in Materials and Methods, misdosing was identified in 2 animals (Table 1), and both of these animals were excluded from PK determinations leaving n = 3 mice for PK evaluation for that sex and strain. Both mice were included in all other data collections and analyses.

Differences in PK levels between male and female C57BL/6 mice (Figure 3A) were due to one male C57BL/6J mouse with high buprenorphine plasma level of Bup-ER compared with other male mice and one female with low Bup-ER levels compared with other female mice of that strain (male mouse no. 11, female mouse no. 14; Table 6).

Bup-ER levels from the cytokine analyses study with PK levels evaluated in 6 male C57BL/6J mice (Figure 3E) revealed levels that were below the 1 ng/mL threshold at 72 hours and decreased into picogram levels at 168 hours (7 days) after dosing. Although these PK data are from the same sex of one strain, it indicates that Bup-ER levels continue to decline over several days and should be well below threshold levels at 7 days after dosing. Seven days is the typical recovery period following invasive surgical manipulations such as orthotopic surgical procedures.

Statistical analyses on PK parameters showed that formulation is a significant factor for both T_max and AUC (P < 0.001 for both). The IR formulation has significantly smaller T_max and AUC compared with the ER formulation.

Cytokine evaluations.

Cytokine analyses were performed in a separate study in male C57BL/6J mice as described in Materials and Methods (Table S1). Cytokine/chemokine levels were compared between Bup-ER at 24 and 168 hours, polymer vehicle at 24 and 168 hours, saline vehicle at 24 and 168 hours, Bup-ER compared with polymer vehicle at 24 or 168 hours, and Bup-ER compared with saline vehicle at 24 or 168 h.

Several cytokines, including GM-CSF, IFN-γ, IL-1β, IL-4, IL-5, IL-7, IL-10, LIF, MIP-1a, M-CSF, and VEGF, had levels below limits of detection for most samples (data not shown). Cytokine levels for LIX, KC, MCP-1, MIP-2, MIP-β, and MIG were similar across all time points and treatments with no statistically significant differences (data not shown).

Comparison of the compounds at 24 hours compared with 168 hours revealed differences for several cytokines:

1. For Bup-ER at 24 hours compared with 168 hours, significant elevations were seen in 4 of the cytokines tested (G-CSF, IL-17, RANTES, and eotaxin). The normalized mean levels were <10 pg/mL except for eotaxin, with mean levels of 10.13 pg/mL. G-CSF had higher levels at 24 hours compared with 168 hours (P = 0.025), whereas IL-17, RANTES, and eotaxin had higher levels at 168 hours compared with 24 hours (P = 0.0039, P = 0.0039, and P = 0.0163, respectively).

2. For polymer vehicle at 24 h compared with 168 hours, statistically significant elevations were seen in 3 of the cytokines tested (IL- 6, IL-9, and IL-12p40). All of these agents had mean levels <10 pg/mL. IL-6 had higher levels at 24 hours compared with 168 hours (P = 0.0202), whereas IL-9 and IL-12 had higher levels at 168 h compared with 24 hours (P = 0.0433 and P = 0.0384, respectively).

3. For saline vehicle at 24 h compared with 168 hours, statistically significant elevation was seen in IL-2 and IL-13 with normalized mean levels <10 pg/mL, and there were higher levels of IL-2 and IL-13 at 168 hours compared with 24 hours (P = 0.0472 and P = 0.433, respectively).

4. For Bup-ER compared with polymer vehicle at 24 or 168 hours, at 24 hours, statistically significant elevations were seen in IL-3, IL-9, and IL-17 (P = 0.0407, P = 0.033, and P = 0.033, respectively). Normalized mean levels were <10 pg/mL, and IL-9 had higher levels for Bup-ER compared with the polymer vehicle, whereas IL-3 and IL-17 had higher levels for the polymer compared with the Bup-ER at 24 hours. At 168 hours, statistically significant elevations were seen in IL-6 and RANTES with mean levels <10 pg/mL, and both IL-6 and RANTES had higher levels for Bup-ER compared with the polymer vehicle at 24 hours (P = 0.0103 and P = 0.033, respectively).

5. For Bup-ER compared with saline vehicle at 24 or 168 hours, at 24 h, comparison of Bup-ER with saline revealed statistically significant elevations in G-CSF, IL-6, IL-17, IP-10, RANTES, and TNF-α (P = 0.0105, P = 0.0176, P = 0.0330, P = 0.0330, P = 0.0105, and P = 0.0416, respectively) with mean levels <10 pg/mL. G-CSF, IL-6, IP-10, and TNF-α had higher levels for Bup-ER compared with saline, whereas IL-17 and RANTES had higher levels for saline compared with Bup-ER at 24 h. For Bup-ER compared with saline at 168 hours, there were no significant differences for any of the cytokines tested.