Subcutaneous implants for long-acting drug therapy in laboratory animals may generate unintended drug reservoirs.

BACKGROUND: Long-acting therapy in laboratory animals offers advantages over the current practice of 2-3 daily drug injections. Yet little is known about the disintegration of biodegradable drug implants in rodents.
OBJECTIVE: Compare bioavailability of buprenorphine with the biodegradation of lipid-encapsulated subcutaneous drug pellets.
METHODS: Pharmacokinetic and histopathology studies were conducted in BALB/c female mice implanted with cholesterol-buprenorphine drug pellets.
RESULTS: Drug levels are below the level of detection (0.5 ng/mL plasma) within 4-5 days of implant. However, necroscopy revealed that interstitial tissues begin to seal implants within a week. Visual inspection of the implant site revealed no evidence of inflammation or edema associated with the cholesterol-drug residue. Chemical analyses demonstrated that the residues contained 10-13% of the initial opiate dose for at least two weeks post implant. DISCUSSION: The results demonstrate that biodegradable scaffolds can become sequestered in the subcutaneous space.
CONCLUSION: Drug implants can retain significant and unintended reservoirs of drugs.

Sustained release formulations offer several advantages over immediate release dosing practices in humans. When drugs are supplied in acute oral form or by injection, blood levels can rise above and fall below optimal therapeutic values with each dose. There are management challenges associated with 4-8 hour dose requirements. Compliance studies show that patients frequently under-medicate themselves (refs). They find it inconvenient, or forget the daily requirements. These management challenges are exacerbated in veterinary pharmacology at two levels. Pet owners universally appreciate the difficulties of giving oral medications to cats and dogs. Second, laboratory animals are difficult to restrain for intraperitoneal (IP) or subcutaneous (SC) injections. Thus, there are numerous reasons to investigate whether sustained release drugs can be optimized for clinical efficacy in animals.[12] Long-acting drugs reduce the number of times an animal needs to be restrained for medication. Decreased handling reduces stress, reduces opportunities for iatrogenic injuries and increases the harmonization of animal models with human therapy.[34]

We elected to investigate the properties of cholesterol-based sustained delivery vehicles in mice. New biodegradable delivery platforms frequently focus on lipid platforms.[5] Kent described an implantable cholesterol matrix that trapped large molecules in a porous structure.[6] Insulin and growth hormone, for example would be flushed from the pores by interstitial fluids.[6] Pontani and Misra described a cholesterol pellet for the long-term delivery of drugs to treat chronic pain and opiate addiction.[7] The latter seemed to provide a useful model system to examine the delivery of antibiotics, anti-inflammatory drugs and analgesics in surgically-treated small animals. We chose buprenorphine as a model drug. It has a high therapeutic index.[8] The drug is generally recommended for analgesia in animals.[910] We are unaware of toxicity concerns associated with SC implants of cholesterol.

In this report, we tested the hypothesis that blood levels of the drug would accurately predict disintegration rates of the drug-implant vehicle. We could not confirm this hypothesis.

Materials and Methods

Animals

Studies were approved by the JHU Institutional Animal Care and Use Committee. BALB/cNCrl female mice, 18-20 g, were obtained from Charles River Laboratories (Wilmington MA). Mice were maintained in Association for Assessment and Accreditation of Laboratory Animal Care accredited JHU facilities. They were housed at a density of 4-5 mice/cage in Smart Bio-Pak cages in ventilated cage racks (Allentown NJ), and received ad libitum Teklad Global Rodent Diet 2018 (Harlan, Madison WI) and municipal water delivered by an automated watering system.

Materials

United States Pharmacopeia (USP) grade buprenorphine HCl was a gift of Reckitt-Benckiser Pharmaceuticals (Hull, UK). USP-grade cholesterol and glycerol tristearate were purchased from Sigma-Aldrich (St Louis MO). Lactose and Plasdone K-29-32 were obtained from ISP Pharma (Calvert City KY). Starch was from Cargill (Minneapolis MN).

Methods

Buprenorphine-cholesterol pellets were prepared essentially as described by Pontani and Misra.[7] Drug pellets were prepared under sterile conditions in the Johns Hopkins Cell Processing Facility (Baltimore MD) with a Natoli NP RD10 tablet press (St. Charles MO). Five mg pellets were prepared using 1.25 tons of pressure. This pressure was selected because studies described in Figure 1 demonstrated that higher tableting pressures did not increase the crush resistance of the tablets. Pellets were stored at 3-8° before use. Pellet hardness was measured with a CCS Stokes Model 539 manual hardness tester (Warrington PA). Dissolution rates in saline were estimated by placing pellets on the surface of a 5 mL glass tube containing 2 mL of saline at 37° and recording the time to disintegration of the pellet.

Figure 1

Tableting pressure and compressive strength of cholesterol-buprenorphine pellets. Error bars represent standard deviation in crush pressure for 5 tablets crushed at each pressure

Buprenorphine assays

Serial blood samples were obtained by facial bleeding of the superficial temporal vein.[11] Samples were taken at noon, 23-25 hour intervals after the drug was implanted. Plasma samples were used for buprenorphine measurements by enzyme-linked immunosorbent assay (ELISA). Samples of 5-20 μL of plasma were analyzed in triplicate using a Buprenorphine One-step ELISA kit (International Diagnostic Systems, St Joseph MI). The Manufacture validated the kit for clinical drug studies with a high-performance liquid chromatography-electro spray mass spectrometry (HPLC-ES-MS) procedure. All known cross-reactivity's are reported by the manufacturer at <0.06%, with the exception of norbuprenorphine, which cross-reacts at 1.1%.[12] Standards curves were prepared with five buprenorphine solutions: 0, 0.01, 0.05, 0.1, 0.5, and 1.0 ng/20 μL. Absorbance was recorded at 450 nm (reference wavelength: 650 nm) using a Perkin Elmer Victor 3 model 1420 micro plate reader with Wallac 1420 data manager software.

Explant analysis for residual drug was performed by dissolving the weighed explant in 10 mL of methanol in a glass vial. After sonication for 30 min at room temperature, a 100 μL of the top solution was taken and mixed with 900 μL of 50% acetonitrile by votex-mixing for 10 min. A volume of 100 μL aliquot of the mixed solution was transferred into a 250 μL of polypropylene autosampler vial and sealed with rubber/Teflon crimp cap for HPLC analysis.

Each analytical run consisted of a standard calibration curve with 8 calibrators, zero blank and 3 levels of quality control (QC) samples (run in duplicate). The calibrators and QCs were made fresh for all analytical runs. The accuracy and precision of the assay were assessed by calculating the percentage deviation from the nominal concentrations. Calibrator values that were >15% from the nominal concentration (except the lower limit of quantitation >20%) were not acceptable. The sample extract was subjected to reverse phase HPLC. Separation of the analytes from potentially interfering material was achieved at room temperature using an Agilent Zorbax XDB-C18 (50 × 4.6 mm id) column packed with 5 μm stationary phase. The mobile phase used for the chromatographic separation was composed of acetonitrile/10 mM ammonium acetate (80:20, v/v), and was delivered isocratically at a flow rate of 1 mL/min for 5 min. Under these conditions, the retention time for buprenorphine was 3.3 ± 0.3, min. The column effluent was monitored using a Waters 2487 ultra violet/Vis Dual Wavelength detector, at wavelength of 214 nm.

Surgical procedures

Mice were given IP anesthesia with a solution containing 25 mg/mL ketamine plus 2.5 mg/mL xylazine and 14.25% ethanol in saline. The dose of anesthesia was 0.15 mL/20 g mouse.

Following anesthesia, an approximately 1 cm square of mid dorsal skin was shaved, washed with ethanol, and then coated with betadine. Mice were transferred to a procedural table that was cleaned with 70% ethanol solution and covered with a clean disposable towel. A sterile disposable no. 10 blade was used to make a 4-5 mm incision through the skin only. Bleeding, if any, was controlled with sterile gauze and light pressure. Sterile forceps were used to separate the skin and to create approximately a 2 cm × 4 cm SC pocket. Drug tablets were placed on top of the SC musculature. The skin was then apposed and stapled with 9 mm autoclips (KentScientific Torrington CN).

After the procedure, mice were moved to a holding cage. This cage contained a 37° heating pad covered with a clean disposable towel. The mouse was placed in a clean cage after it regained consciousness, as demonstrated by movement and the absence of signs of distress, which included but were not limited to sluggish movement, abnormal paw movements, efforts to scratch the incision site and cowering in a corner of the holding cage. Mice were housed 4/cage.

Necroscopy to recover residual tablets were performed on mice sedated with carbon dioxide. The heart was exposed. Mice were exsanguinated through cardiac puncture. Explanted tablets were dried overnight at 37°, weighed on an electronic Ohaus Voyager Pro micro balance (Union NJ) and stored in 5 mL of methanol prior to HPCL assay.

Statistics

Analyses of treatment group buprenorphine levels were made using GraphPad Prism Software Version 5.04 (LaJolla CA). Microsoft Excel 2007 was used to generate average and standard deviation (StDev) data.

Results

To insure that the cholesterol-buprenorphine pellets would have the maximum opportunity to break down and dissolve in the SC space, powders were compressed at low pressures. Pellets prepared with 0.5 tons of pressure frequently crumbled when removed from the tableting press. Those prepared with 0.8 tons were easier to harvest but frequently fell apart as the investigator attempted to implant them in the SC space using forceps. Based on these studies, pellets were prepared with 1.25 tons of pressure. However, data in Figure 1 shows that the adhesiveness of the pellets was largely independent of the pressure used to form them. The drug powder itself had little compressive capacity. Absent the addition of excipients commonly used to tablet drugs, pellets prepared from the drug-lipid powder were soft. This sentence does not read well, particularly the first phrase. They were crushed at a pressure of approximately 2 kg. For reference, the pressure to crush uncoated generic aspirin samples ranged from 11-12 kg. Did you do this test or is this a value obtained from the literature? If so, please cite a reference.

The data in Figure 2 illustrate buprenorphine blood level in female mice implanted with 5 mg cholesterol buprenorphine pellets. Buprenorphine levels generally fell from 10 ng/mL on day-1 to below 0.5 ng/mL by day-5. This data suggests that the implants had significantly dissolved and released their drug content by day-5.

Figure 2

Buprenorphine blood levels in female mice at days 1-9 post cholesterol-buprenorphine implant, 4 mg/kg dose. Error bars represent standard deviation for 6, 8, 8, and 8 mice at days 1,3, 5, and 9 respecively

In a subsequent 3-day experiment intended as a pilot histopathology study of implant sites, three mice each were harvested at day-1 and day-3 following the implantation of 5 mg pellets containing 0.080 mg of buprenorphine. Although the investigation confirmed there was no visible edema or inflammation, in each case, a residual implant was visible. The residues were adherent to either the inside of the dermis or the fascia overlaying the muscle. The remaining implants were dissected from the adherent tissue to the extent possible without crushing the soft pellets. The explants were dried overnight at 37°, weighed and analyzed for buprenorphine by HPLC. The data in Table 1 shows the weight of the pellets harvested at day-1 and day-3.

Table 1

Residual weights and drug content of explanted cholesterol-buprenorphine pellets, 4 mg/kg dose

Excipients are frequently added to drug powders destined for tablet or capsule formation to increase flow-ability of the powder in tablet presses and to increase the breakdown of the tablet in body fluids.[1314] To determine whether excipients commonly used in tablet manufacture and considered as having “disintegration” properties could accelerate the dissolution of the cholesterol pellets or increase the rate of drug release, a series of excipient studies was initiated. Cholesterol drug powder were prepared with up to 2-fold concentration of buprenorphine, dry blended with 20-50% excipients by weight and pressed at 1.25 ton pressure into 5 mg chronic release pellets. The final buprenorphine content (0.08 mg) was 1.6% of the pellet weight.

The data in Table 2 demonstrates the modest effect of excipients on disintegration of cholesterol-triglyceride pellets containing 1.6% buprenorphine. Pellets containing the excipients were tested in vitro for their stability in saline. In each case, pellets with excipients dissolved within 8 hours at room temperature. The excipient-free pellet remained macroscopically intact for at least 24 hours. Mice implanted with 5 mg pellets were sacrificed at intervals after the pellet was placed in a SC site. As shown in Table 2, the excipients had little effect on increasing the pellet's crush resistance and less effect on the rate of disintegration compared to pellets without excipients.

Table 2

Limited effect of excipients on pellet biodegradation

Although the addition of excipients did not significantly accelerate pellet degradation, we examined the possibility that excipients could increase drug release to interstitial fluid by diluting the hydrophobicity of the cholesterol pellets. Implants were recovered at day-7 and 14 from sets of two mice with pellets prepared with lactose and starch excipients. At days-7 and 14, the buprenorphine content of the mice implanted with the excipient-free pellets was 11.2 and 13.5% of the 0.08 mg drug load in the pellet. The buprenorphine content at days-7 and 14 for the lactose-modified pellets was 6.4 and 3.2% respectively. Similar results were obtained from the starch-modified pellet. This data suggests that select excipients can promote drug-release from cholesterol implants, but that significant quantities of buprenorphine were present in each case at day-14.

Discussion

In this report, we examined whether blood levels of the drug would accurately predict disintegration rates of a SC drug-implant vehicle. We could not confirm this hypothesis in mice with cholesterol-buprenorphine implants. Pharmacokinetic studies in BALB/c female mice with cholesterol-buprenorphine implants demonstrated peak levels of buprenorphine approximately 8 hours after a drug-bound pellet was implanted into the dorsal SC space. Plasma levels became undetectable within 4-5 days post-implant. However, examination of the implant site frequently revealed a residual pellet for 1-2 weeks after the surgery. Chemical analyses demonstrated that the explants contained 10-13% of the initial dose at day-7. The addition of excipients commonly used to speed the dissolution of oral drug tablets had little effect on in vivo SC disintegration rates. Although buprenorphine loads in the residual drug pellets were small at day-14, the amount was significant. For example, the 3.2% of drug in the day-14 explant was equivalent to an acute 0.1 mg/kg dose of buprenorphine. This amount provides a clinically significant dose of analgesia in mice.[910]

The compelling needs for improved veterinary drug products support further research on chronic release drug systems. Lipid-based delivery vehicles appear to be ideal candidates because they are safe and biodegradable. Carbohydrate and polymer delivery systems also offer attractive candidates for further research. Polymer implants have been described for analgesia and addiction therapy.[1516] Yet, experience in our laboratory and others suggest drug release from polymers may be anomalous.[17] Moreover, polymer residues can induce an inflammatory response in tissues.[18] Regardless of the composition of the delivery system itself, one cannot assume that drug-bound delivery vehicles are safe. Long-term histopathology studies of the SC space are warranted for each drug-bound vehicle.