Drug Deliv. and Transl. Res. (2011) 1:238–246 DOI 10.1007/s13346-011-0032-4
RESEARCH ARTICLE
Pharmacokinetics of UC781-loaded intravaginal ring segments in rabbits: a comparison of polymer matrices Meredith R. Clark & Patrick F. Kiser & Andrew Loxley & Christopher McConville & R. Karl Malcolm & David R. Friend
Published online: 10 May 2011 # Controlled Release Society 2011
Abstract UC781 is a potent and poorly water-soluble nonnucleoside reverse transcriptase inhibitor being investigated as a potential microbicide for preventing sexual transmission of HIV-1. This study was designed to evaluate the in vivo release and pharmacokinetics of UC781 delivered from matrix-type intravaginal ring segments in rabbits. Three polymer matrices (polyurethane, ethylene M. R. Clark (*) : D. R. Friend CONRAD, Department of Obstetrics and Gynecology, Eastern Virginia Medical School, 1911 North Fort Myer Drive, Suite 900, Arlington, VA 22209, USA e-mail:
[email protected] P. F. Kiser Department of Bioengineering, University of Utah, 20 S 2030 E, Room 108, Salt Lake City, UT 84112, USA P. F. Kiser Department of Pharmaceutics and Pharmaceutical Chemistry, University of Utah, Salt Lake City, UT 84112, USA A. Loxley Particle Sciences Inc., 3894 Courtney St, #180, Bethlehem, PA 18017, USA C. McConville : R. K. Malcolm School of Pharmacy, Medical Biology Centre, Queen’s University of Belfast, 97 Lisburn Road, Belfast BT9 7BL Northern Ireland, UK Present Address: C. McConville School of Pharmacy, Faculty of Health Sciences, Curtin University, GPO Box U1987, Perth, WA 6845, Australia
vinyl acetate copolymer, and silicone elastomer) and two drug loadings (5 and 15 mg/segment) were evaluated in at least one of two independent studies for up to 28 days in vivo. Inter-study comparison of in vivo release, vaginal tissue, and plasma concentrations for similar formulations demonstrated good reproducibility of the animal model. Mean estimates for a 28-day in vivo release ranged from 0.35 to 3.17 mg UC781 per segment. Mean proximal vaginal tissue levels (adjacent to the IVR segment) were 8– 410 ng/g and did not change significantly with time for most formulations. Distal vaginal tissue levels of UC781 were 6- to 49-fold lower than proximal tissue levels. Mean UC781 plasma levels were low for all formulations, at 0.09–0.58 ng/mL. All formulations resulted in similar UC781 concentrations in vaginal tissue and plasma, except the low loading polyurethane group which provided significantly lower levels. Loading dependent release and pharmacokinetics were only clearly observed for the polyurethane matrix. Based on these results, intravaginal ring segments loaded with UC781 led to vaginal tissue concentrations ranging from below to approximately two orders of magnitude higher than UC781’s EC50 under in vitro conditions (2.8 ng/mL), with little influence by polymer matrix or UC781 loading. Moreover, these findings support the use of rabbit vaginal pharmacokinetic studies in preclinical testing of microbicide intravaginal rings. Keywords UC781 . Microbicide . Intravaginal ring . Pharmacokinetics
Introduction Vaginal microbicides are a promising female-controlled strategy for preventing sexual transmission of HIV. In 2010,
Drug Deliv. and Transl. Res. (2011) 1:238–246
the proof of concept that such prophylactic methods can successfully reduce the transmission of HIV by at least 39% was demonstrated using an antiretroviral (ARV) vaginal gel containing tenofovir administered before and after coitus [1]. Only months later, these results were bolstered by similar demonstrative proof of effective pre-exposure prophylaxis in men who have sex with men using daily oral tablets containing tenofovir and emtricitibine [2]. As these results have further fueled the need to develop an effective microbicide, there remains much room for improvement in the level of protection that such products can provide. Research and development of new products continues with special emphasis on more potent ARVs, ARV combinations, and alternative dosage forms designed to optimize drug delivery to the sites of action and enhance user adherence and acceptability. UC781, a thiocarboxanilide-based nonnucleoside reverse transcriptase inhibitor (NNRTI), is one such ARV that has undergone extensive preclinical and early clinical evaluation as a potential microbicide candidate. It is a potent inhibitor of HIV-1 replication in cell culture systems [3, 4] and inhibits HIV-1 strains resistant to nucleoside ARVs with a potency similar to that used for inhibition of wildtype virus [5]. Resistance to UC781 is observed when more than one mutation occurs in the NNRTI-binding pocket [6]. The effects of UC781 on mammalian cells in vitro and in cervical explant cultures demonstrate that UC781 prevents HIV-1 infection in immune cells [7] and inhibits direct infection of mucosal tissue [8, 9]. The EC50 of UC781 is 8± 3 nM (2.8±1.1 ng/mL) in MT2 cells [10]. UC781 has been evaluated as a vaginal gel product in a phase I safety study [11]. With limited water solubility (0.17 μg/mL) and a ClogP of about 5.2 [12], UC781 presents a challenge in designing a topical formulation with adequate local tissue bioavailability. Intravaginal rings (IVRs) are sustained/controlled delivery devices that are receiving attention in the microbicide product development field. Their potential to increase microbicide safety and efficacy by providing sustained local delivery of potent ARVs at therapeutic doses is expected to increase user compliance and coital independence compared with more conventional gel-based products. They are already widely accepted for use in other clinical indications, including contraception [13, 14] and estrogen-replacement therapy [15]. For vaginal microbicide delivery, silicone elastomer IVRs releasing the ARV dapivirine are already undergoing clinical investigation as a 1-month dosage form [16]. A phase III efficacy trial of a 25-mg dapivirine silicone elastomer matrix-type IVR is scheduled to begin soon. Other polymeric materials for fabrication of IVRs have been investigated in a bid to extend their utility to a more diverse range of microbicide candidates. Specifically, thermoplastic elastomers, including ethylene vinyl acetate copolymer (EVAc, similar to the commercially available
239
contraceptive Nuvaring®) and aliphatic polyether urethane (PU), are receiving much attention in microbicide product development [17–20]. From a physicochemical perspective, UC781 is considered a suitable candidate for permeation-controlled delivery from IVRs owing to its small size and hydrophobic character [21]. However, preliminary pharmacokinetic studies following administration of a UC781-loaded silicone elastomer, matrix-type ring in pig-tailed macaques (N=2, data not shown) showed poor in vitro/in vivo release correlation. Therefore, formulation development of a UC781-loaded IVR was expanded to include medical grade thermoplastic elastomers. During development of IVRs, it is usual to assess the release rate under in vitro conditions. However, poorly water-soluble compounds like UC781 demonstrate negligible release from IVRs into aqueous dissolution media. The addition of organic solvents, such as isopropyl alcohol, typically increases the in vitro release rate of poorly watersoluble drugs and maintains sink conditions. This nonphysiological situation may be unsuitable in predicting release under in vivo conditions. Therefore, an animal model that is relatively simple to use would enhance the ability to assess the performance of IVRs designed to release poorly water-soluble drugs such as UC781. A rabbit model has previously been reported to assess the in vivo release of steroid hormones from linear silicone devices [22]. In this work, the in vivo release and pharmacokinetic (PK) profiles of UC781-loaded IVR segments were evaluated in two rabbit studies. In the first study, two polymer matrices, PU and EVAc, were compared for 7 and 14 days in vivo at a single drug loading. In a follow-on expanded study, PU and EVAc were also compared to silicone elastomer, each at two drug loadings, for 14 and 28 days in vivo. The expanded study design was intended to provide additional insight into the polymer and drug loading dependence of UC781 in vivo release and PK over a clinically relevant duration of product use (i.e., 28 days) and to confirm observations made in the 7-/14-day study. IVR segments were designed to contain comparable loadings per segment as well as per unit surface area.
Materials and methods Materials Non-micronized and micronized UC781 was provided by CONRAD (Arlington, VA). Tecoflex® EG-85A (for study 1) and ATPU-1 (for study 2) polyether urethanes were provided by Lubrizol (Wickliffe, OH) and DSM PTG (Berkeley, CA), respectively. ATEVA® 2803G ethylene
240
vinyl acetate copolymer was provided by Celanese (Edmonton, AB). MED-6382 medical grade silicone elastomer and tetrapropyl orthosilicate (TPOS) were supplied by Nusil Technology (Carpinteria, USA). Matrix-type IVR segments (1.5 cm long, 4 mm in diameter) were prepared with PU (University of Utah), EVAc (Particle Sciences, Inc.), or silicone elastomer (Queen’s University of Belfast) at UC781 loadings of 15 (studies 1 and 2) and 5 (study 2 only) mg/segment. Methods Fabrication and characterization of UC781-loaded PU IVR segments PU polymer beads were first cryoground using a Fritsch Pulverisette 14 (Goshen, NY) then mixed with nonmicronized UC781 powder in a glass jar on roller for at least 1 h. The drug–polymer compound was extruded at 145°C using a Minilab twin screw extruder (Thermo Haake, Waltham, MA). Note this extrusion temperature exceeded UC781’s melting temperature of 131°C [23] and therefore facilitated the blending of non-micronized UC781 and molten PU to create homogeneous formulations. Resulting rods were then pelletized using a Randcastle PentaDrive pelletizer (Cedar Grove, New Jersey) and reextruded into 4-mm diameter rods. The rods were cut to 1.5 cm, and the edges were smoothed with a rotary sanding tool. The ring segments were handled under clean conditions or were sanitized prior to animal studies by a brief isopropanol (IPA) wipe. Representative samples of ring segments were assayed for drug loading by extraction and reverse-phase HPLC. Extraction of UC781 from PU IVR segments was performed by dissolving a 50-mg portion of an IVR segment in dichloromethane (final volume 5 mL) overnight, then diluting 1 mL of this solution in 9 mL of acetonitrile to precipitate the polymer. The samples were filtered (0.2 μm PTFE) prior to HPLC analysis using a gradient method (acetonitrile/water mobile phase mixture) on an Agilent 1200 equipped with diode array detector and a Zorbax ODS 4.6×250 mm, 5 μm C18 column. Norethindrone was used as an internal standard during the extraction process to account for dilution effects. Fabrication and characterization of UC781-loaded EVAc IVR segments Micronized UC781 was compounded into molten EVAc (Ateva® 2803G, Celanese) for 15 min at 120–140°C in a Rheocord 9000 batch compounder (Haake) with a Banbury mixer attachment. Note that micronized UC781 was used to enhance drug–polymer mixing at extrusion temperatures around the melting temperature of UC781. Cooled drug–
Drug Deliv. and Transl. Res. (2011) 1:238–246
polymer compound was then cryoground in a lab blender with a 1-L stainless steel canister (Waring 700s single speed, VWR) to a size of <4 mm. Next, full-sized IVRs with a 4-mm cross-sectional diameter and a 54-mm-outer diameter were injection molded at 93°C using a lab-scale piston-type unit (model AB-200, AB Instruments) fitted with an aluminum mold. Finally, IVRs were cut into 1.5-cm-length segments, and segment ends were smoothed with a rotary sanding tool, similar to the method described above. The ring segments were sanitized prior to animal studies by a brief IPA wipe. The representative samples of ring segments were assayed for drug loading by extraction and reverse-phase HPLC. Extractions were performed by dissolving the sample in chloroform, precipitating the EVAc from the solution using acetonitrile, then filtering the solution. The chloroform was evaporated, and the samples were reconstituted with 50:50 v/v acetonitrile/water. HPLC was performed on an Agilent 1100 series HPLC with a photodiode array detector using a Halo C18 column (3.0× 150 mm, 2.7 μm, Advanced Materials Technology). UC781 was quantified against an external standard by UV-absorbance (300 nm). Analyses were conducted using a 20-min gradient method (90/10 water/ACN + 0.05% TFA to 25:75 water/ACN + 0.05% TFA). Fabrication and characterization of UC781-loaded silicone IVR segments MED-6382 silicone elastomer base and the cross-linking agent TPOS (ratio 40:1) were blended in a Speedmixer™ (Model DAC 150 FVZK, Hauschild) to produce the silicone elastomer mix. Following the dispersion of micronized UC781 into the silicone elastomer mix, stannous octoate (0.5% w/w) was added and speed mixed for 30 s before injecting the active mix into stainless steel molds of a laboratory-scale, electrically heated reaction injection molding machine (Technigal, Craigavon, UK). The injection mixes were cured at 50°C for 1 h. Rings were then cut into 1.5-cm-length segments and sterilized using dry heat at 100°C for 15 min. Representative samples of ring segments were assayed for drug loading by extraction and reverse-phase HPLC. Extraction was performed by refluxing the ring segment in 100 mL dichloromethane for 2 h, drying a 10-mL aliquot of the dichloromethane solution and reconstituting in 10 mL ethanol. HPLC analysis was performed on an Agilent 1200 series HPLC with an Agilent ZORBAX Eclipse XDB-C18 4.6×150 mm column with a 5-μm particle size. The mobile phase was comprised of 80% HPLC grade methanol and 20% HPLC grade water. The flow rate was 1 mL/min with a 30-μL injection volume and UC781 was detected using a wavelength of 300 nm.
Drug Deliv. and Transl. Res. (2011) 1:238–246
Rabbit PK studies Studies were performed using New Zealand white female rabbits (age 5–8 months). In study 1, 12 animals per group received UC781-loaded (15 mg/segment) IVR segments comprised of either PU or EVAc polymer matrix. In study 2, 12 animals per group received PU, EVAc, or silicone elastomer matrix IVR segments loaded with either 15 or 5 mg UC781 per segment. Animals were administered the following preoperative anesthesia and analgesic treatment: On the day prior to surgery, the animals were treated with a fentanyl (25 μg/h, transdermal) patch that was then left in place for three additional days; on the day of surgery, acepromazine maleate (0.1 mg/kg, subcutaneous), glycopyrrolate (0.01 mg/kg, subcutaneous), ketamine (25 mg/kg, subcutaneous), meloxicam (0.1 mg/kg, intramuscular), enrofloxacin (10 mg/kg, intravenous or intramuscular), and LRS (10–15 mL/kg/h, intravenous) were administered, and the vaginal opening was infused with 2% liquid lidocaine. Treatment with meloxicam (0.1 mg/kg, intramuscular) was also provided daily for three additional days post-surgery. For each animal, one ring segment, lightly lubricated with K-Y Jelly, was inserted ∼8 cm into the anterior vagina following a midline laparotomy to isolate the vagina from surrounding tissue. The IVR segment was anchored in place with a 5-0 Prolene suture through the outer ventral wall. Following surgery, the animals were closely monitored during anesthetic recovery for physiological disturbances including cardiovascular/respiratory depression, hypothermia, and excessive bleeding from the surgical site. Animals were then monitored daily for up to 14 (study 1) or 28 (study 2) days for incision site and clinical observations; body weight was measured and recorded weekly. For pharmacokinetic analysis, plasma and vaginal secretions via 3-mL vaginal lavage (study 1 only) were collected at several time points. Six animals per group were sacrificed at each of two time points per study: 7 and 14 days post-implantation for study 1, and 14 and 28 days post-implantation for study 2. A macroscopic pathology examination was performed at necropsy, and the ring segments were retrieved for residual drug content analysis. Vaginal tissues, split into proximal (i.e., abdominal) and distal (i.e., urinary) vaginal tract sections, were harvested for bioanalysis. Biosamples were extracted and analyzed for UC781 by LC-MS/MS using validated methods described previously [12]. The extraction efficiencies for plasma, vaginal tissue, and vaginal lavage were 38.6– 57.4%, 93.1–99.3%, and 82.3–90.5%, respectively. The lower limits of quantification (LLOQ) for these matrices were 0.05 ng/mL, 1.0 ng/g, and 100 ng/mL, respectively. Analysis of vaginal lavages from study 1 resulted in levels that were below the LLOQ for all samples. See Fig. 1 and
241
Study 1
Day 0
Plasma collection
X
X
7
14
Ring Segment Insertion
X
Necropsy & Ring Segment Removal
Study 2 X Day
0
7
14
X 21
28
Fig. 1 Rabbit PK study designs. Ring segments were surgically inserted in the vagina on day 0. Animals were sacrificed in cohorts of six at two time points per study (study 1: days 7 and 14; study 2: days 14 and 28)
Table 1 for individual study designs and corresponding formulations. Statistical analysis Statistical analyses were performed using two-way analysis of variance (ANOVA) followed by post hoc Tukey t tests. Analyses were performed using Origin 8 software (OriginLab Corporation). If analysis by ANOVA was not possible due to mismatches in group sample size, then paired (for proximal vs. distal tissue comparisons) or unpaired (all other cases) Student’s t tests were performed.
Results Characterization of UC781-loaded IVR segments UC781 was homogenously loaded into all matrix IVR segments and by visual inspection appeared to be dissolved Table 1 Target and actual UC781 loadings in intravaginal ring segments Target UC781 loading (mg/segment)
Study 1 PUa EVAcb Study 2a PU—high loading PU—low loading EVAc—high loading EVAc—low loading Silicone—high loading Silicone—low loading a
N=12
b
N=11
Actual UC781 loading mg/segment (mean±SD)
% w/w
15 15
15.4±0.9 14.9±0.8
7.13% 8.77%
15 5 15
17.8±1.4 5.6±0.6 15.6±0.8
7.53% 2.58% 9.17%
5 15 5
5.2±0.4 13.4±0.3 4.3±0.2
3.06% 5.57% 1.79%
242
in both PU and EVAc matrices (i.e., segments were transparent) and dispersed in silicone elastomer matrix (i.e., segments were opaque). All ring segment types had similar surface areas. For study 1, PU and EVAc ring segments contained similar drug loadings, with mean values of 15.4 and 14.9 mg UC781 per segment, respectively. For study 2, UC781 loading was slightly more variable between PU, EVAc, and silicone elastomer groups, with mean high drug loadings of 17.8, 15.6, and 13.4 mg UC781 per segment and mean low loadings of 5.6, 5.2, and 4.3 mg UC781 per segment, respectively (Table 1). All analyses were made using measured drug loadings (in mg/g, or wt.%) normalized to pre-insertion ring segment weights. Clinical observations Vaginal implantation of the ring segments was well tolerated in both studies, and all animals survived to the scheduled necropsy. No test article-related effects were reported for either study for the following parameters: clinical observations, body weight, and macroscopic pathology examination. Several animals were observed to have red material in the cage pan on the first few days after implantation. In study 1, ten of 24 animals had this finding, whereas only two of 72 animals had this finding in study 2. In most cases, these animals were also reported to have mild vaginal bleeding during the surgical implantation procedure. The finding did not persist beyond day 3 in study 1 and day 2 in study 2, and was therefore considered non-adverse and related to the surgical procedure. Moreover, given the brief duration of the observed post-surgical bleeding, it is unlikely the 7- to 28-day in vivo cumulative release and PK profiles of UC781-loaded ring segments were greatly affected. In vivo release In study 1, analysis of residual UC781 content in IVR segments retrieved from rabbits following 7 and 14 days in vivo was indicative of similar UC781 in vivo cumulative release for both PU and EVAc matrices (mean values of 0.89 and 0.71 mg on day 7 and 1.38 and 1.23 mg on day 14, respectively; Fig. 2a). Applying linear release kinetics to the data of this shorter duration study, daily release rates approximated to 102–128 μg/day for the first 7 days and 88–98 μg/day for the full 14-day duration. In terms of percentage release, PU and EVAc IVR segments released 8.8±2.7% and 7.9±0.9% UC781 (mean±SD) following 14 days in vivo, respectively (Fig. 2b). The longer duration and expanded design of study 2 provided more information regarding in vivo release rates as a function of duration, polymer matrix, and drug loading (Fig. 2c, d). For high loading PU and EVAc IVR segments,
Drug Deliv. and Transl. Res. (2011) 1:238–246
formulations similar to those used in study 1, the amounts of UC781 released were slightly higher than study 1 at day 14 (mean cumulative release values of 1.96 and 1.78 mg UC781, respectively). By day 28, release from both segment types slowed, with an additional 0.44 and 1.39 mg UC781 released over the second half of the study, respectively. These results are consistent with the expected non-linear release kinetics of matrix-type devices [19] and suggest the release rates will continue to decline over even longer durations. Surprisingly, high loading silicone elastomer IVR segments showed no detectable release at day 14 and only 0.35 mg UC781 release at day 28 (Fig. 2c). Analysis of residual drug content in the low loading IVR segment groups indicates that EVAc segments provided greater UC781 release (mean cumulative release values of 1.45 and 1.98 mg UC781 at days 14 and 28, respectively) compared to PU (0.61 and 0.82 mg, respectively) and silicone elastomer (0.52 and 1.06 mg, respectively) matrices. These values are similar to those calculated for the high loading EVAc ring segments; a statistically significant difference was observed at day 28 (p<0.001) but not at day 14 (p=0.23). Low loading PU and silicone IVR segments provided similar cumulative release estimates at both time points. Based on these estimations of in vivo release from the residual drug content analyses of retrieved ring segments, UC781 release from PU matrices appears to be loading dependent at both days 14 (p=0.002) and 28 (p=0.005), and release from EVAc appears to be loading dependent at day 28 only (p<0.001). Although UC781 release from silicone elastomer does not appear to be loading dependent, the higher release observed at day 28 for the low loading silicone elastomer group compared to the high loading group was found to be statistically significant (p=0.001). It should be noted that extraction and quantitation of residual UC781 content was performed using different methods and at different laboratories for each of the formulation types. It is possible that the level of variability within and across testing methods may differ, providing perhaps some explanation to the low release values estimated for the high loading silicone elastomer group. Thus, specific comparisons of estimated in vivo release between formulations, particularly of different polymer matrices, should be made with caution. As a result, statistical analysis between different polymer formulations was not performed. UC781 in vaginal tissue UC781 concentrations associated with rabbit vaginal tissue are presented in Fig. 3. Mean proximal vaginal tissue levels (adjacent to the IVR segment) were consistently higher than those measured distally. A similar observation has been
Drug Deliv. and Transl. Res. (2011) 1:238–246
b PU EVAc
2.0 1.5 1.0 0.5 0.0 7
Total UC781 Released (%)
a Total UC781 Released (mg)
Fig. 2 Estimated in vivo cumulative release of UC781 from matrix-type IVR segments. Total UC781 released in mg (a, c) and percent (b, d) determined for study 1 (a, b) and study 2 (c, d). PU polyurethane, EVAc ethylene vinyl acetate copolymer, Sil silicone elastomer. Mean+SD, N=11–12
243
PU EVAc
15%
10%
5%
0%
14
7
Time (d)
PU - high PU - low EVAc - high EVAc - low Sil - high Sil - low
5 4 3 2 1 0
d Total UC781 Released (%)
Total UC781 Released (mg)
c
14
PU - high PU - low EVAc - high EVAc - low Sil - high Sil - low
50% 40% 30% 20% 10% 0%
28
14
Time (d)
made previously with dapivirine rings in humans [16]. Mean proximal vaginal tissue levels for all groups were
155–410 ng/g in study 1 and 8–178 ng/g in study 2. Mean distal vaginal tissue levels for all groups were 27–41 ng/g
b PU EVAc
1000 100 10 1
10000
UC781 in Distal Vaginal Tissue (ng/g)
10000
UC781 in Proximal Vaginal Tissue (ng/g)
28
Time (d)
a
0.1
PU EVAc
1000 100 10 1 0.1
7
14
7
Time (d)
14
Time (d)
PU - high PU - low EVAc - high EVAc - low Sil - high Sil - low
10000 1000 100 10 1 0.1
d
PU - high PU - low EVAc - high EVAc - low Sil - high Sil - low
10000
UC781 in Distal Vaginal Tissue (ng/g)
c UC781 in Proximal Vaginal Tissue (ng/g)
Fig. 3 Tissue-associated levels of UC781 in the rabbit vagina. Vaginal tissues collected at necropsy from studies 1 (a, b) and 2 (c, d) were analyzed in two sections: proximal (a, c) and distal (b, d) to the IVR segment’s location in vivo. PU polyurethane, EVAc ethylene vinyl acetate copolymer, Sil silicone elastomer. Mean+SD, N=5–6
14
Time (d)
1000 100 10 1 0.1
14
28
Time (d)
14
28
Time (d)
244
Drug Deliv. and Transl. Res. (2011) 1:238–246
(6- to 11-fold lower) in study 1 and 0.8–17 ng/g (6- to 49fold lower) in study 2. Tissue-associated UC781 concentrations did not change significantly with time for all groups in both studies (p> 0.05), although some notable trends were observed in study 2. For low loading PU ring segments, proximal tissue levels decreased 7.7-fold between days 14 and 28, with mean proximal tissue levels of 64.5 and 8.4 ng/g at days 14 and 28, respectively. On the other hand, both loadings of silicone elastomer ring segments showed 5-fold increases in proximal tissue levels (mean values of 23–26 and 122– 125 ng/g at days 14 and 28, respectively). Comparing the effect of different polymer matrices on UC781 tissue levels showed few differences due to considerable variability within groups (coefficients of variation were 51–132% and 21–157% in proximal and distal tissues, respectively). Study 1 tissue results from PU and EVAc IVR segments were similar with no statistically significant differences when compared at a given time or tissue location (proximal vs. distal; p>0.05). In study 2, high loading polymer groups (similar loading as in study 1) also showed no significant differences in tissue-associated UC781 levels (p>0.05). However, at low loadings, PU ring segments provided approximately 14-fold lower proximal vaginal tissue levels of UC781 than either EVAc or silicone elastomer matrices at day 28 (the difference was only statistically significant for PU vs. EVAc; p<0.01). Distal tissue levels of UC781 for the low loading PU group were also significantly lower than for the low loading EVAc group at both time points (p<0.01). The effect of drug loading on tissue-associated UC781 concentrations was also evaluated in study 2. Both EVAc and silicone elastomer ring segments demonstrated no loading dependent differences in tissue levels of UC781. PU ring segments, on the other hand, did demonstrate a clear loading dependency trend, with low loading PU segments resulting in 3–6- and 8–15-fold decreases in mean tissue levels at days 14 and 28, respectively. Differences were determined to be statistically significant at day 28 only (p=0.02).
UC781 plasma levels were relatively low for all formulations tested (0.2–0.55 ng/mL in study 1 and 0.09–0.58 ng/mL in study 2, Fig. 4). Interestingly, these levels are similar in magnitude to plasma levels reported by Nel et al. for women treated for 28 days with 25 mg dapivirine silicone elastomer IVRs (0.16–1.194 ng/mL for dapivirine matrix-type IVRs) [16]. Moreover, rabbits dosed vaginally with 1 mL of 0.25% w/w UC781 gel also showed similarly low plasma levels in the range of 0.1–2 ng/mL [12]. Cmax values were observed within the first 2 days postimplantation for both studies and then subsequently decreased until reaching a steady state of 0.09–0.4 ng/mL around day 7. Although plasma levels of UC781 were maintained at a higher level for the PU group in study 1, the difference may be attributed to a single animal with high (0.46–1.1 ng/mL) plasma levels at extended durations. In study 2, small but notable differences were observed between some formulation groups. Specifically, clear loading dependent plasma levels were observed for PU IVR segments, with 3- and 3.5-fold differences between the two groups at days 1 and 28 (steady state), respectively. To a lesser extent, EVAc demonstrated a 1.5-fold difference in steady state plasma concentrations for the two drug loadings. Silicone elastomer did not demonstrate any loading dependence, consistent with the vaginal tissue data presented above. When comparing steady state UC781 plasma levels for different polymer compositions at a given drug loading, some significant differences were also observed. Specifically, the low loading PU group gave approximately 2-fold lower steady state plasma levels than either the low loading EVAc or silicone groups (p=0.004). For the high drug loading, silicone elastomer gave the lowest levels, with approximately 2-fold lower steady state plasma levels than its equivalent PU and EVAc counterparts (p<0.001). UC781 plasma levels for each group, ranked from highest to lowest, were the following: high loading PU >
PU EVAc
1.0 0.8 0.6 0.4 0.2 0.0 0
7
Time (d)
14
UC781 in Plasma (ng/mL)
b
a UC781 in Plasma (ng/mL)
Fig. 4 UC781 levels in rabbit plasma. a Study 1—mean+SD, N=11–12 for days 0–7; N=5–6 for days 8–14. b Study 2—mean+SD, N=12 for days 0–14; N=6 for days 21 and 28. PU polyurethane, EVAc ethylene vinyl acetate copolymer, Sil silicone elastomer
UC781 in plasma
PU - high PU - low EVAc - high EVAc - low Sil - high Sil - low
1.0 0.8 0.6 0.4 0.2 0.0
0
7
14
Time (d)
21
28
Drug Deliv. and Transl. Res. (2011) 1:238–246
high loading EVAc > low loading EVAc ≈ low loading silicone ≈ high loading silicone > low loading PU.
Discussion To date, only a limited number of studies have reported the preclinical or clinical pharmacokinetics of microbicidereleasing IVRs. However, the measurement of vaginal levels in rabbits following microbicide gel application is routine, and may be useful in establishing ballpark levels for ring devices in the same model. Previous work by Chien et al. presents historical precedence for using rabbits in evaluating sustained drug release from silicone intravaginal devices for both in vivo release determination and plasma PK [22]. Quantitation of drug levels in rabbit vaginal tissues, vaginal fluids, and plasma is considered essential for early preclinical evaluation of microbicide IVRs or IVR segments [24]. Due to limited availability of comparative data sets, a major unresolved issue for the microbicide field is the extent of correlation between PK in rabbits, other animal models, and women. Differences in the composition and volume of vaginal secretions between species are factors that could significantly impact the quantity and the distribution of vaginally administered drugs. Also, the urethral opening in rabbits bisects the vaginal tract with markedly distinct epithelia on either side; the upper 2/3 of the vaginal mucosa is monostratified and ciliated, while the lower 1/3 is squamous and pluristratified [25–27]. The clinical relevance of variable microbicide levels within the vaginal tract is still unknown. This study has demonstrated that surgical vaginal implantation of IVR segments in rabbits was well tolerated, showed good reproducibility in PK measured across two independent studies, and was useful for comparing different IVR compositions. Few differences in vaginal tissue and plasma levels were observed, even when compared to silicone elastomer matrices, which previously provided inadequate release in vivo in a small pilot study in nonhuman primates. Furthermore, very little control over UC781 PK was observed between groups with 3-fold difference in UC781 loadings. Only PU matrices displayed a clear loading dependent trend, which is likely related to differences in polymer solubility of UC781. Based on the observation of steady state UC781 plasma levels after an initial small burst, it is concluded that the mechanism of drug release is partition-controlled due to limiting aqueous solubility [28]. Rabbit vaginal tissue-associated concentrations of UC781 delivered from matrix-type IVR segments ranged from below to approximately two orders of magnitude higher than the in vitro EC50 for UC781 (∼2.8 ng/mL) [10].
245
In comparison, vaginal gels formulated with 0.25% w/w UC781 resulted in considerably higher vaginal tissue concentrations in the rabbit, with median values in the range of 1,000 to 10,000 ng/g [12]. Based on these rabbit PK results, UC781 delivered from a gel formulation may be expected to provide more protective local tissue concentrations than these matrix-type IVR formulations.
Conclusions The in vivo release and pharmacokinetics of UC781 delivered from matrix-type IVR segments were evaluated in two independent rabbit studies. Inter-study comparison of cumulative release in vivo, vaginal tissue, and plasma concentrations of UC781 for similar formulations demonstrated good reproducibility of the animal model. Moreover, PK results revealed few differences between the IVR formulations tested. Specifically, three polymer matrices— PU, EVAc, and silicone elastomer—tested at two UC781 loadings resulted in generally similar vaginal tissue and plasma concentrations out to 28 days in vivo, with the one notable exception of low loading PU matrices providing substantially lower PK levels. Despite low estimated in vivo release from silicone elastomer matrices, UC781 vaginal tissue and plasma PK from these formulations were remarkably similar to EVAc and high loading PU groups. Loading dependent release and PK were only clearly observed for PU matrices. In summary, these findings support the use of rabbit vaginal PK studies in early preclinical testing to support the formulation development of microbicide IVR products, particularly for low solubility drugs in which in vitro/in vivo release correlations may be unknown. Further elicitation of the links between vaginal PK in rabbits and vaginal PK in other animal species (e.g., non-human primate, sheep), and ultimately, humans, is needed. Acknowledgments The support of Missy Peet and Devon Kyle of MPI Research, Inc., in coordination of the rabbit studies and sample bioanalytical analysis, is gratefully acknowledged. This work was funded by the United States Agency for International Development (USAID) under Cooperative Agreement GPO-A-00-08-00005-00. The views of the authors do not necessarily reflect those of USAID.
References 1. Abdool Karim Q, Abdool Karim SS, Frohlich JA, Grobler AC, Baxter C, Mansoor LE. Effectiveness and safety of tenofovir gel, an antiretroviral microbicide, for the prevention of HIV infection in women. Science. 2010;329(5996):1168–74. 2. Grant RM, Lama JR, Anderson PL, McMahan V, Liu AY, Vargas L, et al. Preexposure chemoprophylaxis for HIV prevention in men who have sex with men. N Engl J Med. 2010;363(27):2587–99.
246 3. Balzarini J, Naesens L, Verbeken E, Laga M, Van Damme L, Parniak M, et al. Preclinical studies on thiocarboxanilide UC-781 as a virucidal agent. AIDS. 1998;12(10):1129–38. 4. Balzarini J, Brouwer WG, Dao DC, Osika EM, De Clercq E. Identification of novel thiocarboxanilide derivatives that suppress a variety of drug-resistant mutant human immunodeficiency virus type 1 strains at a potency similar to that for wild-type virus. Antimicrob Agents Chemother. 1996;40(6):1454–66. 5. Borkow G, Arion D, Wainberg MA, Parniak MA. The thiocarboxanilide nonnucleoside inhibitor UC781 restores antiviral activity of 3′-azido-3′-deoxythymidine (AZT) against AZTresistant human immunodeficiency virus type 1. Antimicrob Agents Chemother. 1999;43(2):259–63. 6. Pelemans H, Esnouf R, De Clercq E, Balzarini J. Mutational analysis of trp-229 of human immunodeficiency virus type 1 reverse transcriptase (RT) identifies this amino acid residue as a prime target for the rational design of new non-nucleoside RT inhibitors. Mol Pharmacol. 2000;57(5):954–60. 7. Van Herrewege Y, Michiels J, Van Roey J, Fransen K, Kestens L, Balzarini J, et al. In vitro evaluation of nonnucleoside reverse transcriptase inhibitors UC-781 and TMC120-R147681 as human immunodeficiency virus microbicides. Antimicrob Agents Chemother. 2004;48(1):337–9. 8. Zussman A, Lara L, Lara HH, Bentwich Z, Borkow G. Blocking of cell-free and cell-associated HIV-1 transmission through human cervix organ culture with UC781. AIDS. 2003;17(5):653–61. 9. Fletcher P, Kiselyeva Y, Wallace G, Romano J, Griffin G, Margolis L, et al. The nonnucleoside reverse transcriptase inhibitor UC-781 inhibits human immunodeficiency virus type 1 infection of human cervical tissue and dissemination by migratory cells. J Virol. 2005;79(17):11179–86. 10. Borkow G, Barnard J, Nguyen TM, Belmonte A, Wainberg MA, Parniak MA. Chemical barriers to human immunodeficiency virus type 1 (HIV-1) infection: retrovirucidal activity of UC781, a thiocarboxanilide nonnucleoside inhibitor of HIV-1 reverse transcriptase. J Virol. 1997;71(4):3023–30. 11. Schwartz JL, Kovalevsky G, Lai JJ, Ballagh SA, McCormick T, Douville K, et al. A randomized six-day safety study of an antiretroviral microbicide candidate UC781, a non-nucleoside reverse transcriptase inhibitor. Sex Transm Dis. 2008;35(4):414–9. 12. Clark M, McCormick T, Doncel G, Friend D. Preclinical evaluation of UC781 microbicide vaginal drug delivery. Drug Delivery and Translational Research. 2011;1(2):175–82. 13. Ahrendt HJ, Nisand I, Bastianelli C, Gomez MA, GemzellDanielsson K, Urdl W, et al. Efficacy, acceptability and tolerability of the combined contraceptive ring, NuvaRing, compared with an oral contraceptive containing 30 microg of ethinyl estradiol and 3 mg of drospirenone. Contraception. 2006;74(6):451–7. 14. Novak A, de la Loge C, Abetz L, van der Meulen EA. The combined contraceptive vaginal ring, NuvaRing: an international study of user acceptability. Contraception. 2003;67(3):187–94. 15. Henriksson L, Stjernquist M, Boquist L, Cedergren I, Selinus I. A one-year multicenter study of efficacy and safety of a continuous, low-dose, estradiol-releasing vaginal ring (Estring) in postmeno-
Drug Deliv. and Transl. Res. (2011) 1:238–246
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
pausal women with symptoms and signs of urogenital aging. Am J Obstet Gynecol. 1996;174(1 Pt 1):85–92. Nel A, Smythe S, Young K, Malcolm K, McCoy C, Rosenberg Z, et al. Safety and pharmacokinetics of dapivirine delivery from matrix and reservoir intravaginal rings to HIV-negative women. J Acquir Immune Defic Syndr. 2009;51(4):416–23. Gupta KM, Pearce SM, Poursaid AE, Aliyar HA, Tresco PA, Mitchnik MA, et al. Polyurethane intravaginal ring for controlled delivery of dapivirine, a nonnucleoside reverse transcriptase inhibitor of HIV-1. J Pharm Sci. 2008;97(10):4228–39. Johnson TJ, Gupta KM, Fabian J, Albright TH, Kiser PF. Segmented polyurethane intravaginal rings for the sustained combined delivery of antiretroviral agents dapivirine and tenofovir. Eur J Pharm Sci. 2010;39(4):203–12. Fetherston SM, Malcolm RK, Woolfson AD. Controlled-release vaginal ring drug-delivery systems: a key strategy for the development of effective HIV microbicides. Therapeutic Delivery. 2010;1:785–802. Woolfson AD, Malcolm RK, Toner CF, Morrow RJ, Lowry D, Jamil A, et al. Potential use of vaginal rings for prevention of heterosexual transmission of HIV: a controlled-release strategy for HIV microbicides. American Journal of Drug Delivery. 2006;4(1):7–20. Malcolm K, Woolfson D, Russell J, Tallon P, McAuley L, Craig D. Influence of silicone elastomer solubility and diffusivity on the in vitro release of drugs from intravaginal rings. J Control Release. 2003;90(2):217–25. Chien YW, Mares SE, Berg J, Huber S, Lambert HJ, King KF. Controlled drug release from polymeric delivery devices. III: In vitro–in vivo correlation for intravaginal release of ethynodiol diacetate from silicone devices in rabbits. J Pharm Sci. 1975;64 (11):1776–81. Damian F, Blaton N, Naesens L, Balzarini J, Kinget R, Augustijns P, et al. Physicochemical characterization of solid dispersions of the antiviral agent UC-781 with polyethylene glycol 6000 and Gelucire 44/14. Eur J Pharm Sci. 2000;10(4):311–22. Doncel GF, Clark MR. Preclinical evaluation of anti-HIV microbicide products: new models and biomarkers. Antivir Res. 2010;88 Suppl 1:S10–8. Barberini F, Correr S, De Santis F, Motta PM. The epithelium of the rabbit vagina: a microtopographical study by light, transmission and scanning electron microscopy. Arch Histol Cytol. 1991;54 (4):365–78. Barberini F, De Santis F, Correr S, Motta PM. The mucosa of the rabbit vagina: a proposed experimental model for correlated morphofunctional studies in humans. Eur J Obstet Gynecol Reprod Biol. 1992;44(3):221–7. Rodriguez-Antolin J, Xelhuantzi N, Garcia-Lorenzana M, Cuevas E, Hudson R, Martinez-Gomez M. General tissue characteristics of the lower urethral and vaginal walls in the domestic rabbit. Int Urogynecol J Pelvic Floor Dysfunct. 2009;20(1):53–60. Chien YW, Lambert HJ. Controlled drug release from polymeric delivery devices. II. Differentiation between partition-controlled and matrix-controlled drug release mechanisms. J Pharm Sci. 1974;63(4):515–9.