Preparation and Evaluation of Intravaginal Ring Containing Drospirenone

In the present study, we investigated the feasibility of the vaginal administration of drospirenone silicone IVR. The in vitro release characteristics of matrix-type and reservoir-type IVR were compared under sink conditions in 21 days. At the same time, API excipients compatibility and preformulation study was performed by HPLC, IR, and DSC methods. Biocompatibility of reservoir system was evaluated by tolerability on tissue level in rats. It was found that, under strong light exposure, high temperature, and high humidity conditions, drospirenone and excipients had no significant interactions. The daily release of reservoir-type IVR was about 0.5 mg/d sustaining 21 days, which significantly decreased the burst effect compared with the matrix system. When drospirenone was modified by the PVPk30 in the reservoir system formulation, the daily release rate increased to 1.0 mg/d sustaining 21 days. The cumulative release of reservoir-type IVR was fitted to zero release equation. In addition, biocompatibility of drospirenone IVR system in this dosage is safe. It is feasibility feasibile to further developed for safe, convenient, and effective contraceptive drug delivery with reduced dosing interval.


Introduction
Drospirenone (6b,7b,15b,16b-dimethylene-3-oxo-17apregn-4-ene-21,17-carbolactone) is an analogue of the antimineralocorticoid spironolactone ( Figure 1) that is synthesized from androstenolone. Unlike other progestogens, drospirenone, an analogue of spironolactone, has biochemical and pharmacologic pro�les similar to endogenous progesterone, especially regarding antimineralocorticoid and antiandrogenic activities. Drospirenone counteracts the estrogen-induced stimulation of the reninangiotensin-aldosterone system and blocks testosterone from binding to androgen receptors. Because of these characteristics, it has the potential to reduce body weight, blood pressure, and low-density lipoprotein levels and to enhance high-density lipoprotein levels. As a combination, oral contraceptive, drospirenone with ethinyl estradiol, is effective and has positive effects on weight and lipid levels [1].
Drospirenone has rapid oral absorption, its peak plasma concentration time is about 60-90 min, and the half-life is 30-35 h [2]. is product is to be administrated daily, otherwise it will affect the contraceptive effect. erefore, sustained release of drug is preferred above the daily administration. Intravaginal rings (IVR) can serve as an alternative for the daily administration of tablets [3]. e contraceptive vaginal ring offers effective contraception that is self-administered, requires less frequent dosing than many other forms of contraception, and provides low doses of hormones. NuvaRing (Organon, Oss, e Netherlands), the contraceptive vaginal ring approved for use in the United States, contains etonogestrel and ethinyl estradiol. It is inserted into the vagina for 3 weeks, followed by 1-week ring-free period, and works by inhibiting ovulation. Most women note a bene�cial effect on bleeding pro�les and are satis�ed with NuvaRing [3]. Serious adverse events are rare. In Chile and Peru, progesteroneonly vaginal contraceptive rings are available for nursing women. Recently more studies are ongoing new formulations of intravaginal rings in the world [4][5][6][7][8][9][10].
In the present study we investigated the feasibility of drospirenone IVR formed by hot vulcanization method with silicone elastomer as the matrix polymer and encapsulated by the same kind of polymer.

Drospirenone Solubility in Different Release Medium.
Overdose drospirenone which could keep saturated state in the following solvents was respectively placed into stoppered conical �asks containing 25 mL of a pure water, �aCL solution, PBS (pH 3.8), 0.03% SDS solution, and 0.3% SDS solution on magnetic stirrer (RT5PS25, IKA, Germany) at 37 ∘ C for 24 h. e samples were withdrawn, and subsequently the concentration of drug was assessed by UV-Vis spectrophotometric method. [10]. Differential scanning calorimetry (DSC) was employed using a modulated instrument in which the total heat �ow curve was not resolved into the reversing and nonreversing component. is curve was used to determine the solubility of drospirenone at its melting temperature in cured silicone elastomer. Drug-loaded silicone elastomer samples containing 1.0%, 2.0%, 5.0%, 10.0%, 15.0%, and 20.0% (w/w) drospirenone were moulded into sheets of 1.5 mm thickness with high temperature and high pressure, using Opening Mixing Mills (X(S)K-160, Shanghai wings of rubber and plastics machinery company limited, China). Sample cubes, approximately 5 mg in weight, were cut from the sheets and individually placed in weighed, matched (5 mg) hermetically sealed aluminium DSC test pans (Perkin-Elmer). Measurements were performed over the temperature range 40-250 ∘ C, at a heating rate of 20 ∘ C⋅min −1 , so as to include the melting point of drospirenone (196-200 ∘ C). System calibration was veri�ed using indium. e thermal properties were determined from the resulting DSC thermographs. e measurement of the area under the transition peak provides the direct base for the calculation of the energy associated with the phase transition. In this study the area was determined computationally and the areas determined were converted to the enthalpy of melting. At least three replicates of each sample were performed.

Preparation of Drospirenone-Loaded Sustained Silicone
Intravaginal Rings. To improve the dissolution in vitro of drospirenone in silicon elastomer ring, solid dispersion of drospirenone was prepared by the conventional solvent evaporation method with polyvinylpyrrolidone (PVP k30 ) as carrier [11]. In this system, the solid dispersions of 0.3 : 1 and 1 : 3 wt/wt of drug to carrier were prepared. e mixture of drug and carrier was dissolved in ethanol. e solvent was evaporated under reduced pressure using a vacuum dryer at 70 ∘ C until complete evaporation. To ensure the residual solvent was completely removed, the solid mass was further dried in a vacuum oven at room temperature for 24 to 48 hours or until the weight constant was obtained. e resulting solid was pulverized and sieved. All of the samples were passed through a �ne mesh (150 mm) and stored in desiccated environment until additional study.
Silicon elastomer C6-165 Part A and Part B was blended by Opening Mixing Mills to produce the silicon elastomer mix. e required amount of drospirenone or drospirenone solid dispersion was added to produce active mixture, and then the blends were removed from mixing mills and moulded in the circle mould (made in-house) at 130 ∘ C by �at vulcanizing machine (�LB-25D/�, Shanghai wings of rubber and Plastics Machinery Company Limited, China) for 10 min to produce matrix Drospirenone IVRs. e matrix drospirenone IVR was coated with the same type silicon elastomer to form the reservoir system to control the release rate of the drug.

Compatibility Study of Drospirenone and Excipients.
Drospirenone and PVP k30 , silicone rubber were performed the compatibility test, respectively, under the condition of 60 ∘ C, relative humidity of 92.5%, and illuminated at 4500 Lx respectively for 5 d and 10 d, at which time drug extraction and analysis were performed.   Changes of impurities from �nal treated silicon elastomer ring were extracted and detected by HPLC. e efficiency of the extraction method was about 85% as determined by the controls of blank ring spiked with a known amount of drospirenone. e test samples were dissolved in 2 mL of acetonitrile in a 50 mL centrifuge tube in triplicate followed by precipitation of the silicon elastomer by addition of 25 mL of cosolvent of water : acetonitrile = 1 : 1, v/v using a volumetric �ask. e polymer was separated by centrifugation 2000 rpm for 5 min (Anke TDL-5, Ji'nan Qike equipment limited company, China), and the supernatant was analyzed for drospirenone imprurities change by HPLC. e caps of the centrifuge tubes were closed tightly and wrapped with Te�on tape during the process to minimize evaporation. e efficiency of the extraction of drospirenone from the silicon matrix using this method was determined by quantifying the extracted drospirenone content from controls of blank silicon elastomer pellets to which known amount of drospirenone was added. e mixtures of silicon elastomer pellets and drospirenone were then subjected to the same precipitation process and drospirenone as above. e concentration of �nal samples was controlled to be about 2 mg/mL.  Aer hematoxylin-eosin staining, the slices were observed under a light microscope (×400).

Drospirenone Solubility in Release
Medium. e solubilities in different solvent were investigated in order to select optimal release medium. e solubilities of drospirenone were determined to be 13.0 g/mL, 12.1 g/mL, 8.9 g/mL, 15.2 g/mL, and 392.6 g/mL in pure water, NaCL solution, PBS (pH 3.8), 0.03% SDS solution, and 0.3% SDS solution, respectively. According to sink conditions, the concentration of drospirenone in medium is 12 g/mL in 200 mL release medium if the probable daily maximum quantity of drug is 2.4 mg. Fivefold concentration is 60 g/mL which is far less than 392.6 g/mL. erefore, 0.3% SDS solution was selected for release medium.

Drospirenone Solubility in Silicone
Elastomer. DSC was employed to determine the solubility of drospirenone at its melting temperature in a silicone elastomer IVR device [12]. A DSC thermogram is depicted in Figure 3. No thermal changes were detected for silicone elastomer over the temperature range used. Characteristic endothermal transition peaks were observed for drug-elastomer formulations containing from 1% to 20% drospirenone. Table 3 shows the thermal properties for the various drospirenone in silicone elastomer formulations at the 6 different loading levels. ermograms obtained from the replicates of each formulation showed no noticeable alteration of the position and size of the transition peak, as evident from the low standard deviations for each formulation. In addition, low standard deviations were obtained for the peak and onset temperatures within each formulation, indicating good reproducibility. Table 3 showed that ΔH (enthalpy of melting) increased linearly, with a corresponding increase in drospirenone loadings. An increase in the percentage concentration for incorporated drospirenone from 1% to 20% resulted in an increased trend in the peak temperature of the samples. e minimum peak temperature was observed for specimens containing 1% of drug (193.70 ± 0.20 ∘ C), whereas the maximum value was observed for specimens containing 20% drospirenone (195.07 ± 0.45 ∘ C). e enthalpy changes dΔH), as measured by the area of the Drospirenone transition peak in the thermograms, were determined and used to quantify the amount of dissolved drug in silicone elastomer. A plot of ΔH versus drospirenone (%, w/w loading, C) was linear (ΔH = 48.588C − 0.0992) with a correlation coefficient of 0.9978. e straight line, however, did not pass through the origin. e intercept of the straight line with the -axis (ΔH = 0), which is the value at which no thermal response of drospirenone in the drug-elastomer system was recorded, coincides with the amount of drospirenone dissolved in the elastomer at the melting temperature of the drug. us, the silicone elastomer solubility of drospirenone according to the DSC method was found to be 2.042 mg/g, which could be helpful for interpreting IVR containing drug appearance and the drug release pro�le of IVR. Additionally, no exothermic peak was seen in any of the thermograms suggesting stability of drug loaded silicon elastomer matrix in the temperature range tested.

Appearance of IVR Made in-House
. e �nal ring had the following characteristics: the matrix ring provided 5.0 mm ring cross-sectional diameter, 50.0 mm overall ring diameter, and 3.4 (±0.2) g mean weight of rings. e membrane thickness of reservoir ring was about 1.0 mm. e drospirenone incorporated polymer ring was not optically clearer (opalescent) than the blank silicon elastomer ring prepared by the same process without drug, suggesting that only part of drospirenone was dissolved in the polymer matrix. is phenomenon conforms to a result that the drug content of �nal optimized formulations was all far more than 2 mg/g that is drug solubility in silicone elastomer from the present solubility study.

Stability of Drospirenone under Preparation Conditions.
In order to further study the possibility of drug degradation or oxidation, more information was gathered using FT-IR spectroscopy. Analysis on structure of drospirenone, it was assumed that there may be any possible changes. And this could lead to the peak broadening and shiing of the absorption bands of functional groups. e infrared spectra of nontreated drospirenone, 130 ∘ C and 140 ∘ C heated, and light exposure treated are presented in Figure 4. No changes comparing treated with nontreated samples were observed. e results indicate the stability of drospirenone under high temperature and strong light exposure. e DSC thermogram of drospirenone from 40 to 240 ∘ C showed no detectable degradation ( Figure 5). e melting temperature of drospirenone was observed by an endotherm at 196 ∘ C. No exothermic peak was observed in the temperature ramp from 40 to 240 ∘ C (Figure 2) or in the isothermal step at 196 ∘ C (data not shown) indicating no detectable degradation of drospirenone under prepared conditions (130 ∘ C-140 ∘ C).
e stability of drospirenone was tested at preparing conditions (higher temperatures and strong light exposure conditions). No signi�cant changes of �uantities of impurities and no more new impurities in the HPLC spectra (see Figure 6) of the treated drospirenone sample were observed when compared with that of the nontreated samples.
All the above experiments results con�rmed that drospirenone could be able to withstand high temperature and light exposure IVR process without degrading.

In Vitro Drug Release Studies.
In vitro drug release testing was conducted in a dissolution apparatus, and release medium was 200 mL 0.3% SDS meeting with sink conditions, and release samples were determined by the UV-Vis spectrophotometry.
e mean daily release pro�les of drospirenone from matrix IVR and reservoir-type devices, loaded with 50 mg, 100 mg, and 200 mg, are shown in Figure 7. As can be seen for the matrix-type IVR, a relatively large amount of the drug permeates from the device during the initial release period (a "burst effect") followed by a decreased steady-state level thereaer. e initial burst continues for the �rst 2 days of release, although the day 1 burst is substantially larger than for day 2. e magnitude of the day 1 bursts and the subse�uent steady-state �uxes (mean daily release, mg/day) were increasing with drospirenone content increasing from 50 mg to 200 mg. However, concerning the release pro�les of reservoir-type systems, no signi�cant "burst effect" has been observed and the cumulative drospirenone release obtained was linear with time verifying a zero-order release rate ( Figure 10(b), Table 4) with corresponding decreasing daily release rate for all dosage level compared with matrix type. e result also con�rmed that it is important to note that the coating and layering prevented the initial burst effect and 8 Advances in Pharmacological Sciences T 3: ermal parameters obtained by differential thermal scanning calorimetry of drospirenone in silicone elastomer ( , mean ± SD).

Drospirenone concentration (%)
Enthalpy of melting Δ (J/g) Onset temperature ( ∘ C) Peak temperature ( ∘ C) provided a fairly stable and efficacious release of drug from the ring [10]. Due to the characterization of decreasing the "burst effect" and a zero-order release rate for the reservoir-type, the drug content added from 50 mg to 800 mg has been investigated in this kind of type VCR. From the perspective of process, 800 mg loaded was observed us maximum loading in the silicon elastomer. For a given core/sheath ratio, the only in�uence of increasing drug loading should be to prolong the duration of constant daily release. As can be seen from Figure 8, the release rate was no more increasing with 800 mg compared with 600 mg.
From release pro�le of the matrix type and reservoir type, release �ux aer 3-day could keep constant rate. A rationale for zero-order release can be the low partition coefficient of drospirenone between the polymer and the sink conditions utilized in the release studies. According to the modeling performed by Chien et al. [12], it is possible to obtain a zeroorder release from this system in cases where the thickness of the drug depletion zone is small owing to a small partition coefficient of the drug between the polymer and the sink.
is reservoir type offers advantage over the matrix type by the absence of burst release, but needs higher drug loading requirements. erefore, increasing the release rate of drug with lower drug loading was to be investigated as follows.
As paper [13] reported, the additives putting into polymer formulation could increase daily rate. e hydrophilic excipient selected was PVP k 0 in the present study. e incorporation of PVP into the IVRs had no effect of release. However, drospirenone converted into solid dispersion form with PVP k 0 carrier, and the result had an effect of increasing the release rate while slightly increasing the magnitude burst ( Figure 9 and Table 4). e release rates were not noted to increase linearly with increasing PVP K 0 loading (Figure 9), while the proportion of PVP and drug (1 : 1) had better result.
e incorporation of PVP k 0 into the IVRs had no effect of release, which is therefore considered as that the incorporation of PVP into the cross linked elastomer network did not provide a more "open" network structure leading to enhanced release rate [13]. e hydrophobic properties might not be modi�ed through the addition of hydrophilic substances PVP into the elastomer mix prior to cross linking/molding as the paper reported. Silicone elastomers are hydrophobic materials with excellent water-repellent characteristics. It is also postulated that solubility of drospirenone has an increase in the PVP-SD, while solubility decreases in the silicone elastomer network. However, this method was particularly useful in enhancing the release of hydrophobic drugs in the silicon elastomer [14].
In order to compare the different types of VCR studied in this paper, the cumulative release of different type of drospirenone IVR was �tted to zero release equation, and the cumulative release equations based on day 1 to day 21 data, along with respective r coefficient are presented in Table 4. Compared with release pro�le of matrix-type, reservoir-type and PVP-SD reservoir loading 200 mg drospirenone, ( Figure  10), the daily release of reservoir system of drospirenone IVR was about 0.5 mg, which sustained 21 days at this release rate and signi�cantly decreased burst effect compared with matrix system. �hen drospirenone is modi�ed with PVP k 0 in the reservoir system formulation, the daily release quantity increased and sustained 21 days at the release rate of about 1.0 mg/d, which had a decreasing burst effect and constant rate compared with matrix ( Extrapolating the cumulative �ux to the matrix, reservoir-type ring, and PVP-SD reservoir-type ring loaded 200 mg shown in Figure 10, we achieved roughly 0.5, 1.0 mg drospirenone daily release during 21 days with low burst effect. It is uncertain what quantity of drospirenone should be released from the IVR in vitro since a disparity oen exists between in vitro release and in vivo vaginal release and subsequent biodistribution. For these reasons, controlled release is critical to the formulation of drospirenone IVR. Given that drospirenone shows initial loading-dependent release in our formulation in vitro, the loading can likely be adjusted to release an amount of drug that is safe and deemed adequate to prevent pregnancy in humans. Furthermore, our IVR design allows for adjustment of the PVP content. is system offers advantage over the reservoir system by having lower drug loading requirements and the absence of burst release. e release rate of drug was observed to be proportional to the drug loading in a certain range.
Although the release rate can be modi�ed by changing the initial loading, this might also change the duration of desired release rate for the system. However, we have demonstrated that sustained zero-order release of drug can be obtained successfully from rings ranging from 0.3 to 1.0 mg/day, depending on drug loading.

Compatibility Study of API and
Excipients. e compatibility of drospirenone with PVP k30 and/or silicone elastomer was tested respectively under the condition of 60 ∘ C, relative humidity of 92.5%, and illuminated at 4500 Lx, respectively, for 5 d and 10 d, at which time drug extraction and analysis was performed. Changes of impurities (single impurities and total impurities) from �nal treated silicon elastomer ring were extracted by solvent, detected by HPLC, and calculated as area percent (%) of total peak area (Tables 5, 6, and 7).
We studied drug degradation using high temperature and humidity conditions since this would be indicative of long-term drug degradation under ambient conditions and may suggest possible degradation mechanisms. e results indicate no signi�cant degradation of drospirenone in polymer at 5 and 10 days when compared to day 0. In addition, there were no new peaks in the sample chromatograms that would indicate possible degradation. Altogether, these results indicate that drospirenone is stable in their respective PVP or silicon elastomer. e results of this study demonstrate  Figure 11 showed the histopathology of the vaginal mucosa aer intravaginal application of IVR. As compared to the control with no treatment, the IVR-treated group showed no visible sign of in�ammation or necrosis. IVR did not affect the morphology of vaginal tissues, which indicated that IVR device was being with good biocompatibility and such drug delivery systems might be safe for vaginal delivery.

Conclusions
Intravaginal ring (IVR) containing drospirenone is one kind of vaginal ring devices, and the drug-loaded core was obtained by using hot vulcanization method with silicone elastomer as the matrix polymer, which was encapsulated by the same kind of polymer membrane. A method for this preparation of drospirenone reservoir system was obtained by adding PVP k30 additives to increase drospirenone solubility. Drospirenone was used as the active pharmaceutical ingredient and silicone elastomer as matrix and controlled release membrane agent. PVP k30 was used as increasing drospirenone solubility and without resulting in the drospirenone degradation detected by HPLC gradient method. Under strong light, high temperature, and high humidity conditions, drospirenone raw materials and formulation excipients PVP k30 and silicon rubber had no signi�cant interactions and indicate good compatibility. It is found that the drospirenone release rate was increasing with the PVP k30 amount at a constant range. Silicone elastomer was also employed as membrane for controlling membrane permeability. In vitro drug release testing was conducted in a dissolution apparatus, the release medium was 200 mL 0.3% SDS meeting with sink conditions, and release samples were determined by the UV-Vis spectrophotometry. e daily release of reservoir system of drospirenone IVR was about 0.5 mg, which sustained 21 days at this release rate and signi�cantly decreased burst effect compared with matrix system. When the drug is modi�ed by PVP k30 in the reservoir system formulation, the daily release quantity increased and sustained 21-days at the release rate of 1.0 mg/d. e cumulative release of reservoir of drospirenone IVR was �tted to zero release equation. In addition, drospirenone IVR system was evaluated by tolerability on tissue level in rat. Drospirenone IVR did not affect the morphology of vaginal tissues. erefore, the dosage form might be further developed for safe, convenient, and effective contraceptive drug delivery with reduced dosing interval. e results of this study demonstrate that drospirenone may be stably formulated into silicon elastomer intravaginal ring for 21-day sustained release.

Con�ict of �nterests
e authors declare that they do not have any con�icts of interest concerning their submitted paper.