In Vitro and in Vivo Evaluations of Buccal Tablet Formulations of Ritodrine Hydrochloride

Abstract

Buccal tablets of ritodrine (RD) hydrochloride (HCl), called RD-HCl, were prepared using the direct compression method with alginate (AL), lactose (LC), magnesium stearate (ST), and microcrystalline cellulose (MC) as excipients. The tablets were evaluated based on hardness, and tablets weighing 80 mg and with hardness of greater than 30 N were chosen as appropriate ones. As a result, tablets composed of RD-HCl (4 mg)/LC (38.5 mg)/ST (0.5 mg)/MC (37 mg) and RD-HCl (4 mg)/AL(7 mg)/LC (28.5 mg)/ST (0.5 mg)/MC (37 mg), called D9 and D10, respectively, were selected. These tablets were further evaluated based on in vitro dissolution and in vivo absorption studies in rats. D9 rapidly released RD, achieved an effective plasma concentration from 15 min to 7 h after its buccal administration, and did not exceed the toxic plasma level of 80 ng/mL. D10 gradually released RD, and maintained an effective concentration from 1 h to 7 h after its buccal administration, without exceeding the toxic plasma level. The absorption was more prolonged in D10 than D9. Their in vivo release was considered to be caused gradually from the amount of RD remaining in the oral cavity at 7 h, in particular D10. The superior retention of D10 in plasma and oral cavity appeared to be related to its higher mucoadhesive properties. Although these results were obtained using rats, they suggest that the chosen tablets should have adequate characteristics from the viewpoints of plasma levels.

Ritodrine (RD) hydrochloride (HCl), called RD-HCl, is essential in the treatment of premature labor or imminent abortion,^1,2) which have been shown to increase the risk of damage to newborn infants.^3,4) Respiratory complications associated with preterm birth can lead to obstructive lung diseases in the later life. RD, a β2-adrenergic agonist, induces relaxation of the uterine smooth muscle, and is selected as a first-line drug in the treatment of premature labor.^5,6) Although RD is known to exhibit highly specific β2-adrenergic actions, its β1-adrenergic function still exists to some extent.²⁾ Therefore, the side effects associated with an increase in the plasma concentration of RD, such as an elevated heart rate and chest pain, must be carefully monitored.^2,7,8)

RD is generally intravenously infused or orally ingested. Also, it is rapidly administered into the systemic circulation via intravenous infusion in urgent cases, such as to inhibit preterm labor.^9,10) The plasma concentration of RD was recently shown to be closely associated with the suppression of preterm labor.^2,8) Furthermore, it had to be maintained above a minimal effective concentration in order to preserve the pregnant state. Although intravenous infusions can maintain effective plasma concentrations of RD, patients are confined to bed rest through the infusion period, which may reduce their QOL. Once the prevention of premature labor is achieved under certain dosing conditions, the administration of RD can be converted from its intravenous infusion to oral ingestion, which is markedly better for the QOL of patients.¹¹⁾

The oral dosage form of RD is a commercially available tablet. Since the tablet can be easily taken and is highly acceptable due to the low burden on patients, its administration can be continued for a long period of time. However, the bioavailability of the oral RD formulation is known to be very low, namely, approximately 30% or less,^12,13) and its plasma half-life of RD is also relatively short at 1–2 h; therefore, the drug has to be taken orally at least 3 times per day.

The disadvantages associated with both the intravenous infusion and tablet formulation of RD, as described above, need to be improved. Oral mucosal absorption has recently attracted attention because it is easy for drug dosing, allows the first-pass effect in the intestine or liver to be avoided,^14,15) enables relatively fast absorption,^16–18) and maintains prolonged systemic levels.¹⁹⁾ Thus, we previously attempted buccal administration as a novel administration route,²⁰⁾ and found that the bioavailability and maintenance of plasma concentrations of RD were superior with buccal administration than with oral dosing. Relatively rapid absorption was also attained with the buccal administration of RD. We developed buccal tablets of RD-HCl for easier handling²¹⁾; tablets with good hardness and fast drug release were produced using alginate (AL), lactose (LC), magnesium stearate (ST), and microcrystalline cellulose (MC) as excipients. MC was found to be crucial for providing good hardness and achieving rapid drug release. However, a detailed formulation study was not performed in our previous study. Therefore, tablet formulations with AL/LC/ST/MC or LC/ST/MC as excipients were herein investigated in detail from the viewpoints of tablet size, hardness, drug release, and the maintenance of plasma concentrations.

MATERIALS AND METHODS

Materials

RD-HCl, phosphoric acid, urethane, sodium alginate of 80–120 cP grade (AL: 80–120 mPa·s for 10 g/L at 20°C), and ST were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan). LC of JP16 grade was supplied by Miyazawa Yakuhin Co., Ltd. (Tokyo, Japan). Avicel PH301, purchased from Asahi Kasei Corp. (Tokyo, Japan), was used as MC. All other chemicals used were of reagent grade.

Animals

Male Wistar rats (7 weeks old, 200–210 g) were purchased from Tokyo Laboratory Animal Science Co., Ltd., and used for in vivo experiments several days after being purchased. Rats were kept on the MF breeding diet supplied by Oriental Yeast Co., Ltd. (Tokyo, Japan) with water ad libitum in a room in which temperature and relative humidity were kept at 23±1°C and 60±1%, respectively. The light–dark cycle was set at 12 h. The experimental protocol was approved by the Committee on Animal Research of Hoshi University (Tokyo, Japan), and animal experiments were performed according to the Guiding Principles for the Care and Use of Laboratory Animals of Hoshi University.

Preparation of Tablets

Disk-shaped tablets were produced by the direct compression method using an SSP-10A manual press (Shimadzu Corp., Kyoto, Japan) with an 8-mm flat face punch and die set. Tablets with no drug (NDn) were initially prepared using AL, LC, ST, and MC as excipients. Each tablet formulation is shown in Table 1. The top 5 formulations were tested in order to clarify the influence of MC on tablet hardness. The bottom 4 formulations were designed to keep the tablet weight at 76 mg by decreasing the amount of LC while increasing that of MC. In all tablets, the power mixture was weighed according to the composition in Table 1, placed into the cylinder, and compressed at 2 kN for 30 s.

Table 1. Formulation and Hardness of Tablets without RD

Formulation	AL (mg)	LC (mg)	ST (mg)	MC (mg)	Hardness (N)a)
ND1	7	48	1	20	11.1±1.5
ND2	7	48	1	30	18.9±0.6
ND3	7	48	1	40	29.7±6.7
ND4	7	48	1	60	52.3±3.6
ND5	7	48	1	80	76.1±6.9
ND6	7	58	1	10	7.6±2
ND7	7	38	1	30	20.7±1.2
ND8	7	28	1	40	32±1.2
ND9	7	18	1	50	33±1.2

a) Mean±S.D. (n=3).

RD-HCl (4 mg) was used in RD-containing tablets (Dn) with AL, LC, ST, and MC being added as excipients. Each tablet formulation is shown in Table 2. In these tablets, the upper 5 formulations were prepared to investigate the influence of MC on tablet hardness. The lower 6 formulations were prepared by reducing the LC amount while increasing that of MC in order to keep the tablet weight at 80 mg. The powder mixture was weighed based on the formulation in Table 2, placed into a cylinder, and then compressed at 2 kN for 30 s.

Table 2. Formulation and Hardness of RD-Containing Tablets Formulation

Formulation	RD (mg)	AL (mg)	LC (mg)	ST (mg)	MC (mg)	Hardness (N)a)
D0	4	7	48	1	0	7±1.5
D1	4	7	48	1	20	21.4±4.7
D2	4	7	48	1	40	35.1±2.7
D3	4	7	48	1	60	50±5
D4	4	7	48	1	80	63.4±7.8
D5	4	7	48.5	0.5	20	24±1.3
D6	4	0	48.5	0.5	27	29.4±1
D7	4	7	38	1	30	23.2±1.2
D8	4	7	38.5	0.5	30	27.5±2.6
D9	4	0	38.5	0.5	37	40.5±2.5
D10	4	7	28.5	0.5	40	39.2±2

a) Mean±S.D. (n=3).

Measurement of Hardness

The side of the tablet was sandwiched between the flat plates of a Kiya-type hardness meter (Fujiwara Scientific Co., Ltd., Tokyo, Japan). Pressure on the tablet was gradually increased, and the force immediately before it was crushed was measured as hardness.

Tablet Selection and Measurement of Physical States

The hardness of RD-containing tablets (Dn) weighing 80 mg was examined, and a hardness of greater than 30 N was used as a requirement. The physical features of the selected tablets were investigated in more detail. Tablets were weighed (n=3) and their thickness was measured using a micrometer (n=6). The drug content per tablet was investigated as follows (n=3). A tablet was placed in 10 mL of the JP16 2nd fluid, that is, 0.05 M phosphate buffer of pH 6.8, and incubated by horizontal shaking at 60 strokes per min at 37°C for 24 h. The resultant solution or suspension was stirred vigorously with a vortex mixer. After centrifugation of the sample, the supernatant was diluted adequately using the 2nd fluid, and analyzed fluorometrically with a FP-777 spectrofluorometer (Jasco Corp., Tokyo, Japan) at an excitation wavelength of 278 nm and emission wavelength of 306 nm. The absolute calibration curve method was used to determine RD concentrations at which a linear calibration was obtained, 0.1–12 µg/mL/mL (R²=0.999).

In Vitro Dissolution Studies

Dissolution studies were conducted on the chosen tablets. Drug release profiles were investigated by the JP16 paddle method using an NTR-3000 dissolution tester (Toyama Sangyo Co., Ltd., Osaka, Japan). JP 2nd fluid (400 mL) was used as a test medium at 37°C, and the paddle was rotated at 50 rpm at a position 2.5 cm from the bottom of the container. After a tablet was placed in the medium, aliquot samples (4 mL each) were withdrawn at the appropriate time points; fresh JP 2nd fluid (4 mL) was added after each sampling. The clear supernatant of the sample was measured fluorometrically in the same manner described above.

In Vitro Mucoadhesion Studies

The mucoadhesive properties of tablets D9 and D10 were investigated in vitro. After the rats were sacrificed by excessive anesthesia with ether inhalation, their oral mucosa tissue (approximately 1 cm in length×1 cm in width×1 mm in thickness) was taken from the buccal mucosa around the bottom lip. The mucoadhesion of the tablet was examined based on the method by Agarwal et al.²²⁾ The opposite side of the mucosal membrane was fixed to the glass plate with glue, and the one flat side of the tablet was also fixed to the glass plate with glue. The glass plate with the mucosal membrane fixed to the table so that the mucosal surface was set upward. After 50 µL of saline was dropped on the mucosal membrane, the tablet surface was set to the downward direction, and brought into contact with the mucosal membrane, and 0.49 N was added to the tablet/membrane composite for 30 s. After the removal of the pressure, the tablet holder was lifted with the increase in force at 0.06 N/min. The force immediately before the tablet was detached from the mucosal membrane was determined as mucoadhesive strength.

In Vivo Absorption Studies

Buccal absorption was examined for the chosen tablets. After rats were anesthetized by an i.p. injection of 1.2 mL of a 25% (w/v) urethane solution in saline, they were fixed on their backs. A sample (0.3 mL) of blood was collected prior to the administration of tablets. After one tablet had been placed on the right cheek mucosa of the rat, 0.1 mL saline was placed on the mucosa near the bottom of the inserted tablet, and the tablet was gently pressed against the mucosal surface at the administered site. Blood samples (0.3 mL) were withdrawn via the jugular vein at the appropriate time points. Plasma was obtained by centrifuging blood samples at 1400×g for 5 min. Plasma (0.1 mL) was mixed with 0.1 mL of 1 M carbonate buffer (pH 9.5) and stirred vigorously. A total of 0.5 mL ethyl acetate was then added, shaken vigorously, and centrifuged at 1500×g for 10 min. The resultant organic phase (0.3 mL) was collected, and the solvent was evaporated under nitrogen gas. The residue was dissolved in 0.1 mL of the HPLC mobile phase. Forty microliters of the solution was injected onto the HPLC column to determine the RD concentration.

Seven hours after the buccal administration of RD, the whole tablet residue in the oral cavity was collected using a small spatula, and transferred into a glass tube. Saline was added to the residue to adjust the volume to 8 mL. A portion (0.4 mL) was collected after vigorous shaking. It was treated in a similar manner to the plasma sample. The volume ratio of sample/1 M carbonate buffer/ethyl acetate at 1/1/5 (v/v) was used in the extraction. The organic phase (0.5 mL) was obtained after centrifugation of the operated mixture, and dried under nitrogen gas. The residue was dissolved in the mobile phase (0.5 mL), diluted appropriately, and the final sample (40 µL) was injected onto the HPLC column. The amount of RD in the residue in the oral cavity was calculated from the RD concentration.

HPLC Assay

A HPLC assay of RD was performed at room temperature using a Shimadzu LC-6AD pump equipped with a Shimadzu RF-10AXL fluorescence detector and a Shimadzu C-R7A chromatopac. A YMC Pack ODS-AM column (6 mm inner diameter ×150 mm length; YMC Co., Ltd., Kyoto, Japan) was used as an analytical column. A mixture of 0.05% phosphoric acid and acetonitrile (5 : 1, v/v) was used as the mobile phase. Detection was performed with the fluorescent detector set at excitation and emission wavelengths of 278 nm and 306 nm, respectively. The concentration of RD was determined by the absolute calibration curve method.

Statistical Analysis

The unpaired t-test was used for statistical analyses, and a significant difference was set as p<0.05.

RESULTS AND DISCUSSION

Tablet Formulations and Hardness

As reported previously,²¹⁾ although AL/LC tablets could be formed adequately by the direct compression method, their hardness was insufficient. The addition of MC improved the strength of tablets. The addition of MC also provided tablets with disintegration properties, thereby enhancing drug dissolution. Although the addition of MC to AL/LC tablets enhanced tablet characteristics such as hardness and drug release, these effects had not been investigated in detail. The weight of these tablets, approximately 120 mg, was also slightly large; therefore, a smaller tablet size was desirable.

In the present study, first, the effects of MC on tablet strength were examined in the absence of RD. The influence of adding MC was investigated by comparing the 5 formulations. Tablets of a size that was easy to place on the oral mucosa were considered desirable for buccal dosing. Thus, the weight of tablets was set at 76 mg, which was achieved by reducing the amount of LC while increasing that of MC. Tablet hardness of greater than 30 N was generally required to prevent it from breaking with normal handling. As shown in Table 1, a larger amount of MC was needed than LC to achieve this hardness. ND8 and ND9 were found to be acceptable.

RD-containing tablets (Dn) were subsequently produced by referring to the above estimation of the relationship between the formulation and hardness of tablets with no RD, and the hardness of the obtained tablets was then investigated. The addition of MC influenced tablet hardness in a similar manner to that of tablets with no RD. D1 and the bottom 6 formulations in Table 2 were prepared with a tablet weight of 80 mg. In D1 vs. D5 and D7 vs. D8, 0.5 mg of ST appeared to increase tablet hardness more than 1 mg of ST. A comparison of D8, D9, and D10 revealed elevations in tablet hardness with increases in the amount of MC. When the tablet hardness was compared between D5 and D6, or D8 and D9, the increase in the MC amount was observed to raise the tablet hardness. MC appeared to contribute to the increase in tablet strength, while AL did not seem to work on the elevation of tablet harness. In fact, as reported previously,²¹⁾ tablets, made using AL as a major component, showed low tablet strength. D9 tablets could be formed well even in the absence of AL, and this was attributed to the good binding ability of MC. Since hardness greater than 30 N was required for tablet strength, D9 and D10 were selected as tablets with a weight of 80 mg, thereby fulfilling the requirements.

In Vitro Characteristics of Chosen Tablets

The physical characteristics of the chosen tablets, D9 and D10, were investigated in detail. The amount of RD contained in these tablets needed to be determined in order to exactly calculate the ratio of RD released. The physical features of these tablets are shown in Table 3. The tablet weight indicated nearly 99% of the whole amount was recovered. From the content of RD in each tablet, approximately 95% of the drug added was recovered, which indicated that approximately 5% was lost in the tableting process. This might be related to the fact that RD tended to attach to the metal or glass surface. In the drug release test, the mean content was used as 100% in the calculation of release percentage. These physical features were considered to be sufficient from the viewpoint of recovery ratios.

Table 3. In Vitro Characteristics of Chosen Tablets, D9 and D10

Formulation	Tablet weighta) (mg)	Tablet thicknessb) (mm)	Drug contenta) (mg)	Mucoadhesive strengtha) (N)
D9	78.95±0.28	1.12±0.02	3.80±0.02	0.012±0.001
D10	78.98±0.17	1.12±0.01	3.82±0.01	0.035±0.024

a) Mean±S.D. (n=3). b) Mean±S.D. (n=6).

In vitro release was investigated using the JP16 2nd fluid. The release profiles of D9 and D10 were obtained, as shown in Fig. 1. D9 disintegrated quickly within 10 min, resulting in the rapid release of RD. The mean release percentage from D9 was 76% at 20 min and 83% at 40 min. The release hardly increased after 40 min, which was assumed as follows; 1) the main reason was considered to be because the drug would still remain in the disintegrated tablet to a small extent, not released completely. 2) The other reason, considered to be minor, was that small loss of RD might happen in the carry and store in the glass bottle because RD tend to attach to the surrounding surface of glass or metal as stated above. On the other hand, D10 swelled in the first stage and took more than 1 h to disintegrate completely. RD was released gradually, with 63% being released 1 h after the start of the test. At 2 h, the average release reached approximately 76% for D10 (data not shown in Fig. 1). These release features demonstrated that D9 was useful for fast drug absorption due to the quick release of RD, while D10 was more suitable for prolonged drug absorption because of the gradual release of RD.

Fig. 1. Release Profiles of RD from Chosen Tablets, D9 and D10

Results are expressed as the mean±S.D. (n=3).

The tablets D9 and D10 were investigated for their mucoadhesive properties. The mucoadhesive strength is shown in Table 3. The strength of D10 tended to be greater than that of D9. Although D9 absorbed water quickly from the membrane surface, the tablet wet surface did not become viscous, and disintegrated rapidly. After the mucoadhesion test, the tablet surface was not sticky. On the other hand, the wet surface of D10 was viscous, and the surface after the mucoadhesion test showed sticky characteristics. As AL is known as a bioadhesive polymer, it was considered to contribute importantly to mucoadhesive properties of D10.

In Vivo Evaluation of Chosen Tablets

Plasma RD concentrations were monitored for 7 h after the buccal administration of D9 and D10. When these tablets were placed on the buccal mucosa, difficulties were encountered when fixing it to the site. Therefore, a small volume of saline (0.1 mL) was added around the placement site. However, D9 did not adhere well to the buccal mucosa. Since AL, which has a mucoadhesive function, was not contained in D9 tablets, these tablets could not be fixed well to the mucosa. D9 quickly became wet, disintegrated within 30 min of its buccal administration, and the resultant powder dispersed gradually over the oral cavity. On the other hand, D10 was fixed well to the buccal mucosa following the addition of saline at the tablet placement site due to mucoadhesive characteristics of AL. D10 became wet, swelled, and remained at the administration site to a large extent even after 2 h.

Plasma concentration–time profiles were obtained for RD after the buccal administration of D9 or D10, as shown in Fig. 2. For D9 tablets, the drug was rapidly absorbed. The mean plasma concentration of RD was 19 ng/mL at 15 min, which was higher than the minimal effective RD concentration of 15 ng/mL, though this level was the concentration for humans. The plasma concentration of RD increased gradually until 2 h, and was then slowly eliminated. The plasma concentration of RD was still higher than the minimal effective level after 7 h. Thus, D9 maintained an effective concentration of RD from 15 min to 7 h after its buccal administration. The plasma concentration of RD obtained following the buccal administration of D10 was lower than the minimal effective level in the initial periods, 15 and 30 min, but the level reached a concentration of greater than 15 ng/mL after 1 h. The RD concentration increased until 4 h, and was then kept at the same level for another 3 h. D10 maintained an effective RD level from 1 to 7 h after its buccal administration. A mono-exponential extrapolation of the decay curves from 4 to 7 h revealed that D9 and D10 kept the plasma concentration of RD above the minimal effective level (15 ng/mL) until 10 h and 30 h after their buccal administration, respectively. D10 was expected to maintain the RD concentration for a much more prolonged period than D9. In addition, the plasma concentration varied more widely with D10 than D9. As D9 disintegrated more simply in vivo than D10, D9 was considered to exhibit less variability in the in vivo release, leading to less variation of the plasma concentration. The variation in plasma concentration regarding D10 might indicate that the dose of RD should be increased in order to ensure the plasma level of more than 15 ng/mL. The plasma concentration–time curve after the intravenous (i.v.) administration of RD (1 mg/kg), as reported previously,²⁰⁾ was superimposed onto the results obtained for D9 and D10 (Fig. 2). The plasma concentration of RD was greater than 180 ng/mL 5 min after its i.v. administration, and decreased rapidly to less than 15 ng/mL after 4 h. An RD concentration of greater than 80 ng/mL had been reported to cause toxic side effects, such as an elevated heart rate, chest pain, and shortness of breath, whereas less than 15 ng/mL may not be able to maintain the pregnancy.²⁾ Therefore, although the i.v. administration of RD achieved the highest exposure, it was considered to be inadequate from the viewpoints of toxic side effects and efficacy.

Fig. 2. Plasma Concentration–Time Profiles after Buccal Administration of Chosen Tablets, D9 and D10

Regarding intravenous data, the observed (▲; mean±S.E. (n=4)) and calculated (---) values are plotted. D9: Mean±S.E. (n=3), D10: Mean±S.E. (n=4), The dashed-dotted line shows the level of 15 ng/mL.

Pharmacokinetic parameters were calculated based on each profile, and the results obtained are shown in Table 4, in which the area under the curve (AUC), mean residence time (MRT), and variance of residence time (VRT) were calculated for 0–7 h using the data observed based on the trapezoidal rule with the program MULTI.²³⁾ In order to compare profiles with the intravenous infusion of RD, the calculated plasma concentration–time profile after its i.v. administration at 1 mg/kg, reported previously,²⁰⁾ was described in Fig. 2 with a dotted line.   

(1)

in which C_{calc, iv}(t) was the calculated concentration (ng/mL) at time t (h). Since Eq. 1 was well fitted to the mean observed profile, it was used to calculate the AUC for 0–7 h after the i.v. administration of RD at 1 mg/kg.

The C_max values of both D9 and D10 indicated that plasma concentrations were maintained below toxic levels, while the MRT and VRT values showed that absorption was more prolonged than the intravenous infusion. The T_max value was significantly larger in D10 than in D9 (p<0.05), while the MRT value was slightly larger in D10 than in D9 (p>0.05). Therefore, the absorption of D10 was more prolonged than that of D9, which was attributed to differences in tablet characteristics. Namely, mucoadhesive properties were observed for D10, but not for D9. D9 released RD rapidly, whereas D10 gradually released RD (Fig. 1). These physical and biological features were considered to be associated with the greater T_max and MRT values of D10. Absolute bioavailability (B.A.) values were calculated using the following equation.   

(2)

in which AUC(buccal) and AUC(i.v.) represented the AUC values of buccal (20 mg/kg) and i.v. (1 mg/kg) administrations, respectively. The AUC(i.v.) value for 0–7 h was 206.2 ng·h/mL from Eq. 1. As the present results were calculated for the data from 0–7 h, the B.A. values of D9 and D10 were considerably low (Table 4). The B.A. values of D9 and D10 from 0–∞ were assumed to be larger by the following estimation based on the mono-exponential extrapolation of the plasma profiles. Namely, when the mono-exponential extrapolation to the infinite time was conducted, and then the mean AUC(0–∞) values in D9 and D10 were estimated, they were calculated as 596 and 1272 ng·h/mL, respectively. The i.v. mean AUC(0–∞) value was calculated as 210 ng·h/mL. Although these were values calculated by simple extrapolation, the B.A. values from 0–∞ were estimated to be 14 and 30% for D9 and D10, respectively. On the other hand, the previous paper²⁰⁾ indicated that the intragastric B.A. value from 0–∞ was estimated to be approximately 4%. The longer monitoring would give a more exact bioavailability for each formulation.

The wet powder products remaining in the oral cavity 7 h after the buccal administration of D9 and D10 were recovered. The amounts of RD in the remaining products are shown in Fig. 3. The mean amount of RD remaining was very low, at less than 1%, for D9, while it was approximately 22% for D10; however, this difference was not significant due to the large variations noted for D10. These results were considered to be attributed to the physical and biological features of each tablet. D9 disintegrated quickly, rapidly released RD, and did not adhere to the mucosa, resulting in RD diffusing almost completely over the oral mucosa. Therefore, RD was considered hardly to be recovered from the remaining powder of D9. D10 was mucoadhesive and remained around the administration site to a fairly large extent even after its disintegration. AL suppressed powder diffusion because of its viscous characteristics and sticky to the mucosa even after the collapse of the tablet. Thus, RD was considered to remain to a certain extent in the wet powder after the disintegration of D10. Individual variabilities in saliva volumes and tablet fixation were related to the large variations observed in the remaining amounts of D10. As only the remaining wet powder in the oral cavity was collected roughly in this experiment, more RD should exist in the oral cavity. Therefore, Fig. 3 only suggested that in vivo release should be slower with D10 than D9.

Table 4. Pharmacokinetic Parameters of Chosen Tablets, D9 and D10

Formulation	Cmax (ng/mL)	Tmax (ng/mL)	AUC (0–7 h) (ng·h/mL)	MRT (0–7 h) (h)	VRT (0–7 h) (h2)	B.A. (0–7 h) (%)
D9	27.1±0.64	1.7±0.3	137.4±15.7	3.3±0.3	4.3±0.1	3.3
D10	30.9±13.7	4.8±0.8	153.6±79.2	4.2±0.3	4.1±0.3	3.7

AUC, MRT, VRT and B.A. were calculated for 0–7 h by the trapezoidal rule based on the data in Fig. 2. B.A. was calculated using Eq. 2, in which the AUC (i.v.) value was 206.2 (ng·h/mL). The results are expressed as the mean±S.E. (n=3 for D9, 4 for D10). *: p<0.05 vs. D9.

Fig. 3. Amount of RD Recovered 7 h after Buccal Administration of Chosen Tablets, D9 and D10

Results are expressed as the mean±S.E. (n=3 for D9, 4 for D10).

As to the retention of RD in the mucosal tissue, it was estimated from the absorption studies in the application of the cotton ball loaded with RD solution.²⁰⁾ The cotton ball loaded with RD solution was applied for 1 h and then removed. The drug concentration at the administration site and in the mucosal tissue was measured immediately and 3 h after the removal of cotton ball. At that time, the drug concentration in the oral cavity was quickly eliminated extensively 3 h after the removal of cotton ball (data not shown). On the other hand, the mucosal concentration did not change so much between immediately and 3 h after the removal of cotton ball; the concentration of RD in mucosal tissue was 78 µg/g immediately after cotton ball removal, and 43 µg/g 3 h after cotton ball removal. This suggested that the mucosal tissue should act as a reservoir of RD in the mucosal absorption.

In conclusion, in the present formulation study, two kinds of tablets, D9 and D10, were chosen as possibly useful buccal tablets. The results of the in vivo study revealed that these tablets achieved effective plasma concentrations of RD over a long period of time without exceeding toxic levels. D9, which did not contain AL, showed rapid absorption, but was expected to be effective for approximately 10 h from the mono-exponential extrapolation of the plasma concentration–time curve. On the other hand, D10, which comprised both AL and MC, reached the minimal effective level at 1 h, and was presumed to maintain an effective RD concentration for a prolonged period (30 h) without exceeding the toxic level. Tablets with AL/LC/ST/MC or LC/ST/MC as excipients may be useful dosage forms for the buccal delivery of RD. Although the in vivo results were obtained in rats, they can provide important information for the consideration of the formulations of buccal dosage forms of RD for clinical use.

Conflict of Interest

The authors declare no conflict of interest.

REFERENCES

• 1) Gross AS, Brown KF, Baird-Lambert JA, Nation RL. Determination of ritodrine in blood and plasma by high-performance liquid chromatography with fluorescence detection. J. Chromatogr. B, 416, 400–408 (1987).
• 2) Konda A, Ito T, Yoshida H, Toda T, Hayakawa T, Inotsume N. Pharmacokinetics of ritodrine diastereomers in patients pregnant with singletons and twins. Eur. J. Clin. Pharmacol., 65, 913–917 (2009).
• 3) Chye JK, Lim CT, Leong HL, Wong PK. Retinopathy of prematurity in very low birth weight infants. Ann. Acad. Med. Singapore, 28, 193–198 (1999).
• 4) Kwinta P, Pietrzyk JJ. Preterm birth and respiratory disease in later life. Expert Rev. Respir. Med., 4, 593–604 (2010).
• 5) Liu A, Liu F, Xu Y, Xu F, Hu W, Guo Q. Determination of ritodrine in human plasma by high-performance liquid chromatography coupled with electrospray ionization tandem mass spectrometry. J. Chromatogr. B: Analyt. Technol. Biomed. Life Sci., 867, 144–148 (2008).
• 6) Marzo A, Coa K, Fontana E, Tavazzi S, Bo LD, Ismaili S, Zava D, Cantoni V, Bertolini A. Pharmacokinetics and pharmacodynamics of ritodrine hydrochloride administered orally and intramuscularly to female healthy volunteers. Arzneimittelforschung, 60, 510–518 (2010).
• 7) Caritis SN, Lin LS, Wong LK. Evaluation of the pharmacodynamics and pharmacokinetics of ritodrine when administered as a loading dose. On establishing a potentially useful drug administration regimen in cases of fetal distress. Am. J. Obstet. Gynecol., 152, 1026–1031 (1985).
• 8) Fujimoto S, Akahane M, Tanaka T, Hanatani K. Pharmacokinetics and materno-fetal influence of ritodrine hydrochloride during the continuous treatment of threatened premature labor. Nihon Sanka Fujinka Gakkai Zasshi, 41, 321–327 (1989).
• 9) Caritis SN, Lin LS, Toig G, Wong LK. Pharmacodynamics of ritodrine in pregnant women during preterm labor. Am. J. Obstet. Gynecol., 147, 752–759 (1983).
• 10) Caritis SN. A pharmacologic approach to the infusion of ritodrine. Am. J. Obstet. Gynecol., 158, 380–384 (1988).
• 11) Schiff E, Sivan E, Terry S, Dulitzky M, Friedman SA, Mashiach S, Sibai BM. Currently recommended oral regimens for ritodrine tocolysis result in extremely low plasma levels. Am. J. Obstet. Gynecol., 169, 1059–1064 (1993).
• 12) Gandar R, de Zoeten LW, van der Schoot JB. Serum level of ritodrine in man. Eur. J. Clin. Pharmacol., 17, 117–122 (1980).
• 13) Essed GG, Struyker Boudier HA, Van Zijl JA. Biopharmaceutical aspects of ritodrine retard in pregnant women. Arch. Int. Pharmacodyn. Ther., 293, 295–300 (1988).
• 14) Dhiman MK, Yedurkar PD, Sawant KK. Buccal bioadhesive delivery system of 5-fluorouracil: optimization and characterization. Drug Dev. Ind. Pharm., 34, 761–770 (2008).
• 15) Semalty A, Semalty M, Nautiyal U. Formulation and evaluation of mucoadhesive buccal films of enalapril maleate. Indian J. Pharm. Sci., 72, 571–575 (2010).
• 16) Bray CL, Jain S, Faragher EB, Myers A, Myers P, MacIntyre P, Rae A, Goldman M, Alcorn M. A comparison of buccal nitroglycerin and sublingual nitroglycerin in the prophylaxis and treatment of exertional (situation-provoked) angina pectoris. Eur. Heart J., 12, (Suppl. A), 16–20 (1991).
• 17) Schols-Hendriks MW, Lohman JJ, Janknegt R, Korten JJ, Merkus FW, Hooymans PM. Absorption of clonazepam after intranasal and buccal administration. Br. J. Clin. Pharmacol., 39, 449–451 (1995).
• 18) Sakata O, Onishi H, Machida Y. Clonazepam oral droplets for the treatment of acute epileptic seizures. Drug Dev. Ind. Pharm., 34, 1376–1380 (2008).
• 19) Mohamed SP, Muzzammil S, Pramod KT. Preparation of fluconazole buccal tablet and influence of formulation expedients on its properties. Yao Xue Xue Bao, 46, 460–465 (2011).
• 20) Sakata O, Onishi H. In vivo absorption study of ritodrine hydrochloride in the buccal administration to rats. Drug Dev. Ind. Pharm., 39, 1414–1420 (2013).
• 21) Onishi H, Yumoto K, Sakata O. Preparation and evaluation of ritodrine buccal tablets for rational therapeutic use. Int. J. Pharm., 468, 207–213 (2014).
• 22) Agarwal V, Mishra B. Design, development, and biopharmaceutical properties of buccoadhesive compacts of pentazocine. Drug Dev. Ind. Pharm., 25, 701–709 (1999).
• 23) Yamaoka K, Tanigawara Y, Nakagawa T, Uno T. A pharmacokinetic analysis program (multi) for microcomputer. J. Pharmacobiodyn., 4, 879–885 (1981).