**4. Discussion**

In current study, we prepared oil formulations for oral administration of paclitaxel and examined their physical properties, pharmacokinetic profiles and antitumor activities.

The formulations made with crystalline and amorphous paclitaxel were administered orally to mice (Figure 4F). The pharmacokinetic profiles were virtually identical for these two formulations. We expected that the crystallinity of the drug would not affect the degree of oral absorption since paclitaxel was dissolved completely in tricaprylin/dichloromethane mixture

Oral formulations with (G2, DHP107, Table 3) or without Tween 80 (G7) were prepared. Pharmacokinetic profiles were different for these two formulations (Figure 4G). When the formulation without Tween 80 was administered, Tmax was 1 h instead of 3, and the AUC

Even though our oral paclitaxel formulations including DHP107 can be taken *per se*, they can also be mixed with water or taste-masking syrups for administration. To examine how the absorption of paclitaxel changes upon feeding the dispersion, DHP107 was mixed with water or syrup at 3:7 by weight. DHP107 was dispersed spontaneously in water to produce emulsion droplets having average diameter of 6-8 μm. By vortexing the mixture of DHP107 and water or syrup for 1 min, the diameters of emulsion droplets were reduced to 6 (polydispersity = 1) or 3 (polydispersity = 1) μm, respectively. By sonicating DHP107/water mixture at 180 KW for 1 min, we also obtained the dispersion with the oil droplets having

When the dispersions were orally administered, pharmacokinetic profiles did not change significantly except for the fact that the AUC values were reduced to *ca.* 40 ~ 70 % of the

To evaluate the antitumor activity of DHP107 in mice, suspension of NCI-H358 cells was injected subcutaneously to Balb/c athymic mice. All of the mice inoculated with the human non-small cell lung cancer cells developed progressively growing tumors. The mice were administered orally with DHP107 (G2, 50 mg/kg) or intravenously with Taxol® (10 mg/kg) for 5 consecutive days 10 days after the inoculation. The entire tumor tissue was removed surgically after the experiment and photographed as shown in Figure 5A. In the group administered with oral DHP107 and intravenous Taxol®, the tumor size was reduced gradually from *ca*. 100 to 15 ± 3 and 19 ± 5 mm3, respectively, in *ca*. 25 days after the administration of the drug and remained unchanged for the duration of the experiment (Figure 5B). The tumor size data for these two groups were indifferent statistically. In the control group administered orally

In current study, we prepared oil formulations for oral administration of paclitaxel and examined their physical properties, pharmacokinetic profiles and antitumor activities.

with the vehicle only, the tumor grew continuously to *ca.* 360 ± 20 mm3 in 45 days.

before adding other ingredients, and therefore, the final products would be identical.

value was reduced to 6.3 μg·h /ml, which was *ca*. 56 % of that for DHP107.

**3.10 Oral administration of the aqueous dispersions of DHP107** 

**3.11 Antitumor efficacy of DHP107 in experimental animals** 

**3.8 Crystalline vs. amorphous paclitaxel** 

**3.9 Formulation without Tween 80** 

1.4 μm (polydispersity = 1) in diameter.

original G2 formulation.

**4. Discussion** 

Fig. 5. Antitumor activity of DHP107 in male Balb/c athymic mice, subcutaneously injected with suspension of NCI-H358 cells (1.2×107 cells/mouse). A) The entire tumor tissue removed surgically after the experiment. B) The volume of tumor for the oral vehicle control (eG2, ), intravenous Taxol® (, 10mg/kg) and oral DHP107 (G2, , 50 mg/kg).

Paclitaxel was commercially available in at least two different polymorphs. Amorphous paclitaxel was readily soluble in our oral formulations at 10 mg/ml when sonicated for 30 s. Crystalline paclitaxel, on the other hand, did not dissolve in the oily formulations directly. We had to dissolve crystalline paclitaxel in the tricaprylin/methylene chloride mixture completely, and to remove the solvent in turn to obtain the oily solution of paclitaxel/tricaprylin before adding monoolein and Tween 80. Even though paclitaxel could also be dissolved in methylene chloride/tricaprylin/monoolein/Tween 80 as well as in methylene chloride/tricaprylin, monoolein and Tween 80 were not added to the mixture until after evaporating the solvent to minimize the oxidation of these materials containing unsaturated hydrocarbons. Pharmacokinetic study showed that AUC values were identical statistically when paclitaxel from different vendors or different preparation processes was used (Figure 4F).

DSC study was performed for the formulations with different triglycerides. In the heating thermograms, two endothermic transitions were observed corresponding to the chain melting transitions of monoolein and triglycerides for T8 and T12 formulations. Chain melting transition of triacetin, tributyrin and tricaproin were not observed since the phase transition temperatures of these triglycerides were below 0 C. There was only a single endothermic transition for the tricaprin/monoolein mixture at the weight ratio of 2:1 since

Development, Optimization and Absorption Mechanism of

injection under fasting and non-fasting conditions, respectively.

Taxol® injection, similar regression rate was expected to a degree.

European Patent 1539099A2, 2005

Vol. 6. pp. 4416-4421

DHP107 reaches the upper intestine.

**5. Conclusions** 

**6. References** 

DHP107, Oral Paclitaxel Formulation for Single-Agent Anticancer Therapy 371

in syrup or in water. From these experiments, we could conclude that the optimum concentration of paclitaxel in the formulation was 10 mg/ml, and the tricaprylin/monoolein/Tween 80/paclitaxel formulation was the most effective when given directly to the non-fasting animals. Under the fasting conditions, Tmax was 3 h. Tmax, however, was 1 h when the animal had free access to food and water. In a separate experiment, we observed that the bile salts helped the micronization and micellization of DHP107. It is possible that the micronization process of DHP107 was fastened by the secreted bile salts already existing under the non-fasting condition. When the formulation was fed to the empty stomach, on the other hand, bile salt secretion would start after the oily

Our liquid type formulations were very effective in delivering paclitaxel orally, easy to prepare, biocompatible, and physically stable. Preclinical studies have demonstrated superior antitumor efficacy and high bioavailability. Most notably, the blood paclitaxel concentration was high after the oral administration even though the P-glycoprotein inhibitors were not coadministered when compared to the formulations reported in the literature (Table 1). The results show that the blood paclitaxel concentration reached as high as *ca*. 3 μg/ml and higher than 1 μg/ml for 4~6 hours in Balb/c mice after oral administration of 50 mg/kg of paclitaxel dose. Also the bioavailability of paclitaxel was *ca*. 14 % and 20 % when compared to Taxol®

Antitumor experiment also showed that the tumor regression rate of the oral DHP107 group (50 mg/kg) was similar to that of the intravenous Taxol® group (10 mg/kg). Also, the tumor size of an established non-small cell lung cancer was significantly reduced after the oral treatment. Considering that the bioavailability of DHP107 was *ca.* 14 ~ 20 % compared to

In conclusion, we prepared oral paclitaxel formulations that do not contain P-glycoprotein inhibitors as active pharmaceutical ingredients. The formulations are liquid at body temperature and can solubilize paclitaxel effectively. The oral bioavailability of paclitaxel was 14 ~ 20 % when compared to the intravenous Taxol® formulation without concomitant administration of P-glycoprotein inhibitors. Preclinical efficacy study on mice showed that the tumor size was reduced significantly for the human non-small cell lung carcinoma. In separate studies, we have determined the tissue distribution of paclitaxel after oral administration (manuscript in preparation) and performed pre-clinical antitumor efficacy studies in mice with several tumor types (manuscript in preparation). Regulatory preclinical

experiments to initiate the clinical evaluations of DHP107 have also been carried out.

Anderson, D. (2005) Reversed liquid crystalline phases with non-paraffin hydrophobes.

Bardelmeijer, H. A., Beijnen, J. H., Brouwer, K. R., Rosing, H., Nooijen, W. J., Schellens, J. H.

M., & van Tellingen, O. (2000) Increased oral bioavailability of paclitaxel by GF120918 in mice through selective modulation of P-glycoprotein. *Clin. Cancer Res.,*

they formed a eutectic mixture (Roh et al., 2004). The phase behavior of monoolein and triglycerides could be explained by the classical binary eutectic phase diagram of two immiscible solid phases and a completely miscible melt.

We could conclude from the DSC data that monoolein and tricaprylin do not mix below *ca.* 30 C where monoolein exists as lamellar crystalline phase. Monoolein and Tween 80 did not mix in this temperature range either. Above 30 C, however, all three ingredients, tricaprylin, monoolein and Tween 80 existed as liquid and mixed homogeneously. Two important things to note were that paclitaxel precipitates were not observed even at 0 C by microscopy in any of the oral formulations and that the formulations could be heated and cooled in cycles between -20 and 40 C without compromising the effectiveness of the drug. These results are significant in that the oral formulation (possibly in a soft capsule) can be stored in the shelf and orally administered in the phase-separated form, but transforms into a homogenous liquid solution by the body heat while traveling inside the gastrointestinal tract.

In our oral formulations, monoolein was included due to its well-known mucoadhesive property (Nielsen et al., 1998). Monoolein and other monoglycerides have been used in oral drug delivery formulations since they can enhance absorption of small molecules and even proteins through the epithelial cells (Ganem-Quintanar et al., 2000; Chung et al., 2002). The absorption enhancing mechanism is not clearly known. Nanopore induction or the membrane perturbation (Anderson, 2005), and the intermediate phases formed in the intestine (Kossena et al., 2005) were considered important. *In vitro* study showed that monoglycerides can enhance the cellular absorption of drugs by inhibition of Pglycoproteins (Konishi et al., 2004). Further studies are required to explain the absorption enhancing mechanism of monoolein unequivocally.

In the previous study, we administered intravesically the dispersion of DHP107 in water to experimental rabbits and observed that the formulation adhered tightly to the bladder mucosa, and paclitaxel penetrated through the physical barrier imposed by the uroepithelium (S. J. Lee et al., 2005). Histological examination of the bladder and other tissues did not reveal any local or systemic toxicity to the rabbits.

Oral administration of formulations containing different triglycerides showed that those with tricaproin, tricaprylin and tricaprin had higher AUC values than others. The formulation with tricaproin had the highest AUC value. We must note that paclitaxel was mixed directly with the vehicles and sonicated for 30 s to obtain the formulations in Table 2. When the preparation process was changed to add and to remove methylene chloride in turn in order to solubilize crystalline as well as amorphous paclitaxel, the AUC value of tricaprylin/monoolein/Tween 80/paclitaxel formulation (G2, DHP107) increased from 8.9 to 11.0 μgh/ml, which was similar to that of tricaproin/monoolein/Tween 80/paclitaxel formulation in Table 2 (T6) statistically. We proceeded with the tricaprylin/monoolein/Tween 80/paclitaxel formulation for further experiments because the toxicity of tricaprylin (LD50 = 3700 mg/kg; intravenous and 26600 mg/kg; oral) was much lower than that of tricaproin (LD50 = 122 mg/kg; intravenous) for mice according to the Material Safety Data Sheets for tricaprylin (T9126) and tricaproin (T4137) provided by Sigma (www.sigma-aldrich.com).

We also performed the pharmacokinetic study with the formulations having different contents of the drug, under fasting or non-fasting conditions, by changing the commercial source of paclitaxel, with or without the emulsifier, Tween 80 and in the form of dispersions in syrup or in water. From these experiments, we could conclude that the optimum concentration of paclitaxel in the formulation was 10 mg/ml, and the tricaprylin/monoolein/Tween 80/paclitaxel formulation was the most effective when given directly to the non-fasting animals. Under the fasting conditions, Tmax was 3 h. Tmax, however, was 1 h when the animal had free access to food and water. In a separate experiment, we observed that the bile salts helped the micronization and micellization of DHP107. It is possible that the micronization process of DHP107 was fastened by the secreted bile salts already existing under the non-fasting condition. When the formulation was fed to the empty stomach, on the other hand, bile salt secretion would start after the oily DHP107 reaches the upper intestine.

Our liquid type formulations were very effective in delivering paclitaxel orally, easy to prepare, biocompatible, and physically stable. Preclinical studies have demonstrated superior antitumor efficacy and high bioavailability. Most notably, the blood paclitaxel concentration was high after the oral administration even though the P-glycoprotein inhibitors were not coadministered when compared to the formulations reported in the literature (Table 1). The results show that the blood paclitaxel concentration reached as high as *ca*. 3 μg/ml and higher than 1 μg/ml for 4~6 hours in Balb/c mice after oral administration of 50 mg/kg of paclitaxel dose. Also the bioavailability of paclitaxel was *ca*. 14 % and 20 % when compared to Taxol® injection under fasting and non-fasting conditions, respectively.

Antitumor experiment also showed that the tumor regression rate of the oral DHP107 group (50 mg/kg) was similar to that of the intravenous Taxol® group (10 mg/kg). Also, the tumor size of an established non-small cell lung cancer was significantly reduced after the oral treatment. Considering that the bioavailability of DHP107 was *ca.* 14 ~ 20 % compared to Taxol® injection, similar regression rate was expected to a degree.
