(T1530-03-15) Enhancing Oral Absorption of a Poorly Water Soluble Anticancer Drug via Prodrug Modification

Purpose: ML329 is a relatively novel small molecule with the potential to treat various types of cancer via a mechanism of action which selectively kills cancer cells. The drug is limited by its poor aqueous solubility (BCS class II) which makes delivering a reasonable dosage extremely challenging. Prodrug modification of this drug was used to provide improved solubility. Prodrugs have little or no pharmacological activity when administered to the body but are converted to an active parent compound in vivo. As a guiding example Valdecoxib is a nonsteroidal anti-inflammatory drug that selectively inhibits the COX-2 receptor. It contains the same sulfonamide functional group exhibited by ML329. Valdecoxib has been modified via acylation of the sulfonamide functional group to produce Parecoxib, leading to a modification of the pKa of the sulfonamide group, making the molecule ionizable and more soluble at physiological pH. This modification allowed for formulations which were used for parenteral administration of Parecoxib and hence, delivery of Valdecoxib. A practically identical modification on ML329 produced a similar acyl prodrug. The first goal of this work is to determine differences of in vitro properties, such as aqueous solubility and partition coefficient, between ML329 and its prodrug. The second goal of the project is to determine if the modification into a prodrug enables delivery of the parent via oral and IV administration in mice.

Methods: In vitro aqueous solubility was determined by incubation of excess drug in citrate and phosphate buffer systems spanning a physiologically relevant pH range (3.5-8). Partition coefficient over the same range was determined using a traditional shake-flask method with a biphasic system consisting of citrate or phosphate buffer and 1-octanol. Concentration was determined by UV absorbance via injection into a UPLC. In vivo pharmacokinetic studies were conducted in naïve female Balb/c mice between the ages of 2 to 6 months in a fed state. Eight samples were collected over 8 hours with 3 mice per time point. IV administration was conducted via injection of solution into tail vein and oral administration was conducted via oral gavage. In both studies samples were taken via cardiac puncture following deep anesthetization from isoflurane inhalation. For IV administration both drugs were formulated at 0.3 mg/mL in 75/25 Polyethylene glycol 400/Phosphate buffer (pH 7.4 100 mM). Both drugs were formulated at 1.3 mg/mL in suspensions consisting of 0.25% Methocel E4M (w/v) and 0.05% Tween 80 (w/v) in phosphate buffer (pH 7.4 100 mM) for oral administration. A particle size of 1-10 microns was confirmed via SEM prior to administration. Plasma samples were analyzed for both prodrug and parent concentration via LC/MS.

Results: The solubility of ML329 was determined to be approximately 0.25 µg/mL. The solubility of the unionized ML329 acyl prodrug was determined to be approximately 0.35 µg/mL. Solubility of the ML329 acyl prodrug over a physiologically relevant pH range was determined every half pH unit and is displayed in image 1. A pKa for the acid ML329 acyl prodrug based on solubility was estimated to be 5.1±0.3. This is in agreement with the shape of the partition coefficient plot over the same pH range, which is displayed in image 2. The LogP of ML329 and its acyl prodrug were determined to be 1.9 and 2.4, respectively. Plots of plasma concentration of ML329 v. time are displayed in image 3 for both oral and IV dosing. To estimate total drug exposure area under the curve was determined using the trapezoidal method. In IV dosing the area under the curve of ML329 and its acyl prodrug were determined to be approximately 574.2 ng/mL*hr and 175.0 ng/mL*hr respectively. In oral dosing the area under the curve of ML329 and its acyl prodrug were determined to be approximately 811.1 ng/mL*hr and 445.8 ng/mL*hr respectively.

Conclusion: Acylation of the sulfonamide moiety exhibited by ML329 reduced the sulfonamide pKa and improved upon its aqueous solubility at physiological pH. These improvements in physiochemical properties were predictably similar to the improvement observed in the modification of valdecoxib to parecoxib. While the solubility at physiological pH is increased by about 3 orders of magnitude the apparent logP is likely to offset that with a similar magnitude reduction. Hence, the overall absorption was largely unimpacted. Furthermore, the conversion of the ML329 acyl prodrug into its parent appears to be much slower in vivo as compared to our model compound of parecoxib. This leads to a decreased drug exposure to the parent when dosing the acyl prodrug as compared to dosing the parent alone. For this reason, while the acylation of sulfonamide functional groups shows promise for improvement of aqueous solubility in BCS class II drugs, the rates of conversion in vivo will need to be studied more carefully.

Acknowledgements: We would like to acknowledge the Higuchi Biosciences Center and the Jay Inez fund for supporting this work.

ML329 Acyl prodrug aqueous solubility

ML329 Acyl prodrug LogD plot

ML329 acyl prodrug in vivo data