The colon is an attractive region for local therapy and drug absorption. Colon-specific drug release ensures an increased local drug concentration enabling direct tumor treatment and reduced systemic adverse effects. Colonic diseases range in severity from irritable bowel syndrome and inflammatory bowel disease (IBD), such as ulcerative colitis and Crohn's disease, to colon cancer. Compared with the stomach and small intestine, the colon contains lower levels of luminal and mucosal digestive enzymes (16
). Molecules, such as peptides and proteins that are degraded in the harsh hydrolytic environment of the upper gut, can therefore be absorbed via the less proteolytically active colon (35
). Moreover, colonic drug residence time is relatively long due to a decreased transit rate. These characteristics reduce systemic drug absorption from the colon and the associated adverse effects and improve local therapy due to longer drug exposure at the disease site.
There have been numerous attempts to use the above-described physiological rationale to design delivery systems for the colon-specific drug delivery. The enzymatic activity of colonic bacteria is one of the major targets in the tailor-made synthesis of low molecular weight as well as polymeric colon-specific prodrugs. Dexamethasone- and prednisolone-β-D
-glucosides have been evaluated as prodrugs capable of releasing the active agent due to high glucosidase activity in the colon (36
). Dextranase activity in the colon has been utilized for the release of naproxen from dextran carriers (37
). Due to the high level of relevant enzymatic activity in the colon, natural polysaccharides are being used in the development of solid dosage forms for colon-specific delivery (38
). The potential of the azoreductase activity for colon-specific drug delivery has been also evaluated (13
). Prodrugs, e.g., 5-[4-(2-pyridylaminosulfonyl)phenylazo]-salicylic acid (29
) and 5,5′-azodisalicylic acid (30
) are used clinically. The use of water-soluble copolymers for the colon-specific oral delivery of 5-ASA has been proposed by Brown et al.
), and crosslinked (branched) copolymers for the colonic delivery of proteins were introduced by Saffran et al.
Targeted colon-specific drug delivery systems have been developed based on the observation that cancerous colonic tissue expresses altered glycoproteins. The Thomsen–Friedenreich (TF) antigen (galactose-β1,3-N
-acetylgalactosamine) is masked in healthy tissue by oligosaccharide extension and/or sialylation. During colonic disease states (IBD or cancer) alterations in cell surface carbohydrate expression occur resulting in TF antigen exposure. Lectins, proteins that bind with specificity to carbohydrate moieties, may distinguish healthy and diseased tissue. Peanut agglutinin (PNA) was shown to bind well to the TF antigen and to diseased colon tissue both in free and HPMA copolymer-bound forms (40
). Detailed evaluations of HPMA copolymer-lectin-drug conjugates indicates that they may improve safety and efficacy of therapy for IBD and Barrett's esophagus (40
The HPMA copolymer–9-AC conjugate evaluated in this study () contains an aromatic azo bond combined with a 1,6-elimination 4-aminobenzylcarbamate group. The synthetic chemistry used in the design of this delivery system permits for binding of lectins as targeting moieties (42
). However, in this study we evaluated the nontargeted derivative. Due to the presence of the aromatic azo bond, it is colon specific, and cleavage experiments have verified conjugate stability in simulated upper GI tract conditions (13
There have been several attempts to predict the absorption behavior of colon-specific prodrugs. Using the gastrointestinal-transit-absorption (GITA) model (44
), Yokoe et al.
correctly predicted the absorption behavior of salicylazosulfanilic acid, a prodrug of 5-ASA (21
). We have previously evaluated the biodistribution and pharmacokinetics of the HPMA copolymer–9-AC conjugate in mice (14
). Colon-specific release produced high local concentrations (14
); which may result in enhanced therapeutic efficacy (45
In this work a pharmacokinetic model () was developed to quantitatively describe drug release and disposition in rats after oral administration of the colon-specific HPMA copolymer 9-AC conjugate (). Various pharmacokinetic parameters were applied to mathematically simulate the processes of drug transit, degradation, absorption, distribution, and elimination. The transit and elimination rate constants were assumed to be the same as those of phenol red, a non-absorbable marker, in male Wistar rats (32
). The degradation and absorption rate constants were experimentally determined. As described previously (14
), the degradation rate of the HPMA copolymer 9-AC conjugate was similar to that of small molecules, such as methyl orange. This finding was attributed to the fact that only one or two drugs were attached to a polymer chain, which facilitates side chain cleavage without formation of micelles (46
). The fact that a similar degradation rate constant as for salicylazosulfanilic acid (SASA; 34.5 h−1
) was observed supports this conclusion. We assumed that the degradation rate was proportional to the amount of cecal contents in the incubation mixture. Therefore, the degradation rate constant for in vivo
conditions was obtained by multiplying the in vitro
degradation constant, obtained using diluted cecum contents (), with 6.7 (100/15). A similar approach was used by Yokoe et al.
). The value of the absorption rate constant of 9-AC in the cecum was approximately two-fold higher than that of 5-ASA (0.061 h−1
), because 9-AC is more lipophilic than 5-ASA, and may diffuse more easily across lipid bilayers. The distribution difference between the tissue compartment (peripheral compartment) and central compartment (systemic circulation) is reflected in the k12
ratio. According to our experimental data, the k12
ratio is almost equal to one, which means that the drug deposition in tissue compartment is the same as in the blood. In other words, no drug accumulation occurred in tissue compartment, because the elimination of the drug from the tissue compartment was equal to that of the central compartment.
Pharmacokinetic profiles were notably improved by the oral dosing of colon-specific HPMA copolymer 9-AC compared to free 9-AC (14
). In the polymer group, the expected and observed release of 9-AC in rats were similar to those in mice. Previously, we reported that the Cmax
was lower, and mean residence time (MRT) was longer in the polymer group in mice when compared to free drug group (14
). The lower Cmax
, attributed to a reduced absorption rate for released 9-AC in the cecum and slow transport into systemic circulation may provide a potential safety benefit, since high 9-AC plasma levels have been correlated with increased side effects in clinical trials (47
). The prolonged MRT caused by a long absorption time in the cecum and a slower transit time may result in enhanced antitumor effects. These properties are particularly advantageous, since 9-AC has a cell cycle dependent mechanism.
Colon-specific drug delivery increased the local drug concentration in cecum. As shown in , the predicted amount of 9-AC released in the cecum increased gradually, reached a peak point at approximately 6 h after oral administration (about 60% of total oral dose released), and then decreased. Even 24 h after oral dosing, 7% of the total 9-AC dose was retained in the cecum. Because of the higher transit rate (0.46 h−1
) from SI to cecum, higher degradation rate (30 h−1
) compared to absorption rate (0.13 h−1
); and, miniscule elimination rate (0.00001 h−1
), released 9-AC accumulated in the cecum before reaching its peak concentration. When approaching the peak 9-AC concentration, almost all of the polymer conjugate had reached the cecum and degraded there. With no further conjugate supply, the (released) 9-AC in the cecum was absorbed and continuously eliminated, resulting in a gradual reduction in 9-AC concentration. These results agree well with our previous data (14
), where we reported that in mice the released 9-AC concentration in cecal contents and tissues reached its peak point about 5 h after oral dosing. The mean peak concentrations of released 9-AC in cecal contents and tissues were, respectively, 3.2-fold and 2.2-fold higher than those from oral dosed free 9-AC. These data on site-specific drug accumulation are of importance for local therapy of GI diseases.
Following oral administration of HPMA copolymer 9-AC conjugate, the released 9-AC was metabolized by the cecum and liver prior to reaching the general circulation. The first-pass effects of released 9-AC were quantitatively determined by liver perfusion and in-situ
absorption studies. The following assumptions were used to analyze liver metabolism. It was assumed that the liver is a single well-stirred compartment and that the concentration of unbound drug in the emergent blood is in equilibrium with the unbound drug within the liver. Additionally, due to assumed passive diffusion with rapid equilibrium, the concentrations of unbound drug in venous blood (COut,u
) and in liver (CL,u
) were considered equal. Furthermore, it was assumed that only unbound drug could traverse membranes, and the rate of drug elimination would be a function of the concentration of unbound drug. Based on the well-stirred model assumptions, the hepatic intrinsic clearance (CLh,int
) was calculated at steady state of liver perfusion. The hepatic and cecal availabilities were calculated using experimentally determined CLh,int
and assuming a rat portal blood flow of 9.8 ml/min. The hepatic extraction ratio was smaller than the fraction of unbound drugs, which was attributed to the liver's restrictive clearance allowing only unbound drug to be metabolized at the site of oxidative enzymes and exclusion of protein-bound drug, which does not diffuse through cell membranes of hepatocytes (48
The pharmacokinetic model developed here may both quantitate drug disposition in animals, and additionally predict drug concentrations in humans. Drug disposition information in the colon and systemic circulation are important for understanding not only drug transport, but elucidation of pharmacokinetic-pharmacodynamic relationships that might aid in drug development process. This pharmacokinetic model may provide an assessment of drug exposure in colon tumors where the local drug concentration is most relevant to therapy. By setting a minimum effective 9-AC concentration in colon tumors, in this case 10 nM of 9-AC in active lactone form, the duration of effective tumor concentration can be determined after a single dose. A reliance on pharmacokinetic properties in colon tumors, and preferably, colon tumor-based pharmacokinetic–pharmacodynamic relationships should provide rational means to select the most efficacious drug-dosing regimen. This pharmacokinetic model established for 9-AC disposition in rats can be scaled to predict drug concentrations in humans to aid drug development process and to foster the rational design of drug administration regimens in humans. The human scaling procedure can be accomplished by substitution of a human forcing function that describes the plasma drug concentration-time profile, or by using the human GI transit data monitored by a gamma scintigraphic technique. It has been reported for three drugs, carboplatin, topotecan, and temozolomide, that the hybrid models, originally derived from preclinical data and then scaled to humans, agreed well with the observed data through the use of human forcing functions (50