Pharmaceutical Industry and FDA Perspective

According to the FDA, severe hepatotoxicity is one of the most common causes for pharmaceutical product recalls and labeling changes, and this raises the question of how effective nonclinical and clinical testing are in recognizing such toxicity. Although much attention has been focused on the predictivity of animal models for clinical findings in humans, clinical trials of pharmaceuticals are in fact a relatively poor source of information for ascertaining the probability that an event observed in an animal model is also observed in the human. Of critical importance to avoiding unanticipated toxicity in human is the need to understand why drugs that were judged safe to administer to humans on the basis of animal data sometimes cause unexpected toxicity or drug interactions. In other words, why do animal models in these cases fail to predict human hepatotoxicity? It is probably even more important to understand why these models sometimes fail to identify hepatotoxic potential in humans than to fully define the overall effectiveness of animal studies in predicting human outcome.

Drug-drug interactions are one of the major concerns for clinical practice and the pharmaceutical industry when several drugs are co-administered or during the development of a new drug. During the last 15 to 20 years drug interactions have received increasing attention because a large number of new drugs have been introduced into clinical practice for the treatment of several diseases, as cancer, HIV, and other opportunistic infections continue to involve multi-drug therapy. The majority of drug-drug interactions of clinical significance occur through induction or inhibition of cytochrome P450 enzymes. According to the new FDA draft guidance, a drug that induces a drug-metabolizing enzyme can increase the rate of metabolic clearance of a co-administered drug that is a substrate of the induced pathway [86]. A potential consequence of this type of drug-drug interaction is subtherapeutic blood concentrations. Drug-drug interactions due to induction of drug-metabolizing enzymes have greater clinical significance for those drugs that have a narrow therapeutic window, such as phenytoin, warfarin, and digoxin. Alternatively, the induced metabolic pathway could lead to increased formation of an active compound, resulting in an adverse event.

The classic example of this phenomenon is the bioactivation of acetaminophen to its reactive metabolite N-acetyl-benzoquinoneimine (NAPQI), which is a cytochrome P450-dependent oxidation step. Excessive formation of NAPQI in the liver can cause depletion of glutathione and may result in cell death and hepatotoxicity due to covalent binding to essential cellular macro-molecules and/or other mechanisms such as oxidative stress. It has been shown in animal models that activators of PXR (PCN) or CAR (PB) can enhance the hepatotoxicity of APAP by the increased formation of NAPQI [87,88]. Notably, increased liver damage is not observed in CAR- or PXR-knockout animals, suggesting that receptor activation and enzyme induction play critical roles in the hepatotoxicity of APAP.

In humans, several hormone nuclear receptors that are activated by many drugs and steroids play major roles in the induced expression of CYPs and other drug-metabolizing enzymes that are involved in APAP elimination and toxicity. For example, PXR plays a key role in the regulation of CYP3A4, CYP2C9, CYP2B6, and UGT1A1 as well as transporters such as MDR1. Likewise CAR plays a key role in the regulation of the same enzymes and transport proteins; however, one important difference is that it does not up-regulate CYP3A genes as effectively as PXR [77]. Therefore, unlike in the rodent model, co-administration of APAP with ligands for CAR or PXR might lead to differential toxic side effects and liver damage in humans.

Chronic activation of PXR has been associated with disruption of homeo-stasis of endogenous substrates as well as other adverse physiological effects. For example, rifampicin is one of the most potent human PXR activators and inducers of CYP3A4, CYP2C9, and CYP2B6. Rifampicin can significantly impair the efficacy of a number of drugs that are substrates for these enzymes, such as tamoxifen, ifosfamide, paclitaxel, ethinyl estradiiol, warfarin, and cyclosporine A. In a double-blind crossover study, administration of rifam-picin caused a 96% reduction in the AUC of midazolam in 10 healthy volunteers [89]. Terzolo et al. [90] reported that long-term treatment of tuberculosis by rifampicin could increase steroid clearance and lead to a misdiagnosis of Cushing's syndrome by interfering with the overnight dexamethasone test.

St. John's wort, a herbal remedy and a widely used antidepressant, has not been subjected to the rigorous clinical testing that most drug candidates currently in development receive. A number of case reports have demonstrated drug-drug interactions between St. John's wort and various drugs that are CYP3A4 substrates [91,92]. In women St. John's wort increased the clearance of oral contraceptives, which led to decreased circulating sex steroid levels and the loss of contraceptive efficacy [93]. Co-administration of St. John's wort also reduces the blood levels of HIV protease inhibitors and immunosuppressant drugs [94,95]. Recently Kliewer and colleagues demonstrated that these St. John's wort related drug interactions are due to activation of human PXR and subsequent induction of CYP3A4 expression [96]. Another example is paclitaxel, which is a widely used antineoplastic agent, an efficient human PXR activator, and a substrate of CYP3A4 and CYP2C8. The therapeutic efficacy of this drug might be limited by autoinduced metabolism. A paclitaxel analogue that has similar antineoplastic activity but does not activate PXR has been shown to have superior pharmacokinetic properties [7].

The co-administration of activators of individual or multiple nuclear receptors complicates further the prediction of drug interactions and toxicity in vivo. From the evidence thus far, it is becoming apparent that NR's have a primary function in the regulation of specific target genes, but they also appear to play a secondary or supporting role in the regulation of others. As such, there is much to be learned regarding the overlapping specificity of ligands and DNA response elements for individual receptors, as well as the manner in which they interact to affect gene expression in a synergistic or antagonistic fashion. Just as important, it is now understood that the molecular basis for the species differences observed in the xenobiotic-dependent induction of CYP3A genes can be simple mutations in the amino acid sequence of the ligand-binding domain of the receptor. Current induction predictions based on animal data are deficient inherently, and therefore new screening strategies combining reporter assays with expressed human receptors and primary cells of human origin are gaining widespread acceptance in the drug discovery and development process.

Although the role of PXR in the regulation of CYP3A4 and MDR1 has been firmly established, there are a number of unresolved issues remaining about the importance of other orphan receptors in drug-induced gene expression. Recent evidence suggests that PXR, CAR, and GR are the primary players involved in the regulation of a number of other phase I and II enzymes and transporters, whereas PXR appears to be the "master" regulator of cytochrome P450 expression in human liver. The role of the nuclear receptor CAR is ambiguous at this point in time, but it is certain that it is quite distinct from that of its rodent counterparts. The ability to identify the significance of CAR in human liver has been confounded historically by the lack of good humanrelevant in vitro model systems. However, recent breakthroughs in the development of cell-based technologies and receptor-binding assays will enable investigators to determine whether CAR plays a dominant or supporting role in the drug-induced expression of hepatic enzymes and transporters.

The recent advances in orphan nuclear receptor biology have further expanded our ability to predict the potential pharmacological and toxicologi-cal properties of new drugs. Indeed great progress has been achieved in developing high-throughput screening of new chemical entities based on in vitro cell-based transfection assays for those likely to be involved in drug interactions with CYP3A4 substrates. More recently cellular-based assays have been employed to identify nuclear receptor agonists and antagonists as potential drug candidates. It is anticipated that these novel methods may lead to a more rational and molecular-based approach to developing drugs with enhanced therapeutic efficacy and improved safety profiles.

Finally, species differences in the induction of individual or multiple biotransformation and elimination pathways can lead to the production of different metabolite profiles in humans compared to animal models. The FDA

considers that the quantitative and qualitative differences in metabolite profiles are important when comparing exposure and safety of a drug in a non-clinical species relative to humans during risk assessment. When the metabolic profile of a parent drug is similar qualitatively and quantitatively across species, it is generally assumed that potential clinical risks of the parent drug and its metabolites have been adequately characterized during standard nonclinical safety evaluations. However, because metabolic profiles and metabolite concentrations can vary across species, there may be cases when clinically relevant metabolites have not been identified or adequately evaluated during nonclini-cal safety studies. This may occur because the metabolite(s) being formed in humans are absent in the animal test species (unique human metabolite) or because the metabolite is present at much higher levels in humans (major metabolite) than in the species used during standard toxicity testing. Therefore understanding the species differences in the metabolic pathways involved in the clearance and toxicity of drugs, as well as how they are regulated by nuclear receptors, will continue to be a major challenge in the future.

0 0

Post a comment