Hepatotoxicity of Dermatologic Drug Therapy

Hepatic Drug Metabolism

The liver serves as the major site of metabolism for most drugs. Q62.3 The overall goal of this hepatic drug metabolism is to convert pharmacologically active, relatively lipophilic drugs, into inactive, relatively hydrophilic metabolites, to facilitate renal or biliary excretion. With these important functions, it should come as no surprise that the liver could be the ‘victim’ of aberrances in these metabolic processes. Furthermore, the hepatocyte is the hepatic cell of central importance in this drug metabolism; it follows that the hepatocyte is the most common liver cell ‘victimized’ by DILI. Important general concepts regarding hepatic drug metabolic systems are listed in Box 62.1 .

Q62.3 The human body has a complicated and intriguing system for biotransformation (conversion from lipophilic to hydrophilic molecules) and detoxification (conversion of reactive metabolic intermediates to more stable molecular compounds) of a multitude of very heterogeneous pharmaceutical molecules/drugs. These systems for biotransformation and detoxification of drugs logically evolved and developed long before all current drugs were developed. The same systems play an important role in the metabolism of a number of molecules (including steroid molecules, such as corticosteroids (CS), sex steroids, bile acids) and detoxification of various potential environmental insults (such as by ‘quenching’ free oxygen and superoxide radicals).

Increased activity of the biotransformation enzymes (because of cytochrome P-450 [CYP] inducers) can produce an increased quantity of reactive, electron-scavenging metabolic intermediates. The process of converting a pharmacologically active lipophilic drug to a reactive intermediate is known as bioactivation . Likewise, abnormalities of the detoxification systems may allow these reactive intermediates to wreak havoc locally in the liver, or to create problems through indirect mechanisms with important systemic consequences. As long as there is a reasonable balance between bioactivation and detoxification, no significant drug-induced liver toxicity occurs. Spielberg and associates developed the lymphocyte cytotoxicity assay as a potentially important laboratory method of assessing an individual’s detoxification capacity, for a limited number of drugs. The reader is encouraged to learn more about these hepatic metabolic systems from several excellent reviews.

Polymorphisms

Q62.4 The question regarding the varying probability of specific individuals experiencing DILI can in part be answered by the presence of genetic polymorphisms in important hepatic biotransformation and detoxification enzymes ( Table 62.2 ). The tremendous variability of enzyme activity is explained in part by distinct heritable deoxyribonucleic acid (DNA) mutations (polymorphisms) of both phase I and phase II hepatic enzymes. There are five CYP isoforms primarily involved in the pathogenesis of drug-induced hepatotoxicity (CYP1A2, 2C9, 2D6, 2E1, 3A4). CYP2C19 is not known to be responsible for producing reactive metabolites with hepatotoxic potential.

The prototype enzymes with the most significant polymorphisms are CYP2D6, CYP2C9, and CYP2C19. Individuals with ‘low activity’ (‘poor metabolizers’) for CYP2D6 may have at least 50 times less enzyme activity than individuals with the ‘high-activity’ (‘extensive metabolizers’ or ‘ultrarapid metabolizers’) phenotype for the same CYP isoform. Such wide variation in activity can help explain why some individuals experience excessive and prolonged daytime sedation following a single 10-mg dose of doxepin (a CYP2D6 substrate) taken at bedtime, whereas other individuals tolerate and have very little sedation with doxepin doses titrated up to 200 to 300 mg each night. By the same logic, other potentially hepatotoxic drugs that are substrates for CYP2D6 could have excessive drug levels in patients who inherited the low-activity phenotype. Even without genetic polymorphisms, other enzymes (such as CYP3A4) can have significant variability (up to 30-fold) of enzyme activity. There are rare allelic variation (polymorphism) in CYP3A4, but are too uncommon to be of clinical signicance.

Even more important to the topic of drug hepatotoxicity is the potential for CYP enzyme inducers to increase the enzyme activity of various CYP isoforms. Such increased activity of a given phase I oxidative CYP isoform will increase the quantities of various reactive intermediates derived from drug substrates. The risk of DILI is greatest when there is an accompanying genetic (such as N -acetyl transferase 2 (NAT ₂ ) ‘slow acetylators’) or acquired (decreased glutathione levels caused by human immunodeficiency virus [HIV] infection) defect in a phase II conjugation and/or detoxification system.

Significant controversies and unanswered questions abound in this subject area. One such issue is the presence of the slow acetylator phenotype in up to 50% of Caucasians, whereas only a very small percentage of patients with this phenotype experience reactions thought to be caused by this polymorphism (such as drug-induced lupus erythematosus and isoniazid hepatotoxicity). DRESS has been postulated to be caused by genetically based reduced activity of epoxide hydrolase, leading to increased quantities of unstable epoxide intermediates from anticonvulsants, such as phenytoin, phenobarbital, and carbamazepine. This explanation has since been challenged.

General Mechanisms Involved

The most well-accepted hypothesis explaining idiosyncratic DILI has two components. First, there must be the presence of reactive metabolic intermediates, which are electrophilic compounds that rapidly form covalent bonds to various cellular components, if not equally rapidly detoxified. These covalent bonds result in various structural and functional changes, depending on the affected molecule ( Table 62.3 ). These molecular components (with which the reactive intermediates bind) include various proteins (including the CYP isoforms), DNA, and lipids, with consequences, as detailed in Table 62.4 . The presence of CYP inducers can significantly increase the quantities of these electrophilic reactive intermediates ( Table 62.5 ). Second, there generally must be a defect in the cellular detoxification systems for there to be an adequate presence or persistence of these reactive intermediates, to result in significant hepatic pathology. In the absence of the CYP inducers and/or with normal detoxification systems, significant DILI from these metabolic products is distinctly uncommon.

Probably the most important component of the detoxification enzyme repertoire is the glutathione and glutathione S -transferase system (GST). GST is an enzyme with a genetic polymorphism, with glutathione serving as a cofactor and various electrophilic drug metabolites serving as GST substrates. Glutathione can be depleted in the process of electron scavenging, particularly with an unusually heavy load of reactive intermediates, such as with an acetaminophen overdose. The electron scavenging capabilities of this GST/glutathione system can be further depleted by excessive ethanol consumption, HIV infection, and malnutrition. It stands to reason that drug hepatotoxicity, let alone drug reactions in general, will be increased in these clinical settings.

Q62.5 Two potential outcomes are thought to occur when there are excessive quantities of reactive, electrophilic intermediates. First, is the potential for direct toxicity to neighboring cells (most notably the hepatocyte) from protein, DNA, and lipid biomembrane alterations resulting from covalent binding to these reactive intermediates. An important final common pathway leading to cell necrosis or apoptosis is likely a disruption in calcium regulation, leading to alterations in the actin cytoskeleton and changes in cell membrane integrity. Binding of fas (on the cell membrane of ‘damaged’ hepatocytes) with fas ligand (on antigen-processing cells) leads to activation of the caspase cascade, and subsequent apoptosis, is an important pathway of cell destruction as well. This pathway may also be activated through the tumor necrosis factor (TNF)-α receptor. Second, the covalent binding of the reactive intermediates to various proteins leads to the development of new antigens (neoantigens). These neoantigens can include various CYP isoforms that logically are ‘in the neighborhood’, when the reactive intermediates are created. It is suggested that these reactive intermediates cause hepatocyte stress or death leading to release of Damage Associated Molecular Patterns (DAMP), which in turn activate innate immune cells, particularly Kupffer cells in the liver. The activated innate immune cells then release cytokines and chemokines that draw inflammatory cells into the liver, a process thought to be a prerequisite for a targeted adaptive immune attack on the liver. The end results are potentially either a direct humoral or cellular response to these neoantigens or an immunologic cross-reactivity, with structurally similar distant antigens, or both. The complexities of major histocompatibility complex (MHC) molecules involved with antigen presentation, along with the uncertainties, regarding the overall pathogenesis of autoimmune disorders, are beyond the scope of this chapter. Also, given the paucity of dermatologic drugs that induce cholestasis, the mechanisms involved with this process are not discussed in this chapter (see Erlinger and Church articles in the Bibliography).

Toxic Versus Idiosyncratic Reactions

It is worth reminding the reader of an important distinction at this point. Direct drug-induced toxicity to the liver results from excessive drug levels from very few specific drugs. The classic biochemistry example is the effect of carbon tetrachloride on the liver. The classic pharmacologic example is an acetaminophen overdose, although pharmacologic doses of acetaminophen in the presence of a CYP2E1 inducer (chronic excessive ethanol consumption) can produce a similar result. By definition, toxicity here means that virtually every individual will have a reaction to carbon tetrachloride or acetaminophen, if excessive quantities of these molecules are administered. In contrast, the term idiosyncratic suggests that only a distinct minority of individuals will experience DILI from selected drugs, and that the reaction will occur independent of drug dose. Under this heading is where the vast majority of hepatotoxicity caused by drugs occurs. What may be somewhat confusing is that the word toxicity is commonly used in the setting of idiosyncrasy as well. Idiosyncrasy can be divided into ‘metabolic’ idiosyncrasy (because of the local toxic effects of locally formed reactive intermediates) and ‘immunologic’ idiosyncrasy (local damage from an immunologic response, caused by neoantigens formed in response to these reactive intermediates). It is with these two dichotomies clearly in mind (‘toxic’ vs. ‘idiosyncratic’ reactions; ‘metabolic idiosyncrasy’ vs. ‘immunologic idiosyncrasy’) that the reader should continue through this chapter ( Q62.6 Box 62.2 ).