6 There are several other phenotypes that will be discussed in the chapter,

particularly those that deal with drug targets or pharmacodynamic responses.

The vast majority of drug metabolism pathways are not simply, a single enzyme

resulting in one inactive metabolite that is immediately excreted by the body. Instead,

most drugs utilize multiple metabolizing enzyme pathways, rely on several drug

transporters, and have many metabolites of varying activity. Each of these steps may

be subject to variable gene expression, coupled with indirect genetic and

environmental effects on variables such as disease, weight, nutrition, age, and others.

Add this complexity and it becomes evident how complicated pharmacogenomic

interpretations can be.

Pharmacogenomics is unique in the field of genetics in that the representation of a

person’s genotype is noted by a “*,” known as a star allele in the literature. For

example, in many situations, the text representation of normal or “wild-type” variant

status is *1/*1.

7 As other alleles are discovered, they are numbered sequentially *2,

*3, etc. Inevitably, as discoveries are sometimes reported simultaneously around the

world, there is occasionally an overlap of star alleles and the variants represented.

There are several worldwide databases for reporting new pharmacogenomic

variants, including the CYP Allele Nomenclature Database

http://www.cypalleles.ki.se/ and the Database of Genomic Variants

http://dgv.tcag.ca/dgv/app/home.

During drug development and postmarketing analysis, it becomes clear that some

patients will have severe adverse drug reactions (including nonresponse) after

receiving the population-derived standard dose. These adverse reactions occur in the

absence of medication error or lack of proper adherence to the drug. The Institute of

Medicine “To Err is Human” report published in 2000 cited over 2 million reported

adverse drug reactions in the United States per year, resulting in approximately

100,000 deaths at a cost of 20.6 billion annually.

8

It is estimated that 52% of U.S.

adults are taking at least one prescription drug and 12% are taking more than five

prescription drugs. If herbal products and over-the-counter drugs are included, then

the estimates increase to over 80% and 29%, respectively.

6

It is well known that

adverse drug reactions often go unnoticed and even more frequently unreported,

making the actual numbers likely far higher. The clinical application of

pharmacogenomics not only has allowed us to explain some of these historical

reactions, but also helps clinicians predict who may be at higher risk for developing

an adverse reaction if exposed to a drug or drug class.

9 Avoidance of adverse drug

reactions has a measureable impact on a patient’s quality of life, decreases overall

health care costs, and lessens the time burden on health care providers managing the

adverse reaction. However, acceptance of clinical pharmacogenomic testing is not

universal and there remains the challenge of conclusively validating the genetic

variant associated with the adverse reaction.

10

Figure 4-1 Example of a single-nucleotide polymorphism (SNP).

p. 52

p. 53

PHARMACOKINETIC IMPLICATIONS

CASE 4-1

QUESTION 1: A.S. is a 2-year-old diagnosed with active multidrug-resistant tuberculosis (TB). The Centers

for Disease Control and Prevention treatment protocol recommends isoniazid, rifampin, ethambutol, and

pyrazinamide for initial therapy. Her mother is concerned because several family members in their home

country of China had developed serious “liver problems” when they took isoniazid and she insists on the “gene

test” for A.S. What genetic testing is A.S.’s mother referring to and what are the risks of toxicity in relation to

isoniazid therapy?

Isoniazid is part of the four-drug regimen used to treat active TB and latent TB

infection because it is bactericidal against Mycobacterium tuberculosis organisms.

11

One of the most commonly reported side effects leading to premature discontinuation

of the drug is drug-induced liver injury (DILI). Significantly elevated hepatic

enzymes are seen in up to 20% of patients, with progression to hepatitis in a smaller

percentage of those treated.

12

The metabolism of isoniazid is complex, with a key enzyme within the pathway

being N-acetyltransferase-2 (NAT2).

11 NAT2 acetylates both the parent drug and a

hydrazine metabolite, converting the hydrazine metabolite to acetyl hydrazine

(AcHz).

12An alternative metabolic pathway for hydrazine results in a toxic reactive

metabolite. When NAT2 function is decreased, this pathway dominates, resulting in

increased DILI and toxicity.

The production of the NAT2 enzyme is generated from a gene of the same name,

with several known polymorphisms leading to alleles associated with either slow or

rapid acetylation rates. Individuals homozygous for alleles associated with loss of

function are considered “slow acetylators,” those homozygous for alleles associated

with gain of function “rapid acetylators,” and heterozygotes “intermediate

acetylators” (Table 4-1). DILI has been found to be highest in patients who are slow

acetylators.

12

In addition to isoniazid, NAT2 is also a known acetylator of other

drugs and/or metabolites, although the clinical effects of NAT2 genotype on these

drugs are less well-studied, including: sulfonamides (sulfamethoxazole, metabolites

of sulfasalazine), aromatic and aliphatic amines (procainamide, dapsone, metabolite

of clonazepam, mescaline).

12

The “gene test” mentioned by A.S.’s mother is likely a test for the NAT2 gene. A

variety of pharmacogenomic tests are available to examine SNPs in the NAT2 gene

and assign a patient’s genotype, either as part of a more extensive gene panel or as a

targeted assay. Preemptive NAT2 testing prior to isoniazid therapy is not currently

the standard of care; however, evidence suggests that in some populations with a high

prevalence of DILI in slow acetylators, this may be beneficial.

14–18 Asian cohorts

have been found to have the highest rates of DILI in the setting of a slow acetylator

status (see summary of studies in Table 4-2). Although there are no formal dosing

guidelines set for isoniazid therapy in the context of genotype or acetylator status, a

study completed by Azuma et al. with 155 Japanese TB patients proposed a

modified, genotype-based dosing regimen with successful clinically and statistically

significant outcomes (results summarized in Table 4-3).

19

There remain several pending questions as to how to most safely and appropriately

incorporate NAT2 genotype results into a widespread clinical application. For

example, pediatric patients such as A.S. offer a unique set of challenges, because

recommended isoniazid starting doses are within a higher range (10–15 mg/kg), and

dose modifications based on genotype have not been studied in this population.