II-37Invited lectures – abstracts
PHENOTYPE-GENOTYPE RELATIONSHIPS OF
DRUG-METABOLIZING ENZYMES IN HUMAN LIVER
Mike Ufer
Institute of Pharmacology, University Hospital Schleswig-Holstein, Kiel, Germany;
mufer@pharmakologie.uni-kiel.de
INVITED LECTURES – ABSTRACTS
VABLJENA PREDAVANJA – POVZETKI
Most drugs are eliminated from the body by hepatic
biotransformation that is catalyzed by Phase I and II
metabolizing enzymes. The expression levels and ca-
talytic activity of these enzymes is predictive for the
drug clearance and hence for the plasma concentra-
tion obtained following intake of a certain dose. Phase
I metabolic reactions are primarily oxidative process-
es catalyzed by cytochrome P450 (CYP) enzymes,
while Phase II metabolism refers to the conjugation
of the parent compound or more commonly its Phase
I metabolite with glucuronic acid, glutathione or sul-
phate mediated by various transferase enzymes. There
are two principal methods to predict the activity of
drug-metabolizing enzymes in human liver termed
phenotyping and genotyping.
Phenotyping is usually done by oral administration
of a certain probe drug that is selectively metabolised
by the enzyme under study. Subsequently, the plas-
ma or urine concentration of probe drug and gener-
ated metabolite is measured and the so-called meta-
bolic ratio of drug vs. metabolite is calculated. This
ratio is inversely correlated with the current real-time
enzyme activity and allows for the assignment of dif-
ferent phenotypes termed ultrarapid (UM), extensive
(EM), intermediate (IM) and poor metabolizer (PM).
More recently, a number of different cocktail ap-
proaches for CYP phenotyping have been proposed
dealing with the simultaneous administration of mul-
tiple, CYP-specific probe drugs. An alternative ap-
proach comprises the intravenous administration of
a radioactively-labelled probe drug with subsequent
quantification of 14CO2 in exhaled air. However, this
approach has so far only been established for CYP3A4
phenotyping as so-called erythromycin breath test.
Genotyping of drug-metabolizing enzymes has been
greatly facilitated in recent years by the implementa-
tion of high-throughput technologies (e.g. real-time
PCR, pyrosequencing, MALDI-TOF mass-spectrome-
try etc.). Completely inactive P450 alleles have been
reported for CYP2D6, 2C19, 2A6 and 3A5 due to a
lack of enzyme expression or expression of an inac-
tive enzyme thereby leading to a PM phenotype. In
addition, the CYP2D6 and CYP2C19 UM phenotype
may be adequately predicted by the number of ac-
tive alleles determined in the course of genotyping.
However, the clinical usefulness of CYP genotyping
is often hampered by a poor genotype-phenotype
correlation of enzyme activity with respect to the
most common EM and IM phenotypes. This can be
the consequence of drug-drug interactions that may
transform a patient with a normal or so-called wild-
type genotype and hence with an anticipated EM
phenotype to a UM or PM upon exposure to induc-
ers or inhibitors of CYP enzymes, respectively. An-
other reason for a poor genotype-phenotype corre-
lation stems from the differential reduction of en-
zyme activity depending on the substrate metabo-
lised among subjects with heterozygous expression
of a defective allele.
In summary, most genetic polymorphisms leading to
an altered expression and activity of drug-metaboliz-
ing enzymes appear to be known and in many instanc-
es allow for a prediction of the corresponding phe-
notype. Such information is presently considered in
the course of drug development and candidate drugs
selectively metabolised by polymorphic enzymes are
often dropped in early drug screening. Despite an
eventually limited genotype-phenotype correlation of
enzyme activity, it would certainly be advisable to con-
duct a genotyping analysis once in a lifetime and in-
clude such information in the patient’s medical record
to improve drug safety and efficacy.