II-38 Zdrav Vestn 2007; 76: SUPPL I
CYP PHENOTYPING TO ASSESS DRUG-DRUG INTERACTION
Uwe Fuhr
Dept. of Pharmacology, University of Cologne, Gleueler Str. 24, 50931 Köln, Germany; e-mail:
uwe.fuhr@uk-koeln.de
chlorzoxazone over chlorzoxazone in plasma 2 hours
after 250 mg chlorzoxazone (chlorzoxazone inhibits
CYP3A4 and therefore should not be incorporated
in phenotyping cocktails); CYP3A4/5, plasma clear-
ance of midazolam after 2 mg midazolam (all drugs
given orally). If intestinal and hepatic CYP3A4/5 ac-
tivities should be assessed separately, hepatic clear-
ance of 1 mg midazolam i.v. may be used for hepatic
activity and intestinal availability, calculated from 1
mg midazolam i.v. and 2 mg midazolam orally (se-
quential administration of use of mass-labelling), is a
metric for intestinal CYP3A4/5 activity.
If the effect of a drug on CYP phenotype is to be stud-
ied, the extent of intervention should be according to
the clinical situation to be assessed. For chronic drug
treatment as an intervention, administration of the
drug thus should be at the standard therapeutic dose
and last until steady state for both the drug (approxi-
mately 5 half-lives) and, if any, the changes of enzyme
activities is reached. One week appears to be suffi-
cient for enzyme induction. The exposure to the in-
teracting drug should be monitored by quantification
of the drug and its relevant metabolites.
Phenotyping procedures may be combined in a »cock-
tail« to assess a metabolic profile and to simultaneously
quantify the effect of a drug on several cytochrome
P450 enzymes in vivo. In all cocktails, the selection of
metrics is determined by several factors, including the
specific objective of the respective study, the avail-
ability of analytical methods, and the balance between
expense and validation. Many useful cocktails have
been reported without mutual interaction between
probe drugs and with good tolerability. Still, there is
ample space for an improvement by replacement of
individual probe drugs, by use of a better validated
metric, by reducing the doses, and/or by simplifica-
tion of the procedures. Low doses and successful val-
idation of limited sampling strategies make the use of
cocktails increasingly convenient. The identification
of metrics with low intraindividual variability allows
the assessment of the effect of interacting drugs with
a very limited sample size (12 in most cases). Yes, there
is an error in the last sentence, it should be corrected
to »In summary, phenotyping and its use drug cock-
tails is a valuable, safe and scientifically sound tool to
quantify the effect of drugs on the in vivo activity of
CYP enzymes and thus to assess the metabolic inter-
action profile of a drug.«
Phenotyping for drug metabolizing enzymes is de-
fined as measuring its actual in vivo activity in an indi-
vidual. This is done by administration of a selective
substrate for this enzyme and subsequent determina-
tion of appropriate pharmacokinetic parameters, or
by using metabolism of endogenous substrates. The
in vivo effect of a drug on enzyme activity may quan-
titatively be assessed by comparison of phenotypes
in presence and absence of the drug.
Appropriate selection of phenotyping test drugs and
metrics is essential to obtain results applicable for oth-
er substrates of the respective enzyme. Phenotyping
is considered as valid if respective criteria are fulfilled,
including (i) changes in metric when patients are treat-
ed with inhibitors/inducers of the enzyme; (ii) effect
of liver disease if the enzyme is expressed primarily
in the liver; (iii) correlation of metabolite formation
with activity and content of enzyme in tissue prepa-
rations containing the enzyme; (iv) in vitro specifici-
ty of the metabolic step; (v) high contribution of the
metabolic step to overall drug metabolism; (vi) corre-
lation of metric with the partial clearance for the re-
spective specific metabolic step; (vii) good reproduc-
ibility; (viii) correlation of the metric with the AUC of
parent substrate; (ix) correlation of the metric with
other validated metrics; (x) metric reflects known ge-
netic polymorphism; (xi) metric does not depend on
confounding factors such as urinary pH, urinary flow,
renal function. Furthermore, the test drug should be
registered as a drug, and the procedure should be sim-
ple and non-invasive.
Currently, for most of the CYP enzymes important in
drug metabolism there are phenotyping procedures
which provide valid information on individual activi-
ty. The following metrics are recommended based
on the level of validation and on practicability (alter-
native in parentheses): CYP1A2, paraxanthine/caf-
feine in plasma 6 hours after 150 mg caffeine;
CYP2C9, tolbutamide plasma concentration 24 hours
after 125 mg tolbutamide ((S)-warfarin plasma AUC
after a 10 mg dose of racemic warfarin given in com-
bination with 10 mg of vitamin K to prevent anti-
coagulation); CYP2C19, urinary excretion of 4’-OH-
mephenytoin 0–12 hours after 50 mg mephenytoin;
CYP2D6, urinary molar ratio debrisoquine/4-OH-
debrisoquine 0–8 hours after 10 mg debrisoquine
(dextromethorphan plasma AUC after a 30 mg dex-
tromethorphan-HBr); CYP2E1, urinary ratio 6-OH-