International Journal of Clinical Pharmacology and Therapeutics, Vol. 49 No.
8/2011 (491-499)
Influence of renal impairment on the
pharmacokinetics of oral roflumilast: an
open-label, parallel-group, single-center study
Original T.D. Bethke1, M. Hartmann1, A. Hnnemeyer1, G. Lahu1 and C.H. Gleiter2
2011 Dustri-Verlag Dr. K. Feistle
ISSN 0946-1965 1Nycomed GmbH, Konstanz, and 2Department of Clinical Pharmacology,
DOI 10.5414/CP201556 University of Tbingen, Tbingen, Germany
Key words Abstract. Objective: Roflumilast is a
renal impairment novel, orally active, selective phosphodies-
roflumilast human terase 4 inhibitor recently approved in the
pharmacokinetics
European Union for the treatment of severe
COPD. Roflumilast and its metabolites are
mainly (70% of total radioactivity) elimi-
nated via the kidneys as glucuronides. The
potential impact of renal impairment on the
pharmacokinetics of roflumilast and its ac-
tive main metabolite roflumilast N-oxide
were characterized. Materials and methods:
Patients (n = 12) with severe renal impair-
ment (creatinine clearance CLCR < 30 ml/
min/1.73 m; otherwise healthy) and matched
(sex, age, weight, and height) healthy control
subjects (n = 12; CLCR > 80 ml/min/1.73 m)
were enrolled into an open-label, parallel-
group study. Single dose (500 g, p.o.)
pharmacokinetics and safety/tolerability of
roflumilast and roflumilast N-oxide were
compared between both groups. Results: A
minor decrease of exposure (area under the
plasma concentration-time curve from time
zero to infinity (AUC0), maximum plasma
concentration (Cmax)) and a small increase
in elimination half-life (t1/2) of roflumilast
(21%; 16%; +19%, respectively) and
roflumilast N-oxide (7%; ND; +30%,
respectively) were observed in renally
impaired patients compared with healthy
subjects. No relevant differences in safety
and tolerability were observed between
groups. Conclusions: The pharmacokinetic
changes observed in patients with renal
impairment are of small magnitude with-
out clinical importance. A dose adjustment
Received or a change in the administration interval
February 8, 2011; of roflumilast is not necessary in patients
accepted with renal impairment.
March 20, 2011
Correspondence to
Dr. T.D. Bethke Introduction
Nycomed Deutschland
GmbH, Moltkestrasse 4, Roflumilast is a selective phosphodies-
78467 Konstanz,
terase (PDE)-4 inhibitor recently approved
Germany
thomas.bethke@ in the European Union for maintenance
nycomed.com treatment of severe chronic obstructive pul-
monary disease (COPD) associated with
chronic bronchitis in adult patients with a
history of frequent exacerbations as add on
to bronchodilator treatment.
Roflumilast has a high bioavailability
(79%) [1]. In humans, hepatic metabolism of
roflumilast is the principal route of drug elim-
ination. Roflumilast is mainly metabolized
by cytochromes P450 (CYP) 3A4 and 1A2
to the primary active metabolite roflumilast
N-oxide, which has approximately a third of
the potency (half-maximum inhibitory con-
centrations (IC50) of 0.7 nmol/l and 2 nmol/l,
respectively) of the parent drug [2, 3]. Com-
pared with roflumilast, the total exposure to
roflumilast N-oxide is approximately 10-fold
higher and the terminal disposition half-life
(t1/2) approximately 1.5-fold longer (30 h and
17 h, respectively) [1]. The plasma protein
binding of both roflumilast and roflumilast N-
oxide is high (98.9% and 96.6%, respectively)
and independent of concentration at therapeu-
tic doses [4, 5]. Roflumilast N-oxide is me-
tabolized predominantly via CYP 3A4 [3] by
loss of the cyclopropylmethyl moiety or am-
ide cleavage to compounds which are mainly
renally excreted after glucuronidation.
The principal route of roflumilast ex-
cretion differs between species. In dog, rat,
hamster, and mice, most of the dose is ex-
creted via the feces, whereas renal elimina-
tion predominates in minipig, rabbit, and
monkeys [5]. Similarly, data from a mass-
balance study in man show that 70% of the
administered dose is eliminated via the kid-
neys as drug-related material (total radioac-
tivity) [6].
Unchanged parent compound has not
been detected in the urine and roflumilast N-
oxide in only trace amounts [4].
Bethke, Hartmann, Hnnemeyer et al. 492
A decline in renal excretion is not likely
to alter the pharmacokinetics (PK) of roflu-
milast significantly. But characterization of
the PK in patients with severe renal impair-
ment is recommended by current regulatory
guidance [7] even if the drug is eliminated
primarily via hepatic metabolism unless
it also has a relatively wide therapeutic in-
dex (which is not the case with roflumilast
that might potentially cause a higher rate of
adverse reactions e.g., diarrhea, nausea or
headache at increased exposure).
Severe renal impairment can affect the
PK by a variety of mechanisms including
changes in drug distribution by reduced plas-
ma protein binding [8], a decrease of hepatic
metabolism [9, 10] or alterations in activity
of transporter systems [10] in the liver, in-
testine or kidney. A decline in renal function
is observed as a consequence of normal ag-
ing or renal disease. As patients with COPD
are typically of advanced age with numerous
co-morbidities, roflumilast is expected to be
given to patients who may have or develop
renal impairment. Whether posology of the
once-daily roflumilast should be modified
in this subset of patients in order to avoid
potentially excessive plasma concentrations
that could lead to decreased tolerability is
consequently a question of practical interest.
Therefore, the PK was investigated as a pri-
mary objective and safety and tolerability as
a secondary objective of this study.
As the study rationale was to exclude the
effect of renal impairment on the PK, a co-
hort of patients with severe renal impairment
was compared in a reduced study design with
healthy subjects as a within-study control
group. The control subjects were selected in
consideration of factors that might affect the
PK of roflumilast and therefore had to match
patients by (race), gender, age, weight, and
height to be eligible for study inclusion.
Since both roflumilast and roflumilast
N-oxide exhibit linear and time-independent
PK [12], a single-dose study was conducted.
The same dose (500 g) was administered
to all subjects regardless of renal function
as the selected dose was safe and well tol-
erated in previous human studies (maximum
tolerated dose of 1,000 g with single oral
doses) [13], and as the peak concentration of
the first (loading) dose is in most cases not
substantially affected by renal impairment.
Methods
Ethics and study subjects
The study was conducted in accordance
with the principles of the International Con-
ference on Harmonisation guideline of Good
Clinical Practice and the Declaration of Hel-
sinki (Somerset West, Republic of South Af-
rica, October 1996). The study protocol, sub-
ject information and informed consent form
were reviewed and approved the Czech Min-
istry of Health, State Institute for Drug Con-
trol, Prague and by the Ethics Committees
of the General Faculty Hospital, 1st Medical
Faculty of Charles University, Prague and of
the Faculty Hospital, 1st Faculty of Charles
University in Pilsen.
The planned enrolment was a total of 24
healthy male or female adult subjects and
patients with severe renal impairment in an
equal ratio. Caucasian males or females,
18 65 years of age, who had a normal body
weight as defined by a Broca index (weight
[kg]/{height [cm] 100}) of 0.8 1.25; a
negative serology test result for HIV 1/2,
HBsAg, anti HBc, anti HAV, anti HCV; and
who were willing and able to provide in-
formed consent were eligible for inclusion
into the study.
Renally impaired patients and healthy
subjects were selected and classified by their
respective body surface (BSA) normalized
glomerular filtration rate (GFR) as assessed
by endogenous creatinine clearance (CLCR).
Creatinine was assayed with Jaffs method
(Roche-Hitachi 717 analyzer). The CLCR
was measured twice in each subject within 6
weeks prior to inclusion (with subjects being
resident for a 24-h urine collection and one
serum creatinine determination). The first
CLCR measurement was used to classify the
renal function and the second measurement
to confirm the classification and to demon-
strate stable renal conditions (defined as dif-
ference by no more than 25% in renal pa-
tients and 35% in healthy subjects).
Patients were eligible who had severe
renal impairment (as defined by endogenous
CLCR of 10 30 ml min1 1.73 m2 BSA)
secondary to chronic glomerulonephritis, py-
elonephritis or interstitial nephritis; who had
a stable state of renal function; and who had
no clinically significant abnormal laboratory
Pharmacokinetics of roflumilast in renal impairment 493
findings other than those associated with re-
nal disease.
Healthy control subjects were requested
instead to have normal renal function (as de-
fined by endogenous CLCR 80 ml min1
1.73 m2 BSA) and to match with renally
impaired patients by gender, age ( 5 y),
weight ( 10%), and height ( 10%).
Subjects and patients were not eligible
for enrolment, if they had: previously par-
ticipated in any other study within the last 3
months prior to study entry; donated blood
or plasma within the last 2 months; smoked
more than 15 cigarettes/d; a history of alco-
hol abuse; a positive or missing urine test
for drugs of abuse at screening; any signs of
cardiac diseases; arterial hypo- or hyperten-
sion; a history of clinically relevant allergy
or asthma; or a history of major gastro-intes-
tinal surgery including cholecystectomy with
the exception of appendectomy. Females (all
study groups) were not eligible if they were
pregnant or nursing, or had childbearing po-
tential and did not use a reliable method of
contraception.
Patients with severe renal impairment
were required to be on a stable treatment
regimen (at least 4 weeks unchanged dosage
prior to screening) and to have no history of
or ongoing relevant other disease. Patients
with nephrotic syndrome, secondary renal
impairment (carcinoma, tuberculosis), re-
cipients of kidney transplants, and patients
on hemodialysis or steroid treatment were
excluded from study participation.
Healthy subjects taking any previous
medication within 2 weeks prior to inclusion
with the exception of oral contraceptives
were excluded from the study as well as pa-
tients taking any previous enzyme inducing
or inhibiting drugs during the last 2 weeks
before inclusion or as concurrent medication
during the course of the study.
Study design and safety
assessments
This was a prospective, open label, non-
randomized, comparative parallel-group study
conducted at two study sites. The primary
objective of this study was to investigate the
PK of roflumilast in patients with severe re-
nal impairment in comparison to a matched
control group of healthy subjects after single
oral dose administration. Secondary objec-
tives were to assess the safety and tolerabil-
ity of roflumilast and to investigate the PK
of its pharmacologically active metabolite
roflumilast N-oxide.
Subjects and patients were screened for
eligibility up to 4 weeks prior to entry into
the study (medical history, vital signs, physi-
cal examination, safety laboratory, electro-
cardiogram (ECG), recording of demograph-
ic information and current medication).
Subjects enrolled into the clinical trial re-
mained resident from the evening of Day 1
until Day 2 after collection of the 36 h blood
sample and returned to the trial unit up to 96 h
thereafter for taking further blood samples.
The study treatment was administered
at Day 1 as a single oral dose of roflumilast
film-coated tablets (500 g) taken with 200 ml
of still water after a standardized breakfast.
Concomitant use of paracetamol (1 g/d) was
allowed in case of severe headache.
Safety measurements (vital signs, safety
laboratory parameters, ECG) and adverse
events (AEs) were recorded continuously
during the course of the study (at screening,
pre-dose, 1 h, 2-h post-dose, and follow-up
examination at 96-h post-dose). An addi-
tional follow-up examination was scheduled
if deemed necessary for medical reasons as
an outpatient visit within 2 weeks after dis-
charge from the clinical trial unit.
Pharmacokinetic analysis
Venous blood samples (7 ml each) were
taken up to 96 h (at pre-dose, 0.25 h, 0.5 h,
1h, 1.5 h, 2 h, 2.5 h, 3 h, 4 h, 5 h, 6 h, 8 h,
10 h, 12 h, 14 h, 24 h, 30 h, 36 h, 48 h, 60 h,
72 h, and 96 h) after dosing for assay of ro-
flumilast and its primary metabolite. On each
occasion, a 7-ml volume of blood was drawn
into a lithium heparinate Monovette tube
(Sarstedt, Germany) and blood was prompt-
ly processed into plasma by centrifugation.
Specimens were stored frozen at approxi-
mately 20 C until analysis.
Bioanalysis was conducted at Nycomed
Germany (Konstanz) using a validated bio-
analytical assay. Plasma samples were ana-
lyzed for roflumilast and roflumilast N-oxide
by high-performance liquid chromatography
 Bethke, Hartmann, Hnnemeyer et al. 494
Table 1. Demographic and clinical baseline characteristics of healthy sub-
jects and patients with severe renal impairment.
Variablea Healthy subjects Patients with renal
impairment
No. of subjects 12 12
Age (y) 52 (38 62) 56 (34 64)
Body weight (kg) 76.0 (56 95) 76.0 (56 105)
Height (cm) 175 (153 182) 173 (155 188)
Broca 1.09 (0.89 1.16) 1.09 (0.88 1.26)
CLCR (ml/min) 102.3 (89.8 156.0) 25.5 (13.6 31.2)
a
Values are expressed as median (range) unless specified otherwise. y =
years.
(HPLC) separation and post-column pho-
tochemical derivatisation for fluorescence
detection. The lower limits of quantitation
(LLOQ) were 0.085 g/l for roflumilast and
0.5 g/l for roflumilast N-oxide, respectively
(sample volume of 1 ml each). The accep-
tance criteria of day-to-day variation for ac-
curacy of roflumilast and roflumilast Noxide
were 6.34 1.3 and 0.2 11.72, respective-
ly whereas the corresponding ranges for pre-
cision during method validation were 1.96
2.71 and 1.91 8.19, respectively.
PK parameters were derived by standard
non-compartmental analysis using KINTPC
software (version 2.0; Nycomed). Maximum
(peak) plasma drug concentrations (Cmax)
were obtained directly from the measured
concentrations and areas under the plasma
concentration-time curve (AUCs) calculated
by the linear trapezoidal method with extrapo-
lation to infinity (AUC0) from last quantifi-
able concentrations. The t1/2 was calculated
(t1/2 = ln2/ lz) from the elimination constant lz
which was estimated by linear regression of at
least three contiguous logarithmic concentra-
tion data in the descending part of the concen-
tration-time profiles. All PK parameters were
calculated on an individual subject basis.
Statistical analysis
No formal power calculation was made
in this study and the sample size of subjects
to be enrolled into the study was decided on
grounds of expected feasibility.
The safety analysis comprised all sub-
jects who received the trial medication. The
per-protocol analysis of the primary variable
was conducted for all subjects who addition-
ally had sufficient data for evaluation of the
primary variables and who completed the
study without any protocol violation.
Statistical analysis was conducted using
SAS software (version 1.2.2; SAS Institute
Inc., Cary, NC, USA).
Primary variables for the statistical anal-
ysis were AUC0 and Cmax of roflumilast.
Point estimates for the population medians
of Test (patients with renal impairment) to
Reference (healthy subjects) ratios were
calculated and the respective 90%-confi-
dence interval (CI) limits estimated from the
independent t-test after logarithmic transfor-
mation. The Welch modification was used in
case of variance inhomogeneity.
No cut-off limits were pre-specified for
posology adjustment and the decision was to
be made on clinical grounds (e.g. tolerabil-
ity), instead.
Secondary variables including t1/2 and
the time to maximum plasma concentration
(tmax) of roflumilast and the AUC0 and t1/2
of roflumilast N-oxide were analyzed in an
explorative manner. Data of t1/2 were ana-
lyzed in a multiplicative model (as described
above for AUC and Cmax) whereas an addi-
tive model was used for tmax. Descriptive sta-
tistics including median, 68%-range, mean,
standard deviation (SD) or standard error of
mean (SEM), geometric mean, and geomet-
ric 68%-range, if appropriate, were given for
safety data.
Results
Demographics
24 subjects were enrolled and completed
the study. The demographic characteristics
of the study population are presented in
Table 1. The cohorts of healthy subjects and
patients with renal impairment showed no
statistically significant differences (p < 0.05)
for any of the matching characteristics age,
weight, height, and Broca index.
Patients and healthy adults had a median
(range) baseline CLCR (calculated as mean
of 2 measurements) of 25.5 (13.6 31.2) ml/
min and 102.3 (89.8 156.0) ml/min, re-
spectively.
Most patients (11 of 12) took concomi-
tant treatments (average 3.4 drugs) for the
Pharmacokinetics of roflumilast in renal impairment 495
Figure 1. Plasma concentrations (geometric mean;
68%-range) of roflumilast in patients with severe
renal impairment (solid line) and in healthy sub-
jects (dotted line) following a single oral dose of
roflumilast (500 g); semilogarithmic scale.
Figure 2. Plasma concentrations (geometric
mean; 68%-range) of roflumilast N-oxide in pa-
tients with severe renal impairment (solid line) and
in healthy subjects (dotted line) following a single
oral dose of roflumilast (500 g); semilogarithmic
scale.
management of renal disease which com-
prised mainly ACE inhibitors (9 patients,
75%), allopurinol (8 patients, 67%), diuret-
ics (6 patients, 50%), and only infrequently
supplemental calcium carbonate (2 patients,
17%) or phosphate binders (1 patient, 8%).
Three subjects with partially missing data
or those otherwise not analyzable for primary
PK parameters of roflumilast or roflumilast
N-oxide were excluded from the calculation
of respective mean PK data.
Pharmacokinetic results
Compared with healthy subjects, plasma
concentration-time curves of group (geomet-
ric) means were nearly superimposable and
not meaningfully altered in subjects with se-
vere renal impairment. The respective plots
show only a minor decrease of exposure to
both roflumilast (Figure 1) and roflumilast N-
oxide (Figure 2) after single dose-administra-
tion of roflumilast. The roflumilast concentra-
tion in plasma samples of most renal patients
(10 of 12) and healthy subjects (10 of 12)
beyond 60 h were below LLOQ while only 2
of the samples of renal patients, but none of
healthy subjects had plasma concentrations
of roflumilast N-oxide below LLOQ at 96 h.
The geometric mean AUC0 and Cmax
of roflumilast were approximately 20% and
16% lower in patients with renal impair-
ment compared with healthy adults while
t1/2 was 19% higher (Table 2). Similarly, the
geometric mean AUC0 and Cmax of roflu-
milast N-oxide were approximately 7% and
12% lower in patients with renal dysfunction
compared with matched healthy adults while
t1/2 was 30% higher.
Diagrams of AUC in the two cohorts do
not show any relevant dependence of expo-
sure on the renal function for both roflumi-
last and its main metabolite. (Figures 3, 4).
Especially roflumilast N-oxide as the prin-
cipal contributor to the overall pharmaco-
dynamic effect is apparently independent of
the glomerular filtration rate as seen by mean
and range of the AUC in patients and healthy
subjects.
Numerical data of the ratios of the popula-
tion medians and the corresponding 90% CIs
are summarized in Table 3. Point estimates
of exposure to roflumilast and its metabolite
indicate a slight decrease with severe renal
impairment, the half-lives a minor increase.
Safety and tolerability
Single doses of roflumilast were general-
ly well tolerated in the two study groups. No
serious AEs and no AEs leading to prema-
ture discontinuation from the study occurred.
AEs were of mild intensity and unrelated to
the study drug. Only two AEs were reported
 Bethke, Hartmann, Hnnemeyer et al. 496

Table 2. Descriptive statistics of the main pharmacokinetic parameter esti- Discussion

mates of roflumilast and roflumilast N-oxide in healthy subjects and patients
with severe renal impairment. The PK of two-thirds of all drugs de-
pends on renal function [14]. The number of
Variablea Healthy subjects Patients with renal
(n = 12) impairment patients in clinical practice who have at least
(n = 12) some degree of renal insufficiency is high
Roflumilast and an estimated 11.5% of adults with an age
AUC0 (mg h/l) 44.7 (33.5, 59.7) 35.5 (23.7, 53.0) of 20 or older have physiological evidence
Cmax (mg/l) 5.07 (3.62, 7.10) 4.27 (2.74, 6.63) of chronic kidney disease (US National
tmax (h)b 1.75 (1.00, 2.50) 1.50 (0.50, 2.50) Health and Nutrition Examination Survey)
t1/2 (h) 18.5 (11.4, 30.0) 22.1 (14.2, 34.4) [15]. Though numerous drugs require dose
Roflumilast N-oxide adjustment in patients with renal impairment
AUC0 (mg h/l) 461.2 (366.2, 580.7) 428.9 (334.2, 550.5)
in order to avoid toxicity [16], many other
Cmax (mg/l) 7.78 (6.35, 9.53) 6.84 (5.35, 8.74)
b
including commonly prescribed drugs (e.g.,
tmax (h) 13.0 (3.0, 30.0) 12.0 (4.0, 48.0)
clindamycin, minoxidil, amiodarone, fen-
t1/2 (h) 28.7 (22.0, 37.5) 37.4 (21.9, 63.9)
tanyl, rosiglitazone) do not [17].
a
Values are expressed as geometric mean (68% range) (back-transformed This clinical trial investigated in a re-
from the mean SD of the log-transformed data) unless specified otherwise.; duced study design, whether a dose adjust-
b
median and range (min, max). AUC0 = area under the plasma concentra- ment is necessary if roflumilast is adminis-
tion-time curve from time 0 to infinity; Cmax = maximum plasma concentration;
tered to patients with renal impairment. Only
tmax= time to reach Cmax; t1/2 = half-life; SD = standard deviation.
subjects at both extremes of renal function
were enrolled as the PK of roflumilast was
Table 3. Comparison of point estimates [%] and 90% CIs for the ratio of the
treatment group means (healthy subjects and patients with severe renal im-
expected to be independent of the renal func-
pairment) of main pharmacokinetic parameters. tion. Patients with mild or moderate renal
impairment may have been added in an adap-
Variablea Point estimates T/R (%) (90% CI) (%)
tive two-stage design if the assumption was
Roflumilast
79.4 (61.8, 101,9)
not met.
AUC0
Cmax 84.1 (63.9, 110.7)
Though the mechanisms of renal elimina-
tmax 41.7 (83.4, 00.1) tion have not been established for roflumilast
t1/2 119.3 (86.2, 165.0) and roflumilast N-oxide, there are no indica-
Roflumilast N-oxide tions for an active secretion of roflumilast.
AUC0 93.0 (77.5, 111.6) Glomerular filtration is likely to be the pre-
Cmax NR N/A dominant process. In this case, alterations in
tmaxb NR N/A drug clearance are proportional to changes in
t1/2 130.3 (95.1, 178.7) glomerular renal function [18]. Other renal
mechanisms (tubular secretion) are assumed
a
Values are expressed as ratio % (T/R) with renal impairment (Test T) and to decline in parallel with glomerular filtra-
healthy subjects (Reference R). AUC0 = area under the plasma concentra-
tion-time curve from time 0 to infinity; CI confidence interval; Cmax = maximum tion in the intact nephron model and may
plasma concentration; tmax = time to reach Cmax; t1/2 = half-life; N/A = not ap- be ignored [19]. Consequently, patients have
plicable; NR = not reported. been selected for this clinical trial on the ba-
sis of their GFR with CLCR as a surrogate
during the study (2 of 24 study participants, marker for impaired renal function. Though
8.34%): CLCR in patients with severe renal impair-
A healthy subject with a previous medical ment probably overestimates the true GFR
history of a similar episode experienced an because of increased tubular secretion of
atrial arrhythmia (A-V junction rhythm) last- creatinine in the residual nephrons it is more
ing for about 1 h at Day 1 with onset about 30 practical than most other alternatives and
min after his tmax. Low-back pain was report- consequently widely used in clinical practice
ed on study Day 4 by another healthy subject as a measure of renal function.
with a history of recurrent complaints. The study used BSA-normalized CLCR in
Laboratory test results, vital signs, and accordance with European Guideline [20],
ECG did not reveal any clinically significant but estimated CLCR is otherwise fully ac-
average or individual changes during the ceptable and has been used in the analysis
course of the study. since factors changing CLCR affect drug CL
Pharmacokinetics of roflumilast in renal impairment 497
Figure 3. Total exposure (AUC0) of roflumilast
in dependence of glomerular filtration rate in pa-
tients with severe renal impairment and in healthy
subjects (single oral dose of roflumilast 500 g).
Figure 4. Total exposure (AUC0) of roflumilast
N-oxide in dependence of glomerular filtration rate
in patients with severe renal impairment and in
healthy subjects (single oral dose of roflumilast
500 g).
in an analogous fashion without normaliza-
tion [21].
Minor PK differences between healthy
subjects and patients were found in this study.
Both, peak and mean systemic exposures to
roflumilast (16% and 12%, respectively) and
roflumilast N-oxide (21% and 7%, respec-
tively) were lower in the cohort with severe
renal impairment than in matched healthy
controls. This is a common finding with
highly protein-bound drugs as slight protein-
uria secondary to the renal damage will lead
to a concomitant loss of the bound drug. An
alternative explanation for the reduced ex-
posure might be a decreased enteral absorp-
tion as patients with kidney impairment may
expose a lower enteral drug absorption and
bioavailability as a result of gut edema [9]
or pH alterations [11]. Though a prolonged
gastric emptying time is frequently observed
in patients with chronic kidney disease as a
result of gastroparesis [22], data of this study
show no indications of delay in tmax.
The half-life of roflumilast and roflumi-
last N-oxide increased by 19% and 30%,
respectively with severe renal impairment.
This is equivalent to a decreased rate elimi-
nation constant which is dependent on clear-
ance and volume of distribution. A reduced
plasma binding due to a decreased concentra-
tion of plasma albumin in patients with renal
impairment [23] or a displacement of drugs
from plasma protein binding by endogenous
or exogenous substances that would normal-
ly be eliminated by the kidney, but accumu-
late in the blood of patients with poor kidney
function can increase the volume of distri-
bution of drugs [24]. An increased volume
of distribution might explain the observed
change in half-life of roflumilast and roflu-
milast N-oxide in renally impaired patients.
For this study it was expected that even
severe renal impairment may not alter the PK
of roflumilast or its active main metabolite
sufficiently to warrant dosage adjustment.
Therefore, an analysis of the PK data was
conducted by calculating ratios (point esti-
mates) of measurements relevant to thera-
peutic outcome (e.g. AUC0, Cmax, tmax, t1/2)
to show that PK parameters in patients with
severe renal impairment (Test) are similar to
those in healthy adult subjects with normal
renal function (Reference).
No drug-specific cut-offs for posology
adjustments were defined in the protocol
prior to the conduct of this study as informa-
tion for what change in the PK of roflumilast
or roflumilast N-oxide should justify a dose
adjustment was not available from PK-PD
relationships, previous dose-finding studies
or dedicated dose-response studies and the
decision was to be made on clinical ground
such as tolerability.
In absence of specific no effect bound-
aries, a priori boundaries of 70 143% for
Bethke, Hartmann, Hnnemeyer et al. 498
Cmax and 80 125% for AUC0 are recom-
mended [7]. While the 90%CI of AUC and
Cmax of roflumilast are not completely con-
tained in these proposed intervals and the
lower boundary of the 90%CI of AUC of
roflumilast N-oxide (which principally de-
terminates efficacy) is only closely outside,
it is explicitly recognized in the current guid-
ance that unlike bioequivalence studies the
limitation for small clinical sample sizes in
renal impairment studies coupled with high
inter-subject variability may preclude meet-
ing these no-effect boundaries. The small
magnitude of the observed differences in ex-
posure and clearance make a loss in thera-
peutic efficacy in patients with severe renal
impairment unlikely.
In addition, no indications of an altered
safety and tolerability were found in this
study between patients with severe renal im-
pairment compared with healthy adult sub-
jects. Considering these comparable clinical
safety findings along with the small mag-
nitude of the PK differences, the observed
changes seen with declining GFR patients
are considered of no clinical relevance. Ex-
trapolating this conclusion from a reduced
study design to patients with any degree of
renal impairment (mild, moderate or severe),
a dose adjustment is not needed in the sub-
population of renally impaired patients.
Patients with end-stage renal disease on
maintenance intermittent hemodialysis may
require greater doses of certain drugs than
patients with normal renal function and
dialysis can be used for accelerated drug
elimination for management of overdose.
However, the PK of roflumilast in patients
on dialysis has not been investigated. As ro-
flumilast is highly bound to plasma protein,
hemodialysis is not likely to be an efficient
method of its removal. There are no data to
support whether roflumilast is dialyzable by
continuous ambulatory peritoneal dialysis.
Conclusion
This study showed some minor changes
in exposure and half-life of roflumilast and
its primary active metabolite roflumilast
N-oxide, but no clinically important PK al-
terations between patients with severe renal
impairment compared with healthy controls.
Safety and tolerability were not different
between both groups. A dose adjustment of
roflumilast is not warranted with decreasing
creatinine clearance.
Acknowledgments
The authors thank the volunteers and
staff who participated in this study, as well as
Dr. Thomas Papke for supporting the prepa-
ration of this manuscript.
Conflict of interest
This study was conducted by Nycomed
GmbH, Research and Development, Germa-
ny. C.H.G. reports no conflicts of interest in
this work. M.H., A.H., G.L., and T.D.B. are
employees of Nycomed.
References
[1] Bethke TD, Lahu G. High absolute bioavailability
of the new oral phosphodiesterase-4 inhibitor
roflumilast. Int J Clin Pharmacol Ther. 2011; 49:
51-57. PubMed
[2] Bundschuh DS, Eltze M, Barsig J, Wollin L, Hatzel
mann A, Beume R. In vivo efficacy in airway
disease models of roflumilast, a novel orally active
PDE4 inhibitor. J Pharmacol Exp Ther. 2001;
297: 280-290. PubMed
[3] Hatzelmann A, Schudt C. Anti-inflammatory and
immunomodulatory potential of the novel PDE4
inhibitor roflumilast in vitro. J Pharmacol Exp
Ther. 2001; 297: 267-279. PubMed
[4] Hermann R, Nassr N, Lahu G, Pterfai E,
Knoerzer D, Herzog R, Zech K, de Mey C. Steady-
state pharmacokinetics of roflumilast and roflu
milast N-oxide in patients with mild and moderate
liver cirrhosis. Clin Pharmacokinet. 2007; 46:
403-416. PubMed doi:10.2165/00003088-20
0746050-00003
[5] Nycomed. Data on file.
[6] Bethke T, Hartmann M, Zech K, David M, Weimar
C, Wurst W. No dose adjustment of roflumilast in
patients with severe renal impairment [abstract].
Am J Respir Crit Care Med. 2002; 65: A594.
[7] Food and Drug Administration. Guidance for
Industry Drug: Pharmacokinetics in patients with
impaired renal function study design, data
analysis, and impact on dosing and labeling.
Rockville, MD: US Department of Health and
Human Services; 1998.
[8] Reidenberg MM, Drayer DE. Alteration of drug-
protein binding in renal disease. Clin Pharmaco
kinet. 1984; 9 (Suppl 1): 18-26. doi:10.2165
/00003088-198400091-00003 PubMed
[9] Lam YW, Banerji S, Hatfield C, Talbert RL. Principles
of drug administration in renal insufficiency. Clin
Pharmacokinetics of roflumilast in renal impairment 499
Pharmacokinet. 1997; 32: 30-57.
doi:10.2165/00003088-199732010-00002
PubMed
[10] Okabe H, Hasunuma M, Hashimoto Y. The
hepatic and intestinal metabolic activities of P450
in rats with surgery- and drug-induced renal dys
function. Pharm Res. 2003; 20: 1591-1594.
doi:10.1023/A:1026131216669 PubMed
[11] Nolin TD, Naud J, Leblond FA, Pichette V.
Emerging evidence of the impact of kidney
disease on drug metabolism and transport. Clin
Pharmacol Ther. 2008; 83: 898-903. doi:10.1038/
clpt.2008.59 PubMed
[12] Bethke TD, Bhmer GM, Hermann R, Hauns B,
Fux R, Mrike K, David M, Knoerzer D, Wurst W,
Gleiter CH. Dose-proportional intraindividual
single- and repeated-dose pharmacokinetics of
roflumilast, an oral, once-daily phosphodiesterase
4 inhibitor. J Clin Pharmacol. 2007; 47: 26-36.
doi:10.1177/0091270006294529 PubMed
[13] Bethke TD, Hauns B, Zech K, David M, Wurst W.
Dose linearity of roflumilast, a new selective
PDE4 inhibitor [abstract]. Am J Respir Crit Care
Med. 2002; 165: A595.
[14] Seyffart G. Introduction. Drug Dosage in Renal
Insufficiency. Dordrecht: Kluwer; 1991.
[15] Levey AS, Stevens LA, Schmid CH, Zhang YL,
Castro AF III, Feldman HI, Kusek JW, Eggers P,
Van Lente F, Greene T, Coresh J; CKD-EPI (Chronic
Kidney Disease Epidemiology Colla boration). A
new equation to estimate glomerular filtration rate.
Ann Intern Med. 2009; 150: 604-612. PubMed
[16] Verbeeck RK, Musuamba FT. Pharmacokinetics
and dosage adjustment in patients with renal
dysfunction. Eur J Clin Pharmacol. 2009; 65: 757-
773. doi:10.1007/s00228-009-0678-8 PubMed
[17] Kappel J, Calissi P. Nephrology: 3. Safe drug
prescribing for patients with renal insufficiency.
CMAJ. 2002; 166: 473-477. PubMed
[18] Dettli L. Multiple dose elimination kinetics and
drug accumulation in patients with normal and
with impaired kidney function. Adv Biosci. 1969;
5: 39-54.
[19] Tett SE, Kirkpatrick CM, Gross AS, McLachlan
AJ. Principles and clinical application of assessing
alterations in renal elimination pathways. Clin
Pharmacokinet. 2003; 42: 1193-1211. doi:10.21
65/00003088-200342140-00002 PubMed
[20] European Agency for the Evaluation of
Medicinal Products (EMEA). COMMITTEE
FOR MEDICINAL PRODUCTS FOR HUMAN
USE (CHMP). Note for guidance on the
evaluation of the PK of medicinal products in
patients with impaired renal function. London:
CHMP/EWP/225/02; 2004.
[21] Boxenbaum H, Tannenbaum S, Mayersohn M,
Oleson F. Pharmacokinetics tricks and traps: drug
dosage adjustment in renal failure. J Pharm Pharm
Sci. 1999; 2: 2-4. PubMed
[22] Gabardi S, Abramson S. Drug dosing in chronic
kidney disease. Med Clin North Am. 2005; 89:
649-687. doi:10.1016/j.mcna.2004.11.007
PubMed
[23] Rowland M, Tozer T. N. Clinical pharmacokinetics:
Concepts and applications. Philadelphia, PA:
Lippincott Williams & Wilkins; 1995.
[24] Bauer L. A. Applied clinical pharmacokinetics.
New York: McGraw-Hill; 2008.
Copyright of International Journal of Clinical Pharmacology & Therapeutics is the property of Dustri-Verlag
Dr. Karl Feistle GmbH & Co., KG and its content may not be copied or emailed to multiple sites or posted to a
listserv without the copyright holder's express written permission. However, users may print, download, or
email articles for individual use.