Prostaglandins Leukotrienes and Medicine 11: 241-257, 1983

PHARMACOKINETICS OF THE STABLE PROSTACYCLIN

ANALOGUE, NILEPROST, IN THE RAT

W. Krause, P.E. Schulze and M. Totzek

Departments of Pharmacokinetics and Isotope
Chemistry
Schering, Berlin, Miillerstr.170 - 178
(W. Germany)

(reprint requests to W. Krause)

SUMMARY

The synthesis of radioactive nileprost, tritium-labelled

at positions 18 and 19, and its application to the
pharmacokinetics and biotransformation of this chemically
stable prostacyclin derivative in the rat is described.
3
H-Nileprost was absorbed after oral administration with
a half-life of 23 min reaching maximum concentrations in
the plasma 90 min after treatment. After intravenous
injection there was a threephasic decline in plasma
levels with half-lives of 15 min, 0.9 h and 11 h, respec-
tively. Unchanged drug was eliminated with t
Brain levels of drug or metabolites were les&'$hin': ?Ef
corresponding plasma concentrations. In autoradiographs
after i.v. and oral administration a very low volume of
distribution was found with maximum levels in the liver,
the kidney and the stomach mucosa. Nileprost was very
rapidly excreted, mainly by biliary elimination. Ten
metabolites were detected in urine and bile, one of them
being formed3exclusively in the gut wall. The main
fraction of H-activity in urine and bile, however, was
due to unchanged drug.

241
 INTRODUCTION
Prostacyclin has been discovered in blood vessels as a
metabolite of the arachidonic acid cascade (1, 2). Its
pharmacological actions include vasodilation (31, plate-
let aggregation inhibition (41, cytoprotection (5) and
dilation of the bronchi (6). The practical use of prosta-
cyclin, however, is limited by its chemical and metabolic
instability (7).

Extensive investigations, therefore, have been started in

the search for stable prostacyclin analogues. Further-
more, these compounds should preferably exhibit some
degree of dissociation of the pharmacological profile in
comparison to their natural precursor. One of those
substances synthesized is ciloprost, which is a potent
vasolidator and platelet aggregation inhibitor (8, 9). A
compound active in cytoprotection and dilation of the
bronchi is nileprost, which is currently under investiga-
tion in clinical trials.

The aim of the present study was to investigate the

pharmacokinetics and biotransformation of nileprost in
the rat and to compare the results to those obtained with
prostacyclin and ciloprost.
Furthermore these results are necessary for exactly
interpreting pharmacological and toxicological data of
this synthetic prostaglandin derivative.

MATERIALS AND METHODS

3H-nileprost
I. Synthesis of
Nileprost was tritium-labelled in positions 18 and 19 by
tritiation of an acetylated triple bond containing
analogue.
2.5 mg (5.3 umole) of compound 1 (Fig. 1) was dissolved
in 400 ~1 of a mixture of benzene/ethanol (l:l, v/v),
which had been heated for 30 min with 50 mg of tris-(tri-
phenyl-phosphin)rhodium(I)-chloride before. After adding
5 mg of the rhodium catalyst to the solufion, tritiation
was performed with 92.5 GBq (2.5 Ci) of H for 2 hours.
After removal of the tritium 400 ~1 of ethanol was added
and the catalyst was separated by filtration. Labile
tritium was removed by adding ethanol and subsequent
evaporation to dryness for several times. The residue

242
obtained was taken up in ethanol again and fractionated
by HPLC (systey no. 1) in 10 portions. 2.4 GBq (65 mCi)
of acetylated H-nileprost finally Were recovered. Ester
cleavage was performed in the combined solutions by
adding K CO to give a final pH value of 8 and standing
in a nitad
og n atmosphere at room temperature for 2 hours.
Methanol was evaportated in vacua, the residual solution
was diluted with water and extracted with methylene
chloride. The solution obtained was evaporated to dryness
and the residue purified by HPLC in 10 portions.

Fig. 1: Synthesis of tritium-labelled nileprost

Purity of 3H-nileprost was tested in two HPLC systems on

Lichrosorb RP-18 (lOpm, 250 x 4.6 nun;Knauer, Berlin).
I: water-methanol-acetic acid 30 : 70 : 0.2 Iv/v/v)
flow rate: 2 ml/min

II: water-methanol-acetic acid 40 : 60 : 0.2 (v/v/v)

flow rate: 4 ml/min
and in one TLC system on silica gel 60F254 (Merck,
Darmstadt)

I: methylene chloride-isopropanol 9 : 1 (v/v).

The specific activity of 1.2913GBq/mM (34.9 Ci/mmol) was
determined by combined UV and H-activity measurements.

243
For use in animal experiments the radioactive compound
was diluted with unlabelled substance as specified in
Table 1.

study animals sex &se radioactive PrC route of sampling

(!&kg) cJxaajI;Ci/ trmtmant ednin.

metabolic stability 12 f 200 10 . i.v./p.o. plasma

of the 3H-label

blood, plasma and 84 f 200 20 i.v./p.o. blood, plasma,

brain levels brain

autoradiography 11 f 100 ID0 Uth dsy of i.v./p.o. Plhole body

ww@an=Y

excretion with urine 10 5f+5m 200 10 - i.v./p.o. urine, faces,

and feces carcass

blliary excretion 5 f 200 10 urinary bled- i.v. urine, bile,

dar and bila- feces, carcass
duct cannula-
tion

enlsrohepatic 5 f 0.5 ml of ~a.2 urinary blad- i.d. urine, bile

recirculation radioec- &r and bile- feces, carcass
tive bile duct cannu-
l&ion

lab. 1: Experimental plan

II. Animal experiments

Male and female Wistar rats weighing about 200 g were
used in the experiments, which are sunn@arizedin Table 1.
In these studies the drug was dissolved in 0.5 ml of TRIS
buffer containing 0.606 mg of TRIS and 0.005 ml of
ethanol at a final pH value of 8.3. Whenever blood,
plasma or brain levels were measured, three animals were
sacrified at any sampling interval. For autoradiography,
which was performed according to Ullberg (191, one rat
and in all the other experiments five animals were used
per sampling interval.
Measurement of radioactivity
3H-activity was determined as has been described earlier
(11). However, H20 was separated before any measurements
by freeze-drying of the samples.

244
HPLC determinations

Metabolic patterns were determined by means of reversed-

phase high-performance liquid chromatography on
Lichrosorb RP-18 (10 km, 250 x 4.6 mm, Knauer) using a
linear methanol-water gradient from a ratio of 30 : 70 to
65 : 35 (v/v) in 100 min. 1 ml of acetic acid was added
to the eluent. The flow rate was 2 ml/min. Fractions were
sampled every 30 set and their radioactivity determined
in a liquid scintillation counter.

Pharmacokinetic calculations
For the evaluation of pharmacokinetic rate constants the
concentrations of drug or of labelled compounds were
plotted semilogarithmically versus time. From the termi-
nal slope the elimination rate constant was calculated.
The constants of other phases including the absorption
phase were gained by the feathering method. Excretion
data were plotted as C, -C versus time and evaluated
likewise.

RESULTS
Metabolic stability of the radioactive label
4 h after intravenous injection 0.19 + 0.05 % of the
radioactive dose was measured in the whole plasma. After
oral treatment 0.22 + 0.05 % was determined. 58 % and
45 %, respectively, of these compounds proved to be
tritium water. 24 h after administration 0.15 + 0.04 % of
the dose/plasma volume (i.v. injection) and 0.16 + 0.03 %
of dose/plasma volume (oral treatmegt) was found. 90 % of
the total radioactivity was due to H20.

Under the assumption that about 70 % of the rat consists

of water and that until 24 h3post administration only a
negligibly small portion of H 0 was eliminated, the
whole fraction of tritium wate$ is calculated to be less
fhan 3 % of the administered radioactive dose. The
H-active label synthesized, therefore, seemed to be
suitable for further pharmacokinetic studies. Neverthe-
less, any samples measured were evaporated to dryness
before radioactivity determinations in order to separate
from the tritium water formed.

245
Plasma levels

The time course of the concentration of radioactive

compounds after intravenous and oral administrations is
shown in Fig. 2. Following intravenous injection of
200 @g/kg the plasma level declined from 149 + 45 ng/ml
(calculated as nileprost equivalents) as measured 5 min.
post administration to 34 + 4 ng/ml at 45 min post admi-
nistration. Then the concentration rose again to
56 + 22 ng/ml (1 h post administration) and fell biphasi-
tally with half-lives of 0.9 h and 11 h (Table II),
finally achieving 2 + 1 nglml at 24 h post administra-
tion. The half-life of the first phase (0 - 45 min) was
calculated to be 15 min. The observed rise in concentra-
tion is due to enterohepatic recirculation and shall be
discussed later on.
The plasma level of unchanged drug (Fig. 2) declined with
a half-life of 14 min from 46 ng/ml as measured 5 min
post administration to 3 ng/ml at 1 h after injection.

70.
- 3H-adivHylplamali.v,
60. e--------s drug/m/i.v.

* - - -- %,adivity/~/,,.O.
-.- -‘. Wa&&ylbloodli.v.
‘---- =I,-~tbloodfD.0.
._ ... qj+&&.,b@“,i, v,

_.. ,.....

12 3 4 5 6 7 6 9 10 11 12 hours
L

Fig. 2
3
Concentration of nileprost and of total H-activity (as
ciloprost equivalents) in blood, plasma and brain of rats
after doses of 200 Nglkg.

Following oral treatment with the same dose maximum

levels of labelled compounds of 34 + 16 ng/ml were
achieved 90 min after administration. The half-life of
absorption was 22 min. Disposition of radioactive

246
substances was biphasic with half-lives of 1.5 h and 8 h
(Table II).

Blood levels
The concentration of labelled compounds in the blood
paralleled that of the plasma level (Fig. 2). The ratio
was calculated to be 1.36 + 0.14 (n = 15)
zPlaPdi:elMY ous injection and 1.19 + 0.58 (n = 15)
after oral treatment. Enrichment of rzdioactive substances
in blood particles was not observed.
Brain levels

When considering any central effects or side-effects of

prostaglandin derivatives the concentrations of drug and
metabolites in the brain itself has to be taken into
account. Nileprost was able to pass the blood/braig
barrier after intravenous injection (Fig. 21, but H-ac-
tivity in the brain of the rat was only about 5 % of the
corresponding plasma levels. Disposition of labelled
compounds was biphasic with half-iives of 17 min and
1.3 h (Tab. II). Due to very low H-activities measured
in the brain, metabolic patterns were evaluated only
5 min post administration. The relative amount of un-
changed drus was about double that in the plasma. The
absolute concentration was calculated to be 3 rig/gas
compared to 46 ng/ml measured in the plasma.

I oawle blood olssma brain

i.v. in.lection
tl,2 (a) (0 - 30 min) 15 min 17 min

tl,2 (0) (until 4 h) 1.1 h 0.9 h 1.3 h

tl,2 OJ) (until 24 h) 7h 11 h

p.0. erlninistrstion

$2 Cabs. 1 27 min 22 min

%/2 (-1
tl,2 (0) (until 4 h) 1.8 h 1.5 h

tl,2 Cy) (until 24 h) 8h 7h

3
Tab. II: Half-lives Df absorption and disposition of H-activity in blood, plasma
and brain of female rats after doses of 200 vg/kg.

247
Following oral treatment labelled compounds could be
detected as well. Their concentration, however, never
exceded 1 rig/g..The half-life of disposition was calcula-
ted to be 7 h (Table II).

Whole-body autoradiography

Autoradiographs of pregnant rats after intravenous and

oral administrations are illustrated in Fig. 3 and 4.
Substantial concentrations of radio-labelled compounds
could be detected only in the excretory organs liver and
kidney and in the stomach, especially the stomach wall,
after both ways of treatment. The lungs were slightly
visible after intravenous injection of the drug. Neither.
the brain nor the placenta sr the fetuses were clearly
detectable. Elimination of H-activity was very rapid.
24 h after intravenous and oral administration labelled
compounds were measurable only in the intestine as pre-
formed feces.

Fig. 3

Autoradiograghs of pregnant rats after intravenous doses

of 20 pg of H-nileprost.

248
Fig. 4

Autoragiographs of pregnant rats after oral doses of

20 pg H-nileprost.

Excretion with urine and feces

The main route of excretion of nileprost and its metabo-

lites in the rat was biliary elimination. There were,
however, significant differences between male and female
animals. After intravenous injection of 200 wg/kg to
female rats 27 % of the dose was recovered in the urine
and 59 % of the dose in the feces (Tab. III). Male
animals, on the other hand, excreted 6 % via the kidneys
and 72 % via the bile. Carcasses contained 0.3 % of the
dose in both cases. After oral treatmegt female animals
again had a higher renal excretion of H-active sub-
stances (Table III).
The rate of elimination was relatively high. More than
97 % of urinary excretion was completed within 24 h.
Half-lives, therefore, were not calculated from these
data. Elimination with feces was biphasic with half-lives
of 8 h (1 - 2 d post administration) and 2 d (3 - 8 d
post administration).
Excretion with urine and feces was complete. 10 d after
administration the carcasses contained less than 0.5 % of
the administered dose (Table III).

249
 recovery iflt rwenamly OrfIlly
(X of dose) feamle lmle female male

urine 27.4 + 3.4 5.5 + 0.6 12.1 + 2.8 4.0 2 2.5

feces 59.2 + 3.9 72.0 + 6.7 66.8 + 9.0 77.5 + 12.0

CBCCBSB 0.3 + 0.0 0.3 2 0.0 0.1 + 0.0 0.2 L 0.0

tot51 06.9 + 4.6 77.0 2 6.7 79.1 + 8.6 81.7 + 11.5

Tab. III: Recovery of 31i-activity in urine, Feces and C~PCBBS~Sof male and female rats
ten dsys after i.v. and p.o. achinistretiona of 200 &kg.

The metabolic patterns of urine after both ways of

administration as determined from the urine of female
animals exhibited substantial amounts of unchanged drug
(Fig. 5) - 10 % of the total activity after oral treat-
ment and 36 % after intravenous injection. About 10 meta-
bolites could be detected. The main metabolite after oral
dosing was no. 4 achieving about 28 %, while after
intravenous administration this compound hardly had 3 %
of the total activity. Metabolite no. 10 amounted to 15 %
and 18 %, respectively, after p.o. and i.v. treatment.
The other compounds exhibited practically identical
relative amount after both ways of dosing.

Biliary excretion and intestinal ra-absorption

Renal and biliary excretion data obtained from female
bile-duct cannulated rats supplied with urinary catheters
al-esummarized in Table IV. In these animals elimination
 2 0

8
odadmhmab

2 0

Fig.5

Metabolic pattern of 3H-nileprost in 24 h-rat urine after

intravenous and oral. doses of 200 erg/kg.

via the kidneys was only about l/10 as compared to intact

rats suggesting considerable enterohepatic re-circula-
tion. This could be demonstrated in a second study, when
radioactive bile of the first group was administered
intraduodenaliy to, another group of animals. About 60 %
of the total H-recovery was measured in urine and bile.

251
Fig. 6
3
Metabolic pattern of H-nileprost in rat bile after
intravenous injection of radioactive drug (upper part)
and after intraduodenal administration of radioactive
bile (lower part).

The rest was un-absorbed material in the GI tract. It is

suggested that mainly unchanged drug was re-absorbed from
the intestine, as thearelative amount of nileprost in the
bile amounted ,to 60 % (Fig. 6).

252
The metabolic pattern of the bile obtained after intrave-
nous injection of drug is practically identical to that
after intraduodenal administration of bile - exept for
metabolite no. 10, which is missing in the bile collected
first. This compound, therefore, had been formed exclu-
sively in the gut wall, as it is detected only in those
cases, when the drug molecule had passed the gut wall
before. This is true for oral as well as for intravenous
administration because of the extensive enterohepatic
re-circulation of the drug.
However, whenever this passage of the gut wall was not
possible - for example in bile-duct cannulated animals -
this metabolite was not observed.
The rates of biliary excretion are summized in Table V.
Half-lives of 0.5 h and 4 h, were calculated for the two
phases observed.

r-
a&ninistration drug i.v. drug i.v. blla i.d.
L
pretreetnlent bile duct and urinary bladder
cannulation
r
urine (% of Qse) 21.4 + 3.4 3.9 2 2.3 6.6 + 7.6
bile (% of dose) 03.6 2 5.9 40.3 2 14.5

i feces (Z of dose) 59.2 + 3.9 0.3 2 0.2 1.4 2 2.3

camas (S of dose) 0.3 + 0.0 1.2 + 0.2 35.6 2 11.9

I +
total (X of Qse) 06.9 + 4.6 89.0 + 6.5 03.9 2 7.5
1
Tab. IV: Recovery of 3!+ectivity in urine, bile, feces and carcaeeee of female rats after intravenous
injection of H-nileproat end intraLodma1 a&inistration of radioactive bile.

ackeiniatration drug i.v. drug i.v. bile i.d.

pretreatmmt bile-&t and urinary bladder

cafnwlation

27 + 3 min
$2 CCL)

Bh 3.6 1: 0.6 h 4.8 2 1.2 h

$2 co)

%/2 ‘y’ 2d

Tab. V: Half-livea of biliary elimination of ‘H-active compounds in female rate in comperison to

fecal excretion.

253
 DISCUSSION

In the present study thg synthesis of a radioactive

prostacyclin analogue, H-nileprost, and its application
to the pharmacokinetics and biotransformation of this
compound in the rat has been described. Nileprost was
tritium-labelled in positions 18 and 19 and this label
proved to be metabolically stable in the rat. Onsy 3 % of
the radioactive marker was found in the form of H 0.
Enzymatic degradation by&-oxidation, therefore, c3 nnot
be a major pathway in the metabolism of nileprost in the
rat. This is in contrast to the results obtained with
prostacyclin, where about 19 % of the total radioactivity
excreted via the kidneys was due to a &-oxidation product
(12, 13).

Nileprost was absorbed in the rat after oral administra-

tion of 200 bg/kg with a half-life of 22 min reaching
maximum concentrations after about 15 min. This is
identical to tlLgt = 23 min and t(max) = 2 h as obtained
with clloprost er the same oral dose (13). Disposition
of radioactive compounds from the plasma as calculated
after intravenous injection was three-phasic with half-
lives of 15 min, 0.9 h and 11 h, respectively. With
ciloprost a two-phasic decline with half-lives of 0.7 h
and 24.5 h was found (11). Whereas with nileprost,
however, a rise in plasma levels was observed 1 h after
intravenous injection, this was not measured with
ciloprost. The reason for rising plasm3 concentrations
may be enterohepatic recirculation of H-active
substances. The degree of gastrointestinal reabsorption
is identical for both prostacyclin derivatives (60 %). In
the case of nileprost, however, it is probably only
unchanged drug that is absorbed from the intestine,
whereas after ciloprost treatment only metabolites are
reabsorbed. The volume of distribution of metabolites
normally is lower than that of unchanged drug, and so a
rise in plasma levels should have been more likely to
occur with ciloprost and not with nileprost. The reason
for this difference is in the absolute amount of
re-absorbed substances. As the relative degrees is
identical for both prostacyclin analogues (141, the
absolute amount is much higher in the case of nileprost,
because this compound is excreted mainly by biliary
elimination, whereas ciloprost mostly via the kidneys
(15).
The half-life of unchanged nileprost in the plasma was
found to be 14 min as compared to about 5 min obtained
with ciloprost. Nileprost, therefore, is metabolically
more stable than ciloprost. The relatively short

254
half-life of 14 min, on the other hand, is not only an
expression of rapid biotransformation but of very fast
biliary excretion. This is3clearly shown by the half-life
of biliary elimination of H-active substances, which was
calculated to be about 0.5 h.
Nilaprost is able to pass the blood/brain barrier. The
concentrations in the brain, however, are very low. They
are of the same order of magnitude as observed with
Siloprost and prostacyclin (16). The low levels of
H-activity in the brain can also be seen from the
autoradiographs presented.
Nileprost was excreted mainly by biliary elimination.
Ciloprost, on the other hand, has been reported to be
eliminated mainly by urinary excretion (15). The metaboli.
tes of prostacyclin were measured in urine and feces in
amounts of 33 % and 44 % of dose, respectively (16).
Ciloprost and prostacyclin are totally metabolized in the
rat and practically no unchanged drug is found in the
urine (11, 15). Nileprost, therefore, has to be consi-
dered as being metabolically far more stable than these
two compounds. This stabilization of the prostacyclin
molecule has been obtained by the introduction of a cyano
group at position 5 and a methyl group at position 16.
Due to the cyano group the opening of the prostacyclin
ring system yielding 6-keto PGF is not possible any
longer. The methyl group probab$$ protects from oxidation
of the 15-hydroxyl function and subsequent hydrogenation
of the 13,14-double bond. Therefore, the modifications of
the prostacyclin molecule as found in nileprost, seem to
be a successful1 starting point for further investiga-
tions in this area.

Acknowledgement
The authors thank Dr. W. Skuballa for the synthesis of
compound I and Mr. W. Schwartz for the preparation and
purification of radioactive nileprost.

255
 LITERATURE

(1) Johnson R.A., Morton D.R., Kinner J.H., Gorman

R.R., McGuire J.C., Sun F.F., Whittaker N.,
Bunting S., Salmon J., Moncada S., Higgs E.A. and
Vane J.: The chemical structure of prostaglandin X
(prostacyclin). Prostaglandins -12: 915, 1976.

(2) Moncada S., Higgs E.A. and Vane J.R.: Human

arterial and venous tissues generate prostacyclin
(prostaglandin X) a potent inhibitor of platelet
aggregation. Lancet 1: 18, 1977.

(3) Dusting G.J., Moncada S., Vane J.R.: Prostacyclin

(PGX) is the endogenous metabolite responsible for
relaxation of coronary arteries induced arachi-
donic acid. Prostaglandins -13: 3, 1977.

(4) Moncada S., Gryglewski R.J., Bunting S. and

Vane J.R.: A lipid peroxide inhibits the enzyme in
blood vessel microsomes that generates from
prostaglandin endoperoxides the substances (prosta-
glandin X) which prevents platelet aggregation.
Prostaglandins -12: 715, 1976.

(5) Robert A.: Current history of cytoprotection.

Prostaglandins 21 (Suppl.), 89, 1981.

(6) Spannhake E.W., Hyman A.L. and Kadowitz P.J.:

Bronchoactive metabolites of arachidonic acid and
their role in airway function. Prostaglandins 22
(6), 1033, 1981.

(7) Dusting G.J., Moncada S. and Vane J.R.:

Disappearance of prostacyclin in the circulation
of the dog. Br. J. Pharmacol. 65: 414 P. 1977.

(8) Schror K., Darius H., Matzky R. and Ohlendorf R.:

The antiplatelet and cardiovascular actions of a
new carbacyclin derivative (ZK 36 374) -
equipotent to PGI in vitro. Naunyn Schmiedeberg's
Arch. Pharmacol. 3 16, 252, 1981.

(9) Haberey M., Maa B., Mannesmann G., Skuballa W.,

Town M.-H. and VorbriiggenH.: Cardiovascular
properties of ZK 36 374, a novel, stable
prostacyclin derivative. Therapiewoche 30 (47),
7860, 1981.

256
(10) Ull&rg S.: Studies on the distribution and fate
of S-labelled bynzylpenicilline in body. Acta
Radiol. 118, 1954.
(11) Krause W. and Schubert M.: Pharmacokinetics and
biotransformation of the stable prostacyclin
analogue, ciloprost. II. Blood and plasma levels
and passage of the blood/brain barrier in the rat
Submitted for publication.

(12) Sun F.F. and Taylor B.M.: Metabolism of prosta-

cyclin in rat. Biochemistry 17 (191, 4096 (1978).

(13) Sun F.F., Taylor B.M., Sutter D.M. and Weeks J.R.
Metabolism of prostacyclin. III. Urinary
metabolite profile of 6-keto-PGF,, in rat.
Prostaglandins 17 (51, 753 (1979).

(14) Krause W., in preparation

(15) Krause W., Skuballa W. and Schulze P.E.:

Pharmacokinetics and biotransformation of the
stable prostacyclin analogue, ZK 36 374.
I. Synthesis of a tritium marker and excretion of
H-ZK 36 374 in the rat. Submitted for publication

(16) Taylor B.M. and Sun F.F.: Tissue distribution and

biliary excretion of prostacyclin metabolites in
the rat. J. Pharm. Exp. Ther. 214 (1): 24, 1980.

257