Europ.J.clin.Pharmacol.

8, 249-254 (1975)
© by Springer-Verlag 1975

Pharmacokinetics and Side-Effects of Clonazepam

and Its 7- Amino-Metabolite in Man

O. SjS, E.F. Hvidberg, J. Naestoft and M. Lund

Departments of Neurology and Clinical Chemistry, Glostrup Hospital, and Clinical Pharmacology Research
Unit, Danish State Medical Research Council, Copenhagen, Denmark

Received: August 23, 1974, and in revised form: October 14, 1974

Summary. Clonazepam (CNP) and its principal metabolite in plasma, 7-amino-CNP (ACNP), have been in-
vestigated in a prospective study of 27 newly diagnosed epileptics and correlated with specified
side-effects. At a daily dose of 6 mg, the average plasma levels of both substances were about 50ng/
ml, and individual values ranged from 30 to about 8Ong/ml. There was a linear correlation between
changes in dose and the resulting plasma levels, which indicates first order elimination kinetics.
Side-effects were frequent, but neither their severity nor their occurrence could be related to
plasma levels or to the rate of increase in plasma concentration of the drug. Three out of five pa-
tients who developed serious dysphoria had significantly high CNP levels. The concentration of ACNP
was considerably increased in four patients who subsequently suffered from withdrawal symptoms. Drug
interaction with diphenylhydantoin, i.e. decreased CNP level, was observed in all five patients who
received hoth compounds. In general it is not yet possible to define an upper limit for the plasma
levels of CNP and ACNP at which toxicity occurs. In patients treated with conventional doses of CNP,
measurement of plasma concentration is not required, except in special circumstances, because of the
lack of correlation between plasma level and side-effects.

Key words. Clonazepam, 7-amino-clonazepam, pharmacokinetics, side-effects, man.

Clonazepam (CNP), 5-(2-chlorophenyl)-l,3 dihydro-
7-nitro-2H-benzodiazepin-2-one (RivotrillR), has
an anti-epileptic effect after oral administration.
Its pharmacokinetics in man have only been studied
in few investigations (Eschenhof, 1973; Huang et
al., 1973; Naestoft et al., 1973), of which the
last two alone were concerned with the relation
between plasma level and effects. The occurrence
of side-effects has not previously been correlated
with the kinetics of CNP.
The most prominent side-effects are drowsiness
and disturbances of coordination, often seen
shortly after the start of medication and usually
at a time when the dose is being increased. Thus,
Elian et al.(1973) suggested that the rate at
which the dose of CNP was increased could be cor-
related with the occurrence of its side-effects.
In order to clarify the pharmacokinetics of this
drug and to define the minimum plasma level at
which signs of toxicity would appear during the
usual dosage regimen, a prospective phase 2-type
trial was undertaken. It was based on gas-chroma-
tographic assay of CNP and its major metabolite,
7-amino-clonazepam (ACNP; Larsen and Naestoft,
1974).
Material and Methods
Patients and Design of the Study
27 patients (I0 female and 17 male) with newly
diagnosed epilepsy were studied; their ages ranged
from 12-71 years, average 39 years. As the type of
seizure was considered of little importance in this
study, patients with any kind of epileptic fit
were accepted. Patients in whom an intracranial
tumaur was suspected were excluded. A routine
clinical and biochemical screen showed that none
suffered from any other disease. None of the
patients had been treated before with anticon-
vulsants, no other medicine had been given three
250
to four weeks prior to the study, nor was any
given during the period of the investigation. In-
formed consent was obtained from all the patients,
After initial evaluation in hospital, the pa-
tients stayed there for three weeks. They were.
then kept under observation in the out-patient
clinic for epileptic patients (Department of Neu-
rology) for a period of six to twelve months. The
dose of CNP they received was increased during the
first two weeks in hospital, from 0.5 mg to a
maximum of 6.0 rag CNP per day (Fig. I), which they
took for the remainder of the trial. It was admin-
istered in three daily doses at 8 a.m., 12 noon
and 10 p.m. In some patients the dose had to be
adjusted after several weeks, either because of
side-effects or continued seizures, The adjustments
enabled a correlation to be made between changes
in dose and steady-state plasma concentration.
During the stay in hospital blood samples were
taken three times a week, just before the morning
dose. At the same time side-effects were recorded
(see below).
During the ambulant period, measurement of
plasma levels and recording of side-effects were
done at intervals of two to four weeks, and later
up to twelve weeks. During this period the blood
samples were also obtained just before the morning
dose was given in the clinic.
In order to estimate the variations in plasma
levels between two doses, blood samples were taken
at intervals of 30 to 60 min for a period of eight
hours, after the morning dose on the nineteenth
day of treatment. On the same day urine was collec-
ted for 24 hours from six of the patients; the
completeness of collection was checked by deter-
mination of urinary creatinine content.
Analyses of CNP and its Metabolites in Plasma
The method used was capable of measuring CNP and
several of its metabolites in plasma after a
therapeutic dose. However, only the 7-amino-meta-
bolite was found in appreciable amounts in plasma.
The compounds were extracted with ethyl acetate
from a 2 ml plasma sample. After evaporation of
the organic phase, the residue was dissolved in
hydrochloric acid and washed with hexane. After
neutralization CNP was extracted with toluene and
the metabolites with a mixture of toluene and
ethyl acetate (I:i). The extracts were concentrated
by evaporation and injected into a gas-chromato-
graph equipped with an electron capture detector.
The internal standards employed were Ro 5-4435 for
CNP and acetylated Ro 1043384 for the metabolites.
The method was checked by control measurements of
some plasma samples by mass fragmentography.
In urine CNP, 7-amino-CNP, 7-acetamino-CNP and
their 3-hydroxy-derivatives were measured by a
slight modification of the plasma assay.
The technical details of the analytical proce-
dure have recently been described by Naestoft and
Larsen (1974).
Record of Side-Effects
The following side-effects were sought by exami-
nation and interviews, according to a special
questionnaire: I) Drowsiness; 2) Incoordination
of gait, both according to the patient's own
complaints and as observed by the phycician and
nurses; 3) Euphoria; 4) Dysphoria; 5) Irritability;
6) Changes in libido and sexual potency; 7) Im-
pairment of memory; 8) Visual disorders; 9) Hypo-
tonia; I0) Dizziness; and, II) Abnormal thirst.
The side-effects were scored in the following
manner: I) Drowsiness and 2) Incoordination in
gait on a scale of 0-3, both from the patient's
own complaints and from observed side-effects.
Items 3) - ii) were scored as 0 or I, according
to their absence or presence.
The study was not carried out as a blind trial,
so considerable bias and influence of placebo-
effects on the results must be suspected. In order
to avoid at least some of these effects, the pa-
tients were also evaluated for side-effects before
medication and always by the same physician (O.S.).
Furthermore, the results of the plasma analyses
were not made available to him until after the
trial had been concluded.
Statistical Methods
The main problem, the correlation of slde-effects
with plasma concentration, was examined by a
sequential rank-correlation test, constructed on
the basis of simulation results, and by a sequen-
tial analogy to the Wilcoxon-Mann-Whitney test
with the Lehmann alternative. The remaining prob-
lems were examined by testing a large number of
combinations of variables by non-parametric fixed
sample tests: Spearman rank-correlation test, Krus-
kal-Wallis rank-variance analysis, Wilcoxon-Mann-
Whitney test, chi-square test and the Friedman
rank variance analysis. An IBM 1800 computer was
used.
CNP ng/ml
ACNP ( in ng CNP e q v . / m l )
L11i1111
70-
60' ~--xc.. |I
50- .,.~ IIiiIHIIllII
40"
30.
20. ~.~.~t'" ll¢l/t/iI//t
10" ~..:.~ IIIIHIIIIIII
2 4 6 8 10 12 14 16 1'8 2'0 29-43days
CNP DOSE 2 per day
Fig. I. Average plasma concentration of CNP and
7-amino-CNP in 25 patients during increasing doses.
The concentration of the metabolite (ACNP) is given
in CNP-equivalents to facilitate comparison of the
two substances
 251

Results conjugated forms of 7-amino-CNP and 7-acetamino-

Plasma Levels
The time-courses of the plasma concentrations of
CNP and ACNP in the initial period of CNP treat-
ment are shown in Fig. I. Only 25 patients have
been included, because two dropped out before the
end of the period. Continuous treatment (Fig. 5)
did not seem to alter the plasma level signifi-
cantly.
The time course of the plasma concentrations
during an eight-hour dose interval is shown in
Fig. 2. On average, the peak concentration of CNP
occurred three to four hours after the morning
dose, but in individual patients the time at which
the peak was reached varied from one to five hours
after administration of the drug. The average peak
concentration of the metabolite was at approximate-
ly the same time as that of the parent compound,
and its variation was of the same magnitude. The
ratio between the plasma concentration of CNP and
the metabolite varied from 1:3 to 3:1.
Fig. 3 shows the steady-state plasma concentra-
tion of CNP in 25 patients as a function of the
dose per kg body weight; there is a three-fold
variation in the values.
A significant correlation was found between
changes in the dose of CNP and the alterations so
CNP amounted to 5 to 20 per cent of the dose given.
The 3-hydroxy-derivatives of all three compounds
could only be detected in trace amounts in urine.
Drug Interaction
Five patients'seizures were not satisfactorily con-
trolled by CNP alone. Their treatment was changed
to diphenylhydantoin (DPH) and for a few weeks they
received both drugs in full dosage. During this
period in all five patients a considerable fall in
the plasma concentrations of CNP was noted. The
7-amino metabolite also decreased in the three pa-
tients in whom this compound was measured (Table i).
Side-Effects and Plasma Levels
All patients, except one, complained of side-ef-
fects at various times, most often of drowsiness;
CPN
ng/ml
100-
80.
x
x x

produced in the steady-state levels of CNP (Fig. 4).

60. x x ~ x x
x
Xx
x
x x
Urinary Excretion x
40. x
x
x

The excretion of unchanged CNP in the urine of six x x x

patients was found to be between O.I and 1.4 per 20-

cent of the dose administered, when measured in
24 hour urine samples after the first 19 days of , , , , ;
treatment. The total excretion of CNP and the non- 0 20 20 80 1~ 120 10 ~g/kg

Peroral dose per kg body weight

Pig. 3. CNP plasma level as a function of dose per

CNP ng/ml
kg body weight in steady-state conditions
80. ACNP ( in ng CNP eqv./ml)
x--x CPN
70- ,__, ACPN l
Correlation coefficient : 0.91
60- *40 Regression coefficient : 1.13-+0.21
50-
~ *20
40-
J o
30- n
z -20
o
2o! ._=

lO,

l f
~ -60'
-40
/
800 ;OOl;OO 11oo 12°0 13oo 14oo 15oo' 16oo
2 mg CPN hours

Fig. 2. Variation in average plasma level of CNP -60 -40 -20 0 +20 +40 "
and 7-amino-CNP in 23 patients during an eight- Change in CNP dose per cent
hour dose interval. Vertical lines indicate SD.
The concentration of the metabolite (ACPN) is Fig. 4. Change in CNP plasma level as a function
given in CNP-equivalents (of Fig. I) of change in dose
252

Table I. Plasma levels of CNP and 7-amino-CNP Table 2. Causes of discontinuation of treatment in
during simultaneous treatment with CNP ]3 patients
and DPH
Reason for discon-
Patient Body CNP DPH Days after CNP 7-amino- tinuation
weight dose dose start of ng/ml CNP
kg mg mg DPH ng/ml
treatment

H.J.L. 80 12 400 0 183

7 141 O O ~ ,~ ~ "~

14 54 O ~
>~ -H
~
.~ ~
O "~
~ 0
•~ ~ ~ ~ ~ ~ ~ o
21 81 -~ ~J ~ >'.. "~ ~ ~ 0 0
<~ o~ ~ ~ 0 ~ ~ ~ ~-~
B.M.K. 80 8 400 0 78 73
G.P. 72 F ]] 4,75 X
7 54 40
A.L. 61 F 32 6 X
P.S. 58 6 300 0 65 96
J.W. 30 F 36 4 X
20 27 18 P .M. 29 M 43 8 X
I.H. 53 6 250 0 51 W.J. 49 M 44 6 X
28 28 A.M. 18 F 46 4 X

42 34 L.J. 15 M 7| 6 X
P.S. 41 M 7! 6 X
L.J. 65 6 400 0 50 33
Z.B. 19 F 97 6 X
20 19 iO
B.K. 45 M 147 8 X
H.L. 20 M 176 12 X
disturbances in gait were also complained of by I.H. 30 F 192 6 X X
17 patients. However, in most cases the side-ef-
B.O. 31 M 225 12 X
fects were present to a severe degree only during
the first five to six weeks of treatment. Behav-
ioural changes were noted in several patients.
Some of them were euphoric during the initial pha-
se of treatment and the same patients became CNP ng/rnl
dysphoric later on. Abnormal thirst, visual dis-
orders, dizziness, decreased sexual potency, and 8O 29* 85*
loss of memory were found in 20 -50 per cent of 71"
T
patients. Table 2 shows the 13 patients in whom
60' 192"
treatment was discontinued, in eight of them
because of the side-effects.
-l-
÷
For all patients the occurrence of each indi- 40' •
vidual side-effect, as well as the total score of
side-effects, have been correlated with the plasma
level measured at the same time. Neither during 20'
ii
the phase of increasing dose nor during continuous i-

treatment was there any statistically significant

correlation found. Furthermore, neither the rate
of increase of CNP concentration nor of its meta-
bolite were significantly correlated with the
variations in side-effects (total net scores).
Excessive dysphoria was the reason for stopping
treatment in six of the eight patients in whom
this was done. The dose of CNP was reduced several
times in one of the patients, and her plasma con-
centrations cannot be compared with the rest of
the series. For the remaining five patients this
comparison is shown in Fig. 5. It is apparent that,
at the time of discontinuation, three patients had
significantly higher plasma concentrations of CNP
than the remaining patients.
O
Days 21 22-49 50-79 80-139 140-215
Day of discontinuation
Fig. 5. Plasma concentration of CNP in five pa-
tients in whom medication was discontinued because
of excessive dysphoria (filled circles) compared
with the average plasma concentration in the other
subjects at the same period of treatment (columns).
Horizontal lines indicate the average concentration
in the six patients who make up the fifth column
(i.e. those who continued treatment for at least
140 days). Vertical lines indicate one standard
deviation, n = number of patients

Of the thirteen patients who stopped treatment
for various reasons (Table 2), four developed
pronounced dysphoria, irritability, restlessness,
sleeplessness and tremor of the hands upon ces-
sation of the treatment. Their condition was inter-
preted as withdrawal symptoms. The symptoms dis-
appeared almost immediately, when small doses
(0.25 mg x 3) of CNP were reinstituted. The same
four patients had 3-4 times higher levels of ACNP
in the last blood sample obtained before medica-
tion was stopped than did the remaining eight pa-
tients, who did not experience these specific
symptoms upon withdrawal of CNP. The plasma levels
of CNP itself did not differ.
Discussion
Clonazepam is one of the few benzodiazepine deri-
vatives which exerts an unquestionable anti-epi-
leptic effect after peroral administration (Gastaut
et al., 1970; Eadia, 1973; Lund and Trolle, 1973;
Mikkelsen and Birket-Smith, 1973). There have been
few kinetic studies of CNP in man, mainly because
of the lack of analytical methods. Gas-chroma-
tographic assays are now being published (Huang
et al., 1973; Naestoft and Larsen, 1974), and the
method used in the present investigation has proved
reliable and usable for routine purposes.
No attempt has been made here to study the half-
life of CNP, because of the risk of precipitating
seizures. A few reports (Eschenhof, 1973; Naestoft
et a/.,1973), as well as unpublished investigations
(personal communications), have indicated first-
order elimination kinetics for CNP in man, and a
half-life of 22-44 hours. The first-order kinetics
were confirmed in the present study by the highly
significant correlation between change in dose and
resultant steady-state plasma level of CNP (Fig.4).
According to the present results, CNP is elimi-
nated predominantly by biotransformation, as are
other benzodiazepine derivatives (Mattson, 1972);
less than 2 per cent unchanged CNP was found in
the urine. This is in agreement with the findings
of Eschenhof (1973).
Simultaneous treatment with DPH and CNP appar-
ently lowered the plasma levels of CNP (Table 3),
which might explain the discrepancy between the
plasma levels found by Huang et al. (1973) and
those reported here, after administration of the
same daily dose of CNP. Lower concentrations of
CNP were also found in a group of patients treated
simultaneously with other anti-epileptic drugs
(unpublished observation). If enzyme induction is
the explantion of this phenomenon, some other meta-
bolic pathway than that through the 7-amino deri-
vative is probably involved.
7-amino-CNP was the predominant metabolite in
plasma. Its plasma level was about the same as that
of CNP itself during constant CNP medication; it
appeared very rapidly after treatment commenced
(Fig. I). The kinetics and pharmacodynamic proper-
ties of 7-amino-CNP in man remain to be investiga-
ted.
Several reports about CNP treatment have men-
tioned the relatively high frequency of side-
253
effects as a major problem, although many authors
have claimed that most of the unwanted effects
disappear as treatment is continued. In the present
study, as in many others, the most frequent side-
effects have been various degrees of sedation and
incoordination of gait. Like Bladin (1973), be-
havioural changes were noted in several patients,
but other side-effects were also cormnon in the
present study. It is reasonable to consider that
the frequent interviews and specific examinations
may have influenced the feelings of the patients.
The possibility of exchange of experiences about
side-effects amongst hospitalized patients must
also be considered. These conditions and other
kinds of bias must be taken into account when prob-
lems of side-effects are evaluated, although the
most prominent side-effects were quite characteris-
tic.
Lethovaara (1973) stated that no correlation
appeared to exist between the dose of CNP and the
occurrence of side-effects, as he observed side-
effects in 50 per cent of patients receiving 2 mg
per day, compared with only 25 per cent in patients
on 8 mg per day. In contrast, Elian et al. (1973)
showed that the rate at which the dosage was in-
creased correlated positively with the occurrence
of side-effects.
In the present study, in general it was not pos-
sible to correlate the occurrence or severity of
the various side-effects with the plasma concen-
tration measured at the same time, Nor was a cor-
relation detected between the rate at which plasma
levels increased at the start of treatment and the
occurrence of side-effects. The negative result of
the latter comparison may not be valid for varia-
tions seen when quite different dosage schemes
are employed~ The lack of correlation also applies
to the major metabolite, 7-amino-CNP. It may be
argued that a better correlation between side-ef-
fects and plasma concentration might have been
found later in the day, e.g. around noon, when the
average plasma level was at its peak (cf. Fig. 2).
However, this seems unlikely for several reasons:
first, the pattern and frequency of side-effects
did not vary much throughout the day; and second,
individual variation in the peak level did not dif-
fer from that found in the morning.
Serious dysphoria did appear to be associated
with an elevated level of CNP in some cases. If
this observation is confirmed by prospective
studies, the limit of the plasma level of CNP be-
for toxic effects appear is about 70 ng/ml, at
least in long-term treatment. It is an interesting
observation, but not one readily explained, that
the main metabolite was found in much higher con-
centrations in patients suffering from withdrawal
symptoms. This finding, too, should be thoroughly
investigated as it may be helpful in the predic-
tion and thus the avoidance of withdrawal symptoms
if CNP has to be discontinued.
Except for these special cases, it may be con-
cluded that, if patients are treated with doses of
CNP within the generally accepted limits, moni-
toring of plasma concentrations is not a useful
way to predict or to avoid the side-effects de-
scribed in the present study.
254
Acknowledgements. The statistical analyses were
performed by Engineers M. E. Larsen and S. Larsen,
the Department of Data Processing, Gentofte Hos-
pital, Copenhagen, to whom the authors are most
grateful. Dr. Gordon Johansen made valuable sug-
gestions about the manuscript and Chemical
Engineer N.E. Larsen contributed greatly to the
analyses.
References
Bladin, P.: The use of clonazepam as an anticon-
vulsant - clinical evaluation. Med. J. Aust. I,
683-688 (1973)
Edwards, V.E., Eadie, M.J.: Clonazepam, a clinical
study of its effectiveness as an anticonvulsant.
Pro¢. Austral. Ass. Neurol. 10, 61-66 (1973)
Elian, M., Lund, M. and Melsen, S.: The rate of
dosage increase of clonazepam. Acta neurol.
scand. 49, suppl. 53, 32-35 (1973)
Eschenhof, E.: Untersuchungen Hber das Schicksal
des Antikonvulsivums Clonazepam im Organismus
der Ratte, des Hundes und des Menschen. Arznei-
mittel. - Forsch. 23, 390-400 (1973)
Gastaut, H., Catier, J., Dravet, C., Roger, J.:
Exceptional anticonvulsive properties of a new
benzodiazepine. Mod, Probl. Pharmacopsychiatry:
Epilepsy. 4, 261-269 (1970)
Ghosh, B.K.: Sequential Tests of Statistical Hypo-
theses. London: Addison-Wesley Publishing Com-
pany 1970
Huang, C.Y., Mc Lead, J.G., Sampson, D., Hensley,
J.: Clonazepam in the treatment of epilepsy.
Proc. Austral. Ass. Neurol. 10, 67-74 (1973)
Lehtovaara, R.: A clinical trial with clonazepam
(Ro 5-4023). Acta neurol, scand. 49, suppl. 53,
77-81 (1973)
Lund, M. and Trolle, E.: Clonazepam in the treat-
ment of epilepsy. Acta neurol~ scand. 49, suppl,
53, 82-90 (1973)
Mattson, R.H. In: Antiepileptic Drugs (Ed. Wood-
bury, D.M., Penry, J.K. and Schmidt, R.P.),chap.
57. New York: Raven Press 1972
Mikkelsen, B., Birket-Smith, E.: A clinical study
of the benzodiazepine Ro 5-4023 (clonazepam) in
the treatment of epilepsy. Acta neurol, scand.
49, suppl. 53, 91-96 (1973)
Naestoft, J., Lund, M., Larsen, N.-E., Hvidberg,
E.F.: Assay and pharmacokinetics of clonazepam
in humans. Acta neurol, scand. 49, suppl. 53,
103-108 (1973)
Naestoft, J., Larsen, N.-E.: Quantitative deter-
mination of clonazepam and its metabolites in
human plasma by gas-chromatography. J. Chroma-
togr. 93, 113-122 (1974)
Prof. Dr. E.F. Hvidberg
Clinical Pharmacology
Research Unit
Rigshospitalet
9 Blegdamsvej
DK - 2100 Copenhagen
Denmark