Interferon kinetics and adverse reactions after intravenous,

intramuscular, and subcutaneous injection

Three groups of six subjects each received a single 36 x 106 U dose of recombinant leukocyte A
interferon (rIFN-a.4) as a 40-min infusion, an intramuscular injection, or a subcutaneous
injection. Blood samples were collected at specific times after dosing for analysis of rIFN-aA
serum concentrations by an enzyme immunoassay method, ELISA. The rIFN-aA was rapidly
distributed and moderately eliminated (t1/2 = 5.1 hr) after intravenous infusion. The maximum
concentrations at the end of intravenous infusion were tenfold the maximum concentrations after
intramuscular and subcutaneous injections. Renal tubular secretion or extrarenal elimination
was suggested by clearance values of 1.8 times the glomerular filtration rate. After
intramuscular and subcutaneous injection, rIFN-aA was absorbed slowly (time to reach
maximum concentration ranged from 4 to 8 hr), which resulted in prolonged serum
concentrations. Estimated bioavailability was more than 80% for both intramuscular and
subcutaneous injection shares qualitatively the same adverse reactions, the reactions differ in
severity and duration. The adverse effects appear to be related to route of administration.
of herpes labialis were also noted. There were no significant clinical laboratory abnormalities of
medical concern. Although rIFN-cxA injected by intravenous infusion or intramuscular or
subcutaneous injection shares qualitatively the same adverse reactions, the reactions differ in
severity and duration. The adverse effects appear to be related to route of administration.

Robert J. Wills, Ph.D., Susana Dennis, M.D., Herbert E. Spiegel, Ph.D.,

Donald M. Gibson, B.S., and Paul I. Nadler, M.D. Nutley, N. J., and Madison, Wis.
HoffmanLa Roche Inc., Nutley, and Hazleton Laboratories America, Inc., Madison

Recombinant leukocyte A interferon (rIFN- fections. Preliminary evidence of antitumor ac-

aA) is a single protein moiety of human a-
interferon derived by recombinant DNA tech-
niques" that is being evaluated in disseminated
cancer and viral diseases. An early tolerance
and efficacy trial with single intramuscular
doses ranging from 3 to 198 x 106 U rIFN-aA
in patients with cancer reported acute side ef-
fects resembling those in patients with viral in-
Received for publication Sept. 20, 1983; accepted Dec. 22, 1983.
Reprint requests to: Dr. Robert J. Wills, Department of Pharma-
cokinetics, Biopharmaceutics, and Drug Metabolism, HoffmanLa
Roche Inc., Nutley NJ 07110
tivity was also reported in this study .1-6 More
recently, rIFN-aA in doses ranging from 3 to
136 x 106 U was injected intramuscularly three
times a week for 28 days in patients with dis-
seminated cancer." Side effects similar to those
after single doses were noted. In both the sin-
gle- and multiple-dose trials, only preliminary
kinetic data were obtained after intramuscular
injection .8 Definitive kinetic evaluations of
rIFN-aa after intravenous infusion and subcuta-
neous and intramuscular injections have not yet
been completed. The purpose of our study was
to determine the kinetics of rIFN-aA after a

722
Volume 35
Number 5
Table I. Adverse reactions after a 36 x 106 U dose of rIFN-aA
Route Chills Headache
Intravenous infusion 1/6 5/6
Intramuscular injection 5/6 5/6
Subcutaneous injectiont 4/6 6/6
*Expressed as the number of volunteers with adverse reactions /number of volunteers evaluated at each route of administration.
tDuring infusion.
*Two instances of herpes labialis recurrence.
40-min intravenous infusion and the bioavail-
ability of rIFN-aA after intramuscular and sub-
cutaneous injection.
Methods
Our subjects were 18 healthy men between
19 and 36 yr of age (X =26) and weighing
between 58.1 and 81.6 kg (X = 74). Their good
health was determined by medical history,
physical examination, and clinical laboratory
tests that included a hematologic examination,
urinalysis, and blood chemistry tests. Volun-
teers with a history of gastrointestinal, renal,
hepatic, pulmonary, cardiac, hematologic, or
endocrinologic disease were excluded. Volun-
teers who received any form of medication
within 2 wk of the study or had a history of drug
addiction or alcohol abuse were also excluded.
Twelve hours before starting, all subjects
were confined to the study area. A light snack
was served 10 hr before dosing, after which an
absolute fast was maintained. In the morning,
three groups of six subjects each received
36)< 106 U rIFN-aA as a 40-min intravenous
infusion, an intramuscular injection into the
gluteus muscle, or a subcutaneous injection into
the forearm. In addition, each subject received
650 mg acetaminophen at the time of dosing
and again every 4 to 6 hr for 24 hr to ameliorate
the expected febrile effect. No food or fluids
were taken until the 4-hr blood sample was col-
lected, after which a meal was served. All sub-
jects were confined to the study site until the
36-hr blood sample was drawn. Blood was
drawn before (7 ml) and after the start of dosing
(4 ml) at 30 mm and 1, 1.5, 2, 3, 4, 5, 6, 7, 8,
12, 24, 36, and 48 hr. Additional samples were
drawn at 10, 20, 40, and 50 min after starting
intravenous infusion only. Finally, a 3-ml blood
Interferon kinetics and adverse reaction 723
Adverse reaction*
Myalgias Malaise Nausea Diaphoresis Syncope
3/6 1/6 0/6 1/6t 1/6t
3/6 4/6 0/6 0/6 0/6
5/6 2/6 4/6 2/6 2/6
sample was drawn for antibody testing before
and 1 wk after dosing. Collected blood was
centrifuged and the serum was removed and
stored frozen at 20° until assayed.
Human rIFN-aA was measured by an en-
zyme immunoassay method with a solid-phase
sandwich principle.9 Incubation at room tem-
perature binds rIFN-aA to a polystyrene bead
coated with mouse monoclonal rIFN-aA anti-
body. Binding was subsequently effected with a
second monoclonal mouse antibody with spec-
ificity for a second epitope on rIFN-aA. This
second monoclonal antibody was conjugated to
horseradish peroxidase. After this incubation
step, unbound material was removed by wash-
ing and the activity of peroxidase bound to the
bead was measured by 0-phenylenediamine as
substrate. The resulting color intensity (mea-
sured photometrically) is directly proportional
to the rIFN-aA concentration in the sample.
The reference standard had a specific activity of
1.7 x 109 U/mg protein as determined against
the National Institutes of Health (NIH) inter-
feron standard. The assay sensitivity in serum
was 20 pg/ml rIFN-aA.
Body temperature was recorded at 2, 3, 4, 5,
6, 7, 12, 24, 28, 31, and 35 hr after dosing.
Analysis of variance was performed on the body
temperature data. In addition, subjects were
queried for adverse effects at 2, 6, 25, 31, 36,
and 48 hr after dosing.
The dose (in units) was converted to pico-
grams by multiplying by a factor of 6.0 as de-
termined from the NIH reference standard. The
maximum concentration (Cmax) and the time of
maximum concentration (tmax) were read di-
rectly from the serum concentrationtime data.
Cmax after intravenous infusion was taken as the
concentration at the end of the infusion, tmax ,
724 Wills et al. Clin. Pharmacol. Ther.
May 1984

Table II. Kinetics determined from plasma rIFN-aA concentration-time data

Subject Weight C max tmax AUMC°-" t1/2 Vd
No. (kg) (pg 1ml) (hr) (pg hr 1ml) (pg hr21m1)* (hr) (1)

Intravenous infusion
2 75.7 17,600 19,700 38,600 5.7 21.5
4 81.6 10,700 14,700 35,500 6.1 35.3
9 73.0 15,500 23,000 73,000 8.5 29.8
10
14
17
74.8
70.8
84.4
16,900
13,900
8,570
- 13,300
21,800
13,200
18,500
34,700
50,900
3.8
5.6
3.7
22.6
15.8
63.1
X 13,900 17,600 41,900 5. it 31.4
±SD 3,570 4,430 18,400 17.0
Intramuscular injection (f = 0.83)$
5 79.8 1,500 4.0 11,200 3.5
6 82.6 2,040 3.0 8,960 1.5
8 58.1 2,580 6.0 17,600 2.6
12 70.8 1,960 4.0 11,700 2.9
16 82.1 1,880 3.0 11,300 1.5
18 69.9 2,130 3.0 26,500 4.5
R 2,020 3.8 14,600 2.3t
±SD 352 1.2 6,540
Subcutaneous injection (f = 0.90)$
1 66.2 2,320 8.0 25,200 3.3
3 68.5 1,720 7.0 16,100 2.6
7 73.5 1,540 7.0 13,500 3.9
11 68.9 2,280 6.0 18,700 3.5
13 74.4 1,250 8.0 10,700 4.4
15 78.5 1,260 8.0 11,200 3.7
3C- 1,730 7.3 15,900 3.5t
±SD 477 0.8 5,466
*Corrected for increase in mean residence time.'s
tHarmonic mean t1/2.
*Determined from mean data and used in determination of apparent Vd and CI,.

and the elimination rate constant (J3) was de-
termined by fitting the individual data with
NONLIN." The 3 values after intramuscular
and subcutaneous injection were determined by
fitting the individual data from the terminal
portion of the concentration-time profiles to a
log-linear regression equation by the method of
least squares. The terminal t1/2 was calculated
by dividing 0.693 by (3. The AUC from time
zero to the time of the last measurable concen-
tration point was calculated by trapezoidal sum-
mation. Extrapolation to time infinity (AUC')
was determined by dividing the last measurable
concentration point by /3. The area under the
plasma concentration-time moment curve
(AUMC') was calculated by trapezoidal
summation and extrapolated to time infinity by
the addition of the last measurable moment di-
vided by and the last measurable concentration
divided by 13-squared." The volume of dis-
tribution at steady-state (Vd) was determined
by multiplying the dose (D) by AUMC°-'
and dividing by AUC°-'-squared.1 The Vd
after intravenous infusion was corrected for the
increase in mean residence time as AUMC°-'' -
[(AUC''/2) x infusion time].'5 Total body
clearance (Clb) was calculated as Vd x
(AUC('''/AUMC") after intravenous infu-
sion. The fraction of D systemically absorbed (f)
after intramuscular and subcutaneous injections
was estimated from mean data because separate
subjects were used in each treatment.
Results
Many of the common side effects of rIFN-aA
appeared during the study and are listed by
Volume 35 Interferon kinetics and adverse reaction 725
Number 5

40

395
Clb Vd CIB 39
(ml/mm) (1/kg) (mIlminIkg)
38 5

183 0.284 2.42 37

245 0.435 3.00
37
157 0.408 2.14
271 0.302 3.62 36.5

165 0.223 2.34 36° 1 1 1 1 I 1 1 1 I 1

4 8 12 16 20 24 28 32 318
273 0.748 3.23 TIME IN HOURS SINCE FIRST OBSERVATION

216 0.400 2.79

53.5 0.188 0.58 Fig. 1. Mean change in body temperature (°C) after a
single 36 x 106 U dose of rIFN-aA given as an intra-
venous infusion (dashed line) or an intramuscular
(dotted line) or subcutaneous (solid line) injection.
The symbol X indicates temperatures significant dif-
ferent (P < 0.05) from the temperatures after intra-
venous infusion.

Parameters other than Cmax, AUC', and

route of administration in Table I. The observ-
able side effects were more prevalent and longer
in duration after intramuscular and subcutane-
ous injection than after intravenous infusion.
Elevated temperatures were common after rIFN-
aA was administered by all routes; these data
are summarized in Fig. 1. Body temperatures
were significantly higher after intramuscular
and subcutaneous administrations than after in-
travenous infusion between 3 and 7 hr of dos-
ing. The maximum (X SD) body tempera-
tures were 37.6° ± 0.6°, 39.1° ± 0.7°, and
40.0° -± 0.7° for intravenous infusion and in-
tramuscular and subcutaneous injections. None
of the clinical or laboratory findings resulted in
significant sequelae.
Mean serum rIFN-aA concentrations in six
subjects per route of dosing are plotted in Fig.
2. Kinetic parameters from the individual serum
concentration-time data are listed in Table II.
AUMC° after intravenous infusion were de-
termined by estimates from the NONLIN fit. 12
Serum rIFN-aA concentrations were measurable
through 24 hr for all routes, but the shapes of
these curves were quite different among the three
routes. Serum concentrations increased rapidly
during the 40-mM intravenous infusion. Cmax at
the end of the infusion ranged from 8570 to
17,600 pg/ml. Concentrations then declined
rapidly in a biexponential manner. Serum rIFN-
aA concentration fell from 1/20 to 1/30 its original
level within 3 to 4 hr of stopping the infusion.
This rapid disposition was followed by an appar-
ent terminal elimination phase, with Ph ranging
from 3.7 to 8.5 hr = 5.1 hr). The AUC''''
ranged from 13,200 to 23,000 pg hr/ml. The
Vd ranged from 0.223 to 0.748 //kg and Clb
ranged from 2.14 to 3.62 ml/min/kg (Table II):
In contrast, the intramuscular and subcutane-
ous routes exhibited protracted absorption, which
resulted in slower rises to lower Cma. values.
After intramuscular injection, C,ax ranged from
1500 to 2580 pg/ml at tmax ranging from 3 to 6
hr. The concentrations increased, then exhibited
a short plateau and subsequently declined
monoexponentially. Serum ti/2 ranged from 1.5
to 4.5 hr and AUC°-'' ranged from 8960
to 26,500 pg hr/ml (Table II). In a like man-
ner, Cmax after subcutaneous injection ranged
from 1500 to 2580 pg/ml at tmax ranging from 6
to 8 hr. The concentration profile resembled that
726 Wills et al.
100,000
10,000
1000
k (Fig. 2). After termination of the infusion,
/ \._
100 P
c
10
0 5 10 20
TIME (hours)
Fig. 2. Mean serum rIFN-aA concentrations after a
single 36 x 106 U dose as an intravenous infusion (o)
or an intramuscular (X) or subcutaneous injection
of the intramuscular injection, differing only in
the time to reach C.a., which caused the elimi-
nation phase of the curve to occur later after
dosing. Serum Ph ranged from 2.6 to 4.4 hr and
AUC° ranged from 10,700 to 25,200 pg hr/
ml (Table II).
Apparent f values of the intramuscular and
subcutaneous doses (with the mean intravenous
data [AUC°-1 as the standard for comparison)
were 0.83 and 0.90. Because a complete
crossover was not accomplished, f was not de-
termined in each subject; hence these mean val-
ues are only approximations of f.
Discussion
Overall, the rIFN-aA kinetics we report are
consistent with published data.s 10 There are
small differences in tihs (4 to 8 and 6 to 9 hr)
between our findings and reported results that
have involved patients with proved dissemi-
nated cancer. Other studies conducted in iso-
lated animal fivers''. 7 and kidneys'. 3' 5' 6 sug-
gest that the kidney is the main site for elimina-
tion of a-interferons. In our study, the mean
total Clb of 216 rt. 53.5 ml/min is about 1.8
times the glomerular filtration rate, indicating
that tubular secretion, renal catabolism, or
extrarenal elimination must also be occurring.
Pharmacol. Ther.
May 1984
It was apparent that the rIFN-aA serum con-
centration profiles varied with route of admin-
istration. Intravenous infusion resulted in earlier
(0.67 vs 3.8 and 7.3 hr) and much higher Cmax
(13,900 vs 2020 and 1730 pg/m1) than those
after intramuscular and subcutaneous injection
serum concentrations declined quickly for 3 to 4
hr before exhibiting an apparent terminal elimi-
nation phase, whereas the continued absorption
from the intramuscular and subcutaneous injec-
tion sites maintained serum concentrations that
exceeded those of the intravenous dose after 3
215
to 4 hr.
There also were distinct differences in the
adverse effect profiles, including increased
body temperature, for the routes of injection.
The clinical adverse experiences were least se-
vere and of shortest duration after intravenous
infusion, whereas they were more severe and of
longer duration after the intramuscular injection
and most severe and of longest duration after
the subcutaneous injection (Table I). The dura-
tion of elevated body temperature showed this
pattern with significantly higher and more sus-
tained elevation after intramuscular and subcu-
taneous injections than after intravenous infu-
sion. If one assumes that there is a relationship
between serum concentrations and adverse ef-
fects, it would appear that these effects are not a
simple function of the magnitude of the serum
concentrations; rather, they appear to be a func-
tion of exceeding and maintaining serum con-
centrations above a threshold level. For exam-
ple, if 66 pg/m1 is arbitrarily chosen as the
cutoff concentration (see Fig. 2) for a given
adverse effect, this adverse effect could be ex-
pected to last for 12 hr after an intravenous in-
fusion and for 24 hr after intramuscular and
subcutaneous injection. If 200 pg/m1 were cho-
sen, the effect could be expected to last between
5 to 6 hr after intravenous infusion, for 12 hr
after intramuscular injection, and for almost 24
hr after subcutaneous injection. These projec-
tions are in good general agreement with the
observed adverse effect profiles in our study
and, therefore, support our conjectures of a
"threshold hypothesis" and the idea that ad-
verse effects, including elevated temperatures,
Volume 35
Number 5
are a function of exceeding and maintaining
rIFN-aA serum concentrations above a certain
serum concentration.
The authors thank Mrs. Vera Kucera and Mr.
Frank Hsieh for technical assistance.
References
Benet LZ, Galeazzi RL: Noncompartmental de-
termination of the steady-state volume of distri-
bution. J Pharm Sci 68:1071-1074, 1979.
Bino T, Edery H, Gertler A, Rosenberg H: In-
volvement of the kidney in catabolism of human
leukocyte interferon. J Gen Virol 59:39-45,
1982.
Bino T, Madar F, Gertler A, Rosenberg H: The
kidney is the main site of interferon degredation.
J Interferon Res 2:301-308, 1982.
Bocci V, Pacini A, Bandinelli L, Pessina GP,
Muscettola M, Paulesu L: The role of liver in the
catabolism of human a- and a-interferon. J Gen
Virol 60:397-400, 1982.
Bocci V, Pacini A, Muscettola M, Paulesu L,
Pessina GP, Santiano M, Viano I: Renal filtra-
tion, absorption and catabolism of human alpha
interferon. J Interferon Res 1:347-352, 1981.
Bocci V, Pacini A, Muscettola M, Pessina GP,
Paulesu L, Bandinelli L: The kidney is the main
site of interferon catabolism. J Interferon Res
2:309-314, 1982.
Bocci V, Russi-Sorce M, Cirri G, Rita G: Role
of liver in the inactivation of interferon. Proc
Soc Exp Biol Med 128:62-65, 1968.
Colburn WA, Spiegel HE, Fein S. Dziewanow-
Interferon kinetics and adverse reaction 727
ska Z: Pharmacokinetics of recombinant leuko-
cyte A interferon following single doses to hu-
mans. J Clin Pharmacol Ther 33:250, 1983.
9. Gallati VH: Interferon: Wesentlich vereinfachte,
enzymimmunologische bestimmung mit zwei
monoklonalen antikorpern. J Clin Chem Clin
Biochem 20:907-914, 1982.
10. Gutterman JU, Fein S, Quesada J, Horning SJ,
Levine JF, Alexanian R, Bernhardt L, Kramer
M, Spiegel H, Colburn WA, Trown P, Merigan
T, Dziewanowska Z: Recombinant leukocyte A
interferon: Pharmacokinetics, single-dose toler-
ance, and biologic effects in cancer patients.
Ann Intern Med 96:549-556, 1982.
11. Levy WP, Rubinstein M, Shively J, Del Valle
U, Lai CY, Moschera J, Brink L, Gerber L,
Stein S, Pestka S: Amino acid sequence of a
human leukocyte interferon. Proc Nat! Acad Sci
78:6186-6190, 1981.
12. Metzler CM, Elfring GL, McEwen AJ: A pack-
age of computer programs for pharmacokinetic
modeling. Biometrics 30:562-563, 1974.
13. Riegelman S, Collier P: The application of
statistical moment theory to the evaluation of
in vivo dissolution time and absorption time. J
Pharmacokinet Biopharm 8:509-534, 1980.
14. Sherwin SA, Knost JA, Fein S, Abrams PG,
Foon KA, Ochs JF, Schoenberger C, Maluish
A, Oldham RK: A multiple-dose phase I trial of
recombinant leukocyte A interferon in cancer
patients. JAMA 248:2461-2466, 1982.
15. Straughn AB: Model-independent steady-state
volume of distribution. J Pharm Sci 71:597-598,
1982.