Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
W090/11091 ,~ T/USgOt~1383
FOR~ULATIONS FOR
STABILIZING OF I8M ANTIBODIES
________________ _________ __
Back~round of the Invention
____ ______________________
It is well known that many protein preparations
05 intended for administration to humans require stabil-
izers to prevent denaturation of the proteins,
agglomeration and other alterations to the proteins
prior to the use of the preparation. Many protein
preparations are particularly unstable in dilute
lO solutions. This instability is manifested in the
formation of insoluble particles, and is often increased
when the protein preparation is stored, or shipped.
A major challen-ge -that exists in the field of protein
drugs is in the development of formulations that
15 maintain both protein solubility and activity.
Immunoglobulins, in particular, are recognized as
possessing characteristics that tend to form partic-
ulates in solution, requiring filtration of these
formulations prior to using them for intra~enous in-
20 jection. The formation of protein aggregates andparticulates has long been a problem in the development
of parenteral immunoglobulin products. The admini-
stration of immunoglobulin G (IgG~, for example, was
limited to the intramuscular route be~ause of endogenous
25 anticomplementary activity due to aggregated immuno-
globulin until the recent development of chemically and
enzymatically treated immunoglobulin G. J. E.
Pennington, ~ev. Inf. Dis.,8(4):5371-5373 (1986).
Recent ~odifications in immunoglobulin G formulations
30 have helped to alleviate the problem. J. P. McCue et
WO~0/11091 ~ ? PCT/US90/01383
6~ ~ , ,.
~.
al., Rev. Inf. Dis~, 8(4):5374-5381 (1986). However,
most commercially available formulations now in use
require iltration of the product prior to iniection to
remove these insoluble aggregates or particulates.
05 The immuno~lobulin (IgM) isotype is, the largest of
the immunoglobulins, having a ~olecular wei~ht of
approximately 900,000 daltons. IgM mole!cules tend to be
inherently unstable and precipitate reaclily upon being
subjected to various forms of physical and chemical
10 stress. This characteristic makes the for~ulation of a
stable composition containing IgM intended for
parenteral administration di~ficult.
Summary of the Invention
______ _________________
The invention comprises stabilizing compositions
15 for IgM antibodies. The present compositions contain a
buffer, human serum albumin, sodium chloride, and IgM
antibodies or antibody fragments. The compositions
enhance the stability of IgM antibodies in solution
intended for intravenous administration.
The present compositions can be lyophilized to form
a dry powder. Lyophilization preserves the biological
activity of the IgM antibody, and minimizes formation of
particulates, which can occur in a liquid for~ulation
under physical or chemical stress. The lyophilized
25 product can be readily reconstituted to a particle-free
solution which shows no loss of biological activity, and
which can be adminlstered without prior filtration.
The present liquid and lyophilized formulations
both exhibit superior stabilizing characteristics in
30 terms of minimal protein particle formation, and
. : : : , :
WO90/11091 P~T/US90/01383
~' ^` :1 r~ ~. .~ rl
-3-
preservation o~ immunoreactivity over time, and under
stress conditions such as elevated temperatures, vial
filling and shipping.
The present liquid and lyophilized compositions
05 have boeh been successful in stabilizing IgM antibodies.
The compositions maintain a pareicle-free, stable
solution for injectable monoclonal antibodies and do not
have to be filtered prior to administration. The
lyophilized product, in particular, can be shipped and
stored ~ithout loss of immunoreactivity. Neither
formulation requires refrigeration or other special
handling.
Brlef Descri~tion of the Fi~ures
___________________________ ____
Figure l shows gel filtration HPLC results
15 comparing non-lyophilized and lyophilized/reconstituted
IgM formulations and placebos.
Figure 2 is a diagram of the results of an immuno-
reactivity assay comparing the ability of
non-lyophilized and lyophilized/reconstituted IgM
20 formulations to bind to solid phase lipid A.
Detailed Descri~tion of The Invention
_______________ ____________._________
The compositions of this invention minimize the
~ormation of protein aggregates and particulates in
reagents containing immunoglobulin M (IgM) antibodies
25 and insure that the antibody in solution maintains its
immunoreactLvity over time. The preparation comprises a
sterile, pharmaceutically acceptable solution containing
tromethamine or phosphate buffer, having a neutral or
.
WO90/11091 ~r~:~, PCT/US90/01383
.,, - ' :
~4-
basic p~ (e.g., 6.8 or above), sodium chloride, IgM
antibodies and human serum albu~in.
Buffers ha~e long been used to stabilize the p~ of
antibody products for parenteral injection. Protein
05 solubility in the buffer solution depends upon a number
of factors, such as ionic strength and pH of the
solution.
Buffers which can be used for this formulation
include tromethamine and phosphate buffers ha~ing a
lO neutral or basic pH. Lower p~ formulations showed less
stability, i.e., a higher tendency to form aggregates.
Tromethamine is described in the Merck Index, 10th
edition, ~erck and Co., Inc., Rahway, N.J. The concen-
tration of tromethamine can be from about 5 to about lO0
15 mM, ha~ing a pH from about 8 to about lO.
A phosphate buffer, such as sodium phosphate, can
also be used. A concentration of from about 8:to about
20 mM phosphate can be used in the present composition,
having a pH of rrom about 6.8 to about 7.4.
A stabilizing protein is added to the formulation.
Stabilizing proteins are proteins which increase the
solubility and/or stability of immunoglobulins in
aqueous solutions. For example, when added to an
aqueous solution of im~unoglobulins, these proteins
25 prevent the ammunoglobulins from precipitating out of
the solution, thereby permitting higher concentrations
of immunoglobulins to be solubilized. It has been found
for the present compositions that human serum albumin
(HSA) is a particularly useful stabilizer for IgM for
30 both liquid and lyophilized formulations. HSA is
present in the formulation in an amount of about 2.5 to
'' . ' . ~ '
WO 90/11091 c~ r ,, P.C~ S90/01383
about 10~ by weight per volume. Levels of HSA of from
about 2.5~ (w/v), to about 5~ (w/v), are particularly
effective in maintaining a stable solution of IgM.
In one embodiment of the invention stabilizing
05 reagents for HSA, e.g., sodium caprylate and N-acetyl
tryptophanate, are present in the formation. HSA is
less stable in solution (i.e., more likely to aggregate)
in the absence of these compounds. For example, a 25
solution (w/v) oi` HSA contains 20 mM ~odium caprylate
10 and 20 mM N-acetyl tryptophanate, therefore, 2.5~ (w/v)
HSA added to a formulation includes 2 mM sodium capylate
and 2 mM N-acetyl tryptophanate. Other stabilizing
reagents can be used other than N-acetyl tryptophanate
and sodium caprylate, which are mentioned above for
15 illustrative purposes.
Sodium chloride is added to the present composition
to increase the ionic strength which is required for the
solubility of the IgM proteins. IgM proteins are more
soluble in an aqueous salt solution than in water alone.
20 The amount of sodium chloride added is from about 200 to
about 350 mM. About 270-300 mM sodium chloride is
particularly effective for this purpose.
The present liquid and lyo,ohilized compositions can
be used to stabilize all subclasses of IgM antibodies,
25 as well as I~M. The present compositions are
particularly useful in stabilizing human monoclonal IgM
antibodies.
One embodiment of this invention comprises a
composi~ion containin~ from about 5mM to about 100mM
30 tromethamine (pH 8-10), from about 200 mM to about 300mM
sodiu~ chloride, from about 2.5 to about 5% (w/v) HSA
and from about 2.5 to about 10.Omg/ml IgM antibody.
WO90/1~091 ~ PCT/US90/01383
~ .
-6-
Sodium caprylate in an amount of from about 2 mM eo
about 4 mM, and N-acetyl tryptophanate in an amount of
from about 2 mM eo about 4 mM can, optionally, be
included to stabilize the HSA. A preferred embodiment
05 of the invention comprises about 4.5 mM tromethamiQe (pH
8.5) about 270 mM sodium chloride, about 2.5% (w/v) HSA,
about 5 mg/ml IgM antibodies or antibody fragments, and
about 2 mH each of N-aceeyl tryptophanate and sodium
caprylate. This formulation enhances the stability of
lO immunological activity of the monoclonal antibody, and
prevents the immunoglobulins in solution intended for
intravenous administration to human subjects from
precipitating and forming particulates in the final
product vial. ~
Another embodiment of the present invention com-
prises a composition containing from about 8 mM to about
20 mM of sterile, pyrogen-free sodium phosphate.~pH
6.8-7.4), from about 250 mM to about 350 mM sodium
ohloride, from about 2.5 to about 5~ (w/v) HSA and from
20 about 2.5 to about 10.0 mg/ml IgM antibody or antibody
fragments. Sodium caprylate and N-acetyl tryptophanate
may be included in the formulation in the amount of
about 2 mM to about 4 ~M of each. A preferred
embodiment of this formulation comprises about 8 mH
25 sodium phosphate (pH 7.2), about 270 mM sodium chloride,
about 5.0~ ~w/v) human serum albumin, about 5 mg/ml IgM
an~ibodies or antibody fragments, and about 2 mM each of
sodium caprylate and N-acetyl tryptophanate.
In another embodiment of the present invention, the
30 above formulations can be lyophilized to form a dry,
storable powder, which can be easily reconstituted to a
...... :
,
.
WO90/11091 PCT/US90/01383
~ ?
-7-
particle free solution suitable for intravenous
injection. Lyophilization is a freeze drying process
which is often used in the preparation of pharmaceutical
products to preserve their biological ac~:ivity The
05 liquid composition is prepared, then lyophilized to form
a dry cake-like product. The process generally involves
drying a previously frozen sample in a vacuum to remove
the ice, leaving the non-water components intact, in the
form of a powdery or cake-like substance. The
10 lyophilized product can be stored ~or prolonged periods
of time, and at elevated temperatures, without loss of
biological activity, and can be readily reconstituted
into a particle-free solution by the addition of an
appropriate diluent. An appropriate diluent can be any
15 liquid which is biologically acceptable and in which the
lyophilized powder is completely soluble Water,
particularly sterile, pyrogen-free water, is the
preferred diluent, since it does not include salts or
other compounds which may affect the stability of the
20 antibody. The advanta~e of lyophilization is that the
water content is reduced to levels which greatly reduce
the various molecular events which lead to instability
of the product. The lyophilized product is also better
able to withstand the physical stresses of shipping. The
. 25 reconstituted product is particle free, so it can be
ad~inistered intravenously without prior filtration.
The present invention is further illustrated by the
follow~ng Examples, which is not intended to be limiting
in any way.
,
. :~
WO90/11091 ,~ ~CT/US90/01383
6~ . .
-8-
EXAMPL~ l
_ _ _ _ _ _ __ _
Pre~aration of I~ _Li~_id__n__Lyophilized Formulations
Based On Tromethamine
___________ _________
Lig~id Formulation
05 IgM (HA-IA IgM, lot ~ 012567, Centocor, Inc.,
Malvern, PA) was concentrated to 5.5 mgJml using a
Centriprep 30 Concentrator (Amicon). The concentrated
protein (20 ml) was dispensed into a 25 ml graduated
cylinder and 2 ml of HSA containing sodium caprylate and
l0 N-acetyl tryptophanate (U.S.P. 25~ HSA in 20 mM sodium
caprylate and 20 mM N-acetyl tryptophanate, Armour
Pharmaceutical Co.), tromethamine (50 mM, pH 8.50) and
300 mM sodium chloride (NaCl) were added. The solution
was filtered with a 0.2~ syrin~e filter into a 50 ml
lS centrifuge tube. Sodium azide ( 0.22 ml of l0~
solution) was added to a final concentration of 0.l~.
The final solution, was a clear, light y2110w liquid
composed of 4.95 mg/ml IgM, 45 mM tromethamine (pH
; 8.35), 270 mM sodium chloride, 2.5~ HSA, 2 mM sodium
20 caprylate and 2 mM N-acetyl tryptophanate
A "placebo" formulation was also made, which was
exacely the same as the above formulation except that
the I8M was left out.. Thus, the placebo formulation
contained 45 mM Tris buffer (pH 8.35)~ 270 mM NaCl, 2.5
25 HSA, 2 mM sodium caprylate and 2 mM N-acetyl trypto-
phanate.
Lvu~hilization
The liq~id IgM formulation was dispensed in l ml
increments into 2 ml Type l Tubing vials (~est Co.). A
30 total of 20 vials were filled. The vials were placed in
,
1., i, .
,
WO9~111091 PCT/US90/01383
,~ ~
a lyophiliæer (FTS) havin~ a l' x l' shelf. In order to
generate a full thermal load, the remainder of the shelf
space was loaded with placebo vials.
The vials were capped with 13 mM gray butyl
05 lyophilization closures (#224142, Uheaton). The shelves
of the lyophili7er were prechilled to about 5C, ~ 2C.
The test vials and placebo ~ials were loaded onto a
shelf in a tray, and a sli~ht vaccuum was induced in the
chamber to maintain a good door seal. A eotal of 20
lO vials of product and 355 vials of placebo were filled to
occupy the entire shelf space.
After the vials had reached and maintained 5C for
at least l hour, the shelf surface temperature was set
for about -40~C.- The-vials were-a~owed-to-remain at - `
lS -40C for at least one hour. The condenser was chilled
to about -70C. The pressure in the chamber was reduced
by Means of a mechanical pump to less than 50 Torr. The
shelf surface temperature was re~ulated such that the
product temperature remained between -47C and -42C.
2Q After the product temperature had reached and maintained
the shelf temperature for at least one hour, a mass
spectrum of the residual gasses in the chamber was
rscorded. The shelf surface temperature was then set
for about ~20C. When the temperature reached and
25 maintained +20C for at least 2 hours, the partial
pressures of the residual gasses in the chamber was
recorded.
The chamber was then backfilled with dry nitrogen
to a pressure of about 600 Torr. The product was
30 removed from the dryer and crimp seals were applied to
. .
,
-
.: .
WOg~/11091 PCTtUS90/~1383
1 0 -
the vials. The formulated protein formed a very dense
cake upon freezing.
The lyophilized cakes that were formed did not
possess any crust or glaze on the surface, and were
05 uniform throughout the vial.
Reconstitution of Lvophilized I~M
The crimp seals were removed from t:he vials to
expose the closure, and the closure was removed from one
vial containing the product and one placebo vial. A
lO sterile pipette was filled with l.O ml of
sterile/pyrogen-free (s/pf) water (McGaw), which was
dispensed into the vial holding the lyophilized product.
Once all of the water was injected, the length of ti~e
necessary to dissolve all visually observable material
lS ~as measured:
Reconstitution Times
____________________
Placebo Product
0.5 minutes 0.3 minutes
Visual Examination After_Reconstitution
The vials were held directly in front of a black
background for ~isual examination. This was accom-
plished by placing a light source below the vial so that
the beam o f light proceded upwards through the liquid.
Changes in color, turbidity, flocculation, fine precipi-
25 tation or any other pareiculate matter were examined.
No discernable difference could be seen between the
non-lyophilized formulation and the lyophilized
` . ' - '
, . ` ~
':
. .
: .
::
. ~ ,, ~ . .
WO90/11091 PCT/US90/0l383
;J
J~ j
reconstituted formulation. The results are shown in the
following Table:
Visual Particle & Cosmetic Analvsis
Vial Appearance
05 Place.bo (non-lyophilized) clear, yellowish liquid
Product (non-lyophilized) clear, yellowish liquid
Placebo (lyophilized) clear, yellowish liquid
Product (lyophilized) clear, yellowish liquid
HPLC Gel Filtration
___________________
The lypophilized product and placebo were measured
by HPLC (Waters) gel filtration. Non-lyophilized product ~- --
and placebo were also run. A DuPont Zorbax GF-450 gel
column was equilibrated with a mixture of 0.2M sodium
phosphate buffer (pH 6.8) and 0.3 M NaCl at a flow rate
15 of 1 ml/min. Absorbance wavelength was set for 214 nm.
One ~1 of undiluted sample (lyophilized and
pre-lyophilized product and placebo) was injected onto
the column through an automatic injector and run for 15
minutes. The results, shown in Figure 1, indicated no
20 detectable difference between the pre-lyophilized
placebo and product and lyophilized/reconstitut~ed
placebo and product.
Immunoactivity Assay
_____ _______ _ ___
The immunological activity of the IgM in each
25 formulation was determined using an enzyme-linked
immunoassay to measure binding to solid-phase lipid A.
': .. . .:
,
.:
.
WO90/11091 PCT/US90/01383
r~ ` :.. .
. -12-
A vial of Salmonella minnesota R595 lipid A (List
Biological Laboratories, Inc., Campbell, CA; catalog
#401) was reconstituted to 1 mg/ml with 0.5~ TEA
~triethylamine) in s/pf water. A 10 ~g/ml solution of
05 lipid A was made in a buffer solution consisting o 10
mM HEPES and sterile/pyrogen-free 0.9~ NaCl ~s/pf
saline, McGraw), pH 7.2 (Buffer #1). This formuation
was then dispensed into a PVC microtiter plate (Dynatech
Laboratories, Inc., Chantilly, VA catalog ~011-010- ,
10 2101), 50 ~l/well, and the plate was covered and
incubated overnight at 4C.
The plates were removed from the incubator and
washed three times with s/pf saline, then blocked by
.... ...d.ispensing 200.~1/well-of a buffer consisting of: 10 mM
15 HEPES, s/pf saline and 2~ heat-inactivated FBS, pH 7.2
(Buffer ~2). The plates were covered and incubaeed for
1 hour at 37C. After incubation, the plates were
washed three times with s/pf saline.
Solutions were prepared of the test formulations,
20 an IgM standard and a negative control of human myeloma
IgM (Chrompure Human Myeloma IgM, Jackson Immuno
Research Laboratories, West Grove, PA) at a
concentration of 5.0 ~g/ml.
Buffer #2 was dispensed into the wells in rows B-H
25 (50 ~l/well). The IgM standard was dispensed into row
A, columns 1-3 (100 ~l/well). Test formulations were
dispensed in triplicate in row A, columns 4-12 (100
~l/well). Serial 50~1 dilutions were then made down the
rows of the plate to row H. 50 ~1 of ehe 100 ~1 in row
30 H was discarded, and the negative control was added.
. .
",~:, . , ' . . ' ' '
.
. . , , ~ .
:. ~ : :., :
WO90/11~91 PCT/~'S90/01383
-l3~
The plate was covered and incubated for 2 hours at 37C,
then washed three times with s/pf saline.
Substrate solution was prepared by ,adding l
phosphatase substrate table tsigma Chemicals, Inc., St.
05 Louis, MO) to 5 ml of s/pf water containing a l:500
dilution of an alkaline buffer solution (Sigma
Chemicals, catalog #014-105), and incubated for 20
minutes. The reaction was stopped by adding 50 ~l of 3M
NaOH.
The optical density of the solutions were measured
at 414 nm or using a plate reader. The data were
analyzed using a 4 parameter fit of OD versus
concentration.
The results, shown in Figure 2, indicated no
detectable difference between the activity of the
prelyophilized and lyophilized product.
Tem~erature Stress Testin~
Lyophilized product samples were stored at 4~,
22~C and 40C. The samples were evaluated periodically
for activity and appearance (i.e., particulate
formation). The samples were reconstituted prior to
evaluation. The results are shown in the following
table:
Tem~erature/Time Activity Ap~earance
25 4C/2 months no change clear
22-C/2 months no change clear
40C/2 months slight decrease some particulates4C/5 months no change clear
. ::
: ........................ : , ~ : . . . . .. .
- . : :. ~ ., :
: . . . , : :
:
. .
WO90/11091 PCT~US90/01383
~ ' ,,~
-14-
22C/5 months no change clear
4C/7 months no change clezr
E~uivalents
Those skilled in the art will recognize, or be able
05 to ascertain, using no more than routine experiment-
ation, many equivalents co the specific embodiments of
the invention described herein. These and all other
equivalents are intended to be encompassed by the
following claims.
:. ' ~: : :
,: . ~: ' ' ' ': ' : ' ' ' ' ' .-
:, ~
