Measurement of whole body interleukin-6 (IL-6) production:

From by guest on November 14, 2014. For personal use only.
1995 86: 3123-3131
Measurement of whole body interleukin-6 (IL-6) production:
prediction of the efficacy of anti-IL-6 treatments
ZY Lu, H Brailly, J Wijdenes, R Bataille, JF Rossi and B Klein
Updated information and services can be found at:
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
Information about ordering reprints may be found online at:
Information about subscriptions and ASH membership may be found online at:
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.
From by guest on November 14, 2014. For personal use only.
Measurement of Whole Body Interleukin-6 (IL-6) Production: Prediction of
the Efficacy of Anti-IL-6 Treatments
By Zhao Yang Lu, Hewe Brailly, John Wijdenes, Regis Bataille, Jean-FranGois Rossi, and Bernard Klein
A major limitation on the
therapeutic use of cytokine antagonists is that theamount of cytokine t o be neutralized in vivo
is not presently known. We previously reported that antiinterleukind (IL-6) monoclonal antibody (MoAb) administered t o a patient with multiple myeloma (MM) induced high
amounts of IL-6 t o circulate in the form of monomeric immune complexes. Based on this observation, the present
study developed anew methodology t o estimate daily IL-6
production in 13 patients with MMor renal cancer who received anti-ll-6 MoAb. Treatment was considered effective
when the production of C-reactive protein (CRP) was inhibited. The production of this acute-phase protein by hepatocytes is dependent on the activation of IL-6 gp130 transducer. Inhibition of tumor proliferation was also evaluated
in patients with MM. In 7 of 13 patients whose CRP produc-
tion was completely inhibited (>W%)
and who showed
some antitumoral effects, whole-body 11-6 production in
vivo was less than 18 p g i d (median, 5.7 pg/d; range, 0.5 t o
17.5 pg/d). In the other 6 patients, subtotal inhibition of
CRP production and a lack of antitumoral response were
associated with high IL-6 production (median,180 pg/d;
range, 18 t o 358 pg/d). These in vivo observations were
consistent with mathematical modeling that predicted that
anti-IL-6 MoAb treatment would be efficient only in low IL6 producers.These data indicate the difficulty of neutralizing
IL-6 with a single anti-ll-6 MoAb in vivo and call for new
strategies to avoid accumulation of IL-6in the form
of stable
immune complexes.
0 7995 by The American Societyof Hematology.
(CRP) production or tumor proliferation in patients with
MM. The production of this acute-phase protein by human
hepatocytes is induced by the different cytokines activating
the gp130 IL-6 transducer." Its in vivo production was also
strongly inhibited in various patients treated with anti-IL-6
Our methodology is of general relevance to the evaluation
of CBP treatments involving MoAbs and soluble receptors.
Within 2 days after the beginning of treatment, it is possible
to determine whether cytokine overproduction is a limiting
factor and thus avoid inefficient therapy.
HE INVOLVEMENT OF cytokines in various tumoral,
autoimmune, and inflammatory diseases has led to the
development of cytokine-binding proteins (CBP), eg, monoclonal antibodies (MoAbs) to cytokines or soluble cytokine
receptors, to neutralize cytokine activity in vivo. Clinical
trials with anti-interleukin-6 (IL-6) MoAb have been initiated in patients with multiple myeloma (MM),',' rheumatoid
Castleman's disease: human immunodeficiency virus lymphoma: or renal carcinoma (RC): Anti-tumor necrosis factor (TNF) MoAbs have been used in the treatment of
patients suffering from hairy cell leukemia' or gram-negative
septicemic shock.' Other CBP, such as soluble IL-1 receptor,
soluble TNF receptor, or immunoadhesins derived therefrom, have also been developed for clinical use in humans.'
A major difficulty in the evaluation of the efficacy of cytokine antagonists in vivo is a lack of data about whole-body
production of a cytokine under normal and pathologic conditions. Such information is required to predict the amount of
CBP needed to neutralize the target cytokine in vivo. Indeed,
several groups have reported that CBP can increase cytokine
half-life and induce the circulation of large amounts of cytokine in the form of complexes. In a patient with MM treated
with anti-E-6 MoAb, the half-life of IL-6 in the form of
IL-6/anti-IL-6 complexes was increased 200-f0ld.'~ This
finding was confirmed in animals treated withanti-IL-6
MoAb,".'2 anti-E-3 MoAb," anti-E-4 MoAb,14 anti-IL-7
MoAb,14 or soluble IL-4 receptor."
In the case of high cytokine production in vivo, the concentration of circulating CBP may notbe sufficient to prevent
the disruption of cytokine/CBP complexes by the cell membrane high-affinity cytokine receptor. Several
have suggested that CBP in such cases behaves as a cytokine
agonist in vivo. Once again, knowledge of whole-body cytokine production in a patient to be treated with CBP might
help prevent such adverse effects.
In the present study, this CBP cytokine carrier effect was
used to estimate daily whole-body production of IL-6 in 13
patients with MM or RC treated with murine anti-IL-6
MoAb. We found that low or high IL-6 production in vivo
was strictly correlated with the clinical efficacy of anti-IL6 treatment as determined by inhibition of C-reactive protein
Blood, Vol 86, No 8 (October 15). 1995: pp 3123-3131
Patients. After informed consent was obtained, anti-IL-6 therapy was administered to 6 patients with MM, extramedullary proliferation, and resistance to chemotherapy and to 7 patients with metastatic RC resistant to IL-2 immunotherapy. Patients were treated with
B-E8 anti-IL-6 MoAb (IgG,) prepared by J. Wijdenes (Diachone,
BesanGon, France)," which was administered in a l-hour infusion
every day at 8 AM for periods of 10 to 60 days. Patients with RC
received a daily dose of 20 mgof B-E8 MoAb throughout the
treatment. These doses were doubled (40 rng/day) on day 13 for
patient DE and on day 15 for patient AT because of incomplete
CRP inhibition. Five patients with MM received a daily dose of 20
mg and 1 patient (MU) received a daily dose of 8 rng. For patient
From the Institute for Molecular Genetics, CNRS BP5051, Montpellier, France: the Department of Pathophysiology, Suzhou Medical
College, Suzhou, Jiangsu, Peoples' Republic of China; Immunotech,
Marseilles, France; the Centre Rdgional de Transfusion Sanguine,
BesanGon, France; the Laboratoire d'oncogdnise Immunohdmatologique, Nantes, France: andthe Service des Maladies duSang,
Hdpital Lapeyronie, Montpellier, France.
Submitted December 15, 1994; accepted June 9, 1995.
Address reprint requests to Bernard Klein, PhD, Institute for Molecular Genetics, CNRS, BP 5051, 1919 Route de Mende, 34033
Montpellier Cedex I , France.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
"advertisement" in accordance with l 8 U.S.C. section 1734 solely to
indicate this fact.
0 1995 by The American Society of Hematology.
From by guest on November 14, 2014. For personal use only.
CH, the MoAb dose was reduced to 10 mgld on day 3 of treatment
and was increased to 20 mg on day 11. Neutrophil counts for this
patient decreased on day 3 due to toxic antibiotic therapy and then
increased again once this therapy was discontinued on day 8. The
clinical results for these anti-IL-6 MoAb treatments have been reported
Briefly, complete blockage of tumor proliferation occurred in 2 patients (TEand MU) with MM. In these patients,
the serum concentration of monoclonal Ig was decreased by 50%
and 30% respectively. No antitumoral response was noted in patients
with RC. Serum CRP concentrations (monitored daily) became undetectable in 5 patients with RC (FA, AT, GO, BO, and CA) and in
2 with MM (TEand MU). Immunization to anti-IL-6 murine MoAb
was detected between days 7 and 15 of treatment in 9 of 13 patients.’”
All immunized patients produced anti-idiotypic antibodies specific
for B-E8 MoAb. Two patients (SO and BO) also developed antibodies directed at the murine IgGl isotype. In these 2 patients, B-E8
MoAb cleared rapidly from the circulation and treatment efficacy
was lost. In patients who developed only anti-idiotypic antibodies,
serum levels of B-E8 MoAb were unchanged and CRP production
remained inhibited, indicating that treatment efficacy wasnot affected by the presence of human antibodies to idiotypic determinants
of B-E8 MoAb.
Reagents. Recombinant IL-6 was provided byD. Stinchcomb
(Synergen, Boulder, CO); anti-IL-6 B-E8 and B-E4 MoAb were
provided by J. Wijdenesl’; and anti-IL-6 AH-65MoAb was provided
by H. Brailly (Immunotech, Marseille, France). These three antibodies bindto distinct epitope~.’~~’’
Natural IL-6 was obtained from
supernatants of human peripheral blood monocytes cultured for 24
hours at a concentration of 10’ cells/mL inRPM11640medium
supplemented with 5% fetal calf serum (FCS).
B9 hybridoma assay. Serum IL-Mike bioactivity was evaluated
usingtheB9 hybridoma bioassay.” Five thousand B9 cells were
cultured with serial dilutions of the sample in 96-wel1, flat-bottomed
culture dishes in 100 pL ofRPM1 1640 medium containing 5%
FCS. After an incubation period of 4 days at 37°C in 5% CO,, cells
were exposed to M’M (Sigma, St Louis, MO) for 4 hours at 37°C
as previously described.*’ One unit of B9-stimulating activity (B9SA), defined as the amount inducing half-maximal optical densities,
corresponded to approximately 5 pg of recombinant [L-6.
Affinity columns (protein A-Sepharose, B-E4-Sepharose, B-E8-Sepharose, and anti-IgG[mouse]-Sepharose) were prepared by coupling protein A (Sigma), B-E4 MoAb,
B-E8 MoAb, or rat antimouse IgG antibody (Nordic Immunology,
Amsterdam, The Netherlands) to cyanogen bromide-activated Sepharose 4B (Pharmacia, Uppsala, Sweden). The columns were used as
previously described.“’ The bound material was eluted at acidic pH
(0.1 mol/L glycine-HCL, pH 2.4) according to the manufacturer’s
recommendations, immediately neutralized with I molL NaHCO?,
and then assayed for B9-SA at different dilutions.
Gel jiltration. A 4-mL sample was loaded onto a Sephadex G 200 column (100 cm X 2.6 cm) and chromatographed at 4°C with
50 mmoVL phosphate-buffered saline (PBS; pH 7.2) as eluting
buffer; the flow rate was 12 mL/h and the fraction volume 4 mL.“’
All fractions were assayed for B9-SA at different concentrations.
Purified human IgG, bovine serum albumin (Sigma), ovalbumin, and
equine myoglobulin (Sigma) were used as standard molecules for
estimating the molecular weight of B9-SA. To dissociate expected
immune complexes, the samples were dialyzed for 12 hours with a
0.1 mom glycine-HCI buffer (pH 2.3) and then gel-filtrated with
the same acid buffer. Each fraction was neutralized with 1 molL
NaHC03 and assayed for B9-SA at different concentrations.
Determination ofplasmaIL-6 and plasma B-E8 MoAb concentrations. To determine the plasma IL-6 concentration bound to B-E8
anti-IL-6 MoAb, we used the two anti-IL-6 MoAbs (AH-65 and BE4) that recognize different epitopes on the IL-6 molecule than those
recognized by B-E8 MoAb.”.” AH-65 was used as a catcher and
B-E4 as a tracer. This methodology was previously described in
detail.” Detection of circulating free IL-6inthe
patient (ie, not
boundto B-E8 MoAb) was performed using an AH-65/B-E8 enzyme-linked immunosorbent assay (ELISA). B-E8 MoAb concentration was assayed using an ELISA, as previously described.’
Calculation of daily IL-6 production. On the basis of the mathematical procedure developed below (equation 5), daily determinations of circulating IL-6 (in the form of monomeric immune complexes) and anti-IL-6 MoAb were performed to estimate the overall
daily production of IL-6 induced to circulate by anti-IL-6 MoAb.
This mathematical procedure required the verification of two assumptions, ie, that the half-life of the immune complexes was similar
to that of free MoAb and that no elimination route existed for IL-6
other than immune complexes. The first assumption was likely for
monomeric immune complexes, as previously shown in animal modand was confirmed by monitoring the accumulation kinetics
of anti-IL-6 MoAb and IL-6/anti-IL-6 MoAb monomeric complexes
at the beginning of treatment. As indicated below (see Fig 3), these
accumulation curves were parallel, yielding a common half-life of
3 to 4 days. With respect to the second assumption, the two major
elimination routes for cytokines in vivo are consumption by cells
and renal eliminati~n.’~
When complete blockage of CRP production
in vivo by anti-IL-6 MoAb occurred, it may reasonably be assumed
that anti-IL-6 MoAb prevented IL-6 binding to cell surface receptors
and elimination by renal filtration. Thus, all IL-6 produced in vivo
was induced to accumulate in the form of monomeric immune complexes, so that daily overall IL-6 production should have been close
to the value estimated by the mathematical procedure below. When
CRP inhibition was partial, itmay be concluded that overall IL-6
production in vivo was higher than the estimated value.
Mathematical procedure for calculating overall daily IL-6 production. Let[MoAb,]equalthe
concentration of circulatingMoAb
at the beginning of day,, just before the injection of anti-IL-6 MoAb
on day,; TI,’ = the half-life of circulating MoAb and b = e
MoAb, = the dose of injected MoAb at the beginning of day,; and V
= the volume of injected MoAb diffusion in the body.
Hence, [MoAb,] = MoAbo bN; [MoAb,] = b[MoAb,] + MoAb,
bN; [MoAb,] = b[MoAb.-,] + MoAb,-, b/V (equation I ) . Thus,
[MoAb,] = (bN)(MoAb,-, + bMoAb,-2 + ... + b”’MoAb,J.
When the dose of injected MoAb is constant, MoAb, = MoAb,-,
= MoAbo = MoAb. Thus, [MoAb,] = (MoAbN) b (I + b + b?
+ ... + b”’)(equation 2).
The concentration of circulating MoAb converges to [MoAb,] =
[MoAbN)(b/( 1 -b)] (equation 3).
Equations 2 and 3 canbeusedto
fit theexperimentalcurve of
circulating MoAb and to provide experimental data for b. Hence IL6, equals the overall production of IL-6 trapped by anti-L-6 MoAb
during day,and induced to circulate in the form
of immune complexes;
V = thevolumeof
= theconcentration of circulating L - 6 intheform
of monomeric
immune complexes at the beginning of day,, just before injection of
anti-L-6 MoAb on day,; TiR = the half-life of circulating L-6 in the
= e-(LnZyT’ln
. Thus.
form of monomeric immune complexes; and b’
[IL-6.] = b’[L-6.-,] + IL-6,_,(1+b‘)/2V (equation 4).
For monomeric immune complexes, the half-life is similar to that
of free MoAb, so that b = b’, which is the case for IL-6/B-E8 MoAb
monomeric immune complexes (see above and Fig 1).
By combining equations 1 and 4, IL-6, can be determined by the
equation: IL-6, = 2 MoAb.b([IL-6,] - b’[IL-6.. ,])/([MoAb,l b[MoAb,-,l)(l +b’) (equation 5).
Numerical modeling of B-E8 MoAb ability toneutralize IL-6 binding to its high-afinity receptor. Our numerical modeling procedure
is based on data in the literature describing the interactions of IL-6
with the two chains of its membrane receptor and the soluble forms
From by guest on November 14, 2014. For personal use only.
1 2
Fig 1. Gel filtration of BS-stimulatingactivity
present in plasma of patients treated with anti-lL-6
Mom. Patient BUS plasma was harvested on day
10 of treatment, and natural 11-6 was obtained from
The plasma
the culture
were filtered througha Sephadex G-200column l1 00
x 2.6cm).Elutingbufferwas
50 mmollL PBS (pH
7.2). In someexperiments, the plasmasampleand
the culturesupernatantweredialyzedfor
12 hours m
beforebeingapplied to the Sephadex 200 column
with an 0.1 mol/L glycine-HCL eluting
buffer (pH2.3).
The flow rate was 12 mLlh, and 4 mL fractions were
collected. Each fraction was neutralized
with 1 moll
L NaHC03and assayed for B9 cell stimulating activity
-” Plasma
Activity (pH 7.2)
Activity (pH 2.3)
11-6 (pH 2.3)
11-6 (pH 7.2)
Fraction number
of these two chains. This procedure depends on a novel approach
for the numerical resolution of complex multimolecular equilibria.26
In this model, IL-6 first binds the membrane IL-6 receptor (IL-6R)
or its soluble form (sIL-6R). IL-6AL-6R or IL-6/sIL-6R complexes
then bind the gp130 transducer chain.” These complexes may also
bind the soluble form of gp130 (sgpl30), which behaves as an
antagonist.” B-E8 anti-IL-6 MoAb blocks the binding of IL-6 to IL6R or sIL-6R.I9 Numerical modeling predicts the fraction of gp130
transducer activated by the IL-6AL-6R or IL-6/sIL-6R complexes
as a function of IL-6 and anti-IL-6 MoAb concentrations. The affinities of the interactions of the different components were estimated
from data in the literature or determined by us: IL-6.B-E8 = IL-6
+ B-E8 (kd = 22 pmolk; H. Brailly, unpublished observations);
IL-6.IL-6R = IL-6 + IL-6R (kd = 500 pmol/L) andIL-6AL-6W
gp130 = IL-6AL-6R + gp130 (kd = 5 p r n ~ l / L ) ~IL-6kIL-6R
IL-6 + sIL-6R (kd = 1 nmoVL)and IL-6/sIL-6R/sgp130
sIL-6R + sgpl30 (kd = 10 nmoVL)”; IL-6/sIL-6R/gp130 = IL-6/
Because numerical modeling
sIL-6R + gp130 (kd = 10 pm~l/L).’~
was used to predict the inhibition of CRP production by hepatocytes,
the concentrations of the different components were chosen to fit
those reported for hepatoma cell lines: 5,000 IL-6R and 500 gp130
per hepatocyte were used according to data published by Baumann
et al” and Rose-John et al.” The concentration of hepatocytes in
the liver was assumed to be 5 X 10” cells/mL. The concentrations
of sIL-6R and sgpl30 were the mean values reported in the plasma
and 800 ng/mL (B. Klein,
of patients with MM, ie, 200 r~g/mL’~
unpublished observations).
IL-6 accumulation in the form of monomeric immune complexes during anti-IL-6 MoAbtreatment.
In agreement
with previous reports,2.’oa high activity stimulating B9 hybridoma growth (B9-SA) was found in the plasma of all
patients at the beginning and after discontinuance of antiIL-6 treatment and in some cases throughout the treatment
period (results not shown). Using a methodology previously
reported by us,” we found that this activity consisted of IL6 bound to the murine anti-IL-6 MoAb. In fact, IL-6 was
retained on protein A or antimurine Ig Sepharose columns,
indicating that it was linked to the murine MoAb (results
not shown). It was eluted at peaks of 25 and 180 k D , respectively, after gel filtration with or without acidic conditions
(Fig 1). Finally, 25 kD activity was inhibited by an anti-IL6 MoAb (results not shown).
ELISA for assaying IL-6 complexed to B-E8 MoAbor free
IL-6. Figure 2A shows that an ELISA usingAH-65 and
B-E4 anti-IL-6 MoAb recognized different epitopes on the
IL-6 molecule than those recognized by B-E8 MoAb. Free
IL-6 as well as preformed complexes of IL-6 and B-E8
MoAb diluted in human plasma were detected (sensitivity,
approximately 30 pdmL of IL-6). No activity was detected
when IL-6 was complexed with B-E4 MoAb. Conversely,
AH-65/B-E8 ELISA detected free IL-6 as well as IL-6/BE4 complexes, but not IL-6/B-E8 complexes (sensitivity,
approximately 30 pgImL of IL-6; Fig 2B).
IL-6/B-E8 monomeric immune complexes have the same
half-life as free B-E8 MoAb in vivo. Figure 3 displays the
accumulation kinetics of B-E8 MoAb in patients treated with
anti-IL-6 MoAb (Fig 3A, patients with MM; Fig 3C, patients
with RC). A plateau generally obtained after 6 to 8 days of
treatment indicated a 3- to 4-day half-life for the MoAb. In
2 cases (patient SO with MM and patient BO with RC), a
decrease in plasma MoAb concentrations was observed after,
respectively, 7 and 13 days of treatment due to immunization
against B-E8 MoAb and subsequent clearance of this
MoAb.”In patient DE (RC), an increase inB-E8 MoAb
concentrations occurred on day 13 when injected doses of
B-E8 were doubled (Fig 3C). In patient CH (MM), the variations in B-E8 MoAb concentrations reflected the injection
course. When an AH-65/B-E4 ELISA was used, a high IL6 concentration was detected in the plasma of every patient
treated with B-E8 MoAb. This activity was entirely bound
to B-E8 MoAb, as shown by the gel filtration experiments
described above (Fig 1) and the absence of a signal when
AH-65/B-E8 ELISA was used (results not shown). The accumulation kinetics of IL-6/B-E8 MoAb complexes are indicated in Fig 3B and D. For 10 patients, these curves paralelled those obtained for B-E8 MoAb, showing that IL,-6/B-E8
MoAb half-life in vivo was similar to that of B-E8 MoAb
(ie, 3 to 4 days). In patient CH (MM), there was a threefold
decrease in the concentration of circulating IL-6, in association with severe neutropenia due to toxic antibiotic therapy
From by guest on November 14, 2014. For personal use only.
Concentdon of 11-6 (-1)
IL-6 alone
-8- IL-6 with BE8 (20pghl)
IL-6 with BE4 (20p g h l )
Concentration of 11-6 (pghnl)
Fig 2. Measurement of free 11-6 or IL-6 complexed with anti-ll-6
MoAb by ELISA. A total of 20 pg/mL of &E8 anti-lL-6 MoAb and 10
nglrnL of IL-6 ware added to a pool of plasma from 5 healthy individuals. The IL-6 activity of this preparation was assayed using two ELISA.
In (A), AH65 anti-lL-6 MoAb wasused as a catcher and in B-E4 antiIL-6 MoAb as a tracer; in (B), AH65 anti-IL-6 MoAb was used as a
catcher and B-E8 anti-lL-6 MoAb as a tracer.
(Fig 3B). In patient FA (RC), plasma IL-6 levels decreased
on and after day 4 of treatment, but withno noteworthy
clinical effects. Finally, in patient AT (RC), IL-6 concentrations plateaued only on and after day 12 (Fig 3D).
Overall IL-6 production in patients with M M or RC. B
ecause plasma concentrations of B-E8 MoAb and IL-6IBE8 complexes were determined daily and the in vivo halflives of these MoAb were similar, it was possible to estimate
daily total IL-6 production in patients treated with anti-IL6 MoAb according to the mathematical methodology presented in the Materials and Methods (Fig 4A andB). Interestingly, total IL-6 production was stable for l l of 13 patients
throughout the treatment period. For patient CH (MM), total
IL-6 production was reduced 23-fold (from 46 to 2 pgld)
in association with severe neutropenia related to antibiotic
therapy. Production returned to initial values when neutrophil counts became normal. For patient FA (RC), IL-6 production became progressively undetectable. Although IL6 production was very stable for most patients throughout
treatment, the level was highly variable among patients,
ranging from 2 to 350 pg/d.
Prediction of theefJicacy of anti-IL-6 treatment. We
investigated whether the efficacy of anti-IL-6 treatment was
limited by IL-6 production. Table 1 shows the mean concentrations of circulating B-E8 anti-IL-6 MoAb and IL-6 from
day 7 to 12 of treatment, at which time a steady state was
achieved (Fig 3), as well as the ratio of these concentrations.
Patients were classified according to the latter parameter.
The mean inhibition of CRP production from days 7 to 12
and the mean IL-6 production values from day 2 to 12 of
treatment are also given (Table 1). As indicated in Table l,
CFW production was completely inhibited (>96%) only in
patients producing less than 18 pg of IL-6 per day for whom
the ratio of IL-6 to anti-IL-6 MoAb molar concentrations
was less than 4 X lo-’. The mathematical procedure described in the Materials and Methods enabled usto plot
theroretical curves for the fraction of gp130 transducer
bound toIL-64L-6R or IL-6/sIL-6R complexes for IL-6 concentrations of 0.1 to 1 pmoVL (ie, 2.5 to 25,000 ng/mL) and
MoAb concentrations of 20 to 200 nmoVL (ie, 3 to 30 pgl
mL). Interestingly, the fraction of bound gp130 in the presence of anti-IL-6 MoAb depended only on the IL-6lantiIL-6 MoAb molar ratio, at least within the selected concentration ranges (Fig 5). Moreover, 10-fold variations inthe
concentrations of the different components used in numerical
modeling did not affect the theoretical curve when an sIL6R concentration of 2 nmol/L was used. This sIL-6R value
is the physiologic concentration found in plasma of patients
with MM or RC.34 Numerical analysis wasinverygood
agreement with the observed inhibitions of CRP production
and IL-6/anti-IL-6 MoAb molar ratios in patients, as shown
in Fig 5 , which plots the observed inhibitions of CRP production over the theoretical curve. This analysis confirmed that
there was a critical level in IL-6 production (18 pgld) or the
IL-6/anti-IL-6 MoAb molar ratio (4 X lo-’)). Above this
critical level, it was predictable that anti-IL-6 MoAb would
no longer beefficient in blocking IL-6 activity, whichis
consistent with the lack of complete inhibition of CRP production in 8 patients (Fig 5 and Table 1).
It is noteworthy that the ratios of IL-6 to anti-IL-6 MoAb
molar concentrations on day 2 of treatment were close to
the mean ratios calculated when a steady state was achieved
(Table 1). This finding is not surprising because daily IL-6
production was stable throughout treatment and the kinetics
of IL-6 andanti-IL-6 MoAb accumulation were parallel
(Figs 3 and 4). Thus, for each patient, the doses of anti-IL6 MoAb required to neutralize IL-6 could be predicted as
early as day 2 of treatment.
This report considers the ability of an MoAb to IL-6 to
neutralize this cytokine in patients with MM or RC in vivo.
IL-6 produced in vivo may exist in free form or be bound
to the extracellular matrix35or to plasma proteins.36It can
be eliminated by binding to cell surface receptors and internali~ation,’~
renal filtration, or degradation by proteases (this
last route has not been shown in vivo). Injection of an antiIL-6 MoAb disturbs this distribution pattern. If the antibody
is in sufficient concentration and has a high affinity with IL6, it will bind free IL-6 and disrupt IL-6 binding to plasma
From by guest on November 14, 2014. For personal use only.
Days after treatment
5 201510
Days after treatment
c 100
t 1
CH -x- MU -c- TE
proteins" and probably to the extracellular matrix. In addition, anti-ILd MoAb will block IL-6 binding to cell surface
receptors, and the high-molecular-weight IL-6/anti-IL-6
complex will no longer be eliminated through the kidneys.
By blocking the different IL-6 elimination routes, antiIL-6 MoAb can thus accumulate large amounts of IL-6 circulating in the form of monomeric complexes. We initially
described this phenomenon in humans 2 years ago," and it
was recently confirmed by several groups in animal models
using anti-IL-6 MoAb"." or other CBP.'33'4In this study,
the concentration of circulating IL-6 ranged from 630 to
77,880 pg/mL in the 13 patients during treatment (range
before treatment, 0 to 56 pg/mL). This accumulation of circulating IL-6 in the form of immune complexes was due to
their long half-life (ie, 3 to 4 days), which is at least 200fold higher than that of free IL-6.37 This finding is in
agreement with those of previous reports concerning the
half-life of monomeric immune complexes of hapten-antihapten antibodies in animal models.=
Days after treatmemt
Days after treatment
Fig 3. Plasmaconcentrations
of anti-IL-6 MoAb (A and C) and
11-6 complexed with anti-IL-6
MoAb (Band D). (A) and (B)represent data from patients with
MM and IC) and (D) data from
patients with RC.
+ BO + FA
* GO
+ CA
This accumulation of IL-6 in the form of stable immune
complexes is a serious limitation on treatments with antiIL-6 MoAb. If IL-6 production is too high, the ratio of IL6 to free MoAb molar concentrations will also be high and
the high-affinity IL-6 receptor will disrupt IL-6/anti-IL-6
immune complexes, as recently disc~ssed."j*'~ deal
To with
this complex situation, we developed a numerical model
taking into account the various chains of membrane IL-6
receptors and their soluble forms (IL-6R, gp130, sIL-~R,
and sgpl30). This model shows that the ILd/anti-IL-6 ratio
should be at least lower than 4 molecules of IL-6 to 1,OOO
molecules ofanti-IL-6MoAb,which
was achieved in 7
patients in whom complete inhibition of CRP production
was observed in vivo. In vitro, the production of this acutephase protein by hepatocytes is induced by activation of
gp130 transducer, which is achieved mainly by IL-638 but
also by other gpl30-activating cytokines, ie, ciliary neurotrophic factor, IL-11, leukemia inhibitory factor, or oncostatin M."Our results here show that there was no evidence
From by guest on November 14, 2014. For personal use only.
* BU
-c- LE
8 16
Days after treatment
-e BO
Days after tlwtment
that these other gpl30-activating cytokines are involved in
CRP production in the patients studied. In fact, CRP production was completely inhibited in the 7 patients whose IL-6
activity was predicted to be fully inhibited byanti-IL-6
MoAb. These data confirm the relevance of using CRP production to monitor the efficacy of anti-IL-6 treatment. Moreover, antitumoral effects were observed only in the patients
with MM in this group. For patient MU, a complete inhibition of tumor cell proliferation was found throughout the
2-monthtreatment.' For patient TE, a 50% reduction in
monoclonal Ig concentration was found within 1 month of
treatment? In the other 6 patients with MM, this ratio was
too high (up to 25-fold), allowing only partial inhibition of
CRP production in vivo and providing no antitumoral effects
(Table 1). To model the neutralizing capacity of anti-IL-6
MoAb, we assumed that local concentrations of IL-6, antiIL-6 MoAb, sIL-~R,and sgpl30 were close to their plasma
Fig 4. Daily prodwUon of
IL-6 in patients Mwith
enti-IL-6 MO&. (AI repm
aents data from patients
with MM and (B1 data from
patient. with RC.
concentrations. This was true for hepatic tissue because IL6 was produced in distant tissues (mainly at tumor sites)
in these cancer patients. IL-6 circulated through the blood,
arriving in the liver in the form of immune complexes that
had the same molecular weight and probably the same diffusion ability as free antibody. This explains why the theoretical predictions for inhibition of gp130 transducer activation
by anti-IL-6 MoAb fitted well with observed inhibitions of
CRP in vivo. In tumor sites in which IL-6 is produced and
thus found at higher concentrations than in plasma, the
MoAb concentrations required to neutralize IL-6 should be
higher than those in plasma. In these circumstances, it is
easy to understand why a patient whose CRP production
was only partially inhibited had no chance of responding to
T h i s study emphasizes the difficulty of neutralizing DL-6
by means of a single anti-L-6 MoAb in vivo, even though
From by guest on November 14, 2014. For personal use only.
Table 1. Biologic Parameters of PatientsTreated With Anti-11-6 MOAbS
Anti-IL-6 MoAb
Ratio of IL-6 to AntiIL-6 MoAb Molar
Concentration ( ~ 1 0 " )
Complete inhibition of
CRP (>96%)
15.31 0.72
2.99 2.00
13.00 2.24
20.47 1.30
Partial inhibition of
Ratio of IL-6 to Anti11-6 MoAb Molar
Concentration on
Day 2 (xlO")
5.86 >18.1
of CRP
Daily 11-6
> 180.3
Concentrations of anti-IL-6 MoAb and IL-6 and observed inhibition of CRP are the mean values from days 7 to 12 of treatment when a steady
state was achieved, except for 2 patients (see notes). Daily IL-6 productions are the mean values for the first 12 days of treatment.
* Mean of values of days 4 to 8 because of subsequent immunization.
t Value on day 5 due to subsequent severe neutropenia.
Mean of values of days 2 to 4.
the B-E8 anti-IL-6 MoAb used had a strong affinity with
IL-6 (kd, lo-" m o m ) and was one of the most potent neutralizing anti-IL-6 MoAbs in our hands. Our injected doses
of MoAb (20 mg/d) were already high but should have been
jii :;
obarved Inhibilbn of CRP Serum L&
- Numerical Modding
i 1 1;:
20 i
Fig 5. Mathematical modeling of the inhibition of gp130 transducer activation by antHL-6 MoAb. The ordinate represents the fraction of the gp130 transducer occupied by l"61L-6R or lL-6/slL-6R
complexes as a function of the ratio of IL-6 t o antHL-6 MoAb molar
mathematconcentrations. The curve isthe numerical f~given by the
ical model. The points represent the observed inhibitions of the
plasma CRP levels in patientstreated with anWL-6 MoAbas a function of the o h r v e d ratios ofIL-6 t o anti-IL-6 MoAb molarconcentrations (Table 1).
at least 25-fold greater to neutralize L - 6 in high L-&producing patients. Such predicted doses are far too elevated
for therapeutic use. A newstrategy could consist in injecting
several anti-L-6 MoAbs that recognize different epitopes
on the IL-6 molecule. We have shown in an animal model
that the resulting polymeric complexes in this case are
cleared within 1 hour.39We are currently investigating this
new therapeutic approach in the hope of considerably
reducing the half-life of these immune complexes in
Our findings in these patients treated with anti-IL-6 MoAb
might be of general use for antagonizing cytokines in vivo.
Several antagonists are now being developed using either
MoAb or soluble cytokine receptors linked or not to Fc
fragments of Ig. Recent discussion^'^^^^ and demonstrations
in animal models relative to soluble L - 4 receptor" or various MoAbs to IL-3, IL-4, IL-6, or IL-7''-I4 suggest that these
CBP will be able to protect the cytokine from its classical
elimination routes, allowing it to accumulate in the form of
complexes. The efficacy of treatment with these CBP will
depend on the ratio of cytokine to free CBP and on total
cytokine production in vivo.I7Thus, it is important to have an
estimation of whole-body cytokine production before CBP
injection, which is precisely the objective of the original
methodology described here. For a monomeric cytokine, it
is conceivable to inject a humanized anticytokine antibody
for 2 days. Our results here indicate that the measurements
of circulating anticytokine antibody and cytokine on day 2
will allow an estimation of whole-body cytokine production
predictive of the minimal CBP doses required to neutralize
the cytokine in vivo.
From by guest on November 14, 2014. For personal use only.
1. Klein B, Wijdenes J, Zhang XG, Jourdan M, Boiron JM, Liau-
tard J, Merlin M, Clement C, Morel-Fournier B, Lu ZY, Mannoni
P, Sany J, Bataille R: Murine anti-interleukin-6 monoclonal antibody
therapy for a patient with plasma cell leukemia. Blood 78: 1 198,
2. Bataille R, Barlogie B, Lu ZY, Rossi JF, Lavabre-Bertrand T,
Beck T, Wijdenes J, Brochier J, Klein B: Effects of anti-interleukin-6
(IL-6) murine monoclonal antibody in advanced multiple myeloma.
Blood (in press)
3. Wendling D, Racadot E, Wijdenes J: Treatment of severe rheumatoid arthritis by anti-interleukin 6 monoclonal antibody. J Rheumatol 20:259, 1993
4. Beck JT, Hsu SM, Wijdenes J, Bataille R, Klein B, Vesole
D, Hayden K, Jagannath S, Barlogie B: Alleviation of systemic
manifestations of Castlemans disease by monoclonal anti-interleukin-6 antibody. N Engl J Med 330:602, 1994
S. Emilie D, Wijdenes J, Gisselbrecht C, Jarrousse B, Billaud E,
Blay JY, Gabarre J, Gaillard JP, Brochier J, Raphael M, Boue F,
Galanaud P: Administration of an anti-interleukin-6 monoclonal antibody to patients with acquired immunodeficiency syndrome and
lymphoma: Effect on lymphoma growth and on B clinical symptoms.
Blood 84:2472, 1994
6. Rossi JF, Legouffe E, Blay JY, Fabbro M, Mousseau M, Khenifar B, Negrier S, Liautard J, Mao LQ, Racadot E, Wijdenes J, Philip
T, Brochier 3, Klein B: Hematological effects of in vivo blocking
interleukin-6 (IL-6) with BE-8, a monoclonal antibody (MoAb) antiIL-6. Blood 80:432a, 1992 (abstr, suppl 1)
7. Huang D, Reittie JE, Stephens S, Hoffbrand AV, Brenner MK:
Effects of anti-TNF monoclonal antibody infusion in patients with
hairy cell leukemia. Br J Haematol 81:231, 1992
8. Fischer CJ, Opal SM, Ohalnaut JF, Zimmerman J, Nightingale
P, Harris SJ, Schein RL, Panacek EA, Vincent JL, Foulke GE,
Warren EL, Garrard C, Park G, Bodmer MW, Stephens S, Cohen
J, Van der Linden G, Sadoff JC: Anti-TNF IgG monoclonal antibody
(CB006) in sepsis syndrome: Pharmacokinetics, safety and role of
pretreatment IL-6 as apredictor of time survival. Circ Shock 34: 167,
9. Fernandez-Botran R: Soluble cytokine receptors: Their role in
immunoregulation. FASEB J S:2567, 1991
10. Lu ZY, Brochier J, Wijdenes J, Brailly H, Bataille R, Klein
B: High amounts of circulating interleukin-6 (IL-6) in the form of
monomeric immune complexes during anti-IL-6 therapy. Toward a
new methodology for measuring overall cytokine production in human in vivo. Eur J Immunol 22:2819, 1992
11. May LT, Neta R, Moldawer LL, Kenney JS, Patel K, Sehgal
PB: Antibodies chaperone circulating IL-6-Paradoxical effects of
anti-IL-6 neutralizing antibodies in vivo. J Immunol 151:3225, 1993
12. Heremans H, Dillen C, Put W, Van Damme J, Billiau A:
Protective effect of anti-IL-6 antibody against endotoxin, associated
with paradoxically increased IL-6 levels. Eur J Immunol 22:2395,
13. Tomlinson A, Zitener HJ: Enhancement of the biologic effects
of interleukin-3 in vivo by anti-interleukin-3 antibodies. Blood
82: 1 133, 1993
14. Finkelman FD, MAdden KB, Moms SC, Holmes JM, Boiani
N, Katona IM, Maliszewski CR: Anti-cytokine antibodies as carrier
proteins-Prolongation of in vivo effects of exogenous cytokines by
injection of cytokine-anti-cytokine antibody complexes. J Immunol
151:123S, 1993
IS. Sat0 TA, Widmer MB, Finkelman FD, Madani H, Jacobs CA,
Grabstein KH, Maliszewski CR: Recombinant soluble murine IL-4
receptor can inhibit or enhance IgE responses in vivo. J Immunol
150:2717, 1993
16. Debets R, Savelkoul HF: Cytokine antagonists and their potential therapeutic use. Immunol Today 15:4SS, 1994
17.Klein B, Brailly H: Cytokine-binding proteins: Stimulating
antagonists. Immunol Today 16:216, 1995
18. Baumann H, Gauldie J: The acute phase response. Immunol
Today 15:74, 1994
19. Wijdenes J, Clement C, Klein B, Morel-Fourrier B, VIta N,
Ferrara P, Peters A: Human recombinant dimeric IL-6 binds to its
receptor as detected by anti-IL-6 monoclonal antibodies. Mol Immuno1 28:1183, 1991
20. Legouffe E, Liautard J, Gaillard JP, Rossi JF, Wijdenes J,
Bataille R, KleinB, Brochier J: Human anti-mouse antibody response to the injection of murine monoclonal antibodies against IL6. Clin E X Immunol
98:323, 1994
21. Montero-Julian FA, Liautard J, Flavetta S, Romagnt F, Gaillard JP, Brochier J, Klein B, Brailly H: Immunoassay for human
soluble interleukin-6 receptor in plasma basedon ligandreceptor
interactions. J Immunol Methods 169:1 1 1, 1994
22. Helle M, Boeise L, Aarden LA: Functional discrimination
between interleukin-6 and interleukin-l. Eur J Immunol 18:153S,
23. Bataille R, Jourdan M, Zhang XG, Klein B: Serum levels of
interleukin-6, a potent myeloma cell growth factor, as a reflect of
disease severity in plasma cell dyscrasias. J Clin Invest 84:2008,
24. Plotz PH, Kimberly RP, Guyer RL, Segal DM: Stable model
immune complexes produced by bivalent affinity labeling haptens:
In vivo survival. Mol Immunol 16:721, 1979
25. Konard MW, Hemstreet G, Hersh EM, Mansell PWA, Mertelsman R, Kilitz JE, Bradley EC: Pharmacokinetics of recombinant
interleukin 2 in humans. Cancer Res 50:2009, 1990
26. Barbet J, Le Doussal JM, Gruaz-Guyon A, Martin M, Gautherot E, Delage M: Computer calculation of multiple binding equilibrium isotherms: Application to the binding of bivalent ligands to
antibodies interacting with cell surface Fc-receptors. J Theor Biol
27. Hibi M, Murakami M, Saito M, Hirano T, Taga T, Kishimoto
T: Molecular cloning and expression ofan IL-6 signal transducer,
gp130. Cell 63:1149, 1990
28. Narazaki M, Yasukawa K, Saito T, Ohsugi Y, Fukui H, Koishihara Y, Yancopoulos GD, Taga T, Kishimoto T: Soluble forms
of the interleukin-6 signal transducing receptor component gp130
in human serum possessing a potential to inhibit signals through
membrane anchored gp130. Blood 82: 1 120, 1993
29. Nesbitt JE, Fuller GM: Differential regulation of interleukin6 receptor and gp130 gene expression in rat hepatocytes. Mol Biol
Cell 3:103, 1992
30. Rose-John S, Schooltink H, Lenz D, Hipp E, Dufhues G,
Schmitz H, Schiel X, Hirano T, Kishimoto T, Heinrich PC: Studies
on the structure and regulation of the human hepatic interleukin-6
receptor. Eur J Biochem 190:79, 1990
31. Brailly H, Yasukawa K, Montero-Jullian FA, Nakano K,
Klein B: Soluble IL-6Rbetdgpl30 is present at high concentration
in normal serum as a dimer which binds efficiently to IL-6IslL6Ralpha complexes. Leucocyte Typing V 1994 (in press)
32. Taga T, Narazaki M, Yasukawa K, Saito T, Miki D, Hamaguchi M, Davis S, Shoyab M, Yancopoulos GD, Kishimoto T: Functional inhibition of hematopoietic and neurotrophic cytokines by
blocking the interleukin-6 signal transducer gp130. h o c Natl Acad
Sci USA 89:10998, 1992
33. Baumann H, Isserof H, Latimer JJ, Jahreis GP: Phorbol ester
modulates interleukin-6 and interleukin-] regulated expression of
acute phase plasma proteins in hepatoma cells. J BiolChem
263:17390, 1988
34. Gaillard JP, Bataille R, Brailly H, Zuber C, Yasukawa K,
From by guest on November 14, 2014. For personal use only.
Attal M, M a o N, Taga T, Kishimoto T, Klein B: Increased and
highly stable levels of functional soluble interleukin-6 receptor in
sera of patients with monoclonal gammopathy. Eur J Immunol
23:820, 1993
35. Ramsden L, Rider C : Selective and differential binding of
interleukin-la, IL-lp, L-2,and L - 6 to glycosaminoglycans. Eur J
Immunol 22:3027, 1992
36. MayLT, Santhanam U, Sehgal PB: On the multimeric
nature of natural human interleukin-6. J Biol Chem 266:9950,
37. Castell J V , Geiger T, Andus T, Walter E, Hirano T, Kishimoto
T, Heinrich Pc: Plasma clearance, organ distribution and target cells
of interleukin-6/hepatocyte-stimulatingfactor in the rat. Eur J Biochem 177:357, 1988
38. Castell JV, Gomez-Lechon MJ, David M, Fabra R, Trullenque
R, Heinrich P c : Acute phase response of human hepatocytes: Regulation of acute phase protein synthesis by interleukin-6. Hepatology
12:1179, 1990
39. Montero-Jullian FA, Klein B, Gautherot E, Brailly H: Pharmacokinetic study of anti-interleukin-6 (E-6) therapy with monoclonal antibodies: Enhancement of IL-6 clearance by coktails of
anti-L-6 antibodies. Blood 85:917, 1995