Occurrence of antibodies against natalizumab in relapsing multiple sclerosis patients treated with natalizumab.
ABSTRACT In the clinical trials about 9% of natalizumab treated multiple sclerosis (MS) patients generated anti-natalizumab antibodies, of which 6% were persistent and 3% transient. The occurrence of antibodies reduced serum levels of natalizumab, decreased bio-efficacy, and abrogated the therapeutic efficacy.
The objective was to assess the frequency of anti-natalizumab antibodies in an unselected cohort of patients from four different countries.
We measured anti-natalizumab antibodies in a large cohort of 4881 unselected patients from four MS centres that systematically measured antibodies in patients treated with natalizumab. We applied the same ELISA assay developed by Biogen Idec and used in the pivotal trials of natalizumab.
Antibodies occurred in 4.5% (95% confidence interval, CI: 4.0-5.1%) of the patients, and were persistent in 3.5% (95% CI: 3.0-4.0%) and transient in 1.0% (95% CI: 0.7-1.3%) of the patients. The frequencies of permanently antibody positive patients did not show statistically significant differences between the four centres, whereas the frequencies of transiently antibody positive patients showed some variations.
The frequencies of antibodies appeared to be of the same magnitude in the four centres, but might be less than in the pivotal studies of natalizumab.
Multiple Sclerosis Journal
The online version of this article can be found at:
2011 17: 1074 originally published online 20 April 2011Mult Scler
Nils Koch-Henriksen, Anna Fogdell-Hahn, Kjell-Morten Myhr, Jan Hillert and Ralf Gold
Per Soelberg Sørensen, Poul Erik Hyldgaard Jensen, Aiden Haghikia, Malin Lundkvist, Christian Vedeler, Finn Sellebjerg,
Occurrence of antibodies against natalizumab in relapsing multiple sclerosis patients treated with
can be found at:
Multiple Sclerosis Journal
Additional services and information for
What is This?
- Apr 20, 2011 OnlineFirst Version of Record
- Aug 16, 2011Version of Record >>
by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013 by guest on October 11, 2013msj.sagepub.commsj.sagepub.com msj.sagepub.commsj.sagepub.commsj.sagepub.commsj.sagepub.comDownloaded from Downloaded from Downloaded from Downloaded from Downloaded from Downloaded from
Occurrence of antibodies against
natalizumab in relapsing multiple sclerosis
patients treated with natalizumab
Per Soelberg Sørensen1, Poul Erik Hyldgaard Jensen1,
Aiden Haghikia2, Malin Lundkvist3, Christian Vedeler4,
Finn Sellebjerg1, Nils Koch-Henriksen5, Anna Fogdell-Hahn3,
Kjell-Morten Myhr4, Jan Hillert3and Ralf Gold2
Background: In the clinical trials about 9% of natalizumab treated multiple sclerosis (MS) patients generated anti-
natalizumab antibodies, of which 6% were persistent and 3% transient. The occurrence of antibodies reduced serum
levels of natalizumab, decreased bio-efficacy, and abrogated the therapeutic efficacy.
Objective: The objective was to assess the frequency of anti-natalizumab antibodies in an unselected cohort of patients
from four different countries.
Methods: We measured anti-natalizumab antibodies in a large cohort of 4881 unselected patients from four MS centres
that systematically measured antibodies in patients treated with natalizumab. We applied the same ELISA assay developed
by Biogen Idec and used in the pivotal trials of natalizumab.
Results: Antibodies occurred in 4.5% (95% confidence interval, CI: 4.0–5.1%) of the patients, and were persistent in
3.5% (95% CI: 3.0–4.0%) and transient in 1.0% (95% CI: 0.7–1.3%) of the patients. The frequencies of permanently
antibody positive patients did not show statistically significant differences between the four centres, whereas the fre-
quencies of transiently antibody positive patients showed some variations.
Conclusion: The frequencies of antibodies appeared to be of the same magnitude in the four centres, but might be less
than in the pivotal studies of natalizumab.
disease modifying therapies, immunology, multiple sclerosis, Tysabri, antibodies
Date received: 4th December 2010; revised: 23rd February 2011; accepted: 25th February 2011
The beneficial effects of natalizumab in the treatment of
relapsing–remitting multiple sclerosis (MS) is well doc-
umented, both in the setting of clinical trials1–4and in
Natalizumab is a monoclonal antibody directed
against the a4component of the a4b1integrin (VLA4;
CD49d) expressed on the surface of lymphocytes and
monocytes.9By binding to a4b1integrin, which is a
mediator of transendothelial leukocyte migration, nata-
lizumab blocks the migration of T-cells into the central
nervous system.10Natalizumab was originally gene-
rated in mice, but humanized by grafting the antibody
complementarity determining regions into a human
IgG4 antibody frame, but is still immunogenic like
Copenhagen University Hospital Rigshospitalet, Denmark.
2Department of Neurology,
Neuroscience, Karolinska Institutet, Sweden.
4Department of Neurology, Haukeland University Hospital, University of
5The Danish MS Treatment Register, Copenhagen University Hospital
Professor Per Soelberg Sørensen, MD, DMSc, Danish Multiple Sclerosis
Center, Department of Neurology 2082, Copenhagen University
Hospital Rigshospitalet, DK-2100 Copenhagen, Denmark
Multiple Sclerosis Journal
! The Author(s) 2011
Reprints and permissions:
other biopharmaceuticals.11In the clinical trials about
9% of natalizumab-treated MS patients generated anti-
natalizumab antibodies, and the presence of these
resulted in reduced serum levels of natalizumab,
decreased bio-efficacy, and abrogation of the therapeu-
In the clinical trials, binding anti-natalizumab anti-
bodies were measured using a sandwich/bridging
ELISA method12and a flow-cytometric assay to
detect blocking anti-natalizumab antibodies. The block-
ing assay assessed the ability of natalizumab specific
antibodies to block the binding of biotinylated natali-
zumab to its cognate a4receptor.12All patients with
antibodies measured by ELISA were also antibody-
positive in the blocking assay.12The anti-natalizumab
antibodies in MS patients typically had developed 12
weeks after initiation of treatment.3,4,12The AFFIRM
study showed that 88% of the eventually antibody-posi-
tive patients had developed antibodies when tested 12
weeks after initiation of natalizumab therapy.3,4,12
About 6% of the patients with antibodies to natalizu-
mab were persistently antibody positive, defined as anti-
bodies detected at two or more times with an interval of
at least 42 days,3,4,12whereas about 3% of the antibody-
generating patients were transiently antibody positive
and in these patients the antibodies normally disap-
peared after a few months of treatment.
The occurrence of natalizumab-binding antibodies
outside the setting of a clinical trial is unknown.
Therefore we decided to investigate the occurrence of
natalizumab antibodies in a large cohort of unselected
patients from four MS centres that are systematically
testing all local patients treated with natalizumab for
The study comprised consecutive patients from four
large MS centres at which all patients starting therapy
with natalizumab routinely are tested for the presence
of anti-natalizumab antibodies in the blood. Eligible
patients had been treated with natalizumab for at
least 48 weeks and had had at least two tests for anti-
bodies performed with an interval of at least 12 weeks,
of which one test had been performed six months or
later after start of natalizumab therapy. Blood samples
for antibody measurements were obtained after 12, 24
and 48 weeks or after 24, 36 and 48 weeks from initia-
tion of natalizumab therapy. Patients were included
from June 2006 to March 2010.
All four centres used the same ELISA developed by
Biogen Idec (Maine Biotechnology Services Inc.,
Portland, ME, USA). All laboratories had to analyse
high positive and low positive quality control samples
distributed by Biogen Idec. All patient serum samples
were tested for antibodies against natalizumab using
the ELISA procedure as prescribed by Biogen Idec.
Biogen Idec provided all material for control of the
assay (natalizumab reference standard, biotinylated-
natalizumab, high positive quality control 1 (QC1)
(3mg/ml of the monoclonal antibody 12C4) and low
positive quality control 2 (QC2) (0.5mg/ml of the
monoclonal antibody 12C4).
The negative control (NC) was derived from a pool
of non-treated healthy people serum samples (Pel-freez
Clinical Systems, VI, USA [cat. no. 34005/100]).
Microtitre plates (cat. no. 3590) were from Corning
Laboratories, NY, USA. The buffers used in the
assay were: 0.1M sodium phosphate, pH 8.5 with
2mg/ml bovine serum albumin (BSA) as plate coating
buffer; 0.1M Tris, pH 7.7 with 20% sucrose as plate
Tween-20 as plate washing buffer; and phosphate-buf-
fered saline with 0.5% BSA and 0.05% Tween-20 as
assay diluent. Sucrose (cat. no. 84097), Tween-20 (cat.
no. P2287), and BSA (cat. no. A3059) were obtained
from Sigma Life Science. Tris-HCl, pH 7.5 and PBS (10
times concentrated) were from Gibco (cat. no. 15567-
027 and cat. no. 14200, respectively). Streptavidin-
horseradish peroxidase was from Pierce, IL, USA
(cat. no. 21127), TMB 1 ready to use substrate from
KEM-EN-TEC Diagnostics (cat. no. 4380H) and
STOP solution H2SO4 was from Bie & Berntsen,
Denmark (cat.no. 1.00731).
Wellwash AC (Thermo Electron Corporation, USA)
for washing the plates and Biotech Synergy HT
(Biotech, USA) to measure absorbance at 400 or
All materials and chemicals are described in the
Biogen Idec protocol: ‘Assay procedure to determine
180AP-R.3)’. All samples and controls were tested at
1:10 dilutions in the ELISA. The requirement for a
positive sample by screening was that the analysed sam-
ple’s optical density (OD) value was greater than or
equal to 0.5mg/ml (QC2) of the monoclonal anti-nata-
lizumab antibody (12C4) and the ratio of OD or the
competition result divided by the OD of the screening
result was lower than or equal to 0.5.
Persistently positive patients
patients with at least two positive samples taken at an
interval of at least 12 weeks and no following nega-
tive samples. A transiently positive patient was
defined as a patient with a positive sample followed
by a subsequent negative sample taken at an inter-
val of at least 12 weeks and no following positive
saline with 0.05%
We used Thermo
were defined as
Sørensen et al. 1075
In order to validate the inter-assay variation between
the four laboratories 20 samples chosen from patients
Sclerosis Center, Copenhagen were evaluated blindly
in all four laboratories. The blood sampling had been
performed between June 2006 and June 2009 and had
been selected to include both negative samples (N¼10)
and positive samples (N¼10) with low and high con-
centrations of anti-natalizumab antibodies.
A total of 4873 patients fulfilled the inclusion criteria:
The Norwegian MS Competence Centre, Haukeland
University Hospital, Bergen, Norway: 235 patients;
Bochum MS Clinic, Bochum, Germany: 3526 patients;
Danish Multiple Sclerosis
Denmark: 399 patients; and NAb lab, MS Research
Stockholm, Sweden: 713 patients. Patient demo-
graphics are shown in Table 1.
Information about the anti-IFN-b neutralizing anti-
body (NAb) status was available in the Danish patient
cohort: 98 out of 298 patients (32.9%) had developed
NAbs during IFN-b treatment (defined as at least one
positive NAb test), of whom seven (7.6%) developed
anti-natalizumab antibodies. In comparison, 14 out of
199 (7.0%) IFN-b NAb negative patients developed
In all, 95.5% (95% confidence interval, CI: 94.9–
96.1%) of the patients remained persistently antibody
negative (Table 2), whereas on average 4.5% (95% CI:
4.0-5.1%) were antibody positive at anytime (range
4.0–7.7%). The antibody positive patients included
3.5% (95% CI: 3.0–4.0%) (range 2.4–5.1%) persis-
tently antibody positive, and 1.0% (95% CI: 0.7–
1.3%) (range 0.4–3.0%) transiently antibody positive
The 20 blindly evaluated samples were classified
identically as positive or negative in all cases but one.
This was a sample with low-level antibodies that tested
positive in Copenhagen but negative in the other three
The occurrence of antibodies against natalizumab did
not vary significantly across the four participating cen-
tres (Table 1). The same validated assay was used in all
four laboratories in order to minimize the inter-labora-
tory variation. All four laboratories had blindly been
analysing the same 20 quality control samples from the
Danish Multiple Sclerosis Center, Copenhagen. Indeed,
Table 2. Percentage antibody-positive and antibody-negative patients with 95% confidence intervals (CI) in consecutive patients who
after at least 3 months of natalizumab therapy had two or more tests for antibodies against natalizumab performed with an interval of
at least 3 months
N (%; 95% CI)
N (%; 95% CI)
N (%; 95% CI)
N (%; 95% CI)
October 2006 217 (92.3; 88.9–95.7)18 (7.7; 4.4–11.1)12 (5.1; 2.3–7.9)6 (2.6; 0.3–4.0)
3384 (96.0; 95.3–96.6)
371 (93,0; 90.5–95.5)
680 (95.4; 93.9–97.0)
142 (4.0; 3.4–4.7)
28 (7,0; 4.5–9.5)
33 (4.6; 3.1–6.1)
127 (3.6; 3.0–4.2)
16 (4,0; 2.1–5.9)
17 (2.4; 1.3–3.5)
15 (0.4; 0.2–0.7)
12 (3,0; 1.3–4.7)
16 (2.2; 1.1–3.3)
June–October 2006 4660 (95.5; 94.9–96.1)221 (4.5; 4.0–5.1)172 (3.5; 3.0–4.0) 49 (1.0; 0.7–1.3)
Table 1. Patient demographics
Age (years), median (range)
Disease duration (years), median (range)
Previous IFN-b treatment, proportion (%)
NANA 355/399 (89.0%)575/713 (80.8 %) 930/1112 (83.6)*
1076Multiple Sclerosis Journal 17(9)
the concurrent results of antibody measurement of
these samples indicated that a satisfactory standardiza-
tion was achieved. The discrepancy in the test result of
one test sample close to the OD value of QC2 might be
due to variations between laboratories in the QC2
batches used or different storing and transportation
conditions of reagents and test samples.
Our finding of 4.5% anti-natalizumab antibody pos-
itive patients is lower than the frequencies reported in
the AFFIRM and SENTINEL studies.12
AFFIRM study 57 out of 625 patients (9.1%; 95%
CI: 6.9–11.4%) tested positive of whom 37 patients
(5.9%; 95% CI: 4.1–7.8%) were persistently positive
and 20 patients (3.2%; 95% CI: 1.8–4.6%) were tran-
siently positive.3In the SENTINEL study 12% of the
patients tested positive for antibodies at any time
during the study; approximately 6.5% were persistently
and 5.5% transiently positive.12However, the 95% CI
for persistently antibody positive patients in our study
(3.5%; 95% CI: 3.0–4.0%) and in the AFFIRM study
(5.9%; 95% CI: 4.1–7.8%) were almost overlapping3.
We were able to corroborate the previously reported
observations from two of the participating centres,
Stockholm and Copenhagen, that there seems to be
no association between previous development of
NAbs against IFN-b and occurrence of antibodies to
In the majority of patients in both the AFFIRM
study (88%) and SENTINEL study (96%) anti-natali-
zumab antibodies were detectable by 12 weeks and the
remaining antibody positive patients exhibited detect-
able antibodies by week 24 apart from one single
patient, who first became antibody positive by week
36 after initiation of natalizumab therapy.12In our
study eligible patients should have a test performed at
week 24 or later after treatment start, and all perma-
nently antibody positive patients should in all proba-
bility have been detected and classified as antibody
positive patients. However, if some patients were
tested only once in the first year with a negative test,
and had a second negative test at a time point after
12 months, these persistently antibody negative patients
would not be included in our study. This could intro-
duce a bias towards enrichment of the study with anti-
body positive patients.
Contrary, if some patients had their first anti-nata-
lizumab antibody test performed at six months or later
after the start of natalizumab therapy, these patients
may have been antibody positive before and have
reverted to antibody negativity. We did not require sys-
tematically testing for antibodies at three months after
initiation of natalizumab treatment, and some patients
might have been antibody positive at three months and
antibody negative at six months. These patients would
be false negative patients and should correctly have
been classified as transiently antibody positive. This is
in accordance with our findings of lower frequencies of
transient antibody positive patients, in particular in
Bochum where testing at three months was not rou-
Further, in daily practice natalizumab is discontin-
ued in patients with two positive antibody tests sepa-
rated by an interval of at least 12 weeks, whereas in the
AFFIRM study antibody positive patients continued
therapy and had the possibility of reverting to antibody
negative status.12In the AFFIRM study 19 persistently
positive patients were treated for the full two years of
the study at which time 10 patients (53%) had reverted
to antibody negative status, some of them during the
second year of treatment.12In our study such patients
would have been classified as persistently antibody pos-
itive, and natalizumab treatment would have been ter-
minated. Discontinuation of treatment in persistent
antibody positive patients is recommended based on
the observation in the AFFIRM trial that showed
that the relapse rate in persistent antibody positive
patients was similar to that seen in the placebo treated
patients3. Hence, it cannot be recommended to con-
tinue natalizumab therapy in these patients, who are
virtually untreated, when other effective treatments,
e.g. fingolimod and, in some countries, cladribine, are
In conclusion, the results from our study of close to
5000 patients systematically tested for anti-natalizumab
antibodies confirm the occurrence of antibodies in
patients treated with natalizumab. The frequency of
persistent antibodies might be lower than that of 6%
reported in the pivotal trials of natalizumab.
The authors wish to acknowledge the support of Susan Goelz,
PhD, Director, Neurology at Biogen Idec in establishing the
ELISA assay for natalizumab antibodies.
This research received no specific grant from any funding
agency in the public, commercial, or not-for-profit sectors.
Conflict of interest statement
1. Miller DH, Khan OA, Sheremata WA, Blumhardt LD,
Rice GP, Libonati MA, et al. A controlled trial of natali-
zumab for relapsing multiple sclerosis. N Engl J Med 2003;
2. O’Connor PW, Goodman A, Willmer-Hulme AJ, Libonati
MA, Metz L, Murray RS, et al. Randomized multicenter
trial of natalizumab in acute MS relapses: clinical and
MRI effects. Neurology 2004; 62: 2038–2043.
Sørensen et al.1077
3. Polman CH, O’Connor PW, Havrdova E, Hutchinson M,
Kappos L, Miller DH, et al. A randomized, placebo-con-
trolled trial of natalizumab for relapsing multiple sclerosis.
N Engl J Med 2006; 354: 899–910.
4. Rudick RA, Stuart WH, Calabresi PA, Confavreux C,
Galetta SL, Radue EW, et al. Natalizumab plus interferon
beta-1a for relapsing multiple sclerosis. N Engl J Med
2006; 354: 911–923.
5. Oturai AB, Koch-Henriksen N, Petersen T, Jensen PE,
Sellebjerg F and Sorensen PS. Efficacy of natalizumab in
multiple sclerosis patients with high disease activity: a
Danish nationwide study. Eur J Neurol 2009; 16: 420–423.
6. Putzki N, Kollia K, Woods S, Igwe E, Diener HC and
Limmroth V. Natalizumab is effective as second line ther-
apy in the treatment of relapsing remitting multiple scle-
rosis. Eur J Neurol 2009; 424–426.
7. Putzki N, Yaldizli O, Maurer M, Cursiefen S, Kuckert S,
Klawe C, et al. Efficacy of natalizumab in second line
therapy of relapsing–remitting multiple sclerosis: results
from a multi-center study in German speaking countries.
Eur J Neurol 2009; 17: 31–37.
8. Piehl F, Holmen C, Hillert J and Olsson T. Swedish nata-
lizumab (Tysabri) multiple sclerosis surveillance study.
Neurol Sci 2011; 31(Suppl. 3): S289–S293.
9. Rudick RA and Sandrock A. Natalizumab: alpha
4-integrin antagonist selective adhesion molecule inhibi-
tors for MS. Expert Rev Neurother 2004; 4: 571–580.
10. Stuve O and Bennett JL. Pharmacological properties,
toxicology and scientific rationale for the use of natalizu-
mab (Tysabri) in inflammatory diseases. CNS Drug Rev
2007; 13: 79–95.
11. Sorensen PS, Tscherning T, Mathiesen HK, Langkilde
AR, Ross C, Ravnborg M, et al. Neutralizing antibodies
hamper IFNbeta bioactivity and treatment effect on MRI
in patients with MS. Neurology 2006; 67: 1681–1683.
12. Calabresi PA, Giovannoni G, Confavreux C, Galetta SL,
Havrdova E, Hutchinson M, et al. The incidence and signif-
13. Lundkvist M, Hillert J and Fogdell-Hahn A. No
increased risk for antibody formation against natalizu-
mab in multiple sclerosis patients positive for neutralising
16(Suppl. 10): S296.
14. Sorensen PS, Koch-Henriksen N and Jensen X. PEH.
Neutralizing antibodies against interferon-beta do not
predispose antibodies against natalizumab. Neurology
2011; 76: 759–760.
1078 Multiple Sclerosis Journal 17(9)