Note: Descriptions are shown in the official language in which they were submitted.
CA 02732445 2011-01-28
WO 2010/012834 1 PCT/EP2009/059982
METHOD FOR THE IN VITRO DIAGNOSIS OF STROKE
Field of the invention
The present invention relates to methods and kits for the in vitro diagnosis
of
stroke.
Background of the invention
Stroke, also known as cerebrovascular accident (CVA), is one of the leading
causes of mortality and morbidity with an estimated 700,000 patients diagnosed
with
stroke each year. Stroke currently ranks third in the cause of death in the
U.S.A.
The term "stroke" encompasses two widely different clinical settings which it
is
of the utmost importance to distinguish. Ischemic stroke is thus usually
caused by the
blockage of blood vessels and is best treated by clot dissolving agents, such
as t-PA,
within three hours of symptom onset. In contrast, hemorrhagic stroke is caused
by
bleeding into the brain which forbids any treatment by anti-clotting agents,
which
could prove fatal.
Transient ischemic attack (TIA, often colloquially referred to as "mini
stroke") is
caused by the changes in the blood supply to a particular area of the brain,
resulting
in brief neurologic dysfunction that persists, by definition, for less than 24
hours; if
symptoms persist then it is categorized as a stroke (see e.g. Transient
Ischemic
Attacks: Stroke (CVA): Merck Manual Home Edition). Patients diagnosed with a
TIA
are sometimes said to have had a warning for an approaching stroke. If the
time
period of blood supply impairment lasts more than a few minutes, the nerve
cells of
that area of the brain die and cause permanent neurologic deficit. One third
of the
people with TIA later have recurrent TIAs and one third have a stroke due to
permanent nerve cell loss (Transient ischemic attack Mount Sinai Hospital, New
York). Therefore, the identification of TIA is beneficial because these
patients are at
increased risk of future stroke.
The diagnosis of stroke, and the segmentation between ischemic and
hemorrhagic stroke, in patients which present with symptoms indicative of
stroke,
such as sudden numbness or blindness, confusion, severe headaches, slurred
speech, and partial paralysis, currently essentially relies on computed
tomography
(CT). CT, however, is not completely satisfying since it has an estimated
sensitivity of
less than 26% in diagnosing acute stroke (Chalela et al. (2007) Lancet 369:293-
298),
CA 02732445 2011-01-28
WO 2010/012834 2 PCT/EP2009/059982
which is linked to a very poor performance in detecting ischemic stroke, with
less
than 33% sensitivity (Reynolds et al. (2003) Clin. Chem. 49:1733-1739).
Magnetic
resonance imaging (MRI) has been shown to be superior to CT in diagnosing
acute
stroke (84% sensitivity, Chalela et al. (2007) Lancet 369:293-298), and
particularly
ischemic stroke. However MRI scanners are costly equipments and are not always
available in the emergency room.
Accordingly there is still the need for alternative or complementary methods,
in
particular to CT, for diagnosing stroke and TIA.
In this respect, biochemical markers have been suggested as an aid in
detecting stroke, in particular in view of the early detection of ischemic
stroke.
S-100b (a marker of astrocytic activation) and neuron-specific enolase (NSE)
are among the best characterized such markers (Jauch et al. (2006) Stroke
37:2508-
2513). Heart-type fatty acid binding protein(H-FABP) has also been considered
as a
promising marker (Lescuyer et al. (2005) Mol. Diagn. 9:1-7). However, it seems
that
the discriminatory power offered by these markers individually is not
sufficient to be
of clinical value.
It has thus been suggested to use panels combining several markers, such as
S-100b, the B-type neurotrophic growth factor (BNGF), the von Willebrand
factor
(vWF), matrix metalloproteinase-9 (MMP-9) and monocyte chemotactic protein-1
(MCP-1), for diagnosing ischemic stroke (Reynolds et al. (2003) Clin. Chem.
49:1733-1739). Indeed, this panel was shown to provide a sensitivity of 92% at
93%
specificity for ischemic stroke sample within 6 hours from symptom onset.
Within 3
hours from onset however, sensitivity is of only 87%, which might be due to a
too low
individual sensitivity/specificity of the markers.
Accordingly, there is still the need for alternative markers offering a good
individual sensitivity/specificity ratio to be used in single-marker tests or
to improve
multi-marker panels, either by increasing sensitivity/specificity or by
enabling
reducing the number of markers which levels have to be measured in panels.
proBNP(1-108), a precursor protein of 108 amino acids, is cleaved in vivo to
yield (i) Brain Natriuretic Peptide (also referred to as BNP(32) or simply
BNP), which
consists of the 32 C-terminal amino acids of proBNP(1-108) and (ii) NT-
proBNP(1-
76), which consists of the 76 N-terminal amino acids of proBNP(1-108)
(Giuliani et al.
(2006) Clinical Chemistry 52:1054-61). Biologically, BNP is a blood pressure
regulatory agent which is released mainly from the left cardiac ventricle in
response
CA 02732445 2011-01-28
WO 2010/012834 3 PCT/EP2009/059982
to volume expansion and pressure overload. proBNP(1-108) has been shown to be
circulating in patients with severe heart failure (Hammerer-Lercher et al.
(2008)
Clinical Chemistry 54:5).
Summary of the invention
The present invention arises from the unexpected finding, by the inventors,
that
proBNP(1-108) alone had a high discriminatory power (e.g. 95% sensitivity and
95 %
specificity as measured in plasma samples) in stroke detection.
Thus, the present invention relates to a method for the in vitro diagnosis of
stroke
and Transient ischemic attack (TIA) in an individual, comprising the following
steps:
(a) measuring the level of proBNP(1-108), or of fragments of proBNP(1-108)
comprising a RAPRSP sequence (SEQ ID NO: 1), in a biological sample of the
individual;
(b) comparing the measured level with a cut-off value;
(c) determining therefrom whether a stroke or a TIA has occurred in the
individual.
In an embodiment of the above defined method, the invention more particularly
relates to a method, comprising the following steps:
(a') measuring the level of proBNP(1-108), or of fragments of proBNP(1-108)
comprising a RAPRSP sequence (SEQ ID NO: 1), in a biological sample of the
individual;
(b') measuring the level of at least one further stroke marker in a biological
sample of the individual;
(c') comparing the level of proBNP(1-108), or of fragments of proBNP(1 -108),
and
the level of the at least one further stroke marker, with one or several cut-
off values ;
(d') determining therefrom whether a stroke or a TIA has occurred in the
individual.
In another embodiment of the invention, the above-defined method further
comprises measuring the level of at least one marker of cardiovascular
diseases.
The present invention also relates to a kit for diagnosing stroke, comprising:
- at least one antibody suitable for detecting proBNP(1-108), or fragments of
proBNP(1-108) which comprise a RAPRSP sequence (SEQ ID NO: 1); and
- at least one calibrator comprising proBNP(1 -108), or a fragment of proBNP(1
-108)
which comprise a RAPRSP sequence (SEQ ID NO : 1), preferably at a
concentration
of 1 to 1000 pg/ml.
CA 02732445 2011-01-28
WO 2010/012834 4 PCT/EP2009/059982
The present invention also relates to the use of proBNP(1-108), or of
fragments of proBNP(1-108) which comprise a RAPRSP sequence (SEQ ID NO: 1),
for the in vitro diagnosis of stroke or TIA.
Brief description of the figures
Figure 1 represents the distribution of H-FABP concentrations (vertical axis,
ng/ml)
in Stroke and Control populations.
Figure 2 represents the distribution of proBNP(1-108) concentrations (vertical
axis,
pg/ml) in Stroke and Control populations.
Figure 3 represents the distribution of proBNP(1-108) concentrations (vertical
axis,
pg/ml) in Stroke, TIA and Control populations.
Figure 4 represents the distribution of proBNP(1-108) concentrations (vertical
axis,
pg/ml) in Stroke and Control populations.
Figure 5 represents the distribution of 5-100b concentrations (vertical axis,
ng/ml) in
Stroke and Control populations.
Figure 6 represents the distribution of NSE concentrations (vertical axis,
ng/ml) in
Stroke and Control populations.
Detailed description of the invention
As intended herein "diagnosing" or establishing a "diagnosis" relates to
determining if a stroke has occurred in an individual.
As intended herein "stroke" relates to all cerebrovascular accidents. In
particular, the term "stroke" encompasses acute and chronic stroke as well as
ischemic and hemorrhagic stroke.
Ischemic stroke is characterized by a partial or total occlusion of cerebral
vessels which may lead to infarction and necrosis of cerebral tissues supplied
by
these vessels. In transient ischemic attack (TIA) the occlusion ceases
spontaneously
causing a dysfunction which lasts for no more than 24 hours.
CA 02732445 2011-01-28
WO 2010/012834 5 PCT/EP2009/059982
Hemorrhagic stroke is characterized by an intracerebral haemorrhage
generally from cerebral vessel rupture.
Preferably, the stroke according to the invention is selected from the group
consisting of an ischemic stroke and a hemorrhagic stroke. More preferably,
the
stroke as intended herein is acute ischemic stroke.
Advantageously, the method of the invention provides for early stroke
diagnosis. Early stroke diagnosis is of particular importance in the case of
ischemic
stroke, since it is usually estimated that treating the occluded vessels
within 3 hours
of stroke symptoms onset will prevent most irreversible cerebral damages.
Accordingly, it is preferred that the above defined step (a) is implemented
within 6
hours, more preferably within 3 hours, and most preferably within 2 hours,
after the
onset of at least one symptom indicative of stroke in the individual.
The symptoms indicative of stroke are well known to one of skill in the art
and
notably encompass sudden numbness or blindness, confusion, severe headaches,
slurred speech, and partial paralysis.
The individual is preferably a human.
"proBNP(1-108)" relates to the precursor BNP(32) and of NT-proBNP(1-76).
As intended herein "proBNP(1-108)" encompasses all its natural variants. By
way of
example proBNP(1-108) is represented by SEQ ID NO: 4.
In vivo, proBNP(1-108) is often partially truncated, in particular it is
deleted of
one or more amino acids on the N-terminal side or optionally on the C-terminal
side,
for instance by circulating proteases, to form so-called "proBNP(1-108)
fragments".
An example of such a proBNP(1-108) fragment, the proBNP(3-108) fragment
produced by cleavage by a dipeptidase, is described in Lam et al. (2007) J.
Am. Coll.
Cardiol. 49:1193-1202. It is believed that it is not only proBNP(1-108) which
is of
diagnosis value but also its various natural fragments. Accordingly, the
present
invention not only relies on measuring the level of proBNP(1-108) but also the
level
of fragments of proBNP(1-108).
The expression "proBNP(1-108)" and "fragments of proBNP(1-108)" also
include any polypeptide having been subjected to at least one post-
translational
modification, such as phosphorylation, glycosylation or the like. For example,
Schellenberger et al. (2006) Arch. Biochem. Biophys. 51:160-6 have shown that
proBNP(1-108) is a glycoprotein which is O-glycosylated either entirely or in
part.
CA 02732445 2011-01-28
WO 2010/012834 6 PCT/EP2009/059982
As intended herein the fragments of proBNP(1-108) comprise the RAPRSP
sequence (SEQ ID NO: 1). This RAPRSP sequence harbours the site of proBNP(1-
108) which is cleaved in vivo to yield NT-proBNP(1-76) and BNP(32). As such,
the
RAPRSP is specific of proBNP(1-108) and cannot be found in BNP(32) nor in NT-
proBNP(1-76), which are thus excluded from the definition of the fragments of
proBNP(1-108) according to the invention.
It is further preferred that the fragments of proBNP(1-108) according to the
invention comprise a FGRKMDR sequence (SEQ ID NO: 2). This sequence is
comprised in the BNP(32) part of proBNP(1-108). Even more preferably, the
fragments of proBNP(1-108) comprise the whole sequence of BNP(32) (SEQ ID NO:
3).
The expression "further stroke marker" relates to any biochemical marker,
other than proBNP(1-108) or fragments of proBNP(1-108) as defined above, which
level is indicative of stroke or TIA as defined above.
The expression "marker of cardiovascular diseases" relates to any marker
useful for detecting or diagnosing a cardiovascular disease. Such markers are
well
known to one of skill in the art. Preferably, the marker of cardiovascular
diseases is
selected from the group constituted of the C reactive protein (CRP) and
cardiac
troponin I (cTnl). Measuring the level of a marker of cardiovascular diseases
may be
advantageous in the the method of the invention since it enables excluding
cardiovascular diseases as a cause for variation of the level of proBNP(1 -
108).
Preferably, measuring or determining the level of proBNP(1 -108), of fragments
of proBNP(1-108), of the at least one further stroke marker, or of the at
least one
marker of cardiovascular diseases, is determined using an immunoassay.
As intended herein an "immunoassay" relates to any method wherein
the level of proBNP(1 -108), of fragments of proBNP(1 -108), of the at least
one further
stroke marker, or the at least one marker of cardiovascular diseases, is
determined
using at least one compound (or ligand) specifically binding thereto. The
compound
(or ligand) specifically binding thereto can be of any type but it is
preferred that it is
an antibody, an aptamer, or a peptide obtained by phage display. Immunoassay
methods are well known to one of skill in the art and may be carried out in
accordance with various formats well-known to the one skilled in the art, for
example
in solid or homogeneous phase, in one or two steps, by a sandwich method or by
a
competitive method.
CA 02732445 2011-01-28
WO 2010/012834 7 PCT/EP2009/059982
Preferably, the sandwich method in solid phase between 2 ligands, one being
a capture ligand and the other being a detection ligand, will be used. This
type of
immunoassay is particularly well-known to one skilled in the art. For example,
the
article by Seferian et al. (2007) Clin. Chem. 53:866-873 gives an example of a
sandwich immunoassay (or immunometric assay at 2 sites) for assaying BNP(32)
and proBNP(1-108), each time using a pair of antibodies (an antibody
immobilised in
solid phase and an labelled antibody in detection).
The presence of the antigen in the biological sample is revealed by detection
means, in particular a "detection ligand". A detection ligand, which is
labelled, is able
to bind to the captured antigen, by recognising an epitopic site which is
different from
that recognised by the capture ligand.
The term "labelled" refers both to a direct labelling and to an indirect
labelling
(for example, by means of other ligands, themselves directly labelled, or
using
reagents of a labelled "affinity pair", such as, but not exclusively, the
labelled avidin-
biotin pair, etc.).
In the case of the sandwich method, the capture ligand is preferably selected
in such a way that it specifically recognises an epitope on the natural
antigen of the
patient, whilst the detection ligand is selected preferably in such a way that
it
specifically recognises another epitope on the natural antigen of the patient.
Preferably, the capture ligand is immobilised on a solid phase. By way of non-
limiting examples of solid phase, microplates could be used, in particular
polystyrene
microplates, such as those sold by Nunc, Denmark. Solid particles or beads,
paramagnetic beads, such as those produced by Dynal, Merck-Eurolab (France)
(under the trademark EstaporTM) and Polymer Laboratories, or even polystyrene
or
polypropylene test tubes, glass, plastic or silicon chips, etc. may also be
used.
ELISA assays, radioimmunoassays, or any other detection method may be
used to reveal the presence of formed antigen-antibody complexes. Thus,
different
types of labelling of ligands in particular of antibodies, are possible
(radioactive,
ezymatic, fluorescent, etc.).
The detection may also be carried out by methods based on mass
accumulation, such as surface plasmon resonance (SPR), by piezo-electric
detection, but also by mass spectrometry or any other methods defined as
enabling
the study of a ligand-antigen-type interaction in the absence of a second
labelled
ligand.
CA 02732445 2011-01-28
WO 2010/012834 8 PCT/EP2009/059982
The term "specific", when it refers to recognition of a ligand or binding of a
ligand to a target, means that the ligand interacts with the target without
interacting
substantially with another target which does not structurally resemble the
target.
An "antibody" as intended herein relates to antibodies belonging to any
species, such as human, mouse, rat, rabbit, goat, or camelidae species. The
antibody can also be a chimeric antibody, i.e. an antibody which comprises
parts
originating from different species. Preferred chimeric antibodies are so-
called
"humanized" antibodies, wherein the constant parts (CH and CL) are of human
origin
and the variable parts (VH and VL) are of another species, such as mouse for
instance. The antibody of the invention can be produced by any method known
the
man skilled in the art, such as by animal immunization, or by recombinant or
synthetic methods for instance. Besides, an "antibody" according to the
invention
also encompasses antibody fragments which comprise at least one of the
paratopes
of said antibody, such as Fab, F(ab')2, scFv fragments as well as camelidae
single-
chain antibodies. The antibody of the invention can be a polyclonal antibody,
in
particular a monospecific polyclonal antibody, or a monoclonal antibody.
"Aptamers" are well-known by the one skilled in the art. Aptamers are
compounds of a nucleotide, in particular a ribonucleotide or
desoxyribonucleotide, or
a peptide nature able to bind specifically to a target, in particular a
protein target. The
aptamers of a nucleotide nature and the production thereof are described, in
particular, by Ellington et al. (1990) Nature 346:818-22 and Bock et al.
(1992) Nature
355:564-6. The aptamers of a peptide nature and the production thereof are
described, in particular, by Hoppe-Seyler et al. (2000) J. Mol Med. 78:426-30.
"Phage display" denotes a technique for selecting polypeptide ligands
expressed on the capsid of a bacteriophage and encoded by a nucleic sequence
inserted into the capsid encoding gene. This method is well known by the one
skilled
in the art and is described, in particular, by Scott & Smith (1990) Science
249:386-
390, and Marks et al. (1991) J. Mol. Biol. 222:581-597. Preferably, the
polypeptide
obtainable by phage display is an scFv-type polypeptide (single-chain variable
fragment). This technique is described, in particular, by Winter et al. (1994)
Annu.
Rev. Immunol. 12:433-455.
Preferably, the above-defined immunoassay comprises an antibody targeting
an epitope which comprises the RAPRSP sequence (SEQ ID NO: 1). More
preferably, the antibody is secreted by the hybridoma deposited at the CNCM
CA 02732445 2011-01-28
WO 2010/012834 9 PCT/EP2009/059982
(Collection Nationale de Cultures de Microorganismes, Institut Pasteur, 25,
rue du
Docteur Roux, 75 724 Paris Cedex 15, France) under the Budapest Treaty on
April
29, 2005, under reference number 1-3073. Such an antibody is notably described
in
the international publication WO 2004/014952. This antibody is advantageous,
in that
it enables the specific detection of proBNP(1-108) and of all the fragments of
proBNP(1-108) according to the invention, with the notable exception of
BNP(32),
NT-proBNP(1-76) and their respective fragments, thereby ensuring obtaining the
full
diagnosis benefits of proBNP(1-108) and its various fragments.
Preferably also, the immunoassay comprises an antibody targeting an epitope
which comprises the FGRKMDR sequence. More preferably, the antibody is
secreted
by the hybridoma deposited by Bio-Rad (3 boulevard Raymond Poincare, 92430
Marnes la Coquette, France) at the CNCM (Collection Nationale de Cultures de
Microorganismes, Institut Pasteur, 25, rue du Docteur Roux, 75 724 Paris Cedex
15,
France) under the Budapest Treaty on April 13, 2007, under reference number I-
3746. Such an antibody is notably described in international application
PCT/EP2008/060188.
Advantageously, the antibody targeting an epitope which comprises the
RAPRSP sequence (SEQ ID NO: 1) and the antibody targeting an epitope which
comprises the FGRKMDR sequence are combined in a same immunoassay, thereby
enabling the specific detection of proBNP(1-108) or of fragments of proBNP(1-
108)
according to the invention.
Preferably, where proBNP(1-108), or fragments of proBNP(1-108) are
concerned, the cut-off value as defined above is of at least the mean level of
proBNP(1-108), or of fragments of proBNP(1-108) according to the invention in
biological samples obtained from an apparently healthy population of
individuals.
More preferably, the above-defined cut-off value is of at least the value
corresponding to the 75th percentile, the 95th percentile, or the 99th
percentile of the
levels of proBNP(1-108), or of fragments of proBNP(1-108) according to the
invention
obtained from an apparently healthy population of individuals. Most
preferably, this
cut-off value is of at least 0, 1, 2, 3, 5, 10, 50 or 100 pg/m1. Where the cut-
off value is
of at least 0 pg/m1 this means that the above defined steps a) and b), or a')
and c'), of
the method of the invention consist simply in determining whether proBNP(1-
108), or
fragments of proBNP(1-108), as defined above are present in the biological
sample
of the individual.
CA 02732445 2011-01-28
WO 2010/012834 10 PCT/EP2009/059982
Similarly, where the at least further stroke marker is concerned, the cut-off
value as defined above is preferably of at least the mean level of said at
least one
further stroke marker in biological samples obtained from an apparently
healthy
population of individuals.
It is well within the ordinary skills of one of skill in the art to determine
cut-off
values according to the invention. In particular, care should preferably taken
to
measure the level of proBNP(1 -108), of fragments of proBNP(1-108) according
to the
invention, or of the at least one further stroke marker, in biological samples
which are
of the same nature. As intended herein, "an apparently healthy population of
individuals" relates to individuals which preferably present none of the
symptoms
indicative of stroke as defined above.
Where the level of proBNP(1 -108), or of fragments of proBNP(1 -108), and the
level of the at least one further stroke marker, are compared with one or
several cut-
off values, this means that the level of proBNP(1-108), or of fragments of
proBNP(1-
108), on one hand, and the level of the at least one further stroke marker, on
the
other hand, can be each compared to respective cut-off values, or that they
can be
both compared to a single cut-off value.
It is preferred that the level of the at least further stroke marker or of the
at
least one marker of cardiovascular diseases is measured in the same biological
sample as that in which the level of proBNP(1-108), or of fragments of
proBNP(1-
108), is measured.
As intended herein, the expression "biological sample" includes both the
sample as taken and the sample which has been subjected to various treatments,
in
particular to render it suitable for use in the processes and methods
according to the
invention. The biological sample according to the invention can be of any
type,
however it is preferred that the biological sample is selected from the group
constituted of a blood sample, a serum sample, a plasma sample, a
cerebrospinal
fluid sample, a urine sample and a saliva sample.
CA 02732445 2011-01-28
WO 2010/012834 11 PCT/EP2009/059982
Examples
Example 1
1. Methods
a. Samples
The levels of proBNP(1-108) and H-FABP were determined in 70 serum
samples obtained from individuals in whom a stroke has occurred less than 3
hours
after onset (15 hemorrhagic stroke samples (HM) and 55 ischemic stroke samples
(IM)) and in 148 control serum samples from apparently healthy individuals.
b. Biomarker measurement
The proBNP(1-108) level was determined using the BioPlexTM2200
proBNP(1-108) assay (Bio-Rad).
The BioPlexTM2200 combines multiplex, magnetic bead and flow cytometry
technologies to provide multi-analyte detection on a fully automated random
access
platform. Magnetic particles (8 pm diameter, carboxyl-modified surface) are
dyed with
2 fluorophores (classification dyes, CL1 and CL2) which emit at distinct
wavelengths
and adsorb significantly at 635 nm. The reporter fluorophore, P-phycoerythrin
(PE)
was chosen for its high molar extinction coefficient, quantum yield,
resistance to
photobleaching, lack of self-quenching and stability. The detector
simultaneously
measures light at 3 wavelengths: the 2 classification dyes and the reporter
dye.
The BioPlexTM2200 proBNP(1-108) assay is a two-step sandwich fluorescence
immunoassay. In a first step, the BioPlexTM2200 system combines 50 pL of
patient
sample, magnetic dyed beads coated with the anti-proBNP(1-108) monoclonal
antibody secreted by the hybridoma deposited at the CNCM under reference
number
1-3073 and assay buffer into a reaction vessel. Then, after 11 minutes of
incubation
and wash cycles, anti-human BNP monoclonal antibody secreted by the hybridoma
deposited at the CNCM under reference number 1-3746 conjugated to
phycoerythrin
(PE) is added and incubated for 2 minutes. After removal of excess conjugate,
the
bead mixture is passed through the detector which identifies the dyed beads
and the
amount of antigens captured on the beads by the fluorescence of PE. After
calibration using a set of 6 distinct calibrators, the 3 levels of quality
controls and
patient samples results are expressed in pg/mL.
CA 02732445 2011-01-28
WO 2010/012834 12 PCT/EP2009/059982
Two Quality Control beads are also tested with each sample to enhance the
integrity of the overall system.
The H-FABP level was determined with the human H-FABP ELISA kit from
Hycult biotechnology according to the manufacturer instructions.
c. Statistical analysis
Distribution of biomarkers according to patient status was represented by
boxplot representation, which is a convenient way of graphically depicting
groups of
numerical data through their five-number summaries (the smallest observation,
lower
quartile (Q1), median (Q2), upper quartile (Q3), and largest observation).
Data have been normalized using the Box-Cox method to follow a Gaussian
distribution and enable statistical analysis (Box, G. E. P. and Cox, D. R.
(1964) An
analysis of transformations. JRSS B 26, 211-246).
Differential analysis between the control and stroke samples were done using
the following statistical tests: the Wilcoxon Rank Sum Test (Wilcoxon, F.
(1945).
Individual comparisons by ranking methods. Biometrics, 1, 80-83) which is non-
parametric (requires no assumption of statistical distribution and which can
be an
alternative to Student's t-test) and the Welch's test which is an adaptation
of
Student's t-test intended for use with two samples having possibly unequal
variances.
The analysis also includes the adjusted (corrected) versions of these tests by
the
Benjamini/Hochber method (Benjamini and Y. Hochberg (1995). Controlling the
False
Discovery Rate: a practical and powerful approach to multiple testing. J. R.
Statist.
Soc. B. Vol. 57: 289-300) was also performed.
The diagnostic performance of the markers was characterized by two indices:
sensitivity (ability to detect the diseased population) and specificity
(ability to detect
the control population). The result of a diagnostic test can be further
characterized by
determining the area under the curve (AUC) of a ROC (receiver operating
characteristic) analysis. The ROC curves are a graphical visualization of the
reciprocal relation between the sensitivity (Se) and the specificity (Sp) of a
test for
various concentrations (M.H. Zweig and G. Campbell (1993). "Receiver-operating
characteristic (ROC) plots: a fundamental evaluation tool in clinical
medicine".
Clinical chemistry 39 (8): 561-57).
CA 02732445 2011-01-28
WO 2010/012834 13 PCT/EP2009/059982
2. Results
a) Distribution of biomarkers by boxplot representation
Figures 1 and 2 respectively represent the distribution of H-FABP and proBNP(1-
108) concentrations in Stroke and Control populations.
Table 1: H-FABP and proBNP(1-108) levels in Control and Stroke populations:
Markers Populations Min Q stile Median 95% Cl Quartile Max IQR
H-FABP Control subjects 1.0 2.78 4.27 3.59 - 5.10 7.95 79.4 5.17
Stroke patients 0.2 2.24 3.60 3.07 - 4.21 5.80 61.0 3.55
proBNP Control subjects 0.0 0.40 1.20 0.90 -1.50 2.00 12.9 1.60
Stroke patients 0.0 9.03 31.05 18.15 - 53.10 91.83 1032.0 82.80
ProBNP(1-108) marker shows higher concentration levels in stroke samples
than in control samples. Whereas H-FABP concentrations measured show no
significant difference between the stroke patient samples and the control
samples.
b) Differential analysis
The statistical significance of the difference in H-FABP and proBNP(1-108)
levels between stroke samples and control samples was determined:
Table 2: Differential analysis between concentrations determined from the
Control
and the Stroke population for H-FABP and proBNP(1-108) markers
Markers pWILCOXbrut pWILCOXadj pWELCHbrut pWELCHadj
H-FABP 0.0131 0.016375 0.0006 0.001
proBNP 0.0001 0.000167 0.0001 0.00025
The difference in concentrations between the stroke samples and the control
samples is statistically significant for both markers (proBNP(1-108) and H-
FABP),
however the proBNP(1-108) presents particularly interesting characteristics
with very
low p-values (10-4) (Table 2).
C) ROC curve analysis
CA 02732445 2011-01-28
WO 2010/012834 14 PCT/EP2009/059982
Based on Receiving-Operating-Characteristic (ROC) analysis for clinical status
and as summarized in Table 3, the area under the curve for proBNP(1 -108)
succeed
in discriminating patients diagnosed for stroke from healthy subjects, with a
sensitivity of 90%, a specificity of 85%. Moreover, this discriminatory power
is very
high for both ischemic and hemorrhagic stroke patients with respective AUC of
0.936
and 0.935.
Table 3: Receiving-Operation-Characteristic analysis data for proBNP(1-108)
for
Stroke patients when compared to Control subjects
Cut-off
proBNP Sample sizes determined from Se Sp AUC IC 95%
normalized
values
Ctrl vs Stroke (148) vs (70) 1.189 90% 85% 0.935 0.883 - 0.965
Ctrl vs IM (148) vs (55) 2.409 82% 95% 0.936 0.874 - 0.969
Ctrl vs HM (148) vs (15) 2.579 80% 96% 0.935 0.793 - 0.982
As shown in table 4 below, the H-FABP marker has a low discriminatory power
for the stroke diagnosis with a sensitivity of 57%, a specificity of 54.7%.
Table 4: Receiving-Operation-Characteristic analysis data for H-FABP for
Stroke
patients when compared to Control subjects
Cut-off
H-FABP Sample sizes determined from Se Sp AUC IC 95%
normalized
values
Ctrl vs (148) vs (70) -1.190 57% 55% 0.612 0.532 - 0.687
Stroke
Ctrl vs IM (148) vs (55) -1.00 58% 55% 0.618 0.533 - 0.696
Ctrl vs HM (148) vs (15) -1.149 53% 55% 0.563 0.416 - 0.749
Example 2
1. Methods
a. Samples
The levels of proBNP(1-108) were determined in 23 serum samples obtained
from individuals in whom a stroke has occurred less than 3 hours after onset
(5
CA 02732445 2011-01-28
WO 2010/012834 15 PCT/EP2009/059982
hemorrhagic stroke samples (HM) and 18 ischemic stroke samples (IM)), in 7
serum
samples obtained from individuals in whom a TIA has been clinically diagnosed,
and
in 94 control serum samples from apparently healthy individuals.
b. Marker level determination
The proBNP(1-108) level was determined as indicated in Example 1.
c. Statistical analysis
Boxplot representation of proBNP(1 -108) levels distribution, and ROC analysis
were conducted as indicated in Example 1.
2. Results
a) Distribution of biomarkers by boxplot representation
Figure 3 represents the distribution of proBNP(1-108) concentrations in
Stroke, TIA
and Control populations.
Table 5: proBNP(1-108) distribution values in Control, TIA and Stroke
populations 1st min Q artile Median 95% Cl Quartile Max IQR
Control Subjects samples 0 0,0 0.2 0,0 - 0,8 1,7 35 1.7
Stroke Patients samples 1 3.1 10.3 3.3 -29.5 39.0 114 35.9
TIA Patients samples 3 5.1 7.2 3.4 - 51.7 25.8 52 20.7
ProBNP(1-108) marker shows higher concentration levels in stroke or TIA
patient
samples than in control samples.
b) ROC curve analysis
Based on Receiving-Operating-Characteristic (ROC) analysis for clinical status
and as summarized in Table 6, the area under the curve for proBNP(1 -108)
succeed
in discriminating patients diagnosed for stroke from healthy subjects, with a
sensitivity of 91 %, a specificity of 84%.
CA 02732445 2011-01-28
WO 2010/012834 16 PCT/EP2009/059982
Table 6: Receiving-Operation-Characteristic analysis data for proBNP(1-108)
for
Stroke patients when compared to Control patients
Sample Cut-off determined
proBNP sizes from raw values Se Sp AUC IC 95 /o
Ctrl vs (94) vs 91 0.720-
Stroke (23) 2.43pg/mL % 84% 0.913 0.989
proBNP(1-108) power for diagnosing stroke patients evidenced in Example 1
are thus confirmed on these new pathological and healthy populations.
Example 3
The discriminatory power of proBNP(1-108) was further compared to that of two
reference stroke markers of the prior (5-100b and NSE) art on identical
populations.
1. Methods
a. Samples
The levels of markers were determined in 41 plasma samples obtained from
individuals in whom a stroke has occurred less than 3 hours after onset and in
51
control plasma samples from apparently healthy individuals.
b. Marker level determination
The proBNP(1-108) level was determined as indicated in Example 1.
The NSE level was determined using the Nexus DXTM Neuron Specific
Enolase (NSE) Test kit from Nanogen according to the manufacturer
instructions.
The 5-100b level was determined using the Nexus DXTM 5-100 ELISA Test kit
from Nanogen according to the manufacturer instructions.
c. Statistical analysis
The statistical analysis was conducted as indicated in Example 1.
2. Results
a) Distribution of biomarkers by boxplot representation
CA 02732445 2011-01-28
WO 2010/012834 17 PCT/EP2009/059982
Figures 4, 5 and 6 respectively represent the distribution of proBNP(1 -108),
5-100b
and NSE concentrations in Stroke and Control populations.
Table 7: proBNP(1-108), S100 and NSE distribution values in Control and Stroke
populations Ilst Markers Populations Min Quartile Median 95% Cl Quatile Max
IQR
proBNP Control Subjects 0 0.3 1.4 0.7 to 1.,9 2.5 97 2.3
Stroke Patients 2 35.3 59.6 40.4 to 120.6 203.7 1019 168.3
5100 Control Subjects 0.00 0.000 0.002 0.001 to 0.003 0.003 0.02 0.003
Stroke Patients 0.00 0.003 0.004 0.004 to 0.006 0.006 0.02 0.003
NSE Control Subjects 0.88 2.846 4.414 3.413 to 4.988 6.229 30.33 3.383
Stroke Patients 3.87 6.766 12.085 8.165 to 15.607 20.843 53.26 14.077
ProBNP(1-108) and NSE markers show higher concentration levels in stroke
samples than in control samples. The best discrimination between both
populations
is observed with the proBNP(1-108) marker. Whereas S-100 concentrations
measured show no significant difference between the stroke patient samples and
the
control samples.
b) Differential analysis
The statistical significance of the difference in proBNP(1-108), NSE and S-100
levels in stroke samples and in control sample was determined:
Table 8: Differential analysis between concentrations determined from the
Control
and the Stroke populations for NSE, 5-100 and proBNP(1-108) markers
Markers pWILCOXbrut pWILCOXadj pWELCHbrut pWELCHadj
NSE 0.0001 0.0003 0.0001 0.0003
S-100 0.0023 0.0037 0.0064 0.0102
proBNP 0.0001 0.0003 0.0001 0.0003
NSE and proBNP(1-108) are statistically significant with very low p values (p<
0.001),
C) ROC curve analysis
CA 02732445 2011-01-28
WO 2010/012834 18 PCT/EP2009/059982
Based on Receiving-Operating-Characteristic (ROC) analysis for clinical status
and as summarized in Table 9, the area under the curve for proBNP(1 -108)
succeed
in discriminating patients diagnosed for stroke from healthy subjects, with a
sensitivity of 95%, a specificity of 95% and an AUC of 0.974.
Table 9: Receiving-Operation-Characteristic analysis data for proBNP(1-108),
NSE
and S100 for Stroke patients when compared to Control subjects
Ctrl vs Stroke Sample Cut-off Se Sp AUC IC 95%
sizes
proBNP(1-108) (55) vs (41) 3.26 95% 95% 0.974 0.904 - 0.993
NSE (55) vs (41) 1.682 85% 73% 0.857 0.770 - 0.915
5-100b (55) vs (41) -6.644 39% 89% 0.636 0.504 - 0.751
It can be seen that proBNP(1-108) presents a high discriminatory power over
that offered by NSE and 5-100b to discriminate the Stroke patient from the
Control
subjects.
Example 4
The diagnostic power of proBNP(1-108) was further assessed in patients with
ischemic stroke using the BioPlexTM2200 proBNP(1-108) assay as described in
Example 1.
a) Samples
- 32 citrated plasma samples from patients with ischemic stroke admitted to
the Emergency Department within 3 hours of the Stroke onset were tested. The
stroke severity was assessed by the National Institutes of Health Stroke Scale
(NIHSS).
- 42 citrated plasma samples from apparently healthy blood donor matched by
gender and age with the patients from the Stroke population were tested.
All the citrated plasma samples were stored at -80 C. Prior the analysis, the
samples were thawed and centrifuged at 3000g for 15 min at 4 C.
CA 02732445 2011-01-28
WO 2010/012834 19 PCT/EP2009/059982
b) Results
The distributions of the proBNP(1 -108) values for the Control and the
ischemic
stroke populations are shown in Table 10. The level of proBNP(1-108) was
significantly higher in the ischemic stroke group compared to the control
group
(Mann-Whitney, p<0.0001). The results confirm that the proBNP(1-108) is thus a
useful plasma biomarker for the early diagnosis of Ischemic stroke.
Table 10: proBNP(1-108) concentrations in ischemic stroke and control citrated
plasma samples (minimum, 1st quartile, median, 3rd quartile and maximum
values)
rd
Populations Minimum 1st Quartile Median Quartile Maximum
(pg/mL) (pg/mL) (pg/mL) (pg/mL) (pg/mL)
Control population 0 0 1 2 23
(N=42) (IC95%: 0-2)
Ischemic 71
Stroke population 2 34 (IC95%: 38-145) 219 1019
(N=32)
