WO2013017891A2

WO2013017891A2 - Assay

Info

Publication number: WO2013017891A2
Application number: PCT/GB2012/051885
Authority: WO
Inventors: Lu-Gang YU
Original assignee: The University Of Liverpool
Priority date: 2011-08-04
Filing date: 2012-08-03
Publication date: 2013-02-07
Also published as: GB201113482D0; WO2013017891A3

Abstract

Methods for detecting metastatic cancer in a subject, and products for use in the methods.

Description

ASSAY

Field of the invention

The invention relates to methods for detecting metastatic cancer, in particular, colorectal cancer. The methods can be used, to determine metastasis in cancer patients, and may also be used to screen for early metastatic cancer in asymptomatic subjects. The invention further relates to products such as kits for use in the methods.

Background to the invention

Cancer is a significant cause of morbidity and mortality worldwide. The disease is becoming more prevalent in many countries, increasing the need for reliable methods for screening, monitoring the progression of and treating the disease. Colorectal cancer (C C) in particular is increasing in incidence due in part to changes in diet.

In the diagnosis and treatment of CRC to date, carcinoembryonic antigen (CEA) and carbohydrate antigen 19-9 (CA19-9) have been used as tumour markers, for example, for monitoring tumour progression, for follow-up after surgery and for evaluation of ongoing treatments such as chemotherapy and radiation therapy.

However, despite their usefulness, there remain a significant number of CRC patients who do not show increases in these marker values. Furthermore, elevated levels sometimes result from factors other than the tumour. For example, abnormal CEA level in the blood is not specific for colon cancer nor for malignancy, and is also detected in benign conditions including cirrhosis, inflammatory bowel disease, chronic lung diseases and pancreatitis. CEA is also elevated in up to 19% of smokers and about 3% of healthy control population. Still further, in early stage CRC, their levels are too low to be used as screening markers.

Therefore there remains a need to develop alternative means for monitoring tumours, and patient progress in cancers such as CRC, which are sufficiently sensitive and specific for clinical use, as well as a need to develop efficient screening methods.

Summary of the invention

In addressing this problem, the inventors have focussed upon the galectin family of proteins, and on galectin- 3 and -4 (Gal3 and Gal4) specifically.

Galectins are a family of 15 mammalian galactoside binding proteins expressed by many types of human cells. Earlier studies have reported that Gal3 is significantly increased in the sera of patients having various cancers, such as colorectal (Barrow et al, 2011, Int. J Cancer: 129: 1-8), lung (lurisci et al, Clin Cancer Res, 2000, 6,1389-93), bladder (Sakaki et al, J Med Invest, 2008, 55, 127-132), head and neck (Saussez et al., Oral Oncology, 2008, 44, 86-93), melanoma (Vereecken et al., Melanoma Research, 2009, 19, 316-320; Vereecken et al., Clinical and Experimental Dermatology, 2005, 31, 105-109), pancreatic (Senapati et al., Clin Cancer Res, 2011, 17, 267-274), breast (lurisci et al, Clin

Cancer Res, 2000, 6,1389-93) and thyroid (Saussez et al., Thyroid, 2008, 18(7), 705-712 and Isic et al., J Cancer Res Clin Oncol, 2010, 136, 1805-1812) cancers. For colorectal cancer specifically, four human galectins, galectin-1, -3, -4 and -8, are expressed in the normal human colon and rectum (Barrow et al, 2011, Int. J Cancer: 129: 1-8).

Barrow et al (supra) report changes in expression of some galectins in colorectal cancer tissue. For example, expression of Gall and Gal3 is reported to be increased, and expression of Gal4 and Gal8 decreased. The authors also report higher levels of Gal3 in the colorectal tissue and the circulation of metastatic patients.

Watanabe et al (Watanabe et al 2011, Oncology Reports 25: 1217-1226) report that circulating levels of Gall, Gal3 and Gal4 are significantly higher in CRC patients than in healthy controls. The authors further report that plasma levels of Gall, Gal3 and Gal4 show a significant difference between Stage l/ll CRC patients v healthy controls, and between Stage lll/IV CRC patients and healthy controls (classified according to the UICC TNM Classification of Malignant Tumours) and that there is a correlation between Gal4 levels and tumour progression at the various stages. However, the data presented in that study did not report a significant difference in plasma levels of any of the three galectins between the CRC disease stages, e.g. between Stage lll/IV patients vs Stage 0 patients, or between Stage lll/IV patients vs Stage l/ll patients, i.e. between early and later stage (metastatic) patients.

The present inventors have developed a new test which, unlike the prior art methods, can reliably distinguish metastatic colorectal cancer from non-metastatic colorectal cancer. The test is based on a combination assay for both Gal3 and Gal4 (CG34 assay) which determines the combined concentration of both Gal3 and Gal4 in a blood sample simultaneously. The assay can significantly separate CRC patients having liver metastasis from CRC patients with a only primary tumour (Figure 1). The inventors have shown that the new assay can distinguish between the non-metastatic and metastatic patients with higher degree of significance than the conventional CEA test (Figure 1). Moreover the inventors have shown that the new assay demonstrates sensitivity and specificity suitable for clinical use, and performs better than the CEA assay (in terms of balance of

sensitivity/specificity) (Figure 4). Not only that, but the assay is simple to use and relatively inexpensive.

The inventors found that neither Gal3 nor Gal4 assayed individually, or indeed simple addition of the results obtained for Gal3 and Gal4 when assayed individually, can distinguish metastatic patients at a significant level in the way that the combined assay does (Figures 2 and 5).

Thus the inventors have shown that total Gal3/Gal4 blood concentration, as determined in a combination assay, provides a means of detecting metastatic colorectal cancer specifically. The present methods can be used to determine metastasis in CRC patients, or to screen for early stage but metastatic colorectal tumours in as yet asymptomatic subjects. The inventors believe that the invention may be also applicable to other cancers, in particular breast cancer, because circulating galectin-3 and -4 levels are, like that of CRC, also seen to be significantly increased in those patients.

The choice of treatment for cancers such as colorectal cancer typically depends upon the state of advancement of the cancer (the cancer stage), and in particular, whether and to what extent metastasis is occurring. For example, metastatic patients may require more extensive surgery, and/or the use of radiotherapy or chemotherapy before or after surgery. By providing a new means for accurately detecting metastasis, the present invention also permits better allocation of appropriate treatments to patients.

Accordingly in one aspect the invention provides a method for detecting metastatic cancer in a subject, the method comprising:

a) performing a combined assay for both Gal3 and Gal4 protein in a blood sample taken from the subject, thereby obtaining a value for the combined concentration of Gal 3 and Gal4 in the sample; and

b) comparing the value obtained in the combined assay of step (a) with a predetermined threshold value;

wherein metastatic cancer is indicated if a value is obtained in step (a) which is higher than the predetermined threshold value.

The method may be for determining metastasis in a cancer patient, or for detecting metastatic cancer in an symptomatic subject. The method may be for monitoring for tumour recurrence after surgical removal of a primary tumour. In one aspect, the cancer may be colorectal cancer. In one aspect the method of the invention may further comprise determining a suitable treatment for the subject or patient.

In one aspect, performing the combined assay in step (a) of the method of the invention may comprise:

a) contacting the sample with a binding agent which binds specifically to Gal3 protein and a binding agent which binds specifically to Gal4 protein; and b) determining the combined amount of bound Gal3 and Gal4 protein.

The binding agent which binds specifically to Gal3 protein may comprise an antibody which binds specifically to Gal3 protein or a fragment or derivative of said antibody and/or the binding agent which binds specifically to Gal4 protein may comprise an antibody which binds specifically to Gal4 protein or a fragment or derivative of said antibody.

The combined assay in step (a) of a method of the invention may comprise an ELISA sandwich assay.

In one aspect, the method for detecting cancer in a subject additionally comprises performing one or more additional tests indicative of metastasis in a sample taken from the subject. The one or more additional tests may comprise assaying the sample for one or more of the cytokines IL-6, G-CSF and slCAM-1.

In a further aspect, the invention also provides a kit for use in determining metastatic cancer in a subject, the kit comprising means for performing a combined assay for both Gal3 and Gal4 in a blood sample taken from the subject.

In one aspect the kit may comprise: a) an antibody which binds specifically to Gal3 protein or fragment or derivative of said antibody; and b) an antibody which binds specifically to Gal4 protein or a fragment or derivative of said antibody.

In one aspect a kit according to the invention may be for use in an ELISA sandwich assay to simultaneously determine the concentration of Gal3 and Gal4 in the blood sample.

In one aspect, the kit for determining metastatic cancer in a subject may comprise means for performing one or more additional tests indicative of metastasis in a sample taken from the subject. The one or more additional tests may comprise assaying the sample for one or more of the cytokines IL-6, G-CSF and slCAM-1.

In one aspect, a kit according to the invention may be suitable for use in any of the methods described herein.

In one aspect, a kit according to the invention may be for use in determining metastatic colorectal cancer in a subject.

In one aspect, a kit according to the invention may be for use in monitoring for tumour recurrence in a patient following surgical removal of a primary tumour.

Description of the Figures

Figure 1 - Comparison of carcinoembryonic antigen (CEA) assay (Figure 1A) and CG34 (combined Gal3 and Gal4) assay (Figure IB) results in the serum of healthy subjects (n=31), of C C patients with liver metastasis (n=ll) and of CRC patients without liver metastasis (n=40). The Y axis is presented in a log scale.

Figure 2 - Individual Gal3 and Gal4 tests in healthy subjects and CRC patients with and without liver metastasis. The Y axis is presented in a log scale.

Figure 3 - Reciprocal relationship of the levels of Gal3 and Gal4 in the serum of CRC patients with liver metastasis. In each of Figures 3A to 3D, the serum concentration of two galectins in each of the 11 metastatic CRC patients is shown as a percentage of the median concentration of that galectin across the group of 11 patients. Each line represents one of the 11 patients.

Figure 4 - Table 1: Specificity and sensitivity of the CEA test (at clinically used thresholds) in CRC patients; Table 2: Specificity and sensitivity of the Gal3/4 test at a range of thresholds in CRC patients. Figure 5 - Addition of Gal3 and Gal4 assay results following individual Gal3 and Gal4 assays in healthy subjects, and in CRC patients with and without liver metastasis. The Y axis is presented in a log scale.

Figure 6 - Amino acid sequences of human Gal3 (Accession number: BAA22164.1), human Gal4 (AAB86590.1), human IL6 (Accession number: AAH15511), human G-CSF (Accession number:

P09919), human slCAM-1 (Accession number: CAA41977) and human GM-CSF (Accession number: AAA52578) . The sequences of Gal3 and Gal4 are taken from the NCBI database as at 28 July 2011. The sequences of IL6, G-CSF, slCAM-1 and GM-CSF are taken from the NCBI database as at 31 July 2012 (versions AAH15511.1, P09919.1, CAA41977.1 and AAA52578.1 respectively).

Figure 7 - Table 3: Relationship between circulating galectin-3 and circulating IL-6, G-CSF, GM-CSF and slCAM-1 in colon cancer patients

Description of the sequences

SEQ ID NO: 1 - amino acid sequence of human Gal3. SEQ ID NO: 2 - amino acid sequence of human Gal4. SEQ ID NO: 3 - amino acid sequence of human IL6 SEQ ID NO: 4 - amino acid sequence of human G-CSF SEQ ID NO: 5 - amino acid sequence of human slCAM-1 SEQ ID NO: 6 - amino acid sequence of human GM-CSF Detailed description of the invention

Throughout the description and claims of this specification, the singular encompasses the plural unless the context otherwise requires. In particular, where the indefinite article is used, the specification is to be understood as contemplating plurality as well as singularity, unless the context requires otherwise.

Features, integers, characteristics, compounds, chemical moieties or groups described in conjunction with a particular aspect, embodiment or example of the invention are to be understood to be applicable to any other aspect, embodiment or example described herein unless incompatible therewith or specifically stated otherwise.

Throughout the description and claims of this specification, the words "comprise" and "contain" and variations of the words, for example "comprising" and "comprises", mean "including but not limited to", and are not intended to (and do not) exclude other moieties, additives, components, integers or steps.

Unless otherwise noted, technical terms are used according to conventional usage. Definitions of common terms in molecular biology may be found in Benjamin Lewin, Genes V, published by Oxford University Press, 1994 (ISBN 0-19- 854287-9); Kendrew et al. (eds.), The Encyclopedia of Molecular Biology, published by Blackwell Science Ltd., 1994 (ISBN 0-632-02182-9); and Robert A. Meyers (ed. ), Molecular Biology and Biotechnology: a Comprehensive Desk Reference, published by VCH Publishers, Inc., 1995 (ISBN 1-56081-569-8).

All publications, patents and patent applications mentioned in this specification are herein incorporated by reference in to the specification to the same extent as if each individual publication, patent or patent application was specifically and individually indicated to be incorporated herein by reference. Where this disclosure references Internet sites, the contents of the referenced Internet sites are incorporated herein by reference as of 28 July 2011. References to accession numbers for cytokines IL6, G-CSF, GM-CSF and slCAM-1 are as at 31 July 2012

All references to "detectable" or "detected" are as within the limits of detection of the given assay. As described above, the inventors have devised a new assay for reliably determining metastatic cancer, in particular colorectal cancer, based on simultaneous assay of circulating levels of Gal3 and Gal4 (CG34 assay). The assay provides a simple test for metastasis, which is sensitive and specific enough for use in a clinical setting. The ability to accurately detect metastasis means that the assay can be used to choose the most appropriate treatment for a subject, e.g. whether chemotherapy and/or radiotherapy will be necessary after surgical removal of a tumour. The test can also be used to monitor for recurrence of a tumour, e.g. a CRC tumour, after surgical removal.

Furthermore, the CG34 assay can also be used for early detection of early and otherwise silent (in terms of presenting clinical symptoms) but metastatic tumours, e.g. CRC tumours, in apparently healthy people. In this way, the invention also provides new screening methods for detecting early metastatic tumours, such as early metastatic CRC tumours.

Accordingly, in one aspect, the present invention provides a method for detecting metastatic cancer in a subject, the method comprising: a) performing a combined assay for both Gal3 and Gal4 in a blood sample taken from the subject, thereby obtaining a value for the combined concentration of Gal 3 and Gal4 in the sample; and b) comparing the value obtained in the combined assay of step (a) with a predetermined threshold value; wherein metastatic cancer is indicated if a value is obtained in step (a) which is higher than the predetermined threshold value.

Metastasis generally refers to the process by which cancer spreads from the primary tumour site to distant location(s) in the body and depends on the cancer cells acquiring two separate abilities: increased motility and invasiveness. Metastasis may also be used to refer to a secondary cancer resulting from the spread of the primary tumour.

Colorectal cancer

In a preferred embodiment herein, the metastatic cancer which is detected is colorectal metastatic cancer. Colorectal cancer generally refers to cancer occurring in the colon or the rectum.

A colorectal primary tumour may metastasise to any other tissue site but particularly to the liver and lung.

A number of clinical symptoms of colorectal cancer are currently recognised, including for example: changes in bowel movement, constipation, diarrhea, blood in the stools, bloating, cramps, unexplained weight loss, extreme tiredness and vomiting (see, for example, the NHS bowel cancer website at http://www.nhs.uk/Conditions/Cancer-of-the-colon-rectum-or- bowel/Pages/Symptoms.aspx). Under current practice, clinical diagnosis of colorectal cancer in a subject (typically presenting one or more symptoms) may involve one or more of: digital rectal examination; fecal occult blood test; barium enema; flexible sigmoidoscopy; colonoscopy; biopsy; and/or virtual colonoscopy. Typically, initial diagnosis is based on colonoscopy, often with biopsy. Determining whether metastasis is occurring is an important factor in choice of treatment for a patient. Current treatments for colorectal cancer include: surgery; radiotherapy; and chemotherapy (see, for example, http://www.nhs.uk/conditions/cancer-of-the-colon-rectum-or- bowel/Pages/treatment.aspx). Other treatments under investigation include targeted antibody treatments. The precise treatment regime chosen depends on the type and stage of the cancer. More advanced cancer, e.g. metastatic cancer, will receive more aggressive or extensive treatment.

At present, a number of staging systems for colorectal cancer are known. These include the TNM (Tumour/Node/Metastasis) system (Sobin LH and Wittekind Ch (eds): TNM Classification of malignant tumours. 5^th edition. John Wiley & Sons, Inc., New York 1997 in which patients are classified as Stage 0, I, II, III or IV, according to how far the tumour as penetrated the intestinal wall, whether and how far the cancer has spread to nearby lymph nodes, and whether the cancer has spread to other organs of areas of the body. A higher stage indicates more advanced disease, including metastasis. Another example is the Dukes system (Dollinger, M. (1997). Everyone's Guide to Cancer Therapy. Toronto : Somerville House Books Limited) in which patients are classified as Stage A, B, C, or D, with D indicating the most advanced disease state.

At present, staging of a patient is usually done after the initial surgical examination. However, other factors may be taken into account in determining likely treatment, including type and size of the tumour, age and general health of the patient, or grade of the tumour under the microscope.

Under current practice, once a diagnosis of CRC is made, e.g. on the basis of a biopsy, a patient is usually referred for surgery, which allows a determination of the extent to which the cancer has spread, and assessment of the cancer stage.

Initial surgery is generally also therapeutic, as it may involve removal of the tumour (or as much of it as possible) as well as part of the large intestine, and possibly lymph node tissue, particularly that close to the primary tumour site. For early stage, e.g. non-metastatic cancer, surgical treatment alone may suffice. However, the more advanced the tumour is believed to be, the more tissue which will be taken. Radiotherapy treatment is not often used against colon cancer but may be used against rectal cancers, either before surgery to shrink a tumour, or after surgery to reduce the likelihood of recurrence. Radiotherapy may be used in combination with chemotherapy. Typically, more advanced tumours, e.g. metastatic tumours, are more likely to be treated with radiotherapy.

Chemotherapy treatment may be used against either colon or rectal cancer. Chemotherapy may be applied after surgery, with the aim of removing any remaining cancerous cells (adjuvant therapy), or before surgery to shrink a tumour. Typically, early stage cancer patients, e.g. non-metastatic patients, may not require chemotherapy, whereas later stage patients, e.g. metastatic patients, will receive the treatment. Chemotherapeutic agents currently in use include fluorouracil, which is often used in combination with other cancer drugs such as irinotecan or oxaliplatin. Following surgery, there is a risk of recurrence of the tumour and relapse. Currently, monitoring for recurrence may be done using, for example, the CEA and CA19-9 marker tests described above. Where recurrence is detected, or is believed to be more likely to occur, treatment may be altered accordingly, e.g. in the use of radiotherapy and/or chemotherapy follow-up to surgery. The present methods can be used to determine whether metastasis is occurring, and therefore allow a more informed decision to be taken, possibly at an earlier time, about what treatment regime is best.

Use in C C patients

In one aspect the present method may be used to determine metastasis in a CRC patient.

In this aspect, the colorectal cancer patient is in general a subject who presents with one or more clinical symptoms of colorectal cancer as described herein, and/or has been clinically diagnosed as having a colorectal tumour according to conventional means. In one aspect the patient has been diagnosed on the basis of flexible sigmoidoscopy or colonoscopy. In one aspect, a biopsy has been carried out to determine the presence of a colorectal tumour. Typically the patient is a human. The present methods may be used prior to initial surgery to determine metastasis. In this way, the clinician can better determine the extent of surgery that may be needed, and/or whether radiotherapy and/or chemotherapy should be administered before and/or after surgery.

The present methods may be used after initial surgery (or follow-up) to determine metastasis. This may be in addition to surgical methods for examining the extent of cancer spread described above. The methods may be used to confirm a surgical finding of metastasis or to detect metastasis in instances where this is not detectable surgically. Thus the clinician can better determine whether radiotherapy and/or chemotherapy should be administered after surgery.

The present methods may be used after surgery to detect metastasis indicative of recurrence of the colorectal cancer. Where metastasis is detected before surgery this may indicate that recurrence is more likely, and more aggressive treatment may be selected. Thus the present methods may be used to monitor for cancer recurrence after surgical removal of a primary tumour.

Use for screening asymptomatic subjects

In a further aspect, the present methods may be used to detect metastatic colorectal tumour in an asymptomatic subject.

Tumour cell invasion into the blood/lymphatic circulation is an important earlier step in cancer metastasis. Recent evidence has revealed that tumour cell invasion into the blood circulation (start of metastasis) occurs much earlier than previously thought. For example, when the blood of pancreatic cancer patients with early stage prostate cancer without any sign of detectable metastasis were analysed by a newly-developed and more sensitive microfluidic CTC-chip, invaded tumour cells were detected in the blood of 100% of the patients (Nagrath et al, Nature, 2007, 450, 1235-1239). As the increased levels of circulating Gal3 and Gal4 in the blood circulation are strongly associated with metastasis, it is believed that the combined Gal3/4 test described herein may also be used to detect earlier stage colorectal cancer in subjects with or without clinically-detectable symptoms.

Thus the present assay may be used to detect the presence of an otherwise silent or undetected colorectal tumour by detecting early metastasis of that tumour.

In this aspect of the invention, a subject typically does not display one or more CRC clinical symptoms. For example, the subject may not display any detectable clinical symptoms of CRC. In one aspect, the subject is not clinically diagnosed with CRC. In one aspect the subject does not have a CRC tumour detectable by conventional means such as colonoscopy, sigmoidoscopy or digital examination. In one aspect, the subject is not positive in a CEA or CA19-9 test. The present assay is therefore suitable for use in apparently healthy individuals, for example, as a screen for colorectal cancer. Subjects for screening may include those who are thought to be at particular risk of developing colorectal cancer e.g. elderly people.

Other cancers

The present methods may also be applicable for detection of other metastatic cancers. In one aspect, the methods may used to detect metastasis of cancers in which there are significantly elevated serum levels of Gal3 and Gal4. In one aspect, the methods may be used to detect metastasis of for example, colorectal, lung, bladder, head and neck, melanoma, pancreatic, breast or thyroid cancers. In one aspect, the cancer may be breast cancer.

The methods may be used to detect metastasis in patients already presenting one or more clinical symptoms of the cancer, or in patients who have been clinically diagnosed with the cancer using conventional means. As described above in relation to CRC, the methods may be used to monitor for cancer recurrence after surgical removal of a primary tumour.

The methods may also be used to detect a metastatic tumour in an asymptomatic subject. In this aspect of the invention, a subject typically does not display one or more clinical symptoms of the primary tumour. For example, the subject may not display any detectable clinical symptoms of the primary tumour. In one aspect, the subject is not clinically diagnosed with the cancer.

The present assay is therefore suitable for use in apparently healthy individuals, for example, as a screen for metastatic early cancerous tumours. Subjects for screening may include those who are thought to be at particular risk of developing a given cancer. Choice of treatment for cancer (for example, surgery, radiotherapy, chemotherapy or any combination thereof) is in general influenced by the stage and type of the cancer. Metastatic cancer will usually be treated more aggressively. Accurate and/or early detection of a metastatic tumour allows a more informed treatment choice to be made.

The present methods The present methods for detecting metastatic cancer, e.g. metastatic colorectal cancer, generally comprise subjecting a suitable sample (e.g. a serum sample) from the test individual to a combination assay for Gal3 and Gal4 proteins in order to determine the combined concentration of Gal3 and Gal4 protein in the sample. The combined concentration obtained in the test assay can then be compared with a predetermined threshold concentration value in order to determine the occurrence of metastasis in the individual.

In general the methods comprise assaying for Gal3 and Gal4 proteins at the same time in the same sample. In general both Gal3 and Gal4 are detected simultaneously in the sample. Typically therefore a single assay is carried out to detect both Gal3 and Gal4 in the same sample. Typically both Gal3 and Gal4 are detected in the same reaction vessel. In the assay both Gal3 and Gal4 may be detected using the same detectable label and/or detectable signal. The combination assay may comprise determining a single detectable signal. The single detectable signal may be representative of the total combined concentration of Gal3 and Gal4 in the sample. In one aspect the combination assay comprises a single signal detection step. In one aspect, neither Gal3 nor Gal4 is detected separately from the other in the assay.

The combination assay is used to determine a value for the combined concentration of both Gal3 and Gal4 in the sample. In one aspect, in the present methods neither Gal3 nor Gal4 is quantified separately from the other. Gal3 and Gal4 proteins

As used herein Galectin 3 (Gal3) protein generally refers to a protein having the amino acid sequence of human galectin-3 protein, such as that in SEQ ID NO: 1, or homolog or variant thereof. Homologs and variants are described further herein.

Human galectin-3 protein sequence is known in the art (NCBI Accession no BAA2.2.164.1). Gal3 as used herein may refer to any human galectin-3 sequence which is in the art, or to a homolog or variant thereof.

As used herein Galectin 4 (Gal4) protein generally refers to a protein having the amino acid sequence of human galectin-4 protein, such as that in SEQ ID NO: 2, or homolog or variant thereof. Homologs and variants are described further herein. Human galectin-4 protein sequence is known in the art (NCBI Accession no AAB86590.1). Gal4 as used herein may refer to any human galectin-4 sequence which is in the art, or to a homolog or variant thereof.

Samples

Typically the present assay is carried out in vitro, or ex vivo on a suitable sample. The sample is a biological sample taken from the individual to be tested. Typically this is a blood sample

Measurement may be made in whole blood. However, the blood may be further processed before an assay is performed. For instance, an anticoagulant, such as heparin, citrate, EDTA, and others may be added. Alternatively, the blood sample may be centrifuged or filtered to prepare a plasma or serum fraction for further analysis. Preferably the assay is carried out on a serum sample. Assay formats

Any suitable assay format may be used in the present methods. In one aspect an assay comprises: a) contacting a sample with a binding agent which binds specifically to Gal3 and a binding agent which binds specifically to Gal4; and b) determining the collective or combined amount of bound Gal3 and Gal4.

A binding agent as used herein generally refers to a molecule having an antigen binding site, such as an antibody antigen binding site. A binding agent may comprise an antibody or a fragment or derivative thereof, or a non-antibody molecule having an antigen binding site. Thus for example, a binding agent may comprise an anti-Gal3 antibody or an anti-Gal4 antibody such as any of those described herein. Binding agents are described further herein.

In one aspect, in the above assay step (a) may comprise use of a single binding agent, e.g. a bispecific antibody which binds specifically to both Gal3 and Gal4.

In one aspect, step (b) comprises a single detection step for detecting both bound Gal3 and bound Gal4. In one aspect, bound Gal3 and Gal4 may be detected by means of the same detectable label and/or signal. In one aspect, neither the amount of bound Gal3 nor Gal4 is determined separately from the other.

Binding agent or Gal3 and Gal4 protein may be immobilised on a solid surface. For example, in one aspect, Gal3 and Gal4 protein may be adhered to a surface by binding to an immobilised galectin- binding glycan, as described herein.

In one example, an immunoassay may be used. Immunoassay techniques are known in the art and described for example in Self CH and Cook DB, 1996, Current Opinion in Biotechnology 7: 60-65, the contents of which are incorporated herein by reference. Examples of immunoassays include immunofluorescence techniques known to the skilled technician, enzyme linked immunosorbent assay (ELISA) or radioimmunoassay analyses.

In one aspect, a competitive assay may be employed. Typically this comprises use of competitor protein(s), which competes with the target Gal3 and/or Gal4 for binding to the binding agent(s). The greater the concentration of target Gal3 and Gal4 in a sample, the less competitor bound to the binding agent(s). Generally the competitor comprises a detectable label to allow detection and measurement of bound competitor. Binding agent may also bear a label, for example, such that a signal is produced only when competitor is bound to binding agent. Binding agent may be immobilised as described herein.

In another aspect, a non-competitive assay may be employed. Typically, a test sample is contacted with binding agent(s) and complexes of "bound Gal3-binding agent" and "boundGal4-binding agent" are detected and/or quantified. Examples of such an assay include a sandwich assay, an anti- immune complex assay and an idiometric assay.

For example, a sandwich assay for a target protein typically comprises use of two binding agents One binding agent acts as a capture molecule, and may be immobilised on a solid support. The other binding agent acts as a detector molecule, and may be labelled with a detectable label.

Examples of detectable labels are given herein. In one example, a detector binding agent may comprise one member of a pair of binding molecules (e.g. the biotin member of the biotin/avidin binding pair) and can be detected by contacting with a molecule comprising the other member of the binding pair, e.g. avidin, linked to a detectable label. In one example, both capture and detector binding agents comprise labels which interact to produce a signal only when brought into proximity by binding of the detector binding agent to the already bound target protein. The capture and detector binding agents may recognise the same or different epitopes in a target protein.

In one example, a suitable sandwich assay may comprise:

i) contacting a sample with (immobilised) capture-binding agent which binds specifically to Gal3 and (immobilised) capture-binding agent which binds specifically to Gal4; ii) optionally washing with suitable buffer; iii) contacting the capture-binding agent-Gal3 and capture-binding agent-Gal4 complexes formed in (i) with detector-binding agent which binds specifically to Gal3 and detector-binding agent which binds specifically to Gal4; iv) optionally washing with suitable buffer; and v) detecting bound Gal3 -detector binding agent and bound Gal4 - detector binding agent.

In one aspect, the capture-binding agent which binds specifically to Gal3 may comprise anti-Gal3 antibody, such as antibody AF1154, a fragment or derivative thereof. In one aspect, the capture- binding agent which binds specifically to Gal4 may comprise anti-Gal4 antibody, such as AF1227, or a fragment or derivative thereof In one aspect the detector-binding agent which binds specifically to Gal3 may comprise anti-Gal3 antibody, such as AF1154, or a fragment or derivative thereof. In one aspect the detector-binding agent which binds specifically to Gal4 may comprise anti-Gal4 antibody, such as AF1227 or a fragment or derivative thereof

The capture-binding agent which binds specifically to Gal3 may comprise a different antigen binding site and bind to a different epitope in Gal3 than the detector-binding agent which binds specifically to Gal3. The capture-binding agent which binds specifically to Gal4 may comprise a different antigen binding site and bind to a different epitope in Gal4 than the detector-binding agent which binds specifically to Gal4.

Step (v) may further comprise addition of a detectably-labelled molecule which itself specifically binds to the detector binding-agent, e.g. by means of a binding pair such as biotin-avidin as described above. Thus, for example, detector binding agent, e.g. antibody, may be biotinylated, and the assay may comprise addition of a molecule comprising detectably labelled avidin, e.g. H P- Avidin.

In one aspect, the detector binding agent for Gal3 and the detector binding agent for Gal4 are detectably labelled using the same detectable label. In one aspect, step (v) comprises detection of a single detectable signal. Steps (i) and (iii) of the sandwich assay may comprise incubation, e.g. at room temperature, for a suitable period to allow binding, e.g. l-2h.

The sandwich assay may be an ELISA sandwich assay. An example of a suitable sandwich assay is described in Example 3.

In another example, an anti-immune complex assay for a target protein (Self & Cook supra) typically comprises use of a first binding agent specific for the target protein, and a second binding agent specific for the complex of "target protein + first binding agent". The first binding agent can be used as a capture binding agent, and may be immobilised. The second binding agent can be used as a detector binding agent, and is typically labelled with a detectable label. In one example, both capture and detector binding agents comprise labels which interact to produce a signal only when brought into proximity by binding of the detector binding agent to the complex. In another example, the detector binding agent may be detected using a detectably-labelled molecule which itself specifically binds to the detector binding-agent, e.g. by means of a binding pair such as biotin- avidin as described above for the sandwich assay. In another example, in an idiometric assay for a target protein (Self & Cook supra) a first binding agent is used to capture target protein. Protein-bound capture binding agent is then detected by addition of a reagent that binds to non-bound sites on the protein and prevents binding of a second binding agent.

In any of the described assays, either a "capture" or a "detector" molecule may comprise, instead of a binding agent as described herein which binds specifically to Gal3 and/or Gal4, a galectin binding molecule, e.g. a galactoside-terminated glycan. Galectin binding molecules (GBMs) are described further herein. For example, a sandwich assay such as that described above may comprise: i) contacting a sample with (immobilised) capture-"GBM " which binds Gal3 and Gal4 (as well as other galectins which may be present); ii) optionally washing with suitable buffer; iii) contacting the captured Gal3 and Gal4 from step (i) with detector-binding agent which binds specifically to Gal3 and detector-binding agent which binds specifically to Gal4; iv) optionally washing with suitable buffer; and v) detecting bound Gal3 -detector binding agent and bound Gal4 - detector binding agent. Alternatively, such an assay may comprise:

i) contacting a sample with (immobilised) capture-binding agent which binds specifically to Gal3 and (immobilised) capture-binding agent which binds specifically to Gal4; ii) optionally washing with suitable buffer; iii) contacting the capture-binding agent-Gal3 and capture-binding agent-Gal4 complexes formed in (i) with detector-"GBM " which binds to the Gal3 and Gal4; iv) optionally washing with suitable buffer; and v) detecting bound Gal3 -detector-"GBM " and bound Gal4 - detector-"GBM "

Other assay features described above in relation to capture-binding agents and detector-binding agents, e.g. labelling methods, apply in the same way to the assays using the galectin binding molecules.

Detectable labels for use in the present methods include reporter enzymes such as alkaline phosphatase, horse radish peroxidise (H P) and colorimetric or fluorometric substrates, as well as electrochemical detection methods (Self & Cook supra) and radio-isotopes.

An assay may comprise one or more "reference samples", such as a negative and/or positive control, or one or more standard samples to calibrate the assay. Examples of negative controls include a an unrelated sample known not to contain the target protein under assay, e.g. PBS or water. A positive control typically comprises a preparation of the target protein.

An assay may be calibrated using a dilution series of target protein (e.g. Gal3 and Gal4 protein). In one example, known amounts of Gal3 and Gal4, (e.g. equal amounts, such as 500ng/ml of each) may be combined, and serially diluted to form a dilution series. Preferably these are prepared in the same biological matrix as the anticipated study samples, e.g. human sera. The dilution series is assayed under the same conditions (e.g. component concentrations, incubation times,

temperatures) as the test assay, and the signals obtained may be used to plot a standard curve.

The signal obtained in a test assay can be compared with the signals in the calibration assays (or standard curve) to determine the total combined concentration of Gal3 and Gal4 in the test sample. Typically a test assay is repeated at least 2, 3, or 4 times, on repeat samples. Assay results may then be averaged.

Binding agents, e.g. antibodies, may be used in the present assays, e.g sandwich assay, at any appropriate concentrations. The skilled person will be aware of how to determine suitable concentrations. For example, binding agent, e.g. capture binding agent, may be used at a concentration of 0.1 to 5μg/ml, such as 0.1 to 4, 0.5 to 3, or 0.5^g/ml. For example, capture binding agent may be used at a concentration of 0.1, 0.5, 0.6, 0.8, 1.0, 1.1, 1.15, 1.2, 1.25, 1.3, 1.35, 1.4, 1.5, 2, 2.5 or 3μg/ml. In one aspect, e.g. in a sandwich assay, the concentration of capture binding agent may be <5^g/ml. In one example, binding agent, e.g. detector binding agent, may be used at a concentration of 0.1 to 5μg/ml, such as 0.3 to 5, 0.4 to 4, 0.5 to 3 or 0.5^g/ml. For example, detector binding agent may be used at a concentration of 0.1, 0.5, 0.6, 0.8, 1.0, 1.1, 1.15, 1.2, 1.25, 1.3, 1.35, 1.4, 1.5, 2, 2.5 or 3μg/ml. In one aspect, e.g. in a sandwich assay, the concentration of detector binding agent may be > O^g/ml.

Sensitivity of an assay may be improved by optimising parameters such as the concentration of binding agents used. In one aspect, dilution series of binding agents may also be prepared and used in assays with the dilution series of combined Gal3/Gal4 protein. This enables selection of optimal concentrations of binding agent for use in the assay.

An assay may be homogeneous or heterogenerous. A heterogeneous assay typically comprises a step of phase separation of bound and unbound analyte before signal detection, and may comprise a washing step. An assay may be designed to be useful at "point of need", for example utilising a dipstick or detector strip.

Binding agents

A binding agent for use in the present CG34 assays binds specifically to Gal3, or to Gal4, or to Gal3 and Gal4.

A "binding agent" for use herein generally refers to a molecule having an antigen binding site, such as an antibody antigen binding site. A binding agent may comprise an antibody or a fragment or derivative thereof, or a non-antibody molecule having an antigen binding site. A binding agent may be natural or wholly or partially synthetic, as described herein.

A binding agent for use in the present CG34 assays may have binding specificity for Gal3 protein, for Gal4 protein or for both Gal3 and Gal4 protein, e.g a bispecific antibody. By this is meant that the binding agent will bind specifically to the given protein(s) but will not substantially bind to other proteins, in particular to other human blood or serum proteins which may be present in a test sample, e.g. other galectins. Thus an agent has the ability to differentially bind to Gal3 and/or Gal4 compared to other human blood/serum proteins, and to distinguish between them. Put in another way, a binding agent has a greater relative affinity for Gal3 and/or Gal4 compared to the other proteins, such that the agent can discriminate between the proteins in a clinical assay. In one aspect, a binding agent shows no or minimal cross-reactivity with other blood/serum proteins in the assay format which is to be used, e.g an ELISA assay. In one aspect, a binding agent shows no or minimal cross-reactivity with other blood/serum proteins in a Western Blot. In one aspect, the degree of cross-reactivity is sufficiently low for use in the present assays.

A binding agent for use herein may comprise an antibody.

The basic structural unit of an antibody is generally a tetramer that consists of two identical pairs of immunoglobulin (Ig) chains, each pair having one light and one heavy chain.

Each chain contains distinct sequence domains. The light chain includes two domains, a variable domain (VL) and a constant domain (CL). The heavy chain includes four domains, a variable domain (VH) and three constant domains (CHI, CH2 and CH3, collectively referred to as CH). The variable regions of both light (VL) and heavy (VH) chains determine binding recognition and specificity to the antigen. The constant region domains of the light (CL) and heavy (CH) chains mediate effector functions. Thus these domains confer important biological properties such as antibody chain association, secretion, transplacental mobility, complement binding, and binding to Fc receptors.

The variable domains or regions are the regions of both the light chain and the heavy chain of an Ig that contain antigen-binding sites. A variable region is composed of polypeptide chains containing four relatively invariant "framework regions" (F s) and three highly variant "hypervariable regions" (HVs). Because the HVs constitute the binding site for antigen(s) and determine specificity by forming a surface complementary to the structure of the bound antigen, they are more commonly termed the "complementarity-determining regions," or CDRs, and, proceeding from the N-terminus of a heavy or light chain, are denoted CDR1, CDR2, and CDR3 (see, Sequences of Proteins of Immunological Interest, E. Kabat et al., U.S. Department of Health and Human Services, 1983). A VH domain comprises a set of HCD s, and a VL domain comprises a set of LCDRs. Thus, VH comprises HCDR1, HCDR2 and HCDR3, and VL comprises LCDR1, LCDR2 and LCDR3.

The CDRs are primarily responsible for binding to an epitope of an antigen and the CDR3 comprises a unique region specific for antigen-antibody binding. An antigen-binding site, therefore, may include six CDRs, comprising the CDR regions from each of a heavy and a light chain V region.

A binding agent may comprise a monoclonal antibody or a fragment thereof. Typically a monoclonal antibody refers to an antibody produced by a single clone of cells, e.g. B-lymphocytes or by a cell into which the light and heavy chain genes of a single antibody have been transfected. Monoclonal antibodies are produced by methods known to those of skill in the art.

A binding agent may comprise a polyclonal antibody

Immunoglobulins also exist in a variety of other forms including, for example, Fv, Fab, and (Fab')₂, as well as bifunctional hybrid antibodies and single chains (e.g., Lanzavecchia et al., Eur. J. Immunol. 17:105, 1987; Huston et al., Proc. Natl. Acad. Sci. U.S.A., 85:5879-5883, 1988; Bird et al., Science 242:423-426, 1988; Hood et al., Immunology, Benjamin, N.Y., 2nd ed., 1984; Hunkapiller and Hood, Nature 323:15-16, 1986).

It has been shown that such fragments of a whole antibody can perform the function of binding antigens. In one aspect, a binding agent for use in the present assays may comprise an antibody fragment or derivative thereof, which comprises an antibody antigen binding site. Examples of antigen binding fragments may include:

(i) the fragment antibody (Fab) fragment consisting of VL, VH, CL and CHI domains;

(ii) the Fd fragment consisting of the VH and CHI domains;

(iii) the Fv fragment consisting of the VL and VH domains of a single antibody;

(iv) the dAb fragment, a small monomeric antigen-binding fragment of an antibody, consisting of the VH or VL domain (Ward, E. S. Et al., Nature 341,544-546 1989; Holt et al, Trends in Biotechnology 21,

484-490, 2003);

(v) isolated CDR regions;

(vi) F(ab')₂ fragments, a bivalent fragment comprising two linked Fab fragments;

(vii) single chain Fv molecules (scFv), wherein a VH domain and a VL domain are linked by a peptide linker which allows the two domains to associate to form an antigen binding site (Bird et al, Science,

242, 423-426,1988 ; Huston et al, PNAS USA, 85, 5879-5883,1988) ;

(viii) dsFV, wherein the VL and VH are chain linked by disulfide bonds;

(ix) bispecific single chain Fv dimers (PCT/US92/09965); and (x)"diabodies", multivalent or multispecific fragments constructed by gene fusion (W094/13804; P. Holliger et al, Proc. Natl. Acad. Sci. USA 90 6444-6448,1993).

Other examples of binding fragments are Fab', which differs from Fab fragments by the addition of a few residues at the carboxyl terminus of the heavy chain CHI domain, including one or more cysteines from the antibody hinge region, and Fab'-SH, which is a Fab' fragment in which the cysteine residue(s) of the constant domains bear a free thiol group. (Fab')₂ fragments are the fragments of the antibody obtained by treating whole antibody with the enzyme pepsin without subsequent reduction, or the corresponding structure obtained by genetic engineering. Minibodies comprising an scFv joined to a CH3 domain may also be made (S. Hu et al, Cancer Res., 56, 3055- 3061,1996).

In one aspect, a binding agent may comprise a bispecific (or bifunctional) antibody, in which two different variable regions are combined in the same molecule.

Bispecific antibodies may be conventional bispecific antibodies, which can be manufactured in a variety of ways (Holliger, P. and Winter G. Current Opinion Biotechnol. 4,446-449 (1993)), e.g. prepared chemically or from hybrid hybridomas. Bispecific antibodies may also be any of the bispecific antibody fragments mentioned above. Thus, bispecific antibodies can be constructed as entire IgG, as bispecific Fab'2, as Fab 'PEG, as diabodies or as bispecific scFv.

A binding agent for use herein may comprise a modified antibody or derivative, such as a chimeric antibody. For example, a binding agent may be one in which an antigen binding site of a given antibody (e.g. the VH and/or VL domain, or the CDRs, of an antibody) is fused to another polypeptide (e.g. the constant regions or constant regions plus framework regions, of a different antibody). This can be done by fusing the encoding DNAs.

Chimeric antibodies may be antibodies whose light and heavy chain genes comprise variable and constant regions encoded by variable and constant region genes of different species. Typically such chimeric antibodies are produced by genetic engineering. For example, the variable segments of the genes from a mouse monoclonal antibody can be joined to human constant segments, such as kappa and gamma 1 or gamma 3.

A binding agent may comprise a "humanized antibody". Typically this is an antibody comprising a humanized light chain and a humanized heavy chain immunoglobulin. Humanizing is a technique in which one or more, e.g. all, CDRs, from a non-human "donor" (such as a mouse, rat, or synthetic) antibody is fused to a human framework region. The human antibody providing the framework is termed an "acceptor." Constant regions need not be present, but if they are, they are generally substantially identical to human immunoglobulin constant regions. A binding agent may comprise a human antibody, wherein the light and heavy chain genes are of human origin.

An antigen binding site refers to the part of the binding agent that binds to and is complementary to all or a part of the target antigen (e.g. Gal3 or Gal4). An antibody may only bind to a particular part of an antigen, which is termed an epitope. An antibody antigen binding site may be provided by one or more antibody-variable domains. An antibody antigen binding site may comprise a VH and/or a VL domain as described. An antigen binding site may comprise one or more loop structures which may be analogous to one or more CDRs. An antigen binding site may comprise one or more CDRs, such as at least 1, 2, 3, 4, 5, or 6 CDRs. An antigen binding site may comprise for example, a HCDR3 and/or LCDR3. An antigen binding site may comprise a set of CDRs corresponding to the CDRs in a VL or VH e.g. the set of CDRs which is (HCDRl + HCDR2 + HCDR3) and/or the set of CDRs which is (LCDR1+ LCDR2 + LCDR3).

An epitope refers to a particular site on an antigen which is recognized by an antigen binding site. An epitope may be defined with reference to a particular amino acid sequence in a peptide antigen. Two binding agents, e.g. antibodies are said to bind to the same epitope if each competitively inhibits (blocks) binding of the other to the antigen as measured in a competitive binding assay (see, e.g., Junghans et al., Cancer Res. 50:1495-1502, 1990).

In one aspect, a binding agent for use in the present methods may comprise an antigen binding site in a non-antibody scaffold. An antigen binding site may be formed by positioning one or more CDRs on a non-antibody scaffold, e.g. by grafting in one or more CDRs. An antigen binding site may be formed by rational or random mutation of amino acids in the molecule, often of amino acids in one or more loop structures, or of surface residues, in the scaffold molecule which is/are involved in binding, to provide a particular binding specificity. Such loop structures can be analogous to the CDRs of an antibody.

For example, protein display scaffolds are reviewed in Hosse, R.J. et al, 2006, Protein Science, 15: 14- 27. These include:

(a) scaffolds with a-helical frameworks, such as Affibodies, Immunity proteins (e.g the E coli colicin E7 immunity protein lmmE7), Cytochrome b₅₆₂, peptide a₂p8, repeat proteins;

(b) scaffolds with few or irregular secondary structures, such as insect defensin A, Kunitz domain inhibitors, PDZ domain proteins (e.g. Ras-binding protein AF-6), scorpion toxins (e.g. Charybdotoxin), the plant homeodomain (PHD) finger protein from the transcriptional cofactor Μϊ2β, TEM-1 β- lactamase; and

(c) scaffolds with β-sheet frameworks, such as 10^th fibronectin type III domain (¹⁰Fn3), human cytotoxic lymphocyte associated protein 4 (CTLA-4) (which comprises CDR-like loops similar to antibodies), T-cell receptors, Knottins, Neocarzinostatin (the neocarzinostatin protein component (NCS) has two loops, structurally equivalent to CDR1 and CDR3 of an antibody), carbohydrate binding module 4-2 (CBM4-2, derived from the Rhodothermus marinus xylanase XynlOA)), Tendamistat (an inhibitor of a-amylase from Streptomyces tendae), Lipocalins, or green fluorescent protein (GFP). An example of repeat motif proteins are those with ankyrin repeat domains such as DARPins or leucine rich-repeat proteins (Hosse et al, supra).

Kunitz domain inhibitors include: bovine pancreatic trypsin inhibitor (BPTI); human pancreatic secretory trypsin inhibitor (PSTI); Alzheimers amyloid β-protein precursor inhibitor (APPI); the leech derived trypsin inhibitor (LTDI); the mustard trypsin inhibitor II (MTI II); the periplasmic E coli protease inhibitor ecotin; and the human lipoprotein associated coagulation inhibitor (LACI) .

Knottins include the trypsin inhibitor from the squirting cucumber Ecballium elaterium (EETI-II), the C-terminal cellulose binding domain (CBD) of cellobiohydrolase I from the fungus Trichoderma reesei, and Min-23, a derivative of EETI-II.

Lipocalin proteins may be used as scaffolds. The engineered versions are generally referred to as anticalins, reviewed in Skerra, A., 2008, FEBS Journal 275: 2677-2683. Examples of liopcalin scaffolds include the bilin-binding protein (BBP) from the butterfly Pieris brassicae, human apolipoprotein D (ApoD), and the bovine heart fatty acid-binding protein (FABP). A suitable scaffold into which to graft one or more CDRs may be provided by any domain member of the immunoglobulin gene superfamily.

A scaffold may be a human or non- human protein. In one aspect the scaffold is a human scaffold.

A binding agent may in some instances comprise a detectable label. Binding agents may be immobilised on a solid surface. For example, binding agent may be adsorbed onto or covalently linked to a microtitre plate, dipstick, or other surface, such as beads or particles.

Known binding agents

Any suitable antibody may be used in the present assays.

Antibodies which bind specifically to Gal3 or Gal4 are known in the art and may be used as binding agents herein. For example suitable antibodies are available from R & D Systems (Abingdon, UK), such as antibodies AF1154 and AF1227.

In one aspect, a binding agent, e.g. antibody, which binds specifically to Gal3 may comprise antibody AF1154 (R & D Systems, Abingdon UK) or a derivative or fragment thereof, as described herein. AF1154 comprises human galectin-3 specific IgG (polyclonal), produced in goats immunised with purified E.coli-derived recombinant human Gal3, and purified by human Gal3 affinity chromatography. AF1154 is reported to show approximately 50% cross-reactivity with rmGal3 and less and than 1% cross-reactivity with rhGalectin -1, -2, -4, -7 and -8 in direct ELISAs and western blots.

A binding agent may, for example, comprise the antigen binding site of AF1154 which specifically recognises Gal3. A binding agent may retain the Gal3 binding specificity of AF1154. For example, in one aspect, a binding agent which binds specifically to Gal3 may show less than 1% cross-reactivity with rhGalectin -1, -2, -4, -7 and -8 in direct ELISAs and western blots.

In one aspect, a binding agent, e.g. antibody, which binds specifically to Gal4 may comprise antibody AF1227 (R& D Systems, Abingdon, UK) or a derivative or fragment thereof, as described herein. AF1227 comprises human galectin-4 specific IgG (polyclonal), produced in goats immunised with purified E.coli-derived recombinant human Gal4 (rhGalectin-4), and purified by human Gal4 affinity chromatography. AF1227 is reported to show approximately 25% cross-reactivity with rmGal4, 5% cross-reactivity with rhGalectin -2, and less than 1% cross-reactivity with rhGalectin -1, -3, -7 and -8 in direct ELISAs.

A binding agent may, for example, comprise the antigen binding site of AF1227 which recognises Gal4. A binding agent may retain the Gal4 binding specificity of AF1154.

For example, in one aspect, a binding agent which binds specifically to Gal4 may show 5% or less cross-reactivity with rhGalectin -2, and less than 1% cross-reactivity with rhGalectin -1, -3, -7 and -8 in direct ELISAs.

Galectin-binding molecules (GBMs)

In general, a galectin binding molecule (GBM) as referred to herein comprises a galactoside- terminated glycan. Typically the molecule binds to one or more galectins including Gal3 and/or Gal4. Examples of galactoside terminated glycans include asialo fetuin (Nilsson et al J Biol Chem. 1979 Jun 10;254(ll):4545-53), asialo bovine mucin (Savage et al. Eur J Biochem. 1990 Sep ll;192(2):427-32) and Antarctic fish antifreeze glycoproteins (Campbell et al, I Clin Inv, 2005, 95, 571-76). These are all Thomsen-Friedenreich (TF) (Galbetal,3GalNAc-) expressing glycoproteins (galactoside-terminated glycans) and are all known to be recognized by Gal-3 and -4.

Threshold values

In the present methods, once a value has been obtained for the combined concentration of Gal3 and Gal4 protein in the sample taken from the test subject, this is typically compared with a

predetermined threshold value. If the amount of protein in the sample is higher than the threshold, this is indicative of metastasis in the subject.

A suitable threshold value is generally one which provides an optimal combination of sensitivity and specificity for clinical use. An optimal threshold value can be determined by calculating the sensitivity and specificity of the assay when applied to an appropriate sample set, using a variety of test threshold concentrations.

Typically the sample set which is used to derive the threshold value is derived from a population similar to that in which the test assay is to be used.

Thus, for example, where the assay is intended for use in cancer patients to detect metastasis, a suitable sample set may comprise samples from metastatic cancer patients (expected "positives" in the assay) and non-metastatic cancer patients (expected "negatives" in the assay). "Metastatic" and "non-metastatic" is generally as diagnosed clinically using conventional means.

Sufficient number of samples are included in the set to provide statistical significance.

Sensitivity and specificity at any test threshold value may be calculated using the following equations:

Numbers of True Positives

Sensitivity =

Numbers of True Positives + Numbers of False Negatives Numbers of True Negatives

Specificity =

Numbers of True Negatives + Numbers of False Positives where, in an assay intended for use in cancer patients to detect metastasis :

"True positives" = the number of samples in the metastatic cancer patient group having a Gal3/Gal4 combined concentration above the threshold.

"False positives" = the number of samples in the non-metastatic cancer patient group having a Gal3/Gal4 combined concentration above the threshold.

"True negatives" = the number of samples in the non-metastatic cancer patient group having a Gal3/Gal4 combined concentration below the threshold.

"False negatives" = the number of samples in the metastatic cancer patient group having a Gal3/Gal4 combined concentration below the threshold.

For an assay intended to detect metastatic cancer in as yet symptomatic patients, a sample set may in practice comprise samples from patients clinically diagnosed as having only a primary tumour (expected "positives" in the assay) and healthy subjects (expected "negatives" in the assay). In the above equation in that case: "True positives'^ the number of samples in the primary tumour patient group having a Gal3/Gal4 combined concentration above the threshold.

"False positives" = the number of samples in the healthy subject group having a Gal3/Gal4 combined concentration above the threshold.

"True negatives" = the number of samples in the healthy subject group having a Gal3/Gal4 combined concentration below the threshold.

"False negatives" = the number of samples in the primary tumour patient group having a Gal3/Gal4 combined concentration below the threshold. It is recognised that this is an approximation. "Healthy" subjects may in fact include those with early metastatic but silent tumours, and "primary tumour" patients may in fact be non-metastatic. However, given the rate of incidence of, e.g. colorectal cancer, and the size of cohort and time necessary to implement a full screen and follow-up of healthy individuals, the inventors believe that the above sample set provides a practical means of determining threshold values, sensitivities, and specificities.

In one aspect, an assay preferably has a % sensitivity of at least 40%, such as at least 40%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 97%, or 99%. Alternatively or additionally, an assay may preferably have a % specificity of 40-100%, such as at least 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 97%, or 99% as appropriate

In one aspect, an assay for use in C C patients to detect metastasis is preferably able to separate CRC (liver)metastatic patients from CRC non-metastatic patients at a significance of at least p<0.05, such as at least p<0.04, p<0.03, p<0.02, or p<0.01, as calculated using the Kruskall-Wallis non- parametric ANOVA test.

Examples of test threshold values and of the corresponding sensitivities and specificities obtained in a CG34 assay are given in Table 2 (Figure 4). In one aspect, an assay may make use of any of these threshold values. In one aspect, an assay for determining metastasis in a C C patient may make use, for example, of a threshold of 5-10ng/ml, such as 6, 7, 8 or 9ng/ml. In one aspect an assay has any of the threshold/sensitivity/specificity combinations shown in the Table.

Use of the CG34 assay with other tests/techniques

In one aspect, the results obtained in the present assays provide an indication of metastatic cancer in an individual. Other factors may also be taken into account.

The present CG34 assay may be used alone for detection of metastatic cancer. Alternatively, the CG34 assay may be used in combination with one or more other tests or techniques for detecting cancer metastasis. In one example, the CG34 assay may be used in combination with a test based on assay of the CEA or CA19-9 antigen referred to herein. The one or more other tests may be carried out before or after or simultaneously with the CG34 assay. The one or more other tests may be carried out on the same or a different sample (taken from the subject) to that used for the CG34 assay. In this aspect, a positive result in both the CG34 assay and at least one other test provides an indication of metastatic cancer in the subject.

The CG34 assay may thus be used in combination with one or more assays for one or more other indicators of metastasis. An indicator of metastasis generally comprises a molecule, the

concentration of which in a sample from a subject can be used to determine the presence of metastasis in the subject.

Certain cytokines are associated with metastasis. The inventors have shown a significant correlation between circulating Gal3 levels and three such cytokines, IL6, G-CSF and slCAM-1 (Table 3, Figure 7). Thus, for example, the CG34 assay may be used in combination with one or more assays for one or more cytokine indicators of metastasis, such as interleukin 6 (IL6), soluble intercellular adhesion molecule-1 (s-ICAMl), and granulocyte colony-stimulating factor (G-SCF). Any one or more, such as two or three of these cytokines may be assayed in a test to indicate cancer metastasis.

As used herein IL6 protein generally refers to a protein having the amino acid sequence of human IL6 protein, such as that in SEQ ID NO: 3, or homolog or variant thereof. Homologs and variants are described further herein. Human IL6 protein sequence is known in the art (NCBI Accession no AAH15511). IL6 as used herein may refer to any human IL6 sequence which is in the art, or to a homolog or variant thereof.

As used herein G-CSF protein generally refers to a protein having the amino acid sequence of human G-CSF protein, such as that in SEQ ID NO: 4, or homolog or variant thereof. Homologs and variants are described further herein. Human G-CSF protein sequence is known in the art (NCBI Accession no P09919). G-CSF as used herein may refer to any human G-CSF sequence which is in the art, or to a homolog or variant thereof. As used herein s-ICAMl protein generally refers to a protein having the amino acid sequence of human s-ICAMl protein, such as that in SEQ ID NO: 5, or homolog or variant thereof. Homologs and variants are described further herein. Human s-ICAMl protein sequence is known in the art (NCBI Accession no CAA41977), s-ICAMl as used herein may refer to any human s-ICAMl sequence which is in the art, or to a homolog or variant thereof.

Indicators may be assayed in a sample taken from a test subject, such as any of the samples referred to herein, e.g. a serum sample.

Any suitable assay format may be used. For example, the assay formats described herein in relation to the CG34 assay may be applied for assaying a target indicator molecule, e.g. IL6, s-ICAMl or G- CSF. For example, an assay may comprise an immunoassay, such as ELISA.

An assay may comprise contacting a sample with a binding agent which binds specifically to the indicator molecule of interest, and determining the amount of bound indicator. The concentration of indicator in the sample may then be compared with a predetermined threshold value. An amount of indicator higher or lower than the threshold (depending on the indicator molecule) may indicate metastasis.

Methods for assaying the cytokines IL6, G-CSF and slCAM-1 are known in the art. For example, assay kits are available from & D Systems (Abingdon, UK) as described herein.

Products for use in the present methods

In a further aspect, the present invention provides products, including kits, for use in the present methods. Typically such a product or kit comprises means for simultaneously assaying serum levels of Gal3 and Gal4 as described herein.

In one example a kit comprises one or more binding agents, e.g. antibodies, which bind specifically to Gal3 and one or more binding agents, e.g. antibodies, which bind specifically to Gal4, as described herein. For example, a kit may comprise: a) a binding agent which binds specifically to Gal3; and b) a binding agent(s) which binds specifically to Gal4.

Binding agent (a) and/or (b) may be suitable for use as a capture binding agent as described herein. Binding agent (a) and/or (b) may be suitable for use as a detector binding agent as described herein. Binding agent (a) and (b) may comprise a single binding agent, e.g a bispecific antibody.

In one example a kit may comprise: a) a capture binding agent which binds specifically to Gal3 and/or a detector binding agent which binds specifically to Gal3; and

b) a capture binding agent which binds specifically to Gal4 and/or a detector binding agent which binds specifically to Gal4; where capture and detector binding agents are as described herein in relation to assay methods.

A kit may comprise a galectin-binding molecule (GBM) as described herein, for use as a capture or detector molecule as described herein.

Any one or more of the binding agents in a kit may be detectably labelled. Any one or more of the binding agents in a kit may be immobilised, e.g on a microtitre plate, dipstick, or other surface.

A kit may comprise means for detecting binding agent. For example, where a binding agent or GBM in the kit comprises one member of a pair of binding molecules, e.g. the biotin member of a biotin- avidin binding pair, a kit may comprise detectably labelled avidin, e.g. H P-Avidin. A kit may comprise a developing solution for a detectable signal.

A kit may comprise one or more other components for use in an assay, e.g. buffer, and/or one or more galectin standards for calibration. A kit may comprise instructions for use of the kit components in an assay as described herein.

A kit may comprise means for quantifying assay results, e.g. calibration curves or other calibration data in paper or electronic form. A kit may comprise data concerning threshold values for an assay as described herein, and/or means for comparing assay results with such threshold values, in paper or electronic form.

In one aspect a kit is suitable for use in a sandwich assay, e.g. ELISA sandwich assay, as described herein.

A product or kit may comprise means for assaying one or more other indicators of metastasis in a suitable sample as described herein. For example, a kit may comprise means for assaying one or more cytokine indicators such as one, two or all three of IL6, G-CSF and slCAM-1 in a sample, e.g. a serum sample.

For example, a kit may comprise one or more binding agents, e.g. antibodies, which bind specifically to an indicator molecule. A binding agent may be detectably labelled or immobilised as described herein. A kit may comprise means for detecting such a binding agent, as described herein.

A kit may comprise one or more other components for use in an indicator assay, e.g. buffer, and/or one or more indicator standards for calibration. A kit may comprise instructions for use of the kit components in an assay as described herein.

A kit may comprise means for quantifying indicator assay results, e.g. calibration curves or other calibration data in paper or electronic form. A kit may comprise data concerning threshold values for an indicator assay, and/or means for comparing assay results with such threshold values, in paper or electronic form.

In one aspect a kit is suitable for use in a sandwich assay, e.g. ELISA sandwich assay, as described herein. Sequence homologs and variants

As used herein a homolog or variant of a protein or nucleic acid sequence (e.g. a gene) refers to a protein or nucleic acid sequence that is similar in sequence and in function to the reference sequence. A species homolog refers to a similar sequence (e.g. gene and/or protein) occurring in a different species to the reference sequence.

For any nucleotide or amino acid sequence, homologous sequences may be identified by searching appropriate databases. For example, suitable databases include GenBank (available at

www.ncbi.nlm.nih.gov/Genbank) or UniProt (available at http://www.ebi.ac.uk/uniprot.

Where appropriate, databases can be searched for homologous sequences using computer programs employing various algorithms. Examples of such programs include, among others, FASTA or BLASTN for nucleotide sequences and FASTA, BLASTP, gapped BLAST, and PSI-BLAST for amino acid sequences. FASTA is described in Pearson, W and Lipman, D J, Proc. Natl., Acad. Sci, USA, 85, 2444 2448, 1988. BLASTP, gapped BLAST, and PSI-BLAST are described in Altschul, S F, et al., Basic local alignment search tool, J. Mol. Biol., 215(3): 403 410, 1990, Altchul, S F and Gish, W, Methods in Enzymology, 266, 460 480, 1996, and Altschul, S F, et al., "Gapped BLAST and PSI-BLAST: a new generation of protein database search programs", Nucleic Acids Res. 25:3389 3402, 1997.

In addition to identifying homologous sequences, programs such as those mentioned above typically provide an indication of the degree of homology (or identity) between sequences. Determining the degree of identity or homology that exists between two or more amino acid sequences or between two or more nucleotide sequences can also be conveniently performed using any of a variety of other algorithms and computer programs known in the art. Discussion and sources of appropriate programs may be found, for example, in Baxevanis, A., and Ouellette, B. F. F., Bioinformatics: A Practical Guide to the Analysis of Genes and Proteins, Wiley, 1998; and Misener, S. and Krawetz, S. (eds.), Bioinformatics Methods and Protocols (Methods in Molecular Biology, Vol. 132), Humana Press, 1999.

Calculations of sequence homology or identity (the terms are used interchangeably herein) between sequences may be performed as follows.

To determine the percent identity of two amino acid sequences, or of two nucleic acid sequences, the sequences are aligned for optimal comparison purposes (e.g., gaps can be introduced in one or both of a first and a second amino acid or nucleic acid sequence for optimal alignment and nonhomologous sequences can be disregarded for comparison purposes). In a preferred embodiment, the length of a reference sequence aligned for comparison purposes is at least 30%, preferably at least 40%, more preferably at least 50%, even more preferably at least 60%, and even more preferably at least 70%, 75%, 80%, 82%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, or 100% of the length of the reference sequence. The amino acid residues or nucleotides at corresponding amino acid positions or nucleotide positions are then compared. When a position in the first sequence is occupied by the same amino acid residue or nucleotide as the corresponding position in the second sequence, then the molecules are identical at that position (as used herein amino acid or nucleic acid "identity" is equivalent to amino acid or nucleic acid "homology"). The percent identity between the two sequences is a function of the number of identical positions shared by the sequences, taking into account the number of gaps, and the length of each gap, which need to be introduced for optimal alignment of the two sequences.

The comparison of sequences and determination of percent identity between two sequences can be accomplished using a mathematical algorithm. In one embodiment, the percent identity between two amino acid sequences is determined using the Needleman et al. (1970) J. Mol. Biol. 48:444-453) algorithm which has been incorporated into the GAP program in the GCG software package (available at http://www.gcg.com), using either a BLOSUM 62 matrix or a PAM250 matrix, and a gap weight of 16, 14, 12, 10, 8, 6, or 4 and a length weight of 1, 2, 3, 4, 5, or 6. In one embodiment, the percent identity between two nucleotide sequences is determined using the GAP program in the GCG software package (available at http://www.gcg.com), using a NWSgapdna.CMP matrix and a gap weight of 40, 50, 60, 70, or 80 and a length weight of 1, 2, 3, 4, 5, or 6. A particularly preferred set of parameters (and the one that should be used if the practitioner is uncertain about what parameters should be applied to determine if a molecule is within a sequence identity or homology limitation of the invention) are a BLOSUM 62 scoring matrix with a gap penalty of 12, a gap extend penalty of 4, and a frameshift gap penalty of 5.

The percent identity between two amino acid or nucleotide sequences can be determined using the algorithm of Meyers et al. (1989) CABIOS 4:11-17) which has been incorporated into the ALIGN program (version 2.0), using a PAM120 weight residue table, a gap length penalty of 12 and a gap penalty of 4.

As used herein, a homologous or variant amino acid sequence generally has at least 50%, 60%, 70%, 75%, 80%, 81%. 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89% 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% or more identity with the reference sequence. Thus, for example, a homolog or variant of the Gal3 protein having the amino acid sequence in SEQ ID NO: 1 generally has at least 50%, 60%, 70%, 75%, 80%, 81%. 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89% 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% or more identity with the SEQ ID NO:l sequence. Thus, for example, a homolog or variant of the Gal4 protein having the amino acid sequence in SEQ ID NO: 2 generally has at least 50%, 60%, 70%, 75%, 80%, 81%. 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89% 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% or more identity with the SEQ ID NO:2 sequence. Variants include insertions, deletions, and substitutions, either conservative or non-conservative.

In terms of amino acids, small non-polar, hydrophobic amino acids include glycine, alanine, leucine, isoleucine, valine, proline, and methionine. Large non-polar, hydrophobic amino acids include phenylalanine, tryptophan and tyrosine. The polar neutral amino acids include serine, threonine, cysteine, asparagine and glutamine. The positively charged (basic) amino acids include lysine, arginine and histidine. The negatively charged (acidic) amino acids include aspartic acid and glutamic acid. Therefore by "conservative substitutions" is intended to include combinations such as Gly, Ala; Val, lie, Leu; Asp, Glu; Asn, Gin; Ser, Thr; Lys, Arg; and Phe, Tyr.

Due to the degeneracy of the genetic code, it is clear that any nucleic acid sequence could be varied or changed without substantially affecting the sequence of the protein encoded thereby, to provide a functional variant thereof. A functional variant is one in which the changes made with respect to the reference sequence do not substantially alter protein activity. Preferably as used herein (and unless otherwise specified), Gal3 or Gal4 homologs and variants refer to functional homologs and variants. Preferably as used herein (and unless otherwise specified), homologs and variants of specific cytokines IL6, G-CSF or slCAM-1 refer to functional homologs and variants.

Examples

The following Examples illustrate the invention:

Materials and Methods

Materials Recombinant human galectin-1, -2, -3, -4 and -8, and all the anti-galectin antibodies and biotinylated anti-galectin antibodies were obtained from R & D systems (Abingdon, UK). Anti-Gal3 antibodies were AF1154, and anti-Gal4 antibodies were AF1227 as described herein. Biotinylated forms of the antibodies were used for detection as described below.

Recombinant human IL-6, G-CSF, GM-CSF, ICAM-1 and human cytokine ELISA kits for assay of G-CSF, GM-CSF, IL-6 and ICAM-1, were obtained from R&D Systems (Abingdon, UK).

Serum samples

Example 1-4

Fifty one serum samples from colorectal cancer patients, 40 without clinically detectable metastasis (26 males and 14 females) and 11 with liver metastasis (7 males and 4 females) and 40 serum samples from breast cancer patients (all females) were obtained from CTBRC cancer tissue bank (Liverpool, UK). These serum samples were obtained by CTBRC from patients at time when their primary tumours were removed by surgery at the Royal Liverpool Hospital. The length of survival of these patients in the next 10 years was followed up. Thirty-one serum samples from healthy people (13 male and 19 female) were obtained from Sera Laboratories International (Haywards Heath, UK). The patients' ages were from 25 to 91 (mean age 63) and healthy people from 20 to 51 (mean age 37). Although the healthy controls were younger than the patients, there was no association between age and concentration of any of the galectins (Spearman's rho correlation coefficient: -0.02 and -0.034, with p=0.92 and 0.86 for galectin-13 and -4, respectively).

Example 5 Fifty serum samples from colorectal cancer patients (age 25 to 91), 39 without clinically detectable metastasis (25 males and 14 females) and 11 with liver metastasis (7 males and 4 females) were obtained from CTBRC cancer tissue bank (Liverpool, UK). These serum samples were obtained from patients at the time of primary tumor resection at the Royal Liverpool University Hospital.

Individual assessments of serum galectin-3 and -4 levels

High-binding 96-well plates were coated 2^g/ml anti-galectin-3 or anti-galectin-4 antibody in coating buffer (Na₂C0₃ 1.6g, NaHC0₃ 1.46g in 1L H₂0) overnight at 4°C. The plate was washed twice with washing buffer (0.05% Tween 20 in phosphate buffered saline (PBS)) and incubated with blocking buffer (1% bovine serum albumin (BSA) in PBS) for lhr at room temperature. The wells were washed once with washing buffer.

Serum samples were diluted 1:2 with PBS and applied to the coated wells for 2hrs at room temperature. 500ng/ml of recombinant galectin-3 or -4 were serially diluted and applied to some coated wells also for 2hrs at room temperature, for generation of the standard curve. The wells were washed twice with washing buffer.

Biotinylated anti-galectin-3 or -4 antibody (1.25μg/ml diluted in blocking buffer) were applied to each well for lhr at room temperature. The wells were washed with washing buffer and applied with H P-ExtrAvidin (1:10,000 in blocking buffer) for lhr. After two washes with washing buffer, the wells were applied with SigmaFAST OPD developing solution for approximately lOmins until a yellow colour developed in the standard recombinant galectin wells. The reaction was stopped with 4M sulphuric acid and read at 492nm by a microplate reader.

The serum concentrations of galectin-3 or -4 were calculated from the standard curves derived from the recombinant galectin-3 or -4 assays run in parallel. The results were analysed using Kruskall- Wallis non-parametric ANOVA test.

Individual assessments of serum levels of galectins -1, -2 and -8 Serum levels of these galectins were assessed in the same way as described for galectins -3 and -4, using antibodies appropriate to the particular galectin being assayed.

Simultaneous assessment of serum galectin-3 and -4 levels (CG34 combination assay) High-binding 96-well plates were coated with a combined solution of anti-galectin-3 and anti- galectin-4 antibody (1.25μg/ml for each antibody) in coating buffer (Na₂C0₃ 1.6g, NaHC0₃ 1.46g in 1L H₂0) overnight at 4°C. The wells were washed once with washing buffer.

Serum samples were diluted 1:2 with PBS and applied to the coated wells for 2hrs at room temperature. An equal amount of recombinant galectin-3 and -4 (500ng/ml) were combined, serially diluted and applied to some coated wells also for 2hrs at room temperature for generation of the standard curve. The wells were washed twice with washing buffer.

Equal amount (1.25μg/ml in blocking buffer) of biotinylated anti-galectin-3 and -4 antibodies were combined and applied to each well for lhr at room temperature. The wells were washed with washing buffer and applied with HRP-ExtrAvidin (1:10,000 in blocking buffer) for lhr. After two washes with washing buffer, the wells were applied with SigmaFAST OPD developing solution for approximately lOmins until a yellow colour developed in the standard recombinant galectin wells. The reaction was stopped with 4M sulphuric acid and read at 492nm by a microplate reader.

The combined serum concentration of galectin-3 and -4 was calculated from the standard curve derived from the recombinant galectin-3/-4 assay The results were analysed using Kruskall-Wallis non-parametric ANOVA test.

Assessment of serum Carcinoembrvonic antigen (CEA) levels

Serum CEA levels were assessed with the CEA ELISA Kit obtained from MP Biomedicals (Oakbank, UK).

Statistical analysis

Examples 1-4

Kruskall-Wallis non-parametric ANOVA test (SPSS) was used where appropriate. Example 5

Unpaired t test was used for single comparison, one-way ANOVA followed by Bonferroni for multiple comparisons were used where appropriate. Spearman's rho correlation analysis was used for determining correlation of serum galectin-3 and cytokine levels Analysis of sensitivity and specificity of the ELISA-based CEA and Gal-3/4 assay This was done using the following equations:

Numbers of True Positives

Sensitivity =

Numbers of True Positives + Numbers of False Negatives

Numbers of True Negatives

Specificity =

Numbers of True Negatives + Numbers of False Positives

True positives = CEA or galectin-3/-4 level that was above the designated threshold in the patient group.

False positives = CEA or galectin-3/-4 level that was above the designated threshold in the healthy control group.

True negatives = CEA or galectin-3/-4 level that was below the designated threshold in the healthy control group.

False negatives = CEA or galectin-3/-4 level that was below the designated threshold in the patient group.

For the purpose of comparison, the thresholds of galectin-3/-4 assay was compared to the standard CEA threshold of 5μg/L (5ng/ml).

Individual assessments of serum cytokine levels (Example 5) The concentrations of G-CSF, GM-CSF, IL-6 and slCAM-1 in the human serum from colon cancer patients were determined using G-CSF, GM-CSF, IL-6 and slCAM-1 ELISA kits as per the manufacturer's instructions.

Example 1 - Determining galectin levels in serum samples from healthy subjects and CRC patients

Serum samples were obtained as described from three subject groups: healthy subjects (31 samples); non-metastatic CRC patients (40 samples); and CRC patients showing liver metastasis (11 samples). The samples were assayed by ELISA sandwich assay as described, to determine serum levels of a number of galectins. Serum levels of the galectins were compared between the three subject groups and any differences analysed for statistical significance.

The concentrations of several galectins were found to be significantly higher in the sera of CRC patients than in healthy people. The difference was even more striking when levels in CRC metastatic patients were compared with healthy people. Galectin Fold difference in Statistical Fold difference in Statistical CRC patients significance CRC patients with significance compared to healthy liver metastasis

subjects compared to

healthy subjects

2 1.9x higher P<0.001 2.3x higher P<0.01

3 11.3 x higher P<0.001 31.6x higher P<0.001

4 11. lx higher P<0.001 25.3x higher P<0.001

8 1.84 x higher P<0.01 5.6x higher P<0.01

It was further determined that serum Gal3 level shows a reciprocal relationship with serum Gal4 level but not with any other serum galectins (Galectin-1, -2 or -8) in the serum of CRC patients having liver metastasis (Figure 3). Example 2 - Assays based on individual measurement of Gal3 or Gal4

The data obtained in Example 1 for Gal3 and Gal4 was analysed to determine if there is a statistically significant difference in Gal3 or Gal4 levels between any of the three subject groups.

As shown in Figure 2, the individual Gal3 and Gal4 tests were each able to separate healthy subjects from CRC patients at a significant level. However, neither test alone was able to distinguish between metastatic and non-metastatic CRC patients.

Example 3 - The combined CG34 assay

The serum samples described in Example 1 were assayed for Gal3 and Gal4 simultaneously using an ELISA sandwich assay as described in the Methods section. The samples were also subjected to a conventional CEA assay. Differences in results obtained for the three subject groups were analysed for statistical significance as described in the Methods section. Results are shown in Figure 1.

As shown, the combined Gal3/Gal4 assay (CG34 assay) is able to distinguish not only between healthy vs CRC patients, but also between metastatic vs non-metastatic CRC patients, both at a statistically significant level (Figure IB). The CG34 assay performed better than the CEA assay when separating metastatic vs non-metastatic (see Figs 1A and IB).

Using the data obtained, % sensitivity and % specificity were calculated for the CG34 assay at a number of different threshold values for the combined Gal3/Gal4 concentration (0.5ng/ml, lng/ml, 5ng/ml, lOng/ml and 20ng/ml), as described in the Methods section. Sensitivity and specificity calculations were also made for the CEA assay data, using the clinically used threshold of 5ng/ml. Results are shown in Tables 1 and 2 (Figure 4). As shown, the CG34 assay provides a better test for separating metastatic from non-metastatic patients, in terms of balance of sensitivity and specificity.

As a further comparison the data obtained in the individual Gal3 and Gal4 assays of Example 1 was added together, and the results analysed to test for any statistically significant separation of metastatic and non-metastatic C C patients. As shown in Figure 5, adding the results of the individual assays provided data which could separate healthy subjects from CRC patients, but could not separate CRC patients with liver metastasis from non-metastatic CRC patients at a significant level. Example 4 - Determining galectin levels in serum samples from healthy subjects and breast cancer patients

Serum samples were obtained as described from two subject groups: healthy subjects (31 samples); and breast cancer patients (40 samples). The samples were assayed by ELISA sandwich assay as described, to determine serum levels of a number of galectins. Serum levels of the galectins were compared between the two subject groups and any differences analysed for statistical significance.

The concentrations of galectins 2, 3, 4 and 8 were found to be significantly higher in the sera of breast cancer patients than in healthy people (1.2x, 11.3x, 11. Ox and 1.84x higher respectively).

Example 5 - Relationship between circulating levels of Gal3 and circulating cytokine levels in colorectal cancer patients

The relationship between serum galectin-3 and serum IL-6, G-CSF, GM-CSF and slCAM-1 concentrations was investigated in 50 colorectal cancer patients with or without clinically detectable metastasis.

A significant correlation between circulating galectin-3 concentration was observed with G-CSF level (p<0.05) but not with the other 3 cytokines when all patients were considered together (Table 3). However, when serum galectin-3 and cytokines levels were further analyzed in the patients that did not have metastasis and those with metastasis, significant correlations of galectin-3 levels were observed with G-CSF (p=0.04), IL-6 (p=0.05) and slCAM-1 (p=0.005) in patients with metastasis but not in those without metastasis (Table 3).

Cytokines IL6, G-CSF and slCAM-1 are previously known for their positive influence on cancer metastasis. Thus, the correlations identified by the inventors provide further support for the role of Gal3 in metastasis.

The inventors have also, in other studies, obtained data which indicates that circulating Gal3 at pathological concentrations seen in the sera of cancer patients, induces endothelial secretion of metastasis-promoting cytokines IL-6, slCAM-1, G-CSF and GM-CSF from the blood vascular endothelium in vitro and in mice (data not shown). The secretion of these cytokines has been found to enhance the endothelial expression of cell surface adhesion molecules, causing increased adhesion of cancer cells to the vascular endothelium and also enhance endothelial cell migration and micro-vascular tubule formation during angiogenesis (data not shown). Table 3 shows a significant correlation between three of the four tested cytokines, but not for GM-CSF. Without wishing to be bound by theory, the inventors believe that galectin-3 may only have an indirect effect on the secretion of GM-CSF in the patients, for example, by induction of the secretion of the other cytokines.

Claims

1. A method for detecting metastatic cancer in a subject, the method comprising: a) performing a combined assay for both Gal3 and Gal4 protein in a blood sample taken from the subject, thereby obtaining a value for the combined concentration of Gal 3 and Gal4 in the sample; and

b) comparing the value obtained in the combined assay of step (a) with a predetermined threshold value; wherein metastatic cancer is indicated if a value is obtained in step (a) which is higher than the predetermined threshold value.

2. A method according to claim 1 which is for determining metastasis in a cancer patient.

3. A method according to claim 1 or 2 which is for monitoring for tumour recurrence in a patient following surgical removal of a primary tumour.

4. A method according to claim 1 which is for detecting metastatic cancer in an asymptomatic subject.

5. A method according to any of the preceding claims wherein the cancer is colorectal cancer.

6. A method according to any of the preceding claims which further comprises determining a suitable treatment for the subject or patient.

7. A method according to any of the preceding claims wherein in step (a) the combined assay comprises: a) contacting the sample with a binding agent which binds specifically to Gal3 protein and a binding agent which binds specifically to Gal4 protein; and b) determining the combined amount of bound Gal3 and Gal4 protein.

8. A method according to claim 7 wherein the binding agent which binds specifically to Gal3 protein comprises an antibody which binds specifically to Gal3 protein or a fragment or derivative of said antibody and/or the binding agent which binds specifically to Gal4 protein comprises an antibody which binds specifically to Gal4 protein or a fragment or derivative of said antibody.

9. A method according to any of the preceding claims wherein in step (a) the combined assay comprises an ELISA sandwich assay.

10. A method for detecting metastatic cancer in a subject according to any of the preceding claims which additionally comprises performing one or more additional tests indicative of metastasis in a sample taken from the subject.

11. A method according to claim 10 wherein the one or more additional tests comprise assaying the sample for one or more of the cytokines IL-6, G-CSF and slCAM-1.

12. A kit for use in determining metastatic cancer in a subject, the kit comprising means for performing a combined assay for both Gal3 and Gal4 in a blood sample taken from the subject.

13. A kit according to claim 12 wherein the kit comprises:

a) an antibody which binds specifically to Gal3 protein or fragment or derivative of said antibody; and b) an antibody which binds specifically to Gal4 protein or a fragment or derivative of said antibody.

14. A kit according to claim 12 or 13 which is for use in an ELISA sandwich assay to simultaneously determine the concentration of Gal3 and Gal4 in the blood sample.

15. A kit according to any of claims 12 to 14 which comprises means for performing one or more additional tests indicative of metastasis in a sample taken from the subject.

16. A kit according to claim 15 wherein the one or more additional tests comprise assaying the sample for one or more of the cytokines IL-6, G-CSF and slCAM-1.

17. A kit according to any of claims 12 to 16 which is suitable for use in a method according to any of claims 1 to 11.

18. A kit according to any of claims 12 to 17 which is for use in determining metastatic colorectal cancer in a subject.

19. A kit according to any of claims 12 to 18 which is for monitoring for tumour recurrence in a patient following surgical removal of a primary tumour.