Methods To Predict Progression Of Berret ' S Esophagus To High Grade Dysplasia Esophageal Adenocarcinoma

METHODS TO PREDICT PROGRESSION OF BERRET ' S ESOPHAGUS TO HIGH GRADE DYSPLASIA ESOPHAGEAL ADENOCARCINOMA

Be Id of the Invention The invention relates to the prediction of progression from Barrett's Esophagus to high grade dysplasia or esophageal adenocarcinoma. In particular the invention relates to methods for aiding the prediction of likelihood of this progression.

Background to the invention

Gastroesophageal reflux can lead to Barrett's oesophagus (BE), in which the squamous epithelium of the distal oesophagus is replaced by columnar epithelium. BE is found in 3% of patients undergoing endoscopy for reflux symptoms and 1 % of all those undergoing endoscopy (1 , 2) . The importance of BE lies in its potential to progress, via a metaplasia-dysplasia-adenocarcinoma sequence, to esophageal adenocarcinoma (EAC) . EAC continues to carry a dismal prognosis with an overall five year survival rate of less than 20% (3).

The rate of progression from non-dysplastic BE (NDBE) to EAC is in the region of 0.2-0.5% per year based on recent meta-analyses (4-6), and large population-based studies. This represents a small but significantly increased relative risk over the general population. Counselling individual patients is difficult as the majority of patients with NDBE will not progress to EAC and our ability to predict who is at risk is, at present, severely limited (4). This is a problem in the art.

Endoscopic surveillance programmes, to monitor for progression from NDBE to dysplasia or EAC, are very costly and require highly skilled endoscopists. In addition, they have questionable cost effectiveness and their value in reducing cancer mortality is un proven.

There is a growing body of evidence that the majority of EACs arising in BE occur in association with chromosomal instability, characterised by widespread imbalances in chromosome number (aneuploidy/tetraploidy) and loss of heterozygosity. These karyotypic abnormalities are often associated with the accumulation of mutations in specific tumour suppressor genes and oncogenes associated with cancer initiation and progression (7). The Early Detection Research Network (EDRN) has defined a pathway through which putative biomarkers must be validated prior to being applicable to routine clinical use (8). Several biomarkers which have potential for clinical use in BE have been identified in cross-sectional Phase 2 and 3 studies. Surface expression of cyclin A, a marker of a proportion of proliferating cells, has previously been shown to correlate with the degree of dysplasia within BE, and can be assessed using histochemistry (9) . DNA content abnormalities (aneuploidy/tetraploidy) previously measured by flow cytometry ( 10, 1 1 ) can also be accurately assessed using image cytometry DNA analysis (DNA copy number) , which can be applied to paraffin sections ( 12) . Inactivation of the tumour-suppressor p53 gene has been identified as a key event in the development of EAC and bi-allelic inactivation of p53 has been associated with subsequent development of EAC ( 13) .

Coordinated changes in glycan expression have been shown to occur in the development of EAC. Gene expression analysis indicated that the majority of genes which contributed most to enrichment of glycan pathways in the progression to EAC had a role in the synthesis of Lewis antigens, which are known to alter in other cancers ( 1 6-18) . A Phase 3 study (using immunohistochemistry) of Lewis antigens confirmed that sLea (also known as CA19-9) and Lex (CD15) expression was associated with the degree of dysplasia within BE. The lectins (specific glycan-binding proteins) Wheat germ agglutinin (WGA) and Aspergillus oryzae lectin (AOL) have also been shown by lectin histochemistry to have highly significant decreased binding in the progression to EAC ( 19) .

Objective markers individualising the risk of progression at BE diagnosis would allow clinicians to target interventions such as endoscopic surveillance or minimally invasive therapy to those at high risk, whereas those at low risk could be reassured and avoid unnecessary and costly follow up. However, no such individualised risk assessment is currently possible, and no reliable molecular predictors of progression from BE to EAC are known, which are problems in the art.

The present invention seeks to overcome problem (s) associated with the prior art. Summary of the invention

The risk of progression to esophageal adenocarcinoma (EAC) from Barrett's Esophagus (BE) is low and unpredictable. Accurate risk prediction tools are absent from the prior art.

The inventors considered a wide range of candidate markers including sLea, Lex, sialic acid residues detected by WGA and the Fuc l -6GlcNAc residues detected by AOL to predict progression to EAC in BE. In addition, early molecular changes, seen even prior †o the advent of dysplasia, were considered as possible biomarkers of the risk of progression to EAC. However there was no teaching in the art that any such markers had any predictive value for the key parameter of progression to EAC. The inventors tackled these problems by designing a study to assess the performance of candidate biomarkers in predicting neoplastic progression in BE and to create a biomarker panel as a risk stratification tool.

We disclose a phase 3 population-based study which reveals new risk-stratification biomarker panels for Barrett's esophagus. As a result of this study, the inventors have been able to derive small and focused panels of biomarkers which provide a statistically robust prediction tool for progression to EAC. This new ability to predict progression to EAC is itself a significant advance. Moreover, use of this tool enables further benefits such as the optimisation of surveillance and intervention regimes. The present invention is based on these important findings.

Detailed IJesc ri tion of the Invention

In one aspect, the invention relates to a method for aiding prediction of the likelihood of progression from Barrett's esophagus to high grade dysplasia or esophageal adenocarcinoma in a subject, the method comprising

(a) providing an oesophagal sample from said subject

(b) determining if said sample stains abnormally with Aspe rgillus oryzae lectin;

(c) determining if there is a DNA content abnormality in said sample; and

(d) determining if there is low grade dysplasia in said sample;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, then an increased likelihood of progression is determined.

Aspe rgillus oryzae lectin suitably corresponds to the sequence of EMBL accession number BAB88318 such as

MSTPGAQEVLFRTGIAAVNSTNHLRVYFQDSHGS IRESLYESGWANGTAKNVIAKAKLGTPLAATSKELK NIRVYSLTEDNVLQEAAYDSGSGWYNGALAGAKFTVAPYSRIGSVFLAGTNALQLRIYAQKTDNTIQEYM WNGDGWKEGTNLGVALPGTGIGVTCWRYTDYDGPS IRVWFQTDNLKLVQRAYDPHTGWFKELTTIFDKAP PRCAIAATNFNPGKSS IYMRIYFVNSDNTIWQVCWDHGQGYHDKRTITPVIQGSEIAI I SWEGPELRLYF QNGTYVSAI SEWSWARHGSQLGRRALPPAE An increased likelihood of progression means a likelihood of progression which is higher than the standard or normal risk of progression for an equivalent subject which does not show the specified abnormal markers. In other words, the subject with a higher likelihood of progression has a greater chance of progressing to EAC.

An equivalent subject is suitably matched for as many as possible of the basic criteria such as age, sex, etc. Choice of control subjects is well known in clinical research, and is discussed in more detail below and in the examples section. In another aspect, the invention relates to a method as described above further comprising the step of

(e) determining if said sample stains abnormally for CA19-9;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (e) is abnormal then a further increased likelihood of progression is determined.

CA19-9 is also known as Lewis antigen sLea and is well known in the art; in case any further guidance is needed reference is made to CAS-No.92448-22-1 .

In another aspect, the invention relates to a method as described above further comprising the step of

(f) determining if said sample stains abnormally for Cyclin A;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (f) is abnormal then a further increased likelihood of progression is determined. Cyclin A suitably means NCBI accession number CAA02087. Suitably this has a sequence of

1 mlgnsapgpa trealgaasi aadgaprgpg eyqpgkgsar pttadpgala vlksgnprgl 61 aheqrpktrr vaplkdlpvn dehvtvppwk anskqpafti hvdeaekeaq kkpaesqkie 121 redalafnsa islpgprkpl vpldypmdgs fesphtmdms ivledekpvs vnevpdyhed 181 ihtylremev kckpkvgymk kqpditnsmr ailvdwlvev geeyklqnet lhlavnyidr 241 flssmsvlrg klqlvgtaam llaskfeeiy ppevaefvyi tddtytkkqv lrmehlvlkv 301 ltfdlaaptv nqfltqyflh qqpanckves lamflgelsl idadpylkyl psviagaafh 361 lalytvtgqs wpeslirktg ytleslkpcl mdlhqtylka pqhaqqsire kyknskyhgv 421 sllnppetln 1

In another aspect, the invention relates to a method as described above further comprising the step of

(g) determining if said sample stains abnormally for p53;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (g) is abnormal then a further increased likelihood of progression is determined. p53 suitably means NCBI accession number AAD28535. Suitably this has a sequence of:

1 meepqsdpsv epplsqetfs dlwkllpenn vlsplpsqam ddlmlspddi eqwftedpgp

61 deaprmpeaa prvapapaap tpaapapaps wplsssvpsq ktyqgsygfr Igflhsgtak 121 svtctyspal nkmfcqlakt cpvqlwvdst pppgtrvram aiykqsqhmt evvrrcphhe

181 rcsdsdglap pqhlirvegn Irveylddrn tfrhsvvvpy eppevgsdct tihynymcns

241 scmggmnrrp iltiitleds sgnllgrnsf evrvcacpgr drrtekenlr kkgephhelp

301 pgstkralpn ntssspqpkk kpldgeyftl qirgrerfem frelnealel kdaqagkepg

361 gsrahsshlk skkgqstsrh kklmfktegp dsd

In another aspect, the invention relates to a method as described above further comprising the step of

(h) determining if said sample stains abnormally with Wheat Germ Agglutinin lectin; wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (h) is abnormal then a further increased likelihood of progression is determined.

Wheat Germ Agglutinin lectin suitably means NCBI accession number AAA34257.1

Gl: 1 70668. Suitably this has sequence of

1 qrcgeqgsgm ecpnnlccsq ygycgmggdy cgkgcqngac wtskrcgsqa ggktcpnnhc 61 csqyghcgfg aeycgagcqg gpcradikcg sqaggklcpn nlccsqwgyc glgsefcgeg 121 cqngacstdk pcgkdaggrv ctnnyccskw gscgigpgyc gagcqsggcd gvfaeaiatn 181 stllae In another aspect, the invention relates to a method as described above further comprising the steps of

(f) determining if said sample stains abnormally for Cyclin A;

(g) determining if said sample stains abnormally for p53;

(h) determining if said sample stains abnormally with Wheat Germ Agglutinin lectin; wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (f) is abnormal and (g) is abnormal and (h) is abnormal then a further increased likelihood of progression is determined.

In another aspect, the invention relates to a method as described above further comprising the steps of

(f) determining if said sample stains abnormally for Cyclin A;

(g) determining if said sample stains abnormally for p53;

(h) determining if said sample stains abnormally with Wheat Germ Agglutinin lectin; wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (e) is abnormal and (f) is abnormal and (g) is abnormal and (h) is abnormal then a further increased likelihood of progression is determined. In another aspect, the invention relates to an assay for selecting a treatment regimen, said assay comprising

(a) providing an oesophagal sample from said subject

(b) determining if said sample stains abnormally with Aspe rgillus oryzae lectin;

(c) determining if there is a DNA content abnormality in said sample; and

(d) determining if there is low grade dysplasia in said sample;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, then a treatment regimen of increased surveillance is selected.

Increased surveillance may mean that additional surveillance techniques are applied to the subject, or may mean that additional surveillance events are applied to the subject. Suitably increased surveillance means that additional surveillance events are applied to the subject. In other words, suitably increased surveillance means that the subject is monitored more frequently (i.e. more often) if a treatment regimen of increased surveillance is selected. ' Increased' means a higher (more frequent) rate of surveillance events than the subject was receiving before the assay was carried out. In other words, suitably increased surveillance means a reduction in the interval between surveillance events for that subject. Clearly individual subjects may be under individual surveillance regimens based on their clinical history. Thus, 'increased' surveillance or reduction of the interval between surveillance events must be judged on a subject-by- subject basis rather than an absolute value for (e.g.) surveillance intervals. This is explained in more detail in example 2.

Surveillance is not regarded as a method of treatment of the human or animal body, but is rather regarded as having its natural meaning of monitoring. Thus suitably if the regimen selected relates only to surveillance, then suitably the assay is for selecting a surveillance regimen and may not involve treatment of the human or animal body.

In another aspect, the invention relates to an apparatus or a system which is

(a) configured to analyse an oesophagal sample from a subject, wherein said analysis comprises

(b) determining if said sample stains abnormally with Aspe rgillus oryzae lectin;

(c) determining if there is a DNA content abnormality in said sample; and

(d) determining if there is low grade dysplasia in said sample; said apparatus or system comprising an output module,

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, then said output module indicates an increased likelihood of progression to EAC for said subject.

Suitably the sample comprises formalin fixed paraffin embedded material. Asp e rsi us oryzae lectin

Suitably Asp e rgiBus oryzae lectin staining may be determined by any suitable method known in the art. Suitably it is determined by a method suitable for use on FFPE material.

More suitably step (b) comprises determining the level of staining by Aspe rgiBus oryzae lectin in said sample by grading the intensity of the stain and the percentage of the area stained at said intensity, and producing an H score for said stain.

DNA content abnormality

Suitably determining whether there is a DNA content abnormality may be done by any suitable method known in the art. Suitably it is determined by a method suitable for use on FFPE material.

More suitably step (c) comprises determining the DNA content by image cytometry DNA analysis and inferring from said determination whether said DNA content is abnormal. Using this detection technique has the advantage of overcoming many of the technical difficulties associated with flow cytometric detection, which is the standard in the art.

Low Grade Dysplasia

Suitably determining whether there is low grade dysplasia may be done by any suitable method known in the art. Suitably it is determined by a method suitable for use on FFPE material.

More suitably step (d) comprises histological examination of the sample, and scoring the presence or absence of low grade dysplasia using the Vienna scale.

The Vienna scale is a well-known standard for scoring dysplasia, and is described in detail in the art such as in Schlemper RJ, Riddell RH, Kato Y, e t al The Vienna classification of gastrointestinal epithelial neoplasia. Gut. 2000 August 1 , 2000:47(2) :251 - 5, which is incorporated herein by reference specifically for the technique of scoring lesions according to the Vienna scale.

CA19-9 (Lewis antigen sLea)

Suitably determining whether there is abnormal staining for CA19-9 may be done by any suitable method known in the art. Suitably it is determined by a method suitable for use on FFPE material.

More suitably step (e) comprises immunohistochemical staining of the sample for CA19- 9, grading the intensity of the stain and the percentage of the area stained at said intensity, and producing an H score for said stain.

Cvclin A

Suitably determining whether there is abnormal staining for Cyclin A may be done by any suitable method known in the art. Suitably it is determined by a method suitable for use on FFPE material.

More suitably step (f) comprises immunohistochemical staining of the sample for Cyclin A, determining the percentage of Cyclin A positive epithelial surface cells compared to Cyclin A negative epithelial surface cells, wherein a percentage of 1 % or more Cyclin A positive epithelial surface cells is an abnormal stain. o53

Suitably determining whether there is abnormal staining for p53 may be done by any suitable method known in the art. Suitably it is determined by a method suitable for use on FFPE material.

More suitably step (g) comprises immunohistochemical staining of the sample for p53, and scoring the stain as normal or abnormal wherein either a strong, dark stain compared to background stain or an absence of stain compared to background stain is an abnormal stain.

Wheat Germ Agglutinin lectin

Suitably Wheat Germ Agglutinin lectin staining may be determined by any suitable method known in the art. Suitably it is determined by a method suitable for use on FFPE material. More suitably step (h) comprises determining the level of staining by Wheat Germ Agglutinin lectin in said sample by grading the intensity of the stain and the percentage of the area stained at said intensity, and producing an H score for said stain. In another aspect, the invention relates to use for prognostic or risk prediction applications relating to high grade dysplasia or esophageal adenocarcinoma, of a material which recognises, binds to or has affinity for a polypeptide selected from p53, Cyclin A, or CA19-9, or a fragment, variant or mutant thereof. In another aspect, the invention relates to use of a combination of materials, each of which respectively recognises, binds to or has affinity for one or more of said polypeptide (s), or a fragment, variant or mutant thereof. Suitably said material comprises an antibody or fragment thereof such as an antigen binding fragment thereof.

Antibodies may be polyclonal or monoclonal, and may be multispecific (including bispecific), chimeric or humanised antibodies. Antibodies according to the present invention may be modified in a number of ways. Indeed the term "antibody" should be construed as covering any binding substance having a binding domain with the required specificity. Thus, the invention covers antibody fragments, derivatives, functional equivalents and homologues of antibodies, including synthetic molecules and molecules whose shape mimics that of an antibody enabling it to bind an antigen or epitope.

Examples of antibody fragments, capable of binding an antigen or other binding partner, are the Fab fragment consisting of the VL, VH, CI and CHI domains; the Fd fragment consisting of the VH and CH I domains; the Fv fragment consisting of the VL and VH domains of a single arm of an antibody; the dAb fragment which consists of a VH domain; isolated CDR regions and F(ab')2 fragments, a bivalent fragment including two Fab fragments linked by a disulphide bridge at the hinge region. Single chain Fv fragments are also included.

Antibody fragments, which recognise specific epitopes, may be generated by known techniques. For example, such fragments include, but are not limited to, the F(ab')2 fragments which can be produced by pepsin digestion of the antibody molecule and the Fab fragments which can be generated by reducing the disulfide bridges of the F(ab')2 fragments. Alternative, Fab expression libraries may be constructed (Huse, et al., 1989, Science 246: 1275-1281 ) to allow rapid and easy identification of monoclonal Fab fragments with the desired specificity. In another aspect, the invention relates to use for prognostic or risk prediction applications relating to high grade dysplasia or esophageal adenocarcinoma, of Aspe rgiBus o ryzae lectin or Wheat Germ Agglutinin lectin. In another aspect, the invention relates to use for prognostic or risk prediction applications relating to high grade dysplasia or esophageal adenocarcinoma, of image cytometry DNA analysis.

In another aspect, the invention relates to an assay device comprising a material which binds to or has affinity for a polypeptide selected from p53, Cyclin A, or CA19-9 and which further comprises Aspe rgillus oryzae lectin and/or Wheat Germ Agglutinin lectin.

Suitably said lectin comprises a detectable label. In all cases, if (b) is normal and/or (c) is normal and/or low grade dysplasia is absent, then there is no determination of increased likelihood of progression to EAC. In these embodiments, no increased surveillance would be indicated and no minimally invasive therapy would be indicated. In all cases, CD15 does not have predictive value. CD15 is suitably not assayed in the present invention. Data relating to CD15 are included for comparison purposes to demonstrate the benefits which the predictive markers of the invention provide.

The invention may be used in connection with high grade dysplasia or esophageal adenocarcinoma. Suitably the invention is in connection with high grade dysplasia. This has the advantage of allowing earlier intervention. More suitably the invention is in connection with esophageal adenocarcinoma. This has the advantage of treating a more threatening condition. Definitions

The term 'comprises' (comprise, comprising) should be understood to have its normal meaning in the art, i.e. that the stated feature or group of features is included, but that the term does not exclude any other stated feature or group of features from also being present. Abnormal vs Normal

I† is believed that the skilled reader can determine whether or not the markers are normal or abnormal using the guidance given above. However, in case further explanation is helpful, this categorisation may be carried out as follows:

Each biomarker score is dichotomised into abnormal or normal.

For DNA copy number, p53 and cyclin A this is on the basis of a prio n knowledge (for example for DNA copy number diploid is normal, tetraploidy or aneuploidy is abnormal) .

For CA 19-9, WGA, AOL and CD-I 5, each of the components of the staining score (most suitably each of four components as described below) were dichotomised into normal/abnormal based on the most frequent median distribution amongst controls and then summed to give an overall abnormal score of 0-4 for each biomarker.

The term ' mutant' of a biomarker such as a polypeptide biomarker of the invention should have its normal meaning in the art. Mutants are sometimes referred to as 'variants' or 'alleles' . The key is to detect biomarkers as have been set out herein. The biomarkers may possess individual variations in the form of mutations or allelic variants between individuals being studied. Therefore there may be some degree of deviation from the exemplary SEQ ID NOs provided herein. The SEQ ID NOs provided herein are to assist the skilled reader in identifying and working with the polypeptides/biomarkers of the invention and are not intended as a restricted and inflexible definition of the individual polypeptides being assayed. Thus minor sequence differences between the SEQ ID NOs provided and the actual sequences of the polypeptide biomarkers being detected will be expected within the boundaries of normal variation between subjects. This should not affect the working of the invention. Fragments/Peptides

It will be appreciated by the skilled worker that the details of the biomarkers discussed herein and in particular the sequences presented for them are given to facilitate their detection. The important information being gathered is the presence or absence (or particular level) of the biomarker in the sample being studied. There is no particular requirement that the full length polypeptide be scored. Therefore the invention embraces the detection of fragments of the polypeptide/lectin biomarkers. Moreover, the kits and peptides of the invention may comprise fragments of the polypeptides and need not comprise the full length sequences exemplified herein. Thus a fragment is suitably at least 6 amino acids in length, suitably at least 7 amino acids in length, suitably at least 8 amino acids in length, suitably at least 9 amino acids in length, suitably at least 10 amino acids in length, suitably at least 1 5 amino acids, suitably at least 25 amino acids, suitably at least 50 amino acids, suitably at least 100 amino acids, or suitably the majority of the biomarker polypeptide of interest.

Sample

The sample may be from a subject. The subject is suitably a mammal, most suitably a human.

Suitably the methods do not involve actual collection of the sample. Suitably the sample is an in vitro sample. Methods of the invention are suitably performed on an isolated sample from the subject being investigated. Thus, suitably the methods are methods which may be conducted in a laboratory setting without the need for the subject to be present. Suitably the methods are carried out in vitro i.e. suitably the methods are in vitro methods. Suitably the sample is from a subject having Barrett' s Esophagus.

Suitably the sample comprises material taken from the region of the Barrett' s Esophagus. Suitably the sample comprises material taken from the Barrett's Esophagus segment itself.

Suitably the sample is a biopsy.

Suitably the sample comprises formalin fixed paraffin embedded (FFPE) material. It is an advantage of the invention that the methods can be carried out on FFPE material. Clinically obtained samples are often processed into FFPE material. Therefore the invention may be practiced very widely.

Suitably each of the markers described herein may be detected in FFPE material. Although there are a range of techniques for assaying the markers described, most suitably those techniques which can be carried out on FFPE material are used. This provides the advantage that only one sample is needed in order to assess multiple markers as described herein. Reference Standard

The reference standard typically refers to a sample from a healthy individual i.e. one who does not have EAC. The reference standard may be from a healthy individual who has BE but does not have HGD/EAC, most suitably does not have EAC.

Moreover, controls may be chosen with greater precision depending on which marker is being considered. For example if considering AOL then it may be advantageous to choose a control of BE without dysplasia or EAC. For example if ploidy is being considered then it may be advantageous to choose a control of any normal tissue (normal squamous oesophagus for example) .

The reference standard can an actual sample analysed in parallel. Alternatively the reference standard can be one or more values previously derived from a comparative sample e.g. a sample from a healthy subject. In such embodiments a mere numeric comparison may be made by comparing the value determined for the sample from the subject to the numeric value of a previously analysed reference sample. The advantage of this is not having to duplicate the analysis by determining concentrations in individual reference samples in parallel each time a sample from a subject is analysed.

Suitably the reference standard is matched to the subject being analysed e.g. by gender e.g. by age e.g. by ethnic background or other such criteria which are well known in the art. The reference standard may be a number such as an absolute concentration drawn up by one or more previous studies.

Reference standards may suitably be matched to specific patient sub-groups e.g. elderly subjects, or those with a previous relevant history such as acid reflux or BE.

Suitably the reference standard is matched to the sample type being analysed. For example the concentration of the biomarker polypeptide(s) being assayed may vary depending on the type or nature of the sample. It will be immediately apparent to the skilled worker that the concentration value(s) for the reference standard should be for the same or a comparable sample to that being tested in the method(s) of the invention. For example, if the sample being assayed is from the Barrett's segment then the reference standard value should be for Barrett's segment to ensure that it is capable of meaningful cross-comparison. Suitably the sample type for the reference standard and the sample type for the subject of interest are the same. Where H scores are determined for staining of the sample, determination of whether the stain is normal or abnormal is suitably carried out by comparing the determined H score to a reference standard H score (such as the most frequent median distribution amongst sample (s) from control subjects) and inferring from said comparison whether said staining is abnormal.

Hsk Scoring

It is possible to quantify the risk for an individual subject depending on their particular scores for the various markers in the panels of the invention.

Risk of progression - individual biomarkers

Table 2 illustrates the risk of progression from BE to HGD/cancer for each of the individual biomarkers. Dysplasia in BE patients who progressed (cases) was more frequently diagnosed in the clinical setting, compared with dysplasia scored by two blinded expert gastrointestinal pathologists, although the latter score had superior specificity. Dysplasia, DNA copy number, CA 19-9 and AOL abnormalities were all significantly directly associated with risk of progression, while cyclin A, p53 and WGA abnormalities showed non-significant tendencies towards an increased risk of progression. CD-I 5 abnormalities were not associated with progression risk in this patient group. It should be noted, however, that sensitivity and specificity of the individual biomarkers were low as reflected in the true and false prediction rates (TPR and FPR) . Analysis restricted to cancer outcomes alone resulted in largely similar results to the combined analysis for HGD and EAC, with the exception of p53 which became significantly associated with an increased risk of progression from BE to cancer (OR 1 .95; 95% CI: 1 .04-3.67).

Combination seven biomarker panel model

Table 3 shows effect estimates from the logistic regression models after combining all seven biomarkers (along with dysplasia and year of BE, age and sex adjusted for by study design) into a risk score and shows estimates of the performance of this risk score. The c statistic from the seven biomarker model was relatively high (c statistic = 0.78) and maintained after internal validation using bootstrap sampling techniques (optimism corrected c statistic = 0.71 ) .

It is important to note that this seven biomarker model was able to better discriminate cases and controls, compared with a basic clinical model (containing just dysplasia, age, sex and year of BE), reflected by an improvement in the c statistic of 0.15 (Figure 1 ) . Table 3 also illustrates similar improvements in discrimination slopes (i.e. the difference in mean predicted probability of being a case from the model in the case and controls) for the models containing seven biomarkers (0.22) and from the basic model (0.07) . In a similar analysis restricted to non-dysplastic BE individuals (76% cases and 74% controls), the c statistic for the seven biomarker model was lower than that observed for all BE patients (optimism corrected c statistic = 0.63) but still represented some improvement over the basic model (optimism corrected c statistic = 0.43) . Similarly, the IDI of 0.13 indicated that the seven biomarker model improved discriminative ability compared with the basic model. Reduced biomarker panel model

It was noted that several biomarkers in the full model were not significantly contributing to the overall model. Furthermore, the smaller the biomarker panel the easier it would be to apply this in routine clinical practice. We therefore set out to construct a reduced model by applying a backward selection procedure. Table 4 shows the performance of this reduced model (AOL and DNA copy number) compared with basic model (age, sex and year of BE) and the 7 biomarker model (Figure 1 ) . It can be seen that this model was equivalent to the seven biomarker model in discriminating cases and controls. It should however be noted that there are large confidence intervals for this analysis in view of the small number of patients.

A risk score was then created for individuals who may be abnormal for one, two or all of these markers, as shown in Table 5. For each 1 point increase in risk score, dysplastic BE patients are almost at four-fold increased risk of progressing to develop EAC (OR 3.75; 95% CI: 2.42-5.79), while NDBE patients have a three-fold increased progression risk (OR 2.99; 95% CI: 1 .72-5.20).

Thus in some embodiments the invention relates to a method as described herein, further comprising the step of determining a risk score for the subject. Suitably the risk score is determined from the individual marker/panel results as explained herein.

Suitably the method may further comprise ascribing a risk of progressing to develop EAC. Suitably this risk is derived using the risk score determined from the individual marker/panel results as above. Flirthe r Pin b o dime nts

In another aspect, the invention relates to an assay for selecting a treatment regimen, said assay comprising (a) providing an oesophagal sample from said subject

(b) determining if said sample stains abnormally with Aspe rgillus oryzae lectin;

(c) determining if there is a DNA content abnormality in said sample; and

(d) determining if there is low grade dysplasia in said sample;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, then a treatment regimen of increased surveillance is selected.

Suitably said assay further comprises the step of

(e) determining if said sample stains abnormally for CA19-9;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (e) is abnormal then a treatment regimen of increased surveillance is selected. More suitably a treatment regimen of minimally invasive therapy may be recommended.

Suitably said assay further comprises the step of

(f) determining if said sample stains abnormally for Cyclin A;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (e) is abnormal then a treatment regimen of increased surveillance is selected. More suitably a treatment regimen of minimally invasive therapy may be recommended. Suitably said assay further comprises the step of

(g) determining if said sample stains abnormally for p53;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (e) is abnormal then a treatment regimen of increased surveillance is selected. More suitably a treatment regimen of minimally invasive therapy may be recommended.

Suitably said assay further comprises the step of

(h) determining if said sample stains abnormally with Wheat Germ Agglutinin lectin; wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (e) is abnormal then a treatment regimen of increased surveillance is selected. More suitably a treatment regimen of minimally invasive therapy may be recommended.

In all cases, if (b) is normal and/or (c) is normal and/or low grade dysplasia is absent, then a treatment regimen of normal surveillance is selected. Normal surveillance may vary depending on the subject; suitably normal surveillance means surveillance and the interval already ascribed for said subject.

In another aspect, the invention relates to a method of collecting information useful in determining prognosis for developing EAC, the method comprising (a) providing an oesophagal sample from said subject

(b) determining if said sample stains abnormally with Aspe rgillus oryzae lectin;

(c) determining if there is a DNA content abnormality in said sample; and

(d) determining if there is low grade dysplasia in said sample.

In another aspect, the invention relates to an apparatus or system which is

(a) configured to analyse an oesophagal sample from a subject, wherein said analysis comprises

(b) determining if said sample stains abnormally with Aspe rgillus oryzae lectin;

(c) determining if there is a DNA content abnormality in said sample; and

(d) determining if there is low grade dysplasia in said sample;

said apparatus or system comprising an output module,

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, then said output module indicates an increased likelihood of progression to EAC for said subject.

Said apparatus or system may further be configured such that the analysis further comprises

(e) determining if said sample stains abnormally for CA19-9;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (e) is abnormal then said output module indicates a further increased likelihood of progression is determined.

Said apparatus or system may further be configured such that the analysis further comprises

(f) determining if said sample stains abnormally for Cyclin A;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (f) is abnormal then said output module indicates a further increased likelihood of progression is determined.

Said apparatus or system may further be configured such that the analysis further comprises

(g) determining if said sample stains abnormally for p53;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (g) is abnormal then said output module indicates a further increased likelihood of progression is determined. Said apparatus or system may further be configured such that the analysis further comprises

(h) determining if said sample stains abnormally with Wheat Germ Agglutinin lectin; wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (h) is abnormal then said output module indicates a further increased likelihood of progression is determined.

The invention also relates to kits comprising AOL and/or WGA together with reagents for ICM DNA detection. The invention also relates to said kit (s) further comprising material capable of binding one or more of Cyclin A, p53, CA 1 9-9.

Brief Description of the figures

Figure 1 shows ROC curve comparing 7 biomarker model (Panel A) and 3 biomarker model (Panel B) with basic clinical model (age, sex, presence of dysplasia, year of BE diagnosis) .

Figure 2 shows the clinical algorithm for application of the biomarker panel of the invention to clinical practice. LGD= low grade dysplasia; HGD= high grade dysplasia; AOL=Asp e rgiHus o ryzae lectin. The invention is now described by way of example. These examples are intended to be illustrative, and are not intended to limit the appended claims.

Ekamples

Summary of the Examples:

Methods: A nested case control study within the population-based Northern Ireland BE Register, ( 1993-2005) . Cases (n=89) who progressed to EAC or HGD >6 months following diagnosis of BE were matched to non-progressor controls (n=291 ) for age/sex and year of BE diagnosis. Established (abnormal DNA content, p53 and cyclin A expression) and novel biomarkers (sialyl Lewis A and CD-I 5 expression, Aspergillus oryzae lectin (AOL) binding and Wheat germ agglutinin (WGA) lectin binding were assessed in paraffin embedded tissue from first BE diagnosis. Conditional logistic regression analysis was applied to investigate risk of progression.

Results: After a mean 6.7 (±3.2) years of follow-up, dysplasia and 7 biomarkers contributed to the risk of EAC/HGD in a multivariable analysis. Using a backward selection technique, a panel comprising low grade dysplasia and 2 biomarkers (abnormal DNA ploidy and AOL) was optimal for distinguishing progressors and non- progressors. When combined into a risk score for dysplastic (0-3) or non-dysplastic BE (0- 2), the adjusted OR for progression was 3.74 (95%CI 2.43-5.79) and 2.99 (95%CI 1.72-5.20) for each point increase, respectively.

Conclusion: A panel of 3 biomarkers (sometimes discussed as 2 molecular biomarkers (DNA/AOL) plus dysplasia) can distinguish both dysplastic and non-dysplastic BE patients at higher risk for future development of EAC and HGD.

Ekample 1: Identification of HomaikerRinel

Using the population based Northern Ireland Barrett's Oesophagus Register (NIBR), we constructed a nested case-control study to investigate a panel of established and novel biomarkers on routinely collected paraffin embedded samples. One aim of this study was to determine whether dysplasia, DNA copy number, p53, cyclin A and the more recently described glycan targets (CA19-9, CD-I 5, WGA, AOL), were predictive of future progression to EAC from BE. Another aim was to create a biomarker panel which could be used as a quantitative risk stratification tool.

The methods arising from this study will be useful in a future phase 4 study.

Study de sign

A nested case-control study was conducted within the Northern Ireland Barrett's Oesophagus Registry (NIBR), (20-22). The NIBR is a population-based register of all 9,329 adults diagnosed with columnar-lined oesophagus between 1993 and 2005 throughout Northern Ireland. Within the NIBR, 4,306 patients with a visible columnar lined segment displayed Specialized Intestinal Metaplasia (SIM) on biopsy and, for the purposes of this study, this definition was used for diagnosis of BE. Cases were BE patients from the NIBR who developed EAC, gastric cardia malignancies or HGD at least 6 months after their initial BE diagnosis. Cancer outcomes were identified by matching the NIBR to the Northern Ireland Cancer Registry database of oesophageal and gastric malignancies diagnosed up to 31 st December 2005. Gastric cardia cancers arising in BE patients were considered to be cases, as these are likely to be categorised as EAC cases under TNM7 HGD outcomes were identified as 2 separate clinical diagnoses within 12 months or 3 separate diagnoses of HGD regardless of time period. Each case was matched to up to 5 controls who were BE patients who had not developed EAC or HGD by the study censor date of 31 st December 2005, based on age (±5 years), sex and year of BE diagnosis. Ethical approval was granted by the regional ethics committee in Northern Ireland (REC No 07/NIR02/109) . Hstopatho bgy re vie w

All index and outcome biopsies were reviewed blindly by two independent expert gastrointestinal pathologists (MN & DM), and scored using the Vienna scale (23). Complete agreement was achieved in 84% of samples scored, with discrepancies resolved by discussion and a re-evaluation of the slide where necessary. Cases and controls were excluded if they did not have an index BE biopsy available for review taken >6 months prior to their HGD or cancer diagnosis. Any index BE samples that were scored by pathologists as having either no evidence of SIM or evidence of HGD and EAC (Vienna score of 4-5) at the outset of the study were excluded from the study. The histological diagnosis of the outcome biopsy was confirmed by both pathologists in 88% of EAC, 100% of gastric cardia and 83% of HGD outcome biopsies. Reasons for outcome biopsies not being histologically verified included no biopsy available or no HGD or EAC present in the tissue section cut from the outcome biopsy. Image cyto me try DNA analysis

One 40μιτι section was cut from FFPE tissue and nuclear monolayers were prepared as previously described (24) . The monolayer was then acid hydrolysed and stained with Feulgen-Schiff reagent using standardized methodology (13) . All slides were given unique code identifiers and studied using an automated image cytometric analyser (Room 4, East Sussex, UK) that consists of a microscope (Axioplan 2, Zeiss, Jena, Germany), a 546-nm green filter, and a black-and-white, high-resolution digital camera (AxioCam MRm, Zeiss, Jena, Germany). Optical density and nuclear area were measured and integrated optical density of each nucleus was calculated as previously described (25) . A histogram representing the DNA content was produced and analysed according to European Society for Analytical Cellular Pathology (ESACP) guidelines (26) . DNA Ploidy-related parameters such as DNA index (Dl) and percentages of cells exceeding 5c (5c ER) and 9c (9c ER) were also noted. All histograms were reported blindly by two of three independent observers (JD, DO and MN) . Consensus was reached in all cases.

Immuno histo c he mistry

Immunohistochemistry was performed for p53 (1 /50 dilution, retrieval H I 30min, D07, Leica, Milton Keynes), Cyclin A (1 /50 dilution, retrieval H I 10 min, Leica), Sialyl Lewis a (BOND ready, retrieval H I 20 min, Leica) and CD15 (BOND ready, retrieval H2 20 min, Leica) on 4 μιτι section using the BOND autostainer (Leica, Milton Keynes, UK) following manufacturer instructions. Stained sections were then counterstained with a light haematoxylin stain. stoc he mistry

Slides were de-paraffinised according†o standard procedures, placed in a humidified incubation chamber and 5 Mg/mL of lectin (WGA or AOL) applied, followed by incubation at 37°C for 15 minutes. WGA was obtained as Biotinylated wheat germ from Vector labs Catalog number V0428. AOL was prepared using AOL purchased from Tokyo Chemical Industry UK Ltd L0169, which was labelled with biotin using ProtOn labelling kit PLK-1202. The un-bound lectin was washed off by immersion in running water (2X 30mins), and mounted with Prolong Gold antifade reagent with DAPI (Invitrogen, Paisley, UK). The lectin-stained slides were scanned using Applied Imaging Ariol® (Genetex Ltd, Hampshire, UK) .

p53 was stained with Leica Novocastra antibody clone D07.

CDl 5 was stained with Leica Novocastra antibody Carb-1 PA0039.

CA19-9 (SLa or sLea) was stained with Leica Novocastra antibody 241 :5: 1 :4 PA0424.

Cyclin A was stained with Leica Novocastra antibody clone 6E6.

Sc oring ofhisto - and immuno histoc he mistry

p53 was scored as significant or non-significant. Strong, dark staining or total absent of staining next to normal background stain were significant. The percentage of cyclin A positive compared to negative epithelial surface cells was calculated. A cut off of 1 % was used for significance.

Slides stained for CDl 5, SLa, AOL and WGA were graded for intensity (0-3) and for the % of the area stained at this intensity (0-4) . The H score (0-12), derived by multiplying intensity and area scores was calculated (0-12) (27) and provides more accurate scoring compared to either the intensity or area score alone. Four epithelial compartments were assessed: the apical epithelial membrane (apical part of the epithelial cell membrane which is exposed to the oesophageal lumen); pan membranous, epithelial mucous globule (globular collections of staining within the cytoplasm) and epithelial cytoplasm). Statistic al analysis

Baseline characteristics between cases and controls were compared using independent †-†es†s for continuous variables and chi-squared tests for categorical variables. Conditional logistic regression was conducted to estimate odds ratios for neoplastic progression and corresponding 95% confidence intervals for each biomarker investigated. In analyses where a case failed to be assigned a biomarker score, both the case and its corresponding controls were excluded from that analysis. Multivariate analysis was adjusted for age, sex, year of BE diagnosis, and the presence of dysplasia (as scored by two expert pathologists). The True Positive Rate (TPR) or sensitivity and False Positive Rate (FPR) or 1 -specificity of biomarkers, were calculated by dividing the number of cases and controls, scored as abnormal by the total number of cases and controls scored for that biomarker. Positive Predictive Value (PPV) and Negative Predictive Value (N PV) for the entire cohort of individuals with BE were estimated from the nested case-control design, based upon the prevalence of cancer and HGD outcomes amongst the SIM positive patients at index Barrett' s oesophagus diagnosis from the entire 1993-2005 population-based N IBR cohort (28) .

Logistic regression was used to combine the seven biomarkers along with dysplasia, year of BE, age and sex into a risk score. Each biomarker score was dichotomised into abnormal or normal. For DNA copy number, p53 and cyclin A this was on the basis of a p rio n knowledge (for example for DNA copy number diploid normal, tetraploidy or aneuploidy abnormal) . For CA 19-9, WGA, AOL and CD-I 5, each of the four components of the staining score were dichotomised into normal/abnormal based on the most frequent median distribution amongst controls and then summed to give an overall abnormal score of 0-4 for each biomarker. To evaluate the performance of models, predicted probabilities were used to calculate the area under the receiver operating characteristic (ROC) curves (the c statistic) and to calculate the discrimination slope (the difference in mean predicted probability in the cases compared with the controls) (29) . The performance of a clinical model (containing only dysplasia, year of BE, age and sex) was compared with the performance of a biomarker model containing all seven biomarkers (along with dysplasia year of BE, age and sex ) using the c statistic and integrated discrimination improvement (IDI) (30) . The IDI measures how much the biomarker model leads to increased estimated risks of cancer (or HGD) for cases and decreased estimated risks for controls, compared with the clinical model without biomarkers. Internal validation was conducted on the c statistics using bootstrap methods. Specifically, the model was estimated in a bootstrap sample and the c statistics was calculated in the bootstrap sample and in the original sample. This process was repeated 200 times and the average difference in performance in the bootstrap sample and in the original sample was calculated (the optimism) and subtracted from the apparent performance to estimate the internally validated performance (31 ) .

A reduced model was selected using the stepwise backward selection procedure (using a cut off of 0.5) , retaining age, year of BE and sex in the model. The internally validated c statistic for this model was calculated as described previously but after applying the stepwise selection procedure to each bootstrap sample. Validation of the model was conducted using logistic regression, rather than conditional logistic regression, for simplicity, but estimates from the logistic and conditional logistic models were similar (results not shown). Additional stratified analyses were repeated removing individuals without dysplasia at index biopsy.

Patie nt de mogruphic s

After a mean (SD) follow-up period of 6.7 (3.3) years, 56 EAC, 13 gastric cardia cancers and 25 HGD cases were diagnosed within the study cohort. After review by two expert pathologists, five cases (2 EAC, 1 gastric cardia cancer and 2 HGD cases) were found to have evidence of HGD or EAC at their initial BE diagnosis and were therefore excluded from analysis, leaving 89 cases. Table 1 shows the characteristics of these 89 cases and their 291 matched controls. Cases and controls did not differ by matching criteria (age, sex and year of BE diagnosis), nor laboratory of origin or length of BE segment, although the latter was unknown for approximately half of participants. Significantly more cases were diagnosed as having indefinite or low grade dysplasia at their first BE diagnosis compared with controls, (20.2% versus 2.4% respectively), (pO.001 ) .

Conclusions

This population based phase 3 biomarker study has revealed a new combination panel for assessment of risk to progression from both dysplastic and non-dysplastic BE to EAC. The combination panel of 7 biomarkers represents a significant step to individualise patient risk and is advantageous due to the use of relatively simple techniques that can be carried out on formalin fixed tissue. Furthermore analysis of a simplified 3 biomarker panel model showed a significant stepwise increase risk of progression for AOL, DNA content abnormalities and presence of LGD. This work was strengthened by the study design, using a well described cohort with long term follow up and in which all biomarker analysis was undertaken blinded to the outcome. The NIBR is a valuable population based resource, with over 4,000 adults diagnosed with visible columnar-lined oesophagus and SIM within Northern Ireland (population 1.8 million). The 89 cases of progression to HGD/EAC from BE represent a substantially larger number of endpoints than previous longitudinal studies of biomarkers in BE from other centres (32, 33). The majority of these patients were Caucasian middle aged males, which is concordant with previous epidemiological studies of BE. Importantly samples were collected as part of routine clinical care and hence are applicable†o everyday practice. A weakness of this study is that limited biopsies were available for analysis and hence it was not possible to achieve data for all of the biomarkers in every patient. Of the validated biomarkers, DNA content abnormalities measured by image cytometry were most predictive of HGD/cancer progression (OR = 3.22; 95% CI 1.73- 6.00; PO.001 ) . DNA ploidy abnormalities have been evaluated in prospective trials of patients with BE, representing phase 4 biomarker development (10, 1 1 , 34) . A landmark study by the Reid group demonstrated that patients who had both HGD and aneuploidy or tetraploidy had a five year cancer risk of 66%, compared to 42% with HGD alone and 28% with DNA ploidy abnormalities alone. Patients who had no cytometric abnormality (DNA diploid) and did not have HGD had a five year cancer risk of zero. [Reid et al., 2000b] The same group went on to evaluate the role of a chromosomal instability panel, combining 9pLOH, 1 7pLOH and DNA ploidy abnormalities. [30] The combination of all three was a better predictor of progression to EAC than any one biomarker alone (RR = 38.7; 95% CI = 10.8-138.5; pO.001 ) . These markers are not used in most centres however, as they require a combination of platforms that would be difficult to perform outside of specialist research centres. DNA content flow cytometry particularly is associated with inter laboratory variability, quality control issues and significant set up and running costs (35) . The use of image cytometry rather than flow cytometry in this study therefore represents an advance as image cytometry is cheaper, easier to set up and has potential for automation. Although new single platform techniques to measure chromosomal instability, such as SNP and gene chip arrays (36), are being developed that may provide rapid throughput of FFPE material, the accuracy and cost implications for surveillance programmes remains unclear.

Of the novel biomarkers involved in glycan expression, CA19-9 and the lectin AOL were significant predictors of cancer development, although only AOL was included as a significant independent biomarker of neoplastic progression in the final reduced model. Both CA19-9 and AOL have previously been shown to be upregulated in the progression from metaplasia through dysplasia to cancer in Phase 2 studies (19, 37) . AOL can be readily assessed following lectin histochemistry which is cheaper and simpler to perform than immunohistochemistry and ploidy techniques. However, AOL is also overexpressed in almost a quarter of BE controls who did not progress, thus limiting its specificity. Nevertheless, the intriguing possibility also exists that it may be possible to evaluate AOL positive areas endoscopically, as previously demonstrated for WGA lectin (19) . The strongest predictor of cancer progression was the presence of LGD by 2 expert Gl pathologists (OR = 1 1 .33; 95% CI 3.97-32.36; pO.01 ), although the overall presence in the cohort when assessed >5 years prior to progression was small (2.8%) . In terms of the level of agreement between pathologists, 84% (21 /25 patients) of LGD were scored by both specialist Gl pathologists on first review, with consensus agreement of all 25 samples upon discussion/re-review. This is comparable with the Dutch consensus study (38)and more favourable than the study by Wani et al. (39) . The difficulty in grading LGD and the uncertainty of the risk of cancer in these patients (38, 39) has led to great debate about the advantages and disadvantages of treating LGD, particularly with the advent of minimally invasive endoscopic therapies (40, 41 ). The complexity of the issue is illustrated by the contrasting conclusions of two recent studies. In a Dutch study biopsies from 147 patients with LGD diagnosed by local hospital pathologists were reviewed by two expert Gl pathologists. The initial diagnosis of LGD by the community pathologist was confirmed by one of the specialist Gl pathologists in only 15% of patients (22/147), and the incidence rate of HGD or EAC in this consensus group was high at 13.4 % (95 % CI 3.5 - 23.2) per annum. In contrast, a study performed in the USA concluded that the risk of progression to HGD and EAC was no different from non-dysplastic BE even when consensus was reached. There are methodological differences which may affect these data, but overall it highlights the clinical challenges in the diagnosis of LGD. The possibility of utilising DNA ploidy and AOL analysis to more objectively determine the LGD patients at highest risk of progression requiring intensive surveillance or intervention is therefore appealing.

Abnormal cyclin A and p53 expression measured by IHC did not confer a significantly increased risk of HGD/EAC progression in multivariate analysis. The presence of TP53 did however show a significant risk in EAC alone (OR = 1.95; 95% CI 1.04-3.67). A previous smaller case-control study from the NIBR of 29 patients who progressed to EAC and 6 patients who progressed to HGD, had shown that TP53 expression was associated with a higher increased risk of progression (OR = 1 1 .7, 95% CI, 1.93 - 71 .4) . [41 ] In a more recent case control study of 54 BE patients, 27 of whom progressed to HGD/EAC and 27 non-progressors, moderate p53 overexpression was also associated with a significantly increased risk of progression with a HR of 6.5 (95% CI, 2.5-1 7.1 ) (42) . The interpretation of these results is limited by inter laboratory variability of TP53 IHC (43) . The gold standard for measurement of p53 mutations is gene sequencing, which is more specific than TP53 IHC, although this is not routinely available and would be technically difficult given the small size and heterogeneity inherent to endoscopic biopsy. p53 expression measured by IHC is a relatively simple technique that is widely available in pathology laboratories, and in our phase 3 study we were able to demonstrate a small but significant risk to cancer progression but this was lost when combined with HGD. The utility of p53, when compared to the other biomarkers in our panel, is therefore likely to be limited. Cyclin A negativity, which has been previously been postulated as a marker of low risk of progression was of borderline significance which dropped out on multivariate analysis. It was interesting that the association with cyclin A was stronger in the 'EAC only' outcomes though, suggesting it may be more useful at identifying patients in the later stages of progression from HGD to cancer.

We have demonstrated that the combination of multiple biomarkers confer a higher risk of progression than any one biomarker alone. Using the combination panel, 67% of cases had 2-4 abnormal markers on the index biopsies. In other words, although the presence of a positive biomarker panel confers a greater risk of progression to HGD/EAC in this study, a third of patients without these abnormalities could still progress. This may be explained by sampling error, both from biopsy at endoscopy and when cutting additional sections from archival FFPE tissue blocks. There may also be a temporal effect if the biopsies at baseline were taken too early before the onset of cancer. When we evaluated the risk of progression in HGD/EAC diagnosed within or after the median time to progression of 3.6 years, there was a trend for AOL to be a better marker of later progression than DNA copy number or LGD. Importantly, on adjusted analysis none of the tests for interaction between markers and follow-up time were significant, as the numbers involved were too small for meaningful analysis. Caution must therefore be exercised in the interpretation of a negative biomarker panel and a reduction of the frequency of endoscopic surveillance on the basis of our panel could not be recommended until these markers are investigated in future Phase 4/5 studies.

With the introduction of minimally invasive endoscopic therapy for the ablation of BE, there is an increasing need to risk stratify patients accurately, as treatment may be offered at an earlier stage. The biomarker panel described here therefore has practical clinical implications and further provides a useful algorithm for risk stratification - i. No dysplasia and panel negative

Risk unknown. Continue current surveillance interval.

ii. No dysplasia and panel positive

Moderate - High risk according to panel. Increase surveillance and consider minimally invasive therapy if >2 positive.

iii. Consensus LGD/IND + any one biomarker positive High risk. Consider minimally invasive therapy.

The combination of dysplasia scoring, DNA ploidy status and AOL may provide significant economic savings as the LGD group, which currently equates to an annual surveillance interval according to AGA guidelines, may be more accurately stratified by biomarkers into a moderate risk group with standard surveillance and higher risk group for ablation. This strategy would benefit from being evaluated in a phase V biomarker study before uptake into routine clinical practice. In conclusion this study has demonstrated that a 7 biomarker panel is a clinically applicable and useful tool for risk stratification in BE. Our results further demonstrate that a reduced panel of 3 biomarkers is advantageous over a clinical model using pathological dysplasia diagnosis, age and sex alone in identifying BE patients at risk of developing HGD/EAC. Our results are consistent with previous longitudinal studies of single biomarkers, including TP53 abnormalities and DNA content abnormalities. That our panel can be undertaken on paraffin embedded tissue and was undertaken on clinical rather than research samples increase its wider application. This panel could be applied to biopsies of BE at diagnosis, has the potential to guide surveillance and therapeutic intervention with expected public health and financial benefit.

Ekample 2

Figure 2 presents a method (algorithm) of the invention. In one embodiment the finding of (no dysplasia, but AOL OR aneuploidy positive) could lead to recommendation to have slightly more frequent surveillance e.g. annual rather than 2-3 yearly. However it must be noted that surveillance intervals are being lengthened in general in most current guidelines to 3-5 years for low risk.

Table 1. Characteristics of cases and matched controls

Characteristic Cases Controls p-value

n=89 (%) n=291 (%)

Gender

Male 67 (75.3) 218 (74.9) 0.94

Female 22 (24.7) 73 (25.1 )

Age at BE diagnosis

( me an + SD, ye ars) 63.8 ± 1 1.9 63.8 ± 1 1.3 0.99

Length BE segment Long 43 (48.3) 141 (48.45)

Short 1 (1 .1 ) 9 (3.1 ) 0.59

Unknown 45 (50.6) 141 (48.45)

Laboratory of origin

Altnagelvin 21 (23.6) 46 ( 15.8)

Antrim 19 (21 .3) 79 (27.1 )

Belfast City 1 1 (12.3) 52 ( 1 7.9) 0.27

Craigavon 7 (7.9) 16 (5.5)

Royal Victoria 31 (34.8) 98 (33.7)

Vienna score of index

BE* 71 (79.8) 284 (97.6) O.001

1 18 (20.2) 7 (2.4)

2-3

*Vienna score 1 : intestinal metaplasia; 2-3: indefinite or low grade dysplasia, as scored by two expert gastrointestinal pathologists.

Tkble 2. Adjusted risk of progressing from EEaccoiding to maikeis inlie initial oesophageal biopsies

Marker Cas Contr TPR FPR Estimat Estimat Cancer + Cancer only es ols (95%CI) (95%CI) ed PPV ed HGD Adjusted

N n NPV Adjusted OR (95% CI)

OR (95% CI)

'Standard' Dysplasia

None 56 262 0.31 0.08 0.08 0.98 1.00 1.00

Indefinite/LGD 25 22 (0.21- (0.05- 22.69 (6.47- 24.86 (5.57-

Unknown 8 7 0.42) 0.12) 79.53) 111.04) p fortre nd 7.30 (2.06- 6.03 (1.47- 25.81) 24.79)

0.001 0.007

'Expert' Dysplasia

None 71 284 0.20 0.02 0.15 0.98 1.00 1.00

Indefinite/LGD 18 7 (0.13- (0.01- 11.78 (4.31- 13.22 (3.67- p fortre nd 0.30) 0.05) 32.18) 47.59)

<0.001 <0.001

DNA copy number

Diploid 44 201 0.44 0.15 0.06 0.99 1.00 1.00

Abnormal0 34 35 (0.33- (0.11- 3.22 (1.73- 3.03 (1.57-5.83) p fortre nd 0.55) 0.21) 6.00) 0.001

<0.001

Cyclin A

Negative/Low % 59 205 0.25 0.16 0.03 0.98 1.00 1.00 posi†iveb 20 39 (0.16- (0.12- 1.32 (0.66- 1.73 (0.81-3.69)

Posifive 0.37) 0.21) 2.66) 0.16 p fortre nd 0.43

p53

0 (None) 18 83 1.00 1.00

1 (Focal) 25 103 1.02 (0.48- 0.87 (0.39-1.91)

2 (Diffuse) 20 62 2.16) 1.21 (0.50-2.94)

3 (intense) 15 30 1.39 (0.63- 3.61 (1.39-9.37) p fortre nd 3.09) 0.02

2-3 v.0-1 0.45 0.33 2.12 (0.87- 1.95 (1.04-3.67)

(0.34- (0.28- 5.16)

0.57) 0.39) 0.08

1.60 (0.91- 2.82)

CA19-9 Lec†inc

0-1 scores 33 151 1.00 1.00 abnormal 34 91 1.44 (0.81- 1.38 (0.75-2.54)

2-3 scores 18 29 2.58) 2.49 (0.98-6.31) abnormal 3.32 (1.48- 0.05

4 scores abnormal 0.61 0.44 0.03 0.99 7.43) 1.56 (0.88-2.76) p fortre nd (0.50- (0.38- 0.004

2-4 υ.0-1 0.71) 0.50) 1.77 (1.04- 3.00)

WGA=

0-1 scores 63 1.00 1.00 abnormal 189 1.43 (0.68- 1.54 (0.70-3.36)

2-3 scores 1 3.01) /

abnormal 9.33 (0.69- 0.36

4 scores abnormal63 0.81 0.75 0.02 0.99 126.21) 1.53 (0.70-3.34) p fortre nd (0.70- (0.69- 0.11

2-4 v.0-1 0.88) 0.80) 1.46 (0.69- 3.06)

CD-15C

0-1 scores 63 186 1.00 1.00 abnormal 14 63 0.50 (0.24- 0.52 (0.24-1.16)

2-3 scores 9 15 1.03) 1.37 (0.48-3.92) abnormal 1.84 (0.69- 0.92

4 scores abnormal63 0.27 0.30 0.02 0.98 4.90) 0.70 (0.36-1.36) p fortre nd (0.18- (0.24- 0.77

2-4 v.0-1 0.38) 0.36) 0.74 (0.41- 1.33)

AOLc

0-1 scores 42 177 1.00 1.00

abnormal 39 61 3.1 7 ( 1 .74- 3.38 ( 1 .74-6.55)

2-3 scores 0 2 5.77) /

abnormal / 0.001

4 scores abnormal 0.48 0.26 0.04 0.99 0.002 3.38 ( 1 .74-6.55)

p fortre nd (0.37- (0.21 - 3.10 ( 1 .71 -

2-4 v. 0-1 0.59) 0.32) 5.63)

Legend for Table 2:

TPR: True Positive Rate, equivalent to sensitivity. FPR: False Positive Rate, equivalent to 1-specificity.

5 PPV: Positive Predictive Value. NPV: Negative Predictive Value. TPR, FPR, PPV and NPV refer to analysis of cancer and HGD cases.

Adjustments: age, gender, year BE diagnosis. All other biomarkers were further adjusted for dysplasia.

aAbnormal includes aneuploid, tetraploid, aneuploid & tetraploid and hypodiploid

bLow % positive defined as <1% positive.

cCA19-9, WGA, CD-15 and AOL all had four components of tissue staining assessed: apical membrane, pan membranous, epithelial mucous globules and

10 epithelial cytoplasm.

Biomarker scores considered abnormal if highest score for apical membrane, epithelial mucous globules or epithelial cytoplasm was 0-<4, or pan membranous was 0.

eBiomarker scores considered abnormal if highest score for apical membrane, epithelial mucous globules or epithelial cytoplasm was 4-12, or pan membranous was >0.

Table 3. Adjusted risk of progressing from BE according to markers in the initial oesophageal biopsies

Individuals without dysplasia at Cancer outcomes only

Full sample (51 cases, 118 controls)

initial biopsy (39 cases, 87 controls) (36 cases, 85 controls)

Variable Full model* Full model* Full model*

Cases Controls Cases Controls Cases Controls

adjusted OR adjusted OR adjusted OR (%) (%) (%) (%) (%) (%)

(95%CI) (95%CI) (95%CI)

Dysplasia 8 (22%) 3 (4%) 6.28 (1.08,

10 (20%) 5 (4%) 2.57 (0.62, 10.72)

36.49)

DNA copy number 16 (44%) 14 (16%) 3.91 (1.35, 11.3

25 (49%) 18 (15%) 4.93 (2.09, 11.61) 15 (38%) 13 (15%) 4.46 (1.69, 11.71)

abnormal

Cyclin A abnormal 12 (24%) 24 (20%) 1.29 (0.48, 3.47) 7 (18%) 22 (25%) 0.81 (0.28, 2.37) 8 (22%) 17 (20%) 1.00 (0.30,3.45

P53 abnormal 21 (41%) 33 (28%) 1.61 (0.72, 3.60) 1 (36%) 24 (28%) 1.84 (0.74, 4.58) 15 (42%) 17 (20%) 3.54 (1.28,9.81

CA199 abnormal 30 (59%) 58 (49%) 1.28 (0.54, 3.01) 21 (54%) 42 (48%) 1.15 (0.44, 2.95) 19 (53%) 45 (53%) 0.61 (0.22, 1.71

WGA abnormal 41 (80%) 91 (77%) 0.95 (0.37, 2.47) 29 (74%) 66 (76%) 0.75 (0.27, 2.07) 29 (81%) 66 (78%) 0.99 (0.30, 3.30

CD15 abnormal 13 (25%) 38 (32%) 0.95 (0.36, 2.49) 11 (29%) 25 (29%) 1.04 (0.37, 2.94) 9 (25%) 27 (32%) 0.63 (0.19,2.12

AOL abnormal 25 (49%) 24 (20%) 3.78 (1.69,8.37) 15 (38%) 17 (20%) 2.97 (1.20, 7.35) 16 (44%) 13 (15%) 4.84 (1.67, 14.0

Model peifoimance c statistic

Basic model† 0.62 (0.53, 0.72) 0.52 (0.41 , 0.63) 0.66 (0.54, 0.77

Basic model plus all biomarkers shown 0.78 (0.71 , 0.86) 0.73 (0.64, 0.82) 0.80 (0.72, 0.88

Internally validated c s†a†is†ic§

Basic model† 0.56 (0.47, 0.66) 0.43 (0.32, 0.54) 0.59 (0.47, 0.70

Basic model plus all biomarkers shown 0.71 (0.64, 0.79) 0.63 (0.53, 0.74) 0.70 (0.62, 0.78

Discrimination slope

Basic model† 0.07 0.00 0.09

Basic model plus all biomarkers shown 0.22 0.13 0.25 IDI 0.15 0.13 0.16

* Full model contains age at diagnosis, sex, year of BE and all biomarkers shown.† Basic model contains dysplasia (except in subgroup excluding dysplasia), age at diagnosis, sex and year of BE. § See methods for more details.

Table 4. Adjusted risk of progressing from BEaccoiding to ma rkers in the initial oesophageal biopsies modelbased upon backwairi selection.

Individuals without dysplasia at Cancer outcomes only

Full samp e (71 cases, 197 controls)

initial biopsy (54 cases, 149 controls) (52 cases, 146 controls)

Variable Reduced model* Reduced model* Reduced mo

Cases Controls Cases Controls Cases Controls

adjusted OR adjusted OR adjusted O

(%) (%) (%) (%) (%) (%)

(95%CI) (95%CI) (95%CI)

Dysplasia 14 (20%) 6 (3%) 4.30 (1 .36, 13.65) 10 ( 19%) 4 (3%) 5.63 ( 1 .43, 22.

DNA copy number 32 (45%) 28 ( 14%) 3.05 ( 1 .61 , 5.82) 19 (34%) 23 ( 15%) 3.25 (1 .55, 6.81 ) 20 (38%) 23 (1 6%) 2.73 (1.24, 6.0

abnormal

AOL abnormal 32(45%) 44(22%) 4.27(2.16,8.43) 22(39%) 32(21%) 2.78(1.38,5.59) 22(42%) 27(18%) 3.13(1.46,6.72

Mode 1 pe dbimanc e c statistic

Basic model† 0.63 (0.54, 0.70) 0.51 (0.42,0.60) 0.64 (0.55,0.73)

Basic model plus AOL and IC DNA 0.75 (0.68,0.81) 0.68 (0.59, 0.77) 0.72 (0.64,0.81)

Internally validated c s†atistic§

Basic model† 0.57 (0.48, 0.64) 0.42 (0.33, 0.51) 0.57 (0.48, 0.66)

Basic model plus AOL and IC DNA 0.69 (0.62, 0.75) 0.57 (0.48, 0.66) 0.68 (0.60,0.77)

Discrimination slope

Basic model† 0.08 0.00 0.09

Basic model plus AOL and IC DNA 0.18 0.12 0.15 IDI 0.10 0.12 0.06

* Reduced model contains age at diagnosis, sex, year of BE, AOL and IC DNA.

† Basic model contains (except in subgroup excluding dysplasia), age at diagnosis, sex and year of BE.

§ See methods for more details.

Table 5. Risk scores of BE individuals scored as abnormal for dysplasia, DNA copy number and AOL.

Risk score count Cases Control Cut off True positive rate False positive rate s

AEEE

0 19 126 >0 versus <0 100% 100%

1 35 64 ≥1 versus <1 73% 36%

2 8 7 ≥2 versus <2 24% 4%

3 9 0 ≥3 versus <3 13% 0%

OR per point increase*= 3.74 (2.43, 5.79), PO.001

No n-dysp hstk EE

0 19 98 >0 versus <0 100% 100%

1 33 47 ≥1 versus <1 66% 34%

2 4 4 ≥2 versus <2 7% 3%

OR per point increase*= 2.99 ( 1 .72, 5.20), PO.001

*Adjus†ed for age, sex and year of

References

1. Vial M, Grande L, Pera M. Epidemiology of adenocarcinoma of the esophagus, gastric cardia, and upper gastric third. Be cent Be suits Cancer Bes.2010;182:1-17.

2. Kadri SR, Lao-Sirieix P, O'Donovan M, et al Acceptability and accuracy of a non-endoscopic screening test for Barrett's oesophagus in primary care: cohort study. Bmj.2010;341:c4372.

3. Group MRCOCW. Surgical resection with or without preoperative

chemotherapy in oesophageal cancer: a randomised controlled trial. lancet.2002 May 18:359(9319):! 727-33.

4. Sikkema M, de Jonge PJ, Steyerberg EW, etal Risk of esophageal

adenocarcinoma and mortality in patients with Barrett's esophagus: a systematic review and meta-analysis. Clin Gastmentemltkpatol.2010 Mar;8(3):235-44; quiz e32.

5. Desai TK, Krishnan K, Samala N, e t al The incidence of oesophageal

adenocarcinoma in non-dysplastic Barrett's oesophagus: a meta-analysis. Gut.2011 Oct 13.

6. Yousef F, Cardwell C, Cantwell MM, etal The incidence of esophageal cancer and high-grade dysplasia in Barrett's esophagus: a systematic review and metaanalysis. Am J Epidemiol 2008 Aug l;168(3):237-49.

7. Reid BJ, Li X, Galipeau PC, etal Barrett's oesophagus and oesophageal adenocarcinoma: time fora new synthesis. NatRe v Cancer.2010 Feb;10(2):87-101.

8. Sullivan Pepe M, Etzioni R, Feng 1, et al Phases of biomarker development for early detection of cancer. JNa tl Cane e rhst.2001 Jul 18;93( 14) : 1054-61.

9. Lao-Sirieix P, Lovat L, Fitzgerald RC. Cyclin A immunocytology as a risk stratificataion tool for Barrett's esophagus surveillance. Clin CancerBe s.2007:13(2) :659- 65.

10. Reid BJ, Levine DS, Longton G, et al Predictors of progression to cancer in Barrett's esophagus: baseline histology and flow cytometry identify low- and high-risk patient subsets. Am JGastmenteml.2000 Jul;95(7):l 669-76.

11. Rabinovitch PS, Longton G, Blount PL, e t al Predictors of progression in Barrett's esophagus III: baseline flow cytometric variables. Am JGastmenteml.2001

Nov;96(ll):3071-83.

12. Dunn JM, Mackenzie GD, Oukrif D, etal Image cytometry accurately detects DNA ploidy abnormalities and predicts late relapse to high-grade dysplasia and adenocarcinoma in Barrett's oesophagus following photodynamic therapy. Br J

Cane e r.2010 May 25:102(11 ) :1608-17.

13. Chao DL, Sanchez CA, Galipeau PC, etal Cell Proliferation, Cell Cycle

Abnormalities, and Cancer Outcome in Patients with Barrett's Esophagus: A Long-term Prospective Study. Clinic alCancerBe search.2008 November 1, 2008:14(21 ):6988-95.

14. Casson AG, Tammemagi M, Eskandarian S, et al p53 alterations in oesophageal cancer: association with clinicopathological features, risk factors, and survival. Mo I

Pathol.1998 Apr;51(2):71-9.

15. Ribeiro U, Finkelstein SD, Safatle-Ribeiro AV, et al p53 sequence analysis predicts treatment response and outcome of patients with esophageal carcinoma. Cancer. 1998;83(1):7-18.

16. Kuroki T, Fujiwara Y, Nakamori S, et al Evidence for the presence of two tumour- suppressor genes for hepatocellular carcinoma on chromosome 13q. Br J Cancer.1995 Aug;72(2):383-5.

17. Jorgensen T, Berner A, Kaalhus O, et al Up-regulation of the oligosaccharide sialyl LewisX: a new prognostic parameter in metastatic prostate cancer. CancerBes. 1995 May 1;55(9):1817-9.

18. Futamura N, Nakamura S, Tatematsu M, et al Clinicopathologic significance of sialyl Le(x) expression in advanced gastric carcinoma. Br J Cancer.2000

Dec;83(12):1681-7.

19. Bird-Lieberman EL, Neves AA, Lao-Sirieix P, etal Molecular imaging using fluorescent lectins permits rapid endoscopic identification of dysplasia in Barrett's esophagus. Nature Med.201 l;In Hess. 20. Murray L, Watson P, Johnston B, etal Risk of adenocarcinoma in Barrett's oesophagus: population based study. Bmj.2003 Sep 6:327(7414) :534-5.

21. Coleman H, Bhat S, Murray I, etal Increasing incidence of Barrett's

oesophagus: a population-based study. Europe an Journal of Epidemiology .

2011:26(9) :739-45.

22. Bhat S, Coleman HG, Yousef F, etal Risk of malignant progression in Barrett's esophagus patients: results from a large population-based study. J Natl Cancer Inst. 2011 Jul 6:103(13): 1049-57.

23. Schlemper RJ, Riddell RH, Kato Y, et al The Vienna classification of

gastrointestinal epithelial neoplasia. Gut.2000 August 1, 2000:47(2) :251-5.

24. Pretorius ME, Waehre H, Abeler VM, et al Large scale genomic instability as an additive prognostic marker in early prostate cancer. CeEOncol.2009:31 (4) :251-9.

25. Bondi J, Pretorius M, Bukholm \, et al Large-scale genomic instability in colon adenocarcinomas and correlation with patient outcome. Ap mis.2009 Oc†;117(10):730- 6.

26. Haroske G, Baak JP, Danielsen H, et al Fourth updated ESACP consensus report on diagnostic DNA image cytometry. AnalCe U Pathol 2001 ;23(2):89-95.

27. Cronin J, McAdam E, Danikas A, et al Epidermal growth factor receptor (EGFR) is overexpressed in high-grade dysplasia and adenocarcinoma of the esophagus and may represent a biomarker of histological progression in Barrett's esophagus (BE). Am J Gastroenterol.2011 Jan;106(l):46-56.

28. Baker SG, Kramer BS, Srivastava S. Markers for early detection of cancer:

statistical guidelines for nested case-control studies. EMC Med Pes Me thodol.2002:2:4.

29. Steyerberg EW. Clinical prediction models: A practical approach to

development, validation and updating: Springer; 2008.

30. Pencina MJ, D'Agostino RB, Sr., D'Agostino RB, Jr., etal Evaluating the added predictive ability of a new marker: from area under the ROC curve to reclassification and beyond. Stat Med.2008 Jan 30:27(2): 157-72; discussion 207-12.

31. Steyerberg EW, Harrell FE, Jr., Borsboom GJ, et al Internal validation of predictive models: efficiency of some procedures for logistic regression analysis. J Clin Epidembl.200] Aug;54(8):774-81.

32. Galipeau PC, Li X, Blount PL, e t al NSAIDs modulate CDKN2A, TP53, and DNA content risk for progression to esophageal adenocarcinoma. PL·SMed.2007

Feb;4(2):e67.

33. Schulmann K, Sterian A, Berki A, etal Inactivation of pi 6, RUNX3, and HPP1 occurs early in Barrett's-associated neoplastic progression and predicts progression risk. Oncogene .2005 Jun 9:24(25) :4138-48.

34. Reid BJ, Prevo LJ, Galipeau PC, et al Predictors of progression in Barrett's esophagus II: baseline 17p (p53) loss of heterozygosity identifies a patient subset at increased risk for neoplastic progression. Am J Gastr enterol 2001 Oc†;96(10):2839-48.

35. Reid BJ, Blount PL, Rubin CE, etal Flow-cytometric and histological progression to malignancy in Barrett's esophagus: prospective endoscopic surveillance of a cohort. Gastroenterology.1992 Apr;102(4 Pt l):1212-9.

36. Paulson TG, Maley CC, Li X, etal Chromosomal instability and copy number alterations in Barrett's esophagus and esophageal adenocarcinoma. Clin CancerPe s.

2009 May 15;15(10):3305-14.

37. Bird-Lieberman EL, Dunn JM, P. L-S, etal Phase 2 and phase 3 multicentre studies demonstrate the potential forglycans as predictive biomarkers in Barrett's oesophagus Gut.2011 ;60(Suppl I):A169-70.

38. Hvid-Jensen F, Pedersen L, Drewes AM, etal Incidence of Adenocarcinoma among Patients with Barrett's Esophagus. NEnglJMed.2011 Oct 13:365(15): 1375-83.

39. Wani S, Falk GW, Post J, etal Risk Factors for Progression of Low-Grade Dysplasia in Patients With Barrett's Esophagus. Gastroenterology.2011 Oc†;141(4):l 179-86 el.

40. Bulsiewicz WJ, Shaheen NJ. The role of radiofrequency ablation in the management of Barrett's esophagus. Gastrointest Endosc Clin N Am.2011 Jan;21(l ):95- 109. 41 . Spechler SJ, Sharma P, Souza RF, e t al American Gastroenterological

Association medical position statement on the management of Barrett's esophagus. Gastme nte m bgy. 201 1 Mar;140(3) : 1084-91 .

42. Sikkema M, de Jonge PJ, Steyerberg EW, e t al Risk of Esophageal

Adenocarcinoma and Mortality in Patients With Barrett's Esophagus: A Systematic Review and Meta-Analysis. Clin G a stro e nte ra l Hepato l. 2009 Oct 19.

43. Alsner J, Jensen V, Kyndi M, e t al A comparison between p53 accumulation determined by immunohistochemistry and TP53 mutations as prognostic variables in tumours from breast cancer patients. Ac ta Onc o l. 2008:47(4) :600-7.

All publications mentioned in the above specification are herein incorporated by reference. Various modifications and variations of the described aspects and embodiments of the present invention will be apparent to those skilled in the art without departing from the scope of the present invention. Although the present invention has been described in connection with specific preferred embodiments, it should be understood that the invention as claimed should not be unduly limited to such specific embodiments. Indeed, various modifications of the described modes for carrying out the invention which are apparent to those skilled in the art are intended to be within the scope of the following claims.

CI AIMS

1. A method for aiding prediction of the likelihood of progression from Barrett's esophagus to high grade dysplasia or esophageal adenocarcinoma in a subject, the method comprising

(a) providing an oesophagal sample from said subject

(b) determining if said sample stains abnormally with Aspe rgiEus oryzae lectin;

(c) determining if there is a DNA content abnormality in said sample; and

(d) determining if there is low grade dysplasia in said sample;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, then an increased likelihood of progression is determined.

2. A method according to claim 1 further comprising the step of

(e) determining if said sample stains abnormally for CA 19-9;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (e) is abnormal then a further increased likelihood of progression is determined.

3. A method according to claim 1 or claim 2 further comprising the step of

(f) determining if said sample stains abnormally for Cyclin A;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (f) is abnormal then a further increased likelihood of progression is determined.

4. A method according to any of claims 1 to 3 further comprising the step of

(g) determining if said sample stains abnormally for p53;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (g) is abnormal then a further increased likelihood of progression is determined.

5. A method according to any of claims 1 to 4 further comprising the step of

(h) determining if said sample stains abnormally with Wheat Germ Agglutinin lectin; wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (h) is abnormal then a further increased likelihood of progression is determined.

6. A method according to claim 1 further comprising the steps of

(f) determining if said sample stains abnormally for Cyclin A;

(g) determining if said sample stains abnormally for p53;

(h) determining if said sample stains abnormally with Wheat Germ Agglutinin lectin; wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (f) is abnormal and (g) is abnormal and (h) is abnormal then a further increased likelihood of progression is determined.

7. A method according to claim 2 further comprising the steps of

(f) determining if said sample stains abnormally for Cyclin A;

(g) determining if said sample stains abnormally for p53;

(h) determining if said sample stains abnormally with Wheat Germ Agglutinin lectin; wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, and if (e) is abnormal and (f) is abnormal and (g) is abnormal and (h) is abnormal then a further increased likelihood of progression is determined.

8. An assay for selecting a treatment regimen, said assay comprising

(a) providing an oesophagal sample from said subject

(b) determining if said sample stains abnormally with Aspe rgillus oryzae lectin;

(c) determining if there is a DNA content abnormality in said sample; and

(d) determining if there is low grade dysplasia in said sample;

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, then a treatment regimen of increased surveillance is selected. 9. An apparatus or system which is

(a) configured to analyse an oesophagal sample from a subject, wherein said analysis comprises

(b) determining if said sample stains abnormally with Aspe rgillus oryzae lectin;

(c) determining if there is a DNA content abnormality in said sample; and

(d) determining if there is low grade dysplasia in said sample;

said apparatus or system comprising an output module,

wherein if (b) is abnormal and (c) is abnormal and low grade dysplasia is present, then said output module indicates an increased likelihood of progression to EAC for said subject.

10. A method, assay, apparatus or system according to any preceding claim wherein said sample comprises formalin fixed paraffin embedded material.

1 1 . A method, assay, apparatus or system according to any preceding claim wherein step (b) comprises determining the level of staining by Aspe rgillus oryzae lectin in said sample by grading the intensity of the stain and the percentage of the area stained at said intensity, and producing an H score for said stain.

12. A method, assay, apparatus or system according to any preceding claim wherein step (c) comprises determining the DNA content by image cytometry DNA analysis and inferring from said determination whether said DNA content is abnormal. 13. A method, assay, apparatus or system according to any preceding claim wherein step (d) comprises histological examination of the sample, and scoring the presence or absence of low grade dysplasia using the Vienna scale.

14. A method, assay, apparatus or system according to any preceding claim wherein step (e) comprises immunohistochemical staining of the sample for CA19-9, grading the intensity of the stain and the percentage of the area stained at said intensity, and producing an H score for said stain.

15. A method, assay, apparatus or system according to any preceding claim wherein step (f) comprises immunohistochemical staining of the sample for Cyclin A, determining the percentage of Cyclin A positive epithelial surface cells compared to Cyclin A negative epithelial surface cells, wherein a percentage of 1 % or more Cyclin A positive epithelial surface cells is an abnormal stain. 16. A method, assay, apparatus or system according to any preceding claim wherein step (g) comprises immunohistochemical staining of the sample for p53, and scoring the stain as normal or abnormal wherein either a strong, dark stain compared to background stain or an absence of stain compared to background stain is an abnormal stain.

1 7. A method, assay, apparatus or system according to any preceding claim wherein step (h) comprises determining the level of staining by Wheat Germ Agglutinin lectin in said sample by grading the intensity of the stain and the percentage of the area stained at said intensity, and producing an H score for said stain.

18. Use for prognostic or risk prediction applications relating to high grade dysplasia or esophageal adenocarcinoma, of a material which recognises, binds to or has affinity for a polypeptide selected from p53, Cyclin A, or CA19-9, or a fragment, variant or mutant thereof.

19. Use according to claim 18 of a combination of materials, each of which respectively recognises, binds to or has affinity for one or more of said polypeptide(s), or a fragment, variant or mutant thereof.

20. Use for prognostic or risk prediction applications relating to high grade dysplasia or esophageal adenocarcinoma, of Aspe rgiEus oryzae lectin or Wheat Germ Agglutinin lectin.

21 . Use for prognostic or risk prediction applications relating to high grade dysplasia or esophageal adenocarcinoma, of image cytometry DNA analysis.

22. An assay device comprising a material which binds to or has affinity for a polypeptide selected from p53, Cyclin A, or CA19-9 and which further comprises Aspe reikis o ryzae lectin and/or Wheat Germ Agglutinin lectin.

23. Use according to claim 20 or a device according to claim 22 wherein said lectin comprises a detectable label.

24. Use according to claim 18 or claim 19 or a device according to claim 22 wherein said material comprises an antibody or fragment thereof.

25. A method, assay, apparatus or system substantially as described herein.

Download Citation


Sign in to the Lens

Feedback