WO2018158589A1

WO2018158589A1 - Diagnostic and prognostic methods

Info

Publication number: WO2018158589A1
Application number: PCT/GB2018/050544
Authority: WO
Inventors: Andrew FEBER; Timothy Fenton; Kerry Chester
Original assignee: UCL Business Ltd
Current assignee: UCL Business Ltd
Priority date: 2017-03-02
Filing date: 2018-03-02
Publication date: 2018-09-07
Anticipated expiration: 2019-09-02
Also published as: GB201703406D0

Abstract

The present invention relates to methods of diagnosing cervical cancer in an individual, involving determining the methylation status of Methylation Variable Positions (MVPs) in DNA from a sample from the individual and providing a diagnosis based on methylation status data. The invention also relates to methods of treating cervical cancer in an individual comprising providing a diagnosis of cervical cancer by the diagnostic methods defined herein, followed by administering one or more anti¬ cancer agents to the individual. The present invention also relates to methods of classifying a cervical cancer tumor from a patient having cervical cancer as being either a group 1 tumor or a group 2 tumor, wherein patients having a tumor assigned to group 2 are at risk of a poorer outcome following primary treatment regimes compared to patients having a tumor assigned to group 1. Such methods allow more efficient patient risk stratification for the purposes of providing further treatment and/or to help plan and manage patient care. The invention also relates to methylation-discriminatory arrays comprising probes directed to the MVPs defined herein, as well as kits comprising the arrays.

Description

DIAGNOSTIC AND PROGNOSTIC METHODS

Field of the Invention

The present invention relates to methods of diagnosing cervical cancer in an individual, involving determining the methylation status of Methyl ation Variable

Positions (MVPs) in DNA from a sample from the individual and providing a diagnosis based on methylation status data. The invention also relates to methods of treating cervical cancer in an individual comprising providing a diagnosis of cervical cancer by the diagnostic methods defined herein, followed by administering one or more anti- cancer agents to the individual.

The present invention also relates to methods of classifying a cervical cancer tumor from a patient having cervical cancer as being either a group 1 tumor or a group 2 tumor, wherein patients having a tumor assigned to group 2 are at risk of a poorer outcome following primary treatment regimes compared to patients having a tumor assigned to group 1. Such methods allow more efficient patient risk stratification for the purposes of providing further treatment and/or to help plan and manage patient care.

The invention also relates to methylation-discriminatory arrays comprising probes directed to the MVPs defined herein, as well as kits comprising the arrays. Background to the Invention

Cervical cancer has been estimated as one of the leading causes of death from cancer among women worldwide [1], with an estimated 527,000 new cases and 265,000 deaths in 2012 [2]. In Europe, there were 58,400 cases diagnosed and 24,400 deaths in 2012 [2]. In the United States, 12200 new diagnoses and 4200 cancer deaths were reported in 2012 [1].

Consistent with the incidence of the disease, the global market size for cervical cancer screening (Papanicolaou, or "Pap" test) and HPV DNA testing has recently been valued at over $3.2 billion per year, and is predicted to reach $5.6 billion per year by 2024 [3].

Cervical cancer develops from pre-cancerous non-invasive lesions termed cervical intraepithelial neoplasias (CINs) which are scaled according to the severity of the lesion. CINl is classified as a mild dysplasia, CIN2 is classified as a moderate dysplasia and CIN3 is classified as a severe dysplasia.

Existing methods for the diagnosis of cervical cancer are based on cytology (the Papanicolaou, or "Pap" smear test), and more recently on detection of high-risk human papillomavirus (HPV) DNA [4] (e.g. using the "Cobas" test marketed by Roche). The Pap test requires examination of cells by light microscope and characterisation of cells on the basis of morphology. Consquently, the test has inherent limitations in terms of sensitivity and specificity.

Data from large randomised trials indicate the use of HPV testing as the primary screening method at 5-year intervals in women over 30, as it was shown to reduce incidence of cervical cancer by 60-70% when compared with cytology [4,5].

In younger women however, where HPV infection is much more prevalent, HPV testing leads to over-diagnosis and therefore unnecessary treatment of CIN lesions that would naturally regress within two years and never progress to cervical cancer [5, 6]. Conversely, false-negative HPV results from CIN3+ lesions have also been reported. In one such study 5% of CIN3+ lesions, 25% of which progressed to cervical cancer, were negative by HPV testing [7]. Thus superior methods for diagnosing cervical cancer, in terms of improvments to both the specificity and sensitivity of cervical cancer screening tests, are required.

Following diagnosis, cervical cancers are staged according to various criteria including the size of the primary tumour, whether it has invaded surrounding organs or tissue and the presence of local or distant metastases.

Although early stage cervical cancers are treated with curative intent, 10-20% of patients with stage IB-IIA tumours with no apparent lymph node involvement will suffer recurrence following primary surgery or radiotherapy [8].

At present, the gold standard for estimating prognosis in cervical cancer is staging by physical examination and imaging (by CT or, less commonly, by MRI).

However, improved methods for estimating prognosis are needed, in particular to identify those patients with early stage cervical cancer that are at risk of a poor outcome. Improved methods to determine which patients may need more intensive primary treatment, and potentially sparing treatment-associated morbidities in those patients having a low risk of recurrence would be of significant value in managing the care of diagnosed cervical cancer patients.

Several studies have now shown the potential utility of DNA methylation biomarkers in body fluids, including urine [9-17], plasma/serum [18-20], and sputum [21, 22], for the non-invasive detection of cancer. Changes in DNA methylation play a key role in malignant transformation, leading to the silencing of tumor-suppressor genes and overexpression of oncogenes [23]. The ontogenic plasticity and relative stability of DNA methylation makes epigenetic changes ideal biomarkers for diagnosis.

More specifically in terms of cervical cancer, methylation status of genes has been examined as a possible approach to the study of the progression of the disease. For example, Virmani et al [24] examined the methylation status of six genes (pi 6, RARb, FHIT, GSTPl, MGMT, and hMLHl). The authors determined that methylation of RARb and GSTPl were early events in disease progresion, methylation of pl6 and MGMT were intermediate events, and methylation of FHIT was a late, tumor-associated event.

More recently, assessment of the levels of Keratin 17 (KRT17) expression has been proposed as a prognostic biomarker approach [25].

Nevertheless, despite prior studies there remains a need for more accurate and improved diagnostic and prognostic assays and tests for cervical cancer based on biomarker analysis.

Summary of the Invention

Diagnostic methods are provided for the detection of cervical cancer from a biological sample from an individual including, but not limited to, a tissue sample, with high sensitivity and specificity, and which have the potential to reduce the need for smear test in patients undergoing screening for disease.

Thus the invention provides a method of diagnosing cervical cancer in an individual comprising:

(a) providing DNA from a sample from the individual;

(b) determining whether each one of a group of MVPs selected from a panel comprising the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG] is methylated, wherein the group comprises at least 10 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG]; and (c) diagnosing cervical cancer in the individual when the at least 10 MVPs of the group of (b) are methylated.

The group of MVPs may comprise at least 20 of the MVPs identified in SEQ ID

NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when at least 10 of those MVPs are methylated.

The group of MVPs may comprise at least 20 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when at least 15 of those MVPs are methylated.

The group of MVPs may comprise at least 20 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when all 20 of those MVPs are methylated.

The MVPs determined to be methylated may include the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG] and wherein cervical cancer is diagnosed when each one of the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG] is methylated.

The group of MVPs may comprise all 30 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when at least 15 of those MVPs are methylated. The MVPs determined to be methylated may include the MVPs identified in SEQ ID NOS 1 to 15 and denoted by [CG] and wherein cervical cancer is diagnosed when each one of the MVPs identified in SEQ ID NOS 1 to 15 and denoted by [CG] is methylated.

The group of MVPs may comprise all 30 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when at least 20 of those MVPs are methylated. The group of MVPs may comprise all 30 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when all 30 of those MVPs are methylated.

The MVPs determined to be methylated may include the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG]. Cervical cancer may be diagnosed when each one of the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG] is methylated.

The present invention also relates to methods of diagnosing cervical cancer in an individual comprising:

(a) obtaining data which identify whether each one of a group of MVPs

selected from a panel comprising the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG] is methylated, wherein the group comprises at least 20 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG]; and

(b) diagnosing cervical cancer in the individual when at least 20 MVPs of the group of (a) are methylated;

wherein the data were obtained by a method comprising:

i. obtaining DNA from the sample; and

ii. determining whether MVPs are methylated in the DNA.

The present invention also relates to methods of diagnosing cervical cancer in an individual wherein the step of diagnosing cervical cancer in the individual further comprises:

I. stratifying the stage of the tumor; and/or

II. determining the risk of recurrence of the tumor following primary

surgery or radiotherapy; and/or

III. determining the risk of progression to invasive disease; and/or

IV. determining the likely response to treatment therapy; and/or

V. providing a cervical cancer treatment to the patient.

The present invention also relates to methods of classifying a cervical cancer tumor from a patient having cervical cancer as being either a group 1 tumor or a group 2 tumor, wherein patients having a tumor assigned to group 2 are at risk of a poorer outcome following primary treatment regimes compared to patients having a tumor assigned to group 1.

Such methods are based on the detection of differential DNA methylation patterns in cervical cancers from patients that experience either better, or worse than expected survival outcomes despite all being diagnosed with early stage tumours and despite all initially receiving uniform treatment.

Such methods allow more efficient patient risk stratification for the purposes of providing further treatment and/or to help plan and manage patient care. For example, such methods provide increased accuracy for determining the prognosis of patients following surgery to remove their primary tumour.

The prognostic methods and related aspects described herein yield more information and outperform clinical staging. These methods outperform both HPV typing and histological subtyping (i.e squamous versus adenocarcinoma).

Using the methods of the invention it is possible to divide patient tumors into two groups, wherein patients having tumors assigned to different groups will be expected to have dramatically different outcomes during the course of futher treatment.

Thus the invention provides a method of classifying a cervical tumor from a patient having cervical cancer as being either a group 1 tumor or a group 2 tumor, the method comprising:

(a) providing a DNA sample from the patient;

(b) determining whether each one of a group of MVPs selected from a panel

comprising the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG] is methylated or unmethylated, thereby providing a methylation dataset for the group of MVPs in the DNA sample;

(c) applying a binary classification algorithm to the dataset; and

(d) classifying the tumor as either a group 1 tumor or a group 2 tumor based on the results of the classification algorithm.

In such methods the group of MVPs may comprise at least 20 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG]. The group of MVPs may comprise at least 40 of the MVPs, or at least 60 of the MVPs, or at least 80 of the MVPs, or at least 100 of the MVPs, or at least 120 of the MVPs, or at least 140 of the MVPs, or all 145 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

In any of the classification methods of the invention the classification algorithm may be based on a support vector machine model (SVM), a k-nearest neighbours model (KNN), a GLMnet model (GLM), or a Random Forest model (RF). In any of the classification algorithms used, the output is the probability (from 0 to 1) of a given tumour belonging to group 2, where 0 is a 100% probability that the sample belongs to group 1, and 1 is a 100% probability that the sample belongs to group 2, and wherein any sample with a probability of equal to or greater than 0.5 is designated as belonging to group 2 and any sample with a probability of than 0.5 is designated as belonging to group 1.

The prognostic methods and related aspects described herein may also be applied to samples from individuals suspected to have cervical cancer wherein the samples are analysed for diagnostic purposes. Such methods will allow simultaneous diagnosis and patient risk stratification for the purposes of providing further treatment and/or to help plan and manage patient care.

Thus the invention additionally provides methods of managing the treatment or healthcare regime of an individual, comprising:

(a) diagnosing cervical cancer in an individual by performing any of the

diagnostic methods of the invention; and

(b) stratifying the patient into a cervical cancer treatment or healthcare regime group according to whether the patient has either a group 1 tumor or a group 2 tumor by performing any of the classification methods of the invention. The invention additionally provides methods of treating a cervical cancer patient, the methods comprising:

(a) stratifying the patient into a cervical cancer treatment regime group

according to whether the patient has either a group 1 tumor or a group 2 tumor by performing any of the classification methods of the invention; and (b) treating the patient with one or more cervical cancer treatments.

The invention additionally provides methods of treating cervical cancer in a patient comprising:

(a) obtaining DNA from a sample from the patient;

(b) diagnosing cervical cancer in the patient by performing any of the diagnosic methods described herein; (c) stratifying the patient into a cervical cancer treatment regime group according to whether the patient has either a group 1 tumor or a group 2 tumor by performing any of the classification methods described herein; and

(d) administering one or more cervical cancer treatments to the patient.

In any of the prognostic and diagnostic methods described herein, the step of determining whether each one the MVPs is methylated may comprise bisulphite converting the DNA.

In any of the prognostic and diagnostic methods described herein, the step of determining whether each one the MVPs is methylated may comprise:

1) performing a sequencing step to determine the sequence of MVPs;

2) hybridising DNA to an array comprising probes capable of discriminating

between methylated and non-methylated forms of MVPs and applying a detection system to the array to discriminate methylated and non-methylated forms of the MVPs; or

3) performing an amplification step using methylation-specific primers, wherein the status of an MVP as methylated or non-methylated is determined by the presence or absence of an amplified product.

Before the sequencing or hybridization steps, an amplification step may be performed, wherein loci comprising each MVP are amplified. Amplification may be performed by PCR.

A capturing step may be performed before the sequencing or hybridization steps. The capturing step may involve binding polynucleotides comprising the MVP loci to binding molecules specific to the MVP loci and collecting complexes comprising MVP loci and binding molecules; and wherein:

i. the capturing step occurs before the step of bisulphite converting the DNA;

ii. the capturing step occurs after the step of bisulphite converting the DNA but before the amplification or hybridization steps; or

iii. the capturing step occurs after the step of bisulphite converting the DNA and after the amplification step. The binding molecules may be oligonucleotides specific for each MVP, preferably DNA or RNA molecules each comprising a sequence which is

complementary to the corresponding MVP.

The binding molecule may be coupled to a purification moiety.

The purification moiety may comprise a first purification moiety and the step of collecting complexes comprising MVP loci and binding molecules may comprise binding the first purification moiety to substrates comprising a second purification moiety, wherein first and second purification moieties form an interaction complex.

The first purification moiety may be biotin and the second purification moiety may be streptavidin; or the first purification moiety may be streptavidin and the second purification moiety may be biotin.

The step of amplifying loci comprising MVPs may comprise the use of primers which are independent of the methylation status of the MVP.

The step of amplifying loci comprising MVPs may be performed by

microdroplet PCR amplification.

The invention also relates to methylation-discriminatory arrays comprising probes directed to the MVPs defined herein and kits comprising the arrays.

Thus the invention provides an array capable of discriminating between methylated and non-methylated forms of MVPs; the array comprising oligonucleotide probes specific for a methylated form of each MVP in a MVP panel and oligonucleotide probes specific for a non-methylated form of each MVP in the panel; wherein the panel consists of at least 10 MVPs selected from the MVPs identified in SEQ ID NOS 1 to 30, or wherein the panel consists of at least 20 MVPs selected from the MVPs identified in SEQ ID NOS 31 to 175.

The invention also provides kits comprising any of the arrays described herein.

The kits may further comprise a DNA modifying regent that is capable of modifying a non-methylated cytosine in a MVP dinucleotide but is not capable of modifying a methylated cytosine in a MVP dinucleotide, optionally wherein the dinucleotide is CpG. Any of the kits described herein may further comprise forward and reverse primers for amplifying any and all of the MVPs defined herein from a sample of DNA from a patient. Brief description of the Figures

Figure 1. Figure 1 shows standard ROC curves for the 30 MVP signature (all 30 of SEQ ID NOs 1 to 30) in four different training models: machine model (SVM), a k- nearest neighbours model (KNN), a GLMnet model (GLM), and a Random Forest model (RF). Performance was assessed in terms of sensitivity, specificity, and Kappa values in the validation dataset.

Figure 2. Figure 2 shows separate statistical metrics of sensitivity, specificity, and Kappa values for the 30 MVP signature (all 30 of SEQ ID NOs 1 to 30) in the validation dataset for each each of the four different training models: machine model (SVM), a k-nearest neighbours model (KNN), a GLMnet model (GLM), and a Random Forest model (RF).

Figure 3. Figure 3 shows standard ROC curves for a subset of the 30 MVP signature (all 20 of SEQ ID NOs 1 to 20) in the four different training models: machine model (SVM), a k-nearest neighbours model (KNN), a GLMnet model (GLM), and a Random Forest model (RF). Performance was assessed in terms of sensitivity, specificity, and Kappa values in the validation dataset.

Figure 4. Figure 4 shows separate statistical metrics of sensitivity, specificity, and Kappa values for the 20 MVP subset signature (all 20 of SEQ ID NOs 1 to 20) in the validation dataset for each each of the four different training models: machine model (SVM), a k-nearest neighbours model (KNN), a GLMnet model (GLM), and a Random Forest model (RF).

Figure 5. Figure 5 shows a heatmap showing the 30 probes (all 30 of SEQ ID NOs 1 to 30) that display hypomethylation in normal cervical tissue and

hypermethylation in cancer samples.

Figure 6. Figure 6 shows a heatmap showing stratification of TCGA cervical cancer samples into two groups 'HPV16-like' and 'HPV45-like', on the basis of the pattern of DNA methylation at the 145 informative MVP loci (SEQ ID NOs 31 to 175).

Figure 7. Figure 7 shows Kaplan-Meier curves showing overall survival of patients with tumours classified as HPV16-like or HPV45-like in TCGA (discovery) or Norwegian (validation) cohorts. Figure 8. Figure 8 shows shows a Kaplan-Meier curve showing overall survival of patients with tumours classified as FIPV16-like or FIPV45-like on the basis of all 145 MVP loci (SEQ ID NOs 31 to 175).

Figure 9. Figure 9 shows shows a Kaplan-Meier curve showing overall survival of patients with tumours classified as FIPV16-like or FIPV45-like on the basis of a subset of the top 20 MVPs out of the total group of 145 MVP loci (i.e. based on MVPs defined by SEQ ID NOs 31 to 50).

Figure 10. Figure 10 shows histograms displaying the classifier-derived probability that each sample derives from a group 1 tumour (probability of group 2 membership = 0) or group 2 tumour (probability of group 2 membership = 1).

Histograms are shown for each of two independent validation cohorts.

Detailed Description of the Invention Cervical cancer

As discussed above, cervical cancer represents one of the most prevalent groups of cancers in women. Infection by human papillomavirus (HPV) is responsible for around 90% of cervical cancer cases. Consistent with this, HPV vaccines have recently been introduced into public healthcare programmes, and HPV screening tests are becoming routine in the diagnosis of the disease. Other risk factors include smoking, prolonged use of oral contraceptives and a weakened immune system [26].

90% of cervical cancer cases are referred to as squamous cell carcinomas, 10% are adenocarcinomas, and a smaller proportion are other types [26].

Once diagnosed, cervical cancer is staged according to the International Federation of Gynecology and Obstetrics (FIGO) staging convention. The disease is staged as stage 0 (carcinoma in situ) followed by stages IA, IB, IIA, IIB, IDA, IIIB, IVA and IVB according to severity and degree of metastasis [26].

The diagnostic, prognostic and treatment methods described herein are capable of positively identifying malignant cells of all classifications and stages of cervical cancer. Thus, any of the methods described herein may be used to diagnose cervical squamous cell carcinomas and cervical adenocarcinomas. The most preferred patient type to which the diagnostic assays described herein are applicable are humans. The diagnostic assays described herein may also be used to identify cervical cancer in a non-human animal. For example, non-human animals may contain tissue derived from humans, e.g. xenografts. Thus, diagnostic assays may be used to diagnose human cervical cancer in an animal model of human cervical cancer. Typical non-human animals to which the diagnostic assays described herein are applicable are rodents such as rats or mice.

Methylation Variable Positions (MVPs)

Methylation of DNA is a recognised form of epigenetic modification which has the capability of altering the expression of genes and other elements such as

microRNAs [27]. In cancer development and progression, methylation may have the effect of e.g. silencing tumor suppressor genes and/or increasing the expression of oncogenes. Other forms of dysregulation may occur as a result of methylation.

Methylation of DNA occurs at discrete loci which are predominately dinucleotide consisting of a CpG motif, but may also occur at CHH motifs (where H is A, C, or T).

During methylation, a methyl group is added to the fifth carbon of cytosine bases to create methylcytosine.

Methylation can occur throughout the genome and is not limited to regions with respect to an expressed sequence such as a gene. Methylation typically, but not always, occurs in a promoter or other regulatory region of an expressed sequence.

A Methylation Variable Position (MVP) as defined herein is any dinucleotide locus which may show a variation in its methylation status between phenotypes, i.e. between tumour and normal tissue. An MVP is preferably a CpG or a CHH

dinucleotide motif. An MVP as defined herein is not limited to the position of the locus with respect to a corresponding expressed sequence.

Typically, an assessment of DNA methylation status involves analysing the presence or absence of methyl groups in DNA, for example methyl groups on the 5^th position of one or more cytosine nucleotides. Preferably, the methylation status of one or more cytosine nucleotides present as a CpG dinucleotide (where C stands for Cytosine, G for Guanine and p for the phosphate group attached to the backbone between the two) is assessed.

Methylation analysis

For the purposes of methods for diagnosing cervical cancer as described herein and methods for classifying cervical tumors as described herein, by assessing the methylation status of an MVP or determining whether an MVP is methylated or unmethylated it is meant that a determination is made as to whether an MVP was methylated or unmethylated in the starting sample of DNA obtained from the individual prior to subsequent processing.

A panel of 30 MVPs whose methylation status varies as between normal tissue and cervical cancer tissue is provided herein. These 30 MVPs are set out in Table 1 below. For each of the 30 MVPs the CpG dinucleotide, of which the cytosine is subject to differential methylation, is indicated within square brackets.

Thus for the purposes of methods for diagnosing cervical cancer, "assessing the methylation status of an MVP" or "determining whether an MVP is methylated" it is meant that a determination is made as to whether the cytosine of the CpG of the MVP is methylated or unmethylated in the starting sample of DNA obtained from the individual.

For the purposes of methods for diagnosing cervical cancer, an MVP is herein defined as methylated if one or more alleles of that MVP in a sample of genomic DNA from the patient is determined to possess one or more methylated CpG dinucleotide loci.

In addition, a further panel of 145 MVPs is provided herein. These 145 MVPs are set out in Table 2 below (SEQ ID NOS 31 to 175). For each of these 145 MVPs the CpG dinucleotide, of which the cytosine is subject to differential methylation, is indicated within square brackets. These 145 MVPs and subsets thereof may be used for classifying a cervical tumor from a patient having cervical cancer as being either a group 1 tumor or a group 2 tumor as discussed in more detail herein.

Thus for the purposes of methods for classifying a cervical tumor by "assessing the methylation status of an MVP" or "determining whether an MVP is methylated" it is meant that a determination is made as to whether the cytosine of the CpG of the MVP is methylated or unmethylated in the starting sample of DNA obtained from the individual.

For the purposes of methods for classifying a cervical tumor, an MVP is herein defined as methylated if one or more alleles of that MVP in a sample of genomic DNA from the patient is determined to possess one or more methylated CpG dinucleotide loci.

For the purposes of methods for classifying a cervical tumor, once a

determination is made as to the methylation staus at a given MVP locus, a methylation dataset is generated based on the methylation status of all MVPs analysed. A binary classification algorithm is then applied to the dataset to determine whether a cervical tumor from a patient having cervical cancer is either a group 1 tumor or a group 2 tumor.

It is to be appreciated that for each MVP the precise genomic sequence upstream and downstream of the CpG site may vary as between individuals. It is therefore to be understood that "assessing the methylation status of an MVP" or "determining whether an MVP is methylated" as described herein does not require a perfect match between the genomic sequence upstream and downstream of a given CpG site for a given individual and the sequence upstream and downstream of the corresponding CpG site as listed in the sequences herein defined by SEQ ID NO. Rather, it is to be understood that "assessing the methylation status of an MVP" or "determining whether an MVP is methylated" as described herein simply requires a determination as to whether the cytosine of the CpG motif at the genomic locus for that MVP is methylated or unmethylated.

In any of the methods described herein, the MVPs determined to be methylated are methylated relative to normal epithelium control and/or whole blood control.

Binary classification algorithms for use in methods for classifying cervical tumors

Binary classification is a method of classifying samples or patients into two groups on the basis of a classification rule. Instancing a decision whether an item has or has not some quantitative property or some specified characteristic. Several binary classification algorthyms exist, these include, but are not limted to Logistic regression, Support vector machine (SMV), Relevance vector machine (RVM), Perceptron, Naive Bayes classifier, k-nearest neighbors algorithm (KNN), Artificial neural network, Random Forests (RF).

The output of most binary classification algorithms is a continuous probability score. The score indicates the algorithm's certainty that the given observation belongs to the positive class. The decision about whether the observation should be classified as a binary positive or negative is determined by a classification threshold (cut-off). Any observations with scores higher than the threshold are then predicted as the positive class and scores lower than the threshold are predicted as the negative class.

Predictions fall into four groups based on the actual known answer and the predicted answer: true positives, true negatives, false positives and false negatives. In a perfect binary classifier the number of true positives will be 100% (1.0) and the number of false positives will be 0% (0.0).

Identification and assessment of Methyl ation Variable Position (MVP) status

A variety of techniques are available for the identification and assessment of Methylation Variable Positions (MVPs), as will be outlined briefly below. The diagnostic and prognostic methods described herein encompass any suitable technique for the determination of MVP status.

Methyl groups are lost from a starting DNA molecule during conventional in vitro handling steps such as PCR. To avoid this, techniques for the detection of methyl groups commonly involve the preliminary treatment of DNA prior to subsequent processing, in a way that preserves the methylation status information of the original DNA molecule. Such preliminary techniques involve three main categories of processing, i.e. bisulphite modification, restriction enzyme digestion and affinity-based analysis. Products of these techniques can then be coupled with sequencing or array- based platforms for subsequent identification or qualitative assessment of MVP methylation status.

Techniques involving bisulphite modification of DNA have become the most common methods for detection and assessment of methylation status of CpG dinucleotide. Treatment of DNA with bisulphite, e.g. sodium bisulphite, converts cytosine bases to uracil bases, but has no effect on 5-methylcytosines. Thus, the presence of a cytosine in bisulphite-treated DNA is indicative of the presence of a cytosine base which was previously methylated in the starting DNA molecule. Such cytosine bases can be detected by a variety of techniques. For example, primers specific for unmethylated versus methylated DNA can be generated and used for PCR- based identification of methylated CpG dinucleotides. A separation/capture step may be performed, e.g. using binding molecules such as complementary oligonucleotide sequences. Standard and next-generation DNA sequencing protocols can also be used.

In other approaches, methylation-sensitive enzymes can be employed which digest or cut only in the presence of methylated DNA. Analysis of resulting fragments is commonly carried out using microarrays.

Affinity-based techniques exploit binding interactions to capture fragments of methylated DNA for the purposes of enrichment. Binding molecules such as anti-5- methylcytosine antibodies are commonly employed prior to subsequent processing steps such as PCR and sequencing.

Olkhov-Mitsel and Bapat (2012) [27] provide a comprehensive review of techniques available for the identification and assessment of MVP-based biomarkers involving methylcytosine.

For the purposes of assessing the methylation status of the MVP -based biomarkers characterised and described herein, any suitable method can be employed.

Preferred methods involve bisulphite treatment of DNA, including amplification of the identified MVP loci for methylation specific PCR and/or sequencing and/or assessment of the methylation status of target loci using methylation-discriminatory microarrays.

Amplification of MVP loci can be achieved by a variety of approaches.

Preferably, MVP loci are amplified using PCR. MVP may also be amplified by other techniques such as multiplex ligation-dependent probe amplification (MLPA). A variety of PCR-based approaches may be used. For example, methylation-specific primers may be hybridized to DNA containing the MVP sequence of interest. Such primers may be designed to anneal to a sequence derived from either a methylated or non-methylated MVP locus. Following annealing, a PCR reaction is performed and the presence of a subsequent PCR product indicates the presence of an annealed MVP of identifiable sequence. In such methods, DNA is bisulphite converted prior to amplification. Such techniques are commonly referred to as methylation specific PCR (MSP) [28].

In other techniques, PCR primers may anneal to the MVP sequence of interest independently of the methylation status, and further processing steps may be used to determine the status of the MVP. Assays are designed so that the MVP site(s) are located between primer annealing sites. This method scheme is used in techniques such as bisulphite genomic sequencing [29], COBRA [30], Ms-SNuPE [31]. In such methods, DNA can be bisulphite converted before or after amplification.

Preferably, small-scale PCR approaches are used. Such approaches commonly involve mass partitioning of samples (e.g. digital PCR). These techniques offer robust accuracy and sensitivity in the context of a highly miniaturised system (pico-liter sized droplets), ideal for the subsequent handling of small quantities of DNA obtainable from the potentially small volume of cellular material present in biological samples, particularly urine samples. A variety of such small-scale PCR techniques are widely available. For example, microdroplet-based PCR instruments are available from a variety of suppliers, including RainDance Technologies, Inc. (Billerica, MA;

http ://rai ndancetech . com/) and Bio-Rad, Inc. (http://www.bio-rad.com/). Microarray platforms may also be used to carry out small-scale PCR. Such platforms may include microfluidic network-based arrays e.g. available from Fluidigm Corp.

( V- Y- fhi ciu;::' '-Ι - ) .

Following amplification of MVP loci, amplified PCR products may be coupled to subsequent analytical platforms in order to determine the methylation status of the MVPs of interest. For example, the PCR products may be directly sequenced to determine the presence or absence of a methylcytosine at the target MVP or analysed by array-based techniques.

Any suitable sequencing techniques may be employed to determine the sequence of target DNA. In the methods of the present invention the use of high-throughput, so- called "second generation", "third generation" and "next generation" techniques to sequence bisulphite-treated DNA are preferred.

In second generation techniques, large numbers of DNA molecules are sequenced in parallel. Typically, tens of thousands of molecules are anchored to a given location at high density and sequences are determined in a process dependent upon DNA synthesis. Reactions generally consist of successive reagent delivery and washing steps, e.g. to allow the incorporation of reversible labelled terminator bases, and scanning steps to determine the order of base incorporation. Array-based systems of this type are available commercially e.g. from Illumina, Inc. (San Diego, CA;

http://www.illumina.com/).

Third generation techniques are typically defined by the absence of a requirement to halt the sequencing process between detection steps and can therefore be viewed as real-time systems. For example, the base-specific release of hydrogen ions, which occurs during the incorporation process, can be detected in the context of microwell systems (e.g. see the Ion Torrent system available from Life Technologies; http://www.lifetechnologies.com/). Similarly, in pyrosequencing the base-specific release of pyrophosphate (PPi) is detected and analysed. In nanopore technologies, DNA molecules are passed through or positioned next to nanopores, and the identities of individual bases are determined following movement of the DNA molecule relative to the nanopore. Systems of this type are available commercially e.g. from Oxford Nanopore (https://www.nanoporetech.com/). In an alternative method, a DNA polymerase enzyme is confined in a "zero-mode waveguide" and the identity of incorporated bases are determined with florescence detection of gamma-labeled phosphonucleotides (see e.g. Pacific Biosciences; http://www.pacificbiosciences.com/).

In other methods in accordance with the invention sequencing steps may be omitted. For example, amplified PCR products may be applied directly to hybridization arrays based on the principle of the annealing of two complementary nucleic acid strands to form a double-stranded molecule. Hybridization arrays may be designed to include probes which are able to hybridize to amplification products of an MVP and allow discrimination between methylated and non-methylated loci. For example, probes may be designed which are able to selectively hybridize to an MVP locus containing thymine, indicating the generation of uracil following bisulphite conversion of an unmethylated cytosine in the starting template DNA. Conversely, probes may be designed which are able to selectively hybridize to an MVP locus containing cytosine, indicating the absence of uracil conversion following bisulphite treatment. This corresponds with a methylated MVP locus in the starting template DNA.

Following the application of a suitable detection system to the array, computer- based analytical techniques can be used to determine the methylation status of an MVP. Detection systems may include, e.g. the addition of fluorescent molecules following a methylation status-specific probe extension reaction. Such techniques allow MVP status determination without the specific need for the sequencing of MVP amplification products. Such array-based discriminatory probes may be termed methylation-specific probes.

Any suitable methylation-discriminatory microarrays may be employed to assess the methylation status of the MVPs described herein. A preferred methylation- discriminatory microarray system is provided by Illumina, Inc. (San Diego, CA;

http://www.ijlumina.com/'). In particular, the Infinium HumanMethylation450

BeadChip array system may be used to assess the methylation status of diagnostic MVPs for cervical cancer as described herein. Such a system exploits the chemical modifications made to DNA following bisulphite treatment of the starting DNA molecule. Briefly, the array comprises beads to which are coupled oligonucleotide probes specific for DNA sequences corresponding to the unmethylated form of an MVP, as well as separate beads to which are coupled oligonucleotide probes specific for DNA sequences corresponding to the methylated form of an MVP. Candidate DNA molecules are applied to the array and selectively hybridize, under appropriate conditions, to the oligonucleotide probe corresponding to the relevant epigenetic form. Thus, a DNA molecule derived from an MVP which was methylated in the

corresponding genomic DNA will selectively attach to the bead comprising the methylation-specific oligonucleotide probe, but will fail to attach to the bead

comprising the non-methylation-specific oligonucleotide probe. Single-base extension of only the hybridized probes incorporates a labeled ddNTP, which is subsequently stained with a fluorescence reagent and imaged. The methylation status of the MVP may be determined by calculating the ratio of the fluorescent signal derived from the methylated and unmethylated sites.

Because the cervical cancer-specific diagnostic MVP biomarkers defined herein were initially identified using the Illumina Infinium HumanMethylation450 BeadChip array system, the same chip system can be used to interrogate those same MVPs in the diagnostic assays described herein. Alternative or customised arrays could, however, be employed to interrogate the cervical cancer-specific diagnostic MVP biomarkers defined herein, provided that they comprise means for interrogating all MVPs for a given method, as defined herein.

Techniques involving combinations of the above-described methods may also be used. For example, DNA containing MVP sequences of interest may be hybridized to microarrays and then subjected to DNA sequencing to determine the status of the MVP as described above.

In the methods described above, sequences corresponding to MVP loci may also be subjected to an enrichment process. DNA containing MVP sequences of interest may be captured by binding molecules such as oligonucleotide probes complementary to the MVP target sequence of interest. Sequences corresponding to MVP loci may be captured before or after bisulphite conversion or before or after amplification. Probes may be designed to be complementary to bisulphite converted DNA. Captured DNA may then be subjected to further processing steps to determine the status of the MVP, such as DNA sequencing steps.

Capture/separation steps may be custom designed. Alternatively a variety of such techniques are available commercially, e.g. the SureSelect target enrichment system available from Agilent Technologies ( hu;v Λ\ .¾^; k:^;u -n/h -no). In this system biotinylated "bait" or "probe" sequences (e.g. RNA) complementary to the DNA containing MVP sequences of interest are hybridized to sample nucleic acids.

Streptavidin-coated magnetic beads are then used to capture sequences of interest hybridized to bait sequences. Unbound fractions are discarded. Bait sequences are then removed (e.g. by digestion of RNA) thus providing an enriched pool of MVP target sequences separated from non-MVP sequences. In a preferred method of the invention, template DNA is subjected to bisulphite conversion and target loci are then amplified by small-scale PCR such as microdroplet PCR using primers which are independent of the methylation status of the MVP. Following amplification, samples are subjected to a capture step to enrich for PCR products containing the target MVP, e.g. captured and purified using magnetic beads, as described above. Following capture, a standard PCR reaction is carried out to incorporate DNA sequencing barcodes into MVP-containing amplicons. PCR products are again purified and then subjected to DNA sequencing and analysis to determine the presence or absence of a methylcytosine at the target genomic MVP [32].

The MVP biomarker loci defined herein are identified e.g. by Illumina® identifiers (IlmnID). These MVP loci identifiers refer to individual MVP sites used in the commercially available Illumina® Infinium Human Methyl ation450 BeadChip kit. The identity of each MVP site represented by each MVP loci identifier is publicly available from the Illumina, Inc. website under reference to the MVP sites used in the Infinium Human Methyl ation450 BeadChip kit.

Further information regarding MVP loci identification used in Illumina, Inc products is found in the technical note entitled "Technical Note: Epigenetics. CpG Loci Identification. A guide to Illumina' s method for unambiguous CpG loci identification and tracking for the Golden Gate® and Infinium® Assay for Methylation" published in 2010 and found at:

http://www.illumina.com/documents/products/technotes/technote cpg loci iden tification.pdf.

Further information regarding the Illumina® Infinium Human Methyl ation450 BeadChip system can be found at:

http://www.illumina.com/content/dam/illumina- marketine documents/products/datasheets/datasheet humanmethylation450.pdf;

and at:

http://www.illumina.com/content/dam/illumina- marketing documents/products/technotes/technote_hm450_data_analysis_optimization. pdf.

To complement evolving public databases to provide accurate MVP/CpG loci identifiers and strand orientation, Illumina® has developed a method to consistently designate MVP/CpG loci based on the actual or contextual sequence of each individual MVP/CpG locus. To unambiguously refer to MVP/CpG loci in any species, Illumina® has developed a consistent and deterministic MVP loci database to ensure uniformity in the reporting of methylation data. The Illumina® method takes advantage of sequences flanking a MVP locus to generate a unique MVP locus cluster ID. This number is based on sequence information only and is unaffected by genome version. Illumina' s standardized nomenclature also parallels the TOP/BOT strand nomenclature (which indicates the strand orientation) commonly used for single nucleotide polymorphism (S P) designation.

Illumina® Identifiers for the Infinium Human Methyl ation450 BeadChip system are also available from public repositories such as Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo/).

An MVP as defined herein thus refers to the CG dinucleotide motif identified in relation to each SEQ ID NO. and Illumina Identifier (limn ID) as listed in Table 1, wherein the cytosine base of the dinucleotide (noted in bold and square brackets in the sequences listed at Table 1) may (or may not) be modified. Thus by determining the methylation status of a CpG defined by or identified in a given SEQ ID NO., or determining whether such a CpG is methylated, it is meant that a determination is made as to whether the cytosine of the CG dinucleotide motif identified in bold and in square brackets in a sequence shown in Table 1 is methylated or not at one or more loci in the sample of DNA from the individual, accepting that variation in the sequence upstream and downstream of any given CpG may exist due to sequencing errors or variation between individuals.

Diagnostic methods for cervical cancer

Methods are disclosed herein for use in the diagnosis of cervical cancer.

The Inventors have surprisingly discovered a panel of 30 diagnostic MVPs whose methylation status varies as between normal tissue and cervical cancer tissue. These 30 MVPs are set out in Table 1 below. For each of the 30 diagnostic MVPs the CpG dinucleotide, of which the cytosine is subject to differential methylation, is indicated within square brackets (see Table 1). Polynucleotide sequence both upstream and downstream of the relevant CpG dinucleotide is shown for each diagnostic MVP.

The Inventors have discovered that within the panel of 30 diagnostic MVPs, MVPs are methylated in DNA samples from tissue from individuals with cervical cancer, whereas MVPs are unmethylated in DNA samples from tissue from individuals without cervical cancer - the determination of whether a given diagnostic MVP is methylated or unmethylated being made in accordance with the methods defined herein. Thus the disclosed panel provides a pool of MVPs which have high discriminatory power in the diagnosis of cervical cancer.

(a) providing DNA from a sample from the individual;

(b) determining whether each one of a group of MVPs selected from a panel

comprising the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG] is methylated, wherein the group comprises at least 10 of the MVPs identified in

SEQ ID NOS 1 to 30 and denoted by [CG]; and

(c) diagnosing cervical cancer in the individual when the at least 10 MVPs of the group of (b) are methylated.

In any such cervical cancer diagnostic method described herein, the group of MVPs (i.e. those MVPs, the methylation status of which are to be determined) may comprise 11 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG]; or the group may comprise 12 or more, 13 or more, 14 or more, 15 or more, 16 or more, 17 or more, 18 or more, 19 or more, 20 or more, 21 or more, 22 or more, 23 or more, 24 or more, 25 or more, 26 or more, 27 or more, 28 or more, 29 or more, or all 30 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG].

In any of these cervical cancer diagnostic methods, cervical cancer may be diagnosed when each MVP in the group of MVPs analysed are methylated. Thus the group of MVPs (i.e. those MVPs, the methylation status of which are to be determined) may comprise 10 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all 10 MVPs are methylated. The group of MVPs may comprise 11 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

11 MVPs are methylated.

The group of MVPs may comprise 12 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

12 MVPs are methylated.

The group of MVPs may comprise 13 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

13 MVPs are methylated.

The group of MVPs may comprise 14 or more of the MVPs identified in SEQ

ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

14 MVPs are methylated.

The group of MVPs may comprise 15 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all 15 MVPs are methylated.

The group of MVPs may comprise 16 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

16 MVPs are methylated.

The group of MVPs may comprise 17 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

17 MVPs are methylated.

The group of MVPs may comprise 18 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

18 MVPs are methylated.

The group of MVPs may comprise 19 or more of the MVPs identified in SEQ

19 MVPs are methylated.

The group of MVPs may comprise 20 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all 20 MVPs are methylated. The group of MVPs may comprise 21 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

21 MVPs are methylated.

The group of MVPs may comprise 22 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

22 MVPs are methylated.

The group of MVPs may comprise 23 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

23 MVPs are methylated.

The group of MVPs may comprise 24 or more of the MVPs identified in SEQ

24 MVPs are methylated.

The group of MVPs may comprise 25 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all 25 MVPs are methylated.

The group of MVPs may comprise 26 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

26 MVPs are methylated.

The group of MVPs may comprise 27 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

27 MVPs are methylated.

The group of MVPs may comprise 28 or more of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all

28 MVPs are methylated.

The group of MVPs may comprise 29 or more of the MVPs identified in SEQ

29 MVPs are methylated.

The group of MVPs may comprise all 30 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and cervical cancer may be diagnosed when all 30 MVPs are methylated. Preferably, cervical cancer may be diagnosed when all 30 of the MVPs identified in SEQ ID NO S 1 to 30 and denoted by [CG] are methylated.

In any of the methods described above the MVPs determined to be methylated may include the MVPs identi fied in SEQ ID NOS 1 to 10 and denoted by [CG], or may include the MVPs identified n SEQ ID NOS 1 to 11 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 12 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 13 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 14 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 15 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 16 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 17 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 18 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 19 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 20 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 21 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 22 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 23 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 24 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 25 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 26 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 27 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 28 and denoted by [CG or may include the MVPs identified n SEQ ID NOS 1 to 29 and denoted by [CG or may include all of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG].

Any of the diagnoistic methods described herein may be used to diagnose squamous cell carcinoma of the cervix.

Any of the diagnoistic methods described herein may be used to diagnose squamous cell carcinoma of the cervix at any of the FIGO stages of the disease.

Any of the diagnoistic methods described herein may be used to diagnose adenocarcinoma of the cervix. Any of the diagnoistic methods described herein may be used to diagnose adenocarcinoma of the cervix at any of the FIGO stages of the disease.

Thus, any of the diagnoistic methods described herein may be used to diagnose FIGO stage 0 cervical cancer.

Any of the diagnoistic methods described herein may be used to diagnose FIGO stage IA cervical cancer.

Any of the diagnoistic methods described herein may be used to diagnose FIGO stage IB cervical cancer.

Any of the diagnoistic methods described herein may be used to diagnose FIGO stage IIA cervical cancer.

Any of the diagnoistic methods described herein may be used to diagnose FIGO stage IIB cervical cancer.

Any of the diagnoistic methods described herein may be used to diagnose FIGO stage IDA cervical cancer.

Any of the diagnoistic methods described herein may be used to diagnose FIGO stage IIIB cervical cancer.

Any of the diagnoistic methods described herein may be used to diagnose FIGO stage IVA cervical cancer.

Any of the diagnoistic methods described herein may be used to diagnose FIGO stage IVB cervical cancer.

Prognostic methods for cervical cancer

Methods are disclosed herein for use in the prognosis of cervical cancer.

The Inventors have surprisingly discovered a panel of 145 prognostic MVPs whose methylation status may be used to classify tumors into either a group 1 tumor (also referred to as 'HPV-16 like') or a group 2 tumor (also referred to as 'HPV-45 like'). These 145 MVPs are set out in Table 2 below. For each of the 145 prognostic MVPs the CpG dinucleotide, of which the cytosine is subject to differential

methylation, is indicated within square brackets (see Table 2). Polynucleotide sequence both upstream and downstream of the relevant CpG dinucleotide is shown for each diagnostic MVP. The Inventors have discovered that methylation datasets may be generated using the 145 MVPs set out in Table 2 below, or subsets thereof, by identifying in a given patient sample whether each MVP is methylated or unmethylated. Application of a binary classification algorithm to the generated datasets may be used to determine the probability of a patient's tumor being either or a group 1 tumor or a group 2 tumor. Any sample with a probability value of greater than or equal to 0.5 is considered to derive from a group 2 tumour ('HPV45-like'). Any sample with a probability value of less than 0.5 is considered to derive from a group 1 tumour ('HPV16-like'). On this basis cervical cancer patients may be stratified into two groups.

Thus the disclosed panel provides a pool of MVPs which have high

discriminatory power in the classification of cervical cancer for patient stratification.

(a) providing a DNA sample from the patient;

(b) determining whether each one of a group of MVPs selected from a panel

(c) applying a binary classification algorithm to the dataset; and

In such methods the group of MVPs may comprise at least 20 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG]. The group of MVPs may comprise at least 21 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG], or at least 22 of the MVPs, or at least 23 of the MVPs, or at least 24 of the MVPs, or at least 25 of the MVPs, or at least 26 of the MVPs, or at least 27 of the MVPs, or at least 28 of the MVPs, or at least 29 of the MVPs, or at least 30 of the MVPs, or at least 31 of the MVPs, or at least 32 of the MVPs, or at least 33 of the MVPs, or at least 34 of the MVPs, or at least 35 of the MVPs, or at least 36 of the MVPs, or at least 37 of the MVPs, or at least 38 of the MVPs, or at least 39 of the MVPs, or at least 40 of the MVPs, or at least 41 of the MVPs, or at least 42 of the MVPs, or at least 43 of the

MVPs, or at least 44 of the MVPs, or at least 45 of the MVPs, or at least 46 of the

MVPs, or at least 47 of the MVPs, or at least 48 of the MVPs, or at least 49 of the

MVPs, or at least 50 of the MVPs, or at least 51 of the MVPs, or at least 52 of the

MVPs, or at least 53 of the MVPs, or at least 54 of the MVPs, or at least 55 of the

MVPs, or at least 56 of the MVPs, or at least 57 of the MVPs, or at least 58 of the

MVPs, or at least 59 of the MVPs, or at least 60 of the MVPs, or at least 61 of the

MVPs, or at least 62 of the MVPs, or at least 63 of the MVPs, or at least 64 of the

MVPs, or at least 65 of the MVPs, or at least 66 of the MVPs, or at least 67 of the

MVPs, or at least 68 of the MVPs, or at least 69 of the MVPs, or at least 70 of the

MVPs, or at least 71 of the MVPs, or at least 72 of the MVPs, or at least 73 of the

MVPs, or at least 74 of the MVPs, or at least 75 of the MVPs, or at least 76 of the

MVPs, or at least 77 of the MVPs, or at least 78 of the MVPs, or at least 79 of the

MVPs, or at least 80 of the MVPs, or at least 81 of the MVPs, or at least 82 of the

MVPs, or at least 83 of the MVPs, or at least 84 of the MVPs, or at least 85 of the

MVPs, or at least 86 of the MVPs, or at least 87 of the MVPs, or at least 88 of the

MVPs, or at least 89 of the MVPs, or at least 90 of the MVPs, or at least 91 of the

MVPs, or at least 92 of the MVPs, or at least 93 of the MVPs, or at least 94 of the

MVPs, or at least 95 of the MVPs, or at least 96 of the MVPs, or at least 97 of the

MVPs, or at least 98 of the MVPs, or at least 99 of the MVPs, or at least 100 of the

MVPs, or at least 101 of the MVPs, or at least 102 of the MVPs, or at least 103 of the

MVPs, or at least 104 of the MVPs, or at least 105 of the MVPs, or at least 106 of the

MVPs, or at least 107 of the MVPs, or at least 108 of the MVPs, or at least 109 of the

MVPs, or at least 110 of the MVPs, or at least 111 of the MVPs, or at least 112 of the

MVPs, or at least 113 of the MVPs, or at least 114 of the MVPs, or at least 115 of the

MVPs, or at least 116 of the MVPs, or at least 117 of the MVPs, or at least 118 of the

MVPs, or at least 119 of the MVPs, or at least 120 of the MVPs, or at least 121 of the

MVPs, or at least 122 of the MVPs, or at least 123 of the MVPs, or at least 124 of the

MVPs, or at least 125 of the MVPs, or at least 126 of the MVPs, or at least 127 of the

MVPs, or at least 128 of the MVPs, or at least 129 of the MVPs, or at least 130 of the

MVPs, or at least 131 of the MVPs, or at least 132 of the MVPs, or at least 133 of the MVPs, or at least 134 of the MVPs, or at least 135 of the MVPs, or at least 136 of the MVPs, or at least 137 of the MVPs, or at least 138 of the MVPs, or at least 139 of the MVPs, or at least 140 of the MVPs, or at least 141 of the MVPs, or at least 142 of the MVPs, or at least 143 of the MVPs, or at least 144 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG]. The group of MVPs may comprise all 145 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

The group of MVPs may also comprise subsets of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG], as defined further herein.

Any of the classification and prognostic methods described herein may be applied to a squamous cell carcinoma of the cervix.

Any of the classification and prognostic methods described herein may be applied to a squamous cell carcinoma of the cervix at any of the FIGO stages of the disease.

Any of the classification and prognostic methods described herein may be applied to an adenocarcinoma of the cervix.

Any of the classification and prognostic methods described herein may applied to an adenocarcinoma of the cervix at any of the FIGO stages of the disease.

Thus, any of the classification and prognostic methods described herein may be applied to FIGO stage 0 cervical cancer.

Any of the classification and prognostic methods described herein may be applied to FIGO stage IA cervical cancer.

Any of the classification and prognostic methods described herein may be applied to FIGO stage IB cervical cancer.

Any of the classification and prognostic methods described herein may be applied to FIGO stage IIA cervical cancer.

Any of the classification and prognostic methods described herein may be applied to FIGO stage IIB cervical cancer.

Any of the classification and prognostic methods described herein may be applied to FIGO stage IIIA cervical cancer.

Any of the classification and prognostic methods described herein may be applied to FIGO stage IIIB cervical cancer. Any of the classification and prognostic methods described herein may be applied to FIGO stage IVA cervical cancer.

Any of the classification and prognostic described herein may be applied to FIGO stage IVB cervical cancer.

Bioinformatic tools and statistical metrics

Software programs which aid in the in silico analysis of bisulphite converted DNA sequences and in primer design for the purposes of methylation-specific analyses are generally available and have been described previously [33,34,35] .

Sensitivity and specificity metrics for cervical cancer diagnosis based on the

MVP methylation status assays described herein may be defined using standard receiver operating characteristic (ROC) statistical analysis [36] . In ROC analysis 100% sensitivity corresponds to a finding of no false negatives, and 100%> specificity corresponds to a finding of no false positives.

Based on analyses conducted using a panel of 30 MVP biomarkers, a cervical cancer diagnostic assay in accordance with the invention described herein can achieve a ROC sensitivity of 90% or greater, 91% or greater, 92%> or greater, 93% or greater, 94%> or greater, 95% or greater, 96% or greater, 97% or greater, 98% or greater or 99%. The ROC sensitivity may be 100%.

Diagnostic assays in accordance with the invention can achieve a ROC specificity of 50% or greater, 51% or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60%) or greater, 61% or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71%) or greater, 72% or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82%o or greater, 83% or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89% or greater, 90% or greater, 91% or greater, 92% or greater, 93%o or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98% or 99%. The ROC specificity may be 100%. Diagnostic assays in accordance with the invention may have an associated combination of ROC sensitivity and ROC specificity values wherein the combination is any one of the above-listed sensitivity values and any one of the above-listed specificity values.

Thus, the ROC sensitivity may be 100% and the ROC specificity may be 50% or greater, 51%> or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56%o or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61% or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67%o or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72% or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78%o or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83% or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89%o or greater, 90% or greater, 91% or greater, 92% or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

The ROC sensitivity may be 95% or greater and the ROC specificity may be

50%) or greater, 51% or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61%) or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72%o or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83%o or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89% or greater, 90% or greater, 91% or greater, 92% or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

The ROC sensitivity may be 90% or greater and the ROC specificity may be

50%) or greater, 51% or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61%) or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72%o or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83% or greater, 84%> or greater, 85%> or greater, 86%> or greater, 87%> or greater, 88%> or greater, 89%> or greater, 90% or greater, 91% or greater, 92%> or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

The ROC sensitivity may be 85%> or greater and the ROC specificity may be 50%) or greater, 51%> or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61%o or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72%o or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83%o or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89% or greater, 90% or greater, 91% or greater, 92% or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

The ROC sensitivity may be 80% or greater and the ROC specificity may be 50%o or greater, 51% or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61%o or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72%o or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83%o or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89% or greater, 90% or greater, 91% or greater, 92% or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

The ROC sensitivity may be 75% or greater and the ROC specificity may be 50%o or greater, 51% or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61%o or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72%o or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83%o or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89% or greater, 90% or greater, 91% or greater, 92%> or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

The ROC sensitivity may be 70% or greater and the ROC specificity may be 50%) or greater, 51% or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61%o or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72%o or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83%o or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89% or greater, 90% or greater, 91% or greater, 92% or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

The ROC sensitivity may be 65% or greater and the ROC specificity may be 50%o or greater, 51% or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61%o or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72%o or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83%o or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89% or greater, 90% or greater, 91% or greater, 92% or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

The ROC sensitivity may be 60% or greater and the ROC specificity may be 50%o or greater, 51% or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61%o or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72%o or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83%o or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89% or greater, 90% or greater, 91% or greater, 92%> or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

The ROC sensitivity may be 55% or greater and the ROC specificity may be 50%) or greater, 51% or greater, 52% or greater, 53% or greater, 54% or greater, 55% or greater, 56% or greater, 57% or greater, 58% or greater, 59% or greater, 60% or greater, 61%) or greater, 62% or greater, 63% or greater, 64% or greater, 65% or greater, 66% or greater, 67% or greater, 68% or greater, 69% or greater, 70% or greater, 71% or greater, 72%) or greater, 73% or greater, 74% or greater, 75% or greater, 76% or greater, 77% or greater, 78% or greater, 79% or greater, 80% or greater, 81% or greater, 82% or greater, 83%) or greater, 84% or greater, 85% or greater, 86% or greater, 87% or greater, 88% or greater, 89% or greater, 90% or greater, 91% or greater, 92% or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98%, 99% or 100%.

A further metric which can be employed to classify the accuracy of the MVP- based assays is ROC AUC. In ROC analysis, the area under the curve of a ROC plot (AUC) is a metric for binary classification. In a random binary classifier the number of true positives and false positives will be approximately equal. In this situation the AUC score for the ROC plot will be 0.5. In a perfect binary classifier the number of true positives will be 100% and the number of false positives will be 0%. In this situation the AUC score for the ROC plot will be 1.

Based on analyses conducted using the panel of 30 diagnostic MVP biomarkers described herein, a cervical cancer diagnostic assay in accordance with the invention can achieve a ROC AUC of 0.90 or greater, 0.91 or greater, 0.92 or greater, 0.93 or greater, 0.94 or greater, 0.95 or greater, 0.96 or greater, 0.97 or greater, 0.98 or greater, 0.99 or 1.00. Preferably the diagnostic assay can achieve a ROC AUC of 0.98 or greater.

Cervical cancer diagnostic tests based on the 30 diagnostic MVP biomarkers described herein may also be characterised using a Negative Predictive Value (NPV) metric. The NPV is a measure of the proportion of negative results that are true negative results.

Based on analyses conducted using the panel of 30 diagnostic MVP biomarkers, a cervical cancer diagnostic assay in accordance with the invention described herein can achieve an NPV of 90% or greater, 91% or greater, 92%> or greater, 93% or greater, 94%) or greater, 95% or greater, 96%> or greater, 97% or greater, 98% or greater or 99% or 100%. Biological samples

The cervical cancer diagnostic assays described herein may be performed on any suitable biological material obtained from the patient. Preferred biological material is a Liquid Based Cytology (LBC) sample. However, samples of cervical tissue, e.g.

obtained via biopsy or aspirates, or obtained from preserved samples (e.g.

cryopreserved material, tissue sections etc.) may be used. Samples of biological material may also include solid tissue samples, aspirates, samples of biological fluids, blood, serum, plasma, ascitic fluid, lymph, peripheral blood, cerebrospinal fluid, fine needle aspirate, saliva, sputum, bone marrow, skin, epithelial samples (including buccal, cervical or vaginal epithelia) or other tissue derived from the ectoderm, vaginal fluid, semen etc. Tissue scrapes may include biological material from e.g. buccal,

oesophageal, bladder, vaginal, urethral or cervical scrapes. The cells of the sample may comprise inflammatory cells, such as lymphocytes.

Any of the assays and methods described herein may involve providing a biological sample from the patient as the source of patient DNA for methylation analysis.

Any of the assays and methods described herein may involve obtaining patient DNA from a biological sample which has previously been obtained from the patient.

Any of the assays and methods described herein may involve obtaining a biological sample from the patient as the source of patient DNA for methylation analysis. Procedures for obtaining a biological sample from the patient may be noninvasive, such as collecting cells from urine. Alternatively, invasive procedures such as biopsies may be used.

In the methods described herein the level of detection is such that 2 tumor cells may be detected in a sample comprising 150,000 cells or more. In such methods the sample may comprise 160,000 cells or more, 170,000 cells or more, 180,000 cells or more, 190,000 cells or more, 200,000 cells or more, 210,000 cells or more, 220,000 cells or more, 230,000 cells or more, 240,000 cells or more, 250,000 cells or more, 260,000 cells or more, 270,000 cells or more, 280,000 cells or more, 280,000 cells or more, or 300,000 cells or more.

In any such method, the number of tumor cells that can be detected is 10 or more, 20 or more, 30 or more, 40 or more, 50 or more, 60 or more, 70 or more, 80 or more, 90 or more, 100 or more, 200 or more, 300 or more, 400 or more, 500 or more, 600 or more, 700 or more, 800 or more, 900 or more, 1000 or more, 2000 or more, 3000 or more, 4000 or more, 5000 or more, 6000 or more, 7000 or more, 8000 or more, 9000 or more, 10000 or more, 20000 or more, 30000 or more, 40000 or more, 50000 or more, 60000 or more, 70000 or more, 80000 or more, 90000 or more or 100000 or more.

Methods of treatment

The invention encompasses methods wherein following a positive diagnosis of cervical cancer, and/or assignment of a patient's tumor as a group 1 tumor, the patient may be administered of one or more urogical procedures, one or more surgical procedures, one or more chemotherapeutic agents, one or more immunotherapeutic agents, one or more radiotherapeutic agents, one or more small molecule agents, one or more kinase inhibitor agents or any combination of the above.

Urogical procedures include radical trachelectomy or radical hysterectomy resection of cervical tumor

Chemotherapeutic agents include the following. Alkylating agents, which include the nitrogen mustards, nitrosoureas, tetrazines, aziridines, cisplatin and platinum based derivatives, as well as the non-classical alkylating agents. Antimetabolites, which include the anti-folates, fluoropyrimidines, deoxynucleoside analogues and thiopurines. Microtubule disrupting agents, which include the vinca alkaloids and taxanes, as well as dolastatin 10 and derivatives thereof. Topoisomerase inhibitors, which include camptothecin, irinotecan and topotecan. Topoisomerase II poisons, which include etoposide, doxorubicin, mitoxantrone and teniposide. Topoisomerase II catalytic inhibitors, which include novobiocin, merbarone, and aclarubicin. Cytotoxic antibiotics, which include anthracyclines, actinomycin, bleomycin, plicamycin, and mitomycin. Combinations of agents include but are not limited to MVAC (Methotrexate, Vinblastine, Vinblastine and Vinblastine), Gem-Cis (GC) (Gemcitabine and Cisplatin), Lapatinib and gemcitabine.

Ilmmune checkpoint modulators include Ipilimumab, Nivolumab,

Pembrolizumab and Atezolizumab.

Other immunotherapeutics include chimeric antigen receptor (CAR) T-cell therapy, HPV therapeutic vaccines and antibody-drug conjugates (ADC). Antibody- drug conjugates include antibodies conjugated to microtubule disrupting agents and DNA modifying agents as described above.) or currently in clinical development.

Targeted kinase inhibitor therapies include Lapatinib, Dasatinib, Gefinitinib,

Erlotinib and Vemurafenib.

Cancer therapeutic agents are administered to a subject already suffering from a disorder or condition, in an amount sufficient to cure, alleviate or partially arrest the condition or one or more of its symptoms. Such therapeutic treatment may result in a decrease in severity of disease symptoms, or an increase in frequency or duration of symptom-free periods. An amount adequate to accomplish this is defined as

"therapeutically effective amount". Effective amounts for a given purpose will depend on the severity of the disease as well as the weight and general state of the subject. As used herein, the term "subject" includes any human.

Arrays

The invention also encompasses arrays capable of discriminating between methylated and non-methylated forms of MVPs as defined herein; the arrays may comprise oligonucleotide probes specific for methylated forms of MVPs as defined herein and oligonucleotide probes specific for non-methylated forms of MVPs as defined herein.

By "specific" it is meant that the probes comprise sequences which are complementary to those of the oligonucleotides comprising the MVP such they may hybridize, particularly under stringent conditions. In some embodiments the array is not an Illumina Infinium

HumanMethylation450 BeadChip array (Infinium HumanMethylation450 BeadChip array).

Separately or additionally, in some embodiments the number of MVP-specific oligonucleotide probes of the array is less than 482,421, preferably 482,000 or less, 480,000 or less, 450,000 or less, 440,000 or less, 430,000 or less, 420,000 or less, 410,000 or less, or 400,000 or less, 375,000 or less, 350,000 or less, 325,000 or less, 300,000 or less, 275,000 or less, 250,000 or less, 225,000 or less, 200,000 or less, 175,000 or less, 150,000 or less, 125,000 or less, 100,000 or less, 75,000 or less, 50,000 or less, 45,000 or less, 40,000 or less, 35,000 or less, 30,000 or less, 25,000 or less,

20,000 or less, 15,000 or less, 10,000 or less, 5,000 or less, 4,000 or less, 3,000 or less or 2,000 or less.

The invention further encompasses the use of any of the arrays as defined herein in any of the methods which require determining the methylation status of MVPs for the purposes of diagnosing cervical cancer cells in an individual.

Kits

Any of the arrays as defined herein may be comprised in a kit.

The kit may comprise any array as defined herein.

The kit may comprise any array as defined herein together with instructions for use.

The kit may additionally comprise a DNA modifying regent, such as a bisulphite reagent.

The kit may additionally comprise reagents for amplifying DNA, such as primers directed to any of the MVPs as defined herein as identified in SEQ ID NOS 1 to 175 (see Tables 1 and 2).

Methods of determining a methylation profile of a sample

The invention further encompasses a method of determining a methylation profile of a sample from an individual, the method comprising:

i. providing DNA from a sample from the individual; ii. determining whether each one of a group of MVPs selected from a panel comprising the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG] is methylated, wherein the group comprises at least 10 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG]; and

iii. determining whether each one of a group of MVPs selected from a panel

comprising the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG] is methylated, wherein the group comprises at least 20 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG]; and

iv. based on the methylation status of the MVPs of the group, determining a

methylation profile of the sample.

In any of the above-described methods of determining a methylation profile of a sample, the group of MVPs which are selected from a panel comprising the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG] (i.e. the group of MVPs whose methylation status is to be determined) may comprise any number of MVPs as described and defined herein, provided that the group comprises at least 10 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG]. Or the group of MVPs which are selected from a panel comprising the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG] (i.e. the group of MVPs whose methylation status is to be determined) may comprise any number of MVPs as described and defined herein, provided that the group comprises at least 20 of the MVPs identified in SEQ ID NOS 1 to 175 and denoted by [CG].

Furthermore, in any such methods, the methylation status of MVPs may be determined using any of the arrays described herein.

In any of the diagnostic methods described herein, the step of diagnosing cervical cancer in the individual may further comprise:

I. determining the prognosis of a patient; and/or

II. stratifying the grade of the tumor, e.g. following primary surgery or radiotherapy; and/or

III. determining the risk of recurrence of the tumor; and/or

VI. determining the likely response to treatment therapy and/or

IV. providing a cervical cancer treatment to the patient. The invention also encompasses a method of determining the risk of the development of cervical cancer in an individual, the method comprising:

(a) providing DNA from a sample from the individual;

(b) determining whether each one of a group of MVPs selected from a panel

comprising the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG] is methylated, wherein the group comprises at least 25 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG]; and

(c) based on the methylation status of the MVPs of the group, determining the risk of the development of cervical cancer in the individual.

In any such method of determining the risk of the development of cervical cancer in an individual, the group of MVPs which are selected from a panel comprising the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG] (i.e. the group of MVPs whose methylation status is to be determined) may comprise any number of MVPs as described and defined herein.

The invention also encompasses the use of a group of MVPs in the diagnosis of cervical cancer in an individual, in determining the risk of the development of cervical cancer in an individual or predicting the prognosis of an individual with cervical cancer.

In any such use, the group of MVPs are selected from a panel comprising the

MVPs identified in SEQ ID NOS 1 to 30 or 31 to 175 and denoted by [CG].

Examples Materials and methods

Introduction

Emerging techniques that utilize next-generation DNA sequencing platforms hold particular promise for the development of highly sensitive epigenetic biomarker panels. For example, the microdroplet-based PCR amplification of bisulphite converted DNA, followed by next-generation sequencing of the amplified target loci developed by RainDance Technologies [37] enables the sensitive, specific and simultaneous amplification of up to 20,000 bisulfite-converted target loci. Highly parallel microdroplet-based PCR amplification of bisulphite-converted DNA has shown utility in the validation of epigenetic alterations in a range of tissues [32, 38, 39].

To derive a sensitive assay for the detection and prognosis of cervical cancer, the inventors have performed studies of genome-wide methylation in cervical cancer. From this, a panel of cervical- specific epigenetic biomarkers have been defined and the sensitivity and specificity of a 30 loci diagnostic panel and a 145 loci prognostic panel using RainDrop-BS [32] have been validated for the detection and prognosis of cervical cancer with a high degree of diagnostic precision.

Study Population

Genome-wide DNA methylation profiling was performed on DNA from 97 cervical cancer samples and 29 matched normal cervix tissues collected from

Haukeland University Hospital (Bergen, Norway - 'validation cohort ) and a further 57 cervical cancer samples from the Medical University of Innsbruck (Innsbruck, Austria - 'validation cohort 3'). DNA methylation data from The Cancer Genome Atlas (TCGA) cervical cancer project (https://gdc cancer.gov/ - 'discovery cohort') and from a further 270 cervical cancer samples (GEO accession number GSE68339 - 'validation cohort 2') were also included in the analysis.

DNA extraction and quantification

DNA was extracted using either Qiagen QIAamp tissue DNA isolation kit, PureLink DNA extraction kit (Life Technologies) or proteinase K digestion followed by chloroform extraction / ethanol precipitation. DNA was quantified by

spectrophotometry (Nanodrop 1000) and fluorometry (Qubit ds DNA HS assay, Invitrogen, UK).

Methylation Array

500 ng of DNA was bisulphite converted and hybridised to the Infinium 450K

Human Methylation array and processed in accordance with the manufacturer's recommendations. DNA bisulphite conversion was carried out using the EZ DNA Methylation kit (Zymo Research) as per the manufacturer's instructions. Samples were processed in a single batch. R statistical software (version 2.14.0 [40]) was used for the subsequent data analysis. The ChAMP pipeline was used to extract and analyze data from iDat files, samples were normalised using BMIQ [41, 42]. Raw β values

(methylation value) were subjected to a stringent quality control analysis as follows: samples showing reduced coverage were removed and only probes with detection levels above background across all samples were retained (detection P < 0.01). DMRs (Differentially Methylated Regions) were determined using Lasso [43,44].

Example 1: Materials and methods for the development of a panel of methylation biomarkers for use in the diagnosis of cervical cancer

DNA extraction and quantification

DNA was extracted using a DNeasy blood and tissue kit (Qiagen, UK) in accordance with the manufacturer's instructions. DNA was quantified by

spectrophotometry (Nanodrop 1000) and fluorometry (Qubit ds DNA HS assay, Invitrogen, UK). Methylation Array

Genome-wide methylation profiles were obtained from 77 cervical cancer and 29 normal cervix samples using bead arrays (Illumina Infinium 450k). 500 ng of DNA was bisulphite converted and hybridised to the Infinium 450K Human Methylation array and processed in accordance with the manufacturer's recommendations. DNA bisulphite conversion was carried out using the EZ DNA Methylation kit (Zymo

Research) as per the manufacturer's instructions. Samples were processed in a single batch. Publicly available DNA methylation data for samples profiled by The Cancer Genome Atlas (TCGA) cervical cancer project was downloaded and analysed. R statistical software (version 2.14.0 [35]) was used for the subsequent data analysis.

Raw IDAT files were between-array- normalised using the FunNorm function in the minfi Bioconductor package (Aryee, Jaffe et al. 2014), with initial noob correction of intensities using dye-swap (Triche, Weisenberger et al. 2013). Beta values were then retrieved and corrected for probe-type differences using BMIQ normalisation

(Teschendorff, Marabita et al. 2013) using 10,000 reference probes for EM fitting.

To define a DNA methylation biomarker panel, those loci which were determined to be methylated (β >60%) in at least 50% of cancers and unmethylated in normal urothelium (β <20%) were identified. By "methylated (β >50%)", as discussed herein in relation to the development of this initial DNA methylation biomarker panel, it is meant that for any given locus, >50% of cells in a patient sample are determined to be methylated with respect to that MVP.

Example 2: Development of a panel of methylation biomarkers for use in the diagnosis of cervical cancer

29 normal samples were compared to 101 cancer samples profiled on the Illumina Infinnium 450k array platform as described above, with delta beta-cutoffs at 0.65 between normal and cancer, yielding a set of 30 methylation variable positions (MVPs), as set out in Table 1 above (SEQ ID NOs 1 to 30). All 30 MVPs were determined to be methylated in cervical cancer samples and unmethylated in samples from normal tissue.

The performance of subsets of the 30 MVP signature, and the entire 30 MVP signature, were then validated in independent data consisting of TCGA (The Cancer Genome Atlas) and other normal samples (n = 18) by training four different models using cross validation to identify the optimal parameters by grid search. The four different training models used were a support vector machine model (SVM), a k-nearest neighbours model (KNN), a GLMnet model (GLM), and a Random Forest model (RF). Performance was assessed in terms of sensitivity, specificity, and Kappa values in the validation dataset.

Figure 1 shows standard ROC curves for the 30 MVP signature (all 30 of SEQ ID NOs 1 to 30) in each model, showing a high degree of sensitivity and good specificity. Figure 2 shows separate statistical metrics for each each model. Figure 3 shows standard ROC curves for a subset of the 30 MVP signature in each model, showing a high degree of sensitivity and good specificity. In this case the MVPs analysed were all 20 of SEQ ID NOs 1 to 20. Figure 4 shows separate statistical metrics for each each model.

Figure 5 shows a heatmap demonstrating that probes corresponding to MVPs defined by SEQ ID NOs 1 to 30 display hypomethylation in normal cervical tissue and hypermethylation in cervical cancer samples.

Example 3: Materials and methods for the development of a panel of methylation biomarkers for use in the prognosis of cervical cancer patient outcome

Assembly of datasets for prognostic classification

Level 3 RSEM upper-quartile normalised count values were downloaded from the TCGA data portal for 307 CESCs and parsed into a feature-by-sample matrix using an in-house package. Methylation beta values were extracted for this subset of samples from the discovery set established for the identification of Pan-tissue UPV methylome profiles. Clinical data were isolated from the TCGA data portal. The analysis was conducted on tumours histologically classified as Cervical Squamous Cell Carcinomas. Expression modelling

Cox regression modelling indicated significant differences in survival between patients having 'HPV45-like' and 'HPV16-like' tumors, especially amongst early stage samples, with HPV18+ tumours comprising an intermediate group. To explore the extent and significance of molecular heterogeneity associated with UPV-type tumors, comparisons were therefore performed between HPV45-like' and 'HPV16-like' tumours.

RNA seq RSEM fractional counts were quantile normalised and filtered to remove low-expressed genes (less than 1 count per million in more than a quarter of samples). Limma voom was then used to estimate mean-variance precision weights for robust linear modelling. Differentially expressed genes were identified at a mean fold- change of 2 and a BH-corrected false-discovery rate of 0.01. Methylation modelling

Differentially methylated probes were identified using a custom R wrapper to limma with correction for mean-variance trend upon empirical bayes modelling to account for the heteroscedasticity of beta-values. Significant MVPs were defined at a mean delta-beta of 0.1 at a two-step BH adjusted false discovery rate of 0.01. Small beta-value thresholds were used to permit more sensitive detection of methylation differences.

The identification of DMRs was carried out using fDMR, and DMRs were required to have at least 3 mvp's at a false discovery rate of 5% and an average deltaBeta difference of 0.1 and a DMR FDR of 0.05. Both these signatures were reduced to features associated with differential expression at an FDR of 0.01 and a fold change of 2, following limma-trend analysis of log2-cpm matrices reduced to genes in the signatures.

Joint clustering of the FIPV45-16 gene expression signature and the functional MVP signature was carried out using a random forest proximity matrix to serve as the distance matrix across the dataset of stage I and stage II tumours. PAM Consensus Clustering was then carried out with 2:4 candidate clusters to identify the most robust partitioning of samples using a custom R wrapper calling clustering and robustness evaluation functions from the clusterCons R package.

Example 4: Development of a panel of methylation biomarkers for use in the prognosis of cervical cancer patient outcome

Aims of the study

The inventors sought to define the DNA methylation landscape of cervical cancer and to identify epigenetic markers of prognosis and to understand the biological differences underlying the differential clinical behaviours of disease subtypes. General summary of the methods of the study

Genome-wide methylation profiles were obtained from 77 cervical cancer and 29 normal cervix samples using bead arrays (Illumina Infinium 450k). Publicly available DNA methylation data for samples profiled by The Cancer Genome Atlas (TCGA) cervical cancer project was analysed and integrated with HPV type information gained from RNA-sequencing data. Data on HPV type and DNA methylation were used to identify a methylation signature of 145 differentially methylated loci that defined 'HPV16-like' and 'HPV45-like' tumours. Clinical data were integrated with these findings to examine survival in either molecular subtype using multivariate analysis. Methylation data were used to estimate differences in infiltrating immune cell fractions between the subgroups using a novel computational approach, which we term

' Methyl CIBERSORT' . The influence of different HPV types on DNA methylation was investigated using keratinocytes (NIKS) transduced with HPV16, HPV18 or HPV45 genomes.

Data analysis

Linear-modelling was carried out to identify probes differentially methylated between 'HPV16-like' and 'HPV45-like' tumour clusters at an FDR of 0.01 (Two-step BH corrected), dB > 0.3. Significant MVPs were then used to train an assortment of classifiers with the tuning parameters set out in the Table below using the caret R package using 10 iterations of 10 fold Cross- Validation.

Model Parameters and Ranges

K Nearest Neighbours (kNN) K (number of nearest neighbours) - 1,3,5

Gradient Boosted Machine Shrinkage - 0.01,0.1,0.5,1 ; Minimum observations in node (gbm) = 10, interaction depth - 5,10 , number of trees - 150,300

Nearest Shrunken Centroid Threshold - 0.01,0.1,1,3,5,10,20.

Glmnet (Elastic Net) Alpha - 0,0.2,0.4,0.6,0.8,1 ; Lambda - 0, 0.01, 0.02, 0.03,

0.04, 0.05

SVM (Support Vector Machine) C - 0 to 1, intervals of 0.1.

- Linear Kernel (svmLinear)

Random Forest mtry = 1/3 the number of features.

Table: Machine Learning Models and Tuning Parameters for developing a model to classify cervical cancers into 'HPV16-like' (group 1) or 'HPV45-like' (group 2) tumors. For each model, the best parameters were selected on the basis of maximum

Kappa values from the cross-validation tuning/fitting process. Class allocation for the dataset was made using out-of-fold calls to generate a majority vote. The different models were compared on the basis of Kappa values and how closely they recapitulated the Hazard Ratios of Cox regressions carried out on type-associated clusters to select a model for application to validation datasets. Performance in validation data was then used to select the optimum set of features and the model on the basis of magnitude of separation of KM curves and statistical significance of separation.

Results

Results of the linear-modelling, which was carried out to identify probes differentially methylated between 'FIPV16-like' and 'FIPV45-like' tumours, identified the 145 MVPs listed in Table 2 (SEQ ID NOs 31 to 175).

'FIPV45-like' tumours displayed significantly worse outcomes (TCGA overall survival FIR=12.7, p<9E-07). Tumours in this poor prognosis group displayed reduced CD8+ T-lymphocyte infiltration and increased expression of marker genes associated with the cervical transformation zone. Significant differences were observed in the genome-wide methylation patterns of NIKS transduced with different HPV types.

Application of machine learning classifier algorithms to methylation datasets from TCGA cervical cancer samples was capable of determining the fit of a given tumor sample with either 'HPV16-like' (group 1) or 'HPV45-like' (group 2) methylation characteristics. This allows patient stratification into one of two groups: 'HPV16-like' (group 1) or 'HPV45-like' (group 2). Figure 6 shows a heatmap showing stratification of TCGA cervical cancer samples into two groups 'HPV16-like' and 'HPV45-like', on the basis of the pattern of DNA methylation at the 145 informative MVP loci (SEQ ID NOs 31 to 175). Figure 7 shows Kaplan-Meier curves showing overall survival of patients with tumours classified as HPV16-like or HPV45-like in TCGA (discovery) or Norwegian (validation) cohorts. Figure 8 shows shows a Kaplan- Meier curve showing overall survival of patients with tumours classified as HPV16-like or HPV45-like on the basis of all 145 MVP loci (SEQ ID NOs 31 to 175). Figure 9 shows shows a Kaplan-Meier curve showing overall survival of patients with tumours classified as HPV16-like or HPV45-like on the basis of a subset of the top 20 MVPs out of the total group of 145 MVP loci (i.e. based on MVPs defined by SEQ ID NOs 31 to 50). Figure 10 shows histograms displaying the classifier-derived probability that each sample derives from a group 1 tumour (probability of group 2 membership = 0) or group 2 tumour (probability of group 2 membership = 1). Histograms are shown for each of two independent validation cohorts. Any sample with a probability value of greater than or equal to 0.5 is considered to derive from a group 2 tumour. Any sample with a probability value of greater than or equal to 0.5 is considered to derive from a group 2 tumour ('HPV45-like'). Any sample with a probability value of less than 0.5 is considered to derive from a group 1 tumour ('HPV16-like'). On this basis cervical cancer patients may be stratified into two groups. Patients having a tumor classified as group 2 are more likely to have a significantly worse outcome compared to patients having a tumor classified as group 1. Therefore, group 2 patients could be spared from treatment regimes which are unlikely to provide significant benefit. Conversely, group 1 patients could be provided with treatment regimes which are likely to provide significant benefit. Conclusions

Cervical cancers can be stratified according to their DNA methylation profiles, into two groups that display differences in HPV type, the immune microenvironment and clinical outcome.

Table 1

The Table below provides a list of 30 methylation variable positions (MVPs or CpGs) as used in the cervical cancer diagnostic methods described herein. Provided for each CpG is the Illumina Identifier (limn ID), the chromosome number (CHR) and chromosome position (MAPINFO) specifying the location of the CpG in the human genome, a forward sequence encompassing each CpG and a corresponding S ID number. The CG dinucleotide motif containing the cytosine which is subject to modification by methylation is identified in bold and within square brackets in each sequence. The CpGs are listed in rank order from 1 to 30. The rank order is in respect of the methylation bet value. Thus, the MVP corresponding to SEQ ID NO: 1 has the highest beta value, whilst the MVP corresponding to SEQ ID NO: 30 has th lowest beta value.

Table 2

The Table below provides a list of 145 methylation variable positions (MVPs or CpGs) as used in the cervical cancer prognostic methods described herein. Provided for each CpG is the Illumina Identifier (limn ID), the chromosome number (CHR) and chromosome position (MAPINFO) specifying the location of the CpG in the human genome, a forward sequence encompassing each CpG and a corresponding S ID number. The CG dinucleotide motif containing the cytosine which is subject to modification by methylation is identified in bold and within square brackets in each sequence. The CpGs are listed in rank order from 1 to 30. The rank order is in respect of the methylation be value. Thus, the MVP corresponding to SEQ ID NO: 1 has the highest beta value, whilst the MVP corresponding to SEQ ID NO: 30 has th lowest beta value.

References

1. Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA: A Cancer Journal for Clinicians 2012, 62, pp 10-29.

2. Cancer Research UK statistics: http ://www. cancerresearchuk. org/health- professional/cancer-statistics/statistics-by-cancer-type/cervical-cancer, and references therein. 3. Market report published by Grand View Research Inc., June 2016 :

http://www.grandviewresearch.com/industiy-analysis/hpy-testing-pap-test-market and references therein.

4. Cuzick J, Clavel C, Petry KU, Meijer CJ, Hoyer H, Ratnam S, Szarewski A, Birembaut P, Kulasingam S, Sasieni P, Iftner T. Overview of the European and North

American studies on HPV testing in primary cervical cancer screening. Int. J. Cancer 2006, 119(5), ppl095-1101.

5. Ronco G, Dillner J, Elfstrom KM, Tunesi S, Snijders PJ, Arbyn M, Kitchener H, Segnan N, Gilham C, Giorgi-Rossi P, Berkhof J, Peto J, Meijer CJ, et al. Efficacy of

HPV-based screening for prevention of invasive cervical cancer: follow-up of four European randomised controlled trials. Lancet. 2014, 383(9916), pp524-532.

6. Ronco G, Giorgi-Rossi, Carozzi F, Confortini M, Dalla Palma P, Del Mistro A, Ghiringhello B, Girlando S, Gillio-Tos A, De Marco L, Naldoni C, Pierotti P, et al.

Efficacy of human papillomavirus testing for the detection of invasive cervical cancers and cervical intraepithelial neoplasia: a randomised controlled trial. Lancet Oncol. 2010, 11(3), pp249-57.

7. Rao J, Escobar-Hoyos L, and Shroyer KR. Unmet clinical needs in cervical cancer screening. MLO MedLab Obs. 2016, 48(1), p8. 8. Friedlander M, Grogan M: Guidelines for the Treatment of Recurrent and Metastatic Cervical Cancer. The Oncologist 2002, 7, pp342-347. 9. Beukers W, Hercegovac A, Vermeij M, Kandimalla R, Blok AC, van der Aa MM, Zwarthoff EC, Zuiverloon TC: Hypermethylation of the poly comb group target gene PCDH7 in bladder tumors from patients of all ages. J Urol 2013, 190:311-316.

10. Hoque MO, Kim MS, Ostrow KL, Liu J, Wisman GB, Park HL, Poeta ML, Jeronimo C, Henrique R, Lendvai A, et al: Genome-wide promoter analysis uncovers portions of the cancer methylome. Cancer Res 2008, 68:2661—2670.

11. Kandimalla R, van Tilborg AA, Zwarthoff EC: DNA methylation-based biomarkers in bladder cancer. Nat Rev Urol 2013.

12. Kandimalla R, Masius R, Beukers W, Bangma CH, Orntoft TF, Dyrskjot L, van Leeuwen N, Lingsma H, van Tilborg AA, Zwarthoff EC: A 3-plex methylation assay combined with the FGFR3 mutation assay sensitively detects recurrent bladder cancer in voided urine. Clin Cancer Res 2013, 19:4760-4769.

13. Yu J, Zhu T, Wang Z, Zhang H, Qian Z, Xu H, Gao B, Wang W, Gu L, Meng J, et al: A novel set of DNA methylation markers in urine sediments for sensitive/specific detection of bladder cancer. Clin Cancer Res 2007, 13 :7296-7304. 14. Friedrich MG, Weisenberger DJ, Cheng JC, Chandrasoma S, Siegmund KD, Gonzalgo 438 ML, Toma MI, Huland H, Yoo C, Tsai YC, et al: Detection of methylated apoptosis-associated genes in urine sediments of bladder cancer patients. Clin Cancer Res 440 2004, 10:7457-7465. 15. Hoque MO, Begum S, Topaloglu O, Chatterjee A, Rosenbaum E, Van Criekinge W, Westra WH, Schoenberg M, Zahurak M, Goodman SN, Sidransky D: Quantitation of promoter methylation of multiple genes in urine DNA and bladder cancer detection. J Natl Cancer Inst 2006, 98:996-1004.

16. Su SF, de Castro Abreu AL, Chihara Y, Tsai Y, Andreu-Vieyra C, Daneshmand S, Skinner EC, Jones PA, Siegmund KD, Liang G: A panel of three markers hyper- and hypomethylated in urine sediments accurately predicts bladder cancer recurrence. Clin Cancer Res 2014, 20: 1978-1989.

17. Andersson E, Steven K, Guldberg P: Size-based enrichment of exfoliated tumor cells in urine increases the sensitivity for DNA-based detection of bladder cancer. PLoS

One 2014, 9:e94023.

18. Fackler MJ, Lopez Bujanda Z, Umbricht C, Teo WW, Cho S, Zhang Z,

Visvanathan K, Jeter S, Argani P, Wang C, et al: Novel methylated biomarkers and a robust assay to detect circulating tumor DNA in metastatic breast cancer. Cancer Res 2014, 74:2160-2170.

19. Philipp AB, Stieber P, Nagel D, Neumann J, Spelsberg F, Jung A, Lamerz R, Herbst A, Kolligs FT: Prognostic role of methylated free circulating DNA in colorectal cancer. Int J Cancer 2012, 131 :2308-2319.

20. Campan M, Moffitt M, Houshdaran S, Shen H, Widschwendter M, Daxenbichler G, Long T, Marth C, Laird-Offringa IA, Press MF, et al: Genome-scale screen for DNA methylation-based detection markers for ovarian cancer. PLoS One 2011, 6:e28141.

21. Leng S, Do K, Yingling CM, Picchi MA, Wolf HJ, Kennedy TC, Feser WJ, Baron AE, Franklin WA, Brock MV, et al: Defining a gene promoter methylation signature in sputum for lung cancer risk assessment. Clin Cancer Res 2012, 18:3387- 3395. 22. Belinsky SA: Gene-promoter hypermethylation as a biomarker in lung cancer. Nat Rev Cancer 2004, 4:707-717.

23. Kulis M, Esteller M: DNA methylation and cancer. Adv Genet 2010, 70:27 - 56. 469

24. Virmani AK, Muller C, Rathi A, Zoechbauer-Mueller S, Mathis M, Gazdar AF. Aberrant methylation during cervical carcinogenesis. Clin. Cancer Res. 2001, 7(3), pp584-589.

25. Escobar-Hoyos, LF, Yang J, Zhu J, Cavallo JA, Zhai H, Burke S, Roller A, Chen EI and Shroyer KR: Keratin 17 in premalignant and malignant squamous lesions of the cervix: proteomic discovery and immunohistochemical validation as a diagnostic and prognostic biomarker. Modern Pathology 2014, 27(4): pp 621-630.

26. https://www.cancer.gov/types/cervical/patient/cervical-treatment-pdq

27. Olkhov-Mitsel, E and Bapat, B: Strategies for discovery and validation of methylated and hydroxymethylated DNA biomarkers. Cancer Medicine 2012, 1(2): 237-260.

28. Herman, J. G, Graff, J. R., Myohanen, S., Nelkin, B. D. & Baylin, S. B.:

Methylation-specific PCR: a novel PCR assay for methylation status of CpG islands. Proc. Natl Acad. Sci. USA 1996, 93 : 9821-9826.

29. Frommer, M. et al. : A genomic sequencing protocol that yields a positive display of 5-methylcytosine residues in individual DNA strands. Proc. Natl Acad. Sci. USA 1992, 89: 1827-1831. 30. Xiong, Z. & Laird, P. W.: COBRA: a sensitive and quantitative DNA methylation assay. Nucleic Acids Res . 1997, 25: 2532-2534. 31. Gonzalgo, M. L. & Jones, P. A.: Rapid quantitation of methylation differences at specific sites using methylationsensitive single nucleotide primer extension (Ms- SNuPE). Nucleic Acids Res. 1997, 25: 2529-2531.

32. Paul DS, Guilhamon P, Karpathakis A, Butcher LM, Thirlwell C, Feber A, Beck S: Assessment of RainDrop BS-seq as a method for large-scale, targeted bisulfite sequencing. Epigenetics 2014, 9. 33. Singal, R. & Grimes, S. R.: Microsoft Word macro for analysis of cytosine methylation by the bisulfite deamination reaction. Biotechniques 2001, 30: 116-120.

34. Anbazhagan, R., Herman, J. G., Enika, K. & Gabrielson, E. : Spreadsheet-based program for the analysis of DNA methylation. Biotechniques 2001, 30: 110-114.

35. Li, L. C. & Dahiya, R.: MethPrimer: designing primers for methylation PCRs. Bioinformatics 2002, 18: 1427-1431.

36. Eng, J: Receiver Operating Characteristic Analysis: A Primer. Academic Radiology 2005, 12(7): 909-916.

37. Tewhey R, Warner JB, Nakano M, Libby B, Medkova M, David PH,

Kotsopoulos SK, Samuels ML, Hutchison JB, Larson JW, et al: Microdroplet-based PCR enrichment for large-scale targeted sequencing. Nat Biotechnol 2009, 27: 1025- 1031.

38. Guilhamon P, Eskandarpour M, Halai D, Wilson GA, Feber A, Teschendorff AE, Gomez V, Hergovich A, Tirabosco R, Fernanda Amary M, et al: Meta-analysis of IDH-mutant cancers identifies EBFl as an interaction partner for TET2. Nat Commun 2013, 4:2166. 39. Komori HK, LaMere SA, Torkamani A, Hart GT, Kotsopoulos S, Warner J, Samuels ML, Olson J, Head SR, Ordoukhanian P, et al: Application of microdroplet PCR for large-scale targeted bisulfite sequencing. Genome Res 2011, 21 : 1738-1745. 40. R: A language and environment for statistical computing. 485

41. Morris TJ, Butcher LM, Feber A, Teschendorff AE, Chakravarthy AR, Wojdacz TK, Beck S: ChAMP: 450k Chip Analysis Methylation Pipeline. Bioinformatics 2013. 42. Teschendorff AE, Marabita F, Lechner M, Bartlett T, Tegner J, Gomez-Cabrero D, 488 Beck S: A beta-mixture quantile normalization method for correcting probe design bias in Illumina Infinium 450 k DNA methylation data. Bioinformatics 2013, 29: 189-196. 43. Morris TJ, Butcher LM, Feber A, Teschendorff AE, Chakravarthy AR, Wojdacz TK, Beck S: ChAMP: 450k Chip Analysis Methylation Pipeline. Bioinformatics 2014, 30:428-430.

44. Butcher LM, Beck S: Probe Lasso: A novel method to rope in differentially methylated regions with 450K DNA methylation data. Methods 2015, 72:21-28.

Claims

1. A method of diagnosing cervical cancer in an individual comprising:

(a) providing DNA from a sample from the individual;

2. The method according to claim 1, wherein the group of MVPs comprises at least 20 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when at least 10 of those MVPs are methylated.

3. The method according to claim 2, wherein the group of MVPs comprises at least 20 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when at least 15 of those MVPs are methylated.

4. The method according to claim 3, wherein the group of MVPs comprises at least 20 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when all 20 of those MVPs are methylated.

5. The method according to any preceding claim, wherein the group of MVPs comprises all 30 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when at least 15 of those MVPs are methylated.

6. The method according to claim 5, wherein the group of MVPs comprises all 30 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when at least 20 of those MVPs are methylated.

7. The method according to claim 6, wherein the group of MVPs comprises all 30 of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG], and wherein cervical cancer is diagnosed when all 30 of those MVPs are methylated.

8. The method according to any one of claims 2 to 4, wherein the MVPs determined to be methylated include the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG].

9. The method according to claim 4, wherein the MVPs determined to be methylated include the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG] and wherein cervical cancer is diagnosed when each one of the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG] is methylated.

10. The method according to claim 5, wherein the MVPs determined to be methylated include the MVPs identified in SEQ ID NOS 1 to 15 and denoted by [CG] and wherein cervical cancer is diagnosed when each one of the MVPs identified in SEQ ID NOS 1 to 15 and denoted by [CG] is methylated.

11. The method according to claim 6, wherein cervical cancer is diagnosed when each one of the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG] is methylated.

12. A method of classifying a cervical tumor from a patient having cervical cancer as being either a group 1 tumor or a group 2 tumor, the method comprising:

(a) providing a DNA sample from the patient;

(b) determining whether each one of a group of MVPs selected from a panel

(c) applying a binary classification algorithm to the dataset; and (d) classifying the tumor as either a group 1 tumor or a group 2 tumor based on the results of the classification algorithm.

13. The method according to claim 12, wherein the group of MVPs comprises at least 20 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

14. The method according to claim 12, wherein the group of MVPs comprises at least 40 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

15. The method according to claim 12, wherein the group of MVPs comprises at least 60 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

16. The method according to claim 12, wherein the group of MVPs comprises at least 80 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

17. The method according to claim 12, wherein the group of MVPs comprises at least 100 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

18. The method according to claim 12, wherein the group of MVPs comprises at least 120 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

19. The method according to claim 12, wherein the group of MVPs comprises at least 140 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

20. The method according to claim 12, wherein the group of MVPs comprises all 145 of the MVPs identified in SEQ ID NOS 31 to 175 and denoted by [CG].

21. The method according to any one of claims 12 to 20, wherein the group of MVPs includes the MVPs identified in SEQ ID NOS 31 to 50 and denoted by [CG].

22. The method according to any one of claims 17 to 20, wherein the group of MVPs includes the MVPs identified in SEQ ID NOS 31 to 130 and denoted by [CG].

23. The method according to any one of claims 12 to 22, wherein the classification algorithm is based on a support vector machine model (SVM), a k-nearest neighbours model (KNN), a GLMnet model (GLM), or a Random Forest model (RF); preferably wherein for each classification algorithm used the output is the probability (from 0 to 1) of a given tumour belonging to group 2, where 0 is a 100% probability that the sample belongs to group 1, and 1 is a 100% probability that the sample belongs to group 2, and wherein any sample with a probability of equal to or greater than 0.5 is designated as belonging to group 2 and any sample with a probability of than 0.5 is designated as belonging to group 1.

24. The method according to claim 23, wherein the classification algorithm is based on a support vector machine model (SVM) - Linear Kernel (svmLinear), and wherein the parameters are C - 0 to 1, intervals of 0.1.

25. The method according to claim 24, wherein the classification algorithm is based on a k-nearest neighbours model (KNN), and wherein the parameters are K (number of nearest neighbours) - 1,3,5.

26. The method according to claim 24, wherein the classification algorithm is based on a GLMnet model (GLM) (Elastic Net) and wherein the parameters are Alpha - 0,0.2,0.4,0.6,0.8,1; Lambda - 0, 0.01, 0.02, 0.03, 0.04, 0.05.

27. The method according to claim 24, wherein the classification algorithm is based on a Random Forest model (RF) and wherein the parameters are mtry = 1/3 the number of features.

28. The method according to any one of the preceding claims, wherein the step of determining whether each one the MVPs is methylated comprises bisulphite converting the DNA.

29. The method according to any one of the preceding claims, wherein the step of determining whether each one the MVPs is methylated comprises:

a) performing a sequencing step to determine the sequence of MVPs;

b) hybridising DNA to an array comprising probes capable of discriminating

c) performing an amplification step using methylation-specific primers, wherein the status of an MVP as methylated or non-methylated is determined by the presence or absence of an amplified product.

30. The method according to claim 29(a) or 29(b), wherein before the sequencing or hybridization steps an amplification step is performed, wherein loci comprising each MVP are amplified.

31. The method according to claim 29(c) or claim 30, wherein the amplification step is performed by PCR.

32. The method according to claim 30 or claim 31, wherein a capturing step is performed before the sequencing or hybridization steps, and wherein the capturing step involves binding polynucleotides comprising the MVP loci to binding molecules specific to the MVP loci and collecting complexes comprising MVP loci and binding molecules; and wherein:

i. the capturing step occurs before the step of bisulphite converting the DNA;

ii. the capturing step occurs after the step of bisulphite converting the DNA but before the amplification or hybridization steps; or iii. the capturing step occurs after the step of bisulphite converting the DNA and after the amplification step.

33. The method according to claim 32, wherein the binding molecules are oligonucleotides specific for each MVP, preferably DNA or RNA molecules each having a sequence which is complementary to the corresponding MVP.

34. The method according to any one of claims 32 to 33, wherein the binding molecule is coupled to a purification moiety.

35. The method according to claim 34, wherein the purification moiety comprises a first purification moiety and the step of collecting complexes comprising MVP loci and binding molecules comprises binding the first purification moiety to substrates comprising a second purification moiety, wherein first and second purification moieties form an interaction complex.

36. The method according to claim 35, wherein the first purification moiety is biotin and the second purification moiety is streptavidin; or wherein the first purification moiety is streptavidin and the second purification moiety is biotin.

37. The method according to any one of claims 30 to 36, wherein the step of amplifying loci comprising MVPs comprises the use of primers which are independent of the methylation status of the MVP.

38. The method according to any one of claims 32 to 37, wherein the step of amplifying loci comprising MVPs is performed by microdroplet PCR amplification.

39. The method according to any one of the preceding claims wherein the biological sample obtained from the individual is a sample of urine, blood, serum, plasma or cell- free DNA.

40. The method according to any one of the preceding claims, wherein the method achieves a ROC sensitivity of 95% or greater and a ROC specificity of 90% or greater; preferably a ROC sensitivity of 96% and a ROC specificity of 97%.

41. The method according to any one of claims 1 to 40, wherein the method achieves a ROC AUC of 0.90 or greater, 0.91 or greater, 0.92 or greater, 0.93 or greater,

0.94 or greater, 0.95 or greater, 0.96 or greater, 0.97 or greater, 0.98 or greater, 0.99 or

1, preferably the method achieves a ROC AUC of 0.98 or greater.

42. The method according to any one of claims 1 to 40, wherein the method achieves a negative predictive value (NPV) of 90% or greater, 91% or greater, 92% or greater, 93% or greater, 94% or greater, 95% or greater, 96% or greater, 97% or greater, 98% or greater or 99% or 100%.

43. A method of diagnosing cervical cancer in an individual comprising:

(a) obtaining data which identify whether each one of a group of MVPs

wherein the data were obtained by a method comprising:

i. obtaining DNA from the sample; and

ii. determining whether MVPs are methylated in the DNA.

44. The method according to claim 43, wherein the MVPs determined to be methylated include the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG] and wherein cervical cancer is diagnosed when each one of the MVPs identified in SEQ ID NOS 1 to 20 and denoted by [CG] are methylated.

45. The method according to claim 44, wherein the MVPs determined to be methylated include the MVPs identified in SEQ ID NOS 1 to 25 and denoted by [CG] and wherein cervical cancer is diagnosed when each one of the MVPs identified in SEQ ID NOS 1 to 25 and denoted by [CG] are methylated.

46. The method according to claim 45, wherein the MVPs determined to be methylated include the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG] and wherein cervical cancer is diagnosed when each one of the MVPs identified in SEQ ID NOS 1 to 30 and denoted by [CG] are methylated.

47. The method according to any one of claims 1 to 27 and 43 to 46, wherein the step of diagnosing cervical cancer in the individual further comprises:

I. stratifying the stage of the tumor; and/or

II. determining the risk of recurrence of the tumor following primary

surgery or radiotherapy; and/or

III. determining the risk of progression to invasive disease; and/or

IV. determining the likely response to treatment therapy; and/or

V. providing a cervical cancer treatment to the patient.

48. A method of managing the treatment or healthcare regime of an individual, the method comprising:

(a) diagnosing cervical cancer in an individual by performing the method of any one of claims 1 to 11 and 43 to 46; and

(b) stratifying the patient into a cervical cancer treatment or healthcare regime group according to whether the patient has either a group 1 tumor or a group 2 tumor by performing the classification method of any one of claims 12 to 28.

49. A method of treating a cervical cancer patient, the method comprising: (a) stratifying the patient into a cervical cancer treatment regime group according to whether the patient has either a group 1 tumor or a group 2 tumor by performing the classification method of any one of claims 12 to 28; and

(b) treating the patient with one or more cervical cancer treatments.

50. A method of treating cervical cancer in a patient comprising:

(a) obtaining DNA from a sample from the patient;

(b) diagnosing cervical cancer in the patient by performing the method of any one of claims 1 to 27 and 43 to 46;

(c) stratifying the patient into a cervical cancer treatment regime group

according to whether the patient has either a group 1 tumor or a group 2 tumor by performing the classification method of any one of claims 12 to 28; and

(d) administering one or more cervical cancer treatments to the patient.

51. The method according to claim 49 or claim 50 wherein the choice of said one or more cervical cancer treatments is determined by whether the patient has a group 1 tumor or a group 2 tumor.

52. The method according to any one of claims 43 to 51, wherein the step of determining whether each one of the MVPs in the group of MVPs is methylated is performed by the method of any one of claims 28 to 39.

53. The method according to any one of claims 43 to 52, wherein a patient having a group 1 tumor is treated with a cervical cancer therapeutic treatment.

54. The method according to claim 53, wherein the cervical cancer therapeutic treatment comprises one or more of surgery, radiotherapy, a cytotoxic chemotherapy (e.g. cisplatin), an immune-modulation therapy (e.g. administration of Ipilimumab, Nivolumab, Pembrolizumab or another immune checkpoint inhibitor or agonist), a chimeric antigen receptor (CAR) T-cell therapy, a targeted kinase inhibitor therapy (e.g. Lapatinib), an antibody-drug conjugate (ADC) and/or an HPV therapeutic vaccine.

55. The method according to any one of claims 1 to 27, 43 to 47 and 50 to 54, further comprising staging the cervical cancer as stage 0, stage IA, stage IB, stage IIA, stage IIB, stage IIIA, stage IIIB, stage IVA or stage IVB cervical cancer.

56. The method according to any one of the preceding claims, wherein the cervical cancer is a cervical squamous cell carcinoma.

57. The method according to any one of claims 1 to 55, wherein the cervical cancer is a cervical adenocarcinoma.

58. An array capable of discriminating between methylated and non-methylated forms of MVPs; the array comprising oligonucleotide probes specific for a methylated form of each MVP in a MVP panel and oligonucleotide probes specific for a non- methylated form of each MVP in the panel; wherein the panel consists of at least 10 MVPs selected from the MVPs identified in SEQ ID NOS 1 to 30, or wherein the panel consists of at least 20 MVPs selected from the MVPs identified in SEQ ID NOS 31 to 175.

59. An array according to claim 58, provided that the array is not an Infinium HumanMethylation450 BeadChip array, and/or provided that the number of MVP - specific oligonucleotide probes of the array is less than 482,421, preferably 482,000 or less, 480,000 or less, 450,000 or less, 440,000 or less, 430,000 or less, 420,000 or less, 410,000 or less, or 400,000 or less.

60. An array according to claim 58 or claim 59, wherein the panel consists of at least 15 MVPs selected from the MVPs identified in SEQ ID NOS 1 to 30.

61. An array according to claim 60, wherein the panel consists of at least the MVPs identified in SEQ ID NOS 1 to 15.

62. An array according to claim 58 or claim 59, wherein the panel consists of at least 20 MVPs selected from the MVPs identified in SEQ ID NOS 1 to 30.

63. An array according to claim 62, wherein the panel consists of at least the MVPs identified in SEQ ID NOS 1 to 20.

64. An array according to claim 58 or claim 59, wherein the panel consists of at least 25 MVPs selected from the MVPs identified in SEQ ID NOS 1 to 30.

65. An array according to claim 64, wherein the panel consists of at least the MVPs identified in SEQ ID NOS 1 to 25.

66. An array according to claim 58 or claim 59, wherein the panel consists of all 30 MVPs identified in SEQ ID NOS 1 to 30.

67. An array according to claim 58 or claim 59, wherein the panel consists of at least 20 MVPs selected from the MVPs identified in SEQ ID NOS 31 to 175; preferably at least 30 MVPs, at least 40 MVPs, at least 50 MVPs, at least 60 MVPs, at least 70 MVPs, at least 80 MVPs, at least 90 MVPs, at least 100 MVPs, at least 110 MVPs, at least 120 MVPs, at least 130 MVPs, at least 140 MVPs, or all 145 MVPs identified in SEQ ID NOS 31 to 175.

68. An array according to claim 67, wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 50; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 60; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 70; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 80; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 90; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 100; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 110; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 120; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 130; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 140; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 150; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 160; or wherein the panel consists of at least the MVPs identified in SEQ ID NOS 31 to 170.

69. An array according to claim 68, wherein the panel consists of each one of the MVPs identified in SEQ ID NOS 31 to 175.

70. An array according to any one of claims 58 to 66, further comprising one or more oligonucleotides comprising a MVP selected from any of the MVPs defined in SEQ ID NOS 1 to 30, wherein the one or more oligonucleotides are hybridized to corresponding oligonucleotide probes of the array.

71. An array according to claim 70, wherein the one or more oligonucleotides comprise at least 10 MVPs selected from the MVPs identified in SEQ ID NOS 1 to 30.

72. An array according to claim 71, wherein the one or more oligonucleotides comprise the MVPs identified in SEQ ID NOS 1 to 10.

73. An array according to claim 70, wherein the one or more oligonucleotides comprise at least 15 MVPs selected from the MVPs identified in SEQ ID NOS 1 to 30.

74. An array according to claim 73, wherein the one or more oligonucleotides comprise the MVPs identified in SEQ ID NOS 1 to 15.

75. An array according to claim 70, wherein the one or more oligonucleotides comprise at least 20 MVPs selected from the MVPs identified in SEQ ID NOS 1 to 30.

76. An array according to claim 75, wherein the one or more oligonucleotides comprise the MVPs identified in SEQ ID NOS 1 to 20.

77. An array according to claim 70, wherein the one or more oligonucleotides comprise at least 25 MVPs selected from the MVPs identified in SEQ ID NOS 1 to 30.

78. An array according to claim 77, wherein the one or more oligonucleotides comprise the MVPs identified in SEQ ID NOS 1 to 25.

79. An array according to claim 78, wherein the one or more oligonucleotides comprise all 30 of the MVPs identified in SEQ ID NOS 1 to 30.

80. An array according to any one of claims 58, 59 and 67 to 69, further comprising one or more oligonucleotides comprising a MVP selected from any of the MVPs defined in SEQ ID NOS 31 to 175, wherein the one or more oligonucleotides are hybridized to corresponding oligonucleotide probes of the array.

81. An array according to claim 80, wherein the one or more oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 50; or wherein the one or more oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 60; or wherein the one or more oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 70; or wherein the one or more oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 80; or wherein the one or more

oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 90; or wherein the one or more oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 100; or wherein the one or more oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 110; or wherein the one or more

oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 120; or wherein the one or more oligonucleotides comprises at least the MVPs identified in

SEQ ID NOS 31 to 130; or wherein the one or more oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 140; or wherein the one or more

oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 150; or wherein the one or more oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 160; or wherein the one or more oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 170.

82. An array according to claim 81, wherein the one or more oligonucleotides comprises each one of the MVPs identified in SEQ ID NOS 31 to 175.

83. A hybridized array, wherein the array is obtainable by hybridizing to an array according to any one of claims 58 to 66 a group of oligonucleotides each comprising a different MVP selected from any of the MVPs defined in SEQ ID NOS 1 to 30, and wherein the group comprises at least 10 oligonucleotides.

84. A hybridized array according to claim 83, wherein the group comprises at least 15 oligonucleotides; preferably wherein the group comprises at least 20

oligonucleotides, or at least 25 oligonucleotides, more preferably wherein the group comprises all 30 oligonucleotides comprising MVPs defined in SEQ ID NOS 1 to 30.

85. A hybridized array, wherein the array is obtainable by hybridizing to an array according to any one of claims 58, 59 and 67 to 69 a group of oligonucleotides each comprising a different MVP selected from any of the MVPs defined in SEQ ID NOS 31 to 175, and wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 60; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 70; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 80; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 90; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 100; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 110; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 120; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 130; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 140; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 150; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 160; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 170.

86. A hybridized array according to claim 85, wherein the group of oligonucleotides comprises each one of the MVPs identified in SEQ ID NOS 31 to 175.

87. A process for making a hybridized array according to claim 83, comprising contacting an array according to any one of claims 58 to 66 with a group of

oligonucleotides each comprising a different MVP selected from any of the MVPs defined in SEQ ID NOS 1 to 30, and wherein the group comprises at least 10 oligonucleotides.

88. A process for making a hybridized array according to claim 84, comprising contacting an array according to any one of claims 58 to 66 with a group of

oligonucleotides each comprising a different MVP selected from any of the MVPs defined in SEQ ID NOS 1 to 30, and wherein the group comprises at least 15 oligonucleotides, preferably wherein the group comprises oligonucleotides comprising the MVPs defined in SEQ ID NOS 1 to 15.

89. A process for making a hybridized array according to claim 84, comprising contacting an array according to any one of claims 58 to 66 with a group of

oligonucleotides each comprising a different MVP selected from any of the MVPs defined in SEQ ID NOS 1 to 30, and wherein the group comprises at least 20 oligonucleotides, preferably wherein the group comprises oligonucleotides comprising the MVPs defined in SEQ ID NOS 1 to 20.

90. A process for making a hybridized array according to claim 84, comprising contacting an array according to any one of claims 58 to 66 with a group of

oligonucleotides each comprising a different MVP selected from any of the MVPs defined in SEQ ID NOS 1 to 30, and wherein the group comprises at least 25 oligonucleotides, preferably wherein the group comprises oligonucleotides comprising the MVPs defined in SEQ ID NOS 1 to 25.

91. A process for making a hybridized array according to claim 84, comprising contacting an array according to any one of claims 58 to 66 with a group of

oligonucleotides comprising each one of the MVPs defined in SEQ ID NOS 1 to 30.

92. A process for making a hybridized array according to claim 85, comprising contacting an array according to any one of claims 58, 59 and 67 to 69 with a group of oligonucleotides each comprising a different MVP selected from any of the MVPs defined in SEQ ID NOS 31 to 175, and wherein the group of oligonucleotides comprises at least 30 of the MVPs defined in SEQ ID NOS 31 to 175, optionally at least 40 of the MVPs defined in SEQ ID NOS 31 to 175, or at least 50 of the MVPs, or at least 60 of the MVPs, or at least 70 of the MVPs, or at least 80 of the MVPs, or at least 90 of the MVPs, or at least 100 of the MVPs, or at least 110 of the MVPs, or at least 120 of the MVPs, or at least 130 of the MVPs, or at least 140 of the MVPs defined in SEQ ID NOS 31 to 175.

93. A process for making a hybridized array according to claim 92, wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 60; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 70; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 80; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 90; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 100; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 110; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 120; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 130; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 140; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 150; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 160; or wherein the group of oligonucleotides comprises at least the MVPs identified in SEQ ID NOS 31 to 170.

94. A process for making a hybridized array according to claim 93, wherein the group of oligonucleotides comprises each one of the MVPs identified in SEQ ID NOS

31 to 175.

95. A kit comprising an array according to any one of claims 58 to 82.

96. The kit according to claim 95, further comprising a DNA modifying regent that is capable of modifying a non-methylated cytosine in a MVP dinucleotide but is not capable of modifying a methylated cytosine in a MVP dinucleotide, optionally wherein the dinucleotide is CpG.

97. The kit according to claim 96, wherein the DNA modifying regent is a bisulphite reagent.