WO2004018711A2

WO2004018711A2 - Diagnostic test

Info

Publication number: WO2004018711A2
Application number: PCT/GB2003/003637
Authority: WO
Inventors: Du Ming-Qing
Original assignee: University College London
Priority date: 2002-08-24
Filing date: 2003-08-20
Publication date: 2004-03-04
Also published as: WO2004018711A3; AU2003255823A1

Abstract

The present invention relates to the prognosis of patients suffering from cervical intraepithelial neoplasia (CIN). Accordingly, the invention provides methods for diagnosing the susceptibility of such a patient to persistence or progression of the CIN. In particular, the invention provides a method of diagnosing susceptibility to persistence or progression of cervical intraepithelial neoplasia (CIN) in an individual suffering therefrom; the method comprising: d) providing a sample of dyskaryotic cells from said individual and a sample of non-dyskaryotic cells from said individual; b) detecting an allelic deletion in one or more genes selected from FHIT, PR, DLEC 1 and TRIM 29 by comparing the FHIT, PR, DLEC 1 and/or TRIM 29 polynucleotides or proteins present in the samples of step a) derived, respectively, from the non-dyskaryotic and dyskaryotic sample, wherein detection of an allelic deletion between the non-dyskaryotic and dyskaryotic samples is correlated with a susceptibility to persistence or progression of CIN.

Description

DIAGNOSTIC TEST

Field of the Invention

The invention relates to the diagnosis of susceptibility to persistence or progression of cervical intraepithelial neoplasia (CIN).

Background of the Invention

Carcinoma of the uterine cervix is the second most common malignancy among women worldwide and its mortality rate is high [Schoell et al, Semin Surg Oncol 1999; 16: 203-211]. Most cervical carcinomas are believed to derive from pre-neoplastic epithelial lesions known as cervical intraepithelial neoplasia (CIN) [Nguyen & Averette, Semin Surg Oncol 1999; 16: 212-216]. CIN is classified in a three-tier system (CIN 1, 2 and 3) in the United Kingdom [Richart, PatholAnn 1973; 8: 301-328] but divided into low and high-grade lesions in North America [Kurman et al, The 1992 National Cancer Institute Workshop. JAMA 1994; 271 : 1866-1869]. In addition, a separate category describing more minor cell changes than CTNl is used both in the United Kingdom (borderline changes) [Parham et al Cytopathology 1992; 3: 85-91] and North America (atypical squamous cells of undetermined significance, ASCUS) [Isacson & Kurman, Contemp Ob/Gyn 1996; 6: 67-72; Korn, Infect Med 1996; 13 : 405-411] . Borderline changes/ASCUS and CIN 1 correspond to low-grade lesions, while CIN2, CLN3 and carcinoma in situ (CIS) correlate with high-grade lesions [Nguyen & Nordqvist, Semin Surg Oncol 1999; 16: 217-221]. Cytologically, CIN1 lesions are characterised by squamous epithelial cells showing mild dyskaryosis, with irregular nuclei occupying less than one half the area of the cytoplasm; CIN2 lesions feature moderately dyskaryotic cells with irregular hyperchromatic nuclei occupying two-thirds of the cytoplasmic area; and CIN3 lesions are composed of severely dyskaryotic cells with large irregular hyperchromatic nuclei occupying more than two-thirds of the cytoplasm.

The Papnicolaou (Pap) smear test has been used as a screening method to identify these pre-invasive lesions and has permitted a dramatic reduction in the cervical cancer mortality rate of up to 70% [Schoell et al, Semin Surg Oncol 1999; 16: 203-211]. However, interpretation of individual CLN lesions, particularly early lesions, and prediction of their clinical behaviour by cytological examination of cervical smears alone is difficult.

Clinically, CLN lesions may have variable behaviour despite morphological homogeneity within the same subgroup. Approximately, two-thirds of CIN1 and CIN2 will fail to progress or will regress spontaneously, whereas one-third will progress into CIN3 or invasive lesions [Cirisano, Semin Surg Oncol 1999; 16: 222- 227]. Currently, there are no biochemical or molecular markers, which can distinguish these lesions with different clinical behaviour. Patients with CIN are subjected to prolonged follow-up with periodical colposcopsy and biopsy [Cirisano, Semin Surg Oncol 1999; 16: 222-227] .

Human papillomaviruses (HPNs) are the most diverse group of DΝA viruses involved in human disease and more than 100 types have been identified based on • their sequence homology of E6 and LI open reading frames. 30 HPNs have been detected in the genital tract and these are further classified as low-risk, intermediate- risk and high-risk types according to their association with different grades of intraepithelial and invasive lesions. The low-risk viruses such as HPN6 and 11 are associated with benign genital warts and CIΝ1 lesions, and are rarely found -in CIΝ2, CIN3 and invasive lesions. By contrast, intermediate-risk (HPN 31, 33, 39, 52 and 58) and high-risk types (HPN16, 18, 45 and 56) are more often seen in CLΝ2, CIN3 and invasive lesions than CIN1 and borderline lesions.

Epidemiological studies strongly indicate that HPN infection is a critical factor in the development of cervical cancer. This is supported by molecular studies of the oncogenic activity of molecules associated with the virus. Two viral oncoproteins namely E6 and E7 have been shown to play an important role in the process of malignant transformation. Both proteins are consistently expressed in cervical cancer cell lines and can immortalise primary cervical epithelial cells and human keratinocytes in vitro. Transgenic mice with E6/E7 of HPN16 show hyperplastic and dysplastic squamous lesions. E7 proteins bind and inactivate the function of the retinoblastoma tumour suppresser protein pRb by targeting it for degradation via the ubiquitin pathway. Similarly, E6 protein has been shown to bind and degrade another tumour suppresser protein p53 and others through E6-associated protein (E6AP), a ubiquitin E3 ligase. Importantly, the E6 protein of high risk HPN, particularly HPN 16 and 18, binds more effectively to p53 and is more potent in its ability to degrade p53. In addition, E6 induces telomerase activity via transcriptional activation of telomerase reverse transcriptase (hTERT) expression. Transcription of E6 and E7 genes depends on the viral E2 protein, which acts as a repressor of early gene transcription and is involved in the control of viral DΝA replication. The E2 gene is commonly disrupted as a result of viral integration, leading to loss of its function.

Despite the strong implication of high risk HPN infection in the genesis of cervical cancer, the value of HPN typing in prediction of the prognosis of individual CIΝ lesions is largely limited due to the following facts: ^• 1) HPN typing does not clearly distinguish HPN infection from HPN associated CTΝ lesions as high risk HPNs are also present in 2% of healthy women;

2) Despite the high prevalence of high risk HPN infection in invasive and high grade CIΝ lesions, they are also present in 10% of borderline and 20% of CLΝ1 lesions;

3) HPN infection is often transient, becoming undetectable by standard molecular biology techniques in 80% of infected individuals after a period of about 12 months;

4) Persistence or progression of HPN infection beyond 12 months is seen in 20% of infected individuals, but less than 1% of these individuals develop cervical cancer within 15 years of infection.

Summary of the Invention

This invention is directed to the diagnosis of the likely outcome of cervical intraepithelial neoplasia (CIΝ). Approximately two-thirds of CIΝ1 and CIΝ2 will fail to progress or will regress spontaneously, whereas one-third will progress into CIN3 or invasive lesions.

The inventor has shown that loss of heterozygosity at microsatellite loci in dyskaryotic cells can be linked to the persistence and/or progression of cervical intraepithelial neoplasia (CIN). These microsatellite loci are believed to correspond to the location of genes involved in CIN, for example tumour-suppressor genes. The present invention therefore provides a diagnostic tool to allow a determination of the likelihood that progression or persistence of neoplasia will occur in an individual suffering from CIN. The diagnostic methods of the present invention have the further advantage that such a determination can be made at a relatively early stage in the progression of the disease.

Accordingly, the invention provides a method of diagnosing susceptibility to persistence or progression of cervical intraepithelial neoplasia (CIN) in an individual suffering therefrom; the method comprising: a) providing a sample of dyskaryotic cells from said individual and a sample of non-dyskaryotic cells from said individual; b) detecting an allelic deletion in one or more genes selected from FHIT, PR, DLEC 1 and TRIM 29 by comparing the FHIT, PR, DLEC 1 and/or TRIM 29 polynucleotides or proteins present in the samples of step a) derived, respectively, from the non-dyskaryotic and dyskaryotic sample, wherein detection of an allelic deletion between the non-dyskaryotic and dyskaryotic samples is correlated with a susceptibility to persistence or progression of CIN. Also provided is a method of diagnosing susceptibility to persistence or progression of cervical intraepithelial neoplasia (CLN) in an individual suffering therefrom, the method comprising: a) providing genomic DNA from a sample of dyskaryotic cells from said individual and from a sample of non-dyskaryotic cells from said individual; b) detecting a loss of heterozygosity in one or more microsatellite DNA markers by comparing the genomic DNA of step a) derived, respectively, from the non- dyskaryotic and dyskaryotic sample, wherein detection of the loss of heterozygosity in the dyskaryotic sample is correlated with a susceptibility to persistence or progression of cervical intraepithelial neoplasia (CIN). The invention also provides kits suitable for use in the methods of the invention. Accordingly the invention provides a kit comprising a panel of two or more pairs of primers wherein each pair of primers is suitable for amplifying a microsatellite DNA marker selected from D3S1300, D3S1260, Dl 1S35 and Dl 1 S528. Also provided is a kit comprising a panel of two or more specific binding agents wherein each binding agent is capable of distinguishing between the normal and allelic deletion forms of a polynucleotide or protein selected from FHIT, PR, TRIM29 and DLEC 1. Description of the Figures

Figure 1 shows GeneScan results from non-dyskaryotic (N) and dyskaryotic (T) cells. LOH +ve = This individual is positive for loss of heterozygosity at this locus. LOH -ve - This individual is negative for loss of heterozygosity at this locus. Homozygous = This individual is homozygous at this locus.

Figure 2 shows the frequency of LOH of microsatellite DNA markers in the different classes of CTN lesions in Example 1. Figure 3A shows the frequency of LOH at >1 or >2 or >3 of the 12 markers examined, in different grades of CIN lesions in Example 1. Figure 3B is a typical case illustration.

Figure 4 shows the correlation between loss of heterozygosity for a selection of loci and the disease follow up in Example 1 : either disease free or disease persistent/progression. Stars indicate those loci at which loss of heterozygosity was seen in a significantly greater proportion of patients that then went on to show disease persistence or progression than in those that became disease free.

Figure 5 shows the frequency of LOH in the three classes of CIN lesions at 12 microsatellite markers investigated in Example 2. Figure 6 shows the frequency of LOH at >1 or >2 or >3 of the 12 markers examined, in different grades of CIN lesions in Example 2.

Figure 7 shows the correlation between loss of heterozygosity for a selection of loci and the disease follow up in Example 2: either disease free (DF) or disease persistent or progression (DP). Data is representative of 117 cases screened, including 25 CINl, 35 CIN2 and 57 CLN3. Stars indicate those loci at which loss of heterozygosity was seen in a significantly greater proportion of patients that went on to show disease persistence or progression than in those that became disease free.

Figure 8 shows the correlation between loss of heterozygosity for four loci and the disease follow up in Example 2: either disease free (DF) or disease persistent or progression (DP). Data is representative of 191 cases screened, including 61 disease free (9CIN1, 20 CIN2 and 32 CIN3) and 130 disease persistent/progression (59 CINl, 44 CIN2 and 27 CIN3). Figure 9 compares the percentage of patients from the group of 191 individuals showing loss of heterozygosity at any 1, any 2, any 3 or all of the four loci D3S1300, D3S1260, D11S35 and D11S528 in different grades of lesions. DF = ^• patients who became disease free after treatment. DP = patients who showed disease progression or persistence.

Figure 10 shows the identification of loci for CIN prognosis A: Frequencies of LOH at 12 microsatellite markers between CIN lesions showing disease free (DF) or disease persistence (DP) after initial treatment from Example 3. The incidence of LOH at D3S1260 (3p22.2) and Dl 1S528 (1 lq23.3) is significantly higher in DP than DF group (p<0.05 for both). D3S1300 (3pl4.2) and Dl 1S35 (1 lq22.1) exhibited the next highest statistical difference between the two groups, albeit not significant. Using a stepwise statistical analysis testing various combination of the 12 markers, the above four markers collectively gave the highest statistical significance (p<0.03) between the two groups and were further analyzed as detailed in Figure 10B .

B: Prognostic value of combined LOH analysis at the above 4 loci. By stepwise testing the sensitivity and specificity of the combined 4 LOH markers in CIN prognosis, i.e. 1 or 2 or 3 or 4 of the 4 loci showing LOH, the best cut-off point is when 2 of the 4 loci showing LOH are applied. At this threshold, between 47-63% of CIN lesions in the DP group but none in the DF group show LOH. Figure 11 shows the results from Example 4 A: Correlation of LOH with CIN grade in Example 4. LOH at D3S1300, D3S1260 and Dl 1S35 correlated positively with the grade, while the incidence of LOH at Dl 1S528 is similar among different CIN lesions. B: Comparison of LOH between CIN lesions showing disease free (DF) or disease persistence (DP) in Example 4.

C: Prognostic value of combined LOH analyses at the above 4 loci. Between 24- 54% CIN lesions of the DP groups can be identified with 100% specificity at the threshold when 2 of the 4 loci showing LOH is applied. Figure 12 shows an example of LOH analysis. The diagnostic biopsy (CIN3) of case 20 shows LOH at D3S1300 and Dl 1S35. Despite treatment, the patient continually presented CIN3. The follow-up biopsy 17 months after diagnosis displays additional LOH at D3S1260. N: normal cells; T: tumor cells.

Detailed Description of the Invention

The present invention lies in changes which occur in some DNA regions in dyskaryotic cells of individuals having CIN. By analysing whether a loss of heterozygosity (LOH) has occurred in specific microsatellite regions, or whether an allelic deletion has occurred in specific genes, the susceptibility of the individual to persistence or progression of neoplasia may be determined.

Generally, the development of human cancer results from the clonal expansion of genetically modified cells, which have acquired a selective growth advantage through accumulated alterations of proto-oncogenes and/or tumor suppressor genes. Somatic inactivation of tumor suppressor genes is usually achieved by intragenic mutations in one allele of the gene and by loss of a chromosomal region spanning the second allele. The steps that lead to homozygosity of a mutant suppressor allele usually involve the flanking chromosomal regions as well. Accordingly, DNA markers mapping to nearby chromosomal sites, which may have shown heterozygosity prior to these steps, will suffer a parallel reduction to homozygosity (or loss of heterozygosity - LOH). Indeed, the repeated observation of LOH of a specific chromosomal marker in cells from a particular type suggests the presence of a closely mapping tumor suppressor gene, the loss of which is involved in tumor pathogenesis.

Cervical Intraepithelial Neoplasia

In the context of the present invention, persistent cervical intraeptithelial neoplasia (CIN) is a CIN which fails to regress spontaneously. For example, the CIN may still be detectable after a period of 6 months, 1 year, 2 years, 3 years, 5 years or more. Progression of CIN refers to a change to a higher classification of neoplasia, for example an increase from CLNl to CLN2, CIN3 or carcinoma, from CLN2 to CIN3 or carcinoma, or from CIN3 to carcinoma. According to the present invention, the susceptibility to persistence or progression of CIN may be determined based on the absence of further treatment.

Alternatively, the susceptibility to persistence or progression of CIN even after treatment may be determined. Treatment encompasses any standard procedure for treatment of CTN. For example, CINl and CIN2 may be treated by surface destruction, for example laser, such as CO₂ laser, vaporisation under colposcopic control after prior histological clarification. Such an approach may be used in the case of benign findings such as papilloma, CINl or CIN2 with ectocervical location, completely visible, after prior biopsy and in a cooperative patient. Alternatively, more serious CLN may be treated by surgical excision in healthy tissue, for example, laser vaporisation after histological exclusion of an invasive lesion, skinning vulvectomy or simple vulvectomy. This may be carried out by loop conisation such as loop excision or large loop excision of the transformation zone (LLETZ), or by conization using a loop, laser or scalpel. This approach is more commonly used in CIN2 (endocervical, CIN3, adenocarcinoma in situ, or in persistent CINl or CIN2 with endocervical extension. In addition to medically performed treatments such as , surgery, trichloroacetic acid, cryotherapy and electrosurgery, treatment includes self treatment, for example using Podophyllotoxin, Imiquimod cream or an interferon beta gel.

Dyskaryotic and non-dyskaryotic cells

Any method known in the art may be used to distinguish cells showing the premalignant and malignant stages of cervical cancer such as dysplasia (described herein as dyskaryotic cells) from normal cervical cells.

The Pap smear test consists of collecting cells from the cervix and vagina, spreading them onto a glass slide, fixing and staining the cells, and analyzing them under a microscope. Cytological features that distinguish dyskaryotic cells, such as mitotically active cells showing enlarged hyperchromatic nuclei and increased nuclear/cytoplasmic ratio, may be identified by visually scanning the entire slide.

Dyskaryotic cells may also be identified by a biochemical screening method, for example by screening for proliferation markers including those that regulate the cell cycle, such as Ki67, cdc6 and Mcm5 and HPN components. Alternatively, cells may be screened for markers of angiogenesis, such as endothelial cell-specific surface proteins, secretory proteins, growth factors, etc. Essentially, the presence in cervical epithelial cells of any biochemical marker, particularly markers of apoptosis and/or angiogenesis, the presence of which in subsuperficial cervical epithelium is correlated with dysplasia can be used to obtain dyskaryotic cells for use in the methods of the present invention. Any other method known in the art suitable for screening cells for features of dysplasia may be used to identify dyskaryotic cells for use in the methods of the present invention.

Non-dyskaryotic cells as used herein are any cells from the individual which do not show features or characteristics of dysplasia as described above. Non- dyskaryotic cells may be obtained from a separate tissue sample from the individual, such as a blood, saliva, cheek cell, epithelial or hair root sample. Non-dyskaryotic cells may also be obtained from the region of the cervix at the same time as the dyskaryotic cells and may be separated from the dyskaryotic cells using a screening method as described above. Cells may be obtained from the region of the cervix by cervical (Pap) smear or any other suitable technique.

Microsatellite markers

Each of the polymorphic microsatellite DNA markers used according to the present invention is detectable with a pair of specific primers having a sequence that is complementary to a genomic DNA sequence flanking the 5' end and the- 3' end respectively of a highly polymorphic microsatellite genomic DNA segment comprising a polymer of oligonucleotide repeats, such as dinucleotide or trinucleotide repeats.

One or more microsatellite DNA markers may be analysed per individual. Preferably more than one marker, such as 2, 3, 4, 5, 10, 15 or more markers may be analysed per individual. For example, one, two or more or all of the markers D3S1260, D3S1300, D3S1285, D3S1289, D3S1566, D3S1611, D5S406, D6S105, D6S265, D6S277, D11S35 and D11S528 may be analysed in an individual. In one aspect, one, two, three or all four of the markers D3S1260, D3S1300, Dl 1S35 and D 11 S528 may be analysed in an individual. In a preferred aspect, at least the four markers D3S1260, D3S1300, Dl 1S35 and Dl 1S528 are analysed in an individual. The microsatellite DNA marker names used herein are the scientific conventional names for which specific pairs of primers have been defined permitting amplification of them, each of the said primers being also useful as a specific probe for detecting the corresponding microsatellite DNA marker. The full sequences of the whole microsatellite DNA markers as well as the full sequences of the amplicons generated using these markers are publicly available from electronic databases such as STS Bank and GenBank (http ://www.ncbi .nlm.nih.gov/GenbanlJindex.html) .

By way of example, suitable loci, locations and primers may be selected from:

A polymorphic microsatellite DNA marker is used to amplify the microsatellite DNA segment which may then be identified by its specific length, for example in a polyacrylamide gel electrophoresis in the presence of urea. The length of the microsatellite marker is determined by the number of oligonucleotide repeats, such as dinucleotide or trinucleotide repeats, that it contains. The number of repeats, and therefore the length, of many microsatellite markers may vary between individuals and between alleles with the genome of an individual.

An individual may carry two identical copies of a microsatellite marker, or two copies having different numbers of repeats. The genotype of an individual for this marker can be visualised by, for example, amplifying the DNA in the region of a microsatellite marker (such as by PCR), separating any DNA amplified on the basis of length (for example in a polyacrylamide gel electrophoresis in the presence of urea, or in a agarose gel), and visualising the DNA (for example using ethidium bromide or by using labelled primers).

Such an assay may be carried out for a single microsatellite DNA marker, or may be used in a high throughput assay by performing multiplex PCR in a single tube. The procedure may be further adapted by automated detection of loss of heterozygosity. An individual carrying two identical copies of a microsatellite marker

(homozygous) will show a single type'of amplified DNA of a particular size. An individual carrying two copies of a microsatellite marker having different numbers of repeats (heterozygous) will show two types of amplified DNA of different sizes.

The analysis of the microsatellite marker may or may not determine the exact identity/sequence of the nucleotides in the marker. It is sufficient merely to analyse the marker in a manner that allows the determination of the length or size of the marker, without a need to determine the identity of the nucleotides. Determination of the exact length of the microsatellite sequence is not required if it is clear whether the individual is homozygous or heterozygous for that particular microsatellite marker. A suitable microsatellite marker for use in a method of the present invention may be identified by determining whether loss in heterozygosity at that candidate microsatellite marker (i) is associated with CIN and (ii) correlates with the persistence and/or progression of the neoplasia, and thereby determining whether the microsatellite marker can be used to diagnose susceptibility to persistence or progression of CIN.

Any microsatellite DNA locus that shows the required association and correlation may be screened for in a method according to the present invention. A - number of chromosomal regions have shown loss of heterozygosity (LOH) in cervical cancer and they include 3pl4.1-p22, 4pl6, 4q21-35, 5pl3-15, 6p213-22, 6q21-25, l lpl5, l lq23, 13ql2.3-ql3, 17pl3.3 and 18ql2.2-22. Preferably, the locus is one that shows loss of heterozygosity in individuals with CINl . Such a locus may be particularly useful in determining the long-term prognosis of an individual who is currently only showing minor effects of CIN, and may be important in determining whether that individual is likely to progress to CIN2 or 3, or to invasive cancer.

A microsatellite DNA locus that is shown to have prognostic value for CIN lesions may be located near or adjacent to a tumor suppressor gene. Identification of suitable microsatellite DNA loci in this way can be used to map and identify a candidate tumor suppressor gene that may be involved in the progression or persistence of the neoplasia.

Loss of heterozygosity is typically determined by comparing the length of the microsatellite marker or of a region of the marker in dyskaryotic and non-dyskaryotic cells of an individual. If the non-dyskaryotic cells of an individual show heterozygosity in a particular microsatellite marker, that marker may be used in the diagnosis of the susceptibility of that individual to persistence or progression of neoplasia. By analysing the DNA of dyskaryotic cells obtained from the individual, a separate genotype for these cells may be obtained. This may be the same or different to that obtained from the non-dyskaryotic cells.

If an individual shows heterozygosity in non-dyskaryotic cells but homozygosity in dyskaryotic cells, that individual is said to show loss of heterozygosity (LOH). This may be the result of loss of a chromosomal region at or near the microsatellite DNA locus.

Loss of heterozygosity at one or more particular microsatellite loci may be used to determine whether the individual is likely to be susceptible to persistence or progression of cervical intraepithelial neoplasia. In particular, loss of heterozygosity at one or more of the loci D3S1300, D3S1260, D11S35 and D11S528, for example at any one, two or three or at all four of these loci, may help determine the prognosis of the individual.

The present invention therefore provides screening methods to determine the susceptibility to persistence or progression of CIN. A screening method of the invention may comprise the steps of:

(a) providing genomic DNA from a sample of dyskaryotic cells from an individual suffering from CIN and from a sample of non-dyskaryotic cells from said individual; and (b) detecting any loss of heterozygosity in a microsatellite DNA marker by comparing the genomic DNA of step (a) derived, respectively, from the dyskaryotic and non-dyskaryotic sample, wherein detection of the loss of hetrozygosity is correlated with a susceptibility to persistence or progression of CIN.

Suitable methods for obtaining samples and for measuring LOH are described above.

In one aspect, the genomic DNA is amplified by at least one pair of primers. The microsatellite DNA marker amplified may be any marker that shows association with CIN and correlates with the persistence and/or progression of the neoplasia. An individual is screened for loss of heterozygosity at a number of loci, for example at each of the loci D3S1300, D3S1260, D11S35 and D11S528.

Gene targets

Bioinformatic analysis of the genomic regions at D3S1300, D3S1260, Dl 1 S35 and Dl 1 S528 has identified genes that may be the target of deletions at these locations. These genes may therefore be involved in the progression or persistence of CIN.

LOCI 52071 (an EST clone), FHIT (fragile histine triad gene) and NPCR (nasopharyngeal carcinoma related protein) are at or in the vicinity of D3S1300. FHIT, sparrning the FRA3B fragile site and the breakpoint of t(3;8) of familial renal carcinoma, has been proposed as a tumor suppresser gene. Deletion and reduced/absent expression of the FHIT gene occur in a wide range of human cancers including cervical cancer. In CIN, LOH and reduced/absent FHIT expression have been shown to positively correlate with the histological grade and correlate each other. Our results reinforce the association of FHIT gene deletion with progression of CIN and further indicate that the gene deletion has prognostic value in the persistence or progression of the CIN. Since FHIT is frequently deleted in a variety of human cancers including cervical cancer, it is most likely the target of deletion detected as loss of heterozygosity at D351300.

Among the genes in the vicinity of D3 SI 260 including ORCTL (organic cationic transporter-like 4), DLEC1 (deleted in lung and esophageal cancer 1) and XYLB (Xylukokinase homolog). DLEC1 may be relevant. AD031 , PR (progesterone receptor) and TRPC6 (transient receptor potential cation channel, superfamily C, member 6) are at or in the vicinity of Dl 1S35. PR is likely to be^' most relevant given its role in inhibition of human endometrial cancer. PR encodes two isoforms, PR-A and PR-B, functioning as ligand activated transcriptional factor. PR-A and PR-B have different transcription activation properties and play distinctive role in different tissues. PR-B acts as a transcriptional activator and is important for normal proliferative response of mammary gland to progesterone, while PR-A functions as transcriptional repressor and is critical for progesterone dependent reproductive responses in uterus and ovary. PR-A represses the activity of PR-B and other steroid receptors including oestrogen receptor alpha, which may underlie the mechanisms of progesterone mediated antiproliferative effect in endometrial cancer. Deletion of the PR gene and loss of its expression are associated with aggressive endometrial cancer and epithelial ovarian tumors. In cervical cancer, reduced PR expression has been found in cervical carcinoma in comparison with normal cervix. Taken together, PR is likely the target of deletion at Dl 1S35. This is in line with the suggestion by both epidemiological and laboratory studies that steroid hormones play a role in development of squamous carcinoma of the cervix. Finally, TRIM29 (tripartite motif-contaiiiing 29) and several hypothetical genes are at D 11 S528. TRIM29 has. multiple zinc finger motifs and a leucine zipper motif and may act as a transcriptional regulatory factor.

The genes FHIT, PR, TRIM29 and DLEC1 may therefore be involved in the persistence and/or progression of CIN. Changes at the locations of these genes, for example deletions of, in, or close to these genes that impair their expression or activity may be responsible for the correlation described herein between loss of heterozygosity at D3S1260, D3S1300, Dl 1S35 and Dl 1S528, and the persistence and progression of CIN. Such a deletion may affect one or both of the two genomic copies of the gene.

The present invention therefore provides methods to determine whether any allelic deletion has occurred in one or more of these four genes in dyskaryotic cells from a CIN patient, when compared with the same genes in non-dyskaryotic cells.

The invention provides a method for diagnosing the susceptibility to persistence or progression of CIN in an individual suffering therefrom, comprising detecting the presence of an allelic deletion in one or more, for example any one, two or three or at all four of the genes selected from FHIT, PR, DLEC1 and TRIM29 by comparing genomic DNA derived from dyskaryotic cells from the individual with genomic DNA derived from non-dyskaryotic cells of the individual. Detection of such an allelic deletion in one of more of these genes indicates that the individual may be susceptible to persistence or progression of CIN.

Such a screening method may alternatively be carried out by comparing a polynucleotide such as DNA, lriRNA or cDNA, or a protein produced by one of these genes, obtained from dyskaryotic cells, with the equivalent protein or expression product obtained from non-dyskaryotic cells.

Suitable cells and samples for use in such methods may be obtained or derived as explained herein.

An allelic deletion is typically detected by comparing the full length sequence of a polynucleotide or protein derived from normal, non-dsykaryotic cells of an individual with the equivalent polynucleotide or protein derived from dyskaryotic cells of the individual.

An allelic deletion may consist of a deletion of all or part of a gene from the genome of the dyskaryotic cells.

An allelic deletion may consist of a deletion of all or part of one copy of a gene from the genome of the dyskaryotic cells. The remaining copy of the gene may be subject to mutation or alteration. An allelic deletion may stop expression of a gene in the dyskaryotic cells, or may lead to lack of function or abnormal functioning of the expressed gene product. The allelic variant may be detected by any suitable method. In one embodiment, an allelic deletion at the FHIT, PR, TRIM29 or DLEC1 locus may be detected by looking for a loss of heterozygosity at a microsatellite marker located at or near that locus. A suitable microsatellite marker may be one of those described above, or may be another marker located at or near the required gene locus. Methods for assessing a loss of heterozygosity are described above.

Alternatively, an allelic deletion may be identified by looking at the gene itself or its expression product. Where the deletion affects both copies of the gene, this will be seen in a total lack of the gene and its expression product in the dyskaryotic cells. Where the deletion affects a single copy of the gene, for example where the remaining copy of the gene contains some other mutation or abnormality, the number of genomic copies of the gene that are present may be quantified. The detection of an allelic deletion according to the invention may comprise contacting a polynucleotide or protein derived from dyskaryotic cells with a specific binding agent capable of binding a normal polynucleotide or protein derived from non-dyskaryotic cells of the same individual, and deteraiining whether the agent binds to the polynucleotide or protein, wherein lack of binding of the agent indicates the presence of an allelic deletion in the dyskaryotic cells.

A specific binding agent is an agent that binds with preferential or high - affinity to the normal protein or polypeptide as found in the non-dyskaryotic cells but does not bind or binds with only low affinity to other polypeptides or proteins. Preferably the binding agent bind with preferential or high affinity to such a normal protein or polynucleotide, but does not bind, or binds with significantly lower affinity to a mutant form of the protein or polynucleotide comprising an allelic deletion. For example, the binding agent may bind to a region of the protein or polynucleotide that is deleted in the allelic deletion form.

The specific binding agent may be capable of specifically binding the amino acid sequence encoded by a variant sequence. For example, the agent may be an antibody or antibody fragment. The detection method may be based on an ELISA system.

The specific binding agent may be a probe or primer. The probe may be a protein (such as an antibody) or an oligonucleotide. The probe may be labelled or may be capable of being labelled indirectly. The binding of the probe to the polynucleotide or protein may be used to immobilise either the probe or the polynucleotide or protein.

In one embodiment a PCR primer is used that primes a PCR reaction only if it binds a polynucleotide containing the normal allele, for example a sequence- or allele-specific PCR system, and the presence of the allelic deletion may be determined by detecting the PCR product. Such a method may use oligonucleotide primers which bind to areas on the polynucleotide deriving from the non-dyskaryotic cells, allowing amplification of a polynucleotide from that region. However the presence of a deletion or other mutation within this region, particularly if it encompasses the binding site of one or both primers, may disrupt amplification. This may be seen as the amplification of a polynucleotide of different size to that produced from non-dyskaryotic cells, or as a failure to amplify anything using those primers based on polynucleotide obtained from dyskaryotic cells. Thus, amplification of a normal length polynucleotide will only occur with a polynucleotide that contains the full-length or normal sequence, and therefore the detection of a shorter or absent product may be used to determine the whether an allelic deletion has occurred at that locus.

A PCR approach can therefore be used where a single copy of the gene of interest is deleted in the dyskaryotic cells. That is, the number of copies of the gene in a cell may be determined by a quantitative PCR method. The "quantity" of the amplified PCR product of interest may be compared against, for example, the "quantity" of another gene from the same cell that is known to be present in two copies. Thus a difference in quantity of the two PCR products may be determined, indicating that one of the DNAs was present in a greater quantity than the other in the cell sample. Methods of quantitative PCR are well known in the art.

The presence of the allelic variant may be determined using a fluorescent dye and quenching agent-based PCR assay such as the Taqman PCR detection system.

The method may be an RFLP based system. This can be used if the presence of the allelic deletion in the polynucleotide creates or destroys a restriction site that is recognised by a restriction enzyme.

The presence of the allelic deletion may be determined based on the change which the presence of the allelic deletion makes to the mobility of the polynucleotide or protein during gel electrophoresis. In the case of a polynucleotide single-stranded conformation polymorphism (SSCP) or denaturing gradient gel electrophoresis

(DDGE) analysis may be used. In another method of detecting the allelic deletion a polynucleotide comprising the polymorphic region is sequenced across the region which contains the allelic deletion to determine the presence of the allelic deletion. The sample is typically processed before the method is carried out, for example DNA extraction may be carried out. The polynucleotide or protein in the sample may be cleaved either physically or chemically, for example using a suitable enzyme. In one embodiment the part of polynucleotide in the sample is copied or amplified, for example by cloning or using a PCR based method prior to detecting the allelic deletion(s).

Screening for HPV infection

In one embodiment, the individual to be screened may suffer from human papillomavirus (HPN) infection. For example, the individual may suffer from an infection by a high risk HPN such as HPV 16, HPN 18, HPN 45 or HPN 56. The screening method of the invention may therefore include the additional step of screening the individual for the presence of an HPN infection. A negative finding when screening for a high risk HPN indicates that the presence of a serious precancerous stage or a carcinoma is extremely unlikely. Suitable methods include colposcopy and molecular biology techniques. HPN 16 infection significantly correlates with LOH at each of the four markers D351300, D351260, Dl 1535 and Dl 15528, particularly D3S1300. HPN16 integrations in cervical cancers preferentially target common fragile sites including FRA3B (3pl4.2) where the FHIT gene and D3S1300 locate, and are accompanied by deletion of cellular genes. Thus, LOH at these loci may be directly attributed to HPN 16 infection.

The molecular biology techniques for HPN detection differ in their sensitivity. The experience of a particular laboratory is often critical for reliable results (particularly with PCR techniques). The classic methods of viral diagnosis such as electron microscopy, cell cultures, and certain immunological methods are not suitable for HPN detection. HPN cannot be cultured in cell cultures. The established method for viral detection as a matter of routine is the hybridization of nucleic acids, either using hybrid capture microplate assay (HC II) or polymerase chain reaction (PCR). The hybrid capture II test (Digene, USA) detects even 1 pg of HPN DΝA/mL; its sensitivity and specificity are almost comparable to PCR. The advantages of this method are the relatively simple handling and good reproducibility of results, which make this test the best standardized HPN detection method. Identification of the exact HPN type is not possible, but only "low-risk" (6, 11, 42, 43, 44) and "high-risk" (16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 68) HPN genotype groups are detected. In PCR, amplification of the viral DΝA occurs first. A sensitivity exceeding that of the hybrid capture can be achieved in appropriately specialized laboratories. Nevertheless, variations in findings among the various laboratories are considerable in some cases. HPN DΝA detection by PCR at a facility specialising in this techniqiie is the method of choice for numerous scientific studies. Primers for the identification or detection of HPN types include:

hi one embodiment, therefore, a screen for the presence of HPN infection and a screen for loss of heterozygosity may be used in a two part test, either at the same time or separately. If the screens are carried out separately, the second screen may be carried out only if the first screen has a positive result in that individual. For example, an individual may only be screened for loss of heterozygosity only if they have already shown a positive result for HPN infection. In a further embodiment, an individual may be screened for loss of heterozygosity if they have shown a persistent HPN infection, i.e. an HPN infection which is still detectable by standard molecular biology techniques more than a year after infection.

Kits

The present invention further provides kits suitable for use in the methods of the invention.

Kits are provided which comprise means for amplifying one or more of the microsatellite markers described herein. For example, the kit may comprise pairs of primers suitable for amplifying the microsatellite markers. In particular, the kit may comprise two or more pairs of primers where each pair of primers is suitable for the amplification of a microsatellite DΝA marker. Primer pairs may be suitable for amplifying, for example, the markers D3S1300, D3S1260, Dl 1S35 and Dl 1S528. In a further aspect, a kit may comprise a panel of pairs of primers which are capable of amplifying all of the markers D3S1300, D3S1260, Dl 1S35 and Dl 1S528.

The invention also provides a kit that comprises means for determining the presence or absence of one or more of the allelic deletions) described herein. In particular, such. means may include a specific binding agent, probe, primer, pair or combination of primers, or antibody, including an antibody fragment, as defined herein which is capable of detecting or aiding detection of such an allelic deletion. The primer or pair or combination of primers may be sequence specific primers which only cause PCR amplification of a polynucleotide sequence comprising the full length gene to be detected, as discussed herein. The kit may comprise two or more such pairs of primers. The kit may also comprise a specific binding agent, probe, primer, pair or combination of primers, or antibody which is capable of detecting the absence of the allelic deletion. In particular, the kit may comprise two or more specific binding agents, each of which is suitable for the detection of an allelic deletion in a gene selected from FHIT, PR, f RTM29 and DLEC1. One or more of the binding agents may be specific for the same protein target. For example, more than one binding agent for one of the proteins, for example FHIT, may be included wherein each of said binding agents binds a different part of the FHIT polynucleotide or protein. The sensitivity of the methods of the invention may therefore be increased.

The kit may additionally comprise one or more other reagents or instruments which enable any of the embodiments of the method mentioned above to be carried out. Such reagents or instruments may include one or more of the following: a means to detect the binding of the agent to the allelic variant, a detectable label such as a fluorescent label, an enzyme able to act on a polynucleotide, typically a polymerase, restriction enzyme, ligase, RNAse H or an enzyme which can attach a label to a polynucleotide, suitable buffer(s) or aqueous solutions for enzyme reagents, PCR primers which bind to regions flanking the allelic variant as discussed herein, a positive and/or negative control, a gel electrophoresis apparatus, a means to isolate DNA from sample, a means to obtain a sample from the individual, such as swab or an instrument comprising a needle, or a support comprising wells on which detection reactions can be carried out. In one aspect, a kit of the invention may also comprise suitable reagents for determining whether a patient carries HPV, for example a high risk HPV type such as HPV 16, HPV18, HPV45 or HPV56. The kit may be, or include, an array such as a polynucleotide array comprising the specific binding agent, preferably a probe, of the invention. The kit typically includes a set of instructions for using the kit.

The following Examples illustrate the invention:

Materials

Cervical smears with a storage time of more than 10 years were retrieved from the Surgical File of Department of Histopathology, University College London (UCL).

The diagnosis of CIN was based on histology. All patients were treated according to the same protocol at UCL Hospital. There was no significant difference in the age, nature of specimen (Pap smear or tissue biopsy) and follow-up time between the groups.

Microdissection and DNA extraction For Pap smears, dyskaryotic cells were identified and marked with a diamond pen on the reverse side of the slide. After removing the coverslip, dyskaryotic and normal epithelial cells were separately microdissected. Similarly, neoplastic and normal cells were microdissected from cervical biopsies. The isolated cells were digested with lOOμg/ml proteinase K in lxPCR buffer containing lOmM Tris-HCl (pH9.0), 50mM KCl and 1.0% Triton X-100 at 56°C for 16-20 hours. The digests were heated at 95 °C for 10 minutes to inactivate proteinase K and centrifuged to remove cell debris. The supernatant was used for PCR.

Polymerase chain reaction (PCR) 12 microsatellite markers were screened by PCR amplification. Primers were designed to immediately flank the tandem repeats, allowing amplification from DNA samples prepared from archival fixed specimens. The primers used for each marker were as follows:

For each pair of PCR primers, the sense primer is labelled with fluorescence. PCR was carried out in a thermal cycler using a "hot-start touch-down" program. PCR products were first confirmed on agarose gels and then analysed on ABI Prism 377 sequencer with GeneScan software. Dyskaryotic and normal cells from the same case were analysed in parallel.

High risk HPNs including 16, 18, 33, 45 and 56 were detected by PCR separately. The primer used for each HPN type were as follows:

The molecular data were generated without the knowledge of clinical follow- up details.

Example 1

LOH in CIN lesions.

12 microsatellite markers were screened in CIΝ lesions from 113 patients including 54 CLΝ3, 34 CIN2 and 25 CINl. The frequency of LOH of these markers in different CIN lesions is shown in figure 2. Several markers including D3S1611, D3S1300, D6S105, D6S265, Dl 1S35 and Dl 1S528 showed frequent LOH in CINl.

Correlation of LOH with CIN grade

The number of loci showing LOH positively correlates with the grade of CIN lesion. Figure 3 A shows the frequency of LOH at >1 or >2 or >3 of the 12 markers examined, in different grades of CIN lesions. Figure 3B is a typical case illustration. In 1988, the CIN2 lesion showed only LOH at one (D3S1289) of the 12 markers examined. In 1990, the CLN2 lesion progressed into a CIN3, which showed LOH at an additional locus (D3S1260). Similarly, the recurrent CIN3 lesion in 1992 displayed LOH at a further. locus (Dl 1S528).

Correlation of LOH with disease follow-up

Patients were divided into two groups according to their clinical follow-up: patients that showed disease free after treatment and those that showed disease persistence or progression after treatment. LOH was correlated with clinical follow- up irrespective of the grade of CIN lesions due to the sample sizes. Six of the 12 markers examined exhibited a much higher frequency of LOH in patients who showed disease persistence or progression than in those who showed disease free (see Figure 4). These markers are D3S1300, D3S1260, D3S1566, D6S265, D11S35 and Dl IS 528. The results indicate that these markers are potentially useful in prediction of the clinical behaviour of CIN lesion and therefore may have huge impact in the clinical management of patients with CIN lesions.

Example 2

LOH in CIN lesions. 12 microsatellite markers were screened for loss of heterozygosity (LOH) in

117 cases of Cervical Intraepithelial Neoplasia (CIN) including 25 CINl, 35 CIN2 and 57 CLN3, which have been followed up for 7-11 years. The frequency of LOH of these markers in different CIN lesions is shown in Figure 5.

Correlation of LOH with CIN grade

The frequency of LOH at these loci varied from 0% to 65% (Figure 5). In general, the frequency of LOH at individual loci and the number of loci showing LOH positively correlated with the cytological grade of CIN lesions (Figures 5 and 6). However, there were exception, for example, D3S1289 showed LOH in 65%> of CIN2 lesions but in only 25% and 32% of CINl and CIN3 respectively. These results indicated that not every LOH marker investigated was associated with disease progression. The ones associated with disease progression may suggest inactivation of a tumour suppressor gene, while those not associated with disease progression may not be related to any important gene and may be the result of a general genomic instability.

Correlation o LOH with disease follow-up

In order to identify the markers that are associated with disease progression,. we divided the 117 cases of CFN lesions into two groups according to their treatment outcome irrespective of their cyto logical grades: disease free group (DF) (those showed disease free (DF) after treatment) and disease persistence group (DP) (those displayed disease persistence or progression (DP) despite treatment), and compared LOH between the two groups. Among the 12 markers investigated, 4 loci (D3S1300, D3S1260, D11S35 and D11S528) showed significant differences between the two groups (Figure 7). These 4 loci were selected for further study.

An additional 74 cases of CIN were screened as described above, making a total group size of 191. These 191 cases included 61 cases (9CIN1, 20CIN2 and 32 CIN3) showing disease free after treatment and 130 cases (59 CINl, 44 CIN2 and 2-7 CIN3) exhibiting disease persistence, or progression despite treatment. The four markers D3S1260, D3S1300, D11S35 and D11S528 were analysed and LOH compared between the disease free and disease persistent/progression groups. The data for this comparison in this larger group of cases is shown in Figure 8.

To determine the prognostic value of these 4 loci, we compared the frequency of LOH occurring at 1 of the 4 markers, 2/4, 3/4 and 4/4 between the DF and DP groups shown in Figure 8 (Figure 9)._, When comparison was made at 1 of the 4 markers showing LOH, the DP group showed significant higher LOH rates (41-57%) than the DF group (20-28%). As a considerable proportion of DF group showed LOH, the prognostic value of using 1 of the 4 markers is limited. When comparison was made at 2 of the 4 markers showing LOH was chosen, 8-20% of the DP group showed LOH, in contrast only 0-3% of the DF group displayed LOH. When comparison was made at 3 of the 4 markers or 4 of the 4 markers showing LOH, only 5% or less of the DP group showed LOH although none of the DF group showed LOH. Thus, the selection of 2 of the 4 markers showing LOH gave the highest power for prediction of disease prognosis in terms of both sensitivity and specificity. By stepwise testing the sensitivity and specificity of the combined 4 LOH markers in disease prognosis, around 20% of high grade CLNs (CIN2 and CIN3) in the disease persistent group could be identified with 100% specificity for CIN2 and 97% specificity for CIN3 when 2 of the 4 markers showing LOH were selected. Based on this, 20% of CIN lesions that showed disease persistence or progression after treatment could be identified.

Example 3

12 microsatellite markers were screenedand correlated LOH with clinical outcome in 97 cases of CIN including 15 CINl, 35 CIN2 and 47 CTN3 using diagnostic Pap smears.

Correlation of LOH with disease follow-up Patients were divided into two groups: those became disease free (DF) after the first treatment and those showed disease persistence (DP) at the same or a higher histological grade despite the initial treatment. As the number of cases examined from each CIN group was relatively small, we correlated LOH with clinical outcome irrespective of CIN grade. Among the 12 markers investigated, D3S1260 and D 11 S528 showed LOH at significantly higher frequencies in the DP than DF group (pO.Ol, Figure 10A). D3S1300 and Dl 1S35 exhibited the next highest statistical difference between the two groups albeit not significant. Using a stepwise statistical analysis testing various combination of the 12 markers, the above four markers collectively gave the highest statistical significance (p<0.03) between the two groups and were selected for further analysis. _, As the value of using these loci alone as prognostic markers was limited, we tested the 4 markers together (Figure 10B). By stepwise examining the sensitivity and specificity of the combined 4 LOH markers, i.e. 1 or 2 or 3 or 4 of the 4 loci showing LOH, the best cut-off point was when 2 of the 4 loci showing LOH was applied. At this threshold, around 47-63% of CIN lesions in the DP group but none in the DF group showed LOH. Example 4

To further confirm the prognostic value of LOH at the markers D3S1300, D3S1260, Dl 1S35 and Dl 1S528 in CIN, we examined a further 116 cases using diagnostic biopsies since it was much easier to microdissect neoplastic cells from biopsies than Pap smears, and correlated LOH with clinical follow-up as above. The data from these additional cases were compatible to those obtained from Pap smears in the earlier studies albeit the LOH frequency in the DP group was slightly higher in the Pap smears than biopsies. This most likely reflected the selection of Pap smears with abundant clumps of dyskaryotic cells, which may bias toward those with "bulky" disease. No statistical differences were found in the age, HPN status, treatment modalities and follow-up time between Pap smears and biopsies.

In total, we examined 213 cases of CIΝ including 71 CLΝ1, 78 CTΝ2 and 64 CIN3. The frequency of LOH at D3S1300, D3S1260 and Dl 1S35 positively correlated with CIN grade, while the incidence of LOH at Dl 1S528 was similar among different CIN groups (Figure 11A). As expected, the frequency of LOH at these loci was significantly higher in DP than DF group (p<0.0005, Figure 1 IB). By combining the 4 markers together as above, 24% of CINl, 27% of CIN2 and 54% of CIN3 of the DP group but none of the DF group showed LOH at 2 or more of the 4 foci examined (Figure 1 IC). Thus, LOH analysis at these loci could identify 24-54% of CIN lesions that showed disease persistence despite treatment.

Overall, the percentage of infection of high risk HPNs was significantly higher in DP (92%) than in DF group (46%) (p<0.0005). Among different high risk HPNs, HPN 16 and double virus infection were significantly associated with DP group (p<0.0005, p<0.05 respectively). Interestingly, HPN16 but not others was significantly associated with LOH at each of the four loci examined (p<0.01), particularly D3S1300 (p<0.0005). Treatment information was available in 130 cases including 59 and 71 from DF and DP group respectively. Cone biopsy was more frequently used in DF than DP group (p<0.05), whereas laser loop excision of the transformation zone (LLETZ) was applied more often to DP (62%>) than DF group (19%)) (p<0.05). There was no significant difference in other treatment modalities, age and follow-up time between the two groups.

*signifιcant difference between DF and DP groups. LLETZ: Laser loop excision of the transformation zone

To further evaluate the association of LOH with clinical outcome of CIN, we examined both diagnostic and follow-up biopsies in 20 cases from DP group. Of these cases, 6 (number 1-6) became disease free after the first follow-up biopsy (2-15 months) and none of them showed accumulation of LOH in the follow-up biopsy. The remaining 14 cases showed disease persistence at the same (6 cases) or higher histological grade (8 cases) during follow-up, and 10 (number 11-20) displayed LOH at additional loci. The average of time to gain LOH at an additional locus was 26 months, ranging from 17 to 84 months. The LOH observed in the diagnostic biopsy was always seen in the follow-up biopsy.

Cases 1-6 became disease free after the first follow-up biopsy; Cases 7-20 continually showed disease persistence or progression after the first follow-up biopsy (data not included because of space limitation). NA: not available; +: LOH positive; -: LOH negative; H: homozygous.

In 6 cases (numbers 13-15, 17-20), the diagnostic biopsy showed no LOH or LOH at only 1 of the 4 loci, but the follow-up biopsy displayed LOH at 2 or more of the 4 loci, reaching the threshold of prognostic significance as detailed above. Interestingly, three of these cases (number 18-20) also showed infection by an additional high-risk HPV.

Claims

1. A method of diagnosing susceptibility to persistence or progression of cervical intraepithelial neoplasia (CIN) in an individual suffering therefrom; the method comprising: b) providing a sample of dyskaryotic cells from said individual and a sample of non-dyskaryotic cells from said individual; b) detecting an allelic deletion in one or more genes selected from FHIT, PR, DLEC1- and TRIM 29 by comparing the FHIT, PR, DLEC1 and/or TRIM 29 polynucleotides or proteins present in the samples of step a) derived, respectively, from the non-dyskaryotic and dyskaryotic sample, wherein detection of an allelic deletion between the non-dyskaryotic and dyskaryotic samples is correlated with a

^' susceptibility to persistence or progression of CIN.

2. A method according to claim 1 comprising the steps of: a) providing genomic DNA from a sample of dyskaryotic cells from said individual and from a sample of non-dyskaryotic cells from said individual; c) detecting an allelic deletion in one or more genes selected from FHIT, PR, DLEC1 and TRIM 29 by comparing the genomic DNA of step a) derived, ^• respectively, from the non-dyskaryotic and dyskaryotic sample, wherein detection of an allelic deletion between the non-dyskaryotic and dyskaryotic samples is correlated with a susceptibility to persistence or progression of CIN.

3. A method according to claim 2 wherein said allelic deletion is detected by comparing the genomic DNA of step a) derived, respectively, from the non-dyskaryotic and dyskaryotic sample, and determining whether a loss of heterozygosity has occurred in one or more microsatellite DNA markers in the dyskaryotic sample wherein detection of the loss of heterozygosity is correlated with a susceptibility to persistence or progression of cervical intraepithelial neoplasia

• (CIN).

4. A method according to claim 3. wherein the genomic DNA of step a) is amplified with at least one pair of primers, wherein the at least one pair of primers amplifies a microsatellite DNA marker.

5. A method according to claim 3 or 4 wherein the microsatellite DNA marker(s) examined are selected from D3S1300, D3S1260, Dl 1S35 and Dl 1S528.

6. A method of diagnosing susceptibility to persistence or progression of cervical intraepithelial neoplasia (CIN) in an individual suffering therefrom, the method comprising: a) providing genomic DNA from a sample of dyskaryotic cells from said individual and from a sample of non-dyskaryotic cells from said individual; b) detecting a loss of heterozygosity in one or more microsatellite DNA markers by comparing the genomic DNA of step a) derived, respectively, from the non- dyskaryotic and dyskaryotic sample, wherein detection of the loss of heterozygosity in the dyskaryotic sample is correlated with a susceptibility to persistence or progression of cervical intraepithelial neoplasia (CIN).

7. A method according to any one of the preceding claims wherein the dyskaryotic cells are obtained from a cervical (Pap) smear.

8. A method according to any one of the preceding claims wherein the individual has been diagnosed with a CINl lesion.

9. A method according to any one of claims 3 to 8 wherein loss of heterozygosity is detected by PCR-based analysis of microsatellite DNA markers using DNA samples prepared from microdissected dyskaryotic and non-dyskaryotic cells.

10. A method according to any one of the preceding claims which, further comprises the step of determining whether said individual is suffering from a human papillomavirus (HPN) infection.

11. A method according to any one of the preceding claims wherein said individual has previously been diagnosed as suffering from an HPN infection.

12. A method according to claim 10 or 11 wherein said HPN is selected from the group consisting of HPN 16 and HPN18.

13. A method according to claim 10 or 11 wherein

(a) a loss of heterozygosity is assessed in the microsatellite marker D3S1300; or

(b) said gene is FHIT and wherein said HPN is HPN- 16.

14. A kit comprising a panel of two or more pairs of primers wherein each pair of primers is suitable for amplifying a microsatellite DΝA marker selected from D3S1300, D3S1260, Dl 1S35 and Dl 1S528.

15. A kit according to claim 14 which comprises a panel of pairs of primers capable of amplifying all of the microsatellite DΝA markers D3 S 1300, D3S1260, Dl lS35 and Dl lS528.

16. A kit according to claim 15 wherein said primers are as given in SEQ ID Νos. l to 8.

17. A kit comprising a panel of two or more specific binding agents wherein each binding agent is capable of distinguishing between the normal and allelic deletion forms of a polynucleotide or protein selected from FHIT, PR, TRIM29 and DLEC 1.