The main objective of this thesis was to estimate the sensitivity and specificity of four screening tests: visual inspection after the application of acetic acid (VIA), VIA_M: combined VIA and VIAM (magnified VIA using an AviScope performed by a doctor), liquid-based cytology (LBC using the AutoCyte-Prep manual system) and HPV testing (using Hybrid Capture II); to detect histologically confirmed high-grade squamous intra- epithelial lesions (HSIL). Participants of the "Tamizaje y Tratamiento Inmediato de lesiones cervico-uterinas" (TATI) project, who signed an informed consent for additional cervical samples, were included in this study. Separate cervical samples for conventional cytology (CC), LBC and Hybrid Capture II (HC-II), were collected by a midwife before applying acetic acid (5%) to the cervix and performing VIA. Women testing positive on VIA were referred to VIAM performed by a doctor who confirmed the midwife diagnosis, and treated lesions with cryotherapy if appropriate, or referred women to colposcopy. Negative women on VIA_M; underwent colposcopy if they had HSIL on CC or LBC; or had second screening if they had any lesser abnormality on LBC or were positive only on HC-II. Of 5565 participants, 104 had histologically confirmed HSIL and an estimated 112 had undetected/unconfirmed HSIL. Sensitivities of VIA, VIA_M, LBC (any abnormality), LBC (high-grade abnormality) and HC-II were 44% (95% confidence interval (CI): 34,59), 31% (CI: 20,50), 69% (CI: 61,78), 35% (CI: 23,56) and 72% (CI: 65,79), respectively. VIA had the lowest specificity among all tests: 77% (CI: 74,77). LBC ?HSIL had the highest PPV: 44% (CI: 34,53). HC-II was more sensitive than the other tests in our study. LBC ?HSIL was more specific than the other tests. VIA had low sensitivity and specificity. VIA could still be an option for developing countries if implemented with regular quality assurance.

Cervical intraepithelial neoplasia (CIN) lesions of the same morphological grade show variable clinical behaviour; some progress and others regress. Currently, there are no biochemical or molecular markers which can distinguish CIN lesions with different prognosis. We have optimised several molecular methods on archival cervical smears and screened a number of molecular markers that may aid prognosis of CIN. By systematically validating different protocols, we have established that crude DNA preparations from a small number of microdissected cells from cervical smears are adequate for various PCR-based investigations. Furthermore, the crude DNA preparations could be further purified and used for PCR-based clonality analysis of the X-linked genes. Using PCR-based clonality analysis of the androgen receptor gene, we have shown that CIN3 and the majority of CIN2 lesions are monoclonal, whereas CINl lesions are polyclonal. Importantly, patients with monoclonal CIN2 show disease persistence or progression, while polyclonal CIN2 regress after treatment and remain negative during follow-up. To further identify molecular markers that are technically easy to apply and potentially suitable for prognosis assessment of all CIN grades, we have investigated the prognostic value of gene deletions. In a pilot study, we screened 12 microsatellite markers, which showed high frequencies of loss of heterozygosity (LOH) in cervical carcinomas, and identified four, including D3S1300 , D3S1260, D11S35 and D11S528, that were significantly associated with CIN persistence or progression. The 4 markers were further investigated in a larger cohort. Combined analysis of LOH at these 4 loci permitted the identification of 22-47% of CIN lesions of various histological grades, that were associated with disease persistence or progression with 100% specificity. LOH at these loci was significantly associated with HPV16 infection. Bioinformatics analysis identified several candidate genes including the fragile histidine triad and progesterone receptor gene that may be the target of deletions.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

Background: In the UK more than 3 million women are screened each year as part of a national screening programme to prevent cervical cancer. Over half of all positive cervical cytology results are reported as 'borderline changes' and 16% of all colposcopy referrals in the NHS cancer screening programme arise from borderline results. As the majority of borderline reports reflect benign changes this results in both considerable overtreatment and a significant burden on the NHS. High-risk strains of the human papillomavirus (HPV) are known to be the causative agent of cancer of the cervix. Aims: To determine if immunocytochemistry for p16 INK4a, a surrogate marker for HPV infection, can identify which borderline cytology results require further investigation, and whether the application of this additional test would be cost-effective. Methods: A prospective observational study in which liquid-based cytology samples from 500 women with borderline changes were reprocessed and immunostained for p161NK4a. The women were followed up via hospital and primary care records for 2.5 years. The costs associated with the additional test were calculated, and balanced against the costs of investigation and treatment, based on an existing protocol for HPV testing. Results: Immunocytochemistry for p16 INK4a showed a Positive Predictive Value (for an outcome of CIN2/moderate dyskaryosis or greater) of 87% and a Negative Predictive Value of 95%. We estimated a cost saving of approximately £28 per sample resulting from fewer colposcopy visits and follow-up cervical cytology. Conclusions: The high positive and negative predictive values demonstrated in this study suggest p16 INK4a immunocytochemistry may be useful as a tool to distinguish those few women with borderline changes who are genuinely in need of further investigation from those who are p161NK4a -negative and can be returned to normal screening programme recall. The introduction of this test could result in considerable cost savings for the NHS, as well as reduced anxiety for women with borderline results.