Department of Obstetrics and Gynaecology University of the Witwatersrand Johannesburg February 2015 A research report submitted to the Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, in partial fulfilment of the requirements for the degree of Masters in Medicine, in the branch of Obstetrics and Gynaecology. / HIV accelerates the development of cervical cancer by up to15 years. South Africa is currently in the midst of an HIV epidemic. With limited facilities for colposcopy it is vital to identify risk factors within the HIV positive population resulting in positive margins after Large Loop Excision of the Transformation Zone (LLETZ) and persistence of cytological abnormalities on follow-up Pap smears. Objective: The primary objective was to determine the patient risk factors, pre and during colposcopy and LLETZ biopsy, which resulted in the histological involvement of margins of the LLETZ biopsy and persistent cervical dysplasia on follow-up Pap smears. Secondary objectives included determining follow up rate of patients at the clinic as well as the correlation between the original Pap smear cytology grade and the histological grade found on histology of the LLETZ biopsy. Methods: A retrospective review of the files of HIV seropositive patients was done at the colposcopy clinic at Charlotte Maxeke Johannesburg Academic Hospital after the roll out of antiretroviral treatment for the period 1 April 2004 to 31 October 2012. Patients with abnormal pap smears during this time were referred to the colposcopy clinic where a colposcopy and LLETZ biopsies were done. Demographic and clinical data in regards to age, gravidity, contraception, CD4 count, antiretroviral usage, and referral time was collected. Data from the clinical description of the colposcopy and histology of the LLETZ biopsy was also collected. Patients followed up again after 6 months when a repeat pap smear was done. The results of these Pap smears were also collected. Data was then analysed and variate and multivariate logistical regression was used to find statistically significant correlations. Results: A total of 480 files were found to have complete clinical records. One hundred and sixty eight (42.71%) patients had both endo and ectocervical margins clear. Predictive factors for the involvement of endocervical margins was the doctor performing the procedure (p-value <0.01) cytology of the original Pap smear (p value <0.01) and the grade of histological abnormality found at time of LLETZ (p-value <0.01). The statistically significant predictive factors for ectocervical margin involvement was the visualization of the transformation zone at colposcopy (p-value <0.01), the size of lesion found at colposcopy (p-value <0.01), the use of combined oral contraceptive pill (OCP) (p-value 0.02) and the histological grade of abnormality found on the LLETZ biopsy. Age, parity, CD4 count, use of antiretroviral drugs, length of time from Pap smear to colposcopy and use of contraception other than OCP were not found to be statistically significant in our sample population for the involvement of either endo or ectocervical margins. Statistically significant risk factors for the recurrence of intraepithelial lesions on follow up Pap smear was having both endo and ectocervical margin involvement on histology (p-value 0.01) The Ectocervical margin alone was found to have a p-value of <0.01. Abnormal cytology on follow up Pap smear was found in 58.69% of patients. The follow up rate at the clinic was 46.04%. Correlation of cytological grade and histological grade of cervical intraepithelial neoplasm in our sample population was found to be adequate (p-value <0.01). Conclusion: Incomplete incision of the intraepithelial lesion was found to be a significant risk factor for the recurrence of cytological abnormality in patients undergoing LLETZ biopsy. Identifying patients at increased risk for recurrence is important to ensure close follow up in this patient population.

Diss. (sammanfattning) Stockholm : Karolinska institutet, 2007. / Härtill 4 uppsatser.