Cervical dysplasia in women infected with the human immunodeficiency virus(HIV): A correlation with HIV viral load and CD4+count

Objectives. The incidence of cervical dysplasia and carcinoma is known to b e increased in HIV-infected women. In addition, there is a positive correla tion between HIV viral load (VL), CD4+ count, and opportunistic infections, as well as the incidence of various malignancies. This study compares HIV VL and CD4+ count with the presence of cervical dysplasia, as well as with the degree of severity of dysplasia. Methods. A retrospective chart review of 350 HIV-infected women with polyme rase chain reaction (PCR) quantitation of viral load was performed to ident ify 82 women with biopsy-proven cervical dysplasia and 25 women without any significant cervical pathology. The highest plasma VL within a year of the patients' cervical pathology and corresponding CD4+ count was selected and compared with cervical pathology. Univariate and multivariate statistical analysis using Student's t test and logistic regression analysis was used t o analyze the significance of other risk factors such as age, race, smoking history, history of illicit drug use, and prior sexually transmitted disea se as well as of viral load and CD4+ count. Results. Of 82 cases of cervical dysplasia, 33 (40.24%) were mild (CIN I), 47 (57.32%) were either moderate or severe (CIN II-III) dysplasia, and 2 de monstrated invasive squamous cell carcinoma (2.44%). A significant statisti cal difference was found when comparing either HIV plasma VL or CD4+ T-cell counts with the presence of cervical dysplasia on biopsy (P < 0.005). Howe ver, only CD4+ count was identified as an independent risk factor for the p resence of cervical dysplasia after multivariate analysis. Conclusion. In our population, there is a significant correlation between V L and CD4+ count and the presence of cervical dysplasia. However, VL does n ot appear to be an independent risk factor for cervical dysplasia in this p opulation of HIV-infected women. (C) 2001 Academic Press.