Colposcopic biopsies versus loop electrosurgical excision procedure cone histology in human immunodeficiency...

To compare the discrepancy between colposcopically directed punch biopsy and excisional cone biopsy in human immunodeficiency-positive (HIV+) vs. HIV-negative (HIV-) women.

We performed a case-control analysis of women treated with excisional cone biopsy after an abnormal colposcopic punch biopsy. Punch and cone biopsy histology were compared in 29 HIV+ (mean CD4 = 251 cells/mm3, 10 with the acquired immunodeficiency syndrome) and 31 HIV- women. Only patients with no prior treatment for cervical dysplasia, satisfactory colposcopy and cervical cytologic smears concordant with colposcopic biopsies were included.

Disagreement between punch biopsy and cone histology was evident in 41% (12/29) of HIV+ patients and 48% (15/31) of seronegative women (chi 2, P = .78). The cone specimen had a higher grade lesion than the punch biopsy in 38% (11/29) of HIV+ patients and 32% (10/31) of seronegative women (P = .65). Overall, patients with HPV, cervical intraepithelial neoplasia (CIN) I or II on punch biopsy had CIN III on 30% of cone biopsies (5/23 HIV+ vs. 9/23 HIV-women, P = .2). In HIV+ women with HPV or CIN I on punch biopsy, 50% (9/18, 95% confidence interval 26-74%) had CIN II or III on the excisional cone vs. 18% (2/11) HIV-patients (Fisher's test, P = .13). However, in HIV+ patients with CIN II or III on cone biopsy, 47% (9/19) had only CIN I or human papillomavirus on punch biopsy as compared to 9% (2/22) HIV-patients (chi 2, P = .01).

Colposcopically directed punch biopsies are poor predictors of cone histology in both HIV+ and HIV-patients. Based on confidence intervals, at least 26% and as many as 74% of HIV+ women with CIN I on punch biopsy may have a significantly worse lesion on cone biopsy despite satisfactory colposcopy. Though CIN I may be observed in immunocompetent women, due to the likelihood of a more advanced lesion, observation may not be justified in HIV+ women.