Abstract
Background/Objectives: When the squamocolumnar junction is not fully visible (type 3 transformation zone), colposcopy-directed biopsy may under-sample endocervical disease. Endocervical curettage (ECC) is recommended in selected settings, but its incremental diagnostic yield in this setting, and whether this yield is concentrated in women with HPV16/18, remains clinically debated. Methods: We performed a retrospective cohort study of women referred to colposcopy because of HPV16/18 positivity regardless of cytology, persistent non-16/18 high-risk HPV positivity, and non-16/18 high-risk HPV positivity with abnormal cytology. Persistent non-16/18 high-risk HPV positivity was defined as repeated positivity on two tests performed at least 6 months apart. Eligible women had type 3 transformation zone documented and underwent paired ectocervical biopsy plus ECC at the same visit; biopsy was obtained in all women, including targeted sampling of the most abnormal ectocervical area when no discrete lesion was evident. Women were stratified by HPV genotype into HPV16/18 and non-16/18 high-risk HPV groups. The primary outcome was index high-grade cervical lesion, defined histologically as CIN2, CIN3, or carcinoma in situ; invasive cervical cancer was excluded. The added diagnostic yield of ECC was defined as ECC-only CIN2+, that is, CIN2+ detected on ECC when biopsy was