Zuchna et al report underestimation of the severity of cervical intraepithelial neoplasia (CIN) in 47% of cases after colposcopic-guided biopsies in patients with suspected high-grade squamous intraepithelial lesions (H-SIL) of the cervix. On the basis of this low correlation between the histology at biopsy vs the histology in the cone specimen the authors conclude that H-SIL cytology can be treated without previous biopsy. We question these findings and the conclusion.

Several investigators have reported that initial biopsy and final histologic diagnosis can differ. Possible reasons for this include misinterpretation of colposcopic findings, inaccurate biopsy, and larger lesions composed of several different epithelial components. Furthermore, in some patients the entire transformation zone is not visualized and there is interobserver or intraobserver variability among pathologists.

At our department, indications for conization are biopsy-verified persistent CIN 1 or 2, CIN 3, adenocarcinoma in situ, and early stromal invasion. Diagnostic conization is also performed in occasional cases with persistent abnormal cytology and negative histology at biopsy. We report our data on comparison between initial histology on guided cervical biopsies and definitive histology of cone specimens between 2007 and 2009. In this series cervical biopsies underestimated the severity of CIN in only 22 (6%) of 380 cones.

We disagree with the recommendation of a see-and-treat strategy in patients with suspected H-SIL (conization only on base of cytology and colposcopy without initial histology) because overtreatment has been reported in up to 83% of such patients. Treatment–ie, conization–at first visit of suspected H-SIL should be used only in exceptional cases. According to European guidelines, cases with discrepancies among cytologic, colposcopic, and histologic findings should be managed with a review of the cytologic and histologic slides and colposcopy including the vagina.

No first-line diagnostic method (cytology, colposcopy, even guided biopsy) can be relied on absolutely for an accurate diagnosis of CIN, but an acceptable correlation of >90% between colposcopic-guided biopsies and cone histology is attainable. Only conization provides a definitive diagnosis of CIN, provided the lesion has been excised completely. Sensitive and specific molecular markers are needed to identify lesions at initial biopsy with significant risk for progression.