Risk of invasive cancer after colposcopy with and without biopsy

Transcription

1 Risk of invasive cancer after colposcopy with and without biopsy Guglielmo Ronco MD Senior Epidemiologist Adele Caprioglio MSc, Giovanni Maina MD Colposcopist, Mario Preti MD Colposcopist City of Health and Science Turin, Italy

2 Disclosures No financial relationships or conflict of interest to disclose

3 Background Colposcopy is used, among other things, to define which women need treatment to prevent an invasive cervical cancer and its features A number of studies showed that by performing random biopsies some 20% more high-grade CIN are detected Only part of high-grade CIN progress to invasion and after long time (about 1/3 in 30 years for CIN3 Mc Credie et al Lancet Oncol 2007) Little is known about the risk of invasive cancer after a colposcopy according to colposcopy s and woman s features

4 Objective Evaluating the ability of colposcopy to stratify women with abnormal cytology as for their risk to develop invasive cervical cancer

5 Setting Organized screening programme active from 1992 Women with abnormal cytology ( ASC-US) referred for colposcopy in pre-defined centres (>80% attend) All cytologies registered All colposcopies (biopies taken and their result) registered Colposcopic impression registered but not used to the moment Treatments registered Cancer Registry covering greater Turin active from 1980s

6 Methods 1 Women having had a first screening test before Dec included Women classified according to screening history and features of their first colposcopy Linkage with cancer register ICC incidence computed from t0 (group specific, see next table) to (a) end of follow-up (Dec,31,2007) or (b) date of death or emigration or (c) date of treatment whichever first Cancers after treatment not considered because study interest not on protection by treatment and clinical follow-up Cancers detected within 6 months of t0 excluded because plausibly already present at t0

7 Methods 2 Classification of women and t0 group Never had abnormal cytology Abnormal cytology ( ASC-US) but no colposcopy Abnormal cytology, colposcopy(ies) without biopsy at the first Abnormal cytology, colposcopy(ies) biopsies <CIN2 only at the first Abnormal cytology, colposcopy(ies) with 1 biopsy(ies) CIN2 at the first t0 Date first cytology in screening programme Date first abnormal cytology Date first colposcopy Date first colposcopy Date first colposcopy

8 Results women included 86 invasive cancers occurring >6 months after t0 Mean follow-up 8.7 years 73% of women with colposcopy had at least 1 repeat 55% of colposcopies had at least 1 biopsy (biopsies taken in colposcopically abnormal areas) 21% (152/725) of the women with a CIN2+ diagnosis had it after the first colposcopy. Consistent with about 80% sensitivity of colposcopy

9 ICC incidence by group No women Personyears ICC Incidence per no abnormal cytology RR p no colposcopy < colposcopy no biopsy at first colposcopy, only biopsies <CIN2 at first colposcopy biopsy CIN2 at first < (0-2.59) <0.0001

10 Women with first colposcopy without biopsy No women Personyears ICC Incidence per RR p Referred for cytology HSIL Referred for cytology < HSIL < RR vs. always normal cytology of colposcopy with no biopsy when excluding HSIL women :

11 Conclusions 1 Women without biopsy plausibly had no clearly abnormal colposcopic finding. Their cancer risk 5-fold lower than that in women with cytology ASC-US and no colposcopy but almost 20 fold higher than that of women with normal cytology Colposcopy able to stratify women with abnormal cytology by risk of ICC but not to identify a group at sufficiently low risk For women with HSIL a histology should be always obtained.

12 Conclusions 2 Women with negative biopsy at very low risk of ICC. Data from other Italian Regions being acquired in order to have a more precise estimate. More detailed analyses taking into account repeated colposcopies planned Results suggest wide use of biopsies