Annual outpatient hysteroscopy and endometrial sampling (OHES) in HNPCC/Lynch syndrome (LS).

BACKGROUND

LS women have a 40-60% lifetime risk of endometrial cancer (EC). Most international guidelines recommend screening. However, data on efficacy are limited.

PURPOSE

To assess the performance of OHES for EC screening in LS and compare it with transvaginal ultrasound (TVS) alone.

METHODS

A prospective observational cohort study of LS women attending a tertiary high-risk familial gynaecological cancer clinic was conducted. LS women opting for EC screening underwent annual OHES and TVS. Histopathological specimens were processed using a strict protocol. Data of women screened between October 2007 and March 2010 were analysed from a bespoke database. Histology was used as the gold standard. Diagnostic accuracy of OHES was compared with TVS using specificity, and positive (PLR) and negative (NLR) likelihood ratios.

RESULTS

Forty-one LS women underwent 69 screens (41 prevalent, 28 incident). Four (three prevalent, one incident) women were detected to have EC/atypical endometrial hyperplasia (AEH), five had endometrial polyps and two had endometrial hyperplasia (EH) on OHES. TVS detected two of four EC/AEH. OHES had similar specificity of 89.8% (CI 79.2, 96.2%), but higher PLR 9.8 (CI 4.6, 21) and lower NLR (zero) compared to TVS: specificity 84.75%(CI 73, 92.8%), PLR 3.28 (CI 1.04, 10.35) and NLR 0.59 (CI 0.22, 1.58). No interval cancers occurred over a median follow-up of 22 months. The annual incidence was 3.57% (CI 0.09, 18.35) for EC, 10.71% (CI 2.27, 28.23) for polyps and 21.4% (CI 8.3, 40.1) for any endometrial pathology.

CONCLUSIONS

Our findings suggest that in LS, annual OHES is acceptable and has high diagnostic accuracy for EC/AEH screening. Larger international studies are needed for confirmation, given the relatively small numbers of LS women at individual centres. It reinforces the current recommendation that endometrial sampling is crucial when screening these women.