A non -invasive method of screening for endometrium cancer by analysis of DNA methylation on cervical smear shows promising results, with a sensitivity of 95 % for two specific biomarkers.
Endometry cancer is one of the most common gynecological cancers. If the prognosis is excellent in early stages with a 5 -year survival of 95% in stage I (tumor limited to the uterine body), it deteriorates at advanced stages, falling at 20% in stage IV (presence of metastases). This means the importance of early screening. Several diagnostic tools are available (imaging, hysteroscopy, cytology, endometrial biopsies), each with drawbacks. Endometrial biopsies, considered the reference technique, nevertheless include 10 % false negatives, and their repetition is unpleasant for patients. Imaging examinations, although informative, are expensive for mass screening, and their specificity is insufficient.
Endometrial cytology consists in collecting exfoliated cells using a swab introduced into the uterine cavity. This method suffers from a limited sensitivity and a delicate interpretation. It cannot therefore be recommended for screening, even in women at risk (patients under tamoxifen for example).
An innovative approach: DNA methylation analysis
Faced with these limitations, the researchers oriented towards the analysis of the methylation of DNA, whose disturbances intervene early in the carcinogenesis process. DNA hyper-methylation, associated with genetic changes, has already been used in the screening of other cancers. This approach has the advantage of being non -invasive, by analysis of the genes made on a sample of urine or cervical smear.
A Chinese study explored the use of detectable methylation biomarkers on simple cervical smear. The authors used 32 histological samples of endometrial cancer, including 19 “fresh” after a hysterectomy and 13 after fixing in paraffin. Have been distinguished types I (7 adenocarcinomas) and II (12 light, serous or undifferentiated cell carcinomas). Cervical smears were taken from 94 women, 19 of which presented endometrial cancer, the others being carrying hyperplasia, intra-epithelial neoplasia (CIN), or with a healthy endometrium.
DNA extracted from the endometrial tissue was treated with bisulfite then analyzed by chain reaction by polymerase in real time (QPCR) with different genes, making it possible to identify specific methylation profiles of endometrium cancer. Several genes have been used as biomarkers, including Rassf1a, Hist1h4f, Septine 9 and Pax1.
Selection of two biomarkers
Among the biomarkers tested, the Rassf1a and Hist14f gene methylation level is significantly higher in endometrium cancer unlike the other two markers.
The analysis of cervical smear confirmed these observations. The methylation of the RASSF1A and Hist1HF4 genes is present in all endometrial cancer samples. The methylation of Rassf1a is found in only 7 of the 34 healthy endometers, that of Hist1hf4 in a single healthy endometrium. In women with hyperplasia or CIN, the results are intermediate between those of the normal endometrium and those of endometrial cancer.
The analysis of the ROC curves and the area under the curve (AUC) revealed that the combination of the two markers reached an AUC of 0.938 for RASSF1A and 0.951 for hist1h4f in the detection of endometrial cancer and pre-cancerous lesions.
These results demonstrate the potential of a non -invasive method based on DNA methylation analysis by QPCR on cervical smear for endometrial cancer screening. This approach could expand the possibilities of conventional cytological screening.
This concept validation study opens up new perspectives in the field of the clinical diagnosis of endometrial cancer, potentially offering a solution to the limits of current methods while maintaining an essential non -invasive character for effective mass screening.
