Endometrial polyp or neoplasia? A case-control study in women with polyps at ultrasound.

OBJECTIVES

To identify factors associated with endometrial neoplasia in women diagnosed with endometrial polyp at transvaginal ultrasound.

METHODS

Within a population of 1390 consecutive patients undergoing hysteroscopy following an ultrasonographic diagnosis of polyps, we compared the cases with a final diagnosis of endometrial neoplasia with controls with benign endometrial polyps. The controls were selected randomly in a ratio of 4 : 1 (controls : cases). Bivariate statistical analysis and multiple logistic regression were used to measure the association between various variables and endometrial neoplasia.

RESULTS

Sixteen cases of endometrial neoplasia were compared to 64 controls with confirmed benign endometrial polyps. All cases of neoplasia were among symptomatic women, while 40.62% of women with benign polyps had been referred to hysteroscopy after a routine ultrasound and were asymptomatic. Women with endometrial neoplasia were significantly older (mean age 64.19 ± 9.382 vs. 52.03 ± 9.846 years; p < 0.001) and had a significantly greater body mass index (median 27.66 vs. 24.59 kg/m(2); p < 0.001). Other factors statistically associated with endometrial neoplasia were postmenopausal status and bleeding as a main symptom. At multivariate analysis with logistic regression, the only factors showing a statistically significant association with endometrial neoplasia were older age (odds ratio 1.102; 95% confidence interval 1.015-1.198) and bleeding (odds ratio 13.7; 95% confidence interval 1.486-126.278).

CONCLUSION

When polyps are diagnosed at ultrasound, bleeding and an older age are independently associated with endometrial neoplasia. A significant proportion of asymptomatic women is referred to hysteroscopy because of a polyp seen at routine ultrasound, although malignancy is highly unlikely in these cases.