Values of narrow-band imaging hysteroscopy in the diagnosis of endometrial hyperplasia and endometrial cancer

Abstract

Objective To explore the values of narrow-band imaging (NBI) hysteroscopy in the diagnosis of endometrial hyperplasia (EH) and endometrial cancer. Methods From June 2015 to January 2017, a total of 419 cases of patients who received hysteroscopy due to abnormal uterine bleeding (AUB) and infertility, the results of ultrasound showed that the endometrium echo was uneven and suspected endometrial lesions in Taihe Hospital of Shiyan City were selected as research subjects. Endometrial lesions of all patients were observed by ordinary white light imaging (WLI) and NBI hysteroscopy, and endometrium biopsy. The sensitivities and specificities of WLI and NBI hysteroscopy in the diagnosis of EH and endometrial cancer, and the sensitivities of these two different modes in the diagnosis of pathological classifications of EH were calculated as the diagnosis of endometrium biopsy was gold standard . And the sensitivities and specificities of WLI and NBI hysteroscopy were compared by chi-square test. This study met the requirements of the World Medical Association Declaration of Helsinki revised in 2013. Informed consent was obtained from every subject and every subject signed the informed consent form. Results Endometrium biopsy was used as the gold standard for clinical diagnosis of endometrial lesions. ①WLI and NBI hysteroscopy for the diagnosis of EH and endometrial cancer: the Kappa values were 0.793 and 0.843, respectively; the sensitivities were 83.5% (111/133) and 91.7% (122/133), respectively; the specificities were 94.8% (271/286) and 93.7% (268/286), respectively. The sensitivity of NBI hysteroscopy in the diagnosis of EH and endometrial cancer was higher than that of WLI hysteroscopy, and the difference was statistically significant (χ2=4.186, P=0.041). However, for the comparison of specificities, the difference was not statistically significant (χ2=0.289, P=0.591). ②WLI and NBI hysteroscopy for the diagnosis of pathological classifications of EH: the sensitivities were 66.9% (89/133) and 71.4% (95/133), respectively, and the difference was not statistically significant (χ2=0.635, P=0.426). ③EH showed uneven or focal thickening on the surface of endometrium under WLI hysteroscopy, with ethmoidal changes in some areas, and occasional distribution of small blood vessels on the surface of endometrium. EH showed a branch or reticular distribution of the thickened endometrium, with a relatively regular shape, relatively uniform blood vessel and no obvious large blood vessels was found under NBI hysteroscopy. Endometrial cancer showed morphologic changes in endometrium, thickening endometrium as gyrus or lobular changes under WLI hysteroscopy, and the surface of endometrium was as rough as the villi, and the texture was bad and crisp under WLI hysteroscopy. Large, naked blood vessels with different diameters, and tumor-like or frog oval blood vessels could be found in endometrial cancer under WLI hysteroscopy. The changes of blood vessels mentioned above were more obvious under NBI hysteroscopy. Conclusions NBI hysteroscopy is more sensitive than WLI hysteroscopy in the diagnosis of EH and endometrial cancer, but there is no difference in the sensitivity of the diagnosis of pathological classifications of EH by WLI and NBI hysteroscopy. For the diagnosis of endometrial cancer, NBI hysteroscopy does not benefit more than ordinary WLI hysteroscopy. NBI hysteroscopy will be limited in the diagnosis of endometrial lesion when there is inflammation, extensive congestion or active bleeding in the endometrium. Key words: Hysteroscopy; Endometrial hyperplasia; Endometrial neoplasms; Narrow-band imaging; White light hysteroscopy; Women