Liquid-based endometrial cytology: its possible value in postmenopausal asymptomatic women

Abstract

Buccoliero AM et al. (1) should be credited for their study on liquid-based endometrial cytology in the management of a subpopulation of asymptomatic postmenopausal women. Their conclusion that liquidbased endometrial cytology may reduce unnecessary, more invasive, and expensive procedures is striking. However, they have obtained a specificity of 94% for endometrial cytology after categorizing the cases studied as nonpathologic and pathologic. If we were to categorize them as malignant and nonmalignant cytology, the resultant sensitivity would only be 83%. Cytohistologic correlation had been possible only in 94 cases (29%) of the study population, a fact that demands concern on the two techniques. Two hundred twenty-two biopsies (69%) were inadequate, a figure which is surprisingly high when compared to other similar studies (2) . It may be related to the prevalence of nonneoplastic or hyperplasic cases or due to the sequence of endometrial sampling: first the cytology and then the biopsy as hypothesized by the authors. A brief note on the time gap between endometrial cytology and biopsy is appreciated. We observed the possibility of bias of the investigators who were keen to highlight the efficacy of endometrial cytology in comparison to biopsy.This exercise adds some valuable information to the excising body of knowledge on managing asymptomatic postmenopausal women with liquid-based endometrial cytology. Low cost, good tolerability, and the fact that it can be carried out as an outpatient procedure (3) makes it an excellent option for women with other comorbid factors. It opens the door for future trends and advances in this method such as immunohistochemical staining rather than routine Papanicolaou stain of the cytology obtained. Gynecologists should be made aware of the advantages of liquid-based endometrial cytology in assessing the endometrium.

Image analysis and multi‐layer perceptron artificial neural networks for the discrimination between benign and malignant endometrial lesions

“…12,29,38,42 The rationale of the current study lied in the fact that carcinogenesis causes destruction of the nuclear-skeletal structures and of the Nuclear morphological characteristics can be evaluated both by conventional and liquid-based cytology; nevertheless, the latter seems to overcome inherent difficulties of the conventional approach, such as remaining cells on the sampling device 9 or the frequent presence of protein, mucus and blood 10 using a standardized procedure regarding fixation, transfer and cytological preparation of the biological material. 7,11,12 Integrating the present findings into current clinical practice, endometrial cytology assisted by ANN-MPL could be an accurate tool, which could allow clinicians to decide which women with endometrial pathology would require more radical procedures, such as dilatation and curettage, and who not. Moreover, cases classified as benign could opt for medical treatment or for followed-up at regular intervals.…”

“…The DNA amount and the chromatin distribution are expressed via nuclear color characteristics and optical density while chromatin distribution by texture characteristics such as margination and heterogeneity. Nuclear morphological characteristics can be evaluated both by conventional and liquid‐based cytology; nevertheless, the latter seems to overcome inherent difficulties of the conventional approach, such as remaining cells on the sampling device or the frequent presence of protein, mucus and blood using a standardized procedure regarding fixation, transfer and cytological preparation of the biological material …”

“…In the case of endometrial cytology, however, several factors are limiting the diagnostic accuracy of the method, including the amount of remaining cells on the sampling device and the presence of protein, mucus and blood, which often contaminates the specimens . Liquid‐based cytology (LBC) represents a novel method of cytological analysis which overcomes these barriers: the technique allows cytomorphological evaluation of nuclear features, including geometric and densitometric analysis …”

“…10 Liquid-based cytology (LBC) represents a novel method of cytological analysis which overcomes these barriers: the technique allows cytomorphological evaluation of nuclear features, including geometric and densitometric analysis. 7,11,12 Artificial neural networks (ANNs) are complex mathematical models inspired from biological neural networks, which are capable to learn and adapt their predictive algorithms after the introduction of new information. 13,14 This potential makes ANNs suitable for the classification and prediction of tasks where various datasets are implemented.…”