A morphometric approach in breast cytology - geometrical descriptors in the differentiation between benign and malignant lesions

Vol. 55 No. 2 Suppl., 2014
This supplement was not sponsored by Outside Organizations.

ROMANIAN JOURNAL of MORPHOLOGY and EMBRYOLOGY

Daniela Mihalache, Simona Eliza Giusca, Raluca Balan, Cornelia Amalinei, Adriana Grigoras, Irina-Draga Caruntu

This study focuses on the analysis of geometric descriptors that can be applied in breast cytology, and their correlation with the qualitative features, with the aim to underline the differences between the benign and malignant cell profile. The morphometric investigation was performed on smears obtained by fine needle aspiration, 10 cases (group 1) diagnosed as benign and 10 cases (group 2) as malignant. For group 2, the malignancy was histopathologically confirmed on the surgical resection specimen. The sequence of automated operation, previously reported by us, permitted the extraction of the following geometrical descriptors: cytoplasmic area, nuclear area, nucleo-cytoplasmic ratio, equivalent diameter and form factor. We analyzed the differences between the benign and malignant morphometric features, and the correlation between the malignant morphometric features and cytological, respectively histological grading. Statistically significant difference in cytoplasmic areas, nuclear areas, value of nucleo-cytoplasmic ratio and equivalent diameter was noted between group I and II. For the form factor, we did not register statistically significant differences. For group 2, the correlation between the morphometric features and cytological grading revealed that the nuclear area is the most valuable descriptor, due to the significant differences between the three successive grades of cytological severity, followed by the cytoplasmic area and equivalent diameter, their numerical values presenting significant differences between cytological grade 1 and 3, and 2 and 3, respectively. The statistical analysis between the morphometric features and histological grading showed that nuclear area and equivalent diameter are the most viable indicators, due to the significant differences present between the three successive grades of pathologic severity, followed by cytoplasmic area (significant differences only for grade 2 versus 3) and for nucleo-cytoplasmic ratio (significant differences only for grade 1 versus 2). The form factor does not provide information that could be correlated with the cytological or histological grading. The defined morphometric features enable the characterization of benign and malignant cells and provide objective criteria that could support a differentiation of benign from the malignant pathology in the cytological diagnosis.

Corresponding author: Irina-Draga Caruntu, Professor, MD, PhD; e-mail: irinadragacaruntu@gmail.com

Download PDF

ROMANIAN JOURNAL of MORPHOLOGY and EMBRYOLOGY

Alina Simona Sovrea, Adina Bianca Bosca, Carmen Georgiu, Anne-Marie Constantin, Mohamed Amine Ben Abdalah, Dan Gheban

Astrocytes represent a heterogeneous population of specialized glial cells responsible not only for accomplishing various important functions in the healthy nervous tissue, but also for reacting to all types of the central nervous system diseases and trauma. GFAP immunostaining is considered to be the most accurate of the routine techniques used for identifying astrocytes; however, silver impregnation techniques, which are inexpensive and approachable, might be a reliable alternative. The present research has brought into question and tried to assess the sensitivity and specificity of these classical methods for identifying and differentiating normal and reactive from tumoral astrocytes. Our study included 10 supratentorial gliomas specimens of various grade and two normal brain samples. We performed a histological study on consecutive seriated sections labeled using four methods: the immunostaining for GFAP (glial fibrillary acidic protein) and the three silver impregnation techniques: Ramon y Cajal, Bielschowsky-Cajal and Gomori. For each tumoral case, two areas were examined: the tumoral parenchyma and the tumor borders (considered as reactive gliosis) and were both compared to healthy brain parenchyma; for each area, three microscopic fields were assessed and two parameters were recorded: a semi-quantitative score (the astrocytes density) and a qualitative score (the color intensity). We used a complex statistical analysis in order to process the data and to compare the diagnostic value of silver impregnation techniques versus GFAP immunostaining (the reference method) in terms of tumoral grading and differentiating tumoral from normal and reactive astrocytes. Our results indicated that there data provided by both GFAP immunostaining and silver impregnation techniques were comparable.

Corresponding author: Alina Simona Sovrea, Associate Professor, MD, PhD; e-mail: a_sovrea@yahoo.com

Download PDF
Download cover
Download contents

Journal archive