After After
Before Before

10042 - CD56, Endometrial Tissue

INTRODUCTION

Uterine natural killer (uNK) cells in the pre-pregnancy endometrium have been associated with recurrent miscarriage. Immunohistochemical (IHC) analysis of endometrial tissue sections has demonstrated increased density of CD56+ CD16- uNK cells in the mid-luteal phase endometrium in women suffering from idiopathic recurrent miscarriage and recurrent implantation failure compared to fertile controls.[See REFERENCES: 1]
This APP measures the ratio of uNK cells among the stromal cells near the epithelial edge. Cells in the glands and in the epithelial edge itself do not contribute to the calculations.
This APP is designed to work on approx. 3 µm sections of endometrial tissue immunostained with anti-CD56 monoclonal antibody and counterstained with Mayer’s haematoxylin.

KEYWORDS

Endometrium, uterus, CD56, uterine Natural Killer cells, uNK, epithelial edge, haematoxylin, immunohistochemistry, quantitative, digital pathology, image analysis.

METHODS

This APP includes 3 separate steps, each step contributing to the final output.
The first step identifies the tissue and the glands, based on intensity and local shape [See REFERENCES: 2]. The tissue is subsequently defined as the initial region of interest (initial ROI), and the glands are excluded from this region [See EXAMPLES: figure 1 and 2]. Manual interaction may be needed to refine the initial ROI e.g. removing additional glands.
The second step identifies the epithelial edge within the tissue based on saturation and local shape. It outlines a final ROI up to 300 µm into the tissue from the epithelial edge, still excluding glands [See EXAMPLES: figure 1, 2, 3 and 4].
The third step identifies uNK cells and stromal cells within the final ROI, and calculates the relevant output variables, including the ratio of uterine Natural Killer (uNK) cells to stromal cells [See EXAMPLES: figure 5 and 6].

QUANTITATIVE OUTPUT VARIABLES

The output variables obtained from this protocol include:

- uNK cell Area: The total area of CD56-positive uNK nuclei

- Total Area: The total area of both stromal and uNK cells

- uNK cell Area per Total Area = (uNK cell Area) / (Total Area): The area of uNK cell nuclei divided by the total area of cells


AUXILIARY APPS (included)

Auxiliary APPs are used for additional process steps, e.g. finding Region of Interest (ROI).

Identify Tissue

The auxiliary APP '01 Find Tissue' can be used to automatically detect the region of interest (ROI). The APP utilizes a Bayesian classifier to detect the tissue areas of interest [See EXAMPLES: figure 1 and 2]. It might, sometimes be needed to manually remove areas of no interest subsequently to e.g. exclude normal and necrotic tissue.

Identify tissue components

The auxiliary APP '02 Find Epitheleal Edge' can be used to automatically identify the epithelial edge within the tissue based on saturation and local shape. It outlines a final ROI up to 300 µm into the tissue from the epithelial edge, still excluding glands.


STAINING PROTOCOL

There is no staining protocol available


ADDITIONAL INFORMATION

The APP utilizes the EngineTM and Viewer software modules, where EngineTM offers an execution platform to expand processing capability and speed of image analysis. Viewer gives a fast review together with several types of image adjustment properties ex. outlining of regions, annotations and direct measures of distance, curve length, radius, etc.
By adding the AuthorTM module the APP can be customized to fit other purposes. AuthorTM offers a comprehensive and dedicated set of tools for creating new fit-for-purpose analysis APPs, and no programming experience is required.

WORKFLOW

Step 1: Load auxiliary APP for tissue detection '01 Find Tissue' and run on the image to process. This outlines the tissue and the glands with a Region Of Interest (ROI) and saves time in the subsequent analysis steps. It may be necesarry to refine the ROI to remove e.g. glands or other residuals.

Step 2: Load auxiliary APP '02 Find Epitheleal Edge' and run on the image to identify the epithelial edge within the tissue based on saturation and local shape. It outlines a final ROI up to 300 µm into the tissue from the epithelial edge, still excluding glands.

Step 3: Load APP quantification protocol '03 Count Cells' and run on the images to process. The third image processing step identifies the uNK and stromal cells, within the specified ROI, and calculates the output variables.


REFERENCES

LITERATURE

1. Endometrial cell counts in recurrent miscarriage: a comparison of counting methods, Josephine A Drury, Helena Nik, Robbert H F van Oppenraaij, A. W. Tang, M. A. Turner, S. Quenby, Histopathology 59, 1156-1162, 4/2, 2011.

2. Learning histopathological patterns, A. Kårsnäs, A. L. Dahl, R. Larsen, Journal of Pathology Informatics, 2(2):12, 2011

RUO
FIGURE 1
FIGURE 1
Endometrial tissue section. Overview image at 0.5x.
FIGURE 2
FIGURE 2
The same image as in Figure 1 after automated segmentation for identification of the region of interest (ROI). The black area is glass background. The blue area is tissue area excluded from the analysis due to distance to the epithelial edge. The red area is the final ROI defined as being tissue within 300µm of an epithelial edge.
FIGURE 3
FIGURE 3
This field of view image of the endometrial tissue section at 20x shows a gland with its lumen (left) and the epithelial edge (right).
FIGURE 4
FIGURE 4
The same image as in Figure 3 after automated exclusion of the gland with its lumen, the epithelial edge and the glass background.
FIGURE 5
FIGURE 5
This field of view image of the endometrial tissue section at 30x shows the excluded, but visible epithelial edge (lower right), and the proximal cells, including uterine Natural Killer cells (dark-brown nuclei due to CD56 IHC) and stromal cells (purple nuclei due to hematoxylin staining).
FIGURE 6
FIGURE 6
The same image as in Figure 5 after automated detection and separation of uNK cell nuclei (red segmentation) and stromal cell nuclei (blue segmentation) within the final ROI. The epithelial edge is excluded.