Mast Cells in the Microenvironment of Hepatocellular Carcinoma Confer Favorable Prognosis: A Retrospective Study using QuPath Image Analysis Software

We focuses on evaluating the distribution of immune cells within the tumor microenvironment. Firstly, we assist the infiltration of immune cells in tumor specimen and adjacent around tumor tissue, and then we determine their prognostic impact in hepatocellular carcinoma. The first challenge is the complexity of the tumor structures, which makes separation between tumor and non-tumor tissue challenging.

In addition, manual research for cell quantification are time consuming. Also, unspecific staining is an inherited problem for all immunohistochemistry methods. Most of prognosis conferred by master cells in the buried tumor area and the inner margin are used for first research on master cells in the hepatocellular carcinoma and other cancers.

Such analysis would profit from using image analysis software for accurate annotation of the region of interest and to correct quantification of the immune cells. To begin, acquire the images of immuno stained HCC sections. Download and open QuPath 0.4.3.

exe file on the computer. Create a new folder with a suitable name for organizing QuPath projects. Click the create project button in the software's upper left corner and select the newly created folder.

Then drag scanned images into the QuPath software window. Once a new window appears, select set image type and click H-DAB, followed by import. To view the imported images, check the list in the left window of the menu and double click on an image to open it.

Select file from the main menu and proceed to save the images as a project. Choose the brush or wand tools in the main menu to annotate the tumor area. In cases of discontinuous tumor regions, annotate each region separately.

Press and hold the control key while selecting all regions. Then right click and choose edit multiple and select merge selected. Afterward, click on automate in the main menu and select show script editor.

Copy the required script, paste it into the script editor, then click run. Alternatively, select the polyline tool from the main menu to accurately draw a border separating the malignant cell nests and the adjacent non-tumor tissue. Then select the border from the main menu.

Choose objects, click annotation, and select, expand annotations. Set the expansion radius to 500 micrometers and choose flat for the line cap. Activate remove interior, and ensure constrained to parent is activated.

Now, right click on the newly created annotation. Select edit single and choose split. To name the annotated regions correctly, right click on the annotation in the list on the left, select set properties and enter the appropriate names for inner and outer margin regions.

Define the remaining tumor area as the tumor center. Extend 500 micrometer wide peri-tumor area adjacent to the outer margin as demonstrated earlier. For detailed pixel analysis, select the rectangle tool from the main menu.

Then annotate a region encompassing all pixel types to be distinguished. Next, navigate to the main menu. Select analyze and click pre-processing.

Then click estimate stain vector. Upon activating the visual stain editor, select auto and click okay. Name the estimated stain vector H-DAB.

For pixel analysis, an alternative method involves using the rectangle tool from the main menu to highlight a small section of a nucleus stained with hematoxylin. Next, access the left menu and select image. Then double click on stain one and select yes.

Again, use the rectangle tool to annotate a small region showing positive staining with DAB. After this, select image from the left menu and double click on stain two, followed by yes. To assess the area fraction of positive immune cells from the main menu, click the classify option, select pixel classification, and click create thresholder.

After that, set the resolution to high, select the DAB channel, and apply a Gaussian pre-filter with a Sigma of 0.5. Adjust the threshold between 0.2 and 0.3. Define pixels above threshold as positive and below threshold as negative.

For the region selection, opt for any annotation, give the classifier a name, then click save, measure, and okay. To document the findings, copy the results displayed in the left side menu and transfer them into a spreadsheet. Select measure from the upper main menu as an alternative method.

Then choose, show annotation measurements. Select copy to clipboard and paste these results into a spreadsheet for a detailed record. Lastly, right click on the image, navigate to multi-view and click close viewer to close the image.

CD117 protein was predominantly observed in the cytoplasmic membranes of rounded cells in tumor stroma and perivascular spaces. CD117-positive cells were also observed in tumor capsules and perivascular areas of the outer margin and peritumor. NKp46 protein was mainly observed in cytoplasmic membranes of rounded cells in sinusoid-like spaces, the stroma of the tumor center, and the inner margin.

NKp46-positive cells were observed in capsules around tumor nests and stroma of portal tracts in the outer margin and peritumor regions. CD1a protein was mostly observed in the cytoplasmic membranes of rounded cells, either scattered or in aggregates in the stroma and sinusoid-like spaces in the tumor center and inner margin regions. CD1a positive cells were found in sinusoids and biliary epithelium in the peritumor region.