6 Ways to Improve the Quality of Quantitative Histopathology Data in 2022
5 December 2022
By Mark Laurie

Article Synopsis: In this article, OracleBio’s Deputy Clinical Operations Manager, Nicole Couper, outlines 6 of the most prevalent issues encountered in Histopathology that hinder or negatively impact data quality.

The generation of high-quality and robust tissue-based quantitative data is significantly dependent on exceptional quality sample preparation and good staining techniques and practices.

Failing to ensure that samples produced are of high quality can negatively impact subsequent image analysis processes.

This can affect the robustness and reliability of data quality, and in some cases, prevent an image analysis study from proceeding, resulting in additional unforeseen costs and delayed timelines for a project.

A considerable number of issues can arise during the analysis of sections prepared of poor quality.

1. Artefacts

Artefacts are described as extraneous objects or defects present in the tissue section that occur as an outcome of the preparation procedure.

Artefacts include folds or tears in the tissue, hair, dust particles, staining deposits or air bubbles and can negatively impact the generation of robust data due to the increased risk of false positive staining detection during image analysis.

It is critical to reduce the introduction or inclusion of potential artefacts during microscope slide preparation and to perform a rigorous assessment of tissue sections during quality control procedures.

Figures demonstrating shadow artefacts present on an IHC stained section (left) and multiple fold artefacts on an IF stained section (right).

2. Edge Effect

A consequence of tissue sections absorbing more stain on the outside of the tissue than the inside, resulting in a ring of non-specific staining around the edge of the tissue is a quality issue known as ‘edge effect’ and may lead to false positive quantification and inaccurate data readouts.

‘Edge effect’ can be the result of inconsistent fixation processing, tissue sections drying out or lifting around the edges. Although repeat staining may not eliminate this problem care with respect to fixation, drying, storage time and environment, for cut slides, should be considered.

Figures exhibiting examples of edge effect on an IHC stained section (left) and IF stained section (right).

3. Intra- and Inter-run staining variability

Intra-run variability is a staining issue that occurs when IHC/ IF stained slides, from the same processing run, display noticeable variation in staining intensities, which are not considered to be the result of the biological variance. This can be caused by incorrect dilution of reagents or machine failure resulting in inconsistent application of reagents.

Inter-run variability occurs when runs performed at different times or on different days result in staining variation across a series of slides or different studies. This variability can be caused by multiple aspects including issues with automated staining platforms or by using different batches of antibodies or reagents.

In image analysis, inter- and intra-run variability occurs can be challenging for image analysis scientists to make accurate comparisons between samples within or across studies, impacting the robustness and reliability of data results.

Automated staining platforms that include environmental controls and utilise standardised reagents can reduce potential intra- and inter-run variability.

Figures showing inter-run staining variability of IHC stained slides on different staining runs.

Figures showing inter-run staining variability of IHC stained slides on different staining runs.

4. Staining gradient across a slide

Another common staining issue observed is the appearance of a staining gradient across a slide, characterised by strong staining at one side of the section fading to weak staining at the opposite side. Staining gradients can be caused by several protocol errors such as incomplete or uneven tissue fixation (potentially a consequence of thick tissue sections), application of reagents, issues with automated staining machines or by failing to keep the slides flat until they have dried.

During quality control reviews, sections displaying acute variations in staining gradients should be excluded. To avoid rejection of sections, repeat staining should be requested and performed before progression to image analysis.

Figures demonstrating staining gradient examples on an IHC stained section (left) and IF stained section (right).

5. Overstaining or Understaining

An additional problem commonly seen with IHC and IF staining is the under or overstaining of marker(s), particularly when staining is performed on a platform that is not environmentally controlled. Staining levels can be affected by numerous factors such as incubation time, incubation temperature and/or antibody concentration.

Staining of very weak or high intensity can produce false negative or false positive results. For example, a dark blue overstained haematoxylin nucleus may be difficult to distinguish from a dark brown DAB-labelled positive nucleus and can mask weakly DAB-labelled positive nuclei, when evaluating colour deconvolution during algorithm development processes. In addition, true morphological staining features can be extremely challenging to differentiate with inappropriate staining levels.

Validation of IHC and IF staining assays should be performed including precision and reproducibility evaluations and testing of counterstain application levels. Once completed, verified processes should be adhered to and regulated for all staining runs performed.

Figures demonstrating staining gradient examples on an IHC stained section (left) and IF stained section (right).

6. Scanning Errors

The digitisation of stained slides can give rise to several possible errors during the scanning process. One potential error includes irregular illumination lines which can be the result of a worn scanner stage, a degrading scanner lamp or an uncalibrated scanner. A considerable issue of digital slide scanning is the production of unfocused regions, predominantly caused by artefacts in or on the slide that can impact the focal plane for a particular field.

There are many preventative measures that can be taken to help reduce the instance of slide scanning errors including, but not limited to, regular maintenance of a slide scanner and system calibrations and implementing a standardised workflow for slide scanning to ensure a consistent approach is followed.

Figures showing scanning errors including out-of-focus regions on an IHC-stained image (left) and scanning banding on an IF-stained image (right).

In Summary

OracleBio offers project management of IHC and IF studies and provides support to reduce the negative impact that histology, staining and scanning issues can have on image analysis studies. On occasion, random artefacts are unavoidable and appear in tissue sections presented for image analysis.

At OracleBio, our dedicated image analysis scientists are trained to identify and exclude these undesirable features prior to analysis of the image. Investing the time to remove artefacts pre-analysis results in improved image analysis and provides better quality data. Manual exclusion of intruding artefacts on images, however, can significantly extend the time required to complete a study. In some cases, where multiple artefacts are present across a large sample set, for example, a Deep Learning algorithm approach can be utilized to recognise and exclude artefacts to obtain reliable and robust data results in a faster manner than manual exclusion.

It is also important to highlight the need for the inclusion of appropriate controls within each study or batch of staining and should be part of the study remit as standard. To establish the impact of staining variance a known positive control should be stained with each run/ batch of images and analysed to compare the coefficient of variation (CoV), based on the positive stain measures being evaluated and/or intensity or optical density. A pre-defined CoV criteria should be established to determine the acceptance level for variability.

OracleBio’s image analysis scientists have extensive experience in quality control checking IHC and IF stained images. We project manage studies to ensure IHC or IF images are of a high-quality standard and suitable for image analysis, facilitating rapid generation of high-quality robust data.

For more information on OracleBio’s image analysis services and project management, please don’t hesitate to get in touch with us.

*Image credit: INCISE project

James Going

About the author: 

Nicole Couper – Deputy Clinical Operations Manager

“With a B.Sc. in Biomedical Science, Nicole is one of OracleBio’s most proficient and respected Scientists. With experience from several years of supporting image analysis at the highest level and a recent promotion, she is now performing a valuable role supporting quality and compliance for our clinical studies within OracleBio as Deputy Clinical Operations Manager.”

Find out more about OracleBio's Pathology Expertise

OracleBio's Quantitative Digital Pathology Services

Stay up to date with OracleBio

Sign up to our mailing list to stay informed about news, webinars, events and more