Guidance Recap Podcast | Use of Whole Slide Imaging in Nonclinical Toxicology Studies: Questions and Answers
Thank you for joining us for another episode of the Guidance Recap Podcast. My name is Kylie Haskins, and I am today’s host. In this episode, I am excited to be talking with Dr. Lynda Lanning, a toxicologic pathologist in CDER’s Office of Translational Sciences. Dr. Lanning will be sharing some thoughts with us on the newly published final guidance titled, “Use of Whole Slide Imaging in Nonclinical Toxicology Studies: Questions and Answers.” Welcome, Dr. Lanning! Thank you for speaking with us today.
To begin today’s episode, can you provide some background for the listener about whole slide imaging and describe the scope of the guidance?
Sure. Whole slide imaging includes the software and hardware used to generate a two-dimensional digital image of a glass histology slide. These slides are used for routine assessment in the generation of the pathology report, and typically, these glass slides would be viewed on a microscope. With whole slide imaging, the slide is put on a scanner, which captures a digital image, and then that image is stored in a software system that pathologists can access anywhere in the world from their computers. So, for example, if there is a pathology peer review in a separate location from the study pathologist, electronically providing the stored image for the review can eliminate the need for shipping the glass slide.
This guidance covers the use of whole slide images evaluated for good laboratory practice (GLP)-compliant nonclinical toxicology studies using non-human specimens. If a nonclinical study is supporting a claim of safety, it must be GLP compliant. The guidance does not cover the use of whole slide imaging for clinical applications. Additionally, the use of whole slide images in casual consultations, opinion exchanges, and/or mentoring among pathologists is not covered by this guidance. Also, while the FDA does recognize the potential of and interest in artificial intelligence, or AI, in this space, the application of that technology is outside the scope of this guidance.
What are some of the reasons FDA issued this guidance?
FDA issued this guidance to provide the Agency’s expectations for use of whole slide imaging in nonclinical toxicology studies to provide clarity for stakeholders and help ensure consistency for the use of these images. For example, FDA would like stakeholders to know that the Agency doesn’t regard whole slide images as specimens or exact copies of the slides and whole slide images don’t have to be retained unless raw data are generated from the images. For example, if the images are used in lieu of glass slides for the primary histopathology assessment, they would be used to generate raw data, because in pathology, raw data is the signed pathology report. To reiterate, if whole slide images are reviewed instead of the original glass slides, the whole slide image files do not need to be retained as study records and archived unless raw data are generated from those images. The guidance notes that the glass slides are study specimens and must be retained as such after study finalization under GLP regulations. In addition, there are some limitations related to the types of scanners, hardware, and software used to generate whole slide images.
Whole slide imaging really took off during the COVID pandemic, as travel was restricted. Pathology peer review was often conducted through the evaluation of whole slide images during this time, while the original assessment was done by a study pathologist using the glass histology slides, which for GLP studies must be archived under Good Laboratory Practice, or GLP regulations. We are happy to note that the evaluation of whole slide images is increasing. This guidance provides our expectations for use of whole slide imaging in nonclinical toxicology studies, and FDA hopes that the document will help support its increased use.
Let’s get into more specifics about the recommendations provided in the guidance. Can the pathologist mark up the whole slide image during the evaluation process?
Pathologists can make annotations and adjustments through the viewing software but should not alter the original image. For example, a pathologist could change the brightness and create annotations through the viewing software such as adding arrows or notes, and we would not consider these actions to be permanent alterations of the image. We don’t need to see documentation for those types of adjustments, and pathologists don’t need to retain them in yet another whole slide image file.
Can you discuss the level of validation recommended for the whole slide imaging technology?
Yes. First, let’s discuss some definitions for use in this discussion. Qualification is the act of proving and documenting that equipment is properly installed, works correctly, and complies with specified requirements. Validation is objective evidence that provides assurance that a specific process will consistently meet predetermined specifications for its intended use.
The entire process used to capture and view the whole slide images, including the software and hardware, should be validated for its specific use to ensure whole slide images can support accurate and precise histopathology assessment in place of the glass histology slides.
What type of documentation is recommended for the use of whole slide images?
FDA expects written procedures to be in place if raw data is going to be generated. These written procedures could include slide scanning, training, software management, and file access and exchange control. All of this should be done to ensure that the slides are being scanned and used in a consistent, understandable, transparent manner.
Additionally, the signed pathology report or peer review statement should say that whole slide images were evaluated in lieu of glass slides for transparency. The FDA wants to understand the manner in which the data was generated. So, if a pathology peer review is performed, the pathology peer review statement should indicate whether glass slides or whole slide images were reviewed.
This guidance is being issued by seven centers. Can you discuss this briefly?
Sure, we always intended that this guidance would be a multicenter effort. Each center regulates different products, and aspects of nonclinical toxicology studies, including the pathology requirements, can vary for each product to best evaluate the potential toxicities and adverse effects. While the expectations for GLP nonclinical toxicology studies vary across the centers, the use of whole slide images is quite similar, and FDA wanted to provide a guidance that communicated the agency’s collective recommendations. Of note, all the centers also came together to issue the FDA guidance Pathology Peer Review in Nonclinical Toxicology Studies: Questions and Answers.
What has changed from the draft version to the final version of the guidance?
The draft version of the guidance was published in April of 2022. As I mentioned earlier, FDA issued this guidance to provide the Agency’s expectations for use of whole slide imaging in nonclinical tox studies. The public comments we received for the draft version of the guidance were overall very positive and expressed appreciation for the direction we provided in the guidance. We also received several comments on the draft version that noted it is burdensome to keep whole slide image files if they aren’t used to generate raw data, as the image files can be enormous. As a result, storage can be a huge issue if a sponsor is conducting a large study with lots of tissues. To address these comments, we limited the expectation for retaining the whole slide images to those used to generate raw data for GLP studies in the final version of the guidance. Several comments requested that we adopt certain language, specifically the term “faithful replica.” We declined to do so, as that would require new terminology and associated standards. Whole slide images are not an exact copy of the glass histology slides, so making up a new term would not provide clarity or a difference in procedures.
For our last question, can you mention a couple of key items that you would especially like the audience to remember?
Yes, I’d like to stress that in terms of using whole slide imaging, FDA does not consider the resulting digital image to be an exact copy of the glass slide. Also, the image should be generated in a validated whole slide imaging system following written procedures, and if it’s used to generate raw data, the image should be retained. Finally, we do understand that industry is moving more and more into whole slide imaging use and AI. We are monitoring it’s use and applicability and will continue to provide our expectations.
Dr. Lanning, thank you for speaking with us about the final guidance on use of whole slide imaging in nonclinical toxicology studies. We have learned so much from your informative discussion on the document. We would also like to thank the guidance working group for writing and publishing this final guidance.
To the listeners, we hope you found this podcast useful. We encourage you to take a look at the snapshot and to read the guidance.