December 1, 2022

Updates to the OECD GLP Working Group’s Frequently Asked Questions

The OECD GLP working group has added new items to their Frequently Asked Questions (FAQ) page for GLP. Lots of updates this year, fourteen new questions in all. Topics run the full gamut, including the usual organizational concerns and clarifications on edge-cases. Here’s my take1Note: my ramblings include lots of rewording and interpretation of the original FAQ items. As such, it may not always reflect the subtlety of the working group’s carefully worded questions and answers!on this year’s additions to the FAQ:

Combining GLP Roles in Small Facilities

Finding the right people for the job is a challenge for any organization. A small GLP facility won’t necessarily have the resources to afford a dedicated archivist or QA. A solution commonly proposed is to have one person wear several hats. So which GLP roles can be combined?

  • As you might expect, the FAQ calls out QA’s need to be independent of study conduct. Furthermore, QA must be under the direct responsibility of TFM. Therefore QA is limited in what additional roles they may take on. And since QA can’t audit themselves, you need to ensure there are other, independent individuals to audit those roles.
  • Net-net, QA is definitely not allowed to be TFM, the study director or study personnel. The FAQ does, however, allow for someone to be QA in one department while participating in a GLP study in another. If you do this, make sure QA’s independence from study conduct is absolutely clear. This could be emphasized in the Org Chart, for example.
  • Other amalgamations (for example TFM + Archivist) should be considered on a case-by-case basis. The FAQ recommends discussing any accumulations of GLP roles with the relevant monitoring authority.

Who (else) goes on the Org Chart?

One of the new questions in the Computerised Systems section suggests that the GLP advisory document #17 requires “any function involved in the validation of a computerised system” to be on the facility’s org chart. Now, I don’t remember the org chart being mentioned at all in that document, and a quick search didn’t help. However I think the wider question of how to properly document validation roles and relationships still stands:

  • Computerised systems validation (CSV) often involves people outside of the GLP facility, for example IT departments or external consultants. Such personnel should have their roles and arrangements fully described somewhere. According to the FAQ this could be in the org chart; in written service level agreements; in TFM delegation statements; or in any combination of these. The FAQ uses “and/or” here, suggesting that facilities do not necessarily need to include all such roles in the org chart.
  • Otherwise, as covered elsewhere in the FAQs, facilities may put any additional functions into their org charts. As long as the named GLP roles are properly represented, of course.

Edge Cases for Test and Reference Items?

Sometime I think that people like to ask questions of the authorities specifically to establish loopholes. Perhaps we see the GLPs as a tax? Take, for example, most of the new questions about test and reference items and samples. These basically boil down to: Do the rules still apply when a test item is ready-to-use? How about ‘formulated’ test or reference items? Are internal standards for bioanalysis considered reference items? As you might expect, in all cases the answer is that yes, the rules still apply. Specifically, the FAQs clarify that:

  • Samples from each batch of “ready-to-use” test item used in a long-term GLP study must be retained in the archives;
  • Internal standards for bioanalysis are considered reference items. Therefore they need to have a valid expiry date, for example based on stability data. Otherwise it is a deviation and the impact on the study should be explained and justified;
  • The rules still apply even if a test or reference item is ‘formulated’. You need to keep an account of its use, for example in a use log. This helps establish that the correct quantities have been administered or applied to the test system.

In a similar vein, if residual samples from bioanalysis need to be archived for study reconstruction then they must follow all the usual rules for archiving samples. For example: Kept to the end of their validated stability period; Stored in conditions that ensure stability; and Addressed appropriately in the study plan and final report.

So no new loopholes for us this year, I’m afraid. Time to move the lab to the Cayman Islands?

Collecting Non-GLP Samples in a GLP Study

This one, brief Q&A could easily be overlooked, but I think it’s an important addition for many sponsors and facilities. Sometimes it just makes sense to take additional samples during a GLP study for exploratory research. This can be critical when there is a limited quantity of test item. And of course we’re always looking for ways to reduce other resource use.

The FAQ gives us some clarity around this. If non-GLP sampling is conducted during a GLP study, then it needs to be transparent and not interfere with the rest of the study. All sampling during the study must be planned, executed and reported in a controlled manner. If study specimens end up being used for non-GLP purposes after study completion, then this should be reflected in a final report amendment.

Digital Histopathology

Digital histopathology is an entirely new topic being addressed in this year’s FAQ. They start with the basic question: Is it allowed in GLP studies?

  • Interestingly, the actual question asked was whether the compliance monitoring authorities object to its use. No clear answer is given here. Instead, the FAQ states that “The OECD Principles of GLP do not preclude the use of digitised histopathology slides in GLP studies for the histopathological assessment of tissue samples.”
  • The answer then goes into some detail about various considerations for it’s use in a GLP study. These amount to treating digital histopathology systems like any other computerized system used in a GLP study. For example:
    • Validating the digital pathology process, its associated equipment, and the process for transferring images for review to ensure data integrity and traceability. I go in-depth on data integrity in this post on ALCOA+;
    • Supporting the process with SOPs and training; and
    • Describing its use in Study Plans and Reports.
  • Should digitised slides used in GLP studies be retained in the archives? If you’ve used them in any way that makes them raw data then yes, they should. And they should be retained in a format that insures their integrity over time and allows for future review. If the study pathologist uses the digitised slides for their initial read, then those slides are raw data. If the peer reviewer uses them for their peer review, then the slides are raw data. The FAQ provides some more guidance on assessing the raw-dataness of digitised slides.

But I Instagrammed the amendment last week!

Sometimes the questions tell you alot more than the answers do. Take, for example, one of the archiving questions. It asks about how the GLPs are applied to the traceability and archiving of study communications through new information technology tools. A valid question. They go on to list examples such as client portals, social media and electronic messaging. And this got me wondering: Are facilities using Twitter, Facebook and Instagram for study communications?

You may think I’m joking, but there are more similarities than differences between these services and, for example, Slack, Sharepoint and Zoom. And these three are definitely used by many companies for project management and team communication. Just because they are being used by many businesses doesn’t mean they are automatically appropriate for use, for example, for critical communication in GLP studies.

Unidirectional by design

One of the reason businesses (and cat-meme-lovers) like these modern tools is because they provide an easy way to combine many forms of communication and related information into one platform. The tool then makes this information discoverable and shareable within that platform. However, these tools are often designed to be unidirectional. They suck in data from many sources, but provide limited ways of getting that information out in its complete (as in ALCOA+) form. This is because they want to keep you on the platform. That’s often at odds with the requirements of a regulated environment.

So how do the GLPs apply? Before using a new technology for study communication, evaluate that it’s appropriate. Document the use of the technology, apply ALCOA+, and archive it like any other electronic data. What if the media cannot be properly archived due to, for example, readability or security issues? The FAQ says that TFM should implement a process for the faithful transcription in a retainable (their word, not mine) format. I say that perhaps you should consider using something else.

Other gems from the updated FAQ

  • A system administrator can conduct electronic archiving of GLP data. The responsibility must be delegated from the archivist, and documentation of the roles and tasks is key. What I like about the FAQ’s answer here is that they highlight that you can’t just ignore what makes a system administrator special. System administrators have unlimited privileges! Therefore you need to have assessments of: a) potential conflicts of interest; and b) the impact that someone with unlimited system privileges may have. Rinse, lather, and apply appropriate risk control procedures as necessary.
  • The FAQs remind us that the OECD GLPs only require TFM to approve SOPs. TFM may delegate, if documented, parts of this responsibility to someone other than QA unless prohibited by the local regulations. Your SOPs would determine any other signatures to be included for SOP review or approval. While the OECD GLP consensus document #4 recommends that QA review all SOPs, this is not an OECD GLP requirement.

And finally, some good news hidden as an FAQ item. The OECD working group is drafting a GLP guidance on Data Integrity! Apparently they will post a draft on the OECD GLP website for public comment when it’s ready. I’m waiting with baited breath.

And so there it is.

I don’t think there are any big surprises in any of the answers of this update. Most clearly fall out of the usual GLP requirements and best practices. I do think we found a few gems in the questions, though! Now for the fine print. Please remember that my ramblings don’t necessarily reflect the subtlety of the OECD working group’s carefully worded questions and answers. I’ve definitely added interpretation and filled in a few gaps. Gaps that were probably intentional in the original.

As always, before taking any action on anything you read here you should definitely: read OECD GLP Working Group’s complete questions and answers; consult the Guidance Documents and Regulations; and of course discuss with your QA! Thanks for reading, and I hope you enjoyed it.

Brendan Hyland

I'm a computer geek, foodie, and father of two. Professionally, I’m a Registered Quality Assurance Professional (RQAP-GLP) and have spent most of my career writing SOPs and designing, implementing and maintaining quality system components for GLP laboratories. I’m also a programmer with a love for strictly-typed functional programming languages. I have developed and validated software systems and tools for data analysis, document control and workflows.

View all posts by Brendan Hyland →