The GLP mindset

In her first column for Bioanalysis Zone, Cathy explores how adopting the mindset and principles of GLP (Good Laboratory Practice) can enhance scientific rigor, reproducibility and troubleshooting in both regulated and non-regulated bioanalytical labs.

Catherine Vrentas
Life Science Lead Associate
Booz Allen Hamilton (VA, USA)

Cathy Vrentas is currently a Life Sciences Lead Associate at Booz Allen Hamilton, where she supports a large portfolio of federally-funded, preclinical through clinical programs in the CGT space and specializes in clinical trials and bioanalysis. Previously, she was a Principal Scientist and managed a team of ~20 scientists at Thermo Fisher Scientific (VA, USA). In this role, she led the development and validation of 100+ assays to assess samples for preclinical programs and clinical trials for pharma and biotech, including work on multiple first-in-human trials for rare diseases and gene therapies. Cathy has experience in regulated immunoassays and cell-based assays for PK, ADA, NAb and biomarker applications, as well as enzymatic assays, oligonucleotide assessments, COVID-19 assays and tissue-based assessments.

Cathy received her BSc in Biochemistry and Molecular Biology from Penn State (PA, USA), a PhD in Cellular and Molecular Biology from the University of Wisconsin-Madison (WA, USA), an MBA from Longwood University (VA, USA), and an MPH in public health practice from Des Moines University (IA, USA). She has mentored over 60 scientists, students and summer interns in laboratory methods over her career and has volunteered for diverse nonprofits including the foster care system, dementia education, prison education, youth science outreach, public health, oyster restoration and health advocacy.

I was well into my career as a Molecular Biologist when I was first introduced to the GLP laboratory—a decade past my PhD, during which I had spent time in the halls of a teaching university and led research on veterinary pathogens as a government Principal Investigator, among other things. I was no stranger to running ELISA plates, preparing calibration curves or designing cell-based assays. And initially, as I completed my first week of training as a scientist at a contract research organization, GLP seemed mostly like a set of rules for documentation.

“Make sure to record information by ALCOA principles (Attributable, Legible, Contemporaneous, Original and Accurate)”, I diligently noted down in my notepad next to my morning coffee. “Make sure to initial and date entries consistent with the Standard Operating Procedures”, I scribbled.

What I didn’t know at the time, however, was that my experience in GLP bioanalysis would remodel my approach as a scientist, as opposed to serving simply as a set of rules or regulations to learn. In fact, I propose that an understanding of the GLP mindset can help us increase the reproducibility of biomedical science across many settings, starting from the early discovery research lab. The original GLP regulations date back to the 1970s, when chemical and pharmaceutical testing, including critical safety testing, was poorly conducted (and even outright fabricated) at laboratories such as Industrial Bio-Test Laboratories [1]. Since then, GLP has played a key role in ensuring that the results of preclinical testing are interpretable, traceable and provide the quality data that is needed to support the progression of therapeutics into clinical studies.

However, the broad issue of reproducibility across scientific research remains, with the need to improve the likelihood that new discoveries can be confirmed and then advanced into products in an industry setting. Working groups, especially across publishers, have come together to advance data reporting standards, prepare recommendations for journal article method sections and outline procedures for statistical review in an attempt to improve reproducibility. Similarly, granting agencies like the National Institutes of Health (MD, USA) have instituted requirements for reagent authentication and rigor. From a day-to-day perspective for the bench scientist, though, the mindset of GLP bioanalysis provides not only improvements in simple documentation for traceability but also the potential for significant enhancements in experimental design and troubleshooting capabilities.

Below, I’ve outlined a few examples of how my time in the bioanalytical lab has molded my scientific mind. The discussion is focused on how this way of thinking can apply to non-regulated settings to improve consistency, traceability and reproducibility across the drug development continuum.

Scientific detective work

We’ve all been there—a completed experiment, but the results just don’t make any sense. I remember my days as a graduate student, learning to probe bacterial gene regulation with tools like long polyacrylamide gels that took hours to prepare and run. Just as the scan appeared on the imager, my advisor would poke his head into the room as I sat puzzled, wondering what might have gone wrong.

What I quickly learned in the GLP bioanalytical lab—even in the non-regulated, early stages of method development—was the superiority of the GLP mindset for the scientific detective, whether you envision yourself more as a Jessica Fletcher or an Olivia Benson. In addition to checking for errors in reagent preparation via documentation, capturing the timing of assay steps, reagent lot numbers and equipment used allows the bioanalytical development (or discovery!) scientist to prepare a grid of the conditions from the experiment in question, which can be easily compared to variables from recent experiments via an Excel™ template. Significant changes, such as a new batch of reagent or substrate with a longer development time, can be quickly highlighted in the grid, allowing the next experiment to assess the impact of these variations on the result.

I’m not proposing that method development or other non-regulated stages of the drug development process involve extensive, GLP-compliant documentation. After all, flexibility and efficiency are important. However, the use of well-crafted templates that can serve as fillable worksheets linked to a reference protocol can be implemented across a vast range of scientific settings, including academic laboratories, and can be simpler and faster than preparing notes for each experiment de novo.  Templates can also be applied for reagent preparation during discovery or development to ensure that the composition of a buffer prepared prior to your experiment is not the culprit. Not only does this approach facilitate troubleshooting, but it also facilitates consistency across days and scientists, increasing the chances that we can make sense of our data.

Where’s that file again?

A second shift in my brain, which was especially important as a people manager, was the power of systems for records. In the GLP laboratory, any reported result can be connected back to all of the raw data as well as the decisions and calculations used to obtain it. These could include the raw files from an ELISA run off of the spectrophotometer, the spreadsheets or Laboratory Information Management Systems (LIMS) used to calculate the averages of two duplicate wells, the assumptions in the calibration curve regression and information about any controls in the experiment. If the system is working correctly, 20 years from now I should be able to look at my data package from an experiment and understand exactly how I obtained all of the conclusions.

Though these principles may seem straightforward, management of collected and analyzed data can quickly get out of control, especially when each scientist in the laboratory uses their own system. In a development or discovery environment, developing a system for file naming and storage and creating a template for standardized calculation for tasks like protein concentration measurement or statistical comparisons of dose groups provides an excellent start. A centralized SharePoint™ can serve as an unofficial archive, with internal policies for the transfer of these files. Lab Managers and leaders, after ensuring that they have collected sufficient feedback from the team, should set the tone and lay the groundwork for these initiatives to make it easy for lab staff—who are already juggling pipetting, weighing, recording and meetings—to use the system.

Managing quality

While there are many potentially challenging situations in the bioanalytical lab, in my opinion, one of the most stressful for the method developer is the incoming Teams message from the validation group or the other lab, stating “the method isn’t behaving the same way in our hands.”

One GLP-inspired approach is to learn from the quality control (QC) and quality assurance (QA) practices of a regulated study, especially during the transition from assay development to regulated validation. One useful, skill-building approach is to integrate peer review at critical steps in method development. For example, a peer scientist can use a checklist against which they review a draft method protocol, using the most recent raw experimental entries (including reagent formulations) as the basis for analysis. Additionally, systems for calculation checks can reduce the incidence of propagating a preparation error in calibration standards that misjudges the sensitivity of a method. Lab staff can further develop skills as they take the lead on creating, reviewing and standardizing the use of templates, including training others. Finally, taking a page from the project management handbook, a requirements table—listing key method requirements like sensitivity, drug tolerance, and/or maximum sample volumes—can be independently compared to supporting development data. These practices adopt principles from the software development practice of Independent Verification & Validation if completed by a neutral reviewer, thereby reducing risk and saving time and money by identifying issues early.

Of course, none of the practices listed above provide a formalized QC or QA function, but instead these quality practices can be fit for purpose at different, non-regulated stages. Soon after I began my bioanalytical lab journey, I realized that much of the data review in discovery and even bioanalytical development involves processed data—bar graphs, comparison tables and figures that are lab-meeting ready. Making sure that raw data and methodology receive a second opinion can prevent downstream issues with brief upstream effort. With scientific integrity on everyone’s minds and having been the subject of the 2024 PharmSci 360 keynote presentation, these practices could also sound alarm bells in cases of intentional fraud before research studies go off to journal publication.

Next steps for the future

6 and a half years after my first foray into this bioanalytical journey, I find myself initialing and dating any changes I’ve made on a document, lab or otherwise without thought—I’m sure many of you share this habit. I hope I’ve convinced fellow bioanalytical scientists of the broad translatability and importance of our skillset. But more than that, we can take the schemas that have become ingrained in our scientific practices and help others apply them to diverse settings in ways that are matched with time constraints and quality needs. Practices that may seem like second nature are not necessarily adopted across relevant opportunities in the biomedical research and development community. As bioanalytical scientists, I call you to action; use your practical knowledge to create systems, share knowledge and practices with leaders in diverse settings and provide opportunities for training and immersion of junior scientists with a diversity of career goals.

References:

1. Schneider, K. Faking it. The Case against Industrial Bio-Test Laboratories. The Amicus Journal. 14–26 (1983).

Interested in our other columns? You can find our collection here.

This article was prepared by the author in their personal capacity. The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy, opinion or position of their employer, Bioanalysis Zone or Taylor & Francis Group.