Unravelling the impact of ICH E6(R3) on Good Clinical Practice

Good Clinical Practice (GCP), the bedrock of ethical and high-quality clinical research, guides operations among sponsor companies, contract research organisations (CROs), investigator sites, and others. However, like other aspects of drug development, it is evolving.

The International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) released the ICH E6(R3) draft guideline for public consultation in May 2023 in collaboration with global health authorities, including the FDA. The draft revises ICH E6(R2), and regulatory agencies are currently reviewing public comments. They anticipate beginning public consultation on Annex II in 2024, with the final ICH E6(R3) guidelines expected to be released in 2025.

E6(R3) aims to advance the themes raised in E6(R2), but make them more flexible, adoptable, and durable over the long term. The overarching goal is to encourage a risk-proportionate approach and a culture of quality that better accommodates ongoing innovation.

It is important to note that E6(R3) is evolving alongside standards like the recently adopted ICH E8(R1), a revision to the General Considerations for Clinical Studies. Indeed, E6(R3) builds beautifully on E8(R1), pulling through concepts including stakeholder engagement, Quality by Design (QbD), and Critical to Quality (CtQ) factors. Together, the two sets of guidelines work hand-in-glove to ensure clinical trials remain fit for purpose - at both the planning stage and throughout execution.

Against this backdrop, we will explore what E6(R3) means for clinical trials moving forwards, as well as how organisations can prepare.

Key concepts

One challenge with E6(R2) is that it sometimes becomes more specific and complicated than necessary. Too often, compliance efforts entail a one-size-fits-all, check-box mentality. E6(R3) attempts to break down unnecessarily rigid barriers to increase agility and ease adoption.

The concept of “Quality Tolerance Limits” (QTLs) offers a prime example.

The QTLs mandated by E6(R2) are not easy to execute. Any deviations from these prescriptive, predefined thresholds must be recorded in the clinical study report. Thus, they have created a perception that regulators expect perfection - which, in turn, has inadvertently infused a “fear factor” into many study operations.

By comparison, E6(R3) now softens and expands the QTL concept into one called “Acceptable Ranges” that allows a broader range of control measures to be applied. By no means does this lessen the vigilance or scientific rigor demanded. However, it enables continual adjustment and realignment, opening the door to greater collaboration and agility.

The updates throughout E6(R3) - and E8(R1), too - make it clear that the guidelines are not about achieving perfection. Rather, they are about protecting what matters most. They embed risk-proportionate approaches to quality management throughout the clinical trial lifecycle.

Risk-Based Quality Management (RBQM) is less about “dotting i’s and crossing t’s” and more about focusing on those processes and data with the most critical impacts on participant safety and study outcomes, including the ability to make data driven decisions.

To that end, E6(R3) also replaces the term “error(s)” used in E6(R2) with the phrase “harms/hazards”. This revision signals that not every error warrants stiff scrutiny. Only issues that present a harm or hazard should rise to the level of robust investigation, root-cause analysis, and preventative actions. For example, if a handful of mild headaches aren’t captured as adverse events in an oncology trial, does this really cause any harms or hazards, or could this be simply corrected?

E6(R3) challenges us as an industry to stop treating every error the same way and instead concentrate on those that are genuinely harmful or hazardous to data quality and participant safety. Every case might not require a root-cause analysis. Still, we must be able to explain what went wrong, the extent of the problem, when it occurred, and how it was addressed.

Implications for innovation

In addition to embracing RBQM, E6(R3) is also largely about sustainability in the face of rapid technological advances and skyrocketing volumes of data.

Consider that any given Phase III study averaged about one million data points a decade ago. Today, the average is roughly 3.5 million data points - and up to six million data points for complex indications such as oncology studies. Furthermore, rather than a single data source - the electronic data capture (EDC) system - studies now commonly manage data from up to half a dozen sources involving multiple vendors.

This explosion of data and technology reinforces the idea that we must take a risk-proportionate approach; it’s not humanly possible to create perfection with so much data from so many sources. Consequently, E6(R3) includes a brand-new section devoted to data governance.

As machine learning, artificial intelligence, and other technologies continue to evolve, E6(R3) emphasises mindfulness about the burdens imposed by data collection activities. In an oncology study, for example, are participants asked to have blood drawn every 20 minutes?

E6(R3) asks us to think critically about whether a study uses appropriate, validated systems and focuses on those things that ensure compliance and the health/integrity of the data. In so doing, it tries to help the industry successfully prepare for the unknown future.

Practical impacts

Admittedly, RBQM entails a substantial mindset shift. Yet, the speed with which COVID-19 vaccines were rolled out testifies to its benefits. Some practical steps organisations can take to prepare for the strategy and collaboration advocated by E6(R3) include:

• Simplify processes. Review all procedural documents, guidelines, etc. Look for places where they’ve been overengineered and strip out the inflexibilities. Align them with a risk-based approach, setting clear pathways for oversight, assessment, and remediation of issues.
• Lean into technology. Appreciate the complexity of clinical trial operations, give some grace to study partners, and increase reliance on technology - like data visualisations, for instance - to ease people’s burdens.
• Build critical thinking. Critical thinking is the essence of RBQM, which starts with strategic questions such as, “Why am I doing this?”, “Does this matter?”, and “Will this materially impact my outcomes?”. RBQM is not easy, but it works when organisations trust and embrace it.

Within each clinical trial, E6(R3) also encourages organisations to work together to:

• Define risk.
• Identify CtQ factors. (In other words, specify what matters most on the premise that, if everything is critical, then nothing is critical.)
• Engage all stakeholders - including CROs, sites, participants, regulators, etc.
• Incorporate stakeholders’ perspectives into clinical trial designs. In addition, QbD requires thinking about the desired quality endpoints from the start, and then designing backwards.
• Conduct frequent team check-ins to assess the study’s progress, work together to fix any errors or issues, and determine if any changes are needed.

The path ahead

E6(R3) recognises that a sanitised clinical trial with perfect data and execution is impossible - and that even striving for perfection is slowing us down.

In many ways, it is “the thinking person’s GCP”. Even as it lessens unnecessary constraints, E6(R3) puts a greater onus on us to think critically about what’s important and potential risks at every step of every study. It advocates for RBQM in all facets of decision-making to ensure safety and protect participants, while also driving efficiency and innovation.

The shift to RBQM may not be simple. However, because it drives a collaborative and patient-centric approach to studies from the start, it is likely to decrease the burden on sites, improve the participant experience, and enable more agile drug development processes. E6(R3) acknowledges that errors and problems occur in the real world - but, ultimately, we are all partners working to improve people’s lives.

About the authors

Nicole Stansbury joined Premier Research in May 2023 as the head of global clinical operations. Prior to joining Premier Research, Stansbury spent 25 years in the CRO industry, where she served in leadership roles including global head of clinical trial management, global head of central monitoring, and head of global clinical performance, a team responsible for SOPs, metrics, training, clinical systems, and clinical quality. Stansbury’s 30 years of industry experience has included positions at the site level and other CRO roles, such as CRA and project manager. Her therapeutic experience has primarily been in dermatology, gastroenterology, urology and women’s health, however, Stansbury has leadership experience overseeing trials in oncology, neuroscience, and general medicine. She has a Bachelor’s degree in Animal Science from Texas A&M University and a Lean Six Sigma Yellow Belt. Stansbury has also served as lead/co-lead for the Association of Clinical Research Organizations’ (ACRO) Risk-Based Monitoring Working Group since 2014, where she works with TransCelerate and global regulatory authorities on driving RBM adoption in the industry.

Madeleine Whitehead is a quality solutions leader at Roche Pharmaceuticals. During her nearly 20 year career in the industry, she has worked in both pharma and CROs as a process and Good Clinical Practice (GCP) specialist. With a robust background in risk-based quality management (RBQM), Whitehead has driven process optimisation and embedded RBQM at Roche. Demonstrating a consistent track record of delivering tangible results and fostering collaboration, she has successfully accelerated clinical and compliance objectives utilising methodologies like Design Thinking and Six-Sigma. Graduating with a BA (Hons) from the University of Nottingham, she brings a strong academic foundation to her professional endeavours. Whitehead deploys her expertise in industry collaborations as co-workstream lead on the TransCelerate ICH E8 and ICH E8 workstream, as well.