Protocol Development #56

Spotlight

[Regulatory Guidance] The FDA published draft guidance for industry on Diversity Action Plans to Improve Enrollment of Participants from Underrepresented Populations in Clinical Studies that replaces the 2022 draft guidance that was released before FDORA. In general, clinical study diversity helps ensure that clinical studies appropriately test the product in a representative sample of the product’s intended use (i.e., target) population. Factors to consider when setting enrollment goals include demographic characteristics (e.g., race, ethnicity, sex, age group), clinical characteristics (e.g., presence of comorbidities, disease etiology), and other characteristics (e.g., access to standard preventive and diagnostic care, access to standard treatments of the clinically relevant population). As the action plans are independent to the protocol, there is no logistical reason why a diversity plan cannot be developed once a protocol has been finalized - so how does this relate to protocols? In Section V part C (measures to meet enrollment goals), the action plan must include an explanation of how the sponsor plans to meet the specified enrollment goals. Examples include:

• Sustained community engagement
• Cultural competency and proficiency training for investigators and sites
• Improving study participant awareness and knowledge of the study
• Reducing participant burden
• Improving access through eligibility criteria
• Employing DCT solutions (when applicable).

From the examples provided, 50% of the recommendations (reducing burden, improving access through eligibility criteria, DCTs) directly impact the protocol. If the protocol has been developed before the diversity plan this would result in potentially extensive rework to revise existing content. An alternative, and more efficient, approach is to have a well-understood diverse target population prior to developing a diversity plan and the protocol. To achieve this, diversity exploration should be initiated in phase 2 studies that would then inform the phase 3 design and lead to parallel diversity plan and protocol development.

Patient Engagement

[Article - Methodology] Alger et al published Patient and public involvement and engagement in the development of innovative patient-centric early phase dose-finding trial designs that success the implementation of PROs in dose-finding studies as "the current landscape reveals a notable gap in understanding how to effectively embed patient and public involvement and engagement (PPIE) within statistical methodology". Using virtual sessions with 9 participants from the UK and Canada, the authors found participants were concerned that self-reported side effects may lead to intervention discontinuation - thus reducing the likelihood that participants accurately report frequency or severity.  Intra-patient escalation, and associated individual tolerability threshold reporting were identified as activities to encourage PRO completion.

[Article - Methodology] Salvaggio et al published How an Innovative Statistical Methodology Enables More Patient-Centric Design and Analysis of Clinical Trials that addresses a common challenge in clinical research, how to determine benefit of an intervention using a single primary endpoint (after all, clinical benefit can be interpreted differently by different stakeholders). The authors propose using generalized pairwise comparisons (GPC) to integrate multiple clinical outcomes into a single assessment. How can this impact protocols? The inclusion of a clinical primary endpoint is intended to provide a objective method of assessment - however, this may not reflect patient preference. Importantly, the authors note a growing number of protocols (using GPC methodology) are being approved by regulatory agencies. I'm not going to pretend I understand the statistical methodology itself, but by incorporating multiple outcomes into a single assessment could mean that study protocols could successfully incorporate patient preference into primary objectives.

Diversity, Equity & Inclusion

[Article - Commentary] Mackall et al published Enhancing pediatric access to cell and gene therapies where they propose a new entity, the Pediatric Advanced Medicines Biotech, to lead development and commercialization of pediatric GCTs outside of traditional biopharmaceutical development models in the US. In the article, the authors summarize the challenges in bringing CAGT to market in general, and  for pediatric populations, including manufacturing challenges, n-of-1 study designs (or accommodating for rare or ultra-rare populations), and licensing/funding. Study design details are light but nevertheless, it highlights some of the obstacles sponsors face when looking to conduct CAGT studies in pediatric populations.

Process & Templates

[Article - Methodology] Iskander et al published Protocol-in-a-Day Workshop: Expediting IRB Approval for Junior and Senior Faculty that details a protocol methodology designed to expedite protocol development and approval for investigator-initiated studies. The protocol in a day (PIAD) workshop focused on 6 components: physics, statistics, research data coordination, finance, scientific correlatives (including imaging, biospecimens and PROs), and scientific review/IRB. Each PI presented their concept to the group and then rotated through 6 break-out sessions to obtain specific feedback. Interestingly, compared to non-PIAD protocols, IRB approval was shorter for both junior and senior faculty members (figure above). In addition, while junior faculty mostly benefited from regulatory, statistics, and IRB review, senior faculty mostly found clinical research finance and statistics to be most helpful based on survey results (Figure 2).

Digital

CIDs/Master Protocols

[Article - Research] Dong et al published Use of Seamless Study Designs in Oncology Clinical Development– A Survey Conducted by IDSWG Oncology Sub-team. The authors received 51 responses to the 16 questions included in an emailing list survey and found 39 (76%) had seamless oncology studies in planning or implementation for registration purposes. The most frequently reported benefits were timeline, cost, other (including probability of technical success), and operational feasibility (figure above). The most frequently reported challenges were other (including pre-specified decision rules), operational feasibility (e.g., DMC or variable sample sizes complicating budgets and timelines), and regulatory advice.

Structured Content

[LinkedIn Post] Val Swisher 's article on What Are You Waiting For? The Tale of Doing Nothing hits home for a variety of reasons. In protocols, redundant content is frequently copied from one protocol to another, inadequate metadata limits reliable source identification, and you are never really sure whether the content approach the team has chosen is truly the most appropriate (i.e., has all options been considered or has the first solution identified been incorporated?). With AI muscling its way into every corner of our lives - what do we do? Hope for an AI "hail mary" that will solve all our woes or structure content to prepare it for more effective AI implementation? What we do know is that AI needs to be trained and adequate training means reducing: content redundancy, conflicting or inaccurate information, poor quality/absent metadata. As Val aptly concludes:

"Rather than continue to suffer from runaway content, get to work on fixing it now, before you have your AI engine. That way, your corpus is ready when your solution arrives."

Protocol Tech

[Article - Case Study] Riaz et al published Applications of Artificial Intelligence in Prostate Cancer Care: A Path to Enhanced Efficiency and Outcomes that discusses the rise of LLMs and AIs and their impact on study design and operations (among other things) via examining available literature related to prostate cancer. An interesting snippet for protocol developers (figure above):

"Some of the key areas where AI can play a role include protocol development, designing simple language consent forms, trial design, and optimization by enriching the enrollment of high-risk patients and creating synthetic control arms, matching patients to trials, data collection, and management regarding clinical outcomes and toxicities, automatic response assessment using RECIST, reporting of serious adverse events (SAEs), predictive analytics, and biomarker discovery using deep learning techniques."

Regarding study execution, outcomes ascertainment and predictive analysis would be incredible tools to support study design and protocol development as these would provide essential feedback loops that could help refine assumptions and improve the probability of technical success. Although we're far from reaching this goal, the opportunity has been presented.

Webinars

TransCelerate is hosting a webinar on Vulcan UDP (Utilizing the Digital Protocol): Collaborating to Accelerate ICH M11 and End User Value on Thurs 11 July 2024 from 15:00 - 16:30 CEST. This webinar will detail Vulcan’s Utilizing the Digital Protocol (UDP), an umbrella project that brings together multiple initiatives to develop a digital representation of a study protocol.