ASA Connect

Dear all,

   ASA BIOP Section Statistics in Pediatric Drug Development Working Group calls for co-leads and interested researchers to join the working group. 

   Six new proposed sub-teams are forming. Please refer below for the descriptions. Each sub-team can have one or two co-leads, who define the research question, lead the sub-team to conduct research, produce publications, and share knowledge in conferences. If you are interested, please reach out to the working group co-chairs: Gamalo, Margaret <Margaret.Gamalo@pfizer.com> and Travis, James <james.travis@fda.hhs.gov>.

   Join the group to move forward the field!

Best,

Jingjing Ye

On behalf of the ASA BIOP Section Statistics in Pediatric Drug Development Working Group

----------------------------------------------------------

1. Pediatric Age Group Inclusion in Clinical Trials

• Adolescents and Young Adults: Statisticians should design trials that include adolescents and young adults by adjusting for age-related factors and ensuring that disease progression models incorporate the developmental stage of the patients. This includes developing statistical methods to account for the variations in treatment response and progression across different pediatric age groups.
• Infants and Children: Statistical models should be expanded to incorporate younger age groups, where development and disease manifestation may differ significantly from both adults and older children. For infants, especially, the inclusion of pharmacokinetic (PK) and pharmacodynamic (PD) modeling becomes crucial, as dosing schedules and drug metabolism differ from older children and adults.
• Statistical Framework: Develop flexible, age-adjusted models that incorporate age-specific parameters such as growth, maturation, and disease progression. Consideration should be given to differences in metabolism, immune response, and other factors across different pediatric subgroups, and how these factors influence the statistical analyses of efficacy and safety.
• Age-Specific Dosage Adjustment: Use pharmacometric approaches to determine appropriate doses for infants, children, and adolescents. Statistical models should account for these changes and help extrapolate findings from older populations to younger ones.

2. Therapeutic Area Prioritization and Age-Inclusive Models

• Therapeutic Areas: Focus on identifying therapeutic areas that span across various pediatric age groups, including infectious diseases, oncology, hematology, and chronic conditions like asthma or diabetes. Statistical models should be tailored to the age-related differences in disease progression and treatment response within these areas.
• For Infants and Children: In diseases like congenital conditions (e.g., hemophilia), or chronic pediatric conditions (e.g., asthma, type 1 diabetes), statisticians should ensure that pediatric age groups are considered in the initial trial designs. Using longitudinal modeling can help capture the evolving nature of these diseases and treatment effects across different pediatric populations.
• Data Integration: Statistical approaches should allow for the integration of data from both adult and pediatric populations to increase statistical power and efficiency. This may involve using methods such as cross-age borrowing or pediatric extrapolation, where data from adults or older children inform the design of trials for younger pediatric groups.
• Statistical Models for Disease Progression: Develop longitudinal models that capture how diseases progress over time in different pediatric subgroups, with a focus on how treatment effects vary across age groups.

3. Collaborative Statistical Approaches for Regulatory Alignment

• Regulatory Bodies and Guidance: Collaborate with regulatory bodies (e.g., FDA, EMA) to advocate for more consistent guidelines that facilitate the inclusion of all pediatric age groups in clinical trials. Statisticians should ensure that their statistical approaches align with regulatory expectations, particularly when proposing age-inclusive trial designs.
• Age-Specific Statistical Methodologies: Work with regulators to develop or refine statistical methodologies that can handle the complexities of including multiple pediatric age groups in a single trial. This could include adjusting for age as a continuous covariate, utilizing Bayesian methods for prior information from adult trials, and refining the use of dose-response models for different age groups.
• Extrapolation Across Age Groups: Statisticians should propose methods for safely extrapolating data from older pediatric age groups to younger ones, based on developmental considerations. This involves using available data from older children and adolescents to predict outcomes for infants or younger children, supported by robust statistical models that account for age-specific factors.

4. Integration of Real-World Data (RWD) and Long-Term Follow-Up

• Real-World Data (RWD): Integrate RWD to support the inclusion of all pediatric age groups, particularly for chronic conditions and diseases with long-term progression. RWD can help generate evidence on treatment efficacy and safety in populations that may be difficult to recruit into randomized controlled trials (RCTs).
• Long-Term Extension Studies: Statisticians should develop plans for long-term follow-up studies that capture both safety and efficacy data for pediatric populations over time. This ensures that the outcomes for all age groups are fully understood and accounted for in the final regulatory submission.

5. Modified Strategy in CDPs

• Early Pediatric Inclusion: Modify traditional CDP strategies to ensure that pediatric populations, including infants, children, and adolescents, are considered early in drug development. Instead of following the conventional route of adult trials first, clinical development plans should integrate pediatric age groups earlier in the process, particularly in conditions where similar mechanisms of action are shared between adults and children.
• Adaptive Trial Designs: Utilize adaptive trial designs that can incorporate multiple pediatric age groups and adjust based on accumulating data. These designs allow for flexibility in the dose escalation and inclusion of different age cohorts as the trial progresses, improving efficiency and ensuring more timely access to medicines across age groups.
• Pediatric-Specific Statistical Considerations: Modify statistical methods to ensure that age-related differences in pharmacokinetics, disease progression, and treatment response are adequately captured. Consider using dose-response models specific to pediatric subgroups and leveraging adaptive Bayesian methodologies for better extrapolation from adults to children.

6. AI in pediatric drug development

·         Simulations and Virtual Populations

Virtual Pediatric Populations: AI can simulate virtual pediatric populations by creating digital twins that represent different age groups, disease severities, and genetic backgrounds. These simulations can model how children across various age groups may respond to treatments, improving the ability to predict treatment

·         Predictive Models for Safety

Safety Profiling: AI can analyze large-scale data to identify potential safety issues in pediatric subgroups, including rare adverse events. By leveraging natural language processing (NLP) on clinical notes and safety reports, AI can help identify signals that indicate age-specific safety concerns.

---------------------------------------------------------------------------------------------------