Let’s Extrapolate!

Global regulators set scientific principles for using adult data in pediatric medicine development 

Simone Langeveld, Clinical Strategy Scientist, Allucent and Richard Veselý, VP Regulatory Strategy 

What connotations does “extrapolation” conjure for you? Something mathematical, obscure, suspicious? Or something liberating, connecting your rational brain with freedom and fantasy?  

You may not realize, but we continuously extrapolate information from known outcomes to make conclusions about unknown situations. Extrapolation helps us to be more efficient with our time and effort. A favorite maxim says, “there are two types of people in the world: one who can extrapolate from incomplete data…”.  

Every day we face new situations and tasks. The forces in our lives – primarily work and family – drive us to make decisions and act quickly.  And when the consequences of those decisions become increasingly weighty, the need to balance expediency with positive outcomes becomes more serious and deliberative. 

Imagine you are a doctor and you have a dying patient in your clinic. There’s no authorized treatment available. And yet, promising data exists for an off-label therapeutic that could cure or extend your patient’s life.  

What if you’re the patient? It’s likely you’ll seek any experimental, alternative, and even unapproved treatments – even if the chances of a cure are infinitesimally small, or there is a risk of doing more harm than good. And if there is an existing treatment available, it’s not uncommon for it to be available only to adults.  

Finally, you’re a regulator. You retain the authority to approve a new drug that could effectively treat the patient. Your focus isn’t on the one patient; it’s on the population. It’s on the data. You’ve seen or heard about the disastrous outcomes when patients get hurt after medications are rushed through regulatory approvals. But you also understand how important your role is in pushing science forward. 

This is when and where extrapolation becomes a critical consideration. Extrapolation is actually critical when supported by science and leveraged from all accumulated experience to distinguish between what is necessary and what is not and, above all, mitigate risk.  

ICH Pediatric Extrapolation Guideline 

We live in a grey world where decisions are rarely black or white. It is important to realize that pediatric clinical research is no exception. There is neither unconditional “yes” nor an absolute “no” to extrapolation. It’s even safe to say that the extremes themselves are exceptional: there is a certain element of using adult data in every pediatric development and extrapolation is almost always “partial”.  

An illustration of this dramatic change in pediatric drug development paradigm can be found  in Draft ICH E11A Pediatric Extrapolation Guideline (released in 2022), which reflects the long term evolution of regulatory thinking over the past decade. The main pillars of the “extrapolation toolkit” are  

• consideration of pediatric extrapolation as a continuum, and  
• identification of the extrapolation as an iterative process.  

In this blog we discuss the proposed guideline so we can understand and appreciate the tools that are now available to minimize the clinical trial burden on the world’s pediatric patient population. 

The draft ICH E11A guideline contains a comprehensive description of the pediatric extrapolation concept, plan, and related considerations. With the goal of harmonizing the use of pediatric extrapolation in clinical development globally, a framework is set based on three parts, which are all connected in an iterative process.  

• First and foremost, the extrapolation concept must be developed by integrating evidence to assess disease similarity and identify knowledge gaps  
• Secondly, the extrapolation plan must be defined based on the extrapolation concept, to generate data supporting efficacy and safety of the product in the target (pediatric) population  
• Last, the extrapolation plan will be executed, generating new data which can inform and influence the extrapolation concept, completing the process 

Pediatric Extrapolation Concept 

The “extrapolation stool” stands on three legs:  

• similarity of the disease between the reference and target populations of patients,  
• pharmacology, and  
• response to treatment  

Similarity of the disease 

It is tempting to resort to rigid thinking, where the disease is either “exactly the same” or “totally different” between the reference and target populations (in most cases between adults and children). In reality, most diseases fall anywhere between these two extremes. This more nuanced understanding invites a focus on similarities that may be present, and to identify possible subgroups where the disease in the reference and target populations is sufficiently similar to each other. An extensive review of available data should be performed while considering the general pathophysiology, definition, and course of the disease. 

Pharmacology 

The next aspect is focused on the treatment: pharmacology. Similarities in absorption, distribution, metabolism, and excretion (ADME) of the studied medicine are important to consider. Any differences in ADME possibly result in differences in the pharmacokinetic (PK) parameters and consequently the exposure. Besides what the body does to the medicine, it is equally important to consider what the medicine does or is intended to do to the body, specifically the mechanism of action. Changing body size, organ maturation, and frequently used concomitant medications may all influence the pharmacology in different age groups. 

Response to treatment  

While it may seem impossible to have data on the response to treatment in the target population when the study treatment has not been tested in the pediatric population, a thorough review of available knowledge can include data on other medicines in the same class, drugs from other classes in the same population, or data obtained from other indications for the same study drug. Evaluation of the available data will contribute to establish an exposure-response relationship between the reference and target populations. As with the pharmacology, developmental and maturational changes may have an extensive effect on the clinical response. A helpful starting point is to focus on the endpoints that are used to measure the response in a clinical trial. Can the same endpoint be used in the reference and target populations? If a biomarker or other surrogate endpoint is favorable in the target population, it is helpful to realize at an early stage so the same endpoint can be incorporated into the clinical trial design for the adult (reference) population. 

Quality of data 

There are numerous types of data from various sources that can be utilized to evaluate the similarity of disease, pharmacology, and response to treatment, as summarized in a comprehensive table in the ICH E11A draft guideline. Real world data is included here, although it should be noted that adequacy and relevance of real world data should first be discussed with the relevant regulatory authorities. All types of data and sources each have their own strengths and weaknesses. Therefore, both the quantity and quality of the data should be assessed next to the similarities and differences between the reference and target populations. Ultimately, the existing evidence must be integrated into one comprehensive presentation: the pediatric extrapolation concept. Any gaps in knowledge must be described, as well as a strategy detailing how these will be addressed. 

Pediatric Extrapolation Plan 

While the extrapolation concept does summarize the existing knowledge, the pediatric extrapolation plan details the relevant studies that are to be conducted to fill the identified gaps.  

Dose selection 

Correct dosing is critical to achieve target exposures and the desired response. An exposureresponse relationship, developed from data in the reference population, can be used to justify exposure ranges to be targeted in pediatric clinical trials. This is the part where modeling, simulation, and model-informed approaches come into play. Techniques such as physiologically based pharmacokinetic (PBPK) modeling and quantitative system pharmacology (QSP) can contribute to accelerating the dose-finding in the pediatric population. If the correct starting dose can be determined based on adult data, dose-finding can and should be avoided in the pediatric population. 

Efficacy 

Apart from the dose selection, efficacy studies may be included in the extrapolation plan to establish efficacy in the target population. The choice of clinical trial design will be based on the extrapolation concept and the scientific questions or gaps in knowledge identified, emphasizing the importance of proper and extensive review of existing data in the reference and target populations before defining the extrapolation plan. In some cases this can be straightforward. For example if the standard of evidence in the reference (adult) population is a single arm trial design, this is likely acceptable in the target (pediatric) population as well. For externally controlled studies, it’s crucial to reach an agreement with the relevant regulatory authorities before initiating the clinical study. Lastly, several statistical tools are described in the E11A draft guideline that can be adopted to increase confidence in the results and robustness of conclusions. 

Safety 

Historically the collection of safety data in the target population was a requirement in most cases, as discussed in ICH guideline E11(R1). The new draft ICH guideline E11A leaves more room for extrapolation of safety data, applying the same principle of using previously generated data from a reference population to guide the plan for data collection in a target population. For reference populations one could think of adolescents enrolled in adult clinical trials, children or adults exposed to the same drug or class of drugs, and populations who have been treated with different dosing regimens or with different indications. In practice, to extrapolate safety data based on findings in a reference population, such as expected adverse events, will be more straightforward than confirming the absence of safety events in the target population. Longer-term safety data may need to be collected in the target (pediatric) population post-approval. 

Inclusion of adolescents in adult trials 

 Enrollment of adolescents into adult clinical trials can accelerate and broaden access to effective therapeutics for the pediatric population. The use of data obtained in suitable cases in the adolescent population should be considered in the context of the extrapolation concept and described in the extrapolation plan, as discussed above. Ethical and operational challenges to consider are described in the E11A draft guideline in detail. 

With this review of the ICH E11A guideline, it’s clear that considerable time and effort is needed to compile a solid extrapolation concept and plan. Modeling and simulation are important parts of this work. Any obvious differences between the reference and target populations should not discourage attempts at extrapolation in the pediatric population. By applying the draft ICH E11A guideline in the future, increasing knowledge of disease, pharmacology, and response to treatment will lead to more opportunities for extrapolation and less unnecessary clinical trials in the pediatric population. 

Extrapolation is sometimes seen as the last resort when pediatric studies seem to be impractical or impossible. This “feasibility” element is of course important, but what is more important is when sufficient data can be obtained using a scientifically valid extrapolation approach and methodology, extrapolation should get priority.  Using extrapolation can spare pediatric patients from unnecessary clinical trials even when they would be normally feasible. 

Allucent brings new therapies to light by solving the distinct challenges of small and mid-sized biotech companies. We’re a global provider of comprehensive drug development solutions, including consulting, clinical operations, biometrics, and clinical pharmacology across a variety of therapeutic areas. Our individualized partnership approach provides experience-driven insights and expertise to assist clients in successfully navigating the complexities of delivering novel treatments to patients. For more details about how the A-Team can support your drug development programs, visit us today.