Drug developers can’t take an adult clinical trial protocol and simply retool it for pediatrics. Yet, in my work at a pediatric clinical research organization (CRO), I frequently see sponsors attempt to do just that. Successfully implementing a pediatric drug development plan poses challenges regarding trial planning, design, and conduct. In this blog post, I’ll discuss these challenges and propose some tips for addressing them.
Considerations for global pediatric studies
To operationalize pediatric clinical study plans, you need to recruit patients and select the right team to conduct the studies. This former task is complicated by the fact that there are simply fewer children able to participate in studies than adults. For example, say you’re working on a drug for premature babies. In the US, 4.2 million children are born each year. But, only 10% (420,000 babies) are born pre-term. Competing studies make the situation even worse as several dozen studies are generally being conducted on that population at any given time in the US.
So, you think to yourself, “Maybe I’ll conduct my pediatric program in Europe!” But, Europeans have fewer kids than Americans. And medical treatment in Europe differs from the US. For example, neonatal intensive care units (NICUs) in Europe average about 14 beds. In the United States, the average NICU has about 60 beds. For this reason, many sponsors prefer to conduct their pediatric trials in the US with some supplemental neonatal work conducted in the European networks.
Investigator and site selection
Likewise, identifying suitable investigators and trial sites is also a challenge. While the pediatric investigator pool is growing in the United States and around the globe, it’s still difficult to find investigators for certain diseases. For example, we just finished a feasibility assessment last year for a clinical trial for a new pain medication. While we surveyed over 800 potential chronic pain centers in North America, only 21 centers were capable of performing this study. Also, when you identify appropriate trial sites, there’s still the issue that they might be hosting competing programs for a similar disease state or patient population. When you’re developing your pediatric trial designs, be cognizant of your inclusion/exclusion criteria and what you’ve negotiated with the FDA or EMA so that you don’t end up with a study design that’s not feasible.
We conduct trials at three types of pediatric research sites:
- Academic experts/“specialty clinics” are generally used for Phase 1–2 studies
- Trained research/”medical practice centers” are used for Phase 2–3 studies
- Condition-of-use/”general clinics” can be used for Phase 3B/4 studies
Sustainable site networks are an emerging concept for supporting pediatric research. In this model, a group of pharmaceutical companies will collaborate to develop a centrally managed network of 30 pediatric clinical research sites across the US, EU, and Latin America. These sites will be supported by auxiliary staff who identify and recruit patients into the network’s trials. This model will help reduce the sites’ administrative burden. It is also a cost-effective model for pharma to decrease start up and enrollment timelines.
Pediatric drug formulation
Ask any parent who has given medication to their child, and they’ll tell you that it’s not easy. There’s a pharmacological rule that if you don’t get the drug to its site of activity, it won’t work. Likewise, if you don’t get the drug in the child, it won’t work either. Clinical trial data integrity is also severely compromised when you try to administer drugs that are inappropriately formulated for that age group.
Well-formulated pediatric drugs facilitate patient adherence and thus support therapeutic efficacy. Oral formulations for children should be able to be administered without water, have a short onset time, and possess high bioavailability. Kids under the age of nine generally can’t swallow tablets or capsules. Crushing tablets or capsules is generally not an appropriate option. So if a health authority has mandated studying a drug that has been formulated as a tablet or capsule in kids between the ages of 6 to 12, you will need to reformulate it.
Infants often receive drugs via nasogastric (NG) tubes. This approach is problematic because the drug typically will adhere to the NG tube’s side wall. Thus, the drug doesn’t get to its site of action. Even if you try to back flush the tube, you still won’t get the drug in the patient at the prescribed dosage.
Liquid formulations can be made more palatable by adding an approved flavoring product like Flavorx. For unflavored medications with poor palatability, you can mix them with food or drink. But, this then requires stability testing to show that the drug isn’t being degraded. Liquids also require more rigorous stability testing because of their shorter half-lives compared to solid dosage forms.
Medication palatability is a complicated factor. “Palatability” is defined as a hedonic reward provided by foods or fluids that are agreeable to the palate in regard to the homeostatic satisfaction of nutritional, water, or energy needs. The palatability of a food or liquid—unlike its flavor or taste—varies with the state of an individual. Palatability is lower after you eat and higher when you’re hungry. It’s not a fixed property either; palatability can be learned.
Drug palatability includes its taste, smell, and mouth feel. Unpalatable drugs often are spit up by kids, which complicates drug administration. The only validated palatability scale was developed at the Hospital for Sick Kids by Freedman and colleagues in 2010. It’s a visual analog scale similar to those used to rate pain.
The 100-mm visual analog scale incorporating facial hedonic features used to evaluate taste: 0 indicates the worst score, 100 the best. Freedman SB, Cho D, Boutis K, et al. Arch Pediatr Adolesc Med 2010:164;696.
Blood sampling in pediatrics
Developing the dosing strategy for pediatric medications requires taking blood samples from patients to understand their pharmacokinetic (PK) profile. This is a big issue in pediatrics—especially for the younger age groups. The study’s protocol sampling schedule should be designed to maximize information from a minimum number of samples. Modeling and simulation approaches―like population pharmacokinetic models―can be helpful in characterizing a drug’s PK from sparse data.
Consider pediatric blood volumes when developing your sampling strategy. The total blood volume of a child can be estimated from his or her age and weight:
- A pre-term baby has about 90 mL of blood per kilogram of body weight
- For term newborns, it is about 80 mL/kg
- For babies aged 1–12 months, it’s 75 mL/kg
- For toddlers up to age 3, it’s 70 mL/kg
- For older children and teens, it’s 65 mL/kg
NIH guidelines state that 3 mL/kg can be drawn for research purposes with no more than 7 mL/kg total being drawn over any six-week time period. Investigators are also encouraged to consider further limiting blood draws in any patients with anemia or low cardiac output. Likewise, the American Academy of Pediatrics recommends a maximum of blood draw volume of 1-2 mL/kg per month or about 3% of body weight. These limitations on blood draw volumes can become an institutional review board (IRB) issue as your program moves into clinical studies.
This is an example of a blood draw schedule for a newborn baby from a pharma company that doesn’t understand the limitations of pediatrics:
- Day -1: Complete blood count (CBC) .5 mL + Chemistry screen (Chem) 0.5 mL + 4.5 mL platelet aggregation testing = 5.5 mL
- Day 0: pK 0.5 mL x 4 = 2.0 mL
- Day 2–4: pK 0.5 x 4 = 6.0 mL
- Day 7: CBC 0.5 mL + Chem 0.5 mL + pK 0.5 mL + 4.5 mL platelet aggregation = 6.0 mL
- TOTAL: 19.5 mL/ 2kg baby = 9.75 mL/kg
Obviously, collecting 9.75 mL/kg from a baby over this time period is unacceptable. This blood sampling schedule is a more appropriate for a pediatric protocol.
Table 1: Summary of blood samples
|Time Point||Volume of Sample
for Pts. < 3 kg,
or 3–5 kg, per Investigator*
|Volume of Sample
for Pts. > 5 kg
|Day 1||Postrandomization, prestudy drug PD sample||reduce to 2.7 mL||4.5 mL|
|10–30 min. postinitial-dose, PK sample||omit 2nd||0.5 mL|
|1–3 hrs. postinitial-dose, PK sample||omit 3rd||0.5 mL|
|6–12 hrs. postinitial-dose, PK sample||omit 4th||0.5 mL|
|12–24 hrs. postinitial-dose, PK sample||omit 5th||0.5 mL|
|After 7 days of
|Postinitial-dose, steady-state PD sample||reduce to 2.7 mL||4.5 mL|
|Postinitial-dose, steady-state PK sample||omit 1st||0.5 mL|
|Safety assessment sample||≤ 1.0 mL||≤ 1.0 mL|
*Stepwise approach will be used to minimize blood sampling for neonates < 3 kg body weight, or 3–5 kg per investigator’s discretion. Order of preferred blood volume reduction is listed above and detailed in protocol section 3.1.
Specimen collection challenges
Another challenge involves where and how you collect your blood samples. If you do a straight stick on a child using an IV access, you need to make sure to flush the lines thoroughly. Heel sticks are another way to get blood, but this approach can cause bruising and pain. Similarly, venous sampling can be painful because pediatric veins are small targets which collapse easily.
I’ve found several tips for making blood collection easier and less painful for patients. Many times, we use EMLA cream—a eutectic mixture of local anesthetics—to numb the site of a blood draw. One caveat is that the anesthetics have a 45 minute onset time so you have to coordinate applying the cream with the clinic. Also, use the smallest gauge needle possible (eg, 23 gauge butterfly) so that the patients don’t feel it as much.
Here are some best practices for non-blood samples. Avoid intrusive samplings such as cerebral spinal fluid, biopsies, bronchial washings, and pH probes. When feasible, use easy to collect specimens like saliva, stool, or urine. These matrices can also be used in PK modeling and simulation to help determine the optimal dose and dosing intervals.
Measuring cognitive function in children
It’s often more difficult to measure clinical outcomes in kids than adults. For example, regulatory bodies may insist that potential for long term cognitive effects is delineated, even in non-neurological drugs. In the case of a sponsor developing an antihypertensive drug for children, the FDA required cognitive function testing on the patients for at least 24 months.
Cognitive testing in children is not a simple undertaking. What cognitive function test do you do? These tests are difficult to perform because children have very short attention spans. They simply cannot sit still for a three-hour computerized cognitive function test. It has to be able to be completed within a short period of time—no more than 30 minutes. So when planning your pediatric studies, remember that you may need to perform complex measurements of long term outcomes.
Take home messages
Pediatric drug development is complex and difficult. Many important research questions remain to be answered.
Fortunately, every pediatric trial presents opportunities for learning. Rigor in study design and execution is essential, but the end goals should be reasonable. And pediatric drug development must be done in partnership with the industry, regulators, academic pediatricians, patients and their families. Finally, we must learn from our failures. As we learn, pediatric clinical trials will provide more valuable and scientifically valid results.
To learn more about addressing the regulatory and practical challenges of pediatric drug development, please watch my webinar. Let me know what you think in the comments section!