PWIV-002 - RECENT TRENDS IN PEDIATRIC DOSING STRATEGY FROM 2000-2019: A NATURAL LANGUAGE PROCESSING/DRUG LABEL ANALYSIS.

K. Cheung¹, H. Ngo², C. Kim², M. Wu², R. Sane²; ¹Genentech, South San Francisco, USA, ²Genentech, South San Francisco, CA, USA.

Background: Pediatric dosing strategy is often different from adult dosing paradigms. The FDA “Pediatric Labeling Changes” database contains valuable information on pediatric study characteristics (e.g. effectiveness, safety, dosing, etc.) described in the product labelings for drug products studied in children. To learn the dosing paradigms from past examples, we applied Natural Language Processing (NLP) to automatically extract information of interest for drugs included in this dataset to study the pediatric dosing approaches of FDA-approved therapeutics.

Methods: We conducted a pediatric drug label analysis by first using an in-house NLP platform to mine the labeling change data submitted under the Best Pharmaceuticals for Children Act and Pediatric Research Equity Act from 2000 to 2019.

Results: Total of 208 discrete drug products were analyzed and summarized by types of molecule, approved or studied adult/pediatric doses, and indications. Our key findings include: 1) antibody drugs are more likely to employ body-size adjusted doses for pediatrics; 2) adolescents receive the same dose as adults in majority of the drugs; 3) despite efforts to study in children, 16% of the molecules were unable to gain approval for pediatric use.

Conclusion: By using an NLP-assisted approach to mine the labeling change data from the last 20 years, we have analyzed and characterized the recent trends in pediatric dosing strategies. The finding that adolescents receive the same dose as adults for most drugs may support dosing justification when including them in adult trials. Totality of our findings can be used in the future to propose a general guideline for pediatric dosing considerations.