Optimal design in pediatric clinical trials: theory and application to the macitentan case study