Background: Growth hormone deficiency (GHD) is a result of low or absent secretion of growth hormone (GH) from the pituitary gland. Somatrogon, a modified hGH, requiring weekly administration, is approved for use in pediatric subjects with GHD in many regions including the the EU and Japan. Somatrogon dose may be modulated based on insulin-like growth factor 1 (IGF-1) concentration. We apply a combined population PK and PKPD model to predict individual IGF-1 levels at 96 h post-dose to support clinical management. Methods: Population PK model for somatrogon had been developed using clinical study data from Phase 2 and Phase 3 studies in adult and pediatric subjects. An indirect-response PKPD model was also developed previously to link drug concentrations with IGF-1 using data from the phase 2 study and long-term extension in pediatric subjects. The combined population PK and PKPD model was translated to R (mrgsolve) and Empirical Bayesian Estimation (EBE) was performed to predict average IGF-1 level at 96 h post-dose for an individual based on demographic inputs (age, height, weight), dose, time on therapy and a single observation of IGF-1 anytime in the 1-week interval post-dose (Kang D., et. al.). Results: EBE provided the value of IGF-1 at 96 h post dose, based on a single observed value of IGF-1 or IGF1 Standard Deviation Score (IGF1SDS) for an individual whose demographic information, dose, time on therapy and a single observation of IGF-1 (or IGF1SDS) is available. An easy-to-use Shiny interface is also provided to calculate IGF-1 or IGF1SDS based on input values. Conclusion: The optimization and associated Shiny interface provide a convenient way to provide input for dose modulation based on individual IGF-1 or IGF-1 SDS level information. This modeling/optimization framework is generalizable to other drugs also.
Kang D., et. al, Standard Error of Empirical Bayes Estimate in NONMEM VI, Korean J Physiol Pharmacol, 16 (97), April 2021, https://dx.doi.org/10.4196/kjpp.2012.16.2.97
