Abstract Disclosure: B. Steffens: None. G. Koch: None. F. Claude: None. F. Bachmann: None. J. Schropp: None. M. Janner: None. D. l'Allemand: None. D. Konrad: None. M. Pfister: None. G. Szinnai: None. Graves’ disease (GD) with onset in childhood or adolescence is a pediatric rare disease (ORPHA:525731) with a ten times lower incidence than in adults. Leading clinical signs of hyperthyroidism are sinus tachycardia, weight loss, tremor, and goiter. First-line treatment are anti-thyroid drugs in order to normalize thyroid function. However, dose finding in pediatric GD is complex due to a broad spectrum of disease severity at diagnosis, and highly variable disease activity during follow-up, especially during puberty. As thyroid hormones have a strong positive chronotropic effect, heart rate (HR) turns out to be a useful clinical marker to monitor thyroid activity under treatment. Our overall aim was to provide a practical pharmacometrics-based (PMX-based) computer model characterizing both, individual FT4 dynamics and the relation between FT4 and tachycardia in children with various disease severity of GD during the first 120 days of treatment. Development of the PMX computer model was based on the non-linear mixed effects approach (i) linking FT4 kinetics with HR dynamics, (ii) accounting for inter-individual variability, and (iii) incorporating individual patient characteristics. Retrospectively collected clinical (resting heart rate during consultation) and laboratory data (FT4) from 41 children and adolescents with GD at four pediatric hospitals in Switzerland (75% female, median age 11.2 [IQR 8.5, 13.5] years) with 187 FT4 measurements and 132 HR measurements, and 124 paired measurements, were available and used for model development. Pediatric patients showed median FT4 of 59.6 [IQR 44.5, 73.0] pmol/l and median HR of 112 [IQR 100, 128] bpm at diagnosis. GD severity groups were defined based on FT4 measurement at diagnosis, resulting in equal numbers of mild (13), moderate (14) and severe (14) GD. We observed a significant difference in HR at diagnosis (p < 0.01) between the three GD severity groups based on FT4 at diagnosis. The final PMX computer model accounted for inter-individual variability and clinically relevant covariate effects such as age, gender, and GD severity, and was able to accurately predict FT4 and HR dynamics for each individual patient during the first 120 days of treatment. A PMX-based computer model that leverages individual HR dynamics can be applied to facilitate personalized pharmacotherapy in pediatric GD and mitigate the risk for under- or overdosing of anti-thyroid drugs in these patients. Prospective randomized validation trials are warranted to further validate and fine-tune such computer-supported personalized dosing in children with Graves’ disease. Presentation: Friday, June 16, 2023
Read full abstract