Director Scientific Affairs Celerion Tempe, Arizona, United States
Statement of Purpose: A mass balance study uses a radiolabeled drug to obtain quantitative and comprehensive information on the absorption, distribution, metabolism and excretion (ADME) of an investigational product. Nonclinical animal mass balance studies help inform the clinical ADME study design. Specifically, the absorbed radioactivity dose, estimated duration of confinement and anticipated main excretion route are needed to support human ADME studies. However, there are several metabolic and technical inter-species factors that may impact the translation of nonclinical to clinical ADME results. Therefore, the aim of this study was to examine the performance of nonclinical models to predict the predominant excretion route of a drug in humans.
Description of Methods & Materials: Radioactivity recovery excretion data were extracted from the FDA clinical pharmacology reviews from Drugs@FDA. Results were compiled from clinical and nonclinical pharmacology dossiers for FDA approved small molecules from 2020-2022. Concordance (%, [95% Confidence Interval]) and correlation (Person R, p value) analysis were calculated to assess nonclinical-clinical performance.
Data & Results: Over the last 3 years, the FDA approved 66 small molecules in which 54 drugs had publicly available mass balance study results reported in their clinical pharmacology review dossier. Drugs indicated for oncology (41%) and classified as kinase inhibitors (33%) represented the largest therapeutic area and class, respectively. The majority of nonclinical mass balance studies used a rat model with intact bile duct (74%) and found that feces or the hepatobiliary pathway was the main route of excretion (80%), while 9% of drugs were predominately excreted via urine and 6% engaged both pathways equally. Interestingly, the human ADME results followed a similar trend with 76%, 15% and 9% of the radioactive label primarily recovered in feces, urine or both, respectively. Moreover, 83% of the drugs analyzed demonstrated concordance between animal and human results. The nonclinical excretion data closely matched human values, and demonstrates excellent positive predictive value (84.8% [71.8%, 92.4%]; 80.0% [37.6%, 99.0%]) and a strong association (R=0.7321, p < 0.0001; R=0.6234, p < 0.0001) for feces and urine radioactivity recovery, respectively. Interestingly, drugs excreted mainly in urine or both pathways were most likely to display discordant results (Figure 1). For instance, the 3 drugs that were predominately excreted in human urine only had minimal radioactivity recovered in rat urine (75.0±8.6% vs 20.1±3.0%, respectively) resulting in substantial inter-species differences. Closer examination into these conflicting cases revealed that animal model sensitivity, species differences in CYP activity, and differing glucuronide metabolite excretion routes contributed to this discrepancy.
Interpretation, Conclusion or Significance: Despite technical and innate differences between nonclinical and human ADME studies, the animal data strongly predicts and correlates with human findings. Differing urine results arose for drugs metabolized by CYP2C19 or glucuronidation. Others have shown that CYP2C19 is >10-fold less active in rats than humans, which may shift the metabolite profile. In addition, rats also tend to readily excrete glucuronide metabolites in bile versus urine. In these cases, the sponsor should consider adding metabolic profiling to the human mass balance study to fully understand the metabolite profile in humans, as it may be quite different from rat models.
