Intricate PK and PD for the novel immunocytokine CEA-IL2v and their pre-clinical to clinical translation
Hans Peter Grimm (1), Flavio Crameri (1), Heather Hinton (2), Dietrich Türck (1), Hanna Silber (1), Benjamin Ribba (1)
Roche Pharmaceutical Research and Early Development, (1) Roche Innovation Center Basel and (2) Roche Innovation Center Zurich
Objectives: To support entry into human and Phase I design of the tumor targeted immunocytokine CEA-IL2v by leveraging non-clinical multi-scale data using mixed effects PK and PKPD modeling.
Methods: PK and PD data from single and multiple dose studies in a total of 40 non-human primates with roughly 12 PK and PD observations per subject were pooled. PK was modeled using several evolving versions of target-mediated disposition in which the IL-2 receptor is the major driver of the clearance of CEA-IL2v. A complicating factor is that the IL-2 receptor is up-regulated by the pharmacological action of CEA-IL2v itself. In a second step, a number of PD measures (including activation markers and cell numbers of several lymphocyte sub-sets) were analyzed to characterize the potency of CEA-IL2v on each of these. All PK and PD analyses were performed using MONOLIX. Model based projection of human exposure was performed mainly based on allometric principles and were incorporated in the design of the First In Human (FIH) study. The preclinical modeling work was applied as a starting point for the analysis of clinical data.
Results: The PK of CEA-IL2v was found to be strongly governed by the IL-2 receptor which in its turn is induced by the pharmacological action of CEA-IL2v itself. This feed-back pattern explains the intricate PK of CEA-IL2v. PD is observed at multiple time scales ranging from a minutes (phosphorylation of STAT5) to hours (up-regulation of CD25 and CD69 on lymphocytes) to days (expansion of cell populations). The PD behavior was captured by turnover models, however complicated by the trafficking of cells between tissues and blood. The relative potency on the various lymphocyte populations was found to be in line with the desired profile for cancer immunotherapy. The projection of human PK and PD, despite some over-prediction of exposures at low doses, allowed a safe and efficient FIH study design.
Conclusions: We have shown how thorough analysis of non-clinical PK and PD data supported the pre-clinical to clinical translation by: 1. providing explanations to complex PK behavior, 2. confirming the relative potency on lymphocyte subpopulations, 3. guiding dose selection and scheduling of the assessments in the FIH study, and finally 4. in prototyping PK and PKPD models that are now employed with human data.