31P - Schedule optimization of a novel tumor-targeted IL-2 variant immunocytokine by integration of human in vivo immune cell kinetics and functional ima...

Date 05 November 2016
Event ESMO Symposium on Immuno-Oncology 2016
Session Lunch and general poster viewing
Presenter Benjamin Ribba
Citation Annals of Oncology (2016) 27 (suppl_8): viii4-viii17. 10.1093/annonc/mdw527
Authors B. Ribba1, C. Boetsch2, T. Nayak3, Z. Xu4, H. Grimm1, H. Silber-Baumann1, J. Saro5, S. Evers5, V. Teichgräber5
  • 1Clinical Pharmacology, Roche Innovation Center Basel, 4007 - Basel/CH
  • 2Clinical Pharmacology, Roche Innovation Center Basel, Basel/CH
  • 3Translational Medicine, Roche Innovation Center Basel, 4007 - Basel/CH
  • 4Clinical Pharmacology, Roche Innovation Center New York, New York/US
  • 5Translational Medicine, Roche Innovation Center Zurich, 8952 - Schlieren/CH



FAP-IL2v is a novel tumor-targeted IL2v designed to overcome limitations of wildtype IL-2. The antigen binding portion targets fibroblast activation protein-alpha (FAP). The IL-2v portion binds to IL-2Rbg but not IL-2Ra, thus preferentially activating CD8+ and NK immune effector cells over Tregs. Phase I data from a sister compound CEA-IL2v showed reduced serum concentrations following multiple dosing. One explanation is that this is due to target-mediated drug disposition (TMDD), caused by IL2v-driven expansion of IL-2R+ peripheral cells. Here, using this assumption, we integrated pharmacokinetic (PK), peripheral pharmacodynamic (PD) and functional CEA-IL2v imaging data to optimize dosing and scheduling of FAP-IL2v for maximal tumor uptake.


PK and immune cell kinetics in peripheral blood (PBMC) were assessed in 40 CEA+ cancer patients treated with CEA-IL2v on weekly (qw) or biweekly (q2w) dosing schedules. An additional 14 patients with both CEA+ and CEA- tumors were enrolled in a PET study with 89Zr-labeled CEA-IL2v. A mathematical model was developed to integrate within a unique quantitative framework PK, PD and imaging data. It was assumed that drug binding to IL-2R+ cells in PBMCs (i.e. CD4, CD8, NK and B cells) induces rapid redistribution and expansion of these cells and that the drug distributes into all tissues including tumor lesions following a passive diffusion process. The model also assumes that properties of binding to CEA will be similar to that of FAP.


In the absence of expansion of IL-2R+ cells in the blood, qw dosing theoretically should result in a doubling of tumor uptake compared to q2w dosing. Model simulations indicate that the actual expansion of IL-2R+ cells negatively affects tumor uptake. However, qw dosing is predicted to still leading to a 90% increase of tumor uptake compared to q2w dosing.


Given our assumptions, these data suggest that increasing the dose at each administration or shortening the time interval between doses can compensate for the reduction of tumor uptake due to TMDD. This modeling framework can be used to optimize dose and schedule of FAP-IL2v in order to maximize drug exposure of the molecule in the tumor microenvironment.

Clinical trial identification

Legal entity responsible for the study





B. Ribba, C. Boetsch, T. Nayak, Z-X. Xu, H-P. Grimm, H. Silber-Baumann, J. Saro, S. Evers, V. Teichgräber: Employee of Roche.