2018 - Montreux - Switzerland

PAGE 2018: Drug/Disease modelling - Oncology
Oleg Demin Jr

Quantitative Systems Pharmacology Modeling of CAR-T Therapy and T Cell Engaging Bispecific Antibodies: B-cell Acute Lymphoblastic Leukemia Case Study

Oleg Demin Jr (1), Antonina Nikitich (1)

(1) InSysBio, Russia

Objectives: T cell engaging bispecific antibodies (T-BsAb) and chimeric antigen receptor T cells (CAR-T) are very promising types of immunotherapy for cancer. It is established that these types of treatment work well in leukemia. Indeed, there are several therapies that have already been approved to treat patients with B-cell acute lymphoblastic leukemia (B-ALL). These therapies are CD3/CD19 bispecific T cell engager (blinatumomab) and anti-CD19 CAR-T (tisagenlecleucel). Moreover, there are a lot of other T-BsAb and CAR-T therapies for B-ALL targeting CD19, CD20 or CD22 in pipelines of pharmaceutical companies. The aims of this work are: (1) to develop quantitative systems pharmacology (QSP) model describing T-BsAb and CAR-T therapies in B-ALL; (2) to evaluate factors affecting patients’ response and non-response to the treatment; (3) to prioritize the potential targets (CD19, CD20, CD22) for B-ALL treatment.   

Methods: QSP model of B-ALL treatment with T-BsAb and CAR-T was developed. The model is based on ordinary differential equations (ODE). It includes description of hematopoiesis of normal B cells (including naïve B cells, CD20 negative and CD20 positive precursors of B cells), leukemic blasts, CD4 and CD8 T cells, different cytokines (IL-2, IL-6, TGF beta, TNF alpha and other) in physiological compartments (bone marrow, blood/plasma). Model describes pharmacokinetics and pharmacodynamics of blinatumomab (CD3/CD19 T-BsAb) [1] and CAR-T targeting CD19 receptor (19-28z CAR-T, 2nd generation of CAR-T including CD28 costimulation domain, developed by Memorial Sloan-Kettering Cancer Center) [2]. Model part describing normal hematopoiesis and B-ALL progression was calibrated on the basis of in vivo data on cells and cytokines level in blood and bone marrow of healthy subjects and relapsed/refractory B-ALL patients. Parameters describing blinatumomab and CAR-T were initially fitted against in vitro data on specific lysis of B-ALL cell lines, cytokines production and activation of T cells during culturing of T cells with target cells in presence of blinatumomab or culturing of CAR-T cells with target cells. Then, part of clinical data was used to re-calibrate some parameters. Other part of clinical data was used for model validation.

Results: The model is able to describe clinical data on dynamics of leukemic blasts, different subsets of T cells and cytokines during treatment of B-ALL patients with blinatumomab and CD19 CAR-T. Model successfully reproduces the clinically observed data on decrease of cytokine release if treatment starts with lower dose of blinatumomab followed by dose increase to the optimal values. It was shown that expression level of target receptor on cancer cells is very important factor affecting treatment success. In general, CD19 is better target than CD20 and CD22 due to the higher expression on leukemic cells. But due to the high inter-patient variability of CD19, CD20 and CD22 expression on leukemic blasts, different receptors could be effective as a target for treatment of particular patients. Expression of CD3 (in case of T-BsAb) and chimeric antigen receptors (in case of CAR-T) are another important factors affecting response/non-response. Model shows that proportion of CD19+ and CD19- leukemic blasts could be used as a predictive biomarker to predict probability of relapse in patients treated with therapies targeting CD19 and to stratify patients.

Conclusions: Developed model accurately captures available clinical data. Model shows the importance of CD19, CD20 and CD22 expression as a potential predictive biomarkers and key biomarkers to choose the right target for the therapy. Model could be used as a tool for optimization of B-ALL patients’ treatment with T-BsAb and CAR-T.



References:
[1] Klinger et al. Blood. 2012 Jun 28;119(26):6226-33
[2] Davila et al. Sci Transl Med. 2014 Feb 19;6(224):224


Reference: PAGE 27 (2018) Abstr 8522 [www.page-meeting.org/?abstract=8522]
Poster: Drug/Disease modelling - Oncology
Click to open PDF poster/presentation (click to open)
Top