Skip to main content
2024 BMES Annual Meeting Main banner
Icon Legend
This session is not in your schedule.
This session is in your schedule. Click again to remove it.
Back

Bioinformatics, Computational and Systems Biology

Bioinformatics, Computational and Systems Biology - Poster Session A

Poster A3 - A quantitative systems pharmacology model for predicting responses to PD-1/PD-L1 checkpoint blockade in classical Hodgkin lymphoma

Thursday, October 24, 2024
10:00 AM – 11:00 AM EST
Location: Exhibit Hall E, F & G

Introduction:  Classical Hodgkin lymphoma (cHL) is a malignant lymphoproliferative disorder characterized by the presence of Hodgkin and Reed-Sternberg (HRS) cells in an inflammatory background. cHL constitutes approximately 15% of all lymphomas, with an annual incidence rate of around 2.6 cases per 100,000 individuals in the United States. The standard of care for cHL involves multi-agent chemotherapy regimens, which are frequently combined with radiotherapy. However, unmet clinical needs of relapsed/refractory disease and treatment-related morbidity remain. Recently, immune checkpoint therapies targeting programmed cell death protein 1 (PD-1) and programmed death-ligand 1 (PD-L1) have emerged as promising options for relapsed/refractory cHL. These therapies counteract the immune evasion mechanisms of HRS cells by inhibiting PD-1 and PD-L1, restoring the immune system's capacity to target malignant cells and reducing treatment-associated toxicity. Nonetheless, prognostication for cHL remains challenging due to its heterogeneous nature and difficulties in identifying predictive biomarkers to determine which patients will respond favorably to checkpoint inhibition. Traditional staging systems often fail to account for factors such as the tumor microenvironment (TME), genetic and epigenetic alterations, and host immune response. These highlight the necessity for more precise and predictive models. Here we offer a quantitative systems pharmacology (QSP) model for cHL that integrates immune-cancer cell interactions, multiple compartments, and patient-derived data. The proposed model aims to identify novel biomarkers, predict the therapeutic efficacy of anti-PD-1/PD-L1 in cHL patients, and address the disparity in relapsed/refractory disease. Our model may yield improved and personalized treatment regimens for cHL patients.

Materials and

Methods:  In this study, a QSP model of cHL was developed using the SimBiology toolbox in MATLAB with the following five compartments: central, peripheral, spleen, non-tumorous lymph nodes (NTLN), and tumor lymph node (TLN) (Figure 1). To accurately and comprehensively represent the pathophysiology of cHL and the TME, the model includes important cell and non-cell species, including but not limited to HRS cells, immune cells, cytokines, tumor-associated antigens, tumor-specific antigens, damage associated molecular patterns (DAMPs), and therapeutic antibodies, featuring their production or administration, distribution, elimination, and molecular interactions (Figures 1-3). Innovations of the model include the incorporation of key species relevant to cHL that previous models failed to account for, such as Epstein-Barr virus (EBV) infection status (Figures 1, 2), the quantification of tumor growth for non-solid tumors such as cHL, the capability to represent co-administered therapies, and the addition of a spleen compartment for T-cell trafficking, improving upon previously modeled anti-tumor immune responses. Parameters of the model were selected through careful and extensive literature review, with sensitivity tests performed to confirm the robustness of the model. Subsequently, a virtual patient (VP) population was generated by defining numerical ranges of patient parameters and varying their values within those defined ranges by the means of statistical distributions to create a multidimensional array of patient parameters, representing a set number of VPs. The QSP model was then run with the VP to simulate patient-specific responses to immune checkpoint therapies for cHL.

Results, Conclusions, and Discussions: The pharmacokinetics of anti-PD-1 and anti-PD-L1 antibodies include their administration into the central compartment through intravenous injection, trafficking between compartments by both diffusion and convection within the lymphatic fluid, and elimination from the central compartment with first-order kinetics. The effect of an 150 mg intrevenous bolus of a therapeutic antibody was simulated over a period of 100 days (Figure 4). In the central compartment, the antibody mass drops sharply following administration (Figure 4) due to its rapid distribution into the other compartments. The antibody exhibits an elimination half-life of approximately 20 days (Figure 4), consistent with published clinical data. Due to the significant role of lymphatic flux in antibody distribution, more antibodies accumulate in TLN and NTLN compartments compared to the peripheral and the spleen compartments (Figure 4).

As we continue to implement the model on a VP cohort, we expect to obtain a comparison of cHL prognoses with FDA-approved anti-PD-1 antibodies (e.g., Nivolumab), anti-PD-L1 antibodies (e.g., Atezolizumab), both antibodies simultaneously, or no intervention, which will inform the efficacies of immune checkpoint therapies for cHL. Furthermore, we will assess the differential effects of patient conditions, including but not limited to cHL subtypes and EBV infection status, on the effectiveness of the therapies. We also aim to identify potential patterns in physiological parameters (e.g., number of T-cell clones, PD-1/PD-L1 expression levels, and metabolic tumor volume) that correlate to anti-tumor immune responses and complete remission. In the future, this model can be easily extended to include other therapies, such as anti-CTLA-4 and Brentuximab vedotin. By integrating the mechanisms of action and interactions of additional checkpoint inhibitors, our model can help in understanding combination therapies' synergistic effects and optimizing therapeutic strategies. Ultimately, the model will predict the outcome of different therapies for individual cHL patients, advising the choice of therapies in a personalized fashion. Additionally, our model can act as a framework for modeling the pathology and treatment of other lymphomas and hematologic malignancies. This adaptability makes our model a powerful tool for advancing personalized medicine, improving outcomes across a range of cancer subtypes, and assessing new drug candidates.

Acknowledgements (Optional):