Quantitative cancer-immunity cycle modeling to optimize bevacizumab and atezolizumab combination therapy for advanced renal cell carcinoma

TL;DR

The paper tackles optimizing combination immunotherapy for advanced RCC under patient heterogeneity by developing a Quantitative Cancer-Immunity Cycle (QCIC) framework. It integrates an ODE-based, multi-scale, stochastic approach with pharmacokinetics to build a virtual patient cohort calibrated to immunohistochemistry data and introduces the Tumor Response Index ($\mathrm{TRI}=\frac{V_{T,after}}{V_{T,before}}-1$) to predict efficacy. Validation against RCC clinical data demonstrates the QCIC’s ability to reproduce immune-phenotype distributions, predict treatment outcomes, and identify Tc cell density as a predictive biomarker, while revealing optimal bevacizumab–atezolizumab dosing strategies and the potential of adaptive dosing. The work provides a pathway toward mechanism-driven, personalized RCC therapy and digital twin frameworks, with implications for dosing optimization and biomarker-driven patient stratification in immunotherapy.

Abstract

The incidence of advanced renal cell carcinoma(RCC) has been rising, presenting significant challenges due to the limited efficacy and severe side effects of traditional radiotherapy and chemotherapy. While combination immunotherapies show promise, optimizing treatment strategies remains difficult due to individual heterogeneity. To address this, we developed a Quantitative Cancer-Immunity Cycle (QCIC) model that integrates ordinary differential equations with stochastic modelling to quantitatively characterize and predict tumor evolution in patients with advanced RCC. By systematically integrating quantitative systems pharmacology principles with biological mechanistic knowledge, we constructed a virtual patient cohort and calibrated the model parameters using clinical immunohistochemistry data to ensure biological validity. To enhance predictive performance, we coupled the model with pharmacokinetic equations and defined the Tumor Response Index (TRI) as a quantitative metric of efficacy. Systematic analysis of the QCIC model allowed us to determine an optimal treatment regimen for the combination of bevacizumab and atezolizumab and identify tumor biomarkers with clinical predictive value. This study provides a theoretical framework and methodological support for precision medicine in the treatment of advanced RCC.

Figures (10)

• Figure 1: QCIC model schematic. The model consists of bone marrow and thymus (compartment $\mathbf{A}$), peripheral blood (compartment $\mathbf{B}$), tumor-draining lymph nodes (compartment $\mathbf{C}$), tumor microenvironment (compartment $\mathbf{D}$), and lymphatic vessels (compartment $\mathbf{E}$).
• Figure 2: Schematic workflow of the data- and model-driven QCIC framework. This study proceeds from model establishment to clinical application. First, initial parameter sets (indicated by colored silhouettes) are generated via Beta distribution sampling and refined through calibration with clinical and experimental data to construct a biologically valid virtual patient cohort. Next, a pharmacokinetic (PK) model is integrated to simulate drug disposition. Virtual patients are then stratified into distinct response subgroups based on heterogeneity in treatment outcomes. Finally, this platform is applied to optimize therapy by comparing fixed-dose versus adaptive administration strategies and to evaluate dynamic tumor biomarkers. RECIST v1.1: Response Evaluation Criteria in Solid Tumors 1.1
• Figure 3: Parameter sensitivity analysis. Sensitivity analysis of the tabulated parameters on immune cells, including $T_h$ (A), $T_r$ (B), $T_C$ (C), and $TAM$ (D), plotted on a logarithmic axis at day 50.
• Figure 4: Comparison of the virtual patient cohort and immunohistochemistry data.(A--C) Box plots. During virtual patient cohort generation, a logarithmic transformation was applied to immune cell subpopulations. Scatter points represent data from real patients and virtual patients, respectively. (D--F) Probability density curve related to immunogenomic data (dark red) and data from randomly generated valid patients (pink bar chart).
• Figure 5: Pharmacokinetic index of atezolizumab and bevacizumab.(A) The range and distribution of popPK parameters. (B) Concentration-time profile of atezolizumab. The green solid line represents the mean plasma concentration, while the light green shaded area indicates the 95% confidence interval. Orange markers represent clinically measured values. (C) Concentration-time profile of bevacizumab. The blue solid line represents the mean plasma concentration, while the light blue shaded area indicates the 95% confidence interval. Purple markers represent clinically measured values.