Introduction/ObjectivesChimeric Antigen Receptor T-cell (CAR-T) therapies are genetically modified cell-based treatments for hematologic malignancies, leveraging engineered T-cells to target cancer antigens. Six CAR-T products have received EU approval, primarily as third- or fourth-line therapies in relapsed or refractory cases. Given the limited patient populations in these indications, model-informed drug development (MIDD) offers potential advantages in addressing data gaps and optimizing dose-response relationships. However, the extent to which modeling and simulation (M&S) methodologies have been integrated into CAR-T regulatory submissions remains unclear. The present study aims to compare the model possibilities available in the literature with those employed in the regulatory assessment of the benefit-risk balance.MethodsTwo separate systematic reviews have been conducted on the role of modelling in the development of CAR-T cells. The first study was situated in the extant literature, focusing on computational models related to autologous for approved CAR-T cells. The second study reviewed the place of modelling and simulation as supporting evidence during regulatory submissions at the EMA. Regulatory questions addressed by M&S were classified into efficacy, safety, pharmacokinetics (PK), and modality of use. M&S methods were identified through targeted screening, and their credibility was evaluated using the structured framework from the ICH M15 guideline.ResultsA review of regulatory documents revealed that the mostly used models were dose-exposure-response (DER) and population PK (popPK) models. These models were utilised for PK characterisation, dose-response evaluations (including efficacy and safety components), and comparison with standard treatment. A review of the literature revealed the existence of more types of models including popPK, Pharmacokinetics/Pharmacodynamics (PK/PD), Physiologically-Based-PK/PD (PB-PK/PD), Exposure-Response (E-R) and two Quantitative Systems Pharmacology (QSP) models. These models were ultimately used to inform different several components of CART cells efficacy and safety, including patient survival. The analysis identified 117 regulatory questions across key categories (modality: 26, benefit-risk: 23, efficacy: 50, safety: 18). M&S approaches were applied in only 11.0% of questions, primarily using dose-exposure-response and population PK (popPK) models for PK characterization, cytokine release syndrome (CRS) prediction, and dose-response evaluations. Model credibility assessments revealed variable regulatory acceptance. ConclusionAlthough M&S methods have significant potential in CAR-T cell development, their use in regulatory submissions remains limited. Most models serve as supportive tools rather than primary evidence for decision-making.
Hanquet et al. (Thu,) studied this question.