CAR-T cell expansion and persistence are critical parameters for therapeutic efficacy and toxicity (Locke et al. 2020). However, CAR-T cells are patient-specific ‘living drugs’ with an unpredictable ability to expand in vivo. Thus, close postinfusion monitoring should be a major prerequisite to better manage this therapy. Critical parameters include CAR-T cell expansion kinetics and phenotype immune reconstitution and serum biomarkers (Fig. 35.1; Kalos et al. 2011; Hu and Huang 2020). Additionally, prospective collection and storage of patient specimens should be planned for future hypothesis-driven studies at specialized research centres. To date, despite the rapid expansion of CAR-T cell therapy, no standard recommendations exist for CAR monitoring, and harmonization of efforts across multiple centres is urgently needed.

Fig. 35.1
figure 1

Schematic overview of monitoring after CAR-T cell therapy

Full size image

Molecular Monitoring of CAR-T Cells via Digital PCR (dPCR)

Most clinically used CAR-T cell products consist of autologous lymphocytes stably transduced with retro- or lentiviral vectors encoding the respective CAR construct. Integrated CAR vectors are commonly detected at the genomic level using real-time quantitative PCR (qPCR) or dPCR. Surprisingly, outside of clinical trials, CAR-specific diagnostic tools were initially missing, requiring the de novo design of lab-made specific assays to enumerate CAR-T cells in vivo (Badbaran et al. 2020; Fehse et al. 2020; Kunz et al. 2020). Despite technological differences, both qPCR and dPCR assays yield robust and accurate results, with limited requirements regarding sample quality (Table 35.1).

Table 35.1 Comparison of molecular monitoring tools
Full size table

dPCR is extremely sensitive and does not rely on standard curves or multiple repetitions. As a limitation, DNA-directed PCR monitoring provides no information on the expression of the CAR construct. However, the expansion of CAR-T cells strongly depends on the interaction of the CAR with its cognate antigen. In accordance, our data have shown excellent correlation of dPCR with flow cytometry (Badbaran et al. 2020) as well as clinical (Ayuk et al. 2021) results. Because flow cytometry-based assays facilitate phenotypic characterization of CAR-T cells, the two methods complement each other well.

Flow Cytometry Monitoring of CAR-T Cells

Identification of CAR-T cells by flow cytometry (CMF) can be performed by using monoclonal antibodies (mAbs) directly recognizing the CAR (idiotype, linker region) or a specific tag included in the CAR construct. Alternatively, indirect detection can be achieved using antigen-Fc chimeric proteins containing the CAR target antigen fused to a human IgG Fc fragment. A secondary staining step is required for the detection of CAR-expressing cells with an anti-Fc or anti-biotin (if the antigen-Fc is biotinylated) mAb labelled with a fluorochrome (Hu and Huang 2020). In practice, outside of clinical trials, patients receiving commercial CD19 CAR-T cells are monitored with biotinylated CD19-Fc proteins in a two-step staining protocol. The advantage of CMF is the possibility of combining CAR staining with other cell surface markers to characterize CAR-T cells in terms of T cell subtype (CD4 and CD8 expression), differentiation (naïve versus memory), and exhaustion (PD1, TIM3, Lag3). In addition, the results can be provided in real time to physicians. The limitation of the technique is the relatively low sensitivity. Below 0.5% of T cells, the reliability of CMF is weak and justifies pursuing monitoring via PCR. Two important pieces of information can be obtained with sequential CMF analysis of CAR-T cells in the peripheral blood after cell infusion: the expansion peak (Cmax, maximum CAR-T cell rate in percentage or absolute value) and the area under the curve of circulating CAR-T cells between D0 and D28 (AUC0-28). These two parameters have been associated with the response and the risk of complications after treatment in B-lymphoid malignancies (Park et al. 2018; Fraietta et al. 2018; Locke et al. 2020; Ayuk et al. 2021). To determine these parameters, the recommended frequency of CAR-T cell monitoring is two or three times a week for the first 2 weeks after CAR-T cell administration, on days 21 and 28, once a month until 3 months and then every 3 months until 1 year (Rubio et al. 2021).

Monitoring of Additional Immune Parameters (Non-CAR-T, B, and NK Cells and Cytokines)

Patients receiving anti-CD19 CAR-T cells might develop prolonged T CD4 lymphopenia as well as B-cell aplasia with severe hypogammaglobulinaemia, making them particularly susceptible to bacterial and viral infections even after haematopoietic recovery (Logue et al. 2021). Therefore, routine immune surveillance of non-CAR-T CD4 and CD8 T cells, B cells, and NK cells and the levels of serum immunoglobulins is recommended during the first year of follow-up.

Many cytokines are produced in large quantities after CAR-T cell administration as a result of activation of T lymphocytes (IL-6, IFN-γ, sIL2-Rα, sIL-6R, GM-CSF, IL-2, and TNF-α), activation and attraction of mono-macrophages (IFNα, IL-1β, IL-6, IL1Rα, IL10, IL-12, IL-13, IL-15, sIL6-R, TNF-α, CXCL10, CCL2, and IL-8) and in response to tissue damage (IL-6, IL-8, G-CSF, and GM-CSF) (Brudno and Kochenderfer 2019). Confounding factors, such as sepsis, degree of CAR-T cell expansion and tumour burden, also impact cytokine levels. Some cytokine signatures have been described to predict the occurrence of cytokine release syndrome (CRS) (Teachey et al. 2016), immune effector cell-associated neurotoxicity syndrome (ICANS) (Santomasso et al. 2018) or the expansion capacity of CAR-T cells in vivo (Kochenderfer et al. 2017). One major limitation in clinical practice is the absence of a validated fast cytokine quantification test predicting severe complications. Therefore, further studies are required in homogeneous groups of patients to determine whether cytokines can predict the occurrence of complications or treatment efficacy. Participation in prospective studies or collection of serum at each time point of CAR-T cell analysis is recommended.

Key Points

Immune monitoring after CAR-T cell therapy should be carefully performed:

  • Medical CAR-T cell products are complex ‘living drugs’ with unpredictable in vivo performance. Thus, the establishment of accompanying diagnostic and research monitoring programmes is a priority for rational development of this approach.

  • Molecular monitoring, especially dPCR, is an excellent, robust, and sensitive tool for real-time/on-site persistence tracking.

  • Flow cytometry is an easy and rapid tool to monitor early CAR-T cell expansion and characterize CAR-T cell phenotype, both of which have been correlated with the response.

  • Routine monitoring of T, B, and NK cell populations and immunoglobulin levels is recommended to evaluate infection risk.

  • Serum collection is recommended to further explore and identify cytokine signatures that enable prediction of complications or response.

  • Efforts to harmonize patient monitoring across multiple centres following CAR-T cell infusion would be desirable (i.e., reference labs, shared databases, and collaborations with dedicated centres for ‘next generation’ research). Successful implementation of these joint efforts will greatly advance our understanding of the biology involved in transferring CAR-T cells and, most importantly, serve our patients.