The start-up “Information Technology for Translational Medicine” (ITTM), a partner of the T2Evolve, developed a digital health web interface for conducting an anonymous survey. The T2Evolve survey content was prepared by the T2Evolve experts participating to the WP5 (a project working package dedicated to “Gold standard analytical methods in manufacture and monitoring of CAR T cells”), and was then shared across the whole consortium to be validated for its comprehensiveness and applicability.

The survey consisted of 36 questions (including both multiple choice and open-ended questions) distributed across six sections ( Supplementary Tables S1 - 2 ) and related to: 1) demographic information ( Supplementary Table S2 section A ), 2) apheresis product collection and quality control procedures ( Supplementary Table S2 section B ), 3) quality control assays for raw materials used in engineered T cell manufacturing ( Supplementary Table S2 section C ), 4) in-process and release quality control assays for the drug product ( Supplementary Table S2 section D ), 5) lympho-depletion regimen ( Supplementary Table S2 section E ), and 6) post-infusion immunomonitoring ( Supplementary Table S2 section F ). Although queries were suitable for both autologous and allogeneic settings, participants exclusively reported experiences with autologous products. Thus, we focused the analysis solely on autologous CAR T-cell products.

The survey was specifically targeted at European stakeholders actively involved in the field of CAR T-cell immunotherapy. The primary audience included biotech, CROs, CMOs, as well as academic institutions, including public and private hospitals, leukapheresis centers, GMP facilities, development labs, and immunomonitoring units. The survey was launched in early February 2022 and concluded at the end of June 2022. It was distributed via a web survey link and quick-response (QR) code provided through the T2Evolve website, and further disseminated through international scientific conferences and email campaigns. Responses were monitored biweekly throughout the five-month period.

The anonymous replies were analyzed using in-house downstream analysis pipelines based on Jupyter Notebook and Python programming language (version: python#3). All anonymous responses are available in Comma-Separated Values (CSV) and JavaScript Object Notation (JSON) text format, with the latter serving as the input format for the downstream analysis pipeline. Results were expressed as an absolute number of respondents or as a percentage of respondents among all participants who answered the given question.

To ensure the validity of the survey results, stringent acceptability criteria were implemented to minimize the risk of representational bias among respondents. Specifically, a maximum acceptability threshold of 5% was enforced for responses received from the same organization within a single country. However, this restriction was not applied when the same organization participated across different countries. No minimum number of respondents was required to represent a country. Each respondent was uniquely identified by three key elements: country, city, and organization name. For the evaluation, only surveys that were at least 50% completed were considered valid, and responses were assessed for consistency with the questions posed.

In 2021, more than 3000 patients had been treated with CAR T-cell therapy in Europe. By comparison, more than twice the number of patients (6,343) were treated in the United States (US). Additionally, in the same year, 16% of patients in Europe and 14% in the US received CAR T-cell therapy in the context of a clinical trial. Regarding geographic distribution, the European Society for Blood and Marrow Transplantation (EBMT) Registry, reported that the access to CAR T-cell therapy is higher in Western Europe than in Eastern Europe in 2022 ( Figure 1A , EBMT website). This scenario appears to align with the distribution of respondents in the current survey, with the exception of UK and Spain, for which only one and three survey participants, respectively, were recorded. Moreover, data from the ClinicalTrials.gov platform indicate that, at the time of the survey, there were 36 interventional clinical trials in Europe related to CAR T therapy, with 86% focusing on hematologic indications (including ALL, lymphoma, multiple myeloma, T-ALL, and AML) and 14% targeting solid tumor indications (such as renal carcinoma, neuroblastoma, sarcoma, and undefined metastatic advanced solid cancers) ( Figure 1B ).

The T2EVOLVE Consortium launched a European anonymized online survey consisting of 36 questions, and covering all the most relevant topics related to CAR T-cell therapy.

A total of 58 participants took part in the survey, including 53 from Europe, 4 from the United States, and 1 from Israel. Since the dissemination plan primarily targeted European countries, the response rates from the US and Israel were deemed too low to accurately represent these countries’ involvement in the field. As a result, respondents from the US and Israel were excluded from the analysis.

The 53 European participants who responded to the T2EVOLVE survey represented 13 European countries. The majority of responses came from Italy (30%, n=16), France (17%, n=9), and Germany (17%, n=9) ( Figure 2 , Supplementary Table S3 ). Among the Italian and French respondents, more than half was concentrated in Rome (9 out of 16) and Paris (6 out of 9), respectively, while respondents from other countries were more evenly distributed ( Supplementary Table S3 ).

A significant majority of respondents (77%, n=41) were affiliated with either public or private hospitals, while 17% (n=9) represented pharmaceutical companies ( Table 1 ). Additionally, 2 respondents (4%) were from CROs or CMOs, and one respondent (2%) was from an applied research organization. In terms of professional role, most respondents were clinicians (42%, n=22), followed by preclinical researchers (30%, n=16) and GMP manufacturing operators (26%, n=14) ( Table 1 ).

The majority of responding centers declared their involvement in the development or administration of CAR T-cells for hematologic malignancies (89%, n=47) ( Figure 3A ), while only 30% (n=16) reported use of CAR T-cells for solid tumors. Notably, 19% of participants (n=10) indicated that their centers were involved in both hematologic and solid tumor therapies. Additionally, 1 center reported using CAR T-cells for the treatment of autoimmune diseases. Interestingly, 45% of respondents (24 out of 53) reported to work with both commercial and investigational CAR T-cells ( Figure 3B ). Among the 45 respondents working with investigational CAR T-cells, 27% of centers (12 out of 45) were exclusively involved in academic clinical trials ( Figure 3C ). The survey questions did not enable the recording of additional details regarding the CAR T-cell product type or the specific indications for which it is used at the respondents’ sites.

Among the 53 respondents, 72% (38 out of 53) indicated that they routinely perform quality control assays for apheresis products. Of these, 76% (29 out of 38) reported conducting sterility tests, viability assays, total cell counts, and assessments of the percentage and absolute count of T-lymphocytes (CD3+ cells) ( Figure 4 ).

In response to an open-ended question regarding improvements needed in the analytical control of apheresis products to standardize practices, 35 respondents highlighted the importance of standardizing both the apheresis process and its associated quality controls. Indeed, they specifically emphasize the importance of standardizing the characterization of the apheresis via flow cytometry. In addition, respondents call for greater consensus on the protocol for apheresis collection in case of cryopreservation and the identification of predictive biomarkers associated with successful manufacturing.

In the context of CAR T-cell manufacturing process, we did not capture the scenario of centers that use only commercial setting, so it is luckily that the data provided in this section refer to the manufacturing of academic products, or of industrial investigational products in early clinical trial development. All respondents reported using viral vector-transduced CAR T-cells, specifically retroviral or lentiviral vectors. Regarding the raw materials, 75% (30 out of 40) of participants in this survey section, reports that they did not perform any analytical tests on the raw materials, relying solely on the certificate of analysis provided by the suppliers.

When it comes to the analytical methods used for the release and characterization of autologous CAR T-cell drug products, only 34% of survey participants (18 out of 53) were directly involved in the manufacturing or release of an engineered autologous T-cell product and provided details on the assays routinely performed. Among these respondents, at least 83% (15 out of 18) conducts safety and identity tests, including viability, sterility, cell count, endotoxin, mycoplasma, and CAR expression by flow cytometry, all in accordance with the European Pharmacopeia and Good Manufacturing Practice (GMP) guidelines. However, only a small portion (16.67%; 3 out of 18) carries out an in-depth characterization of the drug product, focusing on T-cell subsets and activation/exhaustion profiles ( Figure 5 ), and no one of the Respondents is performing integration site analysis.

In response to the open-ended question, regarding the major needs in terms of standardization in the release of engineered T-cell products, 21 respondents emphasize the necessity of establishing gold-standard methods for assays assessing replication-competent lentivirus (RCL) or replication-competent retrovirus (RCR), vector copy number (VCN), potency, and drug product characterization by flow cytometry panels.

Additionally, respondents highlighted the positive impact of developing and validating rapid sterility assays to shorten the release time of autologous CAR T-cell products. They also stressed the importance of identifying and implementing predictive biomarkers closely associated with patient response and potential toxicity.

Regarding lymphodepletion (LD) regimens prior to CAR T-cell infusion, 100% (29 out of 29) of participants report using fludarabine and cyclophosphamide (Flu/Cy) as part of pre-conditioning protocols. Additionally, 3 out of 29 respondents also use bendamustine (of the three centers, one declared the use of only academic products, whereas a second one of industrial CAR T cells in investigational trial), while 2 respondents (one of which declaring the use of only academic products) do not perform LD in patients who are already cytopenic at time of treatment. Lastly, two respondents—possibly from the same institution, as they indicated being from the same city and exclusively infusing industrial CAR T-cells in early development—reported using a different lymphodepletion regimen involving Alemtuzumab.

For the characterization of patient samples collected before LD, 56% of respondents (20 out of 36) reports performing immunophenotyping of peripheral blood leukocytes, 31% (11 out of 36) assesses the expression of the CAR target antigen on tumor cells, and 36% respondents (13 out of 36) does not conduct any analysis prior to LD. Notably, 1 respondent monitors cytokine levels in patient samples before LD.

After LD but before CAR T-cell infusion, 53% of respondents (20 out of 38) reports performing immunophenotyping of leukocytes and lymphocytes in peripheral blood, 24% (9 out of 38) monitored cytokine levels, and 42% (16 out of 38) does not carry out any analysis following LD.

In response to the open-ended question regarding the major needs in terms of standardization of assays in pre- and post- LD timing, 33 out of 38 respondents (86.84%) emphasize the need for consensus on biological markers associated with toxicity. They also underscore the necessity for standardized guidelines on the optimal LD regimen, including the choice of chemotherapy drugs and their respective doses.

Lastly, the survey included 13 additional questions focused on the types of immunomonitoring performed after CAR T-cell infusion by clinical centers across Europe. In particular, 97% (28 out of 29) of respondents in this survey section, reported their experience with CAR T-cell monitoring, while only 3% (1 out of 29) had experience with other engineered T cells, namely T cells engineered with a specific T Cell Receptor (TCR).

The first set of questions aimed to determine the most common methods used for monitoring the persistence of CAR T-cells and the assays used to characterize these cells. The majority of respondents (27 out of 29; 94%) reported using flow cytometry to evaluate the persistence of infused CAR T-cells, regardless the use of commercial or investigational drug products in the academic or industrial setting. Additionally, 45% (13 out of 29) utilizes molecular assessments, including real-time PCR (25%) and digital droplet PCR (ddPCR) (16%). Notably, 34% (10 out of 29) conducts both molecular and flow cytometric evaluations. While 77% of participants limits their quantitative assessments to peripheral blood, 23% extends their analyses to other samples, such as bone marrow, cerebrospinal fluid, lymph nodes, and tumor tissues. Most respondents characterize engineered T cells in detail, focusing on CD4+/CD8+ subsets (80%), memory cell profiles (67%), exhaustion status (60%), activation status (50%), and cytokine production (37%). Interestingly this characterization is performed also by centers that declare the solely use of commercial products. Fewer respondents perform integration analysis (17%), TCR repertoire analysis (17%), or ex-vivo functional assays (13%).

Given the risk for severe toxicity events associated with CAR T-cell therapy, such as cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), early detection and appropriate management are crucial. To this end, 41% of respondents reported conducting additional tests to monitor for CRS and/or ICANS, while 33% retains biological samples for future analysis without performing immediate additional tests. The remaining 26% does not collect or analyze further samples.

In cases of tumor relapse, 52% of respondents undertake additional analytical measures to characterize relapsed tumors, 28% keeps samples for future analysis, and 21% neither conduct further analyses nor collect additional samples.

Regarding potential anti-CAR immunity, which can affect the persistence and efficacy of engineered T cells, only 13% (3 out of 23) of respondents performs tests for Human Anti-Mouse Antibodies (HAMA).

The frequency of the CAR T-cells monitoring significantly varied among respondents, with 43% checking infused CAR T-cells once a week, 21% twice a week, and 11% 3 to 6 times a week. The remaining 25% conducts monitoring sporadically or on clinician demand ( Figure 6A ). The length of the immunomonitoring period across respondents suggests a lack of consensus, ranging from two weeks to years after the CAR T-cell infusion.

Half of the respondents (48%, 12 out of 25) performed additional monitoring following CAR T-cell infusion to investigate CRS/ICANS or other toxicities. The frequency and duration of monitoring for these toxicities varied significantly: 33% adjusts monitoring based on the patient’s clinical status ( Figure 6B ), 8% performs daily blood sampling, 17% monitors 3 to 6 times a week, 17% twice a week, and another 17% once a week. Most respondents monitor patients for 7 to 14 days ( Figure 6C ), though 27% extends monitoring up to one year or based it on the patient’s condition.

When asked about improvements for immunomonitoring in the first 12 months post-infusion, 38 out of 53 respondents (71.7%) emphasize the need to identify and validate specific biomarkers to predict clinical outcomes, toxicity, and early relapse. They also advocate for proficiency testing programs to ensure consistent pharmacokinetic results across laboratories. Some respondents suggest implementing longitudinal sampling and deep immune phenotyping (of apheresis products and blood pre- and post-CAR T-cell infusion) using flow cytometry and single-cell RNA sequencing (scRNAseq). Additionally, they recommend establishing a general flow-cytometry antibody panel to include markers for exhaustion, activation, and differentiation on CAR T-cells, as well as evaluating changes in non-CAR cell populations to monitor bystander effects.