CAR-T cells that target CD19 have become standard treatments of relapsed or refractory large B cell lymphoma (rrLBCL) in 3^rd line or beyond, and have moved up to 2^nd line following the recent reports of two positive randomized trials (1, 2). The safety of CAR-T cells is of major concern, since this new class of anti-tumour therapy may have unknown side effects. The complete risk profile of CD19+ CAR T-cells has not been fully defined by the relatively small registration trials leading to approval of CAR T-cell products. Therefore, health authorities obliged Marketing Authorization Holders (MAH) to report efficacy and toxicity data on predefined numbers of patients treated with commercial CART products in the real world-setting. The EBMT registry that was established forty years ago in Europe to collect data on patients receiving autologous or allogeneic hematopoietic cell transplants, has recently extended its capacity to the capture of long-term follow-up (up to 15 years) of patients treated with CAR-T Cells through the design and implementation of a Cellular Therapy Form (CTF). Primary use of data allows for the conduct of studies such as the one reported in this manuscript, while secondary use of data allows for the conduct of post-authorization safety studies (PASS) and post-authorization efficacy studies (PAES) in partnership with MAH/Pharmas or other stakeholders. Classical side effects of CD19 CAR-T cell products, which are already well described in clinical cohorts, include cytokine release syndrome (CRS), immune cell associated neurotoxicity syndrome (ICANS) and B-cell aplasia. Currently, there is emerging evidence of additional both short-term and medium-term unwanted effects, which were in part unknown at the time of regulatory approval (3–6). There is evidence on neurological events other than ICANS, including ischemic attacks, peripheral neuropathy and Alzheimer’s dementia (7). The patho-mechanism behind this association is not known. In one study major cardiovascular events occurred in 17% of patients till one month after CAR-T cell infusion (8) and in another study frequent adverse events were observed within one year after CAR-T cell infusion (4). The most frequent pulmonary toxicity symptom observed in previous studies was hypoxia, but also pleural effusion, pulmonary embolism, allergic rhinitis and pneumomediastinum were described (4). In a recent study by Wudikarn and colleagues metabolic toxicities after CART have been detected as a frequent complication (4). Typical electrolyte abnormalities were hypokalemia, hypophosphatemia and hypocalcemia. Hyperglycemia and hypoglycemia were also frequent (4, 9, 10).

In the current study, we have used the EBMT CAR-T cell registry (11) to assess the real-world outcome of CD19 CAR-T cell products. Our study aimed at investigating (non-CRS and non-neurotoxicity) organ complications of CD19 CAR-T cell products in adult patients with LBCL reported to the EBMT.

Our present study is the first analysis of EBMT real-world data on other organ complications than CRS and ICANS after anti-CD19 CAR-T cell therapy for LBCL. Our main finding is that despite high rates of CRS in this dataset a low frequency of severe organ complications occurred, marginally contributing to a relatively low NRM. Of note, disease status before CAR-T cell therapy was mainly chemorefractory/progressive disease (75%) in the patient population reported here, which is considerably higher than in the registration trials. Despite this difference, our present results are mostly in line with follow up data of the initial CAR-T cell trials in patients with LBCL, also showing low NRM and manageable safety (9, 14). However, some of the more frequent organ toxicities found in our present study, such as renal, cardiac and hepatic complications, were not reported in the pivotal Zuma-1 (Axi-Cel) and Juliet (Tisa-Cel) trials. This is most likely due to larger patient numbers and a longer follow up period in our real-world analysis. As expected, our patients in this retrospective analysis had received more prior lines of treatment and had a lower performance status when compared to the patient populations in the two registration trials conducted to support marketing approval of Tisa-Cel and Axi-Cel. These different patient characteristics may also be another relevant aspect explaining the occurrence of organ toxicities in real-world conditions.

Recently, a real world data set of patients with LBCL undergoing CAR-T cell therapy in Germany has been published (n=356) (5). When comparing NRM in this national analysis with our current data, differences in NRM become evident. At a median follow up of 11 months, NRM was 10% in the German analysis vs. 4.2% NRM with 12 months median follow up in our current analysis. Interestingly, the reasons for death were also different with bone marrow aplasia and infections as the driving factor for NRM in the recent German analysis (5) and organ toxicities being the major factor for NRM in our current analysis. We recently found in LBCL patients receiving CD19+ CAR-T cells, that severe cytopenia had no significant impact on overall survival (HR 1.13 [95%CI 0.74-1.73] p=0.57) (15). However, patients with severe cytopenia had a poorer progression-free survival (HR 1.54 [95%CI 1.07-2.22] p=0.02) and a higher relapse incidence (HR 1.52 [95%CI 1.04-2.23] p=0.03). Again, the most likely reason for the pronounced differences in outcome between the two analyses is the difference in patient characteristics. The German analyses reported that only 13% of patients in their dataset fulfilled the inclusion criteria of the Zuma-1 trial, indicating a population at very high risk for toxicities and for NRM. Organ complications were not reported in detail in this publication, making it impossible to compare them to our present study. Also recently, a larger real world data set from France was published mainly comparing results from Axi-Cel vs. Tisa-Cel treated patients, but again organ complications were not reported in detail so far (16). In line, reporting of organ complications was not done in detail so far for the recent trials comparing CAR-T cell therapy to standard of care as second line option for large B-cell lymphoma (17, 18).

There are publications from centres in the USA with detailed information on organ toxicities. The MD Anderson Cancer Center in Houston looked specifically at cardiac complications in a retrospective analysis of 165 patients with LBCL undergoing therapy with axicel or tisacel (19). Sixteen percent developed at least one major adverse cardiac event until day +30. Among them, three myocardial infarctions. However, the occurrence of cardiac events was not associated to NRM. The Memorial Sloan-Kettering Cancer Center published even more comprehensive data on organ complications in a smaller series of patients (n=60) undergoing CD19+ CAR-T cell therapy for LBCL (4). At one year follow up, they found severe cardiovascular, pulmonary, hepatic and gastro-intestinal toxicities in 20%, 13%, 13% and 10% of patients respectively. Again, NRM was low with 1.7% at one year and was not associated to the occurrence of organ complications. Table 6 summarizes the occurrence of organ toxicities in major trials on CD19+ CAR-T cells. Although it is impossible to directly compare the previous trial results to our current data it appears that the previously reported complication rate was relatively high. One reason could be the improved clinical management in the more recent period, but we are unable to rule out underreporting of complications in the EBMT data base (see below).

A limitation of our study is that we were not able to perform advanced statistical analyses on the association of severe organ toxicities and NRM because of the low frequencies of these two events, despite the fact that this is one of the largest patient population ever studied. A possible further limitation of our current results is the retrospective data acquisition. We were therefore unable to monitor primary data to exclude a possible underreporting of organ toxicities in the EBMT data file. Finally, the follow up period was limited and is too short to adequately describe the occurrence of later toxicities – e.g. secondary malignancies.

A different and relevant way of looking at our current results is to compare them to toxicity and NRM data in allogeneic stem cell transplantation, which remains a relevant alternative treatment option in patients with relapsed/refractory LBCL. In a recent analysis, we found 12.2% NRM one year after allogeneic stem cell transplantation (22). Although the patient populations can’t be directly compared, the much lower 5.2% NRM one year after CD19+ CAR-T cell therapy in our current study suggests differences in safety in favour of CAR-T cells.

In conclusion, we found a low frequency of severe organ complications and a relatively low NRM after CAR-T cell therapy for LBCL in this international, multicentre analysis. These data demonstrate good safety in the European real-world setting.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving humans were approved by Transplant Complications Working Party of the EBMT, EBMT board. The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation was not required from the participants or the participants’ legal guardians/next of kin in accordance with the national legislation and institutional requirements.

All authors listed have made a substantial, direct, and intellectual contribution to the work and approved it for publication.