Skip to Content
LymphaticsLymphatics
  • Article
  • Open Access

29 June 2026

15 Pages

Adverse Events of CD19- and BCMA-Directed Chimeric Antigen Receptor T-Cell Therapy: An Analysis of the FDA Adverse Events Database

Department of Medicine, University of British Columbia, Vancouver, BC V6T 1Z3, Canada

Abstract

Background: Chimeric antigen receptor T-cell (CAR-T) therapies have transformed the treatment of haematologic malignancies, yet their adverse event (AE) profiles across all approved agents have not been consolidated using a pharmacovigilance methodology in a single comparative analysis. Methods: Using the FDA Adverse Events Reporting System (FAERS) and OpenVigil 2.1, disproportionality analyses for all six FDA-approved CAR-T therapies were performed, stratified by target antigen: anti-CD19 agents (axicabtagene ciloleucel, tisagenlecleucel, lisocabtagene maraleucel, brexucabtagene autoleucel) and anti-BCMA agents (idecabtagene vicleucel, ciltacabtagene autoleucel). For each of the 25 most frequently reported AEs per agent, the report event counts, reporting odds ratio (ROR) with 95% confidence interval, proportional reporting ratio (PRR), and chi-squared statistic were calculated. Results: A total of 36,567 AEs were identified across all six agents. Cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome were the most frequent and most disproportionate AEs across all six therapies, with RORs exceeding 240 in every agent. Idecabtagene vicleucel had the highest ROR for cytokine release syndrome among all agents (ROR 1934.6), while brexucabtagene autoleucel had the highest ROR for immune effector cell-associated neurotoxicity syndrome (ROR 2089.6). Ciltacabtagene autoleucel exhibited a unique neurological toxicity profile, cranial nerve paralysis (ROR 3167.6, PRR 3055.0, chi2 199,190), parkinsonism (ROR 145.5, PRR 136.6, chi2 24,840), Bell’s palsy (ROR 380.4, PRR 370.3), and facial paralysis (ROR 72.7), consistent with the late neurotoxicity syndrome previously characterized, and absent from other agents’ top 25 AE lists. Brexucabtagene autoleucel demonstrated secondary malignancy signals including squamous cell carcinoma of skin (ROR 308.2) and myelodysplastic syndrome (ROR 97.9). Tisagenlecleucel showed the highest hypogammaglobulinemia signal among all agents (ROR 614.1, PRR 536.5, chi2 144,832) alongside a broad pancytopenia profile. Hemophagocytic lymphohistiocytosis was a significant finding with ciltacabtagene autoleucel (ROR 71.0) and brexucabtagene autoleucel (ROR 57.8). Conclusions: CAR-T therapies share class-wide toxicities in cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome, but exhibit clinically important drug-specific and target-class-specific AE profiles. This consolidated FAERS-based analysis corroborates and extends prior pharmacovigilance work by providing direct cross-agent comparisons, and supports the use of agent-tailored monitoring strategies, particularly for the distinctive late neurological toxicity associated with ciltacabtagene autoleucel.

1. Introduction

The advent of chimeric antigen receptor T-cell (CAR-T) therapy has fundamentally transformed the treatment landscape for haematologic malignancies [1,2]. These engineered cellular therapies involve the genetic modification of a patient’s own T-cells to express a synthetic receptor directed against a tumour-associated antigen, enabling potent and targeted anti-tumour immune responses. Since the initial FDA approval of tisagenlecleucel in 2017, six CAR-T therapies have received regulatory approval across two principal antigen targets: CD19, expressed on B-lineage malignancies including diffuse large B-cell lymphoma, acute lymphoblastic leukemia, follicular lymphoma, and mantle cell lymphoma; and B-cell maturation antigen (BCMA), expressed on plasma cells in multiple myeloma [3,4].
Despite their clinical efficacy, CAR-T therapies are associated with a distinct and at times life-threatening toxicity profile that differs substantially from conventional chemotherapy and immune checkpoint inhibitors. The most well-characterized toxicities include cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), which can present acutely and require intensive clinical management [5,6]. However, the full spectrum of AEs across all FDA-approved CAR-T products, including longer-term toxicities such as cytopenias, infections, secondary malignancies, and late-onset neurotoxicity’s, have not been systematically characterized across target classes in a comprehensive pharmacovigilance analysis [2,7,8].
The FDA Adverse Event Reporting System (FAERS) provides a large, publicly available repository of post-marketing pharmacovigilance data enabling identification of disproportionate AE signals. The objective of this study is to ascertain the most frequently reported AEs associated with each FDA-approved CAR-T therapy using FAERS data, reporting the ROR with 95% confidence interval, proportional reporting ratio (PRR), and chi-squared statistic for each AE, and to present findings stratified by therapeutic target class. Prior pharmacovigilance analyses have examined individual CAR-T agents or subsets of the approved products; the present study extends this literature by systematically reporting ROR, PRR, and chi-squared across all six FDA-approved CAR-T therapies in a single analysis stratified by therapeutic target class.

2. Results

There were 13,685 unique adverse event reports for axicabtagene ciloleucel, 13,087 for tisagenlecleucel, 980 for lisocabtagene maraleucel, and 1862 for brexucabtagene autoleucel among the anti-CD19 agents. For the anti-BCMA agents, there were 902 unique reports for idecabtagene vicleucel and 6051 for ciltacabtagene autoleucel. For the anti-BCMA agents, there was a total of 902 AEs reported for idecabtagene vicleucel and 6051 for ciltacabtagene autoleucel. The 25 most frequently reported AEs for each agent, with events, ROR (95% CI), PRR, and chi-squared, are presented in Table 1, Table 2, Table 3 and Table 4 (anti-CD19 agents) and Table 5 and Table 6 (anti-BCMA agents). Cytokine release syndrome and ICANS were the most frequently reported AEs across all six therapies, demonstrating the highest RORs and PRRs consistent with their established status as the principal class-wide toxicities of cellular immunotherapy.
Table 1. Top 25 Most Reported Adverse Events with Axicabtagene Ciloleucel in the FDA Adverse Event Reporting System Database.
Table 2. Top 25 Most Reported Adverse Events with Tisagenlecleucel in the FDA Adverse Event Reporting System Database.
Table 3. Top 25 Most Reported Adverse Events with Lisocabtagene Maraleucel in the FDA Adverse Event Reporting System Database.
Table 4. Top 25 Most Reported Adverse Events with Brexucabtagene Autoleucel in the FDA Adverse Event Reporting System Database.
Table 5. Top 25 Most Reported Adverse Events with Idecabtagene Vicleucel in the FDA Adverse Event Reporting System Database.
Table 6. Top 25 Most Reported Adverse Events with Ciltacabtagene Autoleucel in the FDA Adverse Event Reporting System Database.

2.1. Anti-CD19 CAR-T Cell Therapies

Among the anti-CD19 agents, CRS and ICANS were dominant across all four therapies. Axicabtagene ciloleucel had the largest event burden, with CRS reported in 2012 cases (ROR = 1284.242, PRR = 758.293) and ICANS in 1140 cases (ROR = 2195.754, PRR = 1686.074). A broad neurotoxicity profile was also present for axi-cel, including neurotoxicity NOS (ROR = 164.466, PRR = 147.450, chi2 = 70,612.020), encephalopathy (ROR = 19.494), aphasia (ROR = 14.660), confusional state (ROR = 3.147), and tremor (ROR = 3.055). Myelodysplastic syndrome (ROR = 58.695, PRR = 56.674) and acute myeloid leukaemia (ROR = 22.885) were notable secondary malignancy signals. Tisagenlecleucel demonstrated a broad cytopenia profile with high RORs and PRRs across multiple lineages, and hypogammaglobulinemia had the highest PRR for this AE across all six agents (PRR = 536.484, ROR = 614.106, chi2 = 144,831.660). Lisocabtagene maraleucel showed prominent infectious complications. Brexucabtagene autoleucel had the highest ROR for ICANS among all agents (ROR = 2089.598, PRR = 1468.515, chi2 = 305,837.300) and a distinctive secondary malignancy profile with squamous cell carcinoma of the skin (ROR = 308.203, PRR = 287.588) and basal cell carcinoma (ROR = 74.649).

2.2. Anti-BCMA CAR-T Cell Therapies

Among the anti-BCMA agents, CRS had the highest ROR for any single AE in the entire dataset for ide-cel (ROR = 1934.579, PRR = 843.977, chi2 = 169,836.390). ICANS was also prominent for both agents. Ide-cel demonstrated haematological toxicities including neutropenia (ROR = 18.087, PRR = 16.411, chi2 = 508.513) and anaemia (ROR = 11.847, PRR = 10.990), as well as hypogammaglobulinemia (ROR = 98.060, PRR = 95.680) and myelodysplastic syndrome (ROR = 35.368, PRR = 34.619). Cilta-cel’s most striking finding was cranial nerve paralysis, which had the highest ROR of any AE across all six drugs (ROR = 3167.553, PRR = 3055.046, chi2 = 199,190.440). Parkinsonism was also highly disproportionate (ROR = 145.509, PRR = 136.594, chi2 = 24,840.332), as were Bell’s palsy (ROR = 380.445, PRR = 370.272) and facial paralysis (ROR = 72.714). Hemophagocytic lymphohistiocytosis showed a strong signal with cilta-cel (ROR = 71.030, PRR = 69.470, chi2 = 4520.714).

3. Discussion

Building on prior pharmacovigilance literature, this analysis of all six FDA-approved CAR-T cell therapies identified 36,567 adverse event reports across anti-CD19 and anti-BCMA agents, extending existing safety characterizations by reporting ROR, PRR, and chi-squared in parallel across both target classes within a single unified dataset. Several key themes and novel observations emerge from this dataset.
Cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome were the most frequently reported AEs across all six agents and consistently produced the highest disproportionality signals in the dataset, confirming their status as the defining class toxicities of CAR-T therapy. The ROR for CRS ranged from 242.1 with cilta-cel to 1934.6 with ide-cel, while the ROR for ICANS ranged from 307.4 with cilta-cel to 2195.8 with brexu-cel. The consistency of these signals across agents targeting distinct antigens and manufactured using different constructs underscores that CRS and ICANS are driven by the fundamental biology of engineered T-cell activation rather than target-specific on-tumour effects [2,9,10,11]. Both the ROR and PRR closely tracked each other for these AEs across agents, reflecting the rarity of CRS and ICANS in the broader FAERS drug population and therefore the specificity of the signal to CAR-T products. An interesting trend emerges when comparing CRS and ICANS RORs within individual agents. Ide-cel had the highest CRS ROR (1934.6) but a lower ICANS ROR (1163.6) relative to some CD19-targeting agents, which may reflect the higher CRS rates observed in clinical trials of BCMA-targeting agents in the heavily pre-treated myeloma population, as well as differences in lymphodepletion conditioning intensity [12,13,14]. By contrast, brexu-cel had the highest ICANS ROR among all six agents (2089.6, PRR = 1468.5, chi2 = 305,837), with CRS at a lower but still highly elevated ROR of 834.3. This pattern is consistent with the established observation that brexu-cel, used predominantly in mantle cell lymphoma and B-cell ALL, has higher ICANS rates in clinical trials relative to its CRS burden, potentially reflecting the neuroinvasive biology of these malignancies and the intensity of conditioning [15,16,17]. Liso-cel, by design developed with a defined CD4:CD8 ratio to reduce toxicity, nonetheless demonstrated very high RORs for both CRS (1497.7) and ICANS (1547.2), suggesting that even optimized CAR-T constructs are not exempt from these toxicities in real-world post-marketing practice [18,19].
Beyond ICANS, a broader neurotoxicity spectrum is evident across agents. Neurotoxicity not otherwise classified was significantly reported with all six agents, with RORs ranging from 45.6 with ide-cel to 196.9 with tisagenlecleucel, highlighting that neurological injury extends well beyond the formal ICANS. Axi-cel had the largest absolute burden of neurotoxicity NOS (n = 511, ROR = 164.5, PRR = 147.5, chi2 = 70,612), as well as significant signals for encephalopathy (ROR = 19.5), aphasia (ROR = 14.7), confusional state (ROR = 3.1), and tremor (ROR = 3.1). The breadth and depth of neurotoxic AEs with axi-cel likely reflects both its large volume of post-marketing experience and its association with higher-grade ICANS in pivotal trials.
The most clinically distinctive neurotoxicity finding in this analysis was the late-onset neurological syndrome uniquely associated with ciltacabtagene autoleucel. Cranial nerve paralysis had not only the highest ROR of any AE in the entire dataset (ROR = 3167.6, PRR = 3055.0, chi2 = 199,190), but also the highest absolute event count for this AE of any drug across FAERS, representing a signal of significant strength. Bell’s palsy (ROR = 380.4, PRR = 370.3, chi2 = 27,930), facial paralysis (ROR = 72.7), and parkinsonism (ROR = 145.5, PRR = 136.6, chi2 = 24,840) were similarly striking and absent from the top-25 AE lists of all other agents. The near-perfect concordance of ROR and PRR for these AEs confirms they are not artifacts of differential reporting denominators. These post-marketing signals are consistent with emerging case series and regulatory communications describing a delayed neurotoxicity syndrome following cilta-cel distinct from acute ICANS, characterized by movement disorders, parkinsonism-like features, and cranial neuropathies occurring weeks to months after infusion [20,21,22,23,24]. The mechanistic basis remains incompletely understood but may relate to the prolonged in vivo persistence of BCMA-targeting CAR-T cells with potential off-tumour CNS tropism, possibly driven by low-level BCMA expression on neurons or glia [20,21,25]. Notably, idecabtagene vicleucel, the other approved anti-BCMA agent, did not demonstrate these neurological signals, suggesting the mechanism may be construct-specific rather than a class effect of BCMA targeting, and warranting direct mechanistic comparison between the two agents [20,26].
Hematological toxicities were pervasive across both target classes, though the nature and depth of cytopenia reporting differed importantly across agents. All six drugs demonstrated significant signals for at least one form of cytopenia, consistent with the combined hematotoxic effects of lymphodepleting conditioning regimens, typically fludarabine and cyclophosphamide, and the prolonged immune reconstitution deficit following CAR-T infusion [27,28,29,30]. Tisagenlecleucel exhibited the broadest and most severe cytopenia profile of any agent in this analysis, with high RORs and PRRs across lymphocyte count decreased (ROR = 239.4, PRR = 206.1, chi2 = 64,199), haematocrit decreased (ROR = 191.4, PRR = 168.5, chi2 = 45,383), haemoglobin decreased (ROR = 41.0, PRR = 35.5, chi2 = 10,733), platelet count decreased (ROR = 35.4, PRR = 30.8, chi2 = 8960), white blood cell count decreased (ROR = 32.1, PRR = 28.2, chi2 = 7779), and neutrophil count decreased (ROR = 86.6, PRR = 76.0, chi2 = 21,153). This pattern likely reflects the broad use of tisagenlecleucel across two distinct patient populations, pediatric and young adult B-cell ALL and adult DLBCL, as well as a longer post-marketing history with more accumulated reports. The high chi-squared values for each cytopenia subtype confirm these are robust signals, not noise from small denominators.
Myelodysplastic syndrome emerged as a secondary hematological malignancy signal across three agents: ide-cel (ROR = 35.4, PRR = 34.6, chi2 = 228.6), axi-cel (ROR = 58.7, PRR = 56.7, chi2 = 9176), and most prominently brexu-cel (ROR = 97.9, PRR = 92.3, chi2 = 3867). The presence of this signal across multiple agents and both target classes is notable. MDS following CAR-T therapy may reflect the genotoxic effects of prior alkylating agent therapy in these heavily pre-treated populations, but insertional mutagenesis from lentiviral or retroviral vectors used in CAR construct delivery has also been proposed as a mechanism and remains under active investigation [31,32,33]. The higher MDS signal with brexu-cel may partly reflect the mantle cell lymphoma patient population, who typically have extensive prior exposure to ibrutinib and bendamustine [34].
Hypogammaglobulinemia was a highly disproportionate AE among anti-CD19 agents, with tisagenlecleucel producing the strongest signal in the dataset (ROR = 614.1, PRR = 536.5, chi2 = 144,832), followed by liso-cel (ROR = 174.7, PRR = 167.3, chi2 = 2143) and ide-cel (ROR = 98.1, PRR = 95.7, chi2 = 750). This AE is mechanistically expected with CD19-targeting agents, as durable eradication of both malignant and normal B-cells produces prolonged B-cell aplasia and impaired humoral immunity [35,36]. The higher ROR with tisagenlecleucel relative to other CD19-targeting agents may reflect its longer post-marketing history and therefore more complete capture of this chronic toxicity, which typically manifests months after infusion rather than acutely. The clinical consequences of hypogammaglobulinemia are substantial: patients require regular monitoring of immunoglobulin levels, prophylactic intravenous immunoglobulin replacement, and heightened vigilance for encapsulated bacterial infections. Long-term survivors of CAR-T therapy may require immunoglobulin replacement indefinitely if B-cell recovery fails to occur.
Beyond hypogammaglobulinemia, infectious complications were a significant theme, particularly with liso-cel, which showed notable signals for sepsis (ROR = 5.0), septic shock (ROR = 13.3, PRR = 13.0, chi2 = 66.3), and cytomegalovirus infection (ROR = 28.7, PRR = 28.1, chi2 = 131.1). CMV reactivation in the post-CAR-T setting reflects the profound T-cell-mediated immune suppression from lymphodepletion and the subsequent period of immune reconstitution, and is an under-recognized complication requiring proactive monitoring in CMV-seropositive patients [37,38]. General infection was also significantly reported with cilta-cel (ROR = 3.7, n = 67) and brexu-cel (ROR = 4.3, n = 19). The convergence of lymphodepletion-associated immunosuppression, prolonged hypogammaglobulinemia, steroid use for CRS and ICANS management, and underlying malignancy creates a profound and multifactorial immunocompromised state that clinicians must actively address throughout the post-CAR-T follow-up period.
Secondary malignancies emerged as a signal across multiple agents, though with distinct patterns that appear to reflect both agent-specific biological mechanisms and underlying patient population characteristics. The most striking secondary malignancy profile was observed with brexu-cel, which showed highly disproportionate signals for squamous cell carcinoma of the skin (ROR = 308.2, PRR = 287.6, chi2 = 14,065), basal cell carcinoma (ROR = 74.6, PRR = 71.5, chi2 = 2149), and MDS (ROR = 97.9). These cutaneous malignancies are biologically plausible in this context: brexu-cel is approved for mantle cell lymphoma, a disease predominantly affecting older males with significant cumulative ultraviolet exposure, and the combination of prior alkylating agent immunosuppression and the iatrogenic immunosuppression of the CAR-T procedure likely promotes cutaneous oncogenesis [39,40]. Acute lymphocytic leukemia recurrence was also a distinctive brexu-cel signal (ROR = 125.8, PRR = 123.9), consistent with its indication in relapsed/refractory B-cell ALL. Tisagenlecleucel demonstrated signals for second primary malignancy (ROR = 111.9, PRR = 107.5, chi2 = 9542) and, uniquely, B-cell type acute leukemia (ROR = 1200.9, PRR = 1153.7, chi2 = 87,982), the latter almost certainly reflecting the ALL patient population receiving this therapy rather than a therapy-induced secondary malignancy. However, the presence of treatment failure (ROR = 10.6, n = 95) as a prominent AE with tisagenlecleucel also points to the challenges of durability of response in the ALL population. The FDA’s 2024 decision to add a boxed warning about secondary T-cell malignancies to all CAR-T products reflects growing concern about insertional mutagenesis as a mechanism of therapy-related oncogenesis, and this pharmacovigilance analysis provides population-level support for sustained regulatory vigilance in this domain.
Hemophagocytic lymphohistiocytosis (HLH) was identified as a significant pharmacovigilance signal across multiple agents from both target classes, with cilta-cel producing the strongest signal (n = 69, ROR = 71.0, PRR = 69.5, chi2 = 4521) and brexu-cel also demonstrating a robust signal (n = 14, ROR = 57.8, PRR = 56.7, chi2 = 710). HLH is a life-threatening hyperinflammatory syndrome characterized by uncontrolled macrophage and T-cell activation, multi-organ failure, and profound cytopenias, and it carries high mortality if not promptly recognized and treated. In the post-CAR-T setting, HLH may be particularly difficult to diagnose given the overlapping clinical and laboratory features shared with severe CRS, both syndromes involve hyperferritinemia, cytopenias, and hemodynamic instability, and the absence of universal HLH screening criteria in current CAR-T toxicity management guidelines. The finding that HLH signal strength was highest with cilta-cel among the agents examined is notable. Whether this reflects a true mechanistic predisposition of BCMA-targeting with this specific construct, reporting patterns in the heavily pre-treated myeloma population, or an interaction between cilta-cel’s prolonged persistence and sustained macrophage activation remains unclear [41]. Regardless of mechanism, the convergence of HLH signals across both anti-CD19 and anti-BCMA agents, with high RORs and PRRs in both classes, strongly supports treating HLH as a class-level risk of CAR-T therapy. Physicians may consider maintaining a low threshold for initiating HLH work-up in any patient with refractory or worsening inflammatory symptoms beyond the first week post-infusion, and treatment with anakinra or etoposide-based regimens should be initiated promptly upon diagnosis.
The two N/A chi-squared values for axicabtagene ciloleucel’s CRS and ICANS entries reflect a numerical overflow limitation within OpenVigil 2.1, occurring when observed or expected cell frequencies in the contingency table exceed the tool’s internal computational threshold. This does not indicate an absence of statistical significance; both AEs carry the highest absolute event counts and strongest disproportionality estimates in the dataset.
Several limitations of this analysis require careful consideration. FAERS is a spontaneous reporting database, and reports submitted to it do not establish a causal relationship between the drug and the reported event. The FDA does not require proof of causality, and reports may lack sufficient clinical detail to enable independent assessment. Underreporting is an inherent limitation of spontaneous pharmacovigilance systems, and reporting rates may be disproportionately influenced by the time a product has been on the market, media coverage, institutional reporting practices, and regulatory actions. Agents with longer post-marketing histories, particularly tisagenlecleucel and axi-cel, will naturally accrue more reports and therefore appear to have a broader AE profile than newer agents such as ide-cel and liso-cel, regardless of their true toxicity burden. This temporal confounding should be borne in mind when comparing AE profiles across agents of different approval ages. Confounding by indication is a particularly salient limitation in this analysis. Anti-CD19 and anti-BCMA agents are used in fundamentally different patient populations with distinct baseline comorbidities, prior treatment histories, and malignancy-related complications. AEs such as disease progression, treatment failure, and malignancy recurrence that appear in the FAERS reports may reflect the underlying disease rather than treatment toxicity, inflating disproportionality signals for some AEs. Within the CD19 class, differences in indication, DLBCL versus MCL versus ALL, also produce heterogeneous background populations that cannot be separated within this analysis. Similarly, the very high event counts for some AEs with tisagenlecleucel likely partly reflect the ALL patient population, in which intensive laboratory monitoring generates more granular cytopenia-level reporting. Finally, channelling bias, the tendency for specific patient subgroups to be preferentially treated with certain agents, may create the appearance of agent-specific toxicity signals that in reality reflect patient-level differences.
This comprehensive pharmacovigilance analysis of all six FDA-approved CAR-T cell therapies identified 36,567 adverse event reports across anti-CD19 and anti-BCMA agents, revealing both universal class toxicities and clinically important agent-specific signals. Cytokine release syndrome and immune effector cell-associated neurotoxicity syndrome represent the defining toxicities of CAR-T therapy as a class, with consistently high RORs, PRRs, and chi-squared values across all six agents regardless of target antigen, confirming their mechanistic origin in the biology of adoptive cellular immunotherapy rather than in any target-specific effect.
Within this class-wide framework, several drug-specific signals with direct clinical implications emerged. Ciltacabtagene autoleucel exhibited a late-onset neurological toxicity profile, encompassing cranial nerve paralysis, parkinsonism, Bell’s palsy, and facial paralysis, that was absent from all other agents and produced the highest single-AE ROR in the entire dataset, underscoring the need for prolonged neurological surveillance beyond the acute post-infusion period in patients receiving this agent. Brexucabtagene autoleucel demonstrated a secondary malignancy profile including squamous cell carcinoma of the skin, basal cell carcinoma, and myelodysplastic syndrome that aligns with and extends the evidence base supporting the FDA’s 2024 boxed warning on CAR-T-associated secondary malignancies. Tisagenlecleucel showed the strongest hypogammaglobulinemia signal among all agents, with implications for long-term immunoglobulin replacement and infection surveillance. Hemophagocytic lymphohistiocytosis was identified as a cross-class risk affecting both anti-CD19 and anti-BCMA agents, warranting its inclusion in institutional CAR-T toxicity management protocols.
These findings emphasize that effective management of CAR-T toxicity requires both a class-level foundation and agent-specific vigilance informed by the distinct pharmacovigilance profiles described here. As CAR-T indications continue to expand to earlier lines of therapy and broader patient populations, and as novel constructs targeting additional antigens enter the clinic, ongoing pharmacovigilance and long-term follow-up studies will be essential to characterise the evolving safety landscape of this transformative class of therapies. The present study provides a real-world evidence foundation for such efforts and offers actionable guidance for oncologists, hematologists, and other clinicians involved in the care of patients receiving CAR-T therapy.

4. Methods

The FDA Adverse Event Reporting System (FAERS) serves as a comprehensive repository for documented adverse drug reactions attributed to various pharmaceutical products. Reports are submitted by pharmaceutical manufacturers, healthcare professionals, and consumers in the USA, and FAERS incorporates post-marketing clinical trial data from domestic and international investigations. As all data are publicly available, ethics committee approval was not required.
Using OpenVigil 2.1 (OpenVigil, Kiel, Germany), individual drug queries were performed for all six FDA-approved CAR-T therapies, grouped by antigen target as previously described [42,43]: anti-CD19 agents (axicabtagene ciloleucel, tisagenlecleucel, lisocabtagene maraleucel, brexucabtagene autoleucel) and anti-BCMA agents (idecabtagene vicleucel, ciltacabtagene autoleucel). Reports from the first quarter of 2017 to the first quarter of 2024 were included. For each drug, the 25 most frequently reported AEs were extracted.
Disproportionality analyses were conducted using reporting odds ratios (RORs) and proportional reporting ratios (PRRs). The ROR measures the likelihood of a given AE with drug exposure relative to all other drugs in the database, calculated via a 2 × 2 contingency table. The PRR similarly compares the proportion of a specific AE among all AEs for the drug of interest versus all other drugs. The chi-squared statistic was computed to assess the significance of the PRR. Analyses were restricted to reports in which the drug of interest was designated as the primary suspect agent, excluding concomitant and interacting drug roles. The unit of analysis was the unique adverse event report, defined by caseid after duplicate removal, rather than individual drug-reaction pairs. The comparator population comprised all other drugs in the FAERS database across the same time window. Statistical significance for ROR was defined as a lower-bound 95% confidence interval exceeding 1.00, an ROR greater than 2.00, and a minimum of five or more reports per AE. Statistical significance for PRR was defined as a PRR of 2.00 or greater, a chi-squared statistic of 4.00 or greater, and a minimum of three or more reports, consistent with the Evans criteria. No adjustment for multiplicity was made; given the hypothesis-generating nature of pharmacovigilance disproportionality analysis, correction for multiple comparisons was not applied, and all signals should be interpreted as requiring confirmatory investigation rather than as established causal associations. OpenVigil 2.1 removes duplicate FAERS reports by retaining only the highest case version for each unique caseid, ensuring each adverse event case is counted once. Drug name consolidation is achieved by mapping free-text entries to standardized active substance terms via the WHO Drug Dictionary, collapsing synonym variants into a single query entity.

Funding

This research received no external funding.

Institutional Review Board Statement

Ethics approval and patient consent were not required as the U.S. FDA previously collected all data utilized in this study.

Data Availability Statement

The data are freely available through the FDA Adverse Events Reporting System Database and OpenVigil 2.1.

Conflicts of Interest

The author declares no conflicts of interest.

References

  1. Mahat, U.; Das, A.; Kumar, V.; Prasad, V. Advancing CAR T-Cell Therapy: Evidence-Based Trial Design for Chimeric Antigen Receptor T-Cell Therapy in Oncology. JAMA 2026, 335, 21–24. [Google Scholar] [CrossRef] [Scilit]
  2. Brudno, J.N.; Maus, M.V.; Hinrichs, C.S. CAR T Cells and T-Cell Therapies for Cancer: A Translational Science Review. JAMA 2024, 332, 1924–1935. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  3. Janes, M.J.; Zhao, L.; Bruton, L.G.; Schmidt, A.A.; Parker, J.M.; Wakefield, M.R.; Fang, Y. Rewiring immunity: CAR-T cells as a paradigm shift in hematologic oncology. Crit. Rev. Oncol. Hematol. 2026, 222, 105281. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  4. Parums, D.V. A Review of CAR T Cells and Adoptive T-Cell Therapies in Lymphoid and Solid Organ Malignancies. Med. Sci. Monit. Int. Med. J. Exp. Clin. Res. 2025, 31, e948125. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  5. Zhou, Q.; An, Y.; Zhang, X.; Xiao, X.; Bai, X.; Liu, P.; Pu, Y.; Meng, J.; Zhu, H.; Lyu, C.; et al. Efficacy and safety of tocilizumab in managing cytokine release syndrome after CD19 CAR-T therapy for relapsed or refractory B-cell acute lymphoblastic leukemia. Front. Immunol. 2025, 16, 1530623. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  6. Caserta, S.; Martino, E.A.; Vigna, E.; Bruzzese, A.; Amodio, N.; Lucia, E.; Olivito, V.; Labanca, C.; Mendicino, F.; Morabito, F.; et al. T-Cell–Redirecting Immunotherapies in Relapsed/Refractory Mantle Cell Lymphoma: Current Evidence, Sequencing, and Future Directions. Eur. J. Haematol. 2026, 116, 129–141. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  7. Yu, Z.; Jing, C.; Xie, L.; Min, L.; Li, L.; Wang, Z.; Niu, T. Efficacy and safety of chimeric antigen receptor T-cell in the treatment of hematologic malignancy: An umbrella review of systematic review and meta-analysis. Front. Immunol. 2025, 16, 1608768. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  8. Corraes, A.D.M.S.; Baranwal, A.; Bansal, R.; Gurney, M.; Rosenthal, A.C.; Kharfan-Dabaja, M.; Khurana, A.; Johnston, P.B.; Ansell, S.M.; Kapoor, P.; et al. Factors before leukapheresis that correlate with severe cytopenia and therapy related myeloid neoplasm post CAR-T. J. Clin. Oncol. 2024, 42, 7038. [Google Scholar] [CrossRef] [Scilit]
  9. Diorio, C.; Shraim, R.; Myers, R.; Behrens, E.M.; Canna, S.; Bassiri, H.; Aplenc, R.; Burudpakdee, C.; Chen, F.; DiNofia, A.M.; et al. Comprehensive Serum Proteome Profiling of Cytokine Release Syndrome and Immune Effector Cell-Associated Neurotoxicity Syndrome Patients with B-Cell ALL Receiving CAR T19. Clin. Cancer Res. 2022, 28, 3804–3813. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  10. Gu, T.; Hu, K.; Si, X.; Hu, Y.; Huang, H. Mechanisms of immune effector cell-associated neurotoxicity syndrome after CAR-T treatment. WIREs Mech. Dis. 2022, 14, e1576. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  11. Lee, D.W.; Santomasso, B.D.; Locke, F.L.; Ghobadi, A.; Turtle, C.J.; Brudno, J.N.; Maus, M.V.; Park, J.H.; Mead, E.; Pavletic, S.; et al. ASBMT Consensus Grading for Cytokine Release Syndrome and Neurological Toxicity Associated with Immune Effector Cells. Biol. Blood Marrow Transplant. 2019, 25, 625–638. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  12. Patel, A.; Chaudhari, J.; Mehta, M.; Chaudhari, H.; Muddassir, S.; Seetharaman, K. Assessing the risk of cytokine release syndrome associated with the use of antineoplastic agents: A real-world pharmacovigilance study based on the FDA Adverse Event Reporting System database (FAERS). J. Clin. Oncol. 2025, 43, e23359. [Google Scholar] [CrossRef] [Scilit]
  13. Wesson, W.; Dima, D.; Suleman, N.; Saif, S.I.; Tabak, C.; Logan, E.; Davis, J.A.; McGann, M.; Furqan, F.; Mohan, M.; et al. Timing of Toxicities and Non-Relapse Mortality Following CAR T Therapy in Myeloma. Transplant. Cell. Ther. 2024, 30, 876–884. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  14. Ren, X.; Zhang, G.; Li, G.; Wang, Y. Chimeric antigen receptor T-cell therapy-induced nervous system toxicity: A real-world study based on the FDA Adverse Event Reporting System database. BMC Cancer 2024, 24, 10. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  15. Pleitez, H.G.; Saowapa, S.; Maldonado, A.O.; Kanitthamniyom, C.; Bernal, D.O.; Tijani, L. From Trials to Practice: A 2025 Review of Idecabtagene Vicleucel and Ciltacabtagene Autoleucel Efficacy Across Clinical Studies and Real-World Evidence. Eur. J. Haematol. 2025, 115, 533–546. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  16. van de Donk, N.W.C.J.; Usmani, S.Z.; Yong, K. CAR T-cell therapy for multiple myeloma: State of the art and prospects. Lancet Haematol. 2021, 8, e446–e461. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  17. Rodriguez-Otero, P.; Ailawadhi, S.; Arnulf, B.; Patel, K.; Cavo, M.; Nooka, A.K.; Manier, S.; Callander, N.; Costa, L.J.; Vij, R.; et al. Ide-cel or Standard Regimens in Relapsed and Refractory Multiple Myeloma. N. Engl. J. Med. 2023, 388, 1002–1014. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  18. Abramson, J.S.; Kamdar, M.; Liu, F.F.; Crotta, A.; Previtali, A.; Klijn, S.L.; Wang, P.; Zhang, Y.; Bonner, A.; Lunning, M. Matching-adjusted indirect comparison of lisocabtagene maraleucel versus axicabtagene ciloleucel for second-line treatment of patients with early relapsed or refractory large B-cell lymphoma. Leuk. Lymphoma 2025, 66, 2200–2213. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  19. Lunning, M.A.; Maus, M.V.; Ellithi, M.; Toron, F.; Braun, R.; Wang, L.; Wang, P.; Jones, M.; Elsallab, M. Postmarketing safety profile of chimeric antigen receptor (CAR) T cell therapies in diffuse large B-cell lymphoma (DLBCL): Analysis of real-world (RW) AE reporting from the FDA Adverse Event Reporting System (FAERS). J. Clin. Oncol. 2025, 43, 7028. [Google Scholar] [CrossRef] [Scilit]
  20. van Besien, H.J.; Ozkan, G.; Easton, N.; Tix, T.; Alhomoud, M.; Shouval, R.; Rejeski, K.; Yamshon, S. Non-ICANS Neurologic Toxicity after BCMA CAR T: A systematic review and meta-analysis of 4630 multiple myeloma patients. Blood Adv. 2026. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  21. Saini, M.; Deeksha, F.N.U.; Jandial, A.; Jain, A.; Prakash, G.; Khadwal, A.; Chakraborty, R.; Malhotra, P.; Singh, C. Non-ICANS neurological toxicities associated with BCMA-directed CAR T-cell therapy in patients with plasma cell dyscrasia: A systematic review and pooled analysis. Br. J. Haematol. 2026, 208, 1854–1857. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  22. Hansen, D.K.; Dima, D.; Mian, H.; Devos, J.; Brazauskas, R.; Oloyede, T.; Afrough, A.; Ahmed, N.; Anderson, L.; Banerjee, R.; et al. Safety and efficacy of ciltacabtagene autoleucel for relapsed/refractory multiple myeloma: A CIBMTR study. Blood Cancer J. 2026, 16, 80. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  23. Graham, C.E.; Velasco, R.; Tomas, A.A.; Stewart, O.P.; Dachy, G.; del Bufalo, F.; Doglio, M.; Henter, J.-I.; Ortí, G.; Peric, Z.; et al. Non-ICANS neurological complications after CAR T-cell therapies: Recommendations from the EBMT Practice Harmonisation and Guidelines Committee. Lancet Oncol. 2025, 26, e203–e213. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  24. Lim, K.J.C.; Tan, M.; Parrondo, R.; Chhabra, S.; Dooley, K.; Corraes, A.D.M.S.; Carabenciov, D.; Gertz, M.; Hwa, L.; Haily, S.; et al. Clinical course, risk factors and mitigating strategies for Immune effector cell-associated late onset neurotoxicities after ciltacabtagene autoleucel CAR-T in multiple myeloma. Blood Cancer J. 2025, 16, 18. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  25. Marella, M.; Yao, X.; Carreira, V.; Bustamante, M.F.; Clark, H.B.; Jackson, C.C.; Zudaire, E.; Schecter, J.M.; Glover, T.D.; Shenton, J.; et al. Comprehensive BCMA Expression Profiling in Adult Normal Human Brain Suggests a Low Risk of On-target Neurotoxicity in BCMA-targeting Multiple Myeloma Therapy. J. Histochem. Cytochem. 2022, 70, 273–287, Correction in J. Histochem. Cytochem. 2022, 70, 401. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  26. Hansen, D.K.; Peres, L.C.; Dima, D.; Richards, A.; Shune, L.; Afrough, A.; Midha, S.; Dhakal, B.; Kocoglu, M.H.; Atrash, S.; et al. Comparison of Standard-of-Care Idecabtagene Vicleucel and Ciltacabtagene Autoleucel in Relapsed/Refractory Multiple Myeloma. J. Clin. Oncol. 2025, 43, 1597–1609, Erratum in J. Clin. Oncol. 2025, 43, 1399. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  27. Corona, M.; Brown, S.; Rejeski, K.; Flynn, J.R.; Devlin, S.M.; Raj, S.; de Abia, A.L.; Lin, R.J.; Scordo, M.; Shah, G.L.; et al. Late hematologic toxicity after CAR T-cell therapy in large B-cell lymphoma: Incidence, risk factors, and clinical impact. Bone Marrow Transplant. 2026, 61, 142–148. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  28. Rejeski, K.; Jain, M.D.; Shah, N.N.; Perales, M.A.; Subklewe, M. Immune effector cell-associated haematotoxicity after CAR T-cell therapy: From mechanism to management. Lancet Haematol. 2024, 11, e459–e470, Correction in Lancet Haematol. 2024, 11, e480. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  29. Zhuang, T.; Rupji, M.; Santapuram, P.; Graiser, M.; Romancik, J.; Cohen, J.B. Characterizing outcomes in severe cytopenias by 3 months in BCMA and CD19 CAR-T cell therapies: A single institution experience. J. Clin. Oncol. 2023, 41, e19504. [Google Scholar] [CrossRef] [Scilit]
  30. Nagle, S.J.; Murphree, C.; Raess, P.W.; Schachter, L.; Chen, A.I.; Hayes-Lattin, B.; Nemecek, E.; Maziarz, R.T. Prolonged hematologic toxicity following treatment with chimeric antigen receptor T cells in patients with hematologic malignancies. Am. J. Hematol. 2021, 96, 455–461. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  31. Mohty, R.; Halwani, A.; Badar, T.; Alkhateeb, H.; Shah, M.V.; Qin, H.; Kharfan-Dabaja, M.A. Malignancies after Chimeric Antigen Receptor T Cell Therapy. Transplant. Cell. Ther. 2025, 31, 863–873. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  32. Gazeau, N.; Beauvais, D.; Tilmont, R.; Srour, M.; Ferrant, E.; Safar, V.; Fouillet, L.; Flandrin-Gresta, P.; Gower, N.; Chauvet, P.; et al. Myeloid neoplasms after CD19-directed CAR T cells therapy in long-term B-cell lymphoma responders, a rising risk over time? Leukemia 2025, 39, 1714–1722. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  33. Patel, S.A.; Spiegel, J.Y.; Dahiya, S. Second Primary Cancer After Chimeric Antigen Receptor–T-Cell Therapy: A Review. JAMA Oncol. 2025, 11, 174–181. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  34. Abalo, K.D.; Smedby, K.E.; Ekberg, S.; Eloranta, S.; Pahnke, S.; Albertsson-Lindblad, A.; Jerkeman, M.; Glimelius, I. Secondary malignancies among mantle cell lymphoma patients. Eur. J. Cancer 2023, 195, 113403. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  35. Galli, E.; Fresa, A.; Bellesi, S.; Metafuni, E.; Maiolo, E.; Pansini, I.; Frioni, F.; Autore, F.; Limongiello, M.A.; Innocenti, I.; et al. Hematopoiesis and immune reconstitution after CD19 directed chimeric antigen receptor T-cells (CAR-T): A comprehensive review on incidence, risk factors and current management. Eur. J. Haematol. 2024, 112, 184–196. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  36. Liu, K.J.; Zhang, L.; Zhou, B.; Hong, J.S.; Pak, A.; Yee, A.J.; Raje, N.S.; Frigault, M.J.; Maus, M.V.; Barmettler, S. Characterization of Hypogammaglobulinemia, Infection Incidence, and Mortality in Patients Receiving B-cell Maturation Antigen Chimeric Antigen Receptor (CAR) T-cell Therapy. Transplant. Cell. Ther. 2026; in press. [CrossRef] [Scilit] [PubMed]
  37. Othman, T.; Baird, J.H.; Wang, Y.; Pak, S.; Zhong, H.; Wen, Y.-P.; Shouse, G.; Herrera, A.; Popplewell, L.; Aribi, A.; et al. High incidence of cytomegalovirus infection in patients receiving standard of care CD19- or BCMA-targeted CAR T-cell therapies for hematologic malignancies. Transplant. Cell. Ther. 2026; in press. [CrossRef] [Scilit] [PubMed]
  38. Kaplan, D.; Blyth, E.; Sutrave, G.; Yong, M.K.; Gottlieb, D.J.; Abendroth, A.; Slobedman, B.; Stern, L. Human Cytomegalovirus Infection in Haematopoietic Stem Cell Transplant Recipients and CAR T Cell Recipients—PART 1: Risk Factors, Clinical Impact and Immune Response. Rev. Med. Virol. 2026, 36, e70142. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  39. Umyarova, E.; Pei, C.; Pellegrino, W.; Zhao, Q.; Sharma, N.; Benson, D.; Cottini, F.; Bezerra, E.; Brammer, J.; Bumma, N.; et al. Second primary malignancies following CAR T-cell therapy in patients with hematologic malignancies. J. Hematol. Oncol. 2025, 18, 30. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  40. Shah, B.K.; Khanal, A. Second Primary Malignancies in Mantle Cell Lymphoma: A US Population-based Study. Anticancer Res. 2015, 35, 3437–3440. [Google Scholar] [PubMed]
  41. Zhang, C.; Wu, H.; Cao, T.; Wu, C.; Yan, C.; Qi, W.; Zhang, Y.; Ji, X.; Jing, H.; Peng, Q.; et al. Hemophagocytic Lymphohistiocytosis After CAR T-Cell Therapy: A Systematic Review and Meta-Analysis. Am. J. Hematol. 2026, 1–17. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  42. Frey, C.; Cherniawsky, H.; Etminan, M. Ocular adverse events following CAR-T cell therapy: A pharmacovigilance study and systematic review. Eur. J. Haematol. 2024, 113, 66–71. [Google Scholar] [CrossRef] [Scilit] [PubMed]
  43. Frey, C. Hepatobiliary Adverse Events Associated with Pembrolizumab: A Pharmacovigilance Study from the FDA Adverse Event Reporting System (FAERS) Database. Pharmacoepidemiology 2025, 4, 1. [Google Scholar] [CrossRef] [Scilit]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Article Metrics

Citations

Article Access Statistics

Multiple requests from the same IP address are counted as one view.