Acute lymphoblastic leukemia (ALL) is the most common pediatric cancer, with approximately 75% of cases occurring in patients aged <6 years [1]. During the past few decades, the application of classical multiagent chemotherapy in children with acute B-cell lymphoblastic leukemia/lymphoma has resulted in a marked improvement in survival rate, with the current 5-year overall survival (OS) exceeding 90% in high-income countries [2, 3]. However, survival rates in pediatric patients with relapsed B-cell ALL are significantly low, with a 5-year OS rate of 19–52% [4]. Immune effector cell therapies have significantly improved the survival of patients with relapsed/refractory (R/R) leukemia. Tisagenlecleucel (tisa-cel), a second-generation CD19-directed chimeric antigen receptor (CAR) T-cell product, is an effective and well-tolerated therapy approved by the Food and Drug Administration for use in pediatric and young adult patients with R/R B-cell ALL and in adults with R/R/diffuse large B-cell lymphoma and R/R follicular lymphoma [5]. The pivotal ELIANA clinical trials [6] along with the recently published data [7] demonstrated high long-term (median follow-up of 38.8 mo) relapse-free survival and OS rates of 52% and 63%, respectively, in children with B-cell ALL.

Despite the remarkable clinical success of CAR T-cell therapy, cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), and B-cell aplasia-associated hypogammaglobulinemia have been reported as fatal toxic side effects in pivotal trials [6, 8-10]. CRS and ICANS usually occur within the first eight weeks after infusion and are mostly manageable and reversible with proper supportive care. Nevertheless, it is still imperative that clinicians identify and manage these common toxicities to avoid their life-threatening consequences [11, 12]. To date, numerous studies have addressed early and delayed adverse effects associated with CAR T-cell therapy, and protocols for diagnosing, grading, and managing CRS/ICANS and other toxicities have been developed [13-16]. The guidelines for first-line treatments to mitigate CRS/ICANS are well-established, while second-line treatments, which include anakinra (IL-1 receptor antagonist) [17, 18] and siltuximab (IL-6 receptor antagonist) [13] to treat persistent inflammatory processes, are still under active clinical investigation.

Since the clinical use of CAR T-cell therapy progressed rapidly since its approval in 2017, other delayed toxicities beyond CRS and ICANS have emerged. Among these, hematologic toxicities such as toxicity, anemia, neutropenia, and thrombocytopenia are frequently observed in clinical practice. These common and poorly understood toxicities have been under-reported. Uniform reporting across clinical trials and data collection worldwide is required to understand the mechanisms of hematologic toxicities and associated risk factors. Although CAR T-cell therapy enhances long-term survival, it can also be accompanied by long-term adverse effects that can affect the patient’s chronic health condition.

In this review, the most recent updated data regarding the common toxicities associated with the use of commercially available CAR T-cell products, including CRS, ICANS, and hematologic toxicities, mostly in children and young adults with B-cell ALL.

CRS is a clinical syndrome caused by the rapid release of cytokines from immune effector cells after immuno-therapy such as CAR T-cell therapy. The timing of CRS onset differs across clinical trials; tisa-cel contains 4-1BB costimulatory subunits and has a median onset of three days after infusing patients with B-ALL [6]. The clinical features of CRS include fever, hypoxia, tachycardia, capillary leak, and hypotension [12, 19]. In cases of severe CRS, multi-organ failure, including cardiac arrhythmia, acute renal injury, liver dysfunction, and coagulopathy, can occur and prove fatal. The incidence and severity of CRS vary according to CAR T-cell products, CAR T-cell doses, different malignancies, and disease status at the time of infusion [20]. A prevalence of 70% has been reported in clinical trials of CAR T-cell therapy in children and adults [6]. Although frequent, the symptoms were usually mild-to-moderate and resolved within a few days without requiring intervention. Previous data showed that only 16–48% of children with B-ALL exhibited ≥grade 3 CRS, whereas life-threatening CRS was uncommon (∼7.5%) [11, 21-24].

When CAR T-cells associate with their target (CD19) on lymphoblasts and normal B cells, signaling through this receptor induces marked in vivo activation and proliferation of CAR T-cells, resulting in a cascade of immune stimulation of several cell types, such as macrophages, monocytes, and dendritic cells [25, 26]. This process leads to high serum concentrations of pro-inflammatory cytokines. Notably, interleukin (IL)-6, IL-8, IL-10, IL-15, interferon gamma (IFN-γ), monocyte chemoattractant protein-1, C-reactive protein, and ferritin have been associated with severe CRS [20, 25, 27-32]. However, there is no validated model to predict CRS severity. For patients with a high pre-infusion disease burden (≥5% BM lymphoblasts, CNS3, or non-CNS extramedullary disease), the intensity of lymphodepletion chemotherapy, CAR T-cell doses, pre-existing thrombocytopenia, and comorbidities showed inferior survival and increased toxicities [33].

The first sign of CRS is fever, and mild CRS (grade 1) can be treated. In contrast, treated patients with high CRS grades should be transferred to the local intensive care unit (ICU) (Table 1). Early detection and management of CRS are important to mitigate the risk of fatal adverse events. Once the diagnosis and grading of CRS are made, treatment is initiated with antipyretics, fluid therapy, and vasopressor therapy in cases of hypotension [12]. CRS symptoms and laboratory findings may be misdiagnosed as infection; therefore, all febrile patients should undergo comprehensive infection evaluation and empirical broad-spectrum antibiotics should be initiated [34]. Other differential diagnoses include tumor lysis syndrome and disease progression. When fever persists for ≥3 days with no signs of other illnesses, anti-cytokine therapy should be considered to reduce the exaggerated immune response [13].

Recent studies have shown that IL-1 and IL-6 released by monocytes and macrophages are the major reasons for CAR T-cell-related CRS and ICANS, with monocytes contributing more than macrophages [28, 30]. The use of tocilizumab, a recombinant monoclonal IL-6 receptor antagonist approved for treating CRS in children undergoing CAR T-cell therapy [23, 35], can result in a rapid improvement of CRS symptoms. If the patient’s symptoms did not improve after two doses of tocilizumab, dexamethasone was administered. Corticosteroids are recommended for severe CRS, particularly when associated with neurotoxicity [13]. Furthermore, recent studies have suggested that both tocilizumab and dexamethasone can be used early [36, 37] and do not affect the clinical outcomes of CAR T-cell therapy [8, 15]. Two study groups have reported their experiences with early or prophylactic tocilizumab administration in pediatric patients. The Seattle Children’s Research Institution published the results of tocilizumab use for mild CRS in their PLAT-02 (NCT02028455) trial. Only 15% of patients who received tocilizumab early exhibited severe CRS compared with the control group (30%) [38]. In the ELIANA study (CTL019), 47% of patients required ICU admission to manage severe CRS. Notably, the principal risk factor identified for severe CRS is a high tumor burden (≥40% lymphoblasts) in the bone marrow [23, 24]. Preemptive tocilizumab administration based on disease burden, in which patients with a high tumor burden received a single dose of tocilizumab after developing a high/persistent fever, led to a reduction in the expected incidence of grade 4 CRS [39]. These studies demonstrated that the early and prophylactic use of tocilizumab did not negatively impact the antitumor efficacy of CAR T-cell therapy. However, there is insufficient clinical data for children to support formal recommendations.

Secondary hemophagocytic lymphohistiocytosis (HLH) is a cytokine-driven fulminant hyperinflammatory syndrome associated with immune-induced multiorgan failure [40]. Secondary HLH can occur in patients with severe infection, malignancy, or autoimmune disease, and is rarely associated with allogeneic hematopoietic stem cell transplantation and CAR T-cell therapy. Reports on HLH-like toxicity following CAR T-cell therapy (carHLH) are increasing, with fever, organomegaly, jaundice, gastrointestinal problems, and pulmonary complications being the major clinical features [18, 41-43]. Although the incidence of carHLH is relatively rare (3–4%) [44], it has a high mortality rate. Moreover, it is difficult to distinguish carHLH from severe CRS because of overlapping symptoms and the tendency of CRS to progress into fulminant HLH. Serum ferritin is a suitable and reliable biomarker of HLH and can be used to monitor treatment responses [44]. Neelapu et al. [19] proposed diagnostic criteria for carHLH, which include a peak serum ferritin level of >10,000 μg/L and concomitant organ failure (Table 2).

There is no approved protocol for grading carHLH; its management is independent of grading [14]. Aggressive immunosuppression is required to effectively manage carHLH, and tocilizumab and corticosteroids remain the recommended first-line treatment [13, 44]. Given the fatal nature of fulminant HLH, alternative approaches to control hyperinflammation should be considered in patients resistant to first-line treatment. Anakinra, a recombinant IL-1 receptor blocker, has been approved for treating autoinflammatory disorders in children [45]. Data regarding the clinical use of anakinra in children with carHLH are limited, and treatment guidelines for it have not yet been established. Anakinra is typically added to advanced CRS as a second-line treatment, and studies have shown that it may be effective in controlling CAR T-cell therapy-associated toxicities by suppressing inflammatory pathways [17, 41, 43, 46]. Although the optimal dose and treatment duration has not been determined, a starting dose of 2 mg/kg subcutaneously or intravenously every 6 h (8 mg/kg/day, maximum daily dose of 400 mg) is administered from days to weeks, based on protocols used in autoimmune or rheumatic diseases [47, 48]. Anakinra was suggested to serve as a steroid-sparing agent and the last medication to be weaned off without affecting CAR T-cell efficacy; however, this should be validated in future prospective trials. In addition to anakinra, etoposide can be used to selectively remove activated T-cells, resulting in therapeutic benefits. However, it may also eliminate the patient’s CAR T-cells, thus reducing their clinical efficacy [19, 49].

ICANS is a common and challenging adverse effect of CAR T-cell therapy, which occurs in 25–44% of children with hematologic malignancies [11, 22, 23]. The incidence rate of severe ICANS observed in pivotal studies on tisa-cel was 12–13% [6, 50]. ICANS usually occurs within 7–10 days and sometimes up to 3 weeks after CAR T-cell infusion and can occur concurrently or shortly after CRS [25]. The underlying pathophysiology of ICANS is not fully understood, but it appears to be associated with the release of inflammatory cytokines by macrophages/monocytes, increased vascular permeability, and endothelial activation, resulting in the disruption of the blood-brain barrier (BBB) [51]. Common clinical symptoms of ICANS include depressed consciousness, language disturbances, seizures, motor weakness, and cerebral edema. Dysgraphia has been noted as an early indicator of ICANS [19]. Although most ICANS cases are reversible and manageable, severe presentations, such as rapidly progressive cerebral edema, can be fatal. The clinical risk factors associated with an increased risk of severe ICANS include young age, high disease burden, severe CRS with high ferritin and/or cytokine levels, neurological comorbidities, and high CAR T-cell doses [52-54].

To grade ICANS, the 10-point immune effector cell-associated encephalopathy (ICE) score was used in adult patients. However, this tool is inappropriate for pediatric patients, particularly those aged <12 years or patients with developmental delays [15]. The Cornell Assessment of Pediatric Delirium (CAPD), adapted by Traube et al. [55], is a validated screening tool for recognizing delirium in pediatric patients, and its sensitivity and specificity are highest in patients aged <12 years. The guidelines recommend that delirium assessment should be performed at least twice per day during the high-risk period of ICANS, after 4 weeks of CAR T-cell infusion [13, 15]. Assessing patterns in ICE/CAPD scores within an individual patient is crucial for early intervention. The ICANS grading integrates the ICE (for children aged ≥12 yr) and CAPD (for children aged <12 yr) scores into an overall assessment of neurological function (Table 3

Neuroimaging findings during ICANS in children with ALL are still lacking; therefore, there are no pathognomonic imaging findings related to ICANS [56]. Brain magnetic resonance imaging, electroencephalography, and cerebrospinal fluid (CSF) examination can aid in the differential diagnosis of stroke, hemorrhage, infection, and malignancy. Gaining a clear understanding of the various imaging phenotypes and cytokine levels in CSF will improve timely diagnosis and therapeutic interventions.

Although there is no well-established recommendation to guide the management of ICANS in children with B-ALL [57], supportive care with or without short-term corticosteroids is preferred as the initial treatment. Patients with grade 1 (mild) ICANS were offered supportive care with intravenous hydration and aspiration precautions. Corticosteroids (dexamethasone or methylprednisolone) should be initiated for ≥grade 2 ICANS and continued until symptoms improve to grade 1 and then be rapidly tapered. There is no clear therapeutic role for tocilizumab in ICANS because of its poor BBB penetration [58]; therefore, tocilizumab is only considered in patients with concurrent CRS [13]. Results of two clinical trials, NCT04148430 and NCT04359784, have suggested that when patients’ symptoms deteriorate after the addition of high-dose methylprednisolone, anakinra alleviates clinical symptoms (NCT04148430, NCT04359784). The management recommendations for the ICANS are listed in Table 4.

BCA, which results in the “on-target, off-tumor” activity of CAR T-cells targeting B-cell markers (CD19), provides a surrogate for the persistence of potentially functional CAR T-cells. Patients with ongoing remission show prolonged BCA for months or years; thus, BCA is regarded as an indicator of CAR T-cell treatment effectiveness. However, persistent BCA can lead to hypogammaglobulinemia, an immunological immaturity-related risk factor for sinopulmonary and other infections [59]. According to the ELIANA study, 71% and 59% of pediatric B-ALL patients developed BCA at 12 and 24 months, respectively. In this review, the most frequent grade 3/4 adverse event occurring >1 year after the infusion was infection, with a prevalence of 20.4% [7]. In adults, pathogen-specific antibodies can be detected in the blood due to CD19-negative plasma cells [60]. However, after tissue treatment, plasma cell and serum antibody counts decline progressively [61]. Periodic intravenous immunoglobulin (IVIG) supplementation is recommended for children and young adults during BCA [13-15]. Hill et al. [62] proposed a practical algorithm for hypogammaglo-bulinemia management: first, screening for serum IgG within the first three months after CAR T-cell therapy; second, if patients have IgG ≤400 mg/dL, prophylactic IVIG should be considered; and third, in adults with infection and IgG ≤400 mg/dL, IVIG replacement is the only recommendation three months after infusion, while IVIG replacement is routine in children [13].

Cytopenia is usually observed within 28 days (early cytopenia) after CAR T-cell infusion; however, it may occasionally persist for >90 days (late cytopenia) [63]. CAR T-cell-associated cytopenia is exceedingly common in both CD28-based and 41BB-based costimulatory CAR products, with an incidence range of 50–80% [64]. CAR T-cell-associated cytopenia is biphasic (even triphasic) in nature [65]. With tisa-cel for R/R/B-cell ALL in children and young adults, 43% and 53% of patients had grade 3–4 thrombocytopenia and neutropenia, respectively, which persisted beyond day 30 after infusion [6]. Early cytopenia is attributed to the intensity of bridging chemotherapy, lymphodepletion chemotherapy pre-infusion, severe CRS, and carHLH [63-65]. While standard lymphodepletion chemotherapy generally does not cause prolonged cytopenia, it might exert additional myelosuppression in patients with pre-existing bone marrow infiltrating disorders [66, 67]. High levels of IFN-γ, which are increased in patients with severe CRS or carHLH, may suppress hematopoietic stem cell homeostasis, resulting in stem cell exhaustion [68].

The prolonged duration (>3 mo after infusion) of cytopenia is a unique aspect of the hematologic toxicity of CAR T-cell therapy. Delayed grade 3–4 cytopenia is often observed at a rate of 17–32% after tisa-cel therapy [6, 10]. Although the mechanism of prolonged cytopenia remains unclear, data suggest that baseline cytopenia, severe CRS, and allogeneic hematopoietic stem cell transfusion within 1 year are risk factors associated with late cytopenia [64, 66]. In a retrospective analysis, the levels of stromal cell-derived factor 1 correlated with neutropenia in patients with prolonged cytopenia [64]. More research on bone marrow should be considered in patients with late cytopenia to differentiate it from marrow dysplasia [69]. Since persistent cytopenia may be linked to susceptibility to opportunistic infections and significant morbidity, the use of granulocyte colony-stimulating factor (G-CSF) for hematologic recovery from day 14 after infusion (following the resolution of CRS and ICANS) is recommended according to the institutional protocol [14]. Recent data showed that the early use of prophylactic G-CSF did not negatively impact CAR T-cell expansion or prognosis [70]. While prophylactic treatment against infection, including viruses and Pneumocystis carinii pneumonia, is recommended, there is no consensus on a specific antibacterial or antifungal prophylaxis for patients with late cytopenia [13, 14].

This review summarizes the current knowledge on common complications after anti-CD19 CAR T-cell therapy in children and adults with B-cell malignancies. The management of CAR T-cell toxicities has evolved rapidly, and well-established guidelines based on expert recommendations exist. Future clinical trials should concentrate on the prevention and mitigation of these toxicities by refining CAR products, optimizing CAR T-cell counts, and developing novel therapies. Additionally, prospective and comprehensive studies investigating the late effects and establishing long-term follow-up guidelines for survival are imperative in this field.