Toxicity of CAR T-Cell Therapy for Multiple Myeloma Open Access

In 1987, Kuwana and colleagues [1] depicted the first chimeric receptor, combining immunoglobulin-derived variable regions with the T-cell receptor (TCR). Two years later, in 1989, immunologist Eshhar and colleagues [2] pioneered genetically engineered T cells known as chimeric antigen receptor T cells (CAR-T). They successfully reported the construction of functional chimeric immunoglobulin-TCR genes that combined the extracellular antigen-binding variable domains from an immunoglobulin with the intracellular T-cell signaling domain of a TCR. These genes could be expressed in T cells and recognize antigens in a non-MHC-restricted manner. This groundbreaking development eventually paved the way to produce second-generation CAR-T molecules that involved the addition of a costimulatory signaling domain (either 4-1BB or CD28), and at this time, six of these second-generation CAR-T therapies have been approved for various diseases, including acute lymphoblastic leukemia, non-Hodgkin lymphoma, and most recently multiple myeloma (MM).

MM is the 2nd most common hematologic malignancy and is characterized by uncontrolled proliferation of clonal plasma cells in the bone marrow [3]. A particularly challenging subgroup of myeloma patients to treat are the so-called “triple-class refractory” who have already received the 3 major classes of therapy (an immunomodulatory drug, a proteasome inhibitor, and an anti-CD38 monoclonal antibody) and relapsed again [4, 5]. Based on unprecedented response rates in this difficult-to-treat population, the US FDA has granted approval of two B-cell maturation antigen (BCMA)-directed CAR-T therapies: idecabtagene vicleucel (ide-cel) in March 2021 and ciltacabtagene autoleucel (cilta-cel) in February 2022, specifically for the treatment of relapsed and refractory myeloma patients that are triple class exposed and have relapsed after at least 4 prior lines of therapy. Moreover, recent studies show that these therapies are more effective than standard-of-care options in early relapsed myeloma. As a result, they received FDA approval for use in earlier lines of therapy for relapsed settings, starting April 2024. Concurrently, numerous other products targeting BCMA as well as other novel target antigens are currently under investigation in clinical trials.

Despite the considerable success of CAR-T therapy, it is accompanied by a significant drawback – the potential for substantial toxicities. The most frequently observed adverse events (AEs) associated with CAR-T therapy include cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), non-ICANS neurotoxicity, cytopenias, infections, hypogammaglobulinemia, and hemophagocytic lymphohistiocytosis (HLH)-like [6]. In this review, we will delve into the toxicities related to CAR-T therapy with guidance on current recommendations for management and clinical trial findings.

CAR-T activation triggers local cytokine release, initiating macrophage activation and subsequent release of various cytokines and chemokines including IL-1, IL-6, TNF-α, and IL-10 [7]. This leads to systemic inflammation, increased vascular permeability, and potential neurotoxicity. Macrophages are now recognized as the primary cytokine source in CRS, releasing IL-6 and IL-1, contributing to systemic inflammation and local complications like bone marrow hemophagocytosis [8‒10].

CRS usually manifests as fevers, arthralgias, and myalgias and results from the activation of T cells when they recognize tumor antigens, known as on-target on-tumor toxicity, leading to a transient increase in cytokines [11, 12]. The timing of symptom onset depends on the kinetics of T-cell activation and can vary from early to delayed, but it usually occurs within the 1–2 weeks following CAR-T administration and clearing within a few days to 2–3 weeks after infusion [13]. The three key factors influencing the grading of CRS are fever, hypotension, and hypoxia. However, due to the nonspecific nature of these symptoms, it is crucial to rule out other potential causes, particularly infections. The pivotal trials employed slightly different grading systems: KarMMa and KarMMa-3 used the 2014 criteria by Lee et al. [11] to evaluate ide-cel, while the CARTITUDE-1 and CARTITUDE-4 trials used the 2014 Lee criteria [11] in the phase 1b portion and the American Society for Transplantation and Cellular Therapy (ASTCT) criteria [13] in the phase II and III portions. In practice, both the Society for Immunotherapy of Cancer (SITC) [14], and ASTCT have agreed to utilize the ASTCT grading system for CRS [13]. Table 1 provides a brief overview of these grading systems. Note that grade 5 is defined as CRS-related death not primarily caused by other factors.

In the pivotal KarMMa trial that led to FDA approval of ide-cel, the incidence of CRS was 84% with median time to onset of 1 day. However, only 5% experienced grade 3–4 CRS, and just 1 patient (<1%) had grade 5 (Table 2) [15, 16]. In the KarMMa-3 trial of ide-cel in earlier lines of therapy, the rate of CRS was similar: 88% overall, with 5% grade 3–4, and 1% grade 5 [17]. Furthermore, a retrospective analysis of real world data (RDW) from the US Myeloma Immunotherapy Consortium has been reported combining data from 159 patients at 11 centers across the USA and showed no difference in the rates of CRS (overall 82% with 2% grade 3–4 and 1% grade 5) when using the commercially approved ide-cel product, despite treating a population with a significantly higher rate of comorbidities [18].

In the pivotal CARTITUDE-1 trial that led to approval of cilta-cel in RRMM, CRS occurred in 95% with a median time to onset of 7 days, but grade 3–4 was only seen in 4%, with 1% incidence of grade 5 [19, 20]. When used as second-line therapy in the CARTITUDE-4 study (where extent of tumor burden was less and cytoreduction from bridging therapy was more optimal), the incidence of CRS was reduced to 76% with only 1.1% grade 3–4, with no incidence of grade 5 [21]. The US Myeloma Immunotherapy Consortium also recently presented the first RWD analysis of cilta-cel, and of the 139 patients receiving cilta-cel, the rate of CRS was 81% overall with 7% grade 3 or higher [22].

Managing CRS involves tailoring interventions based on the severity of toxicity, risk factors of the individual patient, and the expected side effect profile of the product infused, prioritizing symptom relief and patient safety for optimal outcomes. Mild CRS (grade 1 and 2) typically involves symptomatic management using antipyretics, intravenous fluids, and supportive measures, especially for ide-cel. However, high tumor burden at baseline has been shown to be associated with severe CRS, and those that progress to grade 2 or higher CRS are more likely to develop severe ICANS or delayed non-ICANS neurotoxicity, especially with cilta-cel. Therefore, for those with high tumor burden or for any patient receiving cilta-cel, many would advise treating earlier and more aggressively. Also, if mild CRS persists despite symptomatic measures (fevers breaking through acetaminophen or lasting more than a day, etc.), it is advised to escalate treatment as if it were severe CRS. Severe CRS (grade 3 and 4 or the scenarios above mentioned for persistent lower grade CRS) warrants starting with an interleukin-6 receptor antagonist like tocilizumab, either alone or in combination with systemic corticosteroids like dexamethasone [11]. Those that respond quickly may only need one dose of tocilizumab and corticosteroids but those that have persistent symptoms may need repeated doses of tocilizumab and corticosteroids or escalation to the next-line therapy.

In some high-risk situations, one may wish to consider intervening before the onset of symptoms from CRS. Prophylactic administration of tocilizumab in CD19-directed CAR-T for lymphoma showed that it was effective at preventing severe CRS; however, ICANS rates were increased [23]. Prophylactic corticosteroids and, more recently, the interleukin 1 receptor antagonist anakinra have also been used in lymphoma CAR-T, but it remains unclear how these strategies might impact disease response in myeloma [24‒26]. The use of corticosteroids as a treatment for CRS and ICANS in CAR-T therapy initially raised concerns about potential inhibitory effects and reduced CAR-T counts [27, 28]. Although studies have revealed that low-dose, short-duration corticosteroids with rapid tapering had no impact on CAR-T expansion and persistence, limited data exist regarding high-dose, long-term corticosteroid effects on CAR-T expansion and persistence [29]. In large B-cell lymphoma, higher cumulative doses and prolonged early corticosteroid use correlated with early progression and shorter survival after CAR-T therapy [30], while another study in MM found no impact on CAR-T clinical efficacy [31]. Despite these contradictory findings across different populations and diseases, the conflicting data emphasize the need for a cautious balance between managing CRS and minimizing potential long-term toxicity, particularly regarding high-dose, long-duration corticosteroid use.

For those with CRS resistant to IL-6 blockade and corticosteroids, or those with rapidly progressive symptoms, anakinra, an IL-1 receptor antagonist, is commonly used as the next-line therapy. A retrospective study of 43 patients with B-cell or plasma cell malignancies experiencing severe and/or refractory CRS/ICANS found that recipients of high-dose anakinra (>200 mg/day intravenously) experienced quicker resolution of CRS/ICANS and lower treatment-related mortality compared to those receiving lower doses (100–200 mg/day subcutaneously or intravenously) [32]. Anakinra’s role is further discussed in the neurotoxicity section.

Various groups have explored potential predictors of CRS. The EASIX (Endothelial Activation and Stress Index) score, calculated using (LDH [U/L] × creatinine [mg/dL])/platelets [10⁹ cells/L], holds promise in identifying CD-19-targeted CAR-T therapy patients at higher risk for severe CRS or ICANS [9, 33]. However, further validation in larger studies is imperative to solidify its utility. Despite this, its relevance in MM patients undergoing BCMA-directed CAR-T therapy remains uncertain [34]. Many have also tried to correlate a relationship between the rate or severity of CRS and the depth of response to CAR-T therapy. However, given that the majority experience low-grade CRS and only a few experience high-grade CRS, drawing conclusions based on the grade proves challenging. Nevertheless, data indicate that CRS correlates with effective CAR-T expansion, and effective CD8 CAR-T expansion, in turn, correlates with response in myeloma [35].

Neurotoxicity represents another common adverse event, though less common in myeloma CAR-T compared to most products used in lymphoma CAR-T. Neurotoxicity may exhibit distinct manifestations that can occur either early (within the first 1–3 weeks following cell infusion) or in a delayed manner (>4 weeks posttreatment) [36, 37]. Although neurotoxicity may be concurrent/overlapping with CRS, it may also be an independent toxicity that does not coincide with CRS [38].

The underlying mechanisms behind neurotoxicity are not fully understood. There is evidence suggesting that cytokine-mediated endothelial activation leads to blood-brain barrier (BBB) disruption and a transient leak of cytokines into the cerebrospinal fluid and brain [36]. Neurotoxicity involving the central nervous system in a nonfocal manner is called ICANS. Other neurotoxicities with more variable and wide-ranging symptoms, such as parkinsonism, cranial nerve palsy, sensory loss, ataxia, peripheral motor or sensory neuropathy, altered mental status, and nystagmus, are often termed atypical neurotoxicity or “non-ICANS neurotoxicity” [37]. Focal manifestations such as movement and neurocognitive treatment emergent adverse events (MNTs) may be due to on-target off-tumor toxicity related to the expression of BCMA in basal ganglia or other neurological tissue [37, 39].

The presentation of neurotoxicity is diverse and can vary between patients. ICANS is a clinical diagnosis, graded using the ASTCT’s 10-point Immune Effector Cell-Associated Encephalopathy (ICE) system, which builds upon the CARTOX-10 tool based on MMSE and was later modified by ASTCT to include a command-following assessment (Tables 3, 4) [13]. ICANS manifests as one or more of a range of neurological symptoms, ranging from minor cognitive shifts to severe conditions like encephalopathy and seizures, which prompted the FDA label prohibiting patients from driving for 2 months after infusion [40]. Initial clinical signs encompass dysgraphia, tremors, impaired attention, and aphasia, with potential progression to confusion, agitation, global aphasia, seizures, and cerebral edema. Headache, although common, is an early finding that lacks specificity. The onset of ICANS varies; it can arise either alongside CRS or days after its resolution [36, 40]. cilta-cel has also been associated with a risk for non-ICANS neurotoxicity including cranial nerve palsy or MNTs, comprising a cluster of movement, cognitive, and personality changes similar to parkinsonism [37].

In the pivotal KarMMa trial evaluating ide-cel in RRMM, 18% of participants (n = 23) experienced ICANS, with median time to onset of 2 days (range, 1–10), with 3% incidence of grade 3 or higher ICANS [15]. Similarly, in the KarMMa-3 trial of earlier line ide-cel, 15% (n = 34) experienced ICANS including 3% with grade 3 or higher severity (Table 2) [17]. In the pivotal CARTITUDE-1 trial that led to approval of cilta-cel in RRMM, 21% (n = 28) of patients experienced neurotoxicity, including 17% (n = 16) with any grade ICANS and 2% with grade 3 or higher ICANS, while 12.4% (n = 12) had non-ICANS neurotoxicity. ICANS and other neurotoxicities were not mutually exclusive, as 8.2% of patients experienced both simultaneously. The median time to onset of ICANS was 8 days (range, 3–12), and 15/16 had concurrent CRS. On other hand, median time to non-ICANS neurotoxicity was 26.5 days (range 11–108) from cilta-cel infusion. The subset (n = 7) with atypical neurotoxicity had facial paralysis, diplopia, cranial nerve palsy, ataxia, peripheral motor, and sensory neuropathy. The remaining 5 had movement and MNTs, with different severity (grade 2 [n = 1], grade 3 [n = 3], fatal grade 5 [n = 1]) [19, 20, 37]. MNTs typically emerged after a delayed period following CAR-T therapy. The spectrum of MNTs spanned from mild, temporary symptoms to enduring impairments, significantly impacting the patient’s quality of life.

Risk factors for MNTs were determined to be high tumor burden, grade 2+ CRS, any grade ICANS, or high CAR-T expansion/persistence [37]. Therefore, mitigation strategies have been implemented in subsequent trials of cilta-cel to try to prevent or minimize MNTs. Strategies include using more aggressive cytoreductive bridging therapy to reduce baseline tumor burden, and early and aggressive treatment of CRS and ICANS, which appears to have reduced rates of neurocognitive MNTs discernably to 1% or lower with these precautions in the CARTITUDE-4 study, though overall non-ICANS neurotoxicity was still seen in 17% including 16 (9.1%) with cranial nerve palsy, and 5 (2.8%) with CAR-T-related peripheral neuropathy [21, 37, 41].

Although the majority of non-ICANS neurotoxicities have been documented after cilta-cel, it is worth mentioning that FDA safety reporting has also identified rare instances of grade 3 parkinsonism following treatment with ide-cel [42]. However, notably no cases of MNTs were reported in the KarMMa or KarMMa-3 trials [15, 17]. It is postulated that the 2 BCMA-binding domains of cilta-cel may be associated with stronger binding to the off-tumor BCMA as compared to the 1 binding domain of ide-cel.

The management of ICANS is tailored to its severity and the presence or absence of CRS. A neurology consultation, including a fundoscopic exam, is recommended. Diagnostic measures may include an MRI of the brain, lumbar puncture, and MRI of the spine if deficits are present; alternatively, a CT scan can be used. A daily 30-min EEG is advised until symptoms resolve [43]. While corticosteroids play a pivotal role, the optimal timing, dose, and duration are yet to be defined. Corticosteroids cross the BBB and are recommended for the initial treatment of ICANS, with a high initial dose and rapid tapering according to the speed of response. Tocilizumab is recommended for patients with concurrent CRS; however, its efficacy in isolated ICANS is hindered by its limited penetration of the BBB due to its large molecular size. Additionally, IL-6 levels increase transiently after tocilizumab [43, 44].

In refractory and severe cases of ICANS (especially grade ≥3), various pharmacological interventions have been suggested for treatment. However, their roles are primarily based on single institution reports, and their overall efficacy remains uncertain [45]. Evidence from a retrospective study [32] supports the efficacy and safety of anakinra in treating both CRS and ICANS, particularly in severe cases. A phase 1 study assessing anakinra’s prophylactic effects in patients with large B-cell lymphoma who received CD-19-directed CAR-T revealed no impact on T-cell expansion kinetics. However, it significantly reduced proinflammatory cytokine levels, corticosteroid treatment duration, and overall incidence of ICANS, while maintaining similar rates of high-grade ICANS [26]. Ongoing clinical trials are further exploring the prophylactic role of anakinra [45]. The efficacy of alternatives like siltuximab, ruxolitinib, cyclophosphamide, intravenous immune globulin, and antithymocyte globulin in treating ICANS remains unclear, with evidence currently limited to case reports [45].

Early identification of patients at risk of developing severe neurotoxicity is critical. In the context of RRMM, various groups have investigated whether baseline laboratory values, including inflammatory biomarkers, are correlated with outcomes after CAR-T therapy. For instance, a single retrospective study revealed an association between a higher risk of ICANS and a higher severity of ICANS with an elevated baseline absolute lymphocyte count. This study did not identify an association between absolute lymphocyte count levels or any tested inflammatory biomarkers and delayed neurotoxicity [46]. Another study demonstrated an association between a higher CAR-HEMATOTOX score before lymphodepleting chemotherapy and grade 3 or higher ICANS [47]. The CAR-HEMATOTOX score is comprised of markers linked to hematopoietic reserve including platelet count, hemoglobin, and absolute neutrophil count, as well as baseline inflammatory markers including C-reactive protein and ferritin. It is important to note that further, larger studies are needed to validate these findings.

Poor count recovery especially in more than 1 cell lineage can lead to increased risk of infections, hemorrhages, increased resource utilization, and decreased quality of life [6]. In general, an increased risk of immune cell-associated hematotoxicity (ICAHT) is associated with factors such as an underlying disease burden, baseline decreased blood counts, bone marrow infiltration, number of prior lines of therapy (including bridging therapy), increased serum C-reactive protein, and ferritin [48]. Although early cytopenias are frequently linked to LDC, it is important to consider immune effector cell-associated HLH-like syndrome (IEC-HS) if blood counts are steadily dropping along with rapidly rising ferritin [49]. For those with prolonged thrombocytopenia and anemia in the setting of good neutrophil recovery, transplant-associated thrombotic microangiopathy must be considered [49]. The risk factors for developing prolonged (30–90 days after infusion) ICAHT include high tumor burden prior to CAR-T, preexisting cytopenias, and high-grade CRS/ICANS [6]. Persistent ICAHT that lasts >90 days post-infusion is mostly seen in the neutrophil and platelet lineage and might warrant a bone marrow biopsy to assess primary disease relapse, secondary marrow neoplasm, or large granular lymphocytosis [48]. During the early cytopenia phase, watchful observation with transfusion support is the mainstay of treatment. Leuko-reduced and irradiated blood products are used for transfusion support [50]. Since severe infection can adversely affect morbidity of CAR-T patients, adequate antimicrobial prophylaxis is warranted.

The use of granulocyte colony-stimulating factor (G-CSF) is controversial in the early post-CAR-T period due to the theoretical risk of worsening CRS/ICANS but is generally felt to be safe as long as the initial CRS has resolved. In the case of ide-cel, it can often be given by days 5–7 or at least by day 14 as the median time to onset of CRS was 1 day (range 1–14 days) with a median duration of 3.5 days (1–51 days) [17]. As for cilta-cel, the median onset of CRS is 8 days (range 1–23 days), with a median duration of 3 days (range 1–17 days), so the use of G-CSF should be delayed until closer to day 14 or after resolution of CRS symptoms [21]. Granulocyte-macrophage colony-stimulating factor should be completely avoided as it may promote inflammatory toxicity and induce neuroinflammation after CAR-T therapy [48]. Prophylactic or early G-CSF administration is generally not recommended in MM CAR-T due to insufficient data. A single-center retrospective involving 47 MM patients revealed that administering prophylactic G-CSF within 2 days of CAR-T infusion did not influence the overall median duration of neutropenia or the infection rates. Surprisingly, the median duration of neutropenia was longer in the prophylactic G-CSF group compared to the nonprophylactic group (6 days vs. 3 days, p value = 0.01). However, it is noteworthy that the early G-CSF group had a significantly shorter duration of antibiotic treatment [51]. Performing the CAR-HEMATOTOX scoring before LDC may guide identification of patients who would likely benefit from early and/or prophylactic G-CSF due to their high scores [47]. Thrombopoietin agonists such as eltrombopag, avatrombopag, and romiplostim can be considered in patients with prolonged thrombocytopenia that might present in the 2nd month post CAR-T therapy. Thrombopoietin agents have been known to work in aplastic anemia and immune thrombocytopenia and can have an impact on all 3 cell lineages through an anti-inflammatory effect [52]. Typically, these therapies are continued for up to 100 days following CAR-T therapy or until transfusion independence.

In cases of severe or prolonged cytopenias despite growth factor and transfusion support, considering a stem cell boost utilizing autologous cryopreserved CD34+ peripheral blood stem cells may also be an option, particularly in myeloma patients where this is more accessible from previous stem cell collection and storage. Multiple studies have validated the advantages of CD34+ stem cell boosts in addressing delayed and persistent ICAHT. Additionally, this boost facilitates the reconstitution of the bone marrow environment, aiding hematopoietic recovery within a span of 2–3 weeks [53].

Infections have always been a risk in myeloma patients due to underlying disease and low immune response to infections and vaccinations. The most common infections are typically bacterial (e.g., pneumonia and sepsis) and viral (e.g., HSV, influenza) [54]. However, fungal infections, though infrequent, are still a possibility [55, 56]. While previous data had shown that older patients are at higher risk for infections due to lower hematopoietic cell reserve, more recent data do not support this claim [57]. A multicenter retrospective study highlighted that earlier infections tend to be more severe and bacterial, whereas those occurring between days 31 and 100 after treatment are generally less severe and could be either bacterial or viral [34]. Within a study involving 52 patients who received standard-of-care ide-cel, infections were most frequent between 8 weeks and 6 months post-infusion. The only significant risk factor for infection was a longer duration between last bridging therapy and LDC, but this factor did not correspond to infection severity. The median time for the first infection to occur was 3.9 months, despite robust prophylaxis against opportunistic infections. Factors such as CRS, ICANS, prior lines of therapy, previous infections, or the use of alkylating agents did not significantly increase the risk of infections in these studies. However, they remain relevant considerations for infection prophylaxis and IgG replacement [58].

A separate retrospective study conducted at UCSF involving 55 patients revealed that the majority of infections took place within the initial 100 days following CAR-T treatment, primarily categorized as grade 2 requiring oral therapy [59]. In another retrospective study across two institutions assessing early (days 0–100) and late (days 101–365) infections, respiratory infections were identified as the most frequent site of infection in the first year after CAR-T infusion, with a prevalence of 31% for early events and 77% for late events. Hypogammaglobulinemia and early respiratory infections emerged as independent risk factors for late respiratory infections [60]. Nevertheless, to prevent infection-related adverse effects, antimicrobial (viral, bacterial, fungal) prophylaxis must begin on the day of CAR-T infusion. Antiviral prophylaxis should be continued for at least 1 year [61, 62]. Pneumocystis jiroveci pneumonia prophylaxis should be initiated around day +30 and continued for at least 6 months and up to recovery of CD4 count. Antibacterial prophylaxis should be continued during early neutropenia and should be considered during recurrent and late neutropenia. Antifungal prophylaxis should be continued until neutrophil recovery and might be extended for at least 1 month post prolonged corticosteroids used for CRS/ICANS management. Vaccination against influenza, coronavirus disease 2019, and pneumonia, among other posttransplant vaccines, should be initiated within 3–6 months post CAR-T therapy. ASH-ASTCT recommends repeating the complete primary coronavirus disease 2019 series with three mRNA vaccines starting at 12 weeks post CAR-T therapy [63].

If there is suspicion of an infection following CAR-T treatment, start empiric broad-spectrum antibiotics and conduct a comprehensive infectious workup, including but not limited to a complete blood count, chest X-ray, blood cultures, and CT scans of the chest/abdomen, along with urine analysis. Antibiotics can be de-escalated based on culture results. Additionally, consider performing cytomegalovirus polymerase chain reaction monitoring and especially testing in patients exhibiting unexplained fever, cytopenias, pneumonitis, hepatitis, or colitis [50].

Hypogammaglobulinemia develops in BCMA-directed CAR-T therapy due to the depletion of both normal and abnormal plasma cells, stemming from prior treatments and the CAR-T therapy itself, leading to plasma cell aplasia [6]. Prophylactic use of intravenous IgG replacement remains controversial. IgG replacement data come largely from our experience with CD19-directed CAR-T therapies, which recommends screening for serum IgG prior to, and in the first 3 months post CAR-T therapy [64]. Repletion in patients who have serum IgG less than 400 mg/dL or have serum IgG between 400 and 600 mg/dL and history of recurrent infections during the first 3 months post CAR-T therapy is reasonable [50]. Following the initial 3 months, it is advisable to monitor patients diligently and contemplate discontinuing IgG replacement, especially for those without additional infection risk factors like prolonged neutropenia or delayed CD4 recovery. For individuals encountering severe hypogammaglobulinemia and dealing with serious, persistent, or recurrent bacterial infections, especially in the sinopulmonary tract, continuing IgG repletion is recommended [64]. This is crucial as hypogammaglobulinemia often persists for 9–12 months post CAR-T therapy and may last much longer [59].

Formerly known as secondary hemophagocytic lymphohistiocytosis/macrophage activation syndrome, IEC-HS manifests as a severe hyperinflammatory state triggered by activated T cells and macrophages, initiating a cytokine storm leading to systemic toxicities and multiorgan failure [65]. Recent ASTCT consensus identifies IEC-HS as an independent hyperinflammatory syndrome, resembling macrophage activation/HLH, linked to immune effector cellular therapy. It displays signs of cytopenias, hyperferritinemia, coagulopathy with low fibrinogen, and/or transaminitis. Early recognition is a key. Diagnosis is established after excluding alternative causes such as CRS, infection, or disease progression. Typically, onset occurs during resolving or resolved CRS or when the inflammatory response worsens post-initial improvement with CRS-directed therapy. While many IEC-HS cases align with preceding CRS, this correlation is not always present [66]. The initial management involves supportive care and addressing the trigger, primarily infection, alongside suppressing the inflammatory response using corticosteroids with consideration for combining it with early use of anakinra [6, 67]. If there’s inadequate response, ruxolitinib may be an effective next line of therapy [68]. Subsequently, immunosuppressive drugs like etoposide, cyclosporine, methotrexate, or cyclophosphamide are considered [66, 69]. However, as controlled clinical trials are lacking, evidence mainly derives from case series, leading treatment decisions to rely significantly on clinical expertise and expert opinions.

MM patients face an increased risk of coagulation dysfunction with CAR-T, possibly due to an imbalanced coagulation environment, platelet dysfunction, disrupted fibrinolysis, and fibrin polymerization. Coagulopathies can present as prolonged prothrombin time, prolonged activated partial thromboplastin time, prolonged thrombin time, decreased fibrinogen, or elevated D-dimer. Changes in coagulation factors typically occur around 6–8 days posttreatment, with fibrinogen reaching its lowest levels around day 12 [70]. A positive correlation exists between bone marrow plasma cell count, baseline thrombocytopenia, and prior lines of therapy with CAR-T-related coagulopathies. Higher CRS grades, leading to increased IL-6, IL-10, and tissue factor levels, are also linked to coagulopathies, likely due to impaired coagulation protein production and endothelial damage [71]. Early coagulopathy management may help prevent CRS-related AEs and mortality. Coagulopathy should be treated to prevent progression to disseminated intravascular coagulation. Management includes controlling CRS, anticoagulation with low-molecular-weight heparin, and replacement with cryoprecipitate or fresh-frozen plasma [72]. Note that delayed coagulopathy may indicate IEC-HS, formerly known as HLH/MAS-like toxicity [14].

In one single-center retrospective study [73] evaluating cardiac toxicity following ide-cel administration, 14% (11 out of 78) experienced cardiac events, encompassing heart failure/ventricular dysfunction and arrhythmias. Notably, preexisting cardiac conditions did not appear to significantly influence the risk of these events, albeit the assessment was limited due to the small cohort and a low number of occurrences. There were no discernible differences in baseline cardiac assessments between individuals who experienced cardiac events and those who did not. Identified risk factors encompassed grade ≥2 CRS, being female, of Black race, having light-chain disease with lambda subtype, an ECOG >1, elevated baseline BNP, and ferritin levels. Most of these events were characterized by atrial and tachyarrhythmias, as well as heart failure. Interestingly, these events manifested within the initial 2 weeks post-CAR-T infusion, either concurrently or within days following CRS. Overall, the etiology of cardiac toxicities is related to the inflammatory state triggered by elevated cytokine levels [74], and higher grades of CRS correlated with a higher risk of any type of cardiovascular toxicity [75]. These CRS-related cardiac toxicities tended to be short-lived and reversible [76]. Within a retrospective study, no acute coronary syndrome events were observed [73]. Overall, the frequency and underlying mechanism of cardiotoxicity align with previous reports and observations from earlier CAR-T studies targeting CD19 [73, 77‒81]. Although no predictive model is currently established, future studies are needed to define the role of cardiac biomarker monitoring and various imaging modalities in patients undergoing CAR-T treatment [82].

While uncommon, other organ toxicity, including pulmonary, renal, liver, and gastrointestinal (GI), mostly occurs concurrently with CRS. The overall safety profile was reassuring, as indicated by reports from the pivotal trial and meta-analysis data. A meta-analysis that analyzed individual case safety reports from the World Health Organization database, along with data from CD-19 CAR-T trials and cohorts in the literature, revealed respiratory failure at 9.0%, nephrotoxicity at 6.0%, and hepatotoxicity at 1.5% [83].

It is known that CRS can cause hypoxia, pulmonary edema, and pneumonitis [84]. In the pivotal KarMMa trial, cough was observed in 20% of patients [15]. Similarly, in the CARTITUDE-1 trial, cough was noted in 24% of patients [19]. Furthermore, in the same trial, two fatal AEs were reported, including lung abscess and respiratory failure [19].

Acute kidney insufficiency (AKI) and electrolyte imbalances are frequent occurrences post-CAR-T therapy and often reversible. There have not been any studies to date looking specifically at BCMA-directed CAR-T-induced AKI; however, based on studies involving CD19-directed CAR-T, it is known that the incidence of AKI is relatively low and not severe [85].

In the pivotal KarMMa trial, mild GI issues like nausea, constipation, or diarrhea, and elevated aspartate aminotransferase were prevalent [15]. Severe grade 3 or 4 diarrhea was reported in only 2%, and one fatality occurred due to GI bleeding [15]. In the CARTITIDE-1 trial, diarrhea was observed in 20% of patients, with 1% experiencing grade 3 or 4 [49].

Overall, late organ toxicity is not a common complication that necessitates long-term management. Most events tend to occur during the acute inflammatory response in the initial days following cell infusion. Ensuring selective candidacy and implementing early diagnosis and management of CRS remain crucial in mitigating morbidity and mortality.

Despite the significant therapeutic impact observed with commercially available CAR-Ts, due to limited patient numbers in studies and short follow-up at FDA approval, long-term outcome data remain undetermined. The FDA requires a long-term follow-up of 15 years for recipients of commercialized genetically engineered cellular therapies to evaluate their safety and efficacy. To address this, the Center for International Blood and Marrow Transplant Research (CIBMTR), in collaboration with pharmaceutical companies, is tracking the long-term safety and outcomes of patients treated with these CAR-T therapies, enabling the capture of long-term data. This registry monitors various CAR-T toxicities including secondary malignancies.

CAR-T therapy theoretically poses a risk for hematologic malignancies like myelodysplastic syndrome (MDS) due to potential adverse gene integration events or due to LDC in the setting of many years of prior therapies [86]. Hematologic secondary malignancies occurred in nine CARTITUDE-1 trial patients who received cilta-cel, comprising one low-grade B-cell lymphoma, six MDS cases, and three acute myeloid leukemia (AML) cases (1 patient with both MDS and AML) [19]. The KarMMa study reported 1 case of MDS as well [87].

A single-center study involving 4 cases of MDS post-CAR-T therapy revealed that all patients had MDS-related genetic changes before CAR-T, with no morphological signs in bone marrow before CAR-T. None developed new mutations post-CAR-T, but the frequency of existing mutations did rise. This suggests that CAR-T therapy might spur the expansion of existing pre-MDS clones rather than instigating new ones [88]. This aligns with prior reports, indicating increased leukemia and MDS rates in MM patients and its precursor, monoclonal gammopathy of undetermined significance [89].

In a recent real-world CIBMTR report studying 603 patients treated with ide-cel, over a median follow-up of 6.6 months, second malignancies occurred in 4.5% (27 patients), including 5 individuals with AML/MDS [90]. The FDA is currently investigating the potential risk of T-cell malignancy post CAR-T therapy. This inquiry stems from a case report detailing CAR-positive T-cell lymphoma emerging after cilta-cel infusion within the CARTITUDE-4 trial. The involvement of CAR-T in this development remains unclear, prompting a need for further investigation into the etiology of this T-cell lymphoma [91]; however, even if related, the incidence is extremely low.

While the occurrence of secondary cancers is infrequent, long-term survivors of CAR-T therapy are advised to undergo vigilant, age-appropriate cancer screenings. Monitoring for deteriorating cytopenias, especially concerning secondary malignancies, is recommended [92].

Understanding the effect of BCMA-directed CAR-T on fertility remains unclear, especially due to complicated factors like older age of most myeloma patients and previous exposure to gonadotoxic treatments such as chemotherapy or radiation [92]. Considering the potential impact of LDC before CAR-T infusion, experts advise discussing fertility preservation options before starting chemotherapy, when feasible and suitable [93]. Due to potential risks of teratogenicity following LDC, and theoretic transferability of CAR-T from mother to child through placenta or breast feeding, the optimal conception time post CAR-T is unknown. Therefore, it is advised to avoid pregnancy for 6–12 months post CAR-T therapy [94]. While specific details on BCMA-directed CAR-T and pregnancies are lacking, there have been documented successful pregnancies following other CAR-T therapies targeting different receptors [94].

There are no data available regarding late autoimmune disorders; however, theoretically, it is a possibility [95]. It has been known that damage-associated molecular patterns (DAMPs) are released in response to cell death and stress [96]. The potential role of tumor cell-specific DAMPs is not yet determined; however, DAMP-mediated sterile inflammation may be a component of many diseases, including autoimmune disorders [97]. Thus, until the availability of long-term data, close monitoring and prompt recognition of autoimmune disorder in this clinical setting are warranted.

In this review, we have summarized the risks and management of various CAR-T-related side effects following BCMA-directed CAR-T therapy in RRMM. As the field of CAR-T therapy progresses, future directions should focus on refining treatment protocols, enhancing patient outcomes, and managing potential long-term side effect. By administering CAR-T earlier in the disease course, before myeloma becomes refractory to all therapies, the options for effective tumor cytoreduction and therefore mitigation of CRS and ICANS appear greater, potentially minimizing comorbidities and prior toxicities. Given that CAR-T composition relies on circulating T-lymphocytes, monitoring T-cell kinetics is crucial. Preclinical research showed chemotherapy like doxorubicin, cytarabine, and cyclophosphamide can impair T-cell function due to damaged mitochondrial energy [98]. Clinical reports suggest a negative impact on apheresis for up to 6–9 months post alkylator treatment, indicating that a shorter washout may harm patient response [99]. Moreover, the role of bridging therapy evaluated in RWD, which showed poorer outcome with intensified/infusional cyclophosphamide combinations [100]. Therefore, understanding how prior treatments can influence CAR-T therapy outcomes and toxicity is essential.

Although CAR-T therapy for myeloma currently only includes BCMA-targeting agents, CAR-T products with novel targets and combinations of targets are under development. These developments may bring additional risks for toxicities specific to the tissue distribution of those antigens, such as skin and nail toxicity, as well as non-ICANS dizziness and cerebellar type toxicity (mostly at high doses of CAR-T) seen in early reports of GPRC5D-targeted CAR-T [101, 102]. Each target is likely to present a unique set of challenges.

In this advancing CAR-T landscape, one emphasis should be on exploring innovative strategies for mitigating and treating associated toxicities. This involves enhancing our understanding of underlying mechanisms and developing targeted interventions and personalized approaches to minimize adverse effects, thus increasing safety and allowing more widespread use of these therapies in the outpatient setting. Additionally, attention should be directed toward refining treatment protocols, identifying reliable biomarkers for predicting toxicity, and establishing robust long-term monitoring systems to ensure sustained safety and efficacy of CAR-T therapies. This comprehensive approach will contribute to continuous improvement and optimization of CAR-T therapy in the future.

A.A. reports research funding from AbbVie, Adaptive Biotech; and advisory board from Bristol-Myers Squibb, Janssen, and Karyopharm. G.K. reports consulting or advisory committees for Bristol Myers Squibb, Cellectar, Sanofi, Janssen, and Arcellx, and research funding from Bristol Myers Squibb, Janssen, and Arcellx. D.K.H. reports research funding from Bristol-Myers Squibb, Janssen, Karyopharm, International Myeloma Society Young Investigator Award, and the Pentecost Family Myeloma Research Center and Adaptive Biotech and consulting or advisory role for Bristol-Myers Squibb, Janssen, Karyopharm, and Pfizer, and member of the Bristol-Myers Squibb IMW Ide-Cel Academic Advisory Board, Bristol-Myers Squibb Multiple Myeloma ASH Steering Committee, and Multiple Myeloma Pfizer Advisory Board, and received net honoraria from OncLive and Survivorship. L.D.A. Jr. reports honoraria and membership on an entity’s Board of Directors or advisory committees for Bristol Myers Squibb, Celgene, GSK, AbbVie, Pharmacyclics, Karyopharm, Janssen, Prothena, Sanofi, Beigene, Cellectar, and Amgen. Remaining authors have no potential conflicts of interest.

None to report.

A.A., P.A., L.T., G.K., A.S., D.K.H., and L.D.A. Jr. performed literature research and drafted the manuscript. All authors provided review and edits and approved the final version of the manuscript. A.A. and L.D.A. Jr. supervised the study.