Abstract
Adoptive immunotherapy has recently garnered widespread interests owing to the successful application of chimeric antigen receptor T cell therapy. CAR-T cells are “living drugs” that can live in patients for several years and act as an effective antitumor agent. Over the last few years, five types of CAR-T cells have been approved by Food and Drug Administration (FDA) for treatment of hematologic malignancies. Despite their impressive clinical efficacy, the current application of CAR-T cell therapy is restricted by the uncontrollable release of cytokines (cytokine release syndrome and cytokine release syndrome) due to serious treatment-related toxicities resulting from synchronous activation and rapid proliferation of CAR-T cells. CRS is the most common toxicity and its severity can range from low-grade physical symptoms to a high-grade syndrome linked with life-threatening multiple organ dysfunction. Treatment-related deaths from severe CRS have been reported, suggesting the importance of appropriate intervention. Gaining a better understanding of CRS and developing new treatments for CRS are active areas of laboratory and clinical research. Herein, we summarize the current studies on prevention and management of CRS to expand the safety and applicability of CAR-T cell therapy in various malignancies.
Introduction
Summary of chimeric antigen receptor (kikCAR)-T cell therapy clinical trials for solid tumors or non-malignant condition.
Compared with traditional cancer therapies, CAR-T cells can be considered “living drugs,” which can last for several years in the body and undergo sequential expansion, contraction, and re-expansion in vivo after (re-) exposure to antigen if they continue to exist in large numbers and have active effector functions. 38 However, in the body, once CAR-T cells start working and activate the cytokine pathways, patients can suffer from serious or even life threatening side effects,39-40 such as cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS). 41 With the widespread application of CAR-T cell therapy, its unique toxicities should be understood. More than 90% of patients treated with CAR-T will develop CRS, and half of the patients will progress to grade 3–4 or even cause death.38,42-44 CRS displays flu-like symptoms with fevers, fatigue, and malaise. However, it can progress to life threatening hypotension, capillary leak, hypoxemia, and end organ dysfunction, which may require intensive care. Successful implementation of CAR-T therapy from the laboratory to the bedside requires an understanding of CRS. A number of recent seminal studies have made key insights in the management of CRS. Herein, we provide a comprehensive overview of pathophysiology and manifestations of CRS, and as well as the latest perspectives on CRS management in preclinical and clinical studies.
Pathophysiology and Manifestations of CRS
Pathophysiology of CRS
To prevent the occurrence of CRS and improve its treatment, its pathophysiological aspect should be understood. However, the mechanism of CRS remained elusive. When the patient is infused, CAR-T cells come into contact with target cells. Then, inflammatory cytokines, such as tumor necrosis factor (TNF)α and interferon (IFN) γ, are released, thereby activating monocytes and macrophages to secrete cytokines including IL-1 and IL-6. In the first few days prior to the development of CRS, IL-6 increased significantly. In addition, IL-6 plays an important role in the formation of characteristic symptoms of severe CRS, such as vascular leakage, activation of the complement and coagulation cascade, and induction of disseminated intravascular coagulation (DIC).45-46 This finding suggests that high levels of IL-6 have a central role in pathogenesis of CRS and may be predictive patients who will develop severe CRS. 47 However, other cytokine levels (including IL-1, IL-2, IFN-γ, GM-CSF, IL-5, IL-8, and IL-10) will also be increased and are closely associated with the severity of CRS.47-52 IL-1, which plays an important role in immune response and generation of inflammation and can activate mast cells, is another important cytokine in the pathophysiology of CRS. Recent studies have demonstrated that IL-1 is a desirable target in the pathophysiology of CRS, and blocking IL-1 receptor with anakinra was effective to protect mice from CRS.47-48
A feature of severe CRS appears to be the activation of endothelial cells. 53 The presence of hemodynamic instability, capillary leak, and a consumptive coagulopathy increased endothelial activation, thereby causing severe CRS. Von Willebrand factor (VWF) is released from the Weibel–Palade bodies after activation of endothelial and plays a key role in the initiation of coagulation. 54 Hay’s group found that patients with severe CRS after CAR-T infusion were accompanied by high concentrations of serum VWF and angiopoietin-2 (Ang-2). 55 In addition, a case study of a severe CRS patient showed that endothelial cells are the main source of IL-6, indicating a direct link between activated endothelial cells and overall production of IL-6. 56 The mechanism that leads to endothelial activation in CRS is still unclear, and further research is required.
Manifestations of CRS
Within the realm of CAR-T, CRS shows different time courses and severity according to the CAR-T structure, tumor type, and patient comorbidity. Fever is usually the first sign of CRS and develops before other symptoms and signs. 57 These fevers often start at a low temperature, but this escalates to levels as high as 105 F/40.5°C. 38 Patients usually experience the first signs of CRS within 1–14 days after CAR-T cell infusion, although some cases show delayed CRS. Following the initial fever, any organ can be affected, and the patient can experience skin rash, nausea, vomiting, respiratory failure, cardiac dysfunction, renal failure, disseminated intravascular, coagulopathy, hepatic dysfunction, and neurological deficit. 58 CRS duration was reported between two and 3 weeks after CAR-T cell infusion. 59
Respiratory symptoms are common in CRS patients. 53 Mild cases may manifest as cough and tachypnea, but can progress to acute respiratory distress syndrome (ARDS) with dyspnea, hypoxemia, and bilateral opacities on chest X-ray. In severe cases, patients may require mechanical ventilation due to respiratory failure. Patients may experience liver failure with elevated serum aspartate aminotransferase, alanine aminotransferase, and bilirubin levels; acute kidney injury with increased serum creatinine levels and decreased urine output; cardiac complications with arrhythmias, heart block, and low ejection fraction; hematologic toxicities with prolongation in the prothrombin time (PT), partial thromboplastin time (PTT) and decreased fibrinogen,39,41,60 and DIC.57-58 Although these end organ toxicities are almost always reversible, maintaining good end organ function is particularly important for patients with ALL for whom allogeneic stem cell transplant is planned once a remission is achieved, following CAR-T cell therapy.
Notably, fulminant hemophagocytic lymphohistiocytosis and/or macrophage activation syndrome (HLH/MAS) can develop as a part of the CRS. 60 HLH is an immune disorder that leads to excessive inflammation and abnormal activation of macrophages and T cells. In the context of CRS, diagnosing HLH/MAS can be difficult. The biological characteristics of severe CRS and HLH/MAS may have similar clinical and laboratory manifestations, including cytopenias, hepatosplenomegaly, coagulopathy with marked hypofibrinogenemia, and hyperferritinemia. 38 The pathology of CRS and HLH/MAS has an overlap, most patients with grade ≥2 CRS satisfy the published HLH/MAS consensus diagnostic criteria, and most patients with grade ≥3 CRS satisfy the definition of CAR-T cell-related HLH/MAS . 60
Clinically available laboratory markers of inflammation, including cytokine profiles (such as IL-6, IFN-γ, IL-8, IL-10, and IL-2) involved in CRS, correlate with the clinical syndrome of CRS.60-63 Studies have shown that IL-6 increases significantly in the first few days prior to CRS, and the peak level of IL-6 after CAR-T cell infusion is related to the CRS severity and ICU admission. 61 The IL-15, INF-γ, IL-10, IL-8, and granzyme B were significantly related to grade 3 or higher CRS. The dynamic change in cytokine levels can be used as a tool to predict the severity of upcoming CRS.38, 55, 61 In addition, C-reactive protein (CRP) and ferritin levels after CAR-T cell therapy are always elevated in patients with CRS.62-64 Clinical trials have shown that monitoring the changes of these cytokines after CAR-T cell infusion can predict the occurrence of CRS. 55 The elevation of CRP ≥20 mg/dl is associated with severe CRS, with a specificity of 100%. 63 Therefore, CRP is considered a valuable biomarker for judging the severity of CRS. 63 However, CRP is not specific to CRS and is elevated for infectious and non-infectious inflammatory causes.
Steps to prevent CAR-T-related CRS
Designing Safe CAR-T Cells
While the second-generation CAR generated considerable response rates, controlling the CAR-T cell’s activity is still a challenge. To improve the efficacy of CAR-T cells and/or control their toxicities, the design of the compound should be optimized.65-70 With the application of tertiary structure prediction program (Phrye2), Chen’s group provided a CAR designed with an optimal length of the hinge and intracellular sequences and generated a new anti-CD19 CAR molecule (CD19-BBz(86)) derived from the CD19-BBz prototype with co-stimulatory 4-1BB and CD3ζ domains. 67 Furthermore, they demonstrated that this new CD19-BBz(86) CAR-T cells express long-lasting antitumor ability and do not cause severe CRS or neurotoxicity,47,67 representing a safe and effective anti-CD19 CAR-T cell therapy. CRISPR/Cas9 is an RNA-based bacterial defense mechanism that requires the transient expression of a nuclease and guides RNA to eliminate foreign DNA at the targeted site. 68 Taking advantage of this tool, Sterner used CRISPR/Cas9-mediated knockout of GM–CSF and showed that the knockout of GM–CSF CAR-T cells produces less GM–CSF while maintaining critical T cell function and effectively enhancing antitumor activity in vivo compared with wild-type CAR-T cells. 69 Eyquem’s group developed uniform CAR expression and enhanced the potency of CAR-T cells with the CRISPR platform. 70
One limitation of the currently used CAR-T cells is the lack of controlled regulation post-dose. Especially when life-threatening side effects occur, no emergency shutdown of CAR-T cells can be performed. Strategies are currently under investigation, including forming bridges (e.g., a low molecular weight adapter, EC17adapter molecule, and switch molecules) between the CAR-T cells and its cancer cell targets.71-74 Yong found that the appearance and intensity of a CRS-like toxicity in mice can be adjusted by controlling the concentration and dose schedule of the adapter that forms a bridge between CAR-T cells and its cancer cell targets. 71 Ma et al. designed and synthesized a “switch” molecule in which anti-CD19 and anti-CD22 antibody fragments are site-specifically modified with FITC using genetically encoded noncanonical amino acids. This method has the precise control over the geometry and stoichiometry of complex formation between CD19−or CD22-expressing cancer cells and a “universal” anti-FITC-directed CAR-T cell. The results showed that this program could improve the safety of CAR-T cells while maintaining the antitumor potential. 74
Advanced insights gained from gene editing technology have generated new forms of interventions and engineering techniques that are expected to break through the toxicity associated with CAR-T cell therapy, such as using synthetic T cell engineering, including SUPRA CAR-T cells, synNotch receptors, and UniCAR, to improve the efficacy and feasibility of off-the-shelf products. 75 Assuming a similar effect can be achieved in humans, the risk of lethal CRS is expected to be reduced and/or eliminated.
Dose Refinement of CAR-T cell
Finding the balance between effectiveness and the potential for CRS appears to be an ideal approach. Clinical studies have demonstrated that adjusting the dose of CAR-T cell therapy may improve its efficacy and decrease the associated toxicities 76-78 Frey described that split-dosing or fractionation of CAR-T cells is effective and less toxic with CD19-positive relapse or refractory ALL. 76 In addition, Park explored dose modifications and outcomes according to baseline disease burden in patients with B-cell ALL. In a phase I clinical trial that included 51 adults with relapsed/refractory B-cell ALL, the patients were categorized and given different doses of CAR-T cells according to their baseline disease burden. The results showed that complete remission was higher and the CRS was milder in patents who adjusted for CAR-T dose according to their baseline disease burden. 77 Previous reports showed that higher doses of CAR-T cells (3×106 CAR-transduced T cells per kg, 2×107 CAR-transduced T cells per kg) were closely related to the occurrence and development of severe CRS.61,79
Higher malignant burden may cause greater antigen stimulation, resulting in sufficient CAR-T cell proliferation to induce remission. The risk-adapted CAR-T cell dose refers to the lower cell dose given to patients with higher disease burden, which may ameliorate toxicity without affecting the therapeutic effect.2,78 In a single-arm cohort study involving 42 patients with refractory and/or relapsed B-cell lymphoblastic leukemia (B-ALL), Pan and colleagues observed that dose reduction (1x105/kg) of CAR-T cells has a high response rate (90%), and only mild to moderate CRS occurred. 78 This finding suggested that lower cell doses given to patients may ameliorate toxicity.
The therapeutic effect of CAR-T cells and the safe dosage should be considered to find a balance between effectiveness and CRS. Thus, adjusting the dosage of CAR-T cells seems to be an effective method. Although the optimal dose of CAR-T cells may be affected by tumor load, baseline disease load, or other factors, multiple clinical trials are required to determine the therapeutic window of CAR-T cell dose to obtain maximal efficacy with minimal toxicity.
GM–CSF antagonism
Multiple clinical trials have been conducted against inflammatory diseases by inhibiting GM–CSF signaling.80-81 As a pro-inflammatory cytokine, GM–CSF is specifically upregulated in a CAR-dependent manner. This condition has led to speculation that GM–CSF may play a key role in monocyte activation, resulting in CRS association with CAR-T cell therapy. Sentman injected C57BL/6 mice with NKG2D CAR-T cells extracted from GM–CSF-deficient wild-type mice. The result showed that C57BL/6 mice had lower amount of cytokines in the serum, consistent with those without evident CRS symptoms. 82 Sachdeva found that monocyte activation occurs specifically via the GM–CSF/GM–CSF Ra receptor axis with a cytokine profiling assay. 83 This finding suggests that GM–CSF from CAR-T cells is necessary for inducing CRS-like responses. Stermer’s group also demonstrated that knocking out GM–CSF during the CAR-T manufacturing process can prevent CRS symptoms. 84 These studies show that GM–CSF genetic inactivation does not impair the antitumor function or proliferative capacity of CAR-T cells in vitro.
Pharmacologic management of CRS
GM-CSF neutralization
In addition to the use of genetic engineering to prevent GM-CSF release, the use of GM-CSF neutralization can also reduce GM-CSF. Sachdeva’s group showed that inactivation of GM-CSF in CAR-T cells by antibody-mediated neutralization or TALEN-mediated genetic modification (TRAC and GMCSF TALEN mRNAs) could evidently decrease the availability of GM–CSF and eliminate macrophage-dependent secretion of CRS biomarkers, including monocyte chemoattractant protein 1 (MCP-1), IL-6, and IL-8. 83 Stermer’s group demonstrated that GM-CSF neutralization with lenzilumab not only maintained CAR-T19 cell function, but also enhanced CAR-T cell proliferation. Moreover, in a patient acute lymphoblastic leukemia xenograft model of CRS, GM-CSF neutralization could prevention of CRS. 84
These studies describe that neutralization of antibody against GM–CSF is a therapeutic approach to prevent CRS. However, many preclinical and clinical trials are required to verify this approach.
Tyrosine kinase inhibitor
In addition to these superior safety strategies, other methods may reduce the occurrence of CRS. Dasatinib, a member of tyrosine kinase inhibitor, is an FDA-approved tyrosine kinase inhibitor for the treatment of t (9; 22) chronic myelogenous leukemia (CML) and Philadelphia chromosome positive ALL. It inhibits T cell activation by obstructing proximal T-cell receptor (TCR) signaling kinases, such as Src, Fym, and Lck.85-88 Researchers hypothesized that dasatinib could inhibit the activation and function of CAR-T cells.89-90 Weber et al. showed that dasatinib could rapidly and reversibly inhibit antigen-induced activation, proliferation, and cytokine production, thereby eliminating the antitumor activity of CAR-T cells and the production of inflammatory cytokines. 91 Mestemann found that dasatinib induces functional inactivation of CD8+ and CD4+ CAR-T cells. The dose of dasatinib can be titrated to achieve partial or complete inhibition of CAR-T cell function. When dasatinib is terminated, the inhibitory effect was immediately and completely reversed, and CAR-T cells restored their antitumor function. 92 These results indicate that dasatinib can be used as a reversible safety switch to inhibit CAR-T cell function in the event of severe or life-threatening CAR-mediated toxicity. Such strategy will likely result in a reduction of CRS and directly control the activity and function of CAR-T cells. Clinical studies are required to examine the effect of dasatinib on the connection of toxicities to CAR-T cell therapeutics.
IL-6 receptor antagonism
IL-6 is an important mediator of the CRS signaling cascade. Substantial evidence has shown that anticytokine therapy targeting IL-6 can effectively ameliorate CRS symptoms. Tocilizumab is a humanized monoclonal antibody (mAb) against the IL-6 receptor (IL-6R). Published studies have shown that tocilizumab could be widely and effectively used in advanced cases, such as CRS of grade 3 or greater and grade 2 CRS with comorbidities; and does have any negative effects on the expansion or persistence of CAR-T cells.57,63 On August 30, 2017, the FDA approved tocilizumab for the treatment of this indication. 93 The recommended dose of tocilizumab for patients weighing <30 kg is 12 mg/kg and 8 mg/kg for patients weighing ≥30 kg. If this dose does not improve the clinical symptoms within 24 h or the patient’s symptoms deteriorate rapidly, then the FDA has approved dosing strategy includes the ability to administer up to three additional doses. 94 The optimal timing and repeated doses of tocilizumab to balance safety and efficacy are unknown because the CRS may be influenced by types of CAR-T cells, receptor target, baseline disease burden, patient age, or other parameters. Whether preventive or early tocilizumab intervention can improve safety without compromising efficacy is still unclear. Thus, for elderly patients and patients with comorbidities, tocilizumab may be administered earlier to avoid unfavorable results. 57 For severe CRS patients, close monitoring and aggressive supportive care should be given in ICU, and repeated doses of tocilizumab should be given when no improvement is observed. 95
Currently, several other IL-6-targeting mAb are introduced in the clinical development stage that may be used in the management of CRS. Siltuximab, which is currently approved by the US FDA for Multicentric Castleman’s disease, 96 is an antibody against IL-6 and binds to human IL-6 with high affinity and can rapidly reverse CRS symptoms. 94 It exhibits a theoretical advantage of blocking the activity of IL-6 more completely than tocilizumab. Neelapu suggested that tocilizumab and siltuximab could be interchangeable in the initial control of CRS. 60 However, data to evaluate the effectiveness of siltuximab in the treatment of CRS are limited, and well-designed clinical trials are required to evaluate the effectiveness of tocilizumab and siltuximab in the treatment of CRS.
Corticosteroid
Corticosteroids, especially systemic adrenal glucocorticoids, have the ability to suppress inflammatory responses and immune system homeostasis in several days and will not cause recurrence of CRS.57, 97 Currently, the recommended dose for dexamethasone is 10 mg intravenous (IV) every 6 h or methylprednisolone 1 mg/kg IV every 12 h in instances of moderate to high-grade CRS. 98 For children, the start recommended dose of methylprednisolone is 1–2 mg/kg per day. 99 However, studies have shown that corticosteroids had detrimental effects on the immunotherapy; they may influence CAR-T expansion and antitumor effect.63, 99 A correlation is found between prolonged corticosteroid administration time and decreased circulating CAR-T cells and increased risk of treatment failure. 63 Moreover, corticosteroids may have unpredictable consequences in patients with cancer. Arbour’s group recommended that corticosteroids should be used cautiously at the time of initiating PD-(L)1 blockade. They found that corticosteroids is connected with poorer prognosis in patients with non-small-cell lung cancer treated with PD-(L)1 blockade. 100 Obradović indicated that the activation of the glucocorticoid receptor promotes breast cancer metastasis and reduces survival. 101 Despite these concerns, glucocorticoid still has an important value in the management of moderate to severe CRS.
IL-1 receptor antagonism
The establishment of mouse models that accurately replicated CRS showed that IL-1 has a significant effect on the development of CRS and is a potential target of toxicity.47-48 IL-1β, a member of the IL-1 family, plays an indirect role in lymphocyte-mediated immunity. In inflammatory disease, by inhibiting the expression of IL-1, inflammatory molecules can be significantly reduced and inflammatory symptoms are alleviated. In addition, in lymphocyte-mediated immunity disease, IL-1 participates in the production of IL-6 and induces the secretion of IL-6 as well as soluble IL-6R (sIL-6R). IL-6 release is also inhibited by blocking IL-1β. 102 These findings suggest that IL-1 is suitable to CRS treatment. 48 When the recombinant IL-1 receptor antagonist anakinra is used in combination with corticosteroids, rapid and sustained remission of systemic juvenile idiopathic arthritis-associated MAS is induced.103–104 Giavridis generated CAR-T cells that could constitutively produce IL-1 receptor antagonist, and found that this novel construct could autonomously prevent CRS-associated mortality in mouse model of CRS. 48
Although anakinra has not been used in the management of CRS in clinical trial, and IL-1β is a fundamental arm of innate immunity, the administration of immunosuppressant during severe CRS may have immunological consequences in patients with acute viral infection; it represents a promising treatment for CRS. Once clinically confirmed, a significant reduction in CRS is expected in the near future.
Conclusion
CRS is one of the most common serious adverse events and the main cause of morbidity after CAR-T cell immunotherapy. Gaining insights from studying the biological mechanism has improved the management of patients with CRS. Lastly, advances in CAR-T bioengineering may allow rapid reversal or complete prevention of CRS in the future. These technologies include(1) constructing transient CAR constructs, such as “biodegradable” mRNA-encoded CARs transferred by electroporation; (2) constructing CAR-T cells with safety or suicide switches; and (3) looking for alternative to CAR-T cells, such as CAR-natural killer (CAR-NK) cells, which may less toxics than CAR-T cells. While the generation of CRS depends on several intrinsic and extrinsic factors, multi-institutional and multi-disciplinary efforts are imperative to generate effective strategies to avoid CRS.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
