Because CRS cannot be decisively differentiated from infection, most centers administer antibiotics for cases of neutropenic fever. However, the use of growth factors should be restricted for the first few weeks. As immune-based cancer therapies become more potent, effective, and more widely available, optimal management of their unique toxic effects becomes increasingly important. Cytokine release syndrome (CRS) is a potentially fatal toxicity that has been observed after the administration of natural and bispecific antibodies and, more recently, after the adoption of T-cell therapies for cancer.
CRS is associated with high circulating levels of several cytokines, such as interleukin (IL) -6 and interferon gamma, and uncontrolled studies show that immunosuppression with tocilizumab, an antibody against the IL-6 receptor, with or without corticosteroids, can reverse the syndrome. However, since early and aggressive immunosuppression may limit the efficacy of immunotherapy, current approaches seek to limit the administration of immunosuppressive therapy to patients at risk of suffering fatal consequences of the syndrome. This report presents a novel system for evaluating the severity of CRS in individual patients and a treatment algorithm for treating CRS based on severity. The goal of our approach is to maximize the chances of obtaining a therapeutic benefit from immunotherapy and, at the same time, to minimize the risk of potentially fatal complications.
of the CRS. The National Cancer Institute's common terminology criteria for adverse events (CTCAE v4.0) contain a classification system that was designed for CRS associated with antibody therapy (Supplementary Table 1, available on the Blood website). We modified this classification system to define mild, moderate, severe and potentially fatal CRS, regardless of the triggering agent, and to guide treatment recommendations (table). Since many patients with chronic rheumatoid arthritis have overlapping symptoms due to fever and neutropenia, an infection, tumor lysis syndrome, or other medical complications, careful consideration of attribution is critical.
This is particularly important since the treatment algorithm is largely based on the grading scheme. and incorporates immunosuppression. One could imagine, for example, a patient with hypotension related to sepsis in whom intervention with tocilizumab or corticosteroids would not be indicated (see below). Therefore, the precise application of this classification system requires clinical judgment to confirm that the symptoms are most likely due to CRS and not to another medical condition.
The clinical symptomatology comprising CRS is an indicator of response in the context of immune-based therapies. It is not clear to what extent the cytokines that mediate symptomatology are necessary to achieve anti-tumor effects. Therefore, the goal of treatment is not to extinguish all evidence of CRS, but to prevent potentially fatal toxicity and, at the same time, to maximize the potential for anti-tumor effects. For this reason, symptomatic treatment of grade 1 CRS (Figure 2, case) is recommended.
We define CRS as grade 2 when the patient develops hypotension that responds to fluids or a vasopressor at low doses or mild respiratory symptoms that respond to low oxygen flow (side effects attributed to tocilizumab in rheumatologic disease, in which the drug is administered chronically, include transaminitis, thrombocytopenia, high cholesterol and low density lipoproteins, 41 Neutropenia has also been infrequently attributed to treatment with tocilizumab, and this seems to resolve with interruption of treatment with the agent. The incidence of viral, bacterial, fungal and mycobacterial infections increases moderately in patients receiving chronic treatment for rheumatologic disease, which forms the basis for a black box warning associated with agent 41. We have not observed acute perfusion toxicities secondary to tocilizumab in patients treated in our studies. Modern anti-tumor immunotherapies hold great promise, but the effective application of this new class of therapies requires doctors to learn to recognize and manage the new toxicities associated with tumor immunotherapy. The primary goal of treating CRS in patients treated with immunotherapy for cancer is to prevent life-threatening CRS, while maintaining the greatest chance of obtaining a beneficial antitumor effect. In this article, we describe a proactive treatment strategy that incorporates a classification system and treatment algorithm designed to deliver early immunosuppression to patients at higher risk while avoiding unnecessary immunosuppression due to the potential risk of diminishing anti-tumor efficacy.
It is necessary to continue working to better understand the pathophysiology of this syndrome, to outline more precisely the aspects of immune activation necessary to obtain anti-tumor effects and to validate the strategies of optimal treatment. Given the limited clinical experience with the syndrome, the possibility of creating a national or international registry may be considered to monitor the results and obtain more information on the spectrum of symptoms and the optimal treatment of CRS. In addition, it is important to emphasize that the severity of CRS is greater in patients with a higher burden of disease. It is expected that the incorporation of immunotherapy into regimens that administer these treatments to patients with a lower burden of disease will considerably reduce the observed toxicity and, in the absence of clinical evidence on CRS, could benefit patients.
Figure 3 represents a hypothetical patient with grade 3 CRS who develops fever on day 3, then hypotension on day 5, which is initially treated with a low-dose vasopressor, and then tocilizumab on day 6 after cardiovascular decompensation. Patients with a high tumor burden and a primary refractory disease are at high risk of relapse in both CRS and lymphoma. All patients with CRS grade 3 should be treated with immunosuppressive agents because of the risk of progression and the possibility of irreversible organ dysfunction, with the objective of preventing progression to grade 4.For example, in DLBCL with CD19 CAR T cells, CAR T cell products co-stimulated with CD28 show higher rates of CRS and ICANS compared to products costimulated with 4-1BB. In some series, peak CRP levels and changes in CRP folds have identified patients at risk of severe CRS. Therefore, we conclude that real-time analysis of a large panel of cytokines will not significantly affect the treatment of individual patients with CRS at this time, and we recommend that treatment decisions are based on clinical parameters.
Although the diagnosis of CRS cannot be established or ruled out using laboratory diagnostics, they can be used to monitor organ dysfunction. Therefore, the possibility of contracting an infection should be considered in all patients who present symptoms of chronic rheumatoid arthritis after obtaining appropriate cultures and initiating empirical antibiotic treatment.
