For many decades, cancer therapy was based mainly on surgery, chemotherapy and radiotherapy. In recent years, therapeutic approaches based on stimulating the patient's immune response have established cancer immunotherapies as new treatment options.1 Immunotherapy with chimeric antigen receptor T (CAR-T) cells, is an innovative technology that has shown promising results in clinical trials in patients presenting with relapsed or refractory B-cell hematologic neoplasms.2,3

The mechanism of action of autologous CAR-T cells is based on modifying the patient's own T cells to express receptors for a target antigen. In this type of treatment, blood is collected through leukapheresis and T cells are genetically modified and reinfused into the patient.4 These modified T cells contain chimeric antigen receptors, which are artificial surface receptors that aim to specifically recognize and attack the patient's tumor.5

The first paper on the chimeric combination of receptors and antibodies was published in 1989 by the Weizmann Institute in Israel.6 Since then effort has been devoted to this area of research, leading to therapeutic success in 2012, when seven-year-old Emily Whitehead was cured of relapsed and refractory B-cell acute lymphoblastic leukemia by CAR-T cell therapy.7,8

The therapy has shown promise and despite advances, many clinical trials involving CAR-T continue to be carried out in order to ensure greater safety and efficacy.9 Common adverse effects related to the use of CAR-T cell therapy, such as cytokine release syndrome (CRS), may limit its use.

Thus, a synthesis of evidence is needed to evaluate the safety of these therapies.10 The aim of this study is to synthesize available evidence in the scientific medical literature on the occurrence and management strategies of CRS in patients with diffuse large B-cell lymphoma (DLBCL) who received CAR-T cell therapy.

Figure 1 shows that the systematic search in the three databases resulted in the identification of 1007 articles. After excluding duplicates and after screening by title and abstract, 76 articles were read in full. Consequently, 19 studies that fulfilled the eligibility criteria were included in this review.

Figure 1.

Flowchart of article selection.

(0.33MB).

The articles included in this review were published between 2017 and 2022, with China (n = 7) producing the most studies, followed by United States of America (USA) (n = 6) and Japan (n = 4). A total of 1193 patients enrolled in 19 clinical trials received CAR-T cell therapy. The sample sizes ranged from 7 to 256 patients across studies.

The ages of the 1193 patients ranged from 17 to 86 years with a mean age of 58 years. Interestingly, the PILOT study by Sehgal et al.26 from the USA that included 61 participants, recruited only over 70-year-old patients. Most studies (n = 11) included patients who had received at least two lines of prior therapy, another five studies required at least one line of prior therapy and three studies tested CAR-T cell therapy as second-line (Table 1).

The characteristics of the CAR-T cell treatment are listed in Table 2. Regarding the CAR-T cell product used, seven studies (37%) did not use commercially described products, five studies (26%) used axicabtagene ciloleucel, four (21%) used lisocabtagene maraleucel, two (11%) used tisagenlecleucel and one (5%) used relmacabtagene autoleucel. The target chimeric antigen receptor antigen in 16 studies (84%) was CD19, in another two studies (11%) it was CD19 plus CD22 and one study (5%) targeted the CD20 antigen.

Lymphodepleting therapy was administered in 100% of patients in 17 studies (89%), one study (5%) performed lymphodepletion in 93% of patients, and the other study (5%) in 87% of patients. The regimen employed for lymphodepletion was fludarabine plus cyclophosphamide in 15 studies (79%), fludarabine plus cyclophosphamide or bendamustine in two studies (11%), fludarabine plus cyclophosphamide and decitabine in one study (5%), and fludarabine plus cyclophosphamide and ifosfamide in one study (5%). The doses and administration regimens were quite variable as can be seen in more detail in Table 3.

Bridging therapy was used in ten studies (53%), the highest frequencies of the use of this therapeutic modality were reported in the articles by Makita et al.18 (where all patients received bridging therapy), by Schuster et al.11 (92% of patients) and by Kamdar et al.25 (63%). The most commonly employed bridging regimen was systemic therapy with glucocorticoids or chemotherapy described in seven studies (37%); one study (5%) employed radiation therapy in addition to systemic therapy, in three studies (16%) bridging therapy was not allowed, two (11%) had no need to use bridging therapy, and seven studies (37%) did not report the regimen that was or would be employed.

The doses of CAR-T applied varied greatly among the studies, 14 studies applied doses based on the weight (kg) of the patient; the doses in these studies ranged from 0.5 × 106 to 3.0 × 108 CAR-T cells/kg. The other five studies applied a fixed average dose that ranged from 91.0 × 106 to 150.0 × 106 CAR-T cells. All patients received a single infusion of CAR-T cells, except in one study where the dose was split over three days in a staggered manner (10%, 30%, and 60% of the total dose).

The outcomes of the studies had a median complete response of 54%; the study with the highest complete response was by Neelapu et al. (the ZUMA-12 study)14 with 78% of patients achieving a complete response. The worst outcome was seen in the report by Kato et al.15 where only 27% of patients achieved a complete response.

Progression-free survival ranged from 5 to 27.6 months, with a median of seven months. In four (21%) studies progression-free survival was not achieved due to the short follow-up period, and two studies (11%) did not report information on progression-free survival.

The overall results of the research related to CRS are presented in Table 4. In this review, the CRS did not receive a standard classification but rather the classification was determined by original articles. It is noteworthy that significant discrepancies may exist between various classification systems currently employed for toxicity related to CAR-T cell therapy. This disparity can have implications on the diagnosis and treatment of CRS.

Of the 1193 patients included in this review, 804 (67%) developed some degree of CRS. The mean frequency of occurrence of any degree of CRS was 72%. In three studies, all patients developed CRS. The study with the lowest occurrence of CRS was Abramson et al.13 in which 113 patients (42%) had some degree of CRS.

Grade 3 CRS was reported in 80 (10%) patients of the 804 patients who evolved with some grade of CRS. The studies with the highest frequencies were Cheng et al.23, Sang et al.29 and Qu et al.28 (Grade 3 CRS in 33%, 28% and 21% of patients, respectively). No occurrence of Grade 3 CRS was reported by four studies.

Grade 4 CRS was reported in 22 (3%) patients of the 804 patients who developed some degree of CRS. Two studies did not report information about the occurrence of Grade 4 CRS. In eight studies, there were no cases of Grade 4 CRS. The highest frequencies of Grade 4 CRS were observed in the studies by Locke et al.17 (14%), Goto et al.20 (11%), Kato et al.15 (8%) and Schuster et al. (8%).11

Only one study recorded death as a result of CRS, with 2% of CRS-associated deaths, that is, 0.25% of the 804 patients who developed some degree of CRS.

The mean time of onset of CRS symptoms ranged from 0.63 to 5 days, with a median of three days. The greatest variation was found in the study by Abramson et al.13 who recorded a range of 1–14 days. Two studies did not report information on the time of symptom onset. The mean duration of CRS ranged from 4 to 16.5 days, with a median of six days. The greatest variation was recorded by Ying et al.16, where the time to solve CRS ranged from 1 to 118 days. Three studies reported no information on the duration of CRS.

The regimen employed for the management of CRS was based on the use of tocilizumab and/or glucocorticoids in 14 studies, two studies used glucocorticoids only, one study used tocilizumab only, and one other study reported using monotherapy or combination therapy with tocilizumab, infliximab, etanercept, and glucocorticoids. One study did not report the therapeutic regimen employed.

Our analysis shows that most patients develop CRS (mean frequency 72%), but only 10% and 3% evolve with Grade 3 or 4 CRS, respectively. Hernani et al.30 presented similar data when they reported that CRS affects from 42 to 93% of CD19 with severe (Grade ≥ 3) CRS in 2–22%.

Analyzing the data obtained in this review there seems to be no correlation between the average age of patients and the development of CRS. An even more curious finding was that the study by Sehgal et al.26, which only enrolled over 70-year-old patients, reported the lowest frequency of CRS of all the included studies. These findings differ from the results of Locke et al.31 who found that older patients with DLBCL had a higher risk of CRS or neurotoxicity syndrome when treated with axicabtagene ciloleucel.

The development of CRS and the type of commercial CAR-T product applied seem to be related. While studies using lisocabtagene maraleucel had an average occurrence of CRS of 45%, those using tisagenlecleucel had an average of 63%, and those using axicabtagene ciloleucel had an average of 90%. Supporting these data, a study conducted by the Spanish Lymphoma and Bone Marrow Transplant Group of patients with refractory DLBCL demonstrated that axicabtagene ciloleucel may be superior in efficacy to tisagenlecleucel, yet more toxic.32

The combination of fludarabine and cyclophosphamide was the most commonly used conditioning therapy. Studies point out that conditioning chemotherapy is essential for the efficacy of adoptive T-cell therapy. The addition to fludarabine, cyclophosphamide has been associated with better CAR-T cell expansion and persistence compared to other regimens, and the combination of these agents represents the preferred conditioning regimen in patients with relapsed or refractory DLBCL.33

A Phase II clinical trial demonstrated that radiotherapy is an optimal volume reducing regimen for the management of patients with DLBCL prior to CAR-T cell therapy and a promising alternative salvage therapy for patients who relapse after CAR-T cell therapy.34 In this review, none of the included studies used radiotherapy as conditioning.

Bridge therapy has varied widely, from standard chemotherapy to targeted therapies (immunomodulators, tyrosine kinase inhibitors), to radiation and finally corticosteroids. Studies show that bridging therapy is used in about half of individuals.33 Cook et al.35 stress that most patients with lymphoma require bridging therapy while waiting for CAR-T cells to be produced due to the long time between the initial evaluation of the patient to the infusion of the cells.

Bridging therapy remains very heterogeneous, and no preferable strategy has been identified to date.36 In this review, we identified one study that used, in addition to systemic therapy, radiation as bridging therapy. Recent research demonstrates that radiation therapy is a safe and effective method for patients with DLBCL, not only because it reduces myelosuppression compared to bridging chemotherapy, but also it potentially improves the tumor microenvironment.37

The conditioning regimen may affect the development of CRS. Wang et al.38 reported low toxicity in treating B-cell non-Hodgkin lymphoma patients with central memory-derived CAR-T cells after autologous transplantation, suggesting that certain T-cell subsets and the choice of conditioning therapy prior to CAR-T infusion contribute to the severity of CRS.

The data set obtained in this review shows a safety profile regarding CRS that is similar to the published literature. As initially discussed, we observed in this review a high rate of CRS, but the portion of patients who evolve to a more severe degree is relatively small. Still in this regard, the number of deaths from CRS was also low. However, it is still necessary to consider that the causes of CRS are very heterogeneous and different classification systems increase the differences in CRS rates between studies.39

The use of tocilizumab appears to be well established as a treatment for CRS. Studies show that tocilizumab, an interleukin-6 blocking therapy, resulted in rapid recovery from CRS in patients with aggressive non-Hodgkin lymphoma treated with CAR-T. Some authors also advocate the early use of tocilizumab as a strategy to reduce the risk of severe CRS further. Steroid therapy may also be effective in attenuating CRS and is currently reserved as second line therapy after tocilizumab failure.39,40

Cook et al.35 point out that future studies should seek to better elucidate the mechanisms of these toxicities and develop prognostic models to identify subsets of patients who will benefit from prophylactic strategies to prevent immune effector cell-associated toxicities. A deeper understanding of CRS will allow the development of new approaches to reduce toxicities and improve the outcomes of CAR-T cell therapy.39

The recent approval of CAR-T cell therapy has marked a new era in cancer treatment. This therapy has been shown to cure some patients and extend survival for many others whose treatment options would otherwise be limited. However, this success comes with challenges as is the case with CRS. The results obtained in this review demonstrate high rates of CRS in DLBCL patients treated with CAR-T cell therapy nevertheless these adverse events are manageable, supporting the conclusion that this therapy is safe in DLBCL patients.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.