Emerging Strategies for the Prevention of Immune Toxicities Associated with T cell–Engaging Cancer Therapies

Summary:

Immune-related toxicities including cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS) are common side effects of bispecific antibody and chimeric antigen receptor (CAR) T-cell therapies of hematologic malignancies. As anti-inflammatory therapy (the standard of care) is variably effective in mitigating these toxicities after onset, here we discuss emerging evidence for shifting the strategy from mitigation to prevention.

INTRODUCTION

In the context of hematologic malignancies, the introduction of bispecific monoclonal antibodies and chimeric antigen receptor (CAR) T-cell therapies has shifted the focus of novel treatment strategies from chemotherapy to immunotherapy. In the past 3 years alone, over 20 new immunotherapies have been approved by the FDA for the treatment of lymphomas, leukemias, and multiple myeloma. Despite the promising efficacy, bispecific antibodies and CAR T cells are associated with a unique and predictable side-effect profile (1). T-cell overactivation can frequently lead to cytokine release syndrome (CRS) and immune effector cell–associated neurotoxicity syndrome (ICANS) after the initial doses (and sometimes occurring or recurring after subsequent doses; ref. 2).

Although the pathophysiology of CRS and ICANS is incompletely understood, both interleukin-1 (IL1) and IL-6 have been demonstrated to be key mediators. Among the clinically approved interleukin receptor antagonists, the IL1 receptor antagonist anakinra and the IL6 receptor-blocking antibody tocilizumab are commonly used to treat immune toxicities in the clinic. In mice, prophylactic tocilizumab and anakinra prevented ICANS and CRS without denting antileukemia CAR T-cell activity (3). This led to the investigation of these drugs to prevent CRS and ICANS in patients with hematologic malignancies receiving immune-engaging therapies.

Management of CRS and ICANS

Although the exact onset varies between patients and individual agents, the highest risk of CRS is early in the treatment course and the cornerstone of CRS management is close monitoring at this stage of treatment. The classic components of CRS include fever, hypotension, and hypoxia and patients should be regularly checked for their presence by trained staff. Patients should also be screened for neurotoxicity at regular intervals using standardized tools such as the Immune effector Cell Encephalopathy (ICE) score. At the first sign of CRS or neurotoxicity, the patient should be graded according to standardized tools and quick action must be taken to manage symptoms to avoid progression to severe manifestations (4).

The standard interventions for CRS and ICANS are determined by the grade of the toxicity and specific manifestations. These interventions include withholding additional doses, acetaminophen, fluids, supplemental oxygen, steroids, and anti-IL6 therapy. For high-grade CRS or ICANS, admission to the intensive care unit and aggressive interventions are indicated. Although grade 3 or higher CRS is rare with the currently approved agents, the potential for severe life-threatening sequelae underscores the importance of optimizing prevention and monitoring.

Strategies in Leukemia

One of the first studies to investigate early tocilizumab intervention was in pediatric patients with acute lymphoblastic leukemia treated with the CAR T-cell therapy tisagenlecleucel (5). Historically with tisagenlecleucel, CRS was observed in 77% of patients including 21% grade 4 CRS among all patients (6). Due to this high rate of severe CRS, investigators looked for novel approaches to mitigate CRS. In one study published by Kadauke and colleagues, tocilizumab was administered in a risk-adapted approach to a subpopulation of 15 patients with a high burden of disease (5). The aim was to prevent progression to severe CRS (grade 3 or higher). These patients were given “early” tocilizumab at the occurrence of two temperatures ≥38.5°C in a 24-hour period at least 4 hours apart. In the overall patient population (including patients with low tumor burden not given “early” tocilizumab), grade 4 CRS occurred in 9% of patients, which is lower than previously reported. Interestingly, this study demonstrated a benefit in the subpopulation deemed to have a high tumor burden who received “early” tocilizumab when compared with historical controls. Although this was not a truly prophylactic approach, the use of “early” tocilizumab was demonstrated to prevent progression to grade 4 CRS without compromising antitumor efficacy. There remains an opportunity to further investigate prophylactic interleukin-blocking therapy to prevent CRS and ICANS in the field of acute leukemias.

Strategies in Lymphoma

In the treatment of non-Hodgkin lymphomas (NHL), both prophylactic IL1 and IL6 blockade have been investigated to reduce the rate and severity of CRS with immune-engaging therapies (7, 8). One of the first studies using this approach included 20 patients treated with tisagenlecleucel or investigational anti-CD19 CAR T-cell therapy who were given a single dose of tocilizumab prophylaxis prior to infusion of CAR T cells (7). There were no grade 3 or higher CRS events, and there was one grade 4 ICANS event. These rates compare favorably with the rates observed in the JULIET trial of tisagenlecleucel in NHL, with grade 3 or higher CRS and neurotoxicity in 22% and 12% of patients, respectively (9). In the prophylactic tocilizumab study, the estimated progression free survival (PFS) and overall survival (OS) were 73% and 83%, suggesting that the efficacy of CAR T-cell therapy may be preserved (7). The authors concluded that this approach was associated with a low incidence of high-grade CRS without an increase in ICANS or diminished efficacy and that further investigation was warranted. A subsequent study published by Park and colleagues in 2023 evaluated anakinra in the prophylactic setting (8). In this trial, 25 patients received either tisagenlecleucel, axicabtagene ciloleucel, or brexucabtagene ciloleucel for NHL and were given prophylactic anakinra continuously on an every 12-hour dosing schedule on day 2 through at least day 10 after CAR T-cell infusion. The primary endpoint of grade 3 or higher ICANS occurred in 9.7% of patients and the rate of grade 3 or higher CRS was 6.4%, both of which compare favorably with historical comparators (9). The overall response rate (ORR) was 77%, the median PFS was 16 months, and the median OS was not reached. This suggests prophylactic anakinra may not compromise antitumor efficacy of CAR T-cell therapies. Taken together, these studies represent emerging data for this novel approach to mitigating CRS in the field of NHL and acute lymphoblastic leukemia.

Strategies in Myeloma

In multiple myeloma, drug development has shifted away from cytotoxic chemotherapy and toward targeted and immune therapies. One target of these novel therapies is B-cell maturation antigen (BCMA), which is expressed exclusively on mature B cells and plasma cells (1). The first FDA approvals for BCMA-targeted drugs were antibody–drug conjugates (ADC) and CAR-modified T cells (10, 11). Despite this progress, relapsed/refractory multiple myeloma (RRMM), which is refractory to proteasome inhibitors, immunomodulatory agents (IMID), and anti-CD38 monoclonal antibodies, remains a challenge to treat. In 2022, teclistamab, the first T-cell redirecting bispecific antibody targeting BCMA and CD3, received FDA approval for the treatment of patients with RRMM who have received at least four prior lines of therapy (2). This was followed in 2023 by a second BCMAxCD3 (elranatamab) and a GPCR5DxCD3 (talquetamab) bispecific antibody. At the time of writing, there are a total of three bispecific antibodies FDA approved for the treatment of RRMM with many more in the drug development pipeline. Early-phase I/II studies of these therapies have shown the potential to induce deep and durable responses. For teclistamab, studies have seen 63% of patients achieving a partial response (PR) or better (ORR) and 39.4% of patients achieving a complete response (CR; ref. 2). This efficacy appears similar for elranatamab with an ORR of 61% and talquetamab with ORR of 68% (12, 13). Among patients who respond to bispecific antibodies, the responses also appear to be durable with a median duration of response of 18.4 months for teclistamab, not reported for elranatamab, and 10.2 months for talquetamab (2, 12, 13).

In multiple myeloma, common strategies used to prevent CRS include step-up dosing, close monitoring/hospitalization, and premedication with early doses when the risk is the highest. Frequently used premedications include acetaminophen, dexamethasone, and diphenhydramine. Despite this approach, CRS is still observed in most patients. Indeed, all three FDA-approved bispecific antibodies recommend at least one period of inpatient monitoring, carry black box warnings, and have risk evaluation and mitigation strategy (REMS) program attempting to mitigate these toxicities. For teclistamab, all grade CRS and neurotoxicity are reported at 72% and 14%, respectively (Table 1). Per the current FDA United States Prescribing Information (USPI), it is recommended to complete step-up dosing in a hospital setting to closely monitor patients for 48 hours after each step-up dose. Similarly, CRS is reported in 58% of patients treated with elranatamab and 78% of patients treated with talquetamab (Table 1). As these and other drugs emerge in the treatment landscape, the prevention, and management of CRS and neurotoxicity will be an important area of clinical interest.

Table 1.

Current literature on prophylactic interleukin blockade prior to administration of immune-engaging therapies, and select comparators without prophylaxis.

	Trudel et al.	Trudel et al.	Majestec-1	Van de Donk et al.	Present study	Caimi et al.	Park et al.
Disease	MM	MM	MM	MM	MM	NHL	NHL
Therapy	Cevostamab (FcRH5xCD3)	Cevostamab (FcRH5xCD3)	Teclistamab (BCMAxCD3)	Teclistamab (BCMAxCD3)	Teclistamab (BCMAxCD3)	CAR T (CD19)	CAR T (CD19)
Prophylaxis intervention	No	Tocilizumab	No	Tocilizumab	Tocilizumab	Tocilizumab	Anakinra
Number of patients	44	28	165	14	31	20	25
CRS %	91%	36%	72%	29%	13%	50%	74%
ICANS %	NR	NR	3%	NR	10%	25%	19%
ORR %	38%	50%	63%	57%a	50%	71%	77%
Age >75%	NR	NR	14%	NR	32%	NR	NRb
EMD – soft tissue%	NR	NR	17%	21%	55%	NA	NA
ECOG 2 +	NR	NR	0.6%	0%	52%	NR	NR
Median prior lines (range)	6 (2–11)	4 (2–7)	5 (2–14)	4 (2–7)	5 (4–12)	4 (2–7)	NR

Abbreviations: ECOG, Eastern Cooperative Oncology Group Performance Status; EMD, extramedullary disease; MM, multiple myeloma; NA, not applicable; NR, not reported.

^aORR with median follow-up only 1.2 months.

^bAge ≥65 is 48%.

In the treatment of multiple myeloma, data from ASH 2022 from 28 RRMM patients treated with the bispecific antibody cevostamab (targeting FcRH5xCD3) indicate that use of prophylactic tocilizumab can reduce the risk of developing CRS significantly, without impacting its antimyeloma activity (Table 1; ref. 14). Subsequently, preliminary data from 14 patients treated with prophylactic tocilizumab prior to teclistamab were presented at the American Society of Clinical Oncology 2023 Annual Meeting, which suggested reduced CRS risk and no new safety signals when tocilizumab was given prophylactically prior to teclistamab (Table 1; ref. 15). We were motivated to investigate the role of a single dose of tocilizumab before teclistamab in a prospectively followed cohort of patients in the real-world setting to further improve the benefit-to-risk balance. We performed a single-center study including all RRMM patients treated with teclistamab in the at the University of Miami Hospital and Clinics, Sylvester Comprehensive Cancer Center between October 25, 2022 (date of FDA approval) and July 21, 2023 (data cutoff). Patients were prospectively followed for safety and efficacy outcomes. The study was approved by the Sylvester Comprehensive Cancer Center Institutional Review Board and data were collected by review of electronic health records.

Patients received teclistamab in accordance with the FDA USPI (i.e., step-up dosing inpatient followed by outpatient treatment). Also, 1 hour prior to receiving the first step-up dose of teclistamab, all patients were given a single dose of intravenous tocilizumab 8 mg/kg (max dose 800 mg). Acetaminophen, dexamethasone, and diphenhydramine were given prior to each of the three doses in the escalation series. All patients received prophylactic antimicrobials including levofloxacin, sulfamethoxazole/trimethoprim, and acyclovir. The primary aim was to investigate the effect on the rate of CRS which was assessed by the standard American Society of Transplantation and Cellular Therapy criteria by Lee and colleagues (4) Other adverse events were graded according to the National Cancer Institute - Common Terminology Criteria for Adverse Events v.5.0 grading criteria.

A total of 31 patients received at least one dose of teclistamab. The median age was 71 (range, 50–84), and 10 (32%) patients were 75 or older. Nine (29%) patients were Black and eight (26%) were Hispanic ethnicity. A total of 24 (77%) patients had extramedullary disease, 17 (55%) of which had soft-tissue plasmacytomas. Eight (26%) patients had at least 60% plasma cell involvement on their most recent bone marrow biopsy. Similarly, nine (29%) patients were known to have a high-risk FISH/cytogenetic profile (t(4;14), t(14;16), and/or del17p). Sixteen (52%) patients had an Eastern Cooperative Oncology Group (ECOG) score of 2 or higher. Twenty-six (84%) met at least one exclusion criterion for the MAJESTEC-1 trial (2). The most common criteria which would have met the MAJESTEC-1 exclusion criteria for our patients included: cytopenias (n = 15; 48%), impaired glomerular filtration rate (GFR, n = 8; 26%), and previous exposure to BCMA-targeted therapy (n = 4; 13%).

With this approach, we observed a low rate of CRS (13%: 95% binomial confidence interval (CI) 4%–30%) and ICANS (10%: 2%–26%), for the entire duration of the step-up dosing time-window performed inpatient (Table 1). All CRS events occurred after either step-up dose 2 (6%: 1%–21%) or step-up dose 3 (6%: 1%–21%) with no recurrent episodes thereafter. All CRS events were grade 1, and no patient required an additional dose of tocilizumab. There was no observed grade 3 or higher CRS or neurotoxicity. One patient experienced a treatment delay due to ICANS; no patients were delayed due to CRS. The treatment delay associated with ICANS occurred after step-up dose 1 and recurred after step-up dose 2. The duration of delay was 1 day and 2 days, respectively. Hypogammaglobinemia (IgG <400 mg/dL) occurred in 21 (68%) patients and documented infections occurred in eight (26%) patients. Neutropenia was observed in 27 (87%) patients, including 20 (65%) with grade 3 or higher neutropenia; these results are similar to the MAJESTEC-1 trial (2). Among patients with a baseline absolute neutrophil count of >1,000 cells/μL who subsequently experienced grade ≥3 neutropenia, the median time to grade ≥3 neutropenia was 7.5 days (range, 1–123). Intravenous immunoglobulin (IVIg) and G-CSF were administered per institutional guidelines. A total of five (16%) patients had a grade 3 infection and types of infection included pneumonia, COVID-19, and urinary tract infections.

At the time of data cutoff, 17 patients remained on therapy. Three (10%) patients died within 30 days, and one (3%) within 60 days secondary to progressive disease. These patients were all heavily pretreated with RRMM and they had received a total of four, nine, five, and six prior lines of therapy, respectively. All four of these patients would have met exclusion criteria for the MAJESTEC-1 trial, three had soft-tissue extramedullary disease involvement, two had high-risk FISH/cytogenetics, two were penta-drug refractory, and one had known central nervous system involvement (2). At the time of data cutoff, we observed an all-cause mortality rate of 19% (6 patients).

One patient had nonsecretory disease. Among 30 patients with secretory disease, the ORR was 15/30 (50%: 31%–69%). Based on standard International Myeloma Working Group (IMWG) response criteria, 9 patients (30%: 15%–49%) achieved a CR, 4 patients (13%: 4%–31%) achieved a PR, 11 patients (37%: 20%–56%) achieved stable disease, and 4 patients (13%: 4%–31%) had progressive disease. The median follow-up time was 109 days (range, 9–225) and the median duration of response was not reached. At the time of data cutoff, 17 (55%) of patients were currently receiving teclistamab therapy. Median time since initiation of teclistamab was 92 days (range, 8–198). Median event-free survival was not reached.

As expected for RRMM patients in the real-world setting, we found the patient disease characteristics in our study to differ from published clinical trials. Indeed, when compared with the MAJESTEC-1 trial (2), our patients had higher median age (71 vs. 64 years), worse ECOG performance status (52% vs. 1% ECOG 2–4), and more extramedullary disease including the presence of soft-tissue plasmacytomas (55% vs. 17%).

DISCUSSION

As immunotherapies are being increasingly used for the treatment of hematologic malignancies, there is an increasing need to mitigate immune-related adverse events. The challenge of CRS is common across hematologic malignancies and is increasing as more of these therapies are available to patients. In leukemia, CRS is seen with the FDA-approved bispecific antibody blinatumomab and with FDA-approved CAR T-cell therapies (tisagenlecleucel and brexucabtagene ciloleucel). In lymphoma, it is seen with the three FDA-approved bispecific antibodies (monsunetuzumab, epcoritamab, and glofitamab) and the four FDA-approved CAR T-cell therapies (tisagenlecleucel, brexucabtagene ciloleucel, lisocabtqgene maraleucel, and axicabtagene ciloleucel). In multiple myeloma, it is seen with the three FDA-approved bispecific antibodies (teclistamab, elranatamab, and talquetamab) as well as the two FDA-approved CAR T-cell therapies (idacabtagene vicleucel and ciltacabtagene autoleucel). In each of these malignancies, CRS is a concern for many emerging therapies in the drug development pipeline, underscoring the importance of limiting its impact on patients.

We observed a substantially lower rate of CRS among patients who received a single dose of prophylactic tocilizumab when compared with published clinical trials without prophylactic tocilizumab (13% vs. 72%). Similarly, van de Donk and colleagues (15) observed CRS in 29% of 14 patients treated with prophylactic tocilizumab prior to teclistamab (15). A summary of prophylactic strategies and their impact on CRS can be seen in Table 1. Given our sicker patient population, treatment responses were slightly poorer compared with early-phase I–II clinical trials of teclistamab (2). The evaluation of prophylactic tocilizumab for the prevention of CRS/ICANS for bispecific antibodies is an area of significant clinical interest given that such strategies may shorten inpatient admission or allow complete outpatient administration. Inherently, since teclistamab was quite recently FDA approved, the follow-up time is short, and the sample size is small. Despite these limitations, the similarity of the preliminary data trends across cohorts in Table 1 suggests that tocilizumab may be effective as a preventative, rather than reactive, measure (2, 14, 15).

When taken together, these small studies warrant further investigation of this approach in larger prospective studies across the spectrum of hematologic malignancies. The prophylactic use of IL1 receptor antagonist and IL6 receptor-blocking antibodies, anakinra and tocilizumab, has the potential to limit or eliminate the need for hospitalization to monitor for CRS and ICANS. In the future, these immunotherapies may be administered in an increasingly outpatient setting. This, in turn, has the potential to lower healthcare resource utilization while improving patient quality of life. Lower rates of all grade CRS and ICANS may translate to a better quality of care for patients. At the present time, many unresolved questions remain for this strategy to mitigate immune-related toxicities. Not the least of which is the potential to decrease efficacy through interference with nonspecific immune system function. This hypothetical concern may be more pronounced in the context of single-dose CAR T-cell infusions when compared with continued weekly or biweekly dosing often seen with the bispecific antibodies. However, this concern remains unresolved with the presently available data. Additional studies are warranted to address these important questions and to further characterize the impact of these therapies on the incidence and severity of CRS and ICANS associated with immune-engaging therapies in the context of hematologic malignancies.

Authors’ Disclosures

A. Kowalski reports personal fees from Pfizer outside the submitted work. B. Diamond reports personal fees from Sanofi and Janssen outside the submitted work. O. Landgren reports grants from Amgen, Janssen, personal fees from Celgene, Karyopharm, Adaptive Biotech, Binding Site, Bristol Myers Squibb, and Pfizer, and other support from Takeda, Janssen, and Merck outside the submitted work. No disclosures were reported by the other authors.