Tocilizumab Dosing for Management of T Cell‐Engaging Bispecific Antibody‐Related CRS in Patients With R/R B‐Cell NHL

Abstract

The recent surge in T‐cell‐engaging and chimeric antigen receptor (CAR) T‐cell therapies is changing the landscape of cancer therapy. Cytokine release syndrome (CRS) is a systemic inflammatory response syndrome that is a well‐known complication of these therapies, of which interleukin‐6 (IL‐6) is a key mediator. Tocilizumab, an IL‐6 receptor (IL‐6R) antagonist, is approved for the management of CAR T‐cell therapy‐induced CRS in adults and in pediatric patients aged ≥ 2 years old. However, the approved dosing schedule was not based on IL‐6R occupancy and may not be the most suitable schedule for T‐cell‐engaging therapies such as bispecific antibodies (bsAb) due to key differences in the levels of released IL‐6, the clinical symptomatology of CRS, and the pharmacology of tocilizumab across settings. In this study, we adapted a previously developed tocilizumab and soluble IL‐6R (sIL‐6R) population pharmacokinetic model to describe and predict tocilizumab concentrations and sIL‐6R occupancy over time in patients with anti‐CD20 bsAb‐induced CRS following tocilizumab dosing. Using this model, which incorporates target binding and receptor occupancy, we propose a new tocilizumab dosing regimen that is based on quantitative clinical pharmacology, cytokine analyses, and clinical practice patterns in patients with relapsed/refractory B‐cell non‐Hodgkin's lymphoma (R/R B‐NHL) treated with the anti‐CD20 bsAb mosunetuzumab or glofitamab. This schedule (up to two 8 mg/kg intravenous tocilizumab doses per CRS event at least 8 hours apart and a maximum of three doses in 6 weeks) can be used to effectively manage acute CRS induced by anti‐CD20 bsAb in patients with R/R B‐NHL.

Tocilizumab dosing for management of T cell‐engaging bispecific antibody‐related CRS in patients with R/R B‐cell NHL.

Study Highlights.

• WHAT IS THE CURRENT KNOWLEDGE ON THE TOPIC?

Although the introduction of T‐cell‐engaging and CAR‐T therapies is changing the cancer treatment landscape, these treatments are associated with CRS, a systemic inflammatory syndrome of which IL‐6 is a key mediator. Tocilizumab is an IL‐6 receptor antagonist, currently approved (among other indications) for the management of CAR T‐cell therapy‐induced CRS in adults and pediatric patients aged ≥ 2 years. The approved schedule is for up to four intravenous infusions of 8 mg/kg every 8 hours for patients with body weight ≥ 30 kg.

• WHAT QUESTION DID THIS STUDY ADDRESS?

Given that the approved tocilizumab dosing schedule for CAR T‐cell therapy‐induced CRS may not be the optimal schedule for T‐cell‐engaging therapies, can we use quantitative clinical pharmacology principles, including target binding and receptor occupancy, to determine the optimal schedule for use with anti‐CD20 bispecific antibodies?

• WHAT DOES THIS STUDY ADD TO OUR KNOWLEDGE?

Using a popPK model previously developed in adult and pediatric patients with rheumatoid arthritis and pooled data from a number of studies of the anti‐CD20 bispecific antibodies glofitamab and mosunetuzumab in patients with R/R B‐NHL, we propose a more fit‐for‐purpose tocilizumab dosing regimen for CRS management in this setting.

• HOW MIGHT THIS CHANGE CLINICAL PHARMACOLOGY OR TRANSLATIONAL SCIENCE?

This regimen (up to two 8 mg/kg intravenous tocilizumab doses per CRS event at least 8 hours apart and no more than three total doses over a consecutive 6‐week period for patients with body weight ≥ 30 kg) can be used to effectively manage acute CRS induced by bispecific antibodies in patients with R/R NHL, while considering safety factors and the unique PK properties and expected activity duration of tocilizumab.

T‐cell‐engaging treatments, such as bispecific antibodies (bsAb) and chimeric antigen receptor (CAR) T‐cell therapies, have seen a recent surge in clinical practice, bringing a transformative shift for cancer treatment. ¹ , ² However, while these new modalities hold great potential for improving patient outcomes, they are often associated with specific on‐target adverse events (AEs), such as cytokine release syndrome (CRS). ³ , ⁴ , ⁵

CRS is a systemic inflammatory response syndrome that is an established complication of T‐cell‐engagers and CAR T‐cell therapies. ⁴ , ⁵ , ⁶ CRS is caused by the over‐activation of T cells and other immune effector cells, leading to a considerable release of cytokines into the bloodstream. Cytokines are signaling molecules that play a crucial role in regulating the immune response. However, excessive levels of cytokines can cause a cascade of harmful effects throughout the body. CRS presents with a range of symptoms from mild and flu‐like (fever, fatigue) to severe and potentially life‐threatening (hypotension, organ dysfunction), which may require hospitalization, thus also posing a substantial economic burden. ⁷ , ⁸ , ⁹ , ¹⁰

Patients with severe CRS present with increased levels of various cytokines, with interleukin‐6 (IL‐6) being a central mediator that contributes to many of its hallmark symptoms. ⁸ However, the pathophysiology of CRS is complex and involves a network of interacting cytokines, including interleukins, interferons, tumor necrosis factor, lymphokines, monokines, and chemokines. ⁷ The incidence of CRS in patients receiving cancer immunotherapies varies widely depending on the type of therapeutic agent, with onset occurring within hours to days or weeks after drug administration. ⁸ Disease burden, patient characteristics, type of therapy, and dosage have previously been reported as risk factors for developing CRS, ⁸ and the severity of CRS depends on the extent and nature of the cytokine elevation, as well as the patient's underlying health conditions.

The management of CRS necessitates a multifaceted approach due to the intricate nature of its causative factors. IL‐6 inhibitors, including tocilizumab (a recombinant, humanized, IL‐6 receptor [IL‐6R] antagonist, monoclonal antibody), have demonstrated efficacy in mitigating CRS severity by impeding IL‐6 activity. ⁸ Corticosteroids, such as dexamethasone, are also frequently employed to suppress the immune response and alleviate inflammation. Beyond IL‐6 inhibitors and corticosteroids, alternative therapeutic interventions may be warranted depending on the specific clinical manifestations and underlying etiology of CRS. These may encompass supportive care measures, including fluid resuscitation, oxygen therapy, and vasopressors, as well as targeted therapies designed to modulate the immune response and avert organ damage. ⁸

Tocilizumab is now a recognized cornerstone in the management of CRS associated with cancer immunotherapies and has been approved, among other indications, for the treatment of severe or life‐threatening CAR T‐cell therapy‐induced CRS in adults and pediatric patients aged ≥ 2 years old. ¹¹ , ¹² However, recent guidance on the management of AEs associated with CAR T‐cell therapies and bsAb ¹³ , ¹⁴ , ¹⁵ , ¹⁶ has led to subjective and heterogeneous use of tocilizumab in clinical practice. The approved tocilizumab dosing schedule for the management of CAR T‐cell therapy‐induced CRS is up to four intravenous infusions of 8 mg/kg every 8 hours, based on a maximum safety concentration threshold of 649 μg/mL that was defined by the US Food and Drug Administration (FDA) as the maximum concentration of the maximum tolerated dose (MTD; 28 mg/kg) at which transient neutropenia was observed in healthy volunteers, and was based on the observed exposure–response relationships for safety. ¹¹ Key clinical pharmacology principles such as drug–target binding/interaction and receptor occupancy were not considered in the development of this schedule, and thus, the optimal tocilizumab dose regimen for the management of CAR T‐cell therapy‐induced CRS is yet to be determined. In addition, given the variability in IL‐6 release and clinical symptomatology of CRS across different treatment settings, ¹⁷ it is inappropriate to extrapolate the dosing regimen from CAR T‐cell therapy to other contexts, such as those involving bsAb. Furthermore, the potential acute safety risk following single CAR‐T administration vs. the prolonged risk following multiple doses of bsAbs should be considered. The pharmacology of tocilizumab is also known to differ significantly between treatment types and indications, such as between chronic inflammatory diseases (e.g., rheumatoid arthritis [RA]) and more acute inflammatory diseases (e.g., COVID‐19), where higher clearance (CL) and relatively lower exposures are observed. ¹⁸ , ¹⁹

Mosunetuzumab and glofitamab are anti‐CD20 bsAbs that can redirect T cells to eliminate B cells and have demonstrated clinical benefit, receiving global approvals for patients with relapsed/refractory (R/R) follicular lymphoma (FL) ²⁰ (mosunetuzumab), or R/R diffuse large B‐cell lymphoma ⁵ or large B‐cell lymphoma arising from FL ²¹ (glofitamab), who have received at least two prior therapies. The dosing schedules for mosunetuzumab and glofitamab are similar, with weekly step‐up doses given for CRS mitigation during Cycle 1, and one fixed dose at the beginning of every 21‐day cycle thereafter. ⁵ , ²⁰ , ²¹ Here, we apply quantitative clinical pharmacology (qCP) principles, including target binding and receptor occupancy, to propose a more fit‐for‐purpose tocilizumab dosing regimen for CRS management in patients treated with anti‐CD20 T‐cell‐engaging therapies, which is consistent with the observed clinical practice patterns in mosunetuzumab‐ and glofitamab‐treated patients across each development program.

MATERIALS AND METHODS

Safety and clinical practice

Pooled safety data from phase 1/2 safety studies evaluating mosunetuzumab and glofitamab in patients with R/R B‐cell non‐Hodgkin's lymphoma (B‐NHL) were analyzed. ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ , ²⁸ Data included the number and severity of CRS events, the number of patients receiving tocilizumab for CRS management, the number of tocilizumab doses relative to the number of CRS events, and CRS resolution over time. CRS severity within these studies was graded using American Society for Transplantation and Cellular Therapy criteria. ⁹

Model design

Five studies were used for the development and external evaluation of the population PK (popPK) soluble IL‐6R (sIL‐6R) model. In all studies, patients who experienced CRS of Grade 2 or above were recommended to receive intravenous tocilizumab, with the dose determined by patient body weight (12 mg/kg for patients < 30 kg, 8 mg/kg for patients ≥ 30 kg; doses exceeding 800 mg per infusion were not recommended). More than one dose could be administered; in studies used for external model evaluation, dosing could be repeated every 8 hours as necessary up to a maximum of four doses.

The concentrations of tocilizumab and sIL‐6R in human serum samples were determined with the established and validated enzyme‐linked immunosorbent assays, which were fully validated to international regulatory standards with a quantitation limit of 0.1 μg/mL for tocilizumab and 1.0 μg/mL for sIL‐6R. The sIL‐6R concentration was the sum of the unbound sIL‐6R and sIL‐6R bound to tocilizumab. Serum samples were obtained according to the following schedule: pre‐dose; end‐of‐infusion; 6, 24, 48, and 72 hours after infusion; Day 8 after infusion; and Week 8 after infusion (tocilizumab only).

popPK model development

The popPK analysis was conducted using non‐linear mixed‐effects modeling with NONMEM software (NONMEM version 7.5.0 [Icon Development Solutions]). During the assembly of the NONMEM data file, data were scrutinized for subjects who had missing covariate information. These subjects were not excluded from the analysis but were assigned the median value of that covariate in the data file or a value based on imputation from their other covariate values (if possible). ²⁹ Missing categorical covariates were assigned a separate category. For missing sampling times or missing concentrations, the corresponding sampling information was omitted. Values below the quantification limit were also excluded from the analysis.

A popPK‐sIL‐6R model of tocilizumab was previously developed by the co‐authors ²⁹ and utilized the indirect‐response model with inhibition of elimination to describe the sIL‐6R time course following tocilizumab administration. This model structure arises from equations of the quasi‐steady‐state (QSS) approximation of the target‐mediated drug disposition (TMDD) model. ³⁰ , ³¹ This joint PK‐sIL‐6R model that applied the QSS approximation of the two‐target (sIL‐6R and membrane‐bound [m]IL‐6R) TMDD model ³² was used to describe concentration time courses of tocilizumab and sIL‐6R.

The tocilizumab population PK‐sIL‐6R model described above was later updated internally by the authors (unpublished model) using data from multiple studies in adult patients with rheumatoid arthritis (RA) and pediatric patients with systemic juvenile idiopathic arthritis (sJIA) and polyarticular‐course juvenile idiopathic arthritis (pcJIA).

Tocilizumab concentrations were described by a two‐compartment model with parallel linear and Michaelis–Menten elimination, and first‐order subcutaneous absorption. Total (unbound and tocilizumab‐bound) sIL‐6R concentrations were described by equations of QSS approximation of the TMDD model (that is equivalent to an indirect response model with inhibition of elimination). ³⁰ , ³² Joint fit of tocilizumab and sIL‐6R data was performed. Parameter estimates for the internally developed PK‐sIL‐6R model are presented in Table S1 .

This unpublished model was applied and adapted to tocilizumab data following tocilizumab administration to mitigate CRS induced by CD20 T‐cell engagers. First, the model with fixed parameters (Table S2 ) was supplemented by CRS effects (i.e., effects of studies in patients with CRS) on model parameters of linear clearance (CL), inter‐compartment clearance (Q), central and peripheral volumes (V _C and V _P), maximum target‐mediated elimination rate (V _M), elimination rate of bound sIL‐6R (k _int), and degradation rate of free (unbound) sIL‐6R (k _deg). Baseline sIL‐6R values (for patients with missing observed baseline) were also re‐estimated. Then, the random effect structure was simplified. Covariate effects not supported by the model were removed; dependencies of CRS effect on clearance and baseline sIL‐6R (for patients with missing baseline) on CRS severity were investigated. The significance level of 0.01 (change of objective function of 6.63 points for 1 estimated parameter) was used to compare models.

The final model was evaluated using diagnostic plots, visual predictive check (VPC), and nonparametric bootstrap procedures.

popPK model applications

The final popPK model was used to simulate typical tocilizumab and sIL‐6R concentration–time courses to predict sIL‐6R occupancy in patients with CRS following a single 8 mg/kg intravenous tocilizumab dose, and to compare these simulations with those from RA patients. ¹¹ Additional simulations were performed to determine sIL‐6R occupancy following repeated tocilizumab dosing over multiple cycles to evaluate the benefits of multiple dosing, while ensuring that the tocilizumab concentration would remain below 649 μg/mL. Downstream IL‐6 and C‐reactive protein (CRP) activity were explored after the administration of various tocilizumab dosing regimens.

RESULTS

Tocilizumab‐related clinical experience in mosunetuzumab and glofitamab studies

Data from 733 patients treated with mosunetuzumab and 772 patients treated with glofitamab across different dose levels from eight phase I/II studies (NCT02500407; NCT03671018; NCT03677141; NCT03677154; NCT03467373; NCT03075696; NCT03533283; NCT04313608) were analyzed (Table S3 ). Of the mosunetuzumab‐ and glofitamab‐treated patients, 232 (31.7% across four studies) and 427 (55.3%) subjects had ≥ 1 CRS event, respectively, of which 36 (15.5%) and 140 (32.8%) patients, respectively, received tocilizumab for CRS management (Table 1 ). Most patients who received tocilizumab received only one dose (91.7% in the mosunetuzumab group, 70.5% in the glofitamab group; one patient from the glofitamab group was excluded from the analysis as the number of tocilizumab doses was unknown) (Table 1 ). CRS in both patient groups was mainly Grade 1 or 2 (mosunetuzumab‐treated patients, 26/36 [72%]; glofitamab‐treated patients, 385/427 [90%]).

Table 1.

Pooled safety data of patients from eight phase 1/2 studies treated with mosunetuzumab or glofitamab across different dose levels and receiving tocilizumab for the management of at least one CRS event

	Mosunetuzumab (N = 733)	Glofitamab (N = 772)	Total (N = 1,505)
Patients with CRS, n (%)	232 (31.7)	427 (55.3)	659 (43.8)
Patients who received tocilizumab for the management of ≥ 1 CRS event, n (%)	36 (15.5)	140 (32.8)a	176 (26.7)
1 dose, n (%)	33 (91.7)	98 (70.5)	131 (74.9)
2 doses, n (%)	1 (2.8)	19 (13.7)	20 (11.4)
3 doses, n (%)	1 (2.8)	15 (10.8)	16 (9.1)
≥ 4 doses, n (%)	1 (2.8)	7 (5.0)	8 (4.6)

CRS, cytokine release syndrome.

^a Number of doses was unknown for one patient in the glofitamab group.

Overall, CRS resolution in the mosunetuzumab group was achieved within 3 days post‐tocilizumab administration in 75.0% of patients, within 7 days in 91.7% of patients, and within 14 days in 97.2% of patients. Similar results were observed in the glofitamab group, where CRS resolution was achieved within 3 and 7 days post‐tocilizumab administration in 66.9% and 88.8% of patients, respectively.

In the mosunetuzumab group, the median duration of CRS was 3.0 (range, 0.0–20.0) days in patients who received tocilizumab for CRS management (n = 36) compared with 1.0 (0.0–19.5) day in patients who did not receive tocilizumab (n = 196). In the glofitamab group, median duration of CRS for patients receiving tocilizumab (n = 140) was 38.7 (range, 3.1–180.5) hours, compared with 17.0 (0.3–318.0) hours in patients who did not receive tocilizumab (n = 287).

The majority of patients experienced a single investigator‐assessed CRS event (mosunetuzumab group, 180/232 [77.6%]; glofitamab group, 250/427 [58.5%]). In the mosunetuzumab group, multiple episodes of CRS were reported in 45 patients (11/36 [30.6%] patients who received tocilizumab and 34/196 [17.3%] patients who did not). In the glofitamab group, multiple episodes of CRS were reported in 177 patients (55/140 [39.3%] patients who received tocilizumab and 122/287 [42.5%] patients who did not).

CRS events by dose cycle and by tocilizumab treatment for patients treated with mosunetuzumab in study NCT02500407 and patients treated with glofitamab in study NCT03075696 are shown in Figure 1 .

Figure 1.

Sunburst plots illustrating CRS events by dose event and tocilizumab treatment status for all patients treated with (a) mosunetuzumab (n = 211; patients with R/R FL from cohort B11 [1/2/60/30 mg] from study NCT02500407) or (b) glofitamab (n = 213; patients from study NCT03075696). BNHL, B‐cell non‐Hodgkin lymphoma; C, Cycle; CRS, cytokine release syndrome; D, Day; FL, follicular lymphoma; PK, pharmacokinetics; popPK, population PK; R/R, relapsed/refractory; sIL‐6R, soluble interleukin‐6 receptor.

Model development

Population PK analysis dataset included data from 67 adult patients with R/R B‐NHL who received tocilizumab for CRS management following treatment with mosunetuzumab or glofitamab in five different studies and who had available tocilizumab PK or sIL‐6R samples (Table 2 ). Of the 67 patients, 46 were treated with glofitamab and 21 were treated with mosunetuzumab. PK and sIL‐6R data from 16 and 45 patients from the NCT02500407 (mosunetuzumab) and NCT03075696 (glofitamab) studies, respectively, were used for the development of the model, while the remaining data, that were made available after completion of the model development, were used for the external model validation (NCT03677141 [n = 3], NCT03671018 [n = 1], NCT03677154 [n = 1], NCT03075696 [n = 1], all treated with mosunetuzumab) (Table 2 ). The model development population included 44 (72.1%) male and 17 (27.9%) female patients with a mean age of 63.0 (range, 30.0–87.0) years and a mean body weight of 75.7 (range, 53.0–128.0) kg (Table S4 ). Model development was performed as outlined in the Methods section. Parameters of the final model are presented in Table S5 . The NONMEM code can be found in the Supplement (Table S6 ).

Table 2.

Patients and tocilizumab PK and sIL‐6R sample numbers from the five studies of glofitamab or mosunetuzumab in R/R B‐NHL for the development and evaluation of the popPK model

Study, n (samples)	Drug	Total	Tocilizumab PK	sIL‐6R
NCT03075696	Glofitamab	46 (408)	46 (224)	42 (184)
NCT02500407	Mosunetuzumab	16 (98)	16 (52)	14 (46)
NCT03677141		3 (18)	2 (4)	3 (14)
NCT03671018		1 (10)	1 (5)	1 (5)
NCT03677154		1 (9)	1 (4)	1 (5)
Total		67 (543)	66 (289)	61 (254)

B‐NHL, B‐cell non‐Hodgkin lymphoma; PK, pharmacokinetics; popPK, population PK; R/R, relapsed/refractory; sIL‐6R, soluble interleukin‐6 receptor.

PK and sIL‐6R data from 61 and 55 patients from the NCT02500407 and NCT03075696 studies, respectively, were used for model development; the remaining data were used for external model validation.

The distributions of the random effects on CL and V_C by study, sex, and study/sex combinations were biased, with lower CL and V_C for females and higher CL and V_C for males from study NCT02500407. However, no mechanistic explanations of these dependencies are available and the sample sizes of the groups are small. Therefore, these observed effects were not introduced in the model.

Goodness‐of‐fit plots for tocilizumab and sIL‐6R indicated a good fit of the model (Figure S1 ). Predicted and observed tocilizumab and total sIL‐6R concentrations in individuals receiving one or two 8 mg/kg tocilizumab doses were in good agreement (Figure S1 ), indicating the adequacy of the model to predict tocilizumab concentrations and sIL‐6R occupancy over time. VPCs for PK and sIL‐6R and predicted‐corrected VPCs provided in Figures S2 – S5 indicated that, overall, the model captured the central tendency and the inter‐individual variability of TCZ‐sIL‐6R pharmacokinetics. Bootstrap parameter medians and confidence intervals were in good agreement with the final parameter estimates and asymptotic confidence intervals (Table S5 ).

qCP analyses

The qCP analysis of tocilizumab was conducted using data from 543 samples (289 PK and 254 sIL‐6R) from a total of 67 adult patients with R/R B‐NHL who received tocilizumab for CRS management following treatment with mosunetuzumab or glofitamab in five different studies (Table 2 ). qCP simulations predicted that, for one tocilizumab dose of 8 mg/kg, the median duration of > 90% sIL‐6R saturation was 28 (range, 14–28) days. The sIL‐6R occupancy at 21 days post‐administration of one and two tocilizumab doses given 8 hours apart (8 mg/kg each) was 98.2% and 99.4%, respectively. Up to two consecutive tocilizumab doses of 8 mg/kg every 8 hours per CRS event with a maximum of three tocilizumab doses within 6 weeks were also shown to maintain the concentration of tocilizumab below the safety threshold recommended by the FDA (649 μg/mL) and ensure an average 90% saturation of sIL‐6R for at least 28 consecutive days (Table 3 ).

Table 3.

Predicted tocilizumab concentrations and sIL‐6R occupancy in patients with R/R B‐NHL treated with glofitamab or mosunetuzumab and receiving one or two 8 mg/kg tocilizumab doses per CRS event

Median (range)	Evaluated tocilizumab dosing regimen
	1 × 8 mg/kg	2 × 8 mg/kg
sIL‐6R ROCC14, %	99 (89.2–99.6)	99.6 (97.5–99.8)
sIL‐6R ROCC21, %	98.2 (9.59–99.5)	99.4 (57.9–99.8)
sIL‐6R SATUR90,28, day	28 (13.8–28)	28 (18.3–28)

B‐NHL, B‐cell non‐Hodgkin lymphoma; CRS, cytokine release syndrome; R/R, relapsed/refractory; ROCC_N, receptor occupancy at Day N (14 or 21); SATUR_90,28, duration of > 90% sIL‐6R saturation in the interval from Day 0–28; sIL‐6R, soluble interleukin‐6 receptor.

Clinical pharmacology

Compared with patients receiving tocilizumab for chronic RA, the CL and volume of tocilizumab distribution were approximately 1.5‐ and 1.2‐fold higher, respectively, in patients with R/R B‐NHL receiving tocilizumab for the management of CRS following the administration of glofitamab or mosunetuzumab (Figure 2 ). PK parameters were comparable across the mosunetuzumab and glofitamab clinical trials.

Figure 2.

Predicted (a) tocilizumab concentration (μg/mL), (b) sIL‐6R concentration (μg/mL), (c) sIL‐6R occupancy (%) and (d) percentage of free sIL‐6R (%) for a typical patient with CRS or RA following a single intravenous tocilizumab dose (8 mg/kg). CRS, cytokine release syndrome; RA, rheumatoid arthritis; sIL‐6R, soluble interleukin‐6 receptor.

The dosing schedules for mosunetuzumab and glofitamab consist of weekly step‐up doses given for CRS mitigation on Day (D) 1, D8, and D15 of Cycle 1, and a fixed dose on D1 of every 21‐day cycle thereafter. ⁵ , ²⁰ , ²³ Comparison of simulated tocilizumab concentration and sIL‐6R occupancy time‐profiles for patients receiving between three and four tocilizumab doses within the first two cycles (42 days/6 weeks) of mosunetuzumab or glofitamab treatment indicated that the tocilizumab concentration approaches the safety threshold (649 μg/mL) without any substantial change in the receptor occupancy profile (Figure 3 ). These results suggest that up to two consecutive tocilizumab doses ≥ 8 hours apart per CRS event, and a maximum of three tocilizumab doses within a 6‐week period, can maintain sustainable sIL‐6R occupancy without tocilizumab accumulation.

Figure 3.

Simulations of tocilizumab concentrations and sIL‐6 receptor occupancy after the administration of (a) four and (b) three doses of tocilizumab (8 mg/kg) for the management of CRS within the first two 21‐day cycles of treatment with mosunetuzumab or glofitamab. The maximum tolerated dose equivalent concentration threshold in healthy volunteers used by the FDA is depicted as a red dashed line. C _max, maximum concentration; CRS, cytokine release syndrome; FDA, US Food and Drug Administration; HV, healthy volunteers; sIL‐6, soluble interleukin‐6; TCZ, tocilizumab.

Downstream cytokine (IL‐6 and CRP) activity

PK/PD analysis of downstream IL‐6 activity in glofitamab‐treated patients receiving tocilizumab for CRS management revealed that serum IL‐6 levels increased immediately after tocilizumab administration, most likely due to a displacement of IL‐6 from its receptors through the inhibitory activity of tocilizumab. IL‐6 concentrations then dissipated rapidly over approximately 8 days (Figure 4 a ). CRP levels, a marker of systemic inflammation, also decreased over time following tocilizumab administration irrespective of the number of doses (Figure 4 b ). Similar results were observed in the mosunetuzumab‐treated population (Figure 4 c,d ).

Figure 4.

Representative (a) serum IL‐6 and (b) CRP profiles of individual glofitamab‐treated patients, and (c) serum IL‐6 (N = 14)^a and (d) CRP profiles (N = 22, first 8 days) of individual mosunetuzumab‐treated patients receiving tocilizumab for CRS management. ^aPatients 104316 and 104893: only first dose data were available. Patient 104801: plot represents two tocilizumab doses administered for the treatment of different CRS events. CRP, C‐reactive protein; CRS, cytokine release syndrome; IL‐6, interleukin‐6.

DISCUSSION

In this study, clinical experience and qCP results based on a mosunetuzumab‐ and glofitamab‐treated population have been leveraged to propose tocilizumab dosing guidance for the management of T‐cell‐engaging bsAb‐related CRS in patients with R/R B‐cell NHL. This regimen (for patients ≥ 30 kg body weight) supports the administration of up to two 8 mg/kg intravenous tocilizumab doses per CRS event at least 8 hours apart, and no more than three total doses over a consecutive 6‐week period, that can maintain sIL‐6R occupancy > 98% for several weeks. This dosing regimen addresses a regulatory question, that is, the difference in risk posed by single CAR‐T administration vs. multiple doses of bsAbs, and presents an alternative to the current approved dosing regimen for tocilizumab that was developed for the management of CAR T‐cell therapy‐induced CRS.

CRS is associated with elevated IL‐6, which is an established causative factor that activates downstream pathways by cis‐ and trans‐signaling. In cis‐signaling, IL‐6 forms a complex with gp130, which binds the mIL‐6R on the surface of immune cells. This results in signal transduction via the Janus kinase (JAK)‐STAT3 pathway, which has multiple effects on the acquired and innate immune systems, which may contribute to CRS development. In trans‐signaling, IL‐6 binds the soluble form of the IL‐6R, resulting in the formation of gp130 dimers on the surface of a variety of cell types, including endothelial cells. This induces downstream JAK‐STAT3 signaling, which has the potential to directly induce a systemic cytokine storm, while also causing an increase in E‐cadherin expression. ³³

Tocilizumab acts as a competitive antagonist for both soluble and membrane‐bound forms of IL‐6R, and therefore inhibits both the cis‐ and trans‐signaling pathways. This suppresses IL‐6 signaling by impeding gp130‐mediated JAK‐STAT3 activation, thus inhibiting cytokine‐mediated inflammation and alleviating CRS.

Tocilizumab is well‐established for the effective treatment of CAR T‐cell therapy‐induced CRS, due to its ability to rapidly manage CRS symptoms without compromising CAR T‐cell therapy function. ⁸ , ³⁴ Studies of CRS following the administration of bsAb have shown similar results. For example, in a recent study of teclistamab, an anti‐B‐cell maturation antigen bsAb, tocilizumab treatment for CRS decreased the likelihood of patients experiencing subsequent CRS events (rate of CRS recurrence: 20.0% in the tocilizumab group vs. 62.2% in the non‐tocilizumab group). Furthermore, tocilizumab did not affect response to teclistamab, with overall response in those patients receiving tocilizumab shown to be comparable with those not receiving tocilizumab (70% vs. 72%). ³⁵ A recent study showed that, in patients receiving tocilizumab (8 mg/kg intravenous; maximum dose 800 mg) as a prophylactic measure prior to treatment with teclistamab, elranatamab, or talquetamab, the rate of CRS was low (14%; Grade 1, 90%) and the rate of recurrence of CRS was 0%. ³⁶ , ³⁷

The time course of CRS is dependent on the treatment received. For example, bsAb‐induced CRS typically occurs within 24 hours of the first full dose (and may recur after subsequent doses with decreasing frequency), whereas CAR T‐cell therapy‐induced CRS has an onset of days to weeks after the first infusion. ³⁸ In a recent study of axicabtagene ciloceucel in patients with R/R B‐NHL (FL or marginal zone lymphoma [MZL]), 82% of patients experienced CRS with a median time to onset of 4 days (interquartile range [IQR] 2–6) in patients with FL and 4 days (IQR 2–7) in those with MZL. ³⁹ However, in a study of epcoritamab in patients with R/R large B‐cell lymphoma, median time to CRS onset was just 0.8 days post‐infusion. ⁴⁰ A recent study of CAR T‐cell therapy‐induced CRS has also demonstrated the importance of timing of tocilizumab intervention, with early tocilizumab administration (within 12 hours of CRS onset) resulting in reduced duration of CRS (18.6 vs. 84.7 hours) when compared with later administration (> 12 hours). ⁴¹ These differences in the time course of CRS induced by bsAb compared with CAR T‐cell therapies indicate the need for a more tailored approach to tocilizumab dosing for these different treatments.

In an analysis conducted by the FDA in patients with hematological malignancies treated with CAR T‐cell therapy and receiving tocilizumab for CRS management, the safety threshold for tocilizumab concentration was established at 649 μg/mL (based on the observed exposure–response relationships for safety in healthy volunteers). ¹¹ Although this MTD approach led to the approval of this tocilizumab dosing regimen for patients receiving CAR T‐cell therapy, the present study indicated that sIL‐6R occupancy of tocilizumab is more appropriate to define a more fit‐for‐purpose tocilizumab dosing regimen for CRS management. Here, simulations of sIL‐6R occupancy indicated that one 8 mg/kg dose of tocilizumab in patients with R/R B‐NHL can induce full sIL‐6R occupancy for 3 weeks, suggesting that the FDA‐recommended maximum dosing is not appropriate for this population as it would result in unnecessarily high tocilizumab concentrations. Moreover, the predicted duration of sIL‐6R saturation and the rapid decline in systemic inflammatory markers (IL‐6, CRP) post‐administration suggested that more than three tocilizumab doses within two consecutive cycles would have only limited benefit in CRS management and symptom resolution. The recommended tocilizumab dosing regimen proposed in this study can both maintain the tocilizumab concentration below the empirically established safety threshold and ensure sustained sIL‐6R occupancy (in patients ≥ 30 kg body weight).

The risk of prolonged IL‐6 suppression, immunosuppression, and increased risk of infection following tocilizumab treatment is acknowledged in the approved label. ¹¹ It is recommended that tocilizumab should not be initiated in patients with an absolute neutrophil count < 2,000 per mm ³ , platelet count <100,000 per mm ³ , or with alanine aminotransferase (ALT) or aspartate aminotransferase (AST) levels > 1.5 times the upper limit of normal. For those patients with severe or life‐threatening CRS (who often have cytopenia or elevated ALT/AST as a result of either lymphodepleting chemotherapy or CRS itself), the decision to administer tocilizumab should remain at the physician's discretion and should balance the benefit of treating CRS that has the potential to become life‐threatening vs. the risk of short‐term treatment with tocilizumab.

In this study, patients with R/R B‐NHL receiving tocilizumab for the management of anti‐CD20 bsAb‐induced CRS showed a higher CL and volume of tocilizumab distribution compared with patients receiving tocilizumab for chronic RA. This effect of acute inflammation on tocilizumab linear CL has been previously observed in patients with COVID‐19, where the linear CL estimate (0.725 L/day) was higher than the estimate in adult RA patients (0.2–0.3 L/day), pediatric sJIA patients (0.17 L/day), and pediatric and adult patients with CAR T‐cell therapy‐induced CRS (0.5 L/day). ¹⁸ , ⁴² The increased tocilizumab CL observed in patients with acute inflammatory conditions may be due to the induction of protein catabolism that leads to lower systemic tocilizumab exposure. ¹⁸

Although CRS can significantly affect tocilizumab PK and sIL‐6R parameters, no differences were observed between data collected in the mosunetuzumab and glofitamab clinical trials, suggesting that tocilizumab PK is similar across the two therapies and giving reassurance that the tocilizumab dosing regimen proposed here is suitable for patients treated with either agent. The proposed dosing regimen reflects current tocilizumab usage patterns in trials evaluating both mosunetuzumab and glofitamab, and is included in the Summaries of Product Characteristics associated with their approvals by the European Medicines Agency. ⁴³ , ⁴⁴ Furthermore, this approach is being applied to tocilizumab given as mitigation or as a prophylaxis measure in patients with multiple myeloma treated with cevostamab. ⁶ Given that cancer development is associated with chronic inflammation conditions, ⁴⁵ the current study also suggests that tocilizumab PK may be comparable across different cancer types. Nevertheless, data presented remain limited and additional studies will be needed to support this hypothesis.

In summary, the tocilizumab dosing schedule developed here (up to two 8 mg/kg intravenous doses per CRS event at least 8 hours apart, and no more than three total doses over a consecutive 6‐week period) can be used to effectively manage acute CRS induced by bsAb in patients with R/R‐NHL, while considering safety factors and the unique PK properties and expected activity duration of tocilizumab.

FUNDING

This study was funded by F. Hoffmann‐La Roche Ltd.

CONFLICTS OF INTEREST

C.J. is an employee of and holds shares and stock options in F. Hoffmann‐La Roche Ltd. L.G. has served as a consultant to and received honoraria from F. Hoffmann‐La Roche Ltd and Genentech, Inc. J.‐F.M. is an employee of Syneos Health France. L.L. is an employee of and holds equity and stock options in F. Hoffmann‐La Roche Ltd. D.J.C. is an employee of F. Hoffmann‐La Roche Ltd, and holds equity in AstraZeneca and F. Hoffmann‐La Roche Ltd. J.R. is an employee of and holds equity in F. Hoffmann‐La Roche Ltd, and is a previous equity holder in F‐Star Therapeutics. M.C.W. is an employee of Genentech, Inc., holds equity in F. Hoffmann‐La Roche Ltd, and holds patents/royalties in Genentech, Inc. A.K. is an employee of Genentech, Inc. and is an equity holder in F. Hoffmann‐La Roche Ltd. D.C.T., F.L., L.R.P., and N.F. are employees of and hold equity in F. Hoffmann‐La Roche Ltd. E.G. and S.G. are employees of F. Hoffmann‐La Roche Ltd.

AUTHOR CONTRIBUTIONS

All authors wrote the manuscript; C.J., D.C.T., L.G., N.F., and S.G. designed the research; A.K., D.J.C., E.G., J.R., L.L., L.R.P., and M.C.W. performed the research; C.J., D.C.T., F.L., J.‐F.M., L.G., N.F., and S.G. analyzed the data.

Supporting information

Data S1

DATA AVAILABILITY STATEMENT

For eligible studies, qualified researchers may request access to individual patient level clinical data through a data request platform. At the time of writing, this request platform is Vivli (https://vivli.org/ourmember/roche/). For up‐to‐date details on Roche's Global Policy on the Sharing of Clinical Information and how to request access to related clinical study documents, see here: (https://go.roche.com/data_sharing). Anonymised records for individual patients across more than one data source external to Roche cannot and should not be linked due to a potential increase in risk of patient re‐identification.