Clinical management of bispecific antibody therapy for lymphoma: A British Society for Haematology Good Practice Paper

This guideline was developed according to the British Society of Haematology (BSH) process, as set out on www.b-s-h.org.uk/guidelines. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) nomenclature was used to evaluate the levels of evidence and to assess the strength of recommendations (see www.gradeworkinggroup.org). A literature review was performed using PubMed (Medline) and Cochrane Library. Search terms used were ((lymphoma) OR (NHL) OR (DLBCL) OR (FL)) AND (("bispecific antibody") OR (bispecific) OR (antibody) OR (BsAb)) AND (("T cell") OR (T-cell) OR (BiTE)) AND ((CD3xCD20) OR (CD20xCD3) OR (CD3) OR (anti-CD3) OR (CD20) OR (anti-CD20)) AND ((safety) OR ("adverse event") OR (AE) OR (tolerability) OR (toxicity)). The search was limited to publications with abstracts, publications in English, studies in humans, Cochrane reviews, trials, clinical answers and special collection content types, published between 15 October 2014 and 15 October 2024. Several conference abstracts, study manuscripts and reviews, identified as relevant by the writing group, were also included. The manuscript was reviewed by the BSH Guidelines Committee Haemato-Oncology Taskforce, the Guidelines Committee and the Haemato-Oncology Sounding Board. It was on the members section of the BSH website for comment. The guideline has also been reviewed by patient representatives nominated by the United Kingdom (UK) charity Lymphoma Action (www.lymphoma-action.org.uk); this organisation does not necessarily approve or endorse the contents. The focus of this BSH good practice paper is to provide practical guidance on patient assessment, treatment administration and toxicity management relating to treatment with anti-CD3 × CD20 bispecific antibodies (BsAb) as currently approved by the Medicines and Healthcare products Regulatory Agency (MHRA) and/or European Medicines Agency (EMA). The role of these antibodies in the treatment pathway of large B-cell lymphoma (LBCL) or follicular lymphoma (FL) is beyond the scope of this paper, but the former is covered in the BSH guideline for the management of relapsed or refractory (R/R) LBCL1. BsAb are a novel class of drugs that activate endogenous immune cells by co-targeting tumour antigens and T or natural killer cells in an Fcγ receptor- and major histocompatibility complex-independent manner.2 Anti-CD3 × CD20 BsAb have demonstrated remarkable activity in patients with heavily pretreated B-cell lymphomas.3-8 Epcoritamab, glofitamab, mosunetuzumab and odronextamab have all shown single-agent efficacy in patients with R/R B-cell lymphomas, including LBCL, FL and mantle cell lymphoma.9-22 Current EMA-approved BsAb for lymphoma and the indications include epcoritamab, glofitamab and odronextamab for diffuse large B-cell lymphoma (DLBCL) and epcoritamab, mosunetuzumab and odronextamab for FL, in patients with R/R disease after two or more prior lines of therapy.23-26 Epcoritamab and glofitamab are also approved for R/R DLBCL in third or later lines by the MHRA and recommended for use in the NHS by National Institute for Health and Care Excellence (NICE) and the Scottish Medicines Consortium (SMC) in the United Kingdom. Overall response rates, complete response rate (CR), duration of response (DoR) and duration of CR reported in the pivotal studies for the approved indications are summarised in Table 1. Of note, these are not directly comparable between studies as patient populations are not matched. Across indications, achievement of CR translates into a longer DoR and progression-free survival. DLBCL FL DLBCL FL LBCL 27 ORR: 59% Median DoR: 20.8 months CR: 41% Median DoCR: 36.1 months FL ORR: 83% Median DoR: 21 months CR: 63% Median DoCR: NR LBCL 28 ORR: 52% Median DoR: 18.4 months CR: 40% Median DoCR: 29.8 months FL ORR: 80% Median DoR: 23 months CR: 60% Median DoCR: NR LBCL ORR: 52% Median DoR: 10 months CR: 31% Median DoCR: 18 months FL ORR: 80% Median DoR: 25 months CR: 73% Median DoCR: NR Each cycle is 28 days C1 D1: 0.16 mg (priming dose) C1 D8: 0.8 mg (intermediate dose) C1 D15: 48 mg (3 mg for FL) C1 D22: 48 mg C2+ C3 weekly dosing (D1, 8, 15, 22) of 48 mg C4–9 two weekly dosing (D1 + 15) of 48 mg C10+ four weekly dosing (D1) of 48 mg Each cycle is 21 days C1 D1: pretreatment with 1000 mg of obinutuzumab C1 D8: 2.5 mg over 4 ha C1 D15: 10 mg over 4 ha C2 D1: 30 mg over 4 ha C3–12 D1: 30 mg over 2 ha Each cycle is 21 days C1 D1: 1 mg over 4 h C1 D8: 2 mg over 4 h C1 D15: 60 mg over 4 h C2 D1: 60 mg over 2 hb C3+ D1: 30 mg over 2 hb Each cycle is 21 days C1 D1: 0.2 mg over 4 h C1 D2: 0.5 mg over 4 h C1 D8 and D9: 2 mg over 4 h C1 D15 and 16: 10 mg over 4 h FL dosing post C1 SUD: C2–4 80 mg every week (D1, 8, 15). C2 D1 over 4 h, if tolerated subsequent doses can be given over 1 h C5+ 160 mg over 1 h every 2 weeks DLBCL dosing post C1 SUD: C2–4160 mg every week (D1, 8, 15). C2 D1 over 4 h, if tolerated subsequent doses can be given over 1 h C5+ 320 mg over 1 h every 2 weeks Patients with CR lasting ≥9 months, dosing frequency can be reduced to 160 mg over 1 h for FL or 320 mg over 1 h for LBCL every 4 weeks 24 h at 2 to 8°C Up to 6 h at room temperature (20–25°C) for the 0.2 mg dose with albumin (Human) Up to 12 h at room temperature (20–25°C) for doses of 0.5 mg or greater C1 D15 for DLBCL Not required for FL patients Subcutaneous administration Step-up dosing Premedications (C1 all patients, C2+ patients who experienced grade 2 or 3 CRS with previous dose): Corticosteroid prophylaxis: C1 D1: pretreatment with obinutuzumab Step-up dosing Premedications: (Anti-pyretic + anti-histamine − prior to each dose) IV corticosteroid (C1–3 all patients, C4+ patients who experienced CRS with the previous dose) Step-up dosing Premedications (C1 + 2 all patients, C3+ patients who experienced any grade CRS with previous dose): Step-up dosing Premedications (C1 and C2D1 + D8; C2D15+ patients who experienced any grade CRS or IRR with previous dose): Corticosteroid prophylaxis: Re-priming cycle (identical to C1 with standard CRS prophylaxis) required: After the re-priming cycle, the patient should resume treatment with day 1 of the next planned treatment cycle (subsequent to the cycle during which the dose was delayed) During step-up dosing (C1): After C2 (30 mg dose): If any dose in C1 delayed for >7 days, the previous tolerated dose should be repeated prior to resuming the planned treatment schedule If a dose interruption occurs between C1 and 2 that results in a treatment-free interval of ≥6 weeks, mosunetuzumab should be administered at 1 mg on day 1, 2 mg on day 8, then resume the planned cycle 2 treatment of 60 mg on day 15 If a dose interruption occurs that results in a treatment-free interval of ≥6 weeks between any cycles in C3 onwards, mosunetuzumab should be administered at 1 mg on day 1, 2 mg on day 8, then resume the planned treatment schedule of 30 mg on day 15 Cycle 1: Cycle 2 to maintenance: Tepkinly/ Epkinly Abbvie Columvi Roche Lunsumio Roche Ordspono Regeneron BsAbs are now being investigated in combination with other therapies in R/R LBCL and FL with promising results.29-38 In addition, clinical trials are evaluating the role of BsAbs in patients with newly diagnosed DLBCL and FL.39-43 In common with chimeric antigen receptor (CAR) T-cell therapy, BsAb treatment has a characteristic and predictable toxicity profile mediated by immune activation and subsequently by B-cell depletion. However, the incidence and severity of cytokine release syndrome (CRS) and immune effector cell–associated neurotoxicity syndrome (ICANS)-like neurotoxicity are much lower with BsAb therapy.44 Timing of CRS onset is also more predictable with BsAbs and is summarised in Table 2. CRS usually occurs within 1–2 days of a BsAb dose and lasts for 1–3 days, with 98% of events resolving. It is typically seen during step-up dosing (SUD) in cycle 1, most commonly after the first dose of glofitamab (C1 D8) and after the first full dose of epcoritamab (C1 D15). CRS is sometimes seen in cycle 2 but is uncommon beyond that. The incidence of any grade CRS in the pivotal trials ranged between 39% and 67% with CRS grade 2 in approximately 16% and grade ≥3 in <5%.23-26 Recurrent CRS with further doses was reported in 33%, typically low grade, with the majority being grade 1.23 Grade 1: 31% Grade 2: 17% Grade 3: 3% Grade 4: 0% Recurrent CRS: 33% Grade 1: 50.3% Grade 2: 13.1% Grade 3: 2.8% Grade 4: 1.4% Any grade 39% Grade 2: 14% Grade 3: 2.3% Grade 4: 0.5% For FL Grade 1: 47% Grade 2: 10% Grade 3: 1.3% Recurrent CRs: 32% For DLBCL Grade 1: 35% Grade 2: 16% Grade 3: 0.8% Recurrent CRS: 20% C1 D1: 6.6% C1 D8: 13% C1 D15: 44% C1 D22: 4.6% C2+: 2.8% Any grade of CRS: C1 D8: 54.5% C1 D15: 33.3% C2: 26.8% C3:: 0.9% C4+: 2% Any grade of CRS: C1 D1: 15% C1 D8: 5% C1 D15: 33% C2: 5% C3+: 1% C1 D1 or D2: 24% C1 D8 or D9: 29% C1 D15 or D16: 26% C2: 22% C3+: 4.6% Median time to onset of CRS from the most recent administered epcoritamab dose was 2 days (1–11 days) The median time to onset after the first full dose was 20.2 h (0.2–7 days) Median time to onset of any grade of CRS from start of infusion (and range): 2.5 mg dose: 12.6 h (5.2–50.8 h) 10 mg dose: 26.8 h (6.7–125.0 h) 30 mg dose: 28.2 h (range: 15.0–44.2 h) Median time to onset from start of infusion (and range): C1 D1: 5 h (1–73 h) C1 D8: 28 h (5–81 h) C1 D15: 25 h (0.1–391 h) C2 D1: 46 h (12–82 h) Median time to onset from end of infusion (and range) across all doses: 19.8 h (3.4 h to 9 days) The median time to onset of CRS from the end of infusion (and range) for: C1 D1 or D2: 6 h (2.4 h to 4 days) C1 D8 or D9: 22 h (3.7 h to 5 days) C1 D15 or D16: 22 h (3.4 h to 9 days) Median duration for any grade of CRS: 2.5 mg dose: 31.8 h (0.5–316.7 h) 10 mg dose: 16.5 h (0.3–109.2 h) 30 mg dose: 18.9 h (range: 1.0–180.5 h) Grade 1: 4.2% Grade 2: 1.2% Grade 3: 0% Grade 4: 0% Grade 5 (fatal): 0.6% Grade 1–2: 5% Grade 3–4: 3% Grade 1–2: 2% Grade 3: 0.1% Grade 4: 0% ICANS-like neurotoxicity17 Any grade: 12% Grade 3+: 3% Neurological toxicity of any grade: headache (13%), dizziness (8%), anxiety (4.3%) and confusional state (3.5%), and encephalopathy (3%) Neutropenia 31% Grade 3–4 neutropenia 23% Neutropenia 40% Grade 3–4 neutropenia 29% Neutropenia 28% Grade 3–4 neutropenia 24% Neutropenia any grade 25% Grade 3–4: 19% Any grade: 44% Grade 3 in 27% Grade 4 in 2.6% 3% (all grade 2) Median time to onset: 17 days (range: 9–34 days) Median duration: 15.5 days (range: 1–50 days) 11.7% any grade 4.8% (grade 2) 2.8% (grade 3) Median time to onset of tumour flare of any grade: 2 days (range: 1–16 days) Median duration: 3.5 days (range: 1–35 days) 4% any grade 1.8% (grade 2) 2.3% (grade 3) Median time to onset: 13 days (range: 5–84 days) Median duration: 10 days (range: 1–77 days) ICANS and ICANS-like symptoms are much less common with BsAbs than with the currently available CAR T-cell products. ICANS incidence rates, median time to onset in relation to dose and duration are summarised in Table 2. The frequency of any grade ICANS ranged between 0% and 8% in the pivotal trials, with most events being grade 1 and self-limiting. With glofitamab, any grade ICANS-like neurological events were reported in 8% with grade ≥3 events in 3%.12 With epcoritamab, any grade ICANS was reported in 6% (grade 1 in 4.2%, grade 2 in 1.2% and grade ≥3 in <1%).25 Although the range for onset of ICANS is much broader than for CRS, median time to first ICANS is typically within 2–3 days of the most recent dose with glofitamab and the first full dose with epcoritamab and mosunetuzumab during cycle 1 SUD. ICANS resolved in 94% with supportive care. The median time to resolution of ICANS was 2–3 days (range: 1–20 days). To date, there is limited published real-world experience of BsAb in R/R LBCL. Initial review of UK real-world experience (RWE) demonstrated that the safety profile appears similar to published data, with 33% of patients experiencing CRS, 15% grade ≥2, with 17% of patients requiring tocilizumab, ICANS in 4%, and 5% requiring intensive care unit (ICU) admission.45 However, response rates were inferior probably reflecting selection of patients with highly aggressive disease kinetics and poor performance status in the real-world setting. Continuing work with RWE collection is required to identify patients unlikely to benefit from BsAb. Strategies to mitigate against these toxicities such as SUD and corticosteroid premedication are routinely incorporated into BsAb administration protocols. Importantly, another notable toxicity of anti-CD3 × CD20 BsAb therapy is B-cell aplasia and increased risk of infections.46 Although CAR T-cell therapy has revolutionised outcomes for many patients with R/R B-cell lymphomas, access is limited due to logistic, geographic and resource constraints. In the UK, European Union and many other territories, CAR T-cell infusion is restricted to selected large haematopoietic stem cell transplantation centres with Foundation for the Accreditation of Cellular Therapy (FACT) or The Joint Accreditation Committee ISCT-Europe & EBMT (JACIE) accreditation. Patients living at a distance from the CAR T-cell delivery centre, those from lower socioeconomic status and minority communities are less likely to receive CAR T-cell therapy.47-50 The manufacturing time for CAR T-cell therapy and risk of manufacturing failure are additional limiting factors, with up to 25% of approved patients not receiving infusion (largely due to clinical deterioration as a result of disease progression while awaiting CAR T-cell manufacture), although the UK experience has shown an improvement in the proportion of infused patients over time.51-54 By contrast, BsAb therapy has the advantage of being 'off-the-shelf' with a manageable toxicity profile and can be delivered in smaller local hospitals without a need for FACT/JACIE accreditation. However, it should be recognised that CAR T-cell therapy has longer follow-up data than BsAb, with a proportion of patients likely to experience cure, whereas the long-term efficacy for BsAbs is yet to be ascertained. An important step to the adoption of BsAb therapy is developing robust pathways for the administration and management of toxicities relating to these products. According to capacity and local circumstances, some smaller hospitals may work in partnership with their nearest referral centre for administering SUD during cycle 1 when there is the highest risk of CRS. The focus of this good practice paper is to provide practical guidance on introducing BsAbs into a clinical service, treatment administration and toxicity management relating to treatment with current EMA and/or MHRA approved anti-CD3 × CD20 BsAb. The role of these antibodies in the treatment pathway of R/R LBCL or FL is beyond the scope of this paper. Patients being considered for BsAb therapy should have an assessment of performance status and baseline blood tests to include full blood count, routine chemistry, serum lactate dehydrogenase, C-reactive protein (CRP) and serum immunoglobulin (Ig) levels. A blood film and flow cytometry should be performed where circulating neoplastic cells are suspected. Full serology for hepatitis B (including core antibody), hepatitis C and human immunodeficiency virus is also mandatory unless performed within the preceding 12 months. Formal investigation to assess cardiac function is not routinely required but should be considered in patients with known or suspected cardiac morbidity and in older patients with previous anthracycline exposure where compromised cardiac function confers additional risks in the context of CRS. A referral to cardiology or a cardio-oncology service should be made for patients with impaired cardiac function. A positron emission tomography/computed tomography (PET-CT) scan prior to starting treatment is recommended as it provides an indication of tumour burden and a baseline for subsequent response assessment. High tumour burden is predictive for increased risk of CRS and inferior outcomes.55 Biopsy of representative tissue for histological confirmation of disease is recommended unless it poses an unacceptable risk. While assessing CD20 expression by immunohistochemistry prior to initiating BsAb therapy is not essential and clinical trials did not routinely require demonstration of the target antigen for patient enrolment, emerging data suggest that CD20 expression does correlate with clinical response and weak or no immunohistochemical expression can be associated with suboptimal response and development of resistance.56, 57 Therefore, caution should be exercised when considering an anti-CD3 × CD20BsAb in the absence of histological CD20 expression. Routine baseline evaluation of handwriting or formal assessment of Immune effector Cell-associated Encephalopathy (ICE) score is not required but may be considered where there is a concern for impaired cognition. An important aspect of developing new therapeutic pathways is education and awareness for all healthcare professionals (HCPs) participating in the care of patients receiving BsAb therapy, including those involved in acute admission pathways of the treatment facility where a patient might present with BsAb toxicity. It is recognised that, in smaller treatment centres, emergency admission will not be directly to a haematology or oncology unit, but HCPs involved in the admission pathway must be cognisant of CRS and ICANS as possible side effects of BsAb treatment and should know who to contact for urgent advice. While it is useful to have ICU physician engagement, the need for ICU utilisation is low, with only a 5% ICU admission rate in the UK RWE.45 Tocilizumab, a monoclonal antibody that blocks the interleukin-6 receptor, is commonly used for the treatment of CRS. A minimum stock of two doses (4 × 400 mg vials) should be routinely and readily available on the haematology ward/emergency medication cupboard. Staff must be aware of the location of tocilizumab within their hospital and have rapid access to it at all times. Pharmacy must devise a system for replenishing tocilizumab stock, and any stock issues should be communicated to the haematology team promptly. It is desirable for each treatment centre to have a team of HCPs who lead on training for staff involved in the management of patients receiving BsAb therapy and providing information for patient and carer education. A comparison of the currently approved anti-CD3 × CD20 BsAb, their administration schedules, management of dose delays, re-priming and CRS mitigation strategies is shown in Table 1. HCPs are also directed to refer to the summary of product characteristics (SmPC) for each product.23-26 There are common themes between treatment schedules of the different BsAb but also notable differences. Epcoritamab is a subcutaneous injection, whereas glofitamab, mosunetuzumab and odronextamab are given by intravenous (IV) infusion. Treatment centres must be familiar with these and assess the impact on their services, including staffing resource, pharmacy aseptic unit capacity, day unit capacity and any mandated in-patient admission requirement. Given aseptic preparation with some BsAb can be demanding, early communication with the hospital's pharmacy aseptic unit is crucial. It is recommended that systemic anti-cancer treatment (SACT) prescriptions are done in advance to ensure enough stock and trained pharmacy staff are available to prepare the BsAb given the short 24–72 h expiry once reconstituted and complex aseptic preparation for certain BsAb (e.g. serial dilution for epcoritamab). Strategies to mitigate against CRS and ICANS such as SUD and steroid premedication are routinely incorporated into BsAb administration protocols as reflected in the respective SmPC. Accelerated SUD strategies aimed at minimising the time taken to reach full therapeutic dose are being explored and seem deliverable without increased risk of CRS or ICANS.58 Additionally, BsAb-specific strategies such as subcutaneous administration for epcoritamab and obinutuzumab pretreatment to deplete circulating B cells prior to glofitamab are designed to reduce the risk of CRS. Corticosteroid prophylaxis is essential during SUD, and dexamethasone appears more effective at reducing the incidence and severity of CRS compared to other corticosteroids.59, 60 Steroid premedication may need to be continued in subsequent cycles depending on individual product guidance and if prior CRS has occurred. Hydrating patients prior to dosing has also been reported as a potential CRS mitigating strategy with epcoritamab.59 The maximum risk period for CRS from treatment initiation varies between the products. An elective admission for a minimum of 24 h following BsAb administration is currently mandated for CRS monitoring during SUD in cycle 1 (C1 D8 for glofitamab and C1 D15 for epcoritamab). For odronextamab, patients in the ELM-2 study were admitted for inpatient monitoring for at least 24 h after each infusion during SUD, that is, C1 D1, C1 D2, C1 D8, C1 D9, C1 D15, C1 D16 and C2 D1. Where patients are hospitalised for monitoring, it is recommended that observations are checked at regular intervals (of at least 4 h). In the absence of CRS, patients can usually be discharged 24 h post-dose. Given that neurotoxicity is infrequent, there is no requirement to perform routine ICE and ICANS evaluations unless there is clinical concern or CRS > grade 2, when ICE scores should be done at regular intervals of 4 h. Given that there is significant steroid exposure during the SUD phase, it is important to pay attention to blood glucose control and its management in diabetic or prediabetic patients. Patients and carers should be given appropriate information on CRS and ICANS, symptoms to look out for and advice on self-monitoring. Figure 1 provides a template that can be adapted by individual centres. It is recommended that patients are instructed to monitor their temperature at home three times per day for 48 h after each SUD until one full dose is administered without CRS. Patients must be provided with a SACT red alert card with instructions on when and whom to call, and when to present directly to the hospital. It is recommended that patients are also provided with a BsAb alert card or a wristband with contact information for the haematology team. For selected high-risk patients, an additional phone call from a HCP at 24 and 48 h after each SUD may be beneficial.44 High-risk patients are those with a high tumour burden, impaired organ function (e.g. impaired renal or cardiac function), older patients (e.g. age >75 years) and those with limited social support. For patients with multiple risk factors, consider in-person clinic review with an HCP at 24 and 48 h after each SUD or even hospitalisation for up to 48 h post-dose. In contrast to CAR T-cell therapy, there is no mandatory requirement for a carer to live with the patient following BsAb therapy.44 A summary of reported frequency of CRS, ICANS and other toxicities is shown in Table 2. CRS is the most frequent acute toxicity mediated by T-cell activation with anti-CD3 × CD20 BsAb therapy. It occurs most frequently during cycle 1 SUD, is usually low grade and has a predictable onset in relation to treatment administration. The most common manifestations of CRS are fever of >38°C, tachycardia, hypotension, chills and hypoxia. Infusion-related reactions (IRR) may be clinically indistinguishable from manifestations of CRS.23-26 Patients should be evaluated for other causes of fever, hypoxia and hypotension, such as infection or sepsis. Grading of both CRS and ICANS is performed as per the American Society for Transplantation and Cellular Therapy (ASTCT) consensus criteria, originally developed for CAR T-cell therapy.61 A recommended algorithm for CRS management for hospitalised patients is shown in Figure 2. Although there are subtle differences in the recommendations for CRS management between product SmPCs, the principles are similar, allowing for a consistent approach applicable to all BsAb in lymphoma. Grade 1 CRS is typically managed with symptomatic measures. Grade 1 CRS persisting for >24 h should be treated with tocilizumab or dexamethasone. Tocilizumab is recommended for patients developing CRS while taking steroid prophylaxis. Earlier use of tocilizumab is recommended for patients with high tumour burden, as they are at an increased risk of higher grade CRS55 and for those with reduced physiological reserve, for example, impaired cardiac or renal function. Given the clinical similarities of CRS and infection, a full septic screen is recommended. Initiate broad-spectrum IV antibiotics, especially for neutropenic patients and/or for CRS grade ≥2. Tocilizumab is recommended for CRS grade ≥2. It is important to note that fever and CRP are unreliable markers of CRS and infection once patients are treated with anticytokine therapy such as tocilizumab. If hypotension or hypoxia persists, consider ongoing CRS and manage accordingly. Always exclude alternative causes for the clinical manifestations. CRS grades 3–4 are very uncommon with BsAb therapy and require admission to a critical care unit. For any grade CRS, treatment with BsAb should be withheld until CRS is completely resolved for at least 72 h. If there is a delay to the next dose, repriming may be required (see Table 1 and SmPC of the product). For patients experiencing CRS grade ≥2, elective admission for monitoring and continuation of steroid prophylaxis are recommended for the next dose. As with the risk of neutropenic sepsis following intensive chemotherapy regimens, selected patients may also benefit from elective readmission for other reasons; for example, if socially isolated and if there is concern about the ability to seek help in the event of CRS at home. As elective hospital admission is currently mandated with most BsAb therapy at defined time points during SUD periods posing the highest risk for CRS, many CRS episodes happen and are consequently managed in the inpatient setting. However, not infrequently, CRS happens outside of this mandated admission period when a patient is at home. The usual recommended protocol for such situations is for patients to attend hospital for immediate assessment. Given the predictable and often low-grade nature of CRS, some institutions are exploring remote management pathways for carefully selected patients to reduce emergency admissions. These pathways require rigorous patient selection, comprehensive patient education and clear protocols for accessing urgent care. Patients with high tumour burden, organ dysfunction, advanced age, limited social support or poor adherence may not be suitable for remote management pathways. Figure 3 shows a potential algorithm for such a pathway. Hospitals offering this home management pathway for CRS should consider supplying patients with a 10-mg dose of dexamethasone to be taken if they develop a fever persisting after a dose of paracetamol. Neurological toxicity seen with BsAb therapy most commonly manifests as headache and dizziness. ICANS-like neurotoxicity with word-finding difficulty, dysgraphia or acute confusional state is almost always low grade; more severe neurotoxicity is rare. The risk of ICANS is highest in cycle 1, but delayed onset may sometimes be seen. If ICANS is suspected, a thorough neurological examination should be performed, including an ICE assessment. ICE scoring and ICANS grading is shown in Figure 4. ICANS grade 1 can be treated with oral dexamethasone. ICANS grade 2 needs treatment with dexamethasone IV and a prophylactic antiepileptic such as levetiracetam. Other causes of neurotoxicity must be excluded, for example, infection, opioid toxicity, medications and metabolic causes. Grades 3 and 4 are rare, should be managed in the ICU setting and alternative aetiologies should also be considered. Dexamethasone IV and anakinra should be used to manage grades 3 and 4 ICANS, but some patients may require high-dose methylprednisolone and siltuximab. A recommended algorithm for ICANS grading and management is shown in Figure 5. IRR manifestations may be clinically identical to CRS but usually occur earlier. If IRR/CRS occur during BsAb infusion, the infusion should be interrupted and symptoms treated. Once the symptoms have resolved, the infusion may be restarted either at the same or slower rate depending on the BsAb, but permanent discontinuation may be required in the case of grade 4 IRR (refer to product SmPC for details). If CRS is suspected, stop the infusion and follow CRS management protocol. Tumour flare reaction (TFR) is a recognised event with BsAb. It manifests as a transient increase in the size of the tumour, sometimes accompanied by pain or fever and may cause compression symptoms if a large tumour is adjacent to a vital structure such as large airways or the superior vena cava. Breathlessness due to new or worsening pleural effusions is also reported. Reported incidence of TFR ranged from 1.6% with epcoritamab (all grade 2) to 11.7% with glofitamab (grade 2 in 4.8% and grade 3 in 2.8%).23-25 Most TFR events occur during cycle 1 with none reported beyond cycle 2. Therefore, the possibility of TFR may need to be considered in some cases when interpreting PET scans during the first two cycles of BsAb therapy. No specific risk factors have been identified for TFR; however, caution is advised in patients with bulky tumours located in close proximity to airways and/or a vital organ. Analgesics and corticosteroids can be given to patients with tumour flare reaction. BsAb treatment does not confer an excess risk of tumour lysis syndrome (TLS) beyond that seen with conventional LBCL therapies. Reported incidence is around 1%–1.5%. Most TLS events occur during cycle 1, with time to onset from treatment initiation varying between BsAb (median of 2 days with glofitamab, 18 days with epcoritamab and on days 2 and 24 with mosunetuzumab).23-26 Prophylaxis, monitoring and treatment are covered by the BSH good practice paper for TLS and should follow standard local/national protocols. In the pivotal trials, the incidence of grade ≥3 neutropenia was reported to range between 19% and 29% and usually responds to granulocyte colony-stimulation factor (G-CSF) support. G-CSF is recommended for all patients with grade 4 neutropenia (<0.5 × 109/L) and should be considered for grade 3 neutropenia (<1.0 × 109/L), especially if it is persistent or worsening. The incidence of febrile neutropenia was 2.1% with glofitamab and 2.4% with epcoritamab.23, 25 Treatment discontinuation due to neutropenia is uncommon but should be considered if grade ≥3 neutropenia associated with infections occurs and does not respond to G-CSF support. In patients remaining well despite neutropenia, it may be optimal to continue BsAb therapy and monitor closely. Infections (not necessarily associated with neutropenia) were reported in 38%–59% of patients with grade ≥3 infections in 15%–33%. Commonly reported severe infections were COVID-19 (32%), clinically diagnosed pneumonia (26%) and sepsis (12%). Severe opportunistic infections reported infrequently include reactivation of herpes simplex virus/varicella-zoster, cytomegalovirus and Epstein–Barr virus as well as severe influenza, toxoplasmosis, Pneumocystis jirovecii pneumonia (PJP) and fungal pneumonia.46 BsAb treatment should be withheld in the presence of active infection. In patients with asymptomatic COVID-19, it may be appropriate to continue BsAb treatment after assessing risk versus benefit on a case-by-case basis.62 Varicella zoster virus (VZV) and PJP prophylaxis are recommended throughout BsAb treatment and should be continued until the CD4 count has recovered to >0.2 × 109/L. Antifungal prophylaxis is not routinely required. The optimal timing and efficacy of vaccinations are yet to be established in these patients, but live vaccines should be avoided. However, it is recommended patients have all routine non-live vaccinations including those for COVID-19, influenza and varicella zoster (where available).44 Anti-CD3 × CD20 BsAb causes profound and prolonged B-cell aplasia and hypogammaglobulinaemia. On continuous therapy with epcoritamab, serum IgG levels dropped by a median of around 20% and remained stable over time.27 With fixed duration glofitamab therapy, recovery of B-cell aplasia and IgG levels was observed starting from around 18 months after the end of treatment.28 A persistent IgG level of <4 g/L is associated with a higher risk of infection. Consider immunoglobulin replacement therapy (IGRT) in patients with recurrent or severe infections despite antimicrobial prophylaxis with IgG levels of <4 g/L. There remains concern that patients initiating BsAb therapy for progressive disease (PD) following CAR T-cell therapy may develop more severe B-cell aplasia and hypogammaglobulinaemia, although further data on this are needed from studies in the real-world setting. Where the patient is in a durable remission, chronic low-level adverse effects such as infections and treatment fatigue can be a challenge with continuous regimens for some patients. For those in a sustained CR, it may be reasonable to consider stopping treatment where the benefits of continuing are outweighed by cumulative toxicity. Current available data suggest that CRs are durable in a proportion of patients even after stopping epcoritamab therapy.21 In pivotal clinical trials, response assessment scans were performed at 6 and 12 weeks from treatment initiation. The majority of responses are seen within this timeframe. In routine clinical practice, a pragmatic approach would be to perform a PET-CT scan after cycle 3 unless there is clinical concern for disease progression, recognising that many patients will not respond. It is important not to prematurely conclude stable disease or PD response on PET scans, especially if performed early after initiating BsAb therapy. TFR is reported with BsAb therapy, is usually early in onset and should be differentiated from true progression. If a patient remains clinically well, it is generally preferred to continue treatment and rescan or biopsy before discontinuing treatment. For patients in a complete metabolic remission (CMR) after three cycles, a repeat PET scan after cycle 12 can be considered to document ongoing response status. For those not in a CMR after three cycles, a further PET-CT scan after another three cycles and thereafter as clinically indicated is recommended. Sridhar Chaganti chaired the writing group. All authors were involved in writing the first draft of the manuscript and in all subsequent revisions. All authors approved the final version for publication. The authors wish to thank Sherwin Barretto PhD, Niche Science & Technology, for help in undertaking the initial literature review. This literature search was funded by BSH. The BSH Haem-Onc task force members at the time of writing these guidelines were Nilima Parry-Jones (Chair), Toby Eyre (Deputy Chair), Dima El-Sharkawi, Rob Sellar, Tracey Chan, Matthew Wilson, Austin Kulasekararaj, Andrew Clark, Joanne Ewing, Andrew McGregor, David Allsup, Faisal Basheer, Kevin Boyd, Christopher Parrish and Daniel Hughes (Trainee member). The authors would like to thank them, the BSH sounding board and the BSH guidelines committee for their support in preparing this guideline. The BSH paid the expenses incurred during the writing of this guidance. All authors have made a declaration of interests to the BSH, which may be viewed on request. Members of the writing group will inform the writing group chair if any new evidence becomes available that would alter the strength of the recommendations made in this document or render it obsolete. The document will be reviewed regularly by the relevant task force and the literature search will be re-run every 5 years to search systematically for any new evidence that may have been missed. The document will be archived and removed from the BSH current guidelines website if it becomes obsolete. Please check the BSH guidelines website (www.b-s-h.org.uk/guidelines) for any addenda that may be produced after the initial publication. While the advice and information in this guidance are believed to be true and accurate at the time of going to press, neither the authors, the BSH nor the publishers accept any legal responsibility for the content of this guidance.