T-cell-based Immunotherapies for Haematological Cancers, Part A: A SWOT Analysis of Immune Checkpoint Inhibitors (ICIs) and Bispecific T-Cell Engagers (BiTEs)

Haematology has been at the vanguard of cancer immunotherapy. Immune checkpoint inhibitors (ICIs), bispecific T-cell engagers (BiTEs), allogeneic haematopoietic stem cell transplantation (allo-HSCT) and donor lymphocyte infusion (DLI), as well as adoptive T-cell therapies outside the setting of allo-HSCT, have been approved for distinct haematologic malignancies producing durable responses in otherwise untreatable patients. Despite recent advances, immunotherapies do not benefit most patients, due to resistance or lack of response, and are only approved in specific settings. Moreover, immunotherapies are expensive and may produce severe immune related adverse reactions. Combination therapy complicates the picture and requires further evaluation. This review considers the current status and future perspectives of ICIs and BiTEs approved for haematological malignancies by analysing their strengths, weaknesses, opportunities and threats (SWOT). The biological rationale for anti-cancer mechanisms, clinical data for specific haematological cancers, efficacy, toxicity, response and resistance profiles, novel strategies to improve these characteristics as well as the potential targets to enhance or expand the application of ICIs and BiTEs are also discussed.

Cancer immunotherapy has revolutionised oncology care, prolonging survival in rapidly fatal diseases. The number of patients eligible for immune-based cancer treatments is increasing, with immunotherapies being adopted in first line setting (1). Novel targets and combination therapies are set to expand cancer immunotherapy applications. Haematology has been central to these advances.
Allogeneic haematopoietic stem cell transplantation (allo-HSCT) was the first clinical application of cancer immunotherapy (1957), while monoclonal antibodies (mAb) were the next success story with the approval of rituximab (anti-CD20 mAb) for B-cell malignancies (1997). These breakthroughs contributed valuable advances to the evolution of cancer immunotherapies. Immune checkpoint inhibitors, developed through mAbs, target T-cells and upregulate anticancer immunity, producing remarkable success in solid and haematologic malignancies. Bispecific T-cell engager (BiTE) antibodies, which redirect T-cells to tumour cells to perform target cell killing, were originally approved for B-cell precursor acute lymphoblastic leukaemia (BCP-ALL), with blinatumomab gaining approval in 2014. Development of novel adoptive T-cell (ATC) therapies in haematology, such as chimeric antigen receptor (CAR)-T-cells, has generated great interest with potential for disease cure.
Yet, despite these advances, several challenges remain. Limited breadth of application, unpredictable efficacy, and limiting toxicity profiles attest the need to drive forward change. This review discusses the strengths, weaknesses, opportunities and threats (SWOT) associated with immune checkpoint inhibitors (ICIs) and BiTEs, providing an up-to-date review of licensed agents for haematological malignancies. The biological rationale for anti-cancer mechanism; clinical data in specific haematological cancers; efficacy, toxicity, response and resistance profiles; novel strategies to improve these characteristics; and potential targets to enhance or expand the application of these agents are discussed.
Reduced T-cell proliferation and cytokine secretion results.

Strengths of ICIs.
Responses in heavily pre-treated/resistant disease. As with other immunotherapies, a major advantage of ICIs is their ability to achieve a response in heavily pre-treated relapsed or refractory patients which is a testament to their therapeutic potency. Indeed, Marjanska et al. demonstrate the efficacy of nivolumab in heavily pre-treated paediatric patients including one patient with stage IV cHL who achieved CR with no significant ARs (58). Among patients with platinumrefractory, recurrent squamous-cell carcinoma of the head and neck, nivolumab produced longer OS and resulted in longer overall survival than treatment with standard, singleagent therapy (methotrexate, docetaxel, or cetuximab) (59).
Durable response. A hallmark of cancer immunotherapy is the durability of response that translates into clinical benefit (60). ICIs can potentially sustain the anti-tumour immune response indefinitely (61) due to T-cell immunologic memory (62)(63)(64).
Relatively well-tolerated. ICIs are tolerated better than chemotherapeutics and do not induce severe myelosuppression or sepsis, thus improving quality-of-life.
Slow response, pseudoprogression, and hyperprogression. ICIs demonstrate different patterns of kinetics and disease progression to chemotherapeutics, producing an initial "tumour flare" termed pseudoprogression. Pseudoprogression (~10%) is a radiologically observed increase in tumour size due to T-cell infiltration, mimicking progressive disease (PD), though remission follows.
Pseudoprogression challenges clinical decision making. Immune-specific related response criteria have been developed (66). These define PD differently, allowing for treatment  beyond initial progression, thus avoiding inappropriate early treatment discontinuation indicated by conventional Response Evaluation Criteria in Solid Tumours (RECIST).
Hyperprogression describes rapid PD after immunotherapy, corresponding to tumour growth and reduced survival. Hyperprogression is observed in 4-29% of ICI-treated solid tumours (67). Predictive markers for progression, hyperprogression and pseudoprogression are needed.
Aside from primary resistance, adaptive and acquired resistance may emerge. Acquired resistance rates vary between tumours. In cHL acquired resistance rates range between 19-57% in studies [reviewed in (61)]. An inverse relationship between PD-1 therapy ORR and acquired resistance indicates disease-specific acquired resistance mechanisms, through this association is absent in cHL (61). Resistance mechanisms and response prediction. In HL, primary ICI resistance mechanisms include: CD8 + T-cell exclusion and increased Tregs within the TME; insufficient T-cell activation by lack of antigen presentation; upregulated indoleamine 2,3-dioxygenase (IDO) metabolism; and augmented immunosuppression via tumour-associated macrophages (TAMs) or natural killer (NK) cells. Acquired resistance occurs via PD-1, LAG-3 and TIM upregulation following anti-PD-1 therapy and increased adenosine levels, though this may present in primary resistance (68,70). IDO, LAG-3 and TIM inhibitors are being investigated in solid tumours as a means to overcome resistance (71)(72)(73). Potential ICI response prediction biomarkers correlating to resistance mechanisms have been identified in solid tumours (74,75). High tumour mutation burden and PD-L1 expression have been shown to be independent predictors of ICI efficacy in various solid cancers according to one meta-analysis (76). Another meta-analysis showed that EGFR mutations are a negative predictor of ICI efficacy in NSCLC (77), while a single-centre study showed that KRAS mutations predict superior response to immunotherapy in NSCLC (78). Differential tumour infiltrating lymphocyte density in metastatic and primary tumour sites could also contribute to ICI response prediction according to one case report (79). PD-L1/2 expression is potentially a novel prognostic predictor according to a recent study (80).
ICI combination therapy with other immunotherapies. Combination therapy decreases resistance rates and improves efficacy. In HL, ipilimumab plus nivolumab demonstrated 74% ORR (81) while this concurrent combination has also demonstrated rapid and deep tumour regression in advanced melanoma with manageable safety profile (82). Combination therapy may enhance the efficacy of developing agents. PD-L1/PD-1 blockade with CD33/CD3 BiTE enhanced T-cell proliferation and IFN-γ production (83). Benefits are also demonstrated with CAR-T cells.
ICI combination therapy with radiotherapy. Radiotherapy promotes tumour-specific antigen presentation. The abscopal effect is a systemic immune mediated regression of nonirradiated lesions distant from the primary irradiation site (94). Preclinical evidence supports combination of stereotactic body radiation therapy (SBRT) with ICIs (95). Radiotherapy sequencing and fractionation alters responses. Radiotherapy combination with anti-CTLA-4 only produced an abscopal effect with fractionation (96). Abscopal response is facilitated via altered antibody response to TAA, modified peripheral blood immune cells, and increased antigen responsiveness (97). Increased tumour-infiltrating lymphocytes were observed in non-irradiated lymph nodes of patients treated with ICI plus radiation (98).

Threats to ICIs.
Despite their clinical efficacy, ICIs' spectrum of use is narrow in haematological malignancies, whereas their role is more prominent in solid tumours. Even though ICIs are a cost-effective, off-the-shelf immunotherapy with universal utility across patients, their use in haematological malignancies will likely be overshadowed by the advent of more novel immunotherapeutic approaches, for example, Tcell-redirecting immunotherapies and adoptive cell therapies.

Bispecific T-cell Engagers (BiTEs)
Biological rationale for anti-cancer mechanism and development.
The concept of selectively targeting tumours via antibodies was proposed over a century ago by Paul Ehrlich (99). Monoclonal antibodies (mAbs) have constituted a weapon in the oncologists' anti-cancer armamentarium since 1997, commencing with the approval of rituximab, a chimeric anti-CD20 agent for low-grade B-cell lymphoma (100). Improvements in antibody engineering technology (101) have enabled scientists to develop bispecific T-cell-redirecting antibodies which bind TAAs, redirecting cytotoxic T-cells to tumours. Bispecific T-cell engagers (BiTEs), consisting of two different single-chain variable fragments (scFvs) derived from the antigen-binding domains of anti-CD3 and anti-TAA antibodies covalently bound via small linker peptides (102).
BiTEs are producible in large quantities through mammalian cell line culture and recombinant single-chain polypeptide secretion (103). Upon simultaneous binding of BiTEs to TAAs and CD3 TCR, a lytic immune synapse forms between T-cells and cancer cells (Figure 4) (104). Simultaneous high affinity binding is facilitated by small size (~55 kDa) and high flexibility of BiTEs due to lack of antibody constant (Fc) regions, which also contributes to reduced half-life and decreased toxicity due to lack of Fc-receptor (FcR) recycling and FcR-mediated effector functions, respectively (105). These characteristics are crucial for in vitro and in vivo efficacy. Tcell-mediated tumour cell killing was observed at very low concentrations (10-100 pg/ml) and low effector-cell to T-cell ratios (<1:90), without immune co-stimulation (106)(107)(108). Hence, blinatumomab ultimately became the first approved BiTE and is still used in ALL subtypes. Several trials are investigating BiTEs in haematological and solid malignancies (109). Mosunetuzumab was recently pre-approved for R/R follicular lymphoma (FL).
Clinical translation. Blinatumomab (Blincyto ® ). In 2014 blinatumomab, an anti-CD19 and anti-CD3 agent, was approved for Ph chromosome (Ph)-negative (110), and subsequently Phpositive (111), B-cell precursor acute lymphoblastic leukaemia (BCP-ALL); nearly 60 years after the first report of human-synthesised bispecific antibodies (112,113). Blinatumomab approval has been expanded to adults and children with R/R BCP-ALL, and adults and children with MRD-positive BCP-ALL in remission.
FDA approval was based on a phase II, open-label, singlearm, multicentre trial of R/R Ph -BCP-ALL patients in which 33% CR and 10% CRh (complete remission with partial haematological recovery) was achieved, with 6.9-month median OS (114). A phase III randomised trial reported similar results with 44% CR/CRh rate and 7.7-month median OS. Blinatumomab was superior to chemotherapy (115). The openlabel, multicentre, single-arm study granting blinatumomab approval for MRD + BCP-ALL evaluated patients experiencing first-or second-time CR with detectable MRD in a >1 in 1,000 bone marrow cells (116). MRD conversion to <0.01% after one blinatumomab cycle was achieved in 85.2% and 72.0% of firstand second-CR patients, respectively, with 35.2-month and 12.3-months median haematologic relapse-free survival, respectively, indicating durable response. At 5 years, 50% remained in remission (117). A phase I/II trial was the first to demonstrate the safety and efficacy of single-agent blinatumomab in paediatric patients with R/R BCP-ALL achieving complete minimal residual disease response (118).
Blinatumomab is generally well-tolerated. Toxicities reflect CD3-activation (109). Cytokine release syndrome (CRS) and neurotoxicity were rare but severe dose-limiting ARs issued Boxed Warnings in addition to pancreatitis, serious infection and sepsis (114,115).
Mosunetuzumab (BTCT4465A). In July 2020, mosunetuzumab, an anti-CD20 and anti-CD3 agent, received pre-approval by the FDA, through breakthrough therapy designation, for the treatment of adults with R/R follicular lymphoma (FL) after at least two prior systemic therapies. The phase I/Ib, multicentre, open-label, dose-escalation study evaluated the safety and pharmacokinetics of mosunetuzumab in 270 heavily pre-treated R/R NHL (119). ORR and CR was observed in indolent (63% and 43%) and aggressive lymphomas (37% and 19%) across doses. CRS and neurological ARs occurred in 29% and 44%, respectively, but only three were grade 3 cases in each. Notably, patients previously receiving CAR-T-cell therapy (n=18) achieved 39% ORR (n=7) and 22% CR (n=4) (120). Investigators observed a lymphocyte expansion, including residual CAR-T-cells, and CRs with and without CAR-T-cell expansions offering potential for mosunetuzumab salvage therapy after CAR-Tcells, though it could potentially be a bridging approach as well by stimulating T-cells.

Strengths of BiTEs.
Superior anti-tumour efficacy. Preclinically, BiTE efficacy is superior to mAbs and other bispecific antibodies (106,121). Higher binding specificity due to two antigens and effector immune cell mediated redirection to tumour cells enhance cytotoxicity. Targeting two pathways improves efficacy and resistance (122). Indeed, in a phase I/Ib study mosunetuzumab demonstrated clinical activity and durable response in R/R Bcell NHL patients who were considered refractory to anti-CD20 therapy and in patients who had relapsed following CD19-directed CAR-T therapy, while the safety profile also appeared favourable compared to standard anti-lymphoma therapies including T-cell directed agents (123).
Lack of MHC and HLA restriction. Lack of major histocompatibility complex (MHC)-and human leukocyte antigen (HLA)-restriction in BiTEs allows for universal offthe-shelf use unlike CAR-T-cells. MHC-independent cancer elimination prevents resistance via MHC-molecule downregulation, loss of MHC-I associated β2-microglobulin or intracellular peptide transporters (124).
Blinatumomab-treatment neurotoxicity (52%) encompasses a spectrum of presentations (109). Most are mild such as grade 2 tremor (17%), while seizures (2%) or encephalopathy (5%) are rare. BiTE neurotoxicity has been attributed to extravasation of adhesive T-cells to the perivascular space in the central nervous system, stimulating endothelial activation which attracts leukocytes, including monocytes, inducing neuroinflammation and neurotoxicity (135). Neurological ARs are reversible on treatment discontinuation and dexamethasone therapy. Similar to CRS, prophylactic dexamethasone and stepwise dosing regimens are advised.
Relapse and resistance. Increased tumour mutational burden negatively correlates to blinatumomab response due to mutation-associated primary resistance (136). Lineage switch to acute myeloid leukaemia (AML) by rearrangement of the myeloid/lymphoid or mixed lineage leukaemia (MLL) lysine (K)-specific methyltransferase 2A (KMT2A) gene expressed on B-cell ALL can cause relapse or resistance (137). TAA target downregulation is a significant cause of blinatumomab resistance (138). Decreased CD19 expression on leukemic blasts prior to, and after blinatumomab therapy confers primary and secondary resistance, respectively (139). CD19 gene mutations and alternate splicing of CD19 mRNA produce truncated receptor variants conferring resistance (140).
New antibody formats. Dual-affinity re-targeting (DART) offers competing diabody format with additional stability through a Cterminal disulphide bridge (144). In vitro CD19xCD3 DARTs outperform BiTEs in cytotoxicity assays (145). DARTs demonstrate higher CD3-association, lower CD19 dissociation, and more efficient T-and B-cell cross-linking. CD19xCD3 DART, duvortuxizumab, demonstrated response in phase I dose escalation. However, high neurotoxicity rates terminated licensing due to high competition against B-cell malignancy therapies. CD32BxCD16 and CD32BxCD79B DARTs provide an alternative T-cell activation mechanism, highlighting bispecific antibody adaptability.

Threats to BiTEs.
Interest in BiTEs has decreased given unsuccessful attempts to translate agents despite numerous trials. Other designs, such as DARTs, have gained interest. Adoptive cell therapies threaten the sustainability of BiTEs. CD19-CAR-T-cell therapy approval for ALL is altering blinatumomab prescribing. Novel adoptive cell therapies are superior and offer durable remission (109).

Discussion
ICIs and BiTEs are exceptional treatments for haematological malignancies. Yet, biotechnological advancements underlying immunotherapeutic development are costly. Thus, close consideration of the strengths, weaknesses, opportunities and threats of each immunotherapeutic modality is essential to direct future research (Table II). Both ICIs and BiTEs are costeffective "off-the-shelf" drugs. ICIs generate durable response in heavily pre-treated and disease refractory patients. However, limited response across patients and diseases, primary and acquired resistance, and rare but severe toxicities, have set ICIs behind novel ATC therapies such as CARs. Nevertheless, lower costs and longer history of approval compared to BiTEs and novel ATC therapies lends to their continued interest. BiTEs represent powerful immunotherapies with superior anti-tumour efficacy to other antibody contracts. However, costly initial synthesis and competitive licensing have restricted their use, though newer constructs may change this. ANTICANCER RESEARCH 41: 1123-1141 (2021) 1134 Table II. Summary of strengths, weaknesses, opportunities and threats associated with immune checkpoint inhibitors and bispecific T-cell engagers.

Conclusion
With new molecular targets being discovered, more progress is to be expected in T-cell-based cancer immunotherapy. The diverse repertoire of molecular targets offers exceptional potential for combination treatments. Clinically, combination immunotherapy is still at its relative infancy with further research necessary to determine how to optimise and translate treatment regimens into routine clinical practice. The potential to combine immunotherapies with chemotherapy, radiotherapy, and targeted molecular therapies is significant and requires systematic investigation.

Conflicts of Interest
The Authors declare that they have no competing interests.

Authors' Contributions
K.S.R. has contributed to reviewing the literature, drafting and revising the article, figure illustrations, and final approval of the version to be published. C.R.T.H. has contributed to revising the article and final approval of the version to be published. M.S. has contributed to revising the article and final approval of the version to be published. J.K.F. has contributed to the conceptualization of the work, revising the article, supervising the work, and final approval of the version to be published.