Building safety into CAR-T therapy

ABSTRACT

Chimeric antigen receptor T cell (CAR-T) therapy is an innovative immunotherapeutic approach that utilizes genetically modified T-cells to eliminate cancer cells using the specificity of a monoclonal antibody (mAb) coupled to the potent cytotoxicity of the T-lymphocyte. CAR-T therapy has yielded significant improvements in relapsed/refractory B-cell malignancies. Given these successes, CAR-T has quickly spread to other hematologic malignancies and is being increasingly explored in solid tumors. From early clinical applications to present day, CAR-T cell therapy has been accompanied by significant toxicities, namely cytokine release syndrome (CRS), immune effector cell-associated neurotoxicity syndrome (ICANS), and on-target off-tumor (OTOT) effects. While medical management has improved for CRS and ICANS, the ongoing threat of refractory symptoms and unanticipated idiosyncratic toxicities highlights the need for more powerful safety measures. This is particularly poignant as CAR T-cell therapy continues to expand into the solid tumor space, where the risk of unpredictable toxicities remains high. We will review CAR-T as an immunotherapeutic approach including emergence of unique toxicities throughout development. We will discuss known and novel strategies to mitigate these toxicities; additional safety challenges in the treatment of solid tumors, and how the inducible Caspase 9 “safety switch” provides an ideal platform for continued exploration.

KEYWORDS:

• CAR-T
• cellular therapy
• inducible caspase 9
• safety switch
• CRS
• ICANS

CAR-T overview

The chimeric antigen receptor (CAR) T-Cell treatment is an immunotherapy strategy that involves genetically engineering autologous or allogeneic T-cells to express a recombinant protein known as a CAR on the cell surface. The CAR is composed of three major components including an (1) ectodomain consisting of an extracellular antigen-binding domain linked via a (2) transmembrane region to an (3) intracellular signaling domain known as the endodomain. The net effect of the CAR-T cell is antigen-specific recognition and cytotoxic killing of any cell expressing the extracellular target antigen in a non-HLA restricted manner.

Transition to clinical use

Despite early interest in a number of solid tumor types including ovarian,¹ kidney,² and prostate³ cancers, it became clear that solid tumors had additional and substantial barriers to CAR-T cell therapy success including (1) effects of the tumor microenvironment (TME) on CAR-T cell-trafficking and persistence in vivo, (2) tumor polyclonality, (3) on-target, off-tumor (OTOT) adverse events. The focus would quickly shift to targeting the CD19 antigen in B-cell malignancies.

To each of these points: hematologic malignancies are simply more accessible to T-cells from a pathophysiologic standpoint than solid tumor cells. Furthermore, CD19 is a B lineage-associated antigen that is stable and expressed at a high copy number by most B-cell malignancies including leukemia and lymphoma. Lastly, OTOT is primarily prolonged B-cell aplasia and secondary hypogammaglobulinemia, which are easily mitigated. In 2011, clinical trials of second-generation CAR-T cells targeting CD19 in chronic lymphocytic leukemia (CLL) and B-cell acute lymphoblastic leukemia (B-ALL) produced unprecedented results including complete remission in heavily pretreated patients demonstrating a key proof of concept.⁴

In the present day, CAR-T cell therapies now have robust and maturing clinical data for their ability to mediate potent antitumor responses in relapsed and/or refractory (r/r) B-Cell leukemias and lymphomas and multiple myeloma leading to FDA approval of multiple CAR-T cell products.

Review of common CAR-T toxicities

Along with impressive response rates, it was quickly discovered that CAR-T therapy carries the risk of potentially severe systemic toxicities defined by profound and generalized immune system activation. A constellation of fever, hypotension, and respiratory distress would come to define the cytokine release syndrome (CRS) and be noted across all early CAR-T trials with rates as high as 56–100% of patients treated with CD19 CAR-T cells for relapsed/refractory B-ALL.⁵ A lesser population of patients would develop nonspecific, variable neurologic toxicities which is now known as immune effector cell neurotoxicity syndrome (ICANS). Diagnosis, grading, and treatment paradigms for both CRS and ICANS have evolved significantly over the years and is now much more standardized and algorithmic based on the grading schema for each with the ASTCT consensus grading scale adapted by most institutions.⁶

We will focus mainly on these acute, inflammatory toxicities of CAR-T therapy in the context of adult patients treated with CD19-targeted commercial products, though we will also look to failed CAR-T trials of the past to illustrate the breadth of potential OTOT toxicities as the clinical application of CAR-T and other adoptive T-cell therapies (i.e. TCR, TILs) expands in the future. Long-term sequalae of commercial CAR-T products are less well described but include increased infectious risk, hypogammaglobulinemia, and persistent cytopenias⁷ which will not be reviewed here.

Cytokine Release Syndrome (CRS)

Following CAR-T cell binding to their target antigen, downstream cell signaling initiates T-cell expansion and massive production of cytotoxic molecules and cytokines including but not limited to TNFα, IFNγ, IL-1β, IL-2, IL-5, IL-6, IL-8, and IL-10, GM-CSF. Furthermore, the rapid death of the target cell elaborates additional cytokines and inflammatory mediators, adding to the cascade. It is important to note that many of these inflammatory mediators, particularly IL-6 (felt to be a major driver of the syndrome), are not produced by either CAR or unmodified T cells, but rather by incited myeloid cells (monocytes, macrophages, etc.) downstream from the activating event.⁸ CRS is the systemic response to the sudden and excessive release of these inflammatory mediators, all of which have a litany of nonspecific and organ-specific effects in part driven by endothelial injury/dysfunction. Post-infusion onset of symptoms is variable, typically peaking at 2–7 days, though delays up to 3 weeks are reported.⁹

Clinically, fever is the hallmark of CRS, though rigors, tachycardia, hypotension, tachypnea, and hypoxemia are also commonly observed. However, all organ systems may be affected by CRS and patients develop rashes, acute renal injury, hepatobiliary injury, arrhythmia, cardiomyopathy,¹⁰ and coagulopathy.¹¹ CRS is mild and reversible in most cases but can progress to refractory distributive shock with multiorgan failure. Additionally, an emergent condition known as immune effector cell-associated HLH-like syndrome (IEC-HS) can develop continuously with or independently of CRS, though the exact pathophysiology is a subject of further study.¹²

Aside from physiologic toxicity, CRS is associated with significant financial toxicity. A recent study assessing the cost associated with CAR-T cell therapy estimated that CRS may cost a patient $36 000 to $50 000 in severe cases, adding to an already expensive treatment with an average cost of $373 000–$475 000 depending on the product.¹³

Neurologic toxicities

ICANS

Neurologic toxicities were previously considered in aggregate with CRS. With increasing CAR-T therapy experience, immune effector cell-associated neurotoxicity syndrome (ICANS), has been developed as a distinct clinical entity that defines neurologic symptoms that can occur after CAR-T infusion. This may co-occur with CRS or more commonly following the resolution of CRS. Usually this occurs within 3–4 weeks of cell infusion, though cases of delayed onset >4 weeks are described. ICANS manifests as difficulty concentrating, confusion, delirium, lethargy, aphasia, tremor, and seizures. In rare cases, malignant cerebral edema can develop and may lead to death. The mechanisms causing ICANS are not completely understood, but important factors probably include release of neurotoxic substances by CAR-T cells, bystander immune cells, and lysed tumor cells; alteration of the blood brain barrier with endothelial activation as a contributor; and potentially the presence of CAR-T cells in the CNS related to CD19 expression on brain mural cells.^14–17

MNTs

A neurologic toxicity distinct from ICANS, termed movement and neurocognitive treatment-emergent adverse events (MNTs) were recently described in patients treated with BCMA CAR-T cell therapy for multiple myeloma. MNT resembles a parkinsonian syndrome though with neither evidence of substantia nigra degeneration on neuropathology nor expression of BCMA in brain parenchyma to invoke an on target, off tumor toxicity. There is no established treatment for this entity as of this writing.¹⁸

Clinical management and prevention of CRS/ICANS

Initial concerns about decreasing CAR-T expansion and efficacy with corticosteroids and/or tocilizumab (IL-6 R monoclonal antibody) use led to a general sentiment of withholding either treatment for CRS/ICANS as long as possible (i.e. until higher CRS grade reached). However, treatment with tocilizumab does not affect CAR-T expansion based on experiences with tisagenlecleucel.¹⁹ Corticosteroids are also not currently shown to affect response rates, or durability of response in clinical trials, though the long-term effects are not determined.²⁰ Taken together, clinical practice has globally shifted to earlier and more aggressive use of tocilizumab and corticosteroids to manage CRS before preventable organ toxicity occurs. Currently, tocilizumab and steroids comprise the standard of care, but there does remain a subset of patients who develop severe CRS that is refractory to these measures. Early identification of this high-risk subset as well as preemptive interventions in this population are an area of ongoing interest.

There are numerous pharmacologic interventions being investigated for severe and refractory CRS/ICANS employing three major strategies: (1) anti-cytokine-directed treatment, (2) T cell ablation, (3) tyrosine kinase inhibitors (TKIs). Anakinra (IL-1 R antagonist), siltuximab (IL-6 antagonist), empalumab (INF-γ-blocking antibody), and lenzilumab (anti GM-CSF antibody)⁸ target various CRS-associated cytokines with the benefit of not directly affecting the CAR-T cell product. T-cell ablation with antithymocyte globulin (ATG), alemtuzumab (anti-CD52), and cyclophosphamide are not well published in CRS/ICANS though are known to have potent T cytotoxic effects. This approach has the obvious disadvantage of destroying the CAR-T product and causing further native immunosuppression. TKIs including dasatinib (BCR-ABL inhibitor), ibrutinib (BTK inhibitor), and ruxolitinib (JAK inhibitor) have shown the ability to decrease secretion of culprit cytokines resulting in reduced CRS severity in preclinical data.^21–24

Among other risk factors, development of CRS/ICANS has been associated with pre-treatment disease burden as well as CAR-T cell dose and peak expansion.⁵ As such, preventative approaches have been explored with these risk factors in mind. Several groups have evaluated the utility of tumor burden-based CAR-T cell dose reduction or split-dose strategies. While low-dose infusions trended toward higher relapse rates,²⁵ a recent literature review suggests that a dose fractionation strategy may mitigate the incidence of CRS and ICANS, though additional studies are needed.²⁶ Other investigators have assessed prophylactic dexamethasone, tocilizumab, and anakinra for all comers regardless of risk factors. In summary, prophylactic dexamethasone and anakinra were useful in decreasing CRS/ICANS with no decrease in efficacy on early follow-up.^27,28 Prophylactic tocilizumab was observed to transiently increase IL-6 levels due to its mechanism of action in receptor blockade. This was felt to correlate with increased rates of ICANS versus the control (41% vs 28%) with a single case of cerebral edema in the prophylaxis group.²⁹ Though the findings are somewhat controversial, most groups avoid prophylactic tocilizumab as a result.

On Target, Off Tumor (OTOT)

While phase 1 trials of first-generation CARs failed to show meaningful clinical tumor response in ovarian cancer,¹ metastatic renal cell carcinoma,^2,30 melanoma, and other tumor types, the field would gain first-hand experience with on-target, off-tumor toxicity (OTOT). There are several documented examples of this across organ systems including B-cell and plasma cell aplasia with CD19 and BCMA targeted CAR-T therapies;⁹ fatal pulmonary toxicity with CAR-T product targeting ERBB2 (HER2/Neu); hepatic toxicity with CAR-T targeting the antigen carboxyanhydrase-IX (CAIX);^2,30 and fulminant myocarditis with a TCR targeting melanoma-associated antigen 3 (MAGEA3).³¹ Apart from B-cell aplasia related to CD19 targeting, the culprit antigens did not signal for toxic potential in pre-clinical testing.

In summary, the most common acute toxicities for CAR-T products in clinical use (CD19, BCMA) are CRS and ICANS. The management of CRS has improved markedly over the years since CAR-T therapies were first introduced and can now be managed relatively easily with early and aggressive use of tocilizumab and corticosteroids. ICANS has become slightly better defined, but the pathophysiology remains clouded. While early and aggressive use of steroids and antiepileptics have improved morbidity and mortality, this remains a substantial safety issue to be addressed. New neurologic toxicities (NMTs) are emerging in the BCMA-targeted CAR-T products that will need to be addressed and raises the specter of more unanticipated neurologic, organ-specific, or long-term toxicities even when the target antigen is not clearly present in the affected tissue. We expect the absolute magnitude of these problems to increase as CD19 and BCMA CAR-T products are approved for new specific indications (i.e. primary CNS lymphoma, CLL) and in earlier lines of therapy for B-cell malignancies and multiple myeloma, respectively.

We also anticipate the morbidity and mortality associated with OTOTs to increase as CAR-T therapies target a wider array of antigens on solid tumors or on hematologic tumors that are known to have heterogeneous or low levels of antigen expression, particularly AML. While no doubt there will be improvements in preclinical safety evaluation of TAAs, it is unlikely that CAR-T developers will be able to anticipate all OTOTs and especially unlikely to anticipate idiosyncratic cross-reactivity events, using the story of MAGEA3 as an example. As such, there remains a great need for rapid and complete eradication of CAR-T products as they are put to clinical use in humans to mitigate known, severe toxicities and protect our patients from unknown and late phase toxicities lying in wait.

Addressing CAR-T toxicity in platform development

There have been a number of innovative approaches to fine-tune the CAR-T cell for safety with several recent reviews provided for reference^32–34 and a table (Table 1) summarizing some of the approaches. The approaches are impressively broad and include either modifying or re-imagining the CAR construct at each of the three basic components (the ectodomain, transmembrane region, and endodomain) or simply by adding on new ancillary features.

Table 1. Genetic engineering approaches to improving CAR-T safety.

Approach	Comment	Reference
Ectodomain modification
scFv affinity	Low affinity scFv mediates tumor killing without host tissue damage	Ghorashian et al.^Citation35
humanized scFv	Humanized scFV targeting CD19 in human pediatric ALL slightly decreased CRS/ICANS with similar efficacy	Myers et al.^Citation36
IL-6 sink	Non-functioning IL-6 receptor serves as a sink for locally elaborated IL-6	Tan et al.^Citation37
Hinge/TM modification
Hu19-CD828z	Decreased rates of CRS/ICANS attributable to hinge/TM alteration rather than humanized scFv alone	Brudno et al.^Citation38
Logic Gating
AND-gate
Split Recognition	Expresses at least two CARs one with CD3z one with costimulatory domain (i.e. CD24, 4-1BB) and can only transduce signal when both are bound to antigen	Lanitis et al.^Citation39
Reverse CAR	Distinct peptide epitopes in place of scFv. An exogenous bi-specific engager is given which has two scFvs: one specific to distinct peptide epitope and the other recognizing a tumor-associated antigen	Kittel-Boselli et al.^Citation40
Avid CAR	Avidity-controlled CAR platform, designed to homodimerize upon dual antigen recognition conferring sufficient avidity for T cell activation	Salzer et al.^Citation41
NOT-gate
iCAR	Contain inhibitory domains derived from PD-1 and CTLA-4. CAR- and iCAR-expressing T cells will kill cells expressing only the CAR antigen while sparing cells harboring both the CAR and iCAR antigens	Fedorov et al.^Citation42
LOH-iCAR (NASCAR)	Pairwise chimeric receptors that, together, are capable of allele sensing and detecting LOH events in cancer	Hwang et al.^Citation43
AND-, AND/NOT- gate
Co-LOCKR	Utilizes a universal CAR specific to a recruitment peptide, which is placed in a ‘Latch’ sequence. The latch is linked to a soluble ‘Cage’ protein that is directed at the first tumor antigen. The recruitment peptide is sequestered by the Cage and is only exposed to engagement by the CAR when a competing ‘Key’ protein binds a second antigen present on the same tumor cell	Lajoie et al.^Citation44
SynNotch	Utilizes a constitutively expressed module which releases a synthetic TF only upon binding of antigen A. Upon translocation to the T cell nucleus, this TF drives the expression of a conventional CAR specific to antigen	Choe et al.^Citation45
SUPRA CAR	Universal, leucine zipper CAR-like receptor (zipCAR), which can bind a corresponding leucine zipper incorporated onto a soluble antigen-specific zipFv protein, enabling multiplexed control of T cell responses	Cho et al.^Citation46
VIPER	As below
Pharmacologic Switch
Dasatinib (OFF)	Administration of dasatinib keeps CAR T cells in an inactive function “off” state	Mestermann et al.^Citation21
Lenalidomide (ON)	lenalidomide-inducible dimerization system with split CARs that required both lenalidomide and target antigen for activation	Jan et al.^Citation47
Lenalidomide (OFF)	lenalidomide-inducible CRL4^CRBN-mediated ubiquitination and proteasomal degradation of a “super” degron-tagged CAR	Jan et al.^Citation47
Tetracycline (ON)	Gene of interest (CAR in this case) is under control of a tetracycline/doxycycline dependent transcriptional regulation	Zhang et al.^Citation48
Protease-based (“SNIP CARs”)
VIPER CAR (ON)	Versatile Protease Regulatable CARs. A protease, NS3 is constitutively expressed in the CAR construct. Only when its inhibitor (grazoprevir) is given can the CAR be assembled and expressed	Li et al.^Citation49
VIPER CAR (OFF)	Split CAR with endodomain held together by NS3-binding peptide and an enzymatically inactive NS3 in series with the signaling domain. CAR is capable of cell signaling until protease inhibitor given which, competitively binds and breaks the linkage	Li et al.^Citation49
CRASH-IT (OFF)	Chemically Regulated SH2-delivered Inhibitory Tail. NS3-degron-PD1 inhibitory tail fusion protein is coexpressed with CAR. The fusion protein has constitutive inhibitory function of the CAR until asunaprevir/grazoprevir is administered.	Sahillioglu et al.^Citation50
Suicide Genes
mAB-mediated
EGFRt	Engineered cells constitutively express truncated EGFR which can be eliminated by administering anti-EGFR mAb, cetuximab.	Paszkiewicz et al.^Citation51
RQ8R	Engineered cells express polypeptide sequence derived from CD34 and CD20 (termed RQR8) which can be bound and eliminated by rituximab	Philip et al.^Citation52
Gene-directed enzyme prodrug therapy (GDEPT)
HSV TK	Renders engineered cells sensitive to ganciclovir which disrupts cell cycling and leads to apoptosis	Bonini et al.^Citation53
Dimerization induced
Inducible Caspase 9	Administration and binding of AP1903 (rimiducid) to FKBP12-F36V leads to dimerization and activation of an inducible caspase 9-based suicide construct (iCaspase 9), which triggers apoptosis of the CAR-T cell.	Di Stasi et al.^Citation54
Environmental Switching
LINTAD	Blue light induced CAR expression	Huang et al.^Citation55
FUS Induced	CARs induced via heat shock protein pathway using focal ultrasound (FUS)	Wu et al.^Citation56
Hypoxia Inducible CARs	Hypoxia-induced CAR expression via HIF pathway	Kosti et al.^Citation57
Inflammation and hypoxia inducible CAR-T (CARTIV)	TME-related hypoxia and inflammatory markers drive CAR expression	Greenshpan et al.^Citation58

A notable example is the integration of logic-gated CAR constructs that utilize AND-, OR-, NOT-gate strategies; of which AND- gating results in improved target specificity to reduce OTOT. The gate is completed either through intrinsic CAR properties or reliance on an exogenous signal (dimerizing agent, small molecule, protease inhibitor, etc.). The synthetic Notch (synNotch) strategy is another combinatorial antigen recognition approach in which a synNotch receptor must be activated by a priming antigen before the authentic CAR is even expressed on the T-cell surface.⁴⁵ Some groups have uncoupled the antigen binding unit from the CAR-T cell altogether creating “universal” CARs that engage the tumor using an exogenous mAb, peptide bridge, or BiTE that can easily be withheld if toxicities arise. Others have created alternative methods of CAR-T cell activity induction/switching using blue light, ultrasound, hypoxia, small molecules, and known pharmacologic (i.e. dasatinib, lenalidomide, doxycycline) agents as switches. Despite these innovative and effective approaches, there is not an “off switch” for CAR-T cells that have already engaged their target. In certain scenarios of life-threatening OTOT, there may be a need for rapid and complete abortion of T-cell activity that can may only be achieved through apoptosis.

Recognizing this need for a rapid and complete “off switch” several groups, including our own, have developed various suicide genes into the CAR construct. For example, Paszkiewicz et al.⁵¹ developed a truncated version of epidermal growth factor receptor (EGFRt) that is co-expressed on CD19 CAR-T cells and leads to effective elimination of CD19 CAR-T cells when an EGFR targeted antibody is administered in a murine model. This fits under the broad umbrella of suicide gene technologies which can be further subdivided by mechanism of action: (1) monoclonal antibody-mediated (as above), (2) gene-directed enzyme prodrug therapy (GDEPT) whereby an inert prodrug is metabolized to a toxin only in the gene-modified cells, and (3) dimerization induced.⁵⁹ Our group has continued to utilize a dimerization induced safety switch for our CAR-T constructs, which we believe addresses a key need in mitigating safety concerns as adoptive cellular therapy applications rapidly expand.

The inducible Caspase9 (iCasp9) safety switch

Originally developed for the mitigation of graft vs. host disease (GVHD) in allogeneic stem cell transplant,⁵⁴ the inducible Caspase 9 system is perfectly suited to manage the significant toxicities of CAR-T cells in their current applications – and likely even more valuable for solid tumor applications where OTOT toxicities may develop as new targets are explored.

Apoptotic genes including caspases eliminate cells by inducing cell death. Chimeric proteins composed of apoptotic pathway components linked to a dimerize-able drug-binding component have been developed.^60,61 Administration of a non-therapeutic small molecule causes dimerization and activates the apoptotic pathway. In the inducible Caspase 9 (iC9)/AP1903 system, the chimeric protein is composed of an FKBP12-F36V domain linked to ΔCaspase 9, which is Caspase without its activator recruitment domain. The FKBP12 carries a substitution (F36V) that allows for the binding of synthetic dimeric ligands, in this case, rimiducid (RIM), which is an otherwise biologically inert and clinically safe compound.⁶²

This platform is on display in an ongoing phase 1/2 trial (NCT03016377) for r/r B-ALL testing the safety and efficacy of autologous T lymphocytes genetically modified to express iC9, truncated low-affinity nerve growth factor receptor (ΔNGFR) (for selection and tracking purposes), and CD19 CAR (encoding 4-1BB). We reported the first case of administering a potentially ablative dose of RIM (0.4 mg/kg) showing that the intervention was able to rapidly eliminate CAR-T cells (>60% within 4 h, and > 90% within 24 h of RIM infusion) along with equally brisk resolution of severe, prolonged corticosteroid refractory ICANS.⁶³

Given this patient’s rapid resolution of symptoms and reduction in iC9 CD19 CAR-T numbers, we sought to explore the effects of lower, potentially non-ablative doses of RIM in patients with CS-nonresponsive ICANS. We recently presented⁶⁴ the clinical and pharmacodynamic correlates in 4 additional patients treated with rimiducid, two of which received full dose RIM 0.4 mg/kg and achieved complete resolution in ICANS by 5 days with rapid reduction in CAR-T cells by 4 h (83.6% and 98.6% reduction). The other two patients received the dose-reduced RIM 0.1 mg/kg with resolution of ICANS within 48 h. These patients also had a significant reduction in CAR-T cells by 4 h (88.6% and 88.1%). Rates of overall response at 4 weeks were 3/4 in RIM-treated subjects vs 9/9 in other reported subjects. No clear association between RIM pharmacokinetics and degree of CAR-T cell ablation has emerged, and it is possible that even lower doses might be necessary to discern this. Decreased response rate in RIM-treated patients raises concern for blunting of antileukemic activity of iC9 CD19 CAR-T cells. Both questions are subjects of active research.

Another notable trial utilizing the iC9 switch for a solid tumor application was recently published.⁶⁵ Del Bufalo et al. reported their findings of a phase 1/2 trial utilizing the iC9 safety switch in a GD2 CAR (encoding 4-1BB) for pediatric neuroblastoma. A total of 27 heavily pretreated children were enrolled and received GD2-CART01. CRS occurred in 20/27 patients (74%) with only one Grade 3 CRS resolving rapidly with tocilizumab. A single patient developed an “altered level of consciousness” and received two doses of RIM (0.4 mg/kg). The workup would reveal brain hemorrhage as the etiology of the neurologic symptoms, but the effects of RIM on the CAR-T product were reported. As previously described in our work, there was a sharp decrease in circulating GD2-CART01 levels at 4 hours after infusion of RIM. Interestingly, a re-expanded population of the GD2-CART01 cells were detected at 6 weeks. This cell population was interrogated and determined to have retained full sensitivity to RIM in vitro suggesting that activation of the suicide gene preferentially eliminates activated gene-modified cells.

This raises the question of whether this re-expanded cell population would be capable of mediating on-target effector functions including both anti-tumor activity and OTOT. Reassuringly, any reemerging activity – favorable or not – would still be under the control of RIM, but further assessment is needed. Taken together, iC9 platform continues to build a portfolio of safe and reliable immune effector cell elimination in human trials spanning both hematologic and solid malignancies, without compromising disease activity.

Conclusion

As existing CAR-T cell trials mature and new trials emerge for both hematologic and solid tumor malignancies, so will new on-target, off-target, and idiosyncratic toxicities. Despite improved clinical management of CRS and ICANS, the perennial risk of unpredictable, untestable, and life-threatening toxicities is one of the biggest barriers to expansion of the field and a significant patient safety risk. It is our hope that the iC9 safety switch along with other promising genetic engineering approaches to CAR-T cell safety will ultimately allay these concerns and allow for the safe and rapid discovery of impactful treatments across all cancer types.

Disclosure statement

N.G serves on the advisory board for Bristol Meyer Squib, Novartis, and Kite Pharmaceuticals

N. G. and B.S. have received previous research funding form Bellicum Pharmaceuticals.

D. P has no interests to declare.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.