Ready or not, here they come: a practical guide to adoptive T-cell therapies for solid tumors for medical oncologists

Abstract

Adoptive cell therapies (ACTs) include chimeric antigen receptors (CARs), bispecific T-cell engagers (BiTES), tumor-infiltrating lymphocytes, and T-cell receptor (TCR) gene-modified T-cells. ACTs have significantly improved patient outcomes associated with hematologic malignancies. Recent advances have demonstrated their potential in solid tumors with promising clinical trial results signaling that they represent paradigm shift in oncologic care. As these therapies become standard care in solid tumors, medical oncologists must become adept at recognizing and managing their unique toxicities including cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS). This review synthesizes the latest clinical data on ACTs in solid tumors, highlighting key findings, and toxicity profiles. In addition, the review provides an overview of early recognition and evidence-based management of CRS and ICANS. Equipping clinicians with the necessary knowledge to navigate toxicity management will be essential in optimizing patient outcomes as ACTs are increasingly adopted in solid tumor oncology.

Implications for Practice.

With adoptive cell therapies (ACTs) emerging as a transformative approach for solid tumors, a comprehensive understanding of their distinct toxicity profiles is imperative for medical oncologists. This review explores the latest clinical trial data and evidence-based strategies for mitigating predominant and potentially life-threatening toxicities associated with ACTs including cytokine release syndrome and immune effector cell–associated neurotoxicity syndrome. By providing a structured framework for early recognition and management, this manuscript aims to enhance clinician preparedness and facilitate the safe integration of ACTs into routine oncologic practice and ultimately improving patient outcomes.

Adoptive T-cell therapies encompass a range of interventions which enhance the ability of T lymphocytes to generate an anti-tumor immune response. These strategies include chimeric antigen receptors (CARs), bispecific T-cell engagers (BiTEs), tumor-infiltrating lymphocytes (TILs), and T-cell receptor (TCR) gene-modified T-cells (Figure 1). Several of these approaches have significantly improved patient outcomes associated with hematologic malignancies and have begun to make forays into solid tumor treatment. Importantly, the toxicity associated with adoptive cell therapies (ACTs) is unique and complex particularly when compared to the more commonly encountered toxicities of immune checkpoint inhibitors in solid tumors. Thus, it is essential for medical oncologists to understand the nature of these therapies, their increasing number of indications, and the management of potential toxicities. Here, we review CAR-T therapies, BiTEs, TILs, and TCRs with a practical focus for medical oncologists.

Figure 1.

Adoptive T-cell therapies for solid tumors. (1) Chimeric antigen receptor T-cell (CAR-T-cell) therapy—CARs are genetically engineered receptors with an extracellular antigen-recognition domain and intracellular signaling domains, connected by hinge and transmembrane regions. Activated CAR-T-cells recognize and bind to tumor-associated antigen (TAA) and exhibit cytotoxic activity via release of perforin and granzyme granules and activation of death receptor pathways (eg, Fas/Fas-L), secretion of interferon-gamma (IFN-γ) and tumor necrosis factor-alpha (TNF-α) promoting natural killer (NK) cells, ultimately lead to cancer cell apoptosis and necrosis. (2) Bispecific T-cell engager (BiTE therapy)—BiTE is a bispecific antibody, comprised of 2 variable fragments from an anti-TAA and an anti-CD3 mAb, with a short linker connecting them in tandem. BiTE molecules crosslink cytotoxic T-cells and cancer cells, independently from major histocompatibility complex (MHC) and costimulatory signals and activate T-cells, subsequently secreting granzyme, perforins and IFN-γ and TNF-α promoting Natural Killer cells, ultimately lead to cancer cell apoptosis and necrosis. (3) Tumor-infiltrating lymphocyte (TIL) therapy—TILs are extracted from a patient’s tumor and expanded ex vivo using IL-2. They are then re-introduced back to lymphodepleted patient, where they proliferate, function as effector cells, and generate immunological memory. Since these T-cells originate from the tumor, a significant proportion is expected to recognize TAA or neoantigens. (4) Engineered T-cell receptor (TCR) cell therapy—Engineered TCR T-cell therapy involves engineering T-cells with receptors that are specific for MHC-presented antigenic peptides. The graphic was created with Biorender.com.

CAR-T therapy

CARs are engineered receptors that enable T-cells to recognize and attack tumor antigens.^1,2 Structurally, CARs consist of 4 key domains: an antigen-binding domain, a hinge region, a transmembrane domain, and an intracellular T-cell signaling domain. CAR-T therapy involves collecting T-cells from the patient’s blood via leukapheresis, genetically modifying them to express CARs, expanding them ex vivo, and re-infusing them into the patient following lymphodepleting chemotherapy. The modified T-cells are enhanced in their ability to target cancer cells by recognizing tumor antigens through the antigen-binding domain, independently of MHC molecules.³ There have been multiple iterations of CAR constructs, with incorporation of co-stimulatory domains (eg, CD28 or 4-1BB), and expression of additional transgenes, such as cytokines (eg, IL-12) or other immunomodulatory molecules, to address challenges of tumor microenvironment, enhance T-cell persistence, proliferation, and increase anti-tumor activity.^4–16

Although CAR-T-cell therapies have transformed the treatment landscape for hematologic malignancies, their development and clinical success in solid tumors have been more limited.¹⁷ Clinical responses of CAR-T-cell therapies in solid tumors remain modest due to multiple challenges, including the scarcity of specific target antigens, the consequent risk of on-target/off-tumor toxicity, inadequate trafficking to tumor sites, limited persistence of CAR-T-cells, loss of effector function, and marked tumor antigen heterogeneity. A particularly formidable obstacle is effectively infiltrating tumor immune microenvironment (TME) in which many tolerogenic immune and stromal cell populations may resist T-cell activation.¹⁸

Cancer-associated fibroblasts (CAFs) constitute a dominant stromal population in many solid tumors and represent a key stromal population that may resist a therapeutic anti-tumor immune response (Figure 2).¹⁹ CAFs promote tumor progression through dynamic crosstalk with malignant and immune cells and by depositing a dense extracellular matrix (ECM) enriched in collagen and hyaluronan.^20,21 This fibrotic network stiffens tissue, increases interstitial pressure, and collapses vasculature, thereby restricting the migration and deep penetration of CAR-T-cells and other therapeutics. In parallel, CAFs secret immunomodulatory mediators such as TGF-β, interleukin (IL)-6, CXCL12, and VEGF that recruit regulatory T-cells and myeloid suppressor cells, exclude effector lymphocytes, and impose metabolic stressors (eg, hypoxia, acidity, and nutrient deprivation) that accelerate T-cell dysfunction.²² These combined structural and biochemical effects reduce the persistence, expansion, and cytolytic function of therapeutic T-cells. Beyond CAFs, other immunosuppressive cell types such as M2 macrophages, Tregs, and myeloid-derived suppressor cells reinforce local immune tolerance by releasing IL-10 and expressing checkpoint ligands such as PD-L1, TIM-3, and LAG-3.¹⁸ Together, these mechanisms constitute “immune-excluded” or “immune-desert” phenotypes characterized by ineffective T-cell trafficking and exhaustion.¹⁸ As such, overcoming CAF-mediated fibrosis and TME-driven immunosuppression remains central to enhancing T-cell infiltration and restoring durable anti-tumor immunity in solid tumors. Recent innovative designs of CAR-T focus on improving T-cell trafficking and infiltration, remodeling stroma (targeting or degrading CAF/ECM components), resisting immunosuppression, broadening antigen recognition and refining manufacturing/dosing to cover the physical and biochemical barriers of solid-tumor microenvironment.¹⁷ Published clinical trials with CAR-T-cell therapy in solid tumors are summarized in Table 1.

Figure 2.

Challenges of chimeric antigen receptor T-cell (CAR-T) therapy in the immunosuppressive tumor microenvironment. Conventional T-cells including CAR-T-cells frequently encounter abnormal vasculature and a dense extracellular matrix (ECM) rich in collagen and hyaluronan deposited by cancer-associated fibroblasts (CAFs). CAFs secrete TGF-β, interleukin (IL)-6, CXCL12, and VEGF, recruiting regulatory T-cells (Tregs) and myeloid-derived suppressor cells (MDSCs) while polarizing macrophages toward an immunosuppressive M2 phenotype. These immune and stromal interactions create fibrotic, hypoxic, and metabolically stressed niches that restrict T-cell trafficking and induce T-cell exhaustion. The graphic was created with Biorender.com.

Table 1.

Published clinical trials utilizing CAR-T-cell therapy for solid tumors and toxicity profile.

Target CAR-T-cells	Target population	Phase	N	Key outcomes	Safety profile
HER2 CAR-T	HER2-positive sarcoma	I23	19	4 patients with SD for 3-14 months, median OS of 10.3 months (95% CI: 5.1-29.1 months)	CRS gr 1/2 = 5%, gr 3/4 = 0% Seizure gr 1/2 = 11%, gr 3/4 = 0% Back pain gr 3/4 = 16%, Liver toxicity gr 1/2 = 32%
HER2 CAR-T	Glioblastoma	I24	17	Median OS of 24.5 months from diagnosis and 11.1 months from first T-cell infusion	No dose-limiting toxicity was observed; however, 2 patients had gr 2 seizures and/or headaches
EGFR CAR-T	NSCLC	I25	9	Median OS of 15.6 months, median PFS of 7.1 months	Fever gr1/2 = 67%, gr3 = 11% Gr 4 = 0%.
EGFR-IL13Ra2 CAR-Ta	Glioblastoma	I26	6	Tumor size reduction at early MRI timepoints (criteria for ORR not met)	CRS gr 1/2 = 100% ICANS gr 1/2 = 83%, gr 3/4 = 17%
EGFRvIII CAR-T	Glioblastoma	I27	10	Median OS ∼8.3 months in 10 infused patients	No reported CRS or ICANS Neurologic events in 3 patients in the context of brain tumors
CARv3-TEAM-E Tb	Glioblastoma	I28	3	Transient radiographic tumor regression observed in 2 out of 3 patients	CRS gr 1/2 = 100% ICANS gr1/2 = 67%, gr 3/4 = 33%
IL-13Ra2 CAR-T	Glioblastoma	I29	1	Regression of all intracranial and spinal tumors, with clinical response ∼7.5 months after initiation of CAR T-cell therapy	CRS gr1/2 = 100% ICANS gr1/2 = 100%
Mesothelin CAR-T (combined with pembrolizumab)	Malignant pleural mesothelioma	I30	27	Median OS of 23.9 months. Two patients with complete metabolic response on PET imaging	CRS gr ½ = 26%, gr 3/4 = 0% Confusion gr 1/2 = 8%, gr ¾ = 0%
GPC3 CAR-T	HCC	I31	13	Two patients with PR, 3-year OS of 10.5%	CRS gr 1/2 = 62%, 1 patient with gr 5 CRS No neurotoxicity observed
CLDN18.2 CAR-T	Gastric and pancreatic cancers	I32	12	ORR of 33.3%, with median PFS of 130 days	There were no serious adverse events, treatment-related death or severe neurotoxicity occurred in the study. All CRS observed were gr 1 or 2
CLDN18.2 CAR-T	GI cancersc	I33	98	ORR of 38.8%, median PFS of 4.4 months (95% CI: 3.7-6.6 months), median OS of 8.8 months (95% CI: 7.1-10.2 months)	CRS gr 1/2 = 96.9%, gr3/4 = 0% GI toxicity gr 1/2 = 93%, gr 3/4 = 73%
CD133 CAR-T	HCC	I/II34	21	ORR of 71%, median OS of 12 months (95% CI: 9.3-15.3 months), median PFS of 6.8 months (95% CI: 4.3-8.4 months)	CRS and ICANS not reported
CEA CAR-T	Colorectal cancer	I35	10	At 30 weeks, there were 7 patients with SD and 2 patients with tumor reduction	Two patients with gr 2 fever Liver toxicity gr 3/4 = 30%
B7-H3 CAR-T	Refractory solid tumor	I36	9	One patient with PR after second CAR T-cell infusion	CRS gr 1/2 = 11%, gr 3/4 = 0%
GD2 CAR-T	Neuroblastoma	I/II37	27	ORR of 63%, 3-year OS for patients receiving the recommended dose was 60% and 3-year EFS was 36%	CRS gr 1/2 = 70%, gr 3/4 = 4% Liver toxicity gr 1/2 = 48%, gr 3/4 = 26% No reported neurotoxic effect
CAIX CAR-T	RCC	I/II38	12	No clinical response recorded	Liver toxicity gr 1/2 = 66%, gr 3/4 = 33%

Abbreviations: CAR-T, chimeric antigen receptor T-cell; CRS, cytokine release syndrome; ICANS, immune effector cell–associated neurotoxicity syndrome; OS, overall survival.

^aTarget EGFR and IL13Ra2.

^bTarget EGFR variant III tumor-specific antigen, as well as the wt EGFR protein, through secretion of a T-cell-engaging antibody molecule (TEAM).

^cGastric carcinoma/gastroesophageal cancers, pancreatic cancers, biliary tract cancers.

Bispecific T-cell engagers

BiTEs are antibody-based molecules classically consisting of 2 distinct single-chain variable fragment (scFv) regions: 1 scFv domain is engineered to bind a tumor-associated antigen (TAA), and the other scFv domain binds and activates CD3 within the TCR complex.^39–41 When both scFv regions are engaged, BiTEs act as a linkage and force T-cell activation and binding tumors cells expressing a specific TAA thus resulting in T-cell mediated cytotoxicity in an MHC-independent manner. Subsequently, cytokines release by activated T-cells are theorized to result in activation of other immune populations potentially further enhancing anti-tumor immune activity.⁴² BiTEs, lacking the Fc region found in antibodies, can reduce the risk of toxicities related to Fc effector functions and improve tumor permeability. However, this comes at the cost of shorter half-life and reduced stability.⁴³ The challenges with BiTE development for solid tumors are 2-fold. First, the immunosuppressive nature of the TME in many solid tumors as well as the short-half-life of BiTEs may hamper the durability of therapy. Second, the expression of TAAs on normal tissue as well as the rapid activation of T-cells via CD3 agonism has the potential to lead to both on-target off-tumor toxicity as well as systemic immune activation.

Catumaxomab, the first CD3/EpCAM BiTE created for the treatment of malignant ascites of epithelial cancers based on an improved puncture-free survival versus paracentesis alone, was approved by the European Medicines Agency in 2009 but later withdrawn due to significant fatal risks of CRS and liver toxicity.⁴⁴ It was thought that the liver toxicity was due to expression of EpCAM in normal Kupffer cells of the liver. Solitomab, another first-generation CD3/EpCAM BiTE, was studied for the treatment of relapsed/refractory advanced-stage solid cancer. Unfortunately, dose-limiting diarrheal and transaminase toxicities precluded dose escalation to potentially therapeutic levels.⁴⁵ Newer generations of BiTE molecules have been designed to target TAAs with minimal expression in normal tissues. Additionally, these molecules are engineered without an Fc domain to further reduce toxicity. There are multiple ongoing studies using BiTEs targeting various solid tumors by varying TAAs. Published clinical trials on BiTEs for solid tumors are summarized in Table 2.

Table 2.

Published clinical trials utilizing BiTE therapy for solid tumors and toxicity profile.

BiTE target	Target population	Phase	N	Reported outcomes	Safety profile
GP100/CD3 (tebentafusp)	HLA-A*02:01-positive Uveal melanoma	III46  ,  47	252	ORR of 11%, median OS of 21.6 months, 3-year survival rate of 27%	CRS gr 1/2 = 88%, gr 3/4 = 1%; 57% of patients experienced CRS within the first 4 weeks of treatment during the dose escalation; the incidence and severity of such events decreased with repeated dosing Rash gr 1/2 = 64%, gr 3/4 = 18%
DLL3/CD3 (tarlatamab)	ES-SCLC	II48	220	ORR of 40%, median PFS of 4.9 months, 9-month OS of 68% in the 10-mg group ORR of 32%, median PFS of 3.9 months, 9-month OS of 66% in the 100-mg group	CRS gr 1/2 = 50%, gr 3/4 = 1% in the 10-mg group CRS gr 1/2 = 55%, gr 3/4 = 6% in the 100-mg group Most CRS occurred on days 1 and 8 of cycle 1 ICANS 1/2 = 8%, gr 3/4 = 0% in the 10-mg group ICANS 1/2 = 23%, gr 3/4 = 5% in the 100-mg group Most ICANS and associated neurologic events occurred after cycle 1 of Tarlatamab, with a median time to onset of 5 days
DLL3/CD3 (tarlatamab)	ES-SCLC	III49	509	Median OS of 13.6 months (95% CI: 11.1 months—NR). ORR of 35%	CRS gr 1/2 = 55%, gr 3 = 1%, no CRS gr 4/5 ICANS gr 1/2 = 6%, 1 patient with ICANS gr 5
DLL3/CD3 (obrixtamig)	DLL-3 positive SCLC, epNEC, and LCNEC-L	I50	168	ORR of 18%, DCR of 42%. DoR of 8.5 months across all tumor types	CRS gr 1/2 = 54%, gr $\ge$3 = 3%. Two fatal AEs were considered to be related to treatment, including ICANS DLT in regimen B2 and confusional state in regimen B3
PSMA/CD3 (pasotuxizumab)	mCRPC	I51  ,  52	47	3/11 patients in the IV cohort had SD, 5/18 patients in the SC cohort had SD	CRS gr 1/2 = 4%, gr 3/4 = 2% Fever any grade = 85%
PSMA/CD3 (acaptamab)	mCRPC	I53  ,  54	133	Median PSA PFS of 3.3 months (95% CI: 3.0-4.9), radiographic PFS of 3.7 months (95% CI: 2.0-5.4)	CRS gr 1/2 = 77%, gr 3/4 = 20%. Most CRS events were seen in treatment cycle 1; incidence and severity decreased at/beyond cycle 2
CD3/MUC17	Advanced liver cancer	II55	11	Median PFS of 4 months, median OS of 13.2 months	Fever = 100% Five patients discontinued treatments due to toxicity
HER2/CD3  (HER2 BATs)	Metastatic breast cancer	II56	32	Median OS of 13.1 months (95% CI: 8.6-17.4)	Fever gr 1/2 = 59%, gr 3/4 = 0% Hypotension gr 1/2 = 44%, gr 3/4 = 0% Headache gr 1/2 = 69%, gr 3/4 = 0%
EGFRvIII/CD3 (AMG 596)	EGFRvIII-positive GBM or malignant glioma	Ib57	14	ORR of 21%	Adverse events occurred in all patients and were serious in 50%. Non resulted in discontinuation. Headache gr 3/4 = 14% Depressed consciousness gr 3/4 = 14%
GPC3/CD3 (ERY974)	Advanced solid tumor	I58	29	One patient with PR, 4 patients with SD	One patient with gr2 CRS and 1 patient with gr3 CRS, with regimen A (dose level 0.81 μg/kg)
MUC16/CD3 (ubamatamab)	Recurrent ovarian cancer	I59	78	ORR of 14.3% and DCR of 57.1% in 42 patients receiving $\ge$1 full dose of $\ge$20 mg	CRS gr 1/2 = 73%, gr 3/4/5 = 0% Most common gr $\ge$ 3 AEs were anemia (24.4%) and abdominal pain (14.3%)
CEA/CD3 (cibisatamab)	Advanced CEA-positive solid tumor	I60	149a 228b	ORR of 4% in S1 and 7% in S2 cohorts.	CRS gr 1/2 = 18.2% (S1 cohort), 9.2% (S2 cohort) CRS gr 3 = 0.7% (S1 cohort), 1.3% (S2 cohort) One patient in S1 experienced CRS gr 5

Abbreviations: BiTE, bispecific T-cell engager; CRS, cytokine release syndrome; DLL3, delta-like ligand 3; HLA, human leukocyte antigen; ICANS, immune effector cell–associated neurotoxicity syndrome; ORR, overall response rate; OS, overall survival; SCLC, small cell lung cancer.

^aCohort (NCT02324257) patients received cibisatamab with or without Obinutuzumab.

^bCohort (NCT02650713) patients received cibisatamab with atezolizumab.

Currently, there are 2 BiTE therapies approved by the FDA for solid tumors. Tebentafusp is a unique bispecific fusion protein consisting of a cloned TCR recognizing gp100 presented by HLA-A*02:01 linked with a CD3 binding domain currently approved for the treatment of unresectable or metastatic uveal melanoma. Tarlatamab is a bispecific DLL3-directed CD3 T-cell engager approved for the treatment of extensive stage small cell lung cancer (ES-SCLC).^46–49,61 Tarlatamab gained accelerated FDA approval in 2024 for ES-SCLC after platinum-based chemotherapy. In the pivotal DeLLphi-301 trial, the overall response rate (ORR) of tarlatamab for relapsed/refractory ES-SCLC following platinum-chemotherapy was 40% with a median duration of response of 9.7 months.⁴⁸ While CRS was common, most cases were low grade and occurred primarily during treatment cycle 1 (Table 2).⁴⁸ Only 3% of patients discontinued tarlatamab due to treatment-related adverse effects. Data from the recent phase 3 randomized trial DeLLphi-304 further validated the significant benefit of tarlatamab in the second-line treatment of ES-SCLC, demonstrating a reduction in risk of death by 40% and extending median overall survival (OS) to 13.6 versus 8.3 months with standard chemotherapy.⁴⁹ Similar toxicity profiles were observed in this phase 3 study.⁴⁹ CRS occurred in 56% of patients treated with Tarlatamab, with the majority of events being grade 1/2 (55%) and only 1% experiencing grade 3 CRS (Table 2). Glucocorticoids were used to manage CRS in 16% of patients, while tocilizumab was administered in 4%.⁴⁹ Immune effector cell–associated neurotoxicity syndrome (ICANS) was reported in 6% of patients, with all cases being grade 1/2 except for 1 fatal (grade 5) event associated with progressive neurologic decline.⁴⁹ Additional studies are ongoing evaluating combining tarlatamab with immune checkpoint inhibitors (DeLLphi-305 trial) thus moving it earlier in the treatment lines of ES-SCLC.⁶²

Engineered TCR T-cell therapy

A cloned TCR may be utilized as an alternative to a CAR construction thus allowing T-cells to target TAAs presented by MHC thus allowing for the recognition of intracellular proteins.⁶³ TCR T-cell therapy addresses a broader range of tumor targets compared to CAR T-cell therapy which are limited to surface antigens. However, engineered TCR T-cells are MHC-restricted which necessitates HLA-matching of potential patients. A range of cloned TCRs and tumor-associated antigens have been identified and early clinical trials have demonstrated potential activity of TCR T-cell therapy against specific solid tumors.^64–66 In addition, CARs are thought to exhibit a greater tendency for tonic signaling, leading to heightened pro-inflammatory cytokine release and an increased incidence of CRS.⁶⁷ Similar to other ACTs, challenges with TCR T-cell in solid tumors include restricted T-cell infiltration, heterogeneous tumor antigen expression, and immunosuppressive TME.⁶⁸ The unique challenge of MHC restriction in this approach also limits its application to patients with compatible HLA types. Pivotal studies using TCR T-cell therapy in solid malignancies have evaluated TCR T-cells targeting MART1 in metastatic melanoma⁶⁹; NY-ESO-1 in synovial sarcoma,⁷⁰ soft tissue sarcoma,⁷¹ and myeloma⁷²; and MAGE-A4 in a range of solid tumors including esophageal, urothelial, osteosarcoma, melanoma, cervical, breast, and non-small cell lung cancer.⁷³ Potential toxicities associated with TCR T-cell therapy include cross-reactivity and off-target effects, CRS, and neurotoxicity. In MAGE-A3 TCR-T-cell studies, life-threatening neurological toxicity and deaths were observed, though to be due to cross reactivity with EPS8L2.^74,75 Recently, Afamitresgene autoleucel (afami-cel), an affinity-enhanced, MAGE-A4-specific TCR therapy, showed early promise in the treatment of refractory synovial sarcoma and myxoid round cell liposarcoma.⁷³ In the phase 2 SPEARHEAD-1 trial, 52 heavily pretreated patients with HLA-A*02 and MAGE-A4-expressing tumors received a single infusion of afami-cel following lymphodepletion.⁷⁶ Afami-cel achieved an ORR of 37% (39% in synovial sarcoma), with several patients with durable response over a median follow-up of 32.6 months.⁷⁶ Of the 19 patients with tumor shrinkage following treatment, 2 had complete response (CR). While CRS was common and occurred in 71% of patients, the majority of cases were low-grade. This study led to FDA approval of afami-cel in advanced synovial sarcoma in August 2024.⁷⁷ Key clinical trials of TCR T-cell therapy for solid tumors are ­summarized in Table S1.

TIL therapy

TIL therapy involves the isolation and ex vivo expansion of TILs.^78,79 These are often obtained from a surgically resected tumor in order to ensure adequate tissue for TIL isolation.^78,79 Tissues are processed and immune cells are cultured and expanded utilizing media containing IL-2 and other cytokines and growth factors over a period of 4-8 weeks.⁷⁹ Once TIL product manufacturing is complete, patients undergo preparatory lymphodepleting chemotherapy. TIL products are subsequently reinfused back into patients in conjunction with high dose of IL-2.^79,80 This process requires significant infrastructure and expertise thus limiting widespread expansion of this approach outside of highly specialized centers. Many patients develop significant toxicities related to high-dose IL-2 and associated capillary leak syndrome which require high level of medical support. There also exists the potential for on-target off-tumor toxicity in normal tissue as well. Furthermore, TIL manufacturing may potentially fail due to insufficient tissue for isolation and expansion of TILs or rapid progression of the patient’s cancer and deteriorating performance status during TIL manufacturing.⁸¹ As TILs have been derived from a patient’s tumor, they may harbor activity against a range of different tumor-associated antigens which may overcome the limitation of single tumor antigen targeting.^79,80,82 TIL therapy has shown efficacy in metastatic melanoma leading to the approval of lifileucel (Amtagvi) by the FDA in March 2024.^83,84 The outcomes of key TIL trials are outlined in Table S2.

Diagnosis, prevention, and management of CRS and ICANS

Cytokine release syndrome

Definition and mechanism of CRS

CRS is a hyper-inflammatory state driven by elevated levels of cytokines and chemokines such as IFN-y, TNF-α, GM-CSF, and various interleukins.^43,85–88 These molecules activate host antigen-presenting cells and T-cells, leading to a cascade of immune responses. Additional factors like IL-15, CRP, ferritin, and various chemokines contribute to the inflammatory milieu. The systemic activation of the vascular endothelium, particularly the release of IL-6, sets up a positive feedback loop exacerbating CRS.^89,90 This endothelial activation results in increased permeability and production of factors like vWF and Ang-2, destabilizing vascular integrity and contributing to hemodynamic instability and capillary leak.⁹¹ Nitric oxide production further amplifies vasodilation and hemodynamic instability. The factors underpinning the variable frequency and severity of CRS across various cell therapy modalities in solid tumors is less well understood and may be multifactorial related to both the specific therapeutic construct as well as underlying tumor and TME.⁸⁵

Incidence of CRS

A meta-analysis of 84 studies using CAR-T-cells reported CRS of all grade in solid tumors to be around 37%, with grade ≥3 of 19%.⁹² Subgroup analysis showed a higher incidence of CRS grade ≥3 in CEA CAR-T for solid tumor.⁹² BiTE treatment, due to its short half-life, can be discontinued without severe outcomes. The incidence of CRS for BiTE are less than that of CAR-T-cell therapy, with tumor burden and initial dose of BiTE to be determinant factors for CRS.⁴³

Symptoms/timeline of CRS

CRS classically presents with fever and fatigue which can progress rapidly to hypotension, tachycardia, and organ dysfunction.⁸⁷ The onset of CRS differs based on the specific CAR T-cell/BiTE product and underlying disease, typically occurring 0-5 days following infusion, with delayed onset up to 3 weeks.⁸⁵ CRS onset for BiTE often occurs earlier due to faster cytokine release kinetics.^39,85 The peak of CRS risk varies with different ACT or CAR products. Tisagenlecleucel’s CRS risks peak at Day 3 and 7 for acute lymphoblastic leukemia (ALL) and diffuse large B cell lymphoma (DLBCL) patients, respectively.^93,94 On the other hand, CRS risk peaks around 2 days post-infusion of Axicabtagene ciloleucel or Brexucabtagene autoleucel.^95–97

Early identification and proactive interventions for individuals at high risk (eg, highly proliferative, aggressive disease, high tumor burden, high-risk CAR T-cell, or BiTE constructs) may improve outcomes. Differentiating CRS from febrile ­neutropenia is crucial particularly given that many patients will be rendered neutropenic by preparatory lymphodepleting chemotherapy. As such, empiric antibiotics should be initiated promptly as well as consideration for anti-virals and anti-fungals when appropriate. Although CRP and ferritin have been proposed as predictive biomarkers for CRS, the low positive predictive value of CRP and its inconsistent elevation prior to onset limit its utility as early indicator of CRS.⁹⁸ Emerging evidence highlights the association between elevated serum IL-6 levels and CRS, positioning IL-6 as a promising biomarker for CRS prediction particularly at centers where this can be rapidly obtained.^89,99

Management of CRS

Management of CRS in patients undergoing CAR-T-cell therapy is guided by a standardized grading system (Table 3).^87,100 There is no comprehensive guideline for management of CRS for BiTE and other cell therapies, although management algorithms are often adapted from those established for CAR T-cell therapy.^46,48 The current CRS grading system was developed by the American Society of Transplantation and Cellular Therapy (ASTCT) (Table 3).⁸⁷ These guidelines identify fever temporally linked to CAR T-cell therapy as a prerequisite for CRS diagnosis. Hypoxia and hypotension are the primary factors determining CRS grade and severity, with management tailored to the assigned grade (Table 3).

Table 3.

CRS grading and management for CAR-T and BiTE therapies.

Grade	Key diagnostic criteria	Common initial measures	CAR-T management	BiTE management
1	Fever $\ge$38C; no hypotension or hypoxia	Antipyretics, IV hydration Infectious workup and management Supportive care	Consider Tocilizumab if fever persists >24 h or high-risk comorbidities	Hold or slow infusion if fever persists during step-up dosing Interrupt step-up dosing; dexamethasone 10 mg IV Q6h if progressive
2	Fever plus hypotension responsive to fluids or hypoxia requiring low-flow O2 $\le$6 L/min	Supplemental O2, IV fluids Supportive care	Tocilizumab 4-8mg/kg (max 800 mg; may repeat Q8h up to 4 doses) ±Corticosteroid (dexamethasone 10-20 mg IV Q6h or methylprednisolone equivalent) if no response	Stop infusion and delay until CRS gr $\le$ 1 Dexamethasone 10 mg IV Q6h per label; tocilizumab if not improving Delay at least 72 h if CRS gr 2 occurs during infusion. Permanently discontinue therapy if CRS is not improved to gr $\le$ 1 within 7d
3	Fever plus hypotension requiring vasopressor(s) or hypoxia requiring high-flow O2 (≥6 L/min, HFNC, NIV)	ICU consult; aggressive fluids; high-flow O2; vasopressor(s) Supportive care	Tocilizumab 4-8mg/kg Q8h up to 4 doses ± Siltuximab Corticosteroid (dexamethasone 10-20 mg IV Q6h or methylprednisolone equivalent) Consider anakinra if refractory	Stop infusion and delay until CRS gr $\le$1 Dexamethasone 10 mg IV Q6h per label Consider tocilizumab 4-8 mg/kg (max 800 mg; may repeat Q8h up to 4 doses) Consider siltuximab/anakinra if refractory Delay at least 72 h if CRS gr 3 occurs during infusion. Permanently discontinue therapy if CRS is not improved to gr $\le$2 within 5d, or to gr $\le$1 within 7d, or if CRS gr 3 during initial dose, or if 2 separate CRS gr 3
4	Life threatening: multiple vasopressors and/or Positive pressure (CPAP, BiPAP, MV)	Full ICU support; invasive monitoring Supportive care	Tocilizumab 4-8 mg/kg Q8h up to 4 doses ± Siltuximab/anakinra Methylprednisolone 1 g/day Consider CAR-T safety switch if available	Dexamethasone 10 mg IV Q6h Tocilizumab 4-8 mg/kg (max 800 mg; may repeat Q8h up to 4 doses) Consider siltuximab/anakinra if refractory Permanently discontinue therapy

Abbreviations: BiTE, bispecific T-cell engager; CAR-T, chimeric antigen receptor T-cell; CRS, cytokine release syndrome.

Patients with grade 1 CRS present with fever without hypotension or hypoxia and may be managed with supportive care including antipyretics, fluids, and close monitoring. Grade 2 CRS, characterized by fever with mild hypotension or hypoxia, may require intravenous (IV) hydration and supplemental oxygen. Tocilizumab, an IL-6 inhibitor, should be considered for grade 2 CRS or higher. The most common side effects of tocilizumab include headache and hypertension, but, rarely, hepatotoxicity ranging from mild transaminase to severe drug-induced liver injury can occur and should be monitored.^101,102 In cases of clinical deterioration or high risk of severe CRS, additional treatment with corticosteroids may be necessary. For grade 3 CRS, which involves fever and hypotension refractory to fluids, ICU management and urgent treatment with tocilizumab and corticosteroids is recommended. Grade 4 CRS is defined as hypotension necessitating multiple vasopressors and/or respiratory distress requiring positive pressure ventilation. Management of grade 4 CRS includes high-dose methylprednisolone (1000 mg/day). In the events where CRS is accompanied by ICANS, corticosteroids are preferred frontline agents given their CNS penetrance in addition to tocilizumab to manage concomitant CRS. Siltuximab, a chimeric immunoglobulin monoclonal anti-IL6 antibody was also shown to effectively manage CRS refractory to tocilizumab.^103,104 Refractory CRS, which does not respond to standard treatments, may be managed with anakinra, an IL-1 receptor antagonist, after tocilizumab and corticosteroids.¹⁰⁵ Other potential therapeutics to reduce CRS include JAK/STAT inhibitors,^106,107 tyrosine kinase inhibitors (TKIs),¹⁰⁸ IFN-gamma-blocking antibody,¹⁰⁹ anti-thymocyte globulin,¹¹⁰ anti-CD52,¹¹¹ cyclophosphamide,¹¹² and adenosine receptor agonists.¹¹³

In contrast to higher risk CAR T and engineered TCR therapy, many BiTE molecules exhibit milder CRS (eg, tarlatamab). CRS in low-risk BiTEs may be managed with dosing interruption and corticosteroid administration.^46,48 Tocilizumab, an IL-6 inhibitor, can be used in combination with corticosteroids and dosing interruption.^114,115

Supportive care is essential in managing CRS, addressing immediate complications, and minimizing the long-term impact on patient well-being.¹¹¹ This includes close monitoring for infections, ensuring adequate nutrition and hydration, and managing pain and other symptoms¹¹¹ (Table 3). A multidisciplinary approach involving specialists in critical care, hematology/oncology, neurology, and infectious diseases is often necessary to provide comprehensive care.¹¹¹

Overall, a comprehensive approach to CRS management, including early recognition, appropriate grading, and timely intervention, is crucial for optimizing outcomes and ensuring the safety of patients undergoing ACTs.

Management of severe and/or refractory CRS

Evidence supporting pharmacologic intervention for severe or refractory for CRS remains limited and are derived primarily from single-institution experiences, case reports, and preclinical data. Currently, additional therapeutics that may be utilized in the management of refractory CRS include anakinra¹¹⁶ (IL-1 inhibitor); ruxolitinib and itacitinib (JAK/STAT inhibitors)^106,107; and dasatinib (multi-targeted TKI).¹⁰⁸ T-cell targeted therapies such as antithymocyte globulin (ATG),¹¹⁰ alemtuzumab (anti-CD52), and cyclophosphamide¹¹² have also been explored for potential use in treatment of relapsed/refractory CRS/ICANS. However, the data supporting the use of each of these agents are limited. Among these agents, anakinra, an IL-1 receptor antagonist capable of crossing the blood–brain barrier, has shown encouraging activity in steroid- and tocilizumab-refractory CRS/ICANS by reducing inflammatory cytokines and promoting clinical improvement.^105,117,118

Potential strategies in preventing and reducing CRS

Strategies have been implemented in efforts to prevent and reduce incidence and/or severity of CRS. The role of prophylactic anti-IL-6 agents and corticosteroids in CRS and/or ICANS is under investigation.^119–121 In general, strategies to reduce CRS incidences include consideration for lower dose of CAR T-cells administered to patients with high disease burden, fractionated CAR T-cell administration and step dosing of the BiTEs.^46,48,100 It is important to note that previous studies showed reduction of CRS rate while preserving treatment efficacy with early/­prophylactic use of corticosteroid or tocilizumab.^122,123

Immune effector cell-associated neurotoxicity syndrome

Definition and mechanism of ICANS

Neurotoxicity is a common complication in CAR T-cell therapy, occurring in up to 64% of cases.^93,94,96,119 Currently, there are several hypotheses on the mechanisms of neurotoxicity (ICANS), including CAR-T-cell trafficking to the CSF, endothelial activation, disruption of the blood–brain barrier, and activation of myeloid cell in the CNS.^{85,116,124,125}

Incidence of ICANS

The incidence of ICANS varies by CAR construct, with high-grade ICANS occurring in approximately 45% of CD28-based CAR T-cell therapies and 13% of 4-1BB-based CAR T-cell therapies. Risk factors for ICANS include concurrent CRS,¹²⁴ pre-existing neurologic co-morbidities,¹²⁶ high disease burden,^124,127 amount of infused CAR-T-cells,¹²⁴ high peak of CAR T-cell expansion,¹²⁷ high pre-treatment LDH,¹²⁸ low platelets or endothelial growth factors,¹²⁷ high serum Ang-2/Ang-1 ratio,¹²⁴ and elevated ferritin within 72 h of CAR T-cell administration.¹²⁷ ICANS was reported with Tarlatamab infusion, occurring in 8% of patients (grade 1/2) at a 10-mg dose, and at a 100-mg dose, in 23% (grade 1/2) and 5% (grade 3/4). Most cases occurred within cycle 1, with a median onset of 5 days.⁴⁸

Symptoms of ICANS

ICANS can present concurrently with CRS, emerge after CRS has resolved, or occur independently, with the latter typically producing milder manifestations. Symptoms typically emerge 4-5 days following CAR T-cell therapy but may be delayed by 3-4 weeks. Early and specific signs include expressive aphasia and handwriting changes, alongside common symptoms like attention impairment and confusion (Table 4).^95,127,129 Importantly, early signs of ICANS may be subtle but can progress rapidly to severe ICANS and thus should be monitored for carefully and treated early. Severe ICANS can progress to coma, seizures, or cerebral edema, with fatal cerebral edema reported in 3% of CD19-directed CAR T-cell therapy cases, often linked with concomitant severe CRS.^124,130 The ASTCT guidelines for grading ICANS combine the CARTOX-10 screening tool with the ICE score, assessing consciousness, seizures, motor function, and signs of elevated ICP (Table 4).⁸⁷ Grade 1 ICANS involves mild symptoms like delayed responses or mild inattention, often seen during CRS (Table 4). Grade 2 includes expressive aphasia and difficulty in communication. Grade 3 indicates severe global aphasia or excessive drowsiness. Grade 4 is characterized by unarousable stupor or coma, often requiring intubation (Table 4).

Table 4.

ICANS grading and management for CAR-T and BiTE therapies.

Grade	ICE score	Key ICE test components	Common initial measures	CAR-T management	BiTE management
1	7-9	Orientation (year, month, city, hospital) = 4 pts Naming 3 objects = 3 pts Follow simple command = 1 pt Write a sentence = 1 pt Count backward from 100 by 10 s = 1 pt	Neuro vitals Q2-4h Seizure prophylaxis (levetiracetam 500-1000 mg BID) EEG, Neuroimaging, Lumbar puncture	Supportive care Tocilizumab only if concurrent CRS	Hold or slow infusion if symptoms worsen Supportive care Interrupt step-up dosing Consider dexamethasone 10 mg IV Q6h if persistent
2	3-6	Same ICE components as above	Continue neuro vitals and seizure prophylaxis EEG, Neuroimaging, lumbar puncture	Supportive care Dexamethasone 10 mg IV Q6h or methylprednisolone equivalent	Hold infusion Dexamethasone 10 mg IV Q6h Delay infusion until ICANS gr $\le$1
3	0-2	Same ICE components as above Any seizure, focal deficit, or raised ICP	Consider ICU management Airway protection Anti-epileptic management Continuous EEG Aggressive ICP management	Supportive care Dexamethasone 10-20 mg IV Q6h or methylprednisolone equivalent Consider anakinra Treat local/focal edema with methylprednisolone	Stop infusion until ICAN gr $\le$1 Dexamethasone 10 mg IV Q6h Permanently discontinue therapy if ICANS is not improved to gr $\le$1 within 7d, OR if ICANS gr 3 reoccurs within 7d of re-initiation
4	0	Same ICE components as above Coma, herniation	ICU management Intubation/ventilation Full neurocritical care Anti-epileptic management Aggressive ICP management	Supportive care Methylprednisolone 1 g/day Add anakinra	Permanently discontinue therapy Methylprednisolone 1 g/day IV × 3 Consider additional therapies

Abbreviations: BiTE, bispecific T-cell engager; CAR-T, chimeric antigen receptor T-cell; CRS, cytokine release syndrome; EEG, electroencephalogram; ICANS, immune effector cell–associated neurotoxicity syndrome; OS, overall survival.

Management of ICANS

Steroids are the primary treatment for isolated ICANS.^85,87 Due to poor blood–brain barrier penetration and risk of worsening neurotoxicity, tocilizumab should only be used in instances of concurrent ICANS and CRS.⁹⁵ Grade 1 ICANS requires careful management due to the potential for rapid progression. IV corticosteroids, supportive care, and close monitoring should be initiated as well as imaging to evaluate for cerebral edema or other ICANS-related changes and to rule out alternative diagnoses like stroke. EEG can help detect subclinical seizures, and neurology consultation may be warranted (Table 4).^85,87,95 Lumbar puncture often reveals elevated protein and lymphocytes in the CSF (Table 4).

Corticosteroids are recommended for Grade 2 ICANS and higher, with high-dose methylprednisolone advised for grade 4 (Table 4). Anakinra has shown clinical benefits in glucocorticoid-resistant ICANS and should be considered early.^{105,117,131,132} Severe ICANS may require ICU care and airway protection.^95,133 Seizure prophylaxis with levetiracetam is recommended for patients with a seizures or pre-existing neurological condition, though the role of routine primary prophylaxis is not clear.^124,127 Active seizures should be managed with benzodiazepines or other antiepileptics. ICANS management in BiTE therapy also includes dosing interruption, similar to that of CRS.^46,48

As previously mentioned, supportive care plays an important role in managing CRS and ICANS, as patients are at a higher risk of increased morbidity and mortality. Consultation with specialists (ICU, neurology, hematology/immunology, ­infectious diseases) and allied health are recommended in supporting patients through toxicity management of cell therapy. A comprehensive grade-based management of CRS and ICANS for both CAR-T and BiTE therapies is summarized in Table 5.

Table 5.

Comprehensive grade-based management of CRS and ICANS for CAR-T and BiTE therapies.

Grade	CRS		ICANS
1	Antipyretics, IV hydration, infectious work-up, continuous vitals/pulse-oximeter.		Neuro checks q2–4h; seizure prophylaxis (levetiracetam 500–1000 mg BID); EEG/neuroimaging if focal deficit.
	CAR-T	Consider tocilizumab 4-8 mg/kg IV (max 800 mg; repeat q8 h up to 4 doses) if fever >24 h/high-risk comorbidities.	Supportive care; tocilizumab only if concurrent CRS.
	BiTE	Hold or slow infusion during step-up dosing; interrupt if progressive; dexamethasone 10 mg IV q6h if persistent.	Hold or slow infusion; supportive care; interrupt step-up dosing; consider dexamethasone 10 mg IV q6h if persistent.
2	Supplemental O₂ (≤6 L/min), IV fluids, telemetry.		Continue neuro monitoring and seizure prophylaxis; EEG/neuroimaging/lumbar puncture as indicated.
	CAR-T	Tocilizumab 4-8 mg/kg IV q8 h (max 4 doses) ± dexamethasone 10-20 mg IV q6 h if no response.	Dexamethasone 10 mg IV q6 h (or methylprednisone equivalent); escalate if no improvement.
	BiTE	Stop infusion; dexamethasone 10 mg IV q6h per label; tocilizumab if not improving; delay ≥72 h if CRS gr 2 during infusion; permanently discontinue if not ≤ gr 1 within 7 days.	Hold infusion; dexamethasone 10 mg IV q6 h; delay restart until ICANS ≤ gr 1.
3	ICU support; aggressive fluids; high-flow O₂ (≥6 L/min), vasopressors.		ICU support; airway protection; continuous EEG; aggressive ICP management.
	CAR-T	Tocilizumab 4-8 mg/kg IV q8 h (×4) ± siltuximab/anakinra; dexamethasone 10-20 mg IV q6 h (or methylpred 1-2 mg/kg/day)	Dexamethasone 10-20 mg IV q6 h or methylprednisone 1-2 mg/kg/day; consider anakinra if steroid-refractory.
	BiTE	Stop infusion; dexamethasone 10 mg IV q6h; tocilizumab 4-8 mg/kg IV q8 h (×4); ± siltuximab/anakinra; delay ≥72 h if gr 3; permanently discontinue if not ≤gr 2 within 5 days or ≤gr 1 within 7 days, or if gr 3 occurs at initial dose, or if 2 separate gr-3 events.	Stop infusion; dexamethasone 10 mg IV Q6h; permanently discontinue if ICANS not ≤grade 1 within 7 days or if grade 3 recurs within 7 days of re-initiation.
4	Full ICU support; invasive monitoring; mechanical ventilation as needed.		Full neurocritical care; intubation/ventilation; aggressive ICP management; anti-epileptics.
	CAR-T	Tocilizumab 4-8 mg/kg IV q8 h (×4) ± siltuximab/anakinra; methylprednisolone 1 g/day IV ×3 days, then taper; consider CAR-T safety switch if available.	Methylprednisolone 1 g/day IV ×3 days, add anakinra if needed.
	BiTE	Permanent discontinuation of therapy; dexamethasone 10 mg IV Q6h, tocilizumab 4-8 mg/kg IV q8 h (×4); ± siltuximab/anakinra	Permanent discontinuation; methylprednisolone 1 g/day IV × 3; consider additional therapies as adjunct.

Abbreviations: BiTE, bispecific T-cell engager; CAR-T, chimeric antigen receptor T-cell; CRS, cytokine release syndrome; EEG, electroencephalogram; ICANS, immune effector cell–associated neurotoxicity syndrome; OS, overall survival.

Conclusion

Adoptive T-cell therapies offer promise in selected solid tumors. While these therapies have demonstrated substantial potential, their success in solid tumors has been constrained by unique challenges, including the immunosuppressive TME, antigen escape, and on-target, off-tumor toxicities. Advances in genetic engineering, tumor antigen and TCR identification, and combination strategies with immune modulators are paving the way for more effective and durable responses in patients with solid malignancies.

Despite these advancements, toxicity management remains a critical area for improvement. CRS, ICANS, and other therapy-specific toxicities such as off-target effects and lymphodepletion-associated complications continue to complicate widespread clinical adoption. Emerging strategies, including optimized cell designs, biomarker-driven monitoring, and prophylactic interventions hold promise for mitigating these adverse events.

Moving forward, the integration of widespread provider education, more effective algorithms for the management of toxicities, and the development of safer and more efficient manufacturing and infusion processes will be essential for broadening the applicability of ACTs. As these therapies evolve, their potential to transform the treatment landscape for solid tumors will depend not only on enhancing efficacy of these treatments but also on minimizing the risk of toxicity in order to ensure that patients can benefit from these cutting-edge innovations with improved safety and quality of life.

Author contributions

Tuan Hoang (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Visualization, Writing—original draft, Writing—review & editing), Sameena Khan (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Writing—original draft, Writing—review & editing), and Adrian Sacher (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Writing—original draft, Writing—review & editing)

Supplementary material

Supplementary material is available at The Oncologist online.

Funding

None declared.

Conflicts of interest

T.H. has no conflict of interest to declare. S.K. has received honorarium and travel expenses from Pfizer, Flatiron Health, BMS, Takeda, Novartis, AstraZeneca. S.K is a clinical trial PI and received funding from OSE pharmaceuticals. A.S. is a clinical trial PI and received institutional research funding from AstraZeneca, Amgen, Genentech, Merck, Lilly, Pfizer, BMS, Spectrum, GSK, Iovance, CRISPR Therapeutics, BridgeBio, HotSpot Therapeutics, AdaptImmune. A.S. is a member of the advisory committee (no personal fees) for Genentech, Amgen, and Merck. A.S. received travel expenses for clinical trial investigator meetings from Amgen, Merck, and Genentech-Roche.

Data Availability

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.