The Next Frontier of Cellular Therapy in Oncology

The National Immune Cell Effector Therapy (ICE-T) Symposium convened on September 27, 2025, in Kansas City, Kansas, bringing together hematologists, oncologists, nurses, and researchers to address the rapidly evolving landscape of immune effector therapies. Sessions spanned chimeric antigen receptor (CAR) T-cell therapy and bispecific T-cell engagers, with emphasis on their expanding role in multiple myeloma, lymphomas, and, increasingly, autoimmune diseases such as systemic lupus erythematosus. Presenters explored both the clinical promise and the practical challenges of these therapies across a wide range of settings. Collectively, the meeting underscored a shift toward making cellular therapy safer, more scalable, and more equitably accessible.

Parkinsonism and Other Non–Immune Effector Cell–Associated Neurotoxicities Related to BCMA-Directed CAR T-Cell Therapy

Beyond ICANS

Recognizing a New Neurotoxicity Spectrum

Chimeric antigen receptor (CAR) T-cell therapy targeting B-cell maturation antigen (BCMA) has transformed the therapeutic landscape of relapsed and refractory multiple myeloma. Although early attention has focused on well-characterized immune-mediated toxicities such as cytokine release syndrome (CRS) and immune effector cell–associated neurotoxicity syndrome (ICANS), emerging clinical experience suggests a broader and less well-defined neurologic spectrum.

Paul Worsley, BS, BA, a second-year medical student at the University of Kansas Medical Center in Kansas City, and colleagues presented a real-world analysis examining non-ICANS neurologic toxicities associated with BCMA-directed CAR T-cell therapies, specifically ciltacabtagene autoleucel (cilta-cel) and idecabtagene vicleucel (ide-cel), during the Second Regional Immune Cell Effector Therapy (ICE-T) Congress, held in September 2025. Their findings highlight a distinct pattern of delayed, often parkinsonian-like symptoms that challenge traditional toxicity frameworks and demand increased clinical awareness.

Defining the Population

Real-World CAR T-Cell Experience

The analysis included 128 patients treated with BCMA-directed CAR T-cell therapy between December 2019 and December 2024. Among these, 82 patients received cilta-cel and 46 received ide-cel, reflecting contemporary treatment distribution in a real-world academic center.

Within this cohort, a subset of patients developed neurologic symptoms not attributable to ICANS, forming the basis of a focused evaluation of these atypical toxicities. A total of 18 patients were identified with clinically significant neurologic manifestations, representing approximately 14% of the treated population (N = 128).

A Distinct Neurotoxicity Profile

Beyond Acute Immune Activation

Unlike ICANS, which typically occurs early and is temporally linked to systemic inflammation, the neurologic symptoms observed in this cohort followed a more delayed and heterogeneous course. Clinical presentations were characterized by movement disorders, including tremor and parkinsonian features, suggesting involvement of extrapyramidal pathways rather than diffuse encephalopathy.

Timeline analyses demonstrated variability in symptom onset and progression, underscoring the unpredictability of this toxicity. Importantly, these manifestations were not consistently associated with severe CRS or ICANS, indicating a potentially distinct underlying mechanism separate from acute cytokine-mediated neuroinflammation.

Product-Specific Patterns

Cilta-cel vs Ide-cel

A notable observation was the apparent difference in neurotoxicity profiles between CAR T-cell products. Of the 18 patients with neurologic symptoms, the majority (n = 14) had received cilta-cel, whereas only 4 cases were observed among ide-cel recipients.

Although the sample size is limited, this distribution raises the possibility of product-specific neurotoxicity risks, potentially related to differences in CAR construct design, antigen-binding characteristics, or cellular expansion kinetics. These findings align with broader clinical observations suggesting a higher incidence of delayed neurologic syndromes with cilta-cel, warranting further mechanistic investigation.

Clinical Characterization

Imaging and Cerebrospinal Fluid Findings

Diagnostic evaluation included neuroimaging and cerebrospinal fluid analysis. MRI findings were variable and did not consistently reveal structural abnormalities, reflecting the functional rather than structural nature of the observed neurologic deficits.

Similarly, lumbar puncture findings were heterogeneous and did not identify a unifying inflammatory or infectious etiology. These results reinforce the complexity of this toxicity and suggest that conventional diagnostic tools may have limited sensitivity in detecting underlying pathophysiology.

Management and Outcomes

A Prolonged Clinical Course

Management of non-ICANS neurologic toxicities was variable and guided by symptom severity, incorporating both supportive care and targeted immunomodulatory strategies. Corticosteroids formed the therapeutic backbone, with 50% of patients (n = 9) receiving steroids alone. Escalation approaches were less common, including corticosteroids combined with intravenous immunoglobulin (IVIG) in 11% (n = 2) and corticosteroids with anti–IL-1 monoclonal therapy in another 11% (n = 2). Notably, 28% of patients (n = 5) did not receive targeted immunomodulatory treatment, reflecting either milder clinical presentations or uncertainty in management.

Clinical outcomes were heterogeneous. Among the 18 affected patients, 11 (61%) experienced eventual symptom resolution, whereas 7 (39%) had persistent neurologic deficits. Recovery was often prolonged, with a median time to resolution of 62 days, substantially longer than typically observed with ICANS. At a median follow-up of 213 days, 56% of patients remained in remission, 11% were alive with relapsed disease, and 33% had died, underscoring both the clinical burden of delayed neurotoxicity and the high-risk nature of this heavily pretreated population.

Clinical Implications

Expanding the Neurotoxicity Framework

This analysis underscores the need to broaden the current conceptual framework of CAR T–associated neurotoxicity. Although ICANS remains the dominant paradigm, clinicians must now recognize a second category of neurologic complications characterized by delayed onset, movement disorders, and prolonged recovery.

The observed incidence of approximately 14% suggests that these events are not rare and may become increasingly relevant as CAR T-cell therapy is deployed earlier in the disease course and in larger patient populations. Early recognition, longitudinal neurologic monitoring, and multidisciplinary management will be essential to mitigate long-term morbidity.

CAR T-Cell Therapy in Autoimmune Disease

From Malignancy to Autoimmunity

CAR T-cell therapy has transformed the treatment of hematologic malignancies, delivering deep and durable responses in diseases once considered refractory. This success has prompted a paradigm shift: leveraging the same technology not to eliminate malignant clones, but to reset dysregulated immune systems in autoimmune disease.

At the ICE-T Congress, Kimberly Liang, MD, MS, an associate professor of medicine in the Division of Allergy, Immunology, and Rheumatology, and director of the Lupus Center and Vasculitis Center at the University of Kansas Medical Center, presented a comprehensive overview of CAR T-cell therapy in autoimmune diseases, focusing on systemic lupus erythematosus (SLE), lupus nephritis, idiopathic inflammatory myopathies, and systemic sclerosis. Her presentation highlighted that a central concept—targeted depletion of autoreactive B cells and plasma cells—may enable an immune system reboot, offering the possibility of sustained, drug-free remission in diseases historically managed with chronic immunosuppression.

Mechanistic Rationale

Targeting the Source of Autoimmunity

Autoimmune diseases such as SLE are fundamentally driven by autoreactive B cells and plasma cells, which produce pathogenic autoantibodies and immune dysregulation. CAR T-cell therapy enables precise, major histocompatibility complex (MHC)–independent targeting of these cells through engineered T-cell receptors, leading to rapid and profound depletion.

Beyond simple cytotoxicity, this approach disrupts key immunologic pathways, including autoantibody production, complement activation, and interferon signaling. Importantly, reconstitution of the B-cell compartment following CAR T-cell therapy appears to favor naive, non–class-switched populations, suggesting a reprogramming of immune tolerance rather than transient suppression.

CD19 CAR T in SLE

Proof of Concept for Immune Reset

Early clinical experience with CD19-directed CAR T-cell therapy in refractory SLE has demonstrated striking efficacy. In small case series and compassionate-use cohorts, patients with severe, treatment-resistant disease achieved rapid clinical improvement, normalization of serologic markers including anti–double-stranded DNA (dsDNA) antibodies and complement levels, and sustained drug-free remission.

In one reported cohort, all treated patients achieved remission, with Systemic Lupus Erythematosus Disease Activity Index 2000 (SLEDAI-2K) scores reduced to 0 and complete resolution of disease activity across organ systems. Notably, these responses were maintained despite re-emergence of B cells, supporting the concept that CAR T-cell therapy induces durable immune reprogramming rather than continuous depletion. These findings are reinforced by longer-term follow-up data extending up to 29 months, in which all treated patients with SLE maintained complete clinical remission with SLEDAI-2K scores of 0, sustained disappearance of anti–dsDNA antibodies, normalization of complement levels, and resolution of proteinuria, highlighting both the depth and durability of immune reset.

Equally important, safety profiles in autoimmune populations appear favorable, with predominantly mild CRS and no significant neurotoxicity, contrasting with experiences in oncology.

Expanding Indications

Beyond Lupus

The therapeutic signal extends beyond SLE. In patients with idiopathic inflammatory myopathies and systemic sclerosis, CAR T-cell therapy has demonstrated meaningful clinical responses for all patients (SLEDAI-2K score, 0), including normalization of muscle enzymes, improvement in functional status, and reduction in systemic disease activity.

Across these conditions, a consistent theme emerges: Patients with severe, refractory disease can achieve rapid and sustained remission, often accompanied by complete discontinuation of glucocorticoids and other immunosuppressive therapies. This raises the possibility that CAR T-cell therapy may shift treatment goals from disease control to true disease modification. In idiopathic inflammatory myopathies, all treated patients achieved major clinical responses per American College of Rheumatology/European Alliance of Associations for Rheumatology criteria within 3 months, accompanied by normalization of creatine kinase levels and sustained recovery of both muscular and extramuscular disease activity. Similarly, patients with systemic sclerosis demonstrated consistent reductions in global and cutaneous disease activity.

BCMA and Dual-Target Strategies Deepening Immunologic Control

BCMA Alone

BCMA-directed CAR T-cell therapy has also demonstrated promising activity in lupus nephritis, supported by early-phase clinical data. In an open-label, interventional single-arm study (NCT06277427) of 7 patients with biopsy-confirmed refractory lupus nephritis, treatment following fludarabine/cyclophosphamide lymphodepletion resulted in rapid and profound immunologic responses. Peripheral B cells were eliminated within 1 month of infusion and typically reconstituted by 3 months. Clinically, disease activity improved markedly, with SLEDAI-2K scores declining from a median of 18 (range, 10-22) at baseline to 0 (range, 0-4) at last follow-up. Five of 7 patients (71%) achieved complete response (CR) per Definition of Remission in SLE criteria. Therapy was well tolerated, with hypogammaglobulinemia as the most common adverse event and only a single case of grade 1 CRS reported.

Dual Target

Building on this approach, dual-target CAR T strategies combining CD19 and BCMA aim to eliminate both autoreactive B cells and long-lived plasma cells. In a phase 1 open-label study (NCT04162353 and NCT05474885) of 13 patients with refractory SLE or lupus nephritis, infusion of compound CAR T cells following lymphodepletion led to rapid and sustained immunologic responses. Autoantibodies became undetectable in 10 of 13 patients within 12 months, complement levels normalized, and B-cell recovery occurred at approximately 6 months. Clinical improvement was substantial, with mean SLEDAI scores decreasing from 10.6 to 2.7 within 3 months, alongside significant renal recovery in patients with lupus nephritis. Notably, 2 patients achieved prolonged medication-free remission lasting up to 46 months. Treatment was well tolerated, with only mild CRS observed.

Safety Profile

A Distinct Advantage in Autoimmune Disease

One of the most notable findings across studies is the relatively favorable safety profile of CAR T-cell therapy in autoimmune disease compared with hematologic malignancies. Patients with SLE and related disorders consistently experience lower rates of severe CRS and minimal neurotoxicity. However, long-term efficacy and safety remain under study.

This difference may reflect lower disease burden, distinct immune environments, or differences in CAR T expansion kinetics. Although hypogammaglobulinemia and transient cytopenias remain common, these effects are generally manageable and limited in duration. Importantly, infection rates appear lower than expected, particularly in the context of discontinuation of chronic immunosuppressive therapies.

Emerging Platforms

Toward Scalable and Rapid CAR T-Cell Therapy

Next-generation CAR T platforms are focused on improving accessibility, safety, and manufacturing efficiency. Rapid manufacturing approaches, including technologies capable of producing CAR T cells within days, are reducing time to treatment and enabling broader clinical application.

Agents such as resecabtagene autoleucel (rese-cel) and rapcabtagene autoleucel represent evolving strategies designed to optimize expansion kinetics, reduce toxicity, and enhance durability. Early data suggest these therapies maintain efficacy while offering improved logistical feasibility, a critical factor for integrating CAR T into routine autoimmune care.

Early-phase trials of CAR T-cell therapy in autoimmune disease follow a consistent model of lymphodepletion followed by autologous CD19– or dual-target CAR T infusion, leading to rapid B-cell depletion, early biomarker normalization, and marked reduction in disease activity. B-cell recovery within 2 to 3 months supports an immune reset rather than sustained depletion. Across studies, safety is favorable with predominantly low-grade CRS and minimal neurotoxicity, whereas responses are often durable and achieved without ongoing immunosuppression.

Clinical Positioning

Not Yet Standard, but Rapidly Evolving

Despite compelling early data, CAR T-cell therapy remains investigational in autoimmune disease. It is not yet incorporated into standard management guidelines for SLE or related conditions, reflecting the need for larger, controlled trials and long-term safety data.

However, the growing number of ongoing phase 1 and 2 studies, coupled with early regulatory discussions, signals rapid momentum. As evidence matures, CAR T-cell therapy may move from compassionate use in refractory disease to earlier-line intervention in selected high-risk patients.

TACTUM-23: Trends in Access to Cellular Therapies in Multiple Myeloma and Beyond

Innovation Without Access

Cellular therapies, including CAR T and T-cell engagers (TCEs), have transformed outcomes in multiple myeloma, offering unprecedented response rates and survival benefits in relapsed and refractory disease. However, these advances have exposed a critical gap between therapeutic innovation and real-world accessibility.

In her presentation at the ICE-T Congress, Aroog Khaliq, MD, internal medicine resident at HCA Overland Park Regional in Kansas, explored the structural, financial, and logistical barriers that limit access to these therapies, emphasizing that clinical efficacy alone does not translate into equitable care. Her discussion reframed the modern challenge in multiple myeloma treatment: not how to treat, but who is able to receive treatment.

A Real-World Case

When Eligibility Is Not Enough

Khaliq illustrated this challenge through a representative patient case. Despite meeting all clinical criteria for CAR T-cell therapy and successfully navigating approval and cell collection, treatment was delayed due to lack of caregiver support. This nonmedical barrier ultimately postponed therapy beyond the expected vein-to-vein timeline, highlighting how social and logistical constraints can disrupt even well-coordinated treatment plans.

This case underscores a central theme: Access to CAR T-cell therapy depends not only on disease biology and treatment eligibility, but also on infrastructure, social support, and system-level coordination.

The TACTUM-23 Survey

Physician Perspectives on Barriers

The TACTUM-23 (Trends in Access to Cellular Therapies in Multiple Myeloma) survey was a 24-question medical survey conducted between June and August 2023. It revealed distinct and quantifiable differences in perceived barriers between referring and treating oncologists. Among referring oncologists, financial toxicity emerged as the most frequently cited barrier to CAR T-cell therapy (45%) and TCEs (43%), followed by relocation requirements (27%) and hospitalization needs (29%), with toxicity concerns cited less frequently (18% and 14%, respectively).

In contrast, treating oncologists identified logistical and system-level constraints as the dominant barriers. Regarding CAR T-cell therapy, limited manufacturing capacity, reflected by the number of available production slots, was the leading concern (69%), followed by insurance and financial barriers (15%) and toxicity monitoring requirements (8%). Regarding TCEs, hospitalization requirements (63%) were the most commonly cited barrier, with additional challenges including the need for frequent dosing at specialized centers (17%) and ongoing toxicity monitoring (13%).

These findings highlight a clear divergence in perspective: Referring physicians emphasize patient-level financial and access barriers, whereas treating physicians are more focused on system capacity and operational constraints. This misalignment suggests that barriers to cellular therapy access are multifactorial and occur at different stages of the treatment pathway, from referral to delivery.

Infrastructure and Role Clarity

Who Treats, Who Refers?

Although more than 70% of referring centers reported readiness to manage complications such as CRS and ICANS, uncertainty remains regarding roles and responsibilities across institutions. Questions surrounding who should administer therapy, manage toxicities, and coordinate follow-up care continue to complicate referral pathways.

These ambiguities contribute to delays in care delivery and underscore the need for clearer, standardized care models to support the expanding use of cellular therapies.

Disparities in Access

The Impact of Social Determinants

Access to CAR T-cell therapy is not evenly distributed. Socioeconomic and geographic factors play a significant role in determining who ultimately receives treatment. Higher social vulnerability is associated with reduced likelihood of accessing cellular therapies, while racial and ethnic minorities remain underrepresented in clinical trials through real-world implementation.

Insurance status further influences access patterns. Differences in coverage among commercial insurance, Medicare, and Medicaid populations reflect systemic inequities that persist even after regulatory approval. Additionally, longer travel distances to treatment centers significantly reduce the likelihood of receiving CAR T-cell therapy, emphasizing the burden of geographic barriers.

Financial and Logistical Burden

The Hidden Costs of Therapy

The cost of CAR T-cell therapy exceeds $400,000 per patient for the cellular product alone, excluding associated expenses such as travel, lodging, and caregiver support. Although assistance programs exist, they often fail to address the full spectrum of patient needs.

Time-sensitive logistics further complicate access. Delays in manufacturing, coordination, and patient readiness can result in missed treatment windows, particularly in patients with rapidly progressing disease. Although recent reductions in monitoring requirements, from 4 weeks to 2 weeks, represent progress, substantial barriers remain.

Global Inequities

A Wider Access Crisis

Globally, access to cellular therapies is even more restricted. Many regions face limited availability of immunotherapies, inadequate infrastructure for cell manufacturing, and inconsistent insurance or reimbursement systems. Even access to noncellular therapies remains variable across countries, reflecting broader disparities in cancer care delivery.

These challenges highlight that improving access is not solely a national issue, but a global priority.

Strategies for Improvement

Toward Equitable Access

Khaliq emphasized several actionable strategies to address these barriers. Standardizing Medicaid coverage across states, expanding the number of CAR T and TCE treatment centers, and improving reimbursement for caregiver support represent key steps toward reducing inequities.

Equally important is increasing awareness of available support programs among referring physicians and streamlining manufacturing and monitoring processes. Collectively, these efforts aim to align therapeutic innovation with equitable delivery.

Advances and Next-Generation CAR T Strategies in Multiple Myeloma

From Breakthrough to Optimization

CAR T-cell therapy has fundamentally transformed the treatment landscape of relapsed and refractory multiple myeloma, delivering response rates and survival outcomes previously considered unattainable. What began as a late-line salvage strategy has rapidly evolved into a central therapeutic pillar, with ongoing efforts focused on improving durability, safety, and earlier integration.

At the ICE-T Congress, Prerna Mewawalla, MD, director of stem cell transplant and cellular therapy and medical director of apheresis at Allegheny Health Network in Pittsburgh, Pennsylvania, presented a comprehensive update on CAR T-cell therapy in multiple myeloma, spanning pivotal trial data, emerging cellular platforms, and evolving strategies designed to overcome resistance and expand access.

Establishing the Standard

Cilta-cel and Ide-cel

Cilta-cel continues to define the upper boundary of efficacy in relapsed and refractory multiple myeloma. In the phase 1/2 CARTITUDE-1 study (NCT03548207), overall response rates (ORRs) reached 98%, with CRs in 82.5% of patients and a median progression-free survival (PFS) of 34.9 months. Long-term follow-up further reinforces durability, with a median overall survival (OS) of 60.4 months and approximately one-third of patients remaining progression-free at 5 years.

These results are complemented by randomized data from the phase 3 CARTITUDE-4 trial (NCT04181827), where cilta-cel significantly outperformed standard-of-care regimens in lenalidomide-refractory disease. The HR for PFS was 0.26, with a 30-month PFS of 59.4% vs 25.7% in the control arm. A corresponding OS benefit was also observed (HR, 0.55), with a 30-month OS of 76.4% vs 63.8%, supporting earlier integration of CAR T-cell therapy in the treatment continuum.

Ide-cel further supports this paradigm shift, as demonstrated in the phase 3 KarMMa-3 trial (NCT03651128). In patients with relapsed/refractory multiple myeloma, after 2 to 4 prior lines of therapy, ide-cel significantly improved PFS compared with standard regimens, with a median PFS of 13.3 months vs 4.4 months (HR, 0.49). ORRs were substantially higher with ide-cel (71.3% vs 41.7%), and deeper responses were reflected in higher CR rates. The median duration of response was also prolonged (14.8 vs 9.7 months), indicating more sustained disease control.

Although OS analysis in the intention-to-treat population was affected by crossover in approximately 56% of patients in the standard arm, the median OS favored ide-cel (41.4 vs 37.9 months), and adjusted analyses further supported a survival benefit (HR, 0.72). Collectively, these findings reinforce the movement of CAR T-cell therapy into earlier lines of treatment.

Emerging CAR T Platforms

Beyond First Generation

Next-generation CAR T platforms are now focused on improving both efficacy and safety profiles. Anitocabtagene autoleucel (anito-cel), evaluated in the phase 2 iMMagine-1 study (NCT05396885), represents a novel BCMA-directed CAR T construct designed to enhance durability while minimizing neurotoxicity. In a heavily pretreated population (n = 117), anito-cel achieved an ORR of 97%, with deep responses including a very good partial response or better in 85% and CR rates of 68% at a median follow-up of 12.6 months. Responses were rapid, with a median time to first response of approximately 1 month. The safety profile was favorable and predictable, with CRS occurring in 85% of patients, predominantly grade 1 and 2, with a median onset of 4 days and short duration. Importantly, ICANS was infrequent (8%) and limited to low-grade events, with no delayed neurotoxicity signals observed. These data suggest that anito-cel may offer a more controlled safety profile while maintaining high efficacy, supporting its potential role as a next-generation BCMA CAR T platform.

Arlocabtagene autoleucel (arlo-cel), targeting GPRC5D, represents an alternative strategy designed to overcome antigen escape following BCMA-directed therapy. In early-phase data, arlo-cel demonstrated high activity in a heavily pretreated population, with ORRs approaching 94% and CR in approximately 71%, alongside measurable residual disease (MRD) negativity of 56.3% at 10^–5 sensitivity. Responses were rapid, with a median time to response of 0.99 months, and continued to deepen over time. The median PFS reached 24.3 months (95% CI, 12.5-not reached), supporting durable disease control. The safety profile was favorable and predictable, with CRS occurring in 84% of patients, limited to grade 1 and 2 events. Hematologic toxicities were common, including neutropenia in approximately 80% of patients, whereas infections occurred in 55% without high-grade events. On-target, off-tumor toxicities, including dysgeusia, nail changes, and skin adverse effects, were observed in approximately one-third of patients. Neurotoxicity was uncommon, with only isolated grade 2 events and no cases of ICANS or treatment-related mortality reported.

These developments highlight a shift toward antigen diversification as a means of sustaining long-term disease control.

Bridging Strategies

Controlling Disease in Real Time

Bridging therapy has emerged as a critical component of CAR T delivery, addressing disease progression during the manufacturing interval. Talquetamab, a GPRC5D-targeted bispecific antibody, has demonstrated particular utility in this setting. In a multicenter cohort of 77 patients, talquetamab bridging achieved a 62% ORR, with most patients receiving biweekly dosing over a median of 22 days. Toxicities were manageable, with no grade 3 or higher CRS, low rates of grade 3 ICANS (2.5%), and predominantly low-grade mucocutaneous adverse events (skin, 41%; oral, 56%; nail, 25%).

Importantly, 95% of patients successfully proceeded to CAR T infusion, with postinfusion response rates reaching 97.5% at day 30. CAR T–related toxicities remained acceptable following bridging (any grade CRS, 65%; grade ≥3 CRS, 3%; ICANS, 8.6%), with no treatment-related mortality. Alternative bridging strategies, including selinexor-based regimens, have also demonstrated feasibility, achieving response rates of 30% to 50% without compromising CAR T manufacturing or T-cell fitness.

Next-Generation Directions

Expanding the Horizon

The field is rapidly evolving beyond single-target autologous therapies. Dual-target CAR constructs combining BCMA and GPRC5D aim to prevent antigen escape and enhance durability. Allogeneic off-the-shelf CAR T-cell products and natural killer cell–based platforms are being developed to address manufacturing delays and improve accessibility.

Simultaneously, CAR T-cell therapy is moving earlier in the treatment paradigm, with ongoing studies evaluating its role in second-line and frontline settings. The expansion of outpatient CAR T programs and the removal of risk evaluation and mitigation strategy (REMS) requirements for approved products further signal a shift toward broader, more accessible implementation.

Nursing and Its Impact on Cellular Therapy Treatments

When Standard Tools Fall Short

As CAR T-cell therapy continues to expand, the safe delivery of treatment depends not only on efficacy but also on the ability to accurately monitor toxicity. ICANS remains a key complication, typically assessed using the Immune Effector Cell–Associated Encephalopathy (ICE) score. However, this standardized tool assumes a baseline level of communication and cognitive function that may not apply to all patients.

At the ICE-T Congress, Karley Bogina, MSN, RN, TCTCN, nurse manager of the hematology and cellular therapies unit at the University of Kansas Health System, presented a practical and clinically relevant challenge: how to safely monitor ICANS in patients who are nonverbal or developmentally delayed. Her work highlights a gap in current clinical practice and proposes a nursing-driven solution to ensure equitable access to cellular therapies.

A Clinical Challenge

The Limits of ICE Scoring

The need for innovation emerged from a real-world case of a young adult patient with relapsed B-cell acute lymphoblastic leukemia (B-ALL) who was nonverbal and developmentally delayed. Although the patient was eligible for CAR T-cell therapy, standard ICE scoring could not be applied, as it relies on orientation, naming, writing, and counting tasks.

Alternative tools, including the Cornell Assessment of Pediatric Delirium, were also inadequate due to similar limitations in assessing communication and awareness. This scenario exposed a critical barrier: Without a reliable method to monitor neurotoxicity, safe CAR T delivery becomes uncertain.

Developing a Solution

The Modified ICE Tool

To address this gap, an interdisciplinary team developed a modified ICE (m-ICE) scoring system tailored to nonverbal and developmentally delayed adult patients. The tool preserved the 10-point structure of the traditional ICE score to maintain compatibility with institutional grading and management algorithms for ICANS.

Implementation was supported through practical strategies, including laminated visual aids and integration into the electronic medical record via standardized documentation templates. Nursing staff were trained through structured, unit-based education, ensuring consistent adoption across care teams.

Feasibility in Practice

A Nursing-Led Implementation

The m-ICE tool was rapidly integrated into clinical workflow and demonstrated strong feasibility. Nursing staff reported that the tool was intuitive, efficient, and comparable in effort to standard ICE scoring. Importantly, it allowed for consistent neurologic assessment without increasing clinical burden.

Minor challenges were noted, primarily related to patient cooperation, but these did not significantly limit usability. Overall, the tool proved adaptable to real-world bedside conditions.

Clinical Application

Monitoring ICANS in a Complex Patient

The tool was prospectively applied throughout the patient’s CAR T treatment course, with assessments performed every 6 hours from infusion through discharge. At baseline, the patient achieved a full score (10/10), establishing an individualized reference point.

During treatment, the patient developed grade 1 ICANS and grade 2 CRS. The m-ICE tool successfully captured subtle neurologic fluctuations, including changes in activity, restlessness, and ability to perform individualized tasks such as identifying time or writing. These findings allowed for continuous monitoring and appropriate clinical management.

Nursing Leadership Perspectives for Cellular Therapy Programs

Beyond Therapy—The System Behind Success

The rapid expansion of cellular therapies has transformed hematologic oncology, but successful implementation extends far beyond the infusion itself. Delivering CAR T-cell therapy safely and efficiently requires a coordinated system of care, supported by robust infrastructure, interdisciplinary collaboration, and continuous education.

At the ICE-T Congress, Stephanie Nigro, MSN, RN, CHPN, OCN, TCTCN, CPHQ, assistant director of nursing in the Hematological Malignancies and Cellular Therapeutics Division at the University of Kansas Cancer Center, presented a leadership-focused perspective on cellular therapy programs, emphasizing that clinical success is inseparable from operational excellence.

Interdisciplinary Coordination

A Foundation for Care Delivery

Effective cellular therapy programs depend on seamless collaboration across multiple teams, including inpatient and outpatient services, apheresis units, pain management, and financial coordination. These programs require highly synchronized workflows, particularly around patient admission, treatment timing, and discharge planning.

Regular multidisciplinary meetings and clear communication pathways are essential to align stakeholders and ensure continuity of care. Without this coordination, even the most advanced therapies risk delays and inefficiencies that can compromise patient outcomes.

Infrastructure Matters

Supporting Complexity at Scale

Cellular therapies introduce unique logistical and clinical demands that require dedicated infrastructure. Hematology and cellular therapy programs must be designed to support high-acuity patients, complex monitoring protocols, and rapid clinical decision-making.

This includes not only physical resources but also standardized processes that enable scalability as program volume grows. Building this infrastructure is critical to transitioning CAR T-cell therapy from specialized centers to broader clinical adoption.

Education as a Continuous Process

Education emerged as a central pillar of successful program development. Given the evolving nature of cellular therapies, ongoing training is required not only for nursing staff but also for ancillary and newly onboarded team members.

Understanding the rationale behind workflows and clinical decisions is particularly important in managing change and ensuring adherence to protocols. Continuous education fosters confidence, improves patient safety, and supports consistent care delivery across teams.

Beyond Efficacy: Redefining Access and Delivery of CAR T-Cell Therapy

Multiple Myeloma and Lymphoma

CAR T-cell therapy has redefined therapeutic expectations across hematologic malignancies, including multiple myeloma and lymphoma. However, despite its remarkable efficacy, its real-world impact has been limited by barriers that extend beyond biology, encompassing regulatory constraints, logistical complexity, and health system fragmentation.

During the ICE-T Congress, M. Bilal Abid, MD, MS, FACP, MRCP, FRCP, associate professor of medicine in hematology and medical oncology at Texas Tech University Health Sciences Center in Lubbock, as well as medical director of the Bone Marrow Transplantation and Cellular Therapy Program and associate director of the Hematology/Oncology Fellowship Program, presented a comprehensive overview of the evolving CAR T landscape following the removal of REMS requirements. His central message was clear: The next phase of CAR T innovation lies not in improving efficacy alone, but in ensuring equitable and scalable access.

The CAR T Journey

A Complex and Fragile Pathway

Unlike conventional therapies, CAR T treatment involves a multistep, highly coordinated process that includes referral to a specialized center, insurance authorization, leukapheresis, manufacturing, lymphodepletion, and eventual infusion. Each step introduces potential delays and points of failure.

Manufacturing remains a major bottleneck. Autologous CAR T production is inherently patient-specific, requiring complex logistics and cold-chain handling. Vein-to-vein timelines can be prolonged, and approximately 10% to 20% of products may fail due to manufacturing or logistical issues. Furthermore, the higher cost of CAR T, which can reach more than $500,000 per treatment, also complicates access to the therapy. These delays are clinically meaningful, particularly for patients with aggressive or rapidly progressing disease who may not be able to wait.

Clinical and Toxicity Challenges

Beyond the Infusion

CAR T-cell therapy also carries significant clinical complexity. CRS and ICANS remain common, requiring prompt recognition and management by experienced teams. Although these toxicities are often less severe in multiple myeloma compared with lymphoma or leukemia, they still necessitate infrastructure capable of inpatient monitoring and escalation of care.

Additional challenges include prolonged cytopenias, hypogammaglobulinemia, B-cell aplasia, and increased infection risk, all of which extend the care continuum well beyond the initial infusion. Survivorship care, therefore, becomes an integral component of CAR T programs, further increasing resource requirements.

Access Barriers

Geography, Infrastructure, and Timing

CAR T delivery has traditionally been restricted to specialized Foundation for the Accreditation of Cellular Therapy (FACT)-accredited academic centers. This centralization creates geographic disparities, as many patients live far from treatment sites or are referred too late in their disease course.

Operational constraints, including limited treatment slots, bed availability, and manufacturing backlogs, further delay access. Even when patients are eligible, these barriers can result in missed therapeutic windows, particularly in those with rapidly evolving disease.

CAR T vs Bispecific Antibodies

A Practical Comparison

Abid highlighted a critical distinction between CAR T-cell therapy and bispecific antibodies in terms of accessibility and scalability. CAR T represents a “long-pathway” treatment, requiring multiple coordinated steps before delivery, including access to a FACT-accredited center, insurance approval, and secure apheresis slots. In contrast, bispecific antibodies are readily available, off-the-shelf therapies that can often be initiated within 1 to 2 weeks.

This difference has important clinical implications. Although CAR T may offer deeper and more durable responses, bispecific antibodies provide a practical alternative for patients requiring immediate treatment or those unable to access specialized centers. Increasingly, these modalities are viewed as complementary rather than competitive.

Models of Care

Redefining Delivery Systems

The removal of REMS requirements has accelerated the development of new care delivery models. Three primary frameworks are emerging as follows:

• Academic center model: Full treatment and monitoring at specialized centers
• Integrated/hybrid model: Shared care between academic and community sites
• Community-focused model: Independent delivery within community settings

The shift toward hybrid and community-based models reflects an effort to bring therapy closer to patients, reduce travel burden, and improve continuity of care. However, these models require careful coordination to maintain safety and quality.

Building Community Programs

Infrastructure and Readiness

Expanding CAR T into community settings requires significant investment in infrastructure and training. Facilities must develop capabilities for both inpatient and outpatient care, establish clear communication pathways with hospital systems, and ensure appropriate staffing and credentialing.

Education is a cornerstone of this transition. Providers, nurses, pharmacists, and ancillary staff must be trained to recognize and manage CAR T–related toxicities. Structured in-service programs and standardized protocols are essential to ensure consistent care delivery.

Workflow Optimization

From Onboarding to Follow-Up

Efficient CAR T and bispecific therapy programs rely on well-defined workflows. These include patient onboarding processes, scheduling strategies, preauthorization pathways, and coordination with pharmacy and inpatient teams.

Standardized protocols for toxicity monitoring, often integrated into electronic medical records, enable real-time decision-making and improve patient safety. Clear delineation of roles among physicians, advanced practice providers, and nursing staff is critical to maintaining workflow efficiency.

REMS Removal

Opportunity and Responsibility

The elimination of REMS requirements represents a pivotal shift in CAR T delivery. By reducing regulatory burden, it allows for expansion into community settings and increases the number of potential treatment centers.

However, this decentralization introduces new challenges. Ensuring consistent quality of care, maintaining safety standards, and supporting less experienced centers are essential to prevent variability in outcomes. The success of this transition will depend on balancing accessibility with clinical rigor.

Operational Pathways

Standardizing Care in the Community

A key component of expanding CAR T-cell and bispecific antibody therapies into community settings is the development of standardized operational workflows. Abid highlighted structured treatment pathways that enable safe delivery outside traditional academic centers, particularly for bispecific antibodies.

Bispecific antibody therapy typically follows a step-up dosing model, with initial doses administered in a monitored setting followed by rapid transition to outpatient care. Patients are admitted for the first and second ramp-up doses, with early discharge and close follow-up, including same-day or next-day assessments and structured phone monitoring. By the third week, treatment is often delivered entirely in the outpatient setting, reflecting a shift toward less resource-intensive care models.

Equally important is the implementation of clearly defined onboarding workflows. These include coordinated scheduling, insurance authorization, pharmacy preparation, and patient education prior to treatment initiation. Responsibilities are distributed across multidisciplinary teams, with nursing staff managing logistics and patient preparation, advanced practice providers overseeing clinical monitoring and toxicity assessment, and physicians guiding treatment decisions and escalation pathways.

Standardized toxicity management protocols are central to this model. Flow-based algorithms for fever and CRS management allow rapid triage and escalation, beginning with symptomatic management and progressing to corticosteroids or hospital evaluation when indicated. Drug-specific protocols further guide supportive care, including antimicrobial prophylaxis and fluid management during treatment.

Together, these structured workflows demonstrate that with appropriate planning, education, and protocolization, complex cellular therapies can be delivered safely and efficiently beyond academic centers.

The Future

Toward Equitable Cellular Therapy

Abid emphasized that the ultimate goal is equitable access to cellular therapies. This includes not only expanding the number of treatment centers but also addressing systemic barriers such as insurance approval, education gaps, and infrastructure limitations.

Emerging approaches, including hybrid care models, decentralized clinical trials, and integration of digital tools such as artificial intelligence (AI), may further streamline care delivery and accelerate adoption. These innovations have the potential to reshape the cellular therapy landscape on a global scale.

Integrated Microfluidic Systems for the Comprehensive Analysis of Liquid Biopsy Samples

Moving Beyond Bone Marrow

The management of hematologic malignancies, particularly multiple myeloma, has traditionally relied on invasive bone marrow sampling and imaging-based assessments. However, these approaches are limited by sampling bias, procedural burden, and inability to support frequent monitoring.

During the ICE-T Congress, Steven Soper, PhD, Foundation Distinguished Professor at the University of Kansas, and principal investigator and director of the Center of BioModular Multi-Scale Systems for Precision Medicine, presented an innovative paradigm centered on liquid biopsy, highlighting how microfluidic technologies can enable sensitive, rapid, and minimally invasive disease monitoring across the cancer care continuum.

The Biology Challenge

Rare Signals in Complex Samples

Liquid biopsy in hematologic malignancies faces a fundamental technical challenge: detecting extremely rare tumor-derived components within a vast background of normal cells. Circulating tumor or plasma cells may be present at levels as low as 1 to 100 cells/mL of blood, whereas cell-free DNA fragments carrying tumor mutations can be exceedingly scarce.

Additionally, no universal marker exists to reliably distinguish all malignant circulating cells from normal hematopoietic cells, and many target populations overlap in size and phenotype with normal cells. These constraints demand technologies with exceptionally high sensitivity, specificity, and dynamic range.

Microfluidics

A Platform for Precision

Microfluidic systems offer a solution by enabling controlled processing of small-volume clinical samples within highly engineered fluidic channels. These platforms integrate multiple analytical steps, including cell capture, enrichment, molecular analysis, and imaging into a single automated system.

Key advantages include reduced processing time, minimized contamination risk, lower reagent consumption, and the potential for point-of-care deployment. Importantly, these systems support reproducible and scalable testing, addressing critical limitations of conventional laboratory workflows.

Integrated Systems

From Capture to Characterization

Soper highlighted a fully automated microfluidic liquid biopsy platform at the University of Kansas Medical Center designed for analysis of circulating tumor cells, exosomes, and cell-free DNA. Using a catch-and-release approach, the system can process 2 mL of whole blood in just 20 minutes, making it markedly faster than conventional workflows.

The platform also demonstrated high analytical sensitivity, with reported detection performance reaching 1 target cell per 20,000 background cells and molecular sensitivity approaching 1 in 10⁷, underscoring its potential for rare-event detection in hematologic malignancies.

Applications in Multiple Myeloma

Toward Real-Time Monitoring

Soper further demonstrated that circulating plasma cells (CPCs) can serve as a dynamic biomarker across the myeloma disease spectrum. CPC burden increased progressively from monoclonal gammopathy of undetermined significance to smoldering myeloma and active disease, with significant differences observed between disease stages, supporting their role in risk stratification and disease monitoring. Importantly, treated patients demonstrated reduced CPC levels, highlighting their potential utility in assessing therapeutic response.

Beyond enumeration, CPCs also enable real-time assessment of treatment efficacy. Using CD138-based capture and microfluidic enrichment, CPCs can be isolated from peripheral blood and analyzed at the single-cell level. This allows longitudinal monitoring without reliance on bone marrow biopsy, offering a minimally invasive approach to track disease burden and response.

Circulating Plasma Cells

A Prognostic and Monitoring Tool

The integration of microfluidic liquid biopsy platforms has the potential to transform clinical practice. By enabling frequent, minimally invasive testing, these technologies support real-time assessment of treatment response, early detection of relapse, and monitoring of minimal residual disease.

Additionally, the scalability and cost-efficiency of microfluidic systems open the possibility of decentralized testing, extending advanced diagnostics beyond specialized centers and into broader clinical settings.

Integrated Microfluidic Systems

From Detection to Molecular Profiling

A key innovation presented was the development of fully integrated, modular microfluidic systems capable of performing multiple analytical steps within a single platform. These systems incorporate sequential modules for cell selection, viability assessment, immunophenotyping, and cytogenetic analysis, enabling comprehensive characterization of circulating tumor cells.

This lab-on-chip approach allows flexible configuration depending on clinical needs, reduces fabrication complexity, and supports high-throughput, scalable testing. The integration of electronic and fluidic components further enhances automation and reproducibility, positioning these platforms for real-world clinical implementation.

Microfluidic FISH

Rapid and Cost-Effective Cytogenetics

Microfluidic-based fluorescence in situ hybridization (FISH) represents a major advancement in cytogenetic testing. Compared with conventional FISH, which typically requires up to 2 days, microfluidic platforms can deliver results within approximately 4 hours, while using only one-hundredth of the reagent volume.

This approach enables simultaneous immunophenotyping and cytogenetic analysis on the same enriched circulating cells, eliminating the need for bone marrow sampling. Clinically relevant abnormalities, such as chromosome 13 deletions, associated with poor prognosis in multiple myeloma, can be detected directly from peripheral blood, supporting both prognostication and disease monitoring.

Toward Clinical Translation

Scalable Precision Diagnostics

The platform’s modular design and compatibility with injection-molded plastics enable large-scale production and cost reduction. Applications extend across hematologic malignancies, including acute myeloid leukemia (AML) and B-ALL, where disease-specific markers (eg, CD19, CD138) can be used for targeted cell selection and molecular profiling.

These advances support a shift toward distributed, minimally invasive diagnostics, where high-sensitivity liquid biopsy testing can be performed rapidly, repeatedly, and potentially at the point of care.

Advances in Bispecific T-Cell Engager Therapies for Multiple Myeloma

A Class-Level Transformation

Bispecific T-cell engagers have rapidly emerged as a central therapeutic strategy in relapsed/refractory multiple myeloma. Rather than focusing on a single agent, Jeries Kort, MD, assistant professor in the Division of Hematologic Malignancies and Cellular Therapeutics in the Department of Internal Medicine at the University of Kansas Medical Center, gave a presentation at the ICE-T Congress that framed the field as a class-wide evolution. Kort centered on 3 key therapies—teclistamab, elranatamab, and talquetamab—each contributing distinct efficacy, toxicity profiles, and strategic roles in treatment sequencing.

All 3 agents operate through a shared mechanism of T-cell redirection, binding CD3 on T cells and tumor-associated antigens such as BCMA or GPRC5D, leading to immune synapse formation and targeted cytotoxicity. The central question is no longer whether these therapies are effective, but how to optimize their use across patients and disease settings.

Comparative Efficacy

Consistency Across BCMA, Expansion Beyond

Four major bispecific antibodies now define the clinical landscape: teclistamab, elranatamab, talquetamab, and linvoseltamab. Across pivotal phase 2 trials, BCMA-directed agents demonstrate remarkably consistent efficacy. Teclistamab (the phase 2 MajesTEC-1 trial; NCT04557098) and elranatamab (the phase 2 MagnetisMM-3 trial; NCT04649359) showed similar response rates of 63% and 61%, respectively. Linvoseltamab (the phase 1/2 LINKER-MM1 trial [NCT03761108]) demonstrated a higher ORR of 71% and encouraging durability, with a 60.4% PFS at 18 months; the median PFS has not yet been reached. The median PFS varies across studies, ranging from 11.4 months with teclistamab to 17.2 months with elranatamab. These differences likely reflect variations in study populations and follow-up rather than true superiority of one agent over another.

Talquetamab expands the therapeutic paradigm beyond BCMA. In the phase 2 MonumenTAL-1 trial (NCT04634552), it achieved ORRs of 69% to 70% in heavily pretreated populations, including patients with high-risk cytogenetics and extramedullary disease. This reinforces a key concept from the presentation: Effective T-cell redirection is not antigen-restricted, and GPRC5D provides a critical alternative target, particularly in the setting of antigen escape.

Dosing strategies also differ meaningfully. Teclistamab is administered weekly, whereas elranatamab and linvoseltamab allow for de-escalation to every-2-week or every-4-week schedules, an important factor in long-term tolerability.

Toxicity

Early Control vs Long-Term Burden

Toxicity with bispecific antibodies follows a clear temporal pattern. CRS is common but predictable, occurring predominantly during cycle 1 and step-up dosing, and is largely low grade. This predictability has enabled standardized mitigation strategies.

One of the most impactful advances is prophylactic tocilizumab. In MajesTEC-1, administration prior to step-up dosing reduced CRS rates from over 70% to approximately 25% without compromising efficacy, supporting a shift toward proactive rather than reactive management.

Neurotoxicity (ie, ICANS) remains uncommon across bispecific therapies, especially compared with CAR T, supporting their broader applicability. In contrast, infection represents the dominant long-term toxicity. This is particularly pronounced with BCMA-targeted therapies due to depletion of plasma cells and B cells, resulting in hypogammaglobulinemia and impaired humoral immunity. These effects are compounded by T-cell dysfunction and cumulative immunosuppression from continuous therapy and supportive medications.

Clinical data show high rates of severe infections, with grade 3 to 4 events reaching up to approximately 60% in some cohorts. Importantly, IVIG supplementation significantly reduces the risk of serious infection, making infection prevention a central component of bispecific therapy rather than adjunctive care.

Dose Optimization

A Shift Toward Sustainable Therapy

Beyond pharmacologic prophylaxis, dosing strategy plays a critical role in toxicity management. In MajesTEC-1, patients transitioning from weekly to every-2-week teclistamab dosing experienced substantially lower rates of severe infections (15.6% vs 33.3%) without loss of efficacy.

This reflects a broader shift in treatment philosophy: from continuous maximal exposure to optimized, sustainable immune engagement that balances efficacy with long-term safety.

Combination Strategies

Depth vs Toxicity Trade-Off

Combination therapy represents a major area of expansion. Teclistamab-based combinations with daratumumab and daratumumab plus pomalidomide achieve high response rates of approximately 76.5% and 85.2%, respectively, with rapid and deep responses.

Elranatamab combinations similarly demonstrate strong efficacy, with elranatamab plus daratumumab achieving an ORR of 70.6% and favorable safety, including absence of high-grade CRS and ICANS in early data.

Talquetamab-based combinations further extend targeting strategies beyond BCMA. However, intensified immune activation introduces significant risk. Dual bispecific therapy (eg, teclistamab plus talquetamab) achieves high response rates but is associated with substantial infection burden, including high rates of grade 3 or higher infections and treatment-related mortality. This underscores one of the key insights of the presentation: Increasing efficacy through combination must be carefully balanced against cumulative toxicity.

Earlier-Line Integration

Redefining the Treatment Paradigm

Bispecific antibodies are now being evaluated earlier in the course of therapy. In newly diagnosed, transplant-ineligible patients, elranatamab combined with daratumumab and lenalidomide achieved an ORR of 97.3%, with rapid responses and high rates of deep remission.

In transplant-eligible settings, early-phase studies demonstrate exceptionally high rates of MRD negativity, suggesting the potential for profound disease control when these agents are used earlier. These findings indicate that bispecific therapy may transition from a late-line salvage strategy to an earlier, potentially time-limited intervention.

Next-Generation Designs

Beyond Bispecifics

Emerging platforms are extending beyond traditional bispecific constructs. Trispecific antibodies targeting BCMA, GPRC5D, and CD3 aim to overcome antigen escape and enhance efficacy. Additional approaches include alternative targets such as FcRH5 and engineered molecules designed to improve binding specificity and reduce cytokine release.

These innovations reflect a broader evolution of T-cell–redirecting therapy, where molecular design will play an increasingly important role in shaping clinical outcomes.

Building the Future of Cellular Therapy: Infrastructure, Integration, and Strategic Growth

Introduction

From Innovation to Implementation

As cellular therapies continue to transform oncology, the next phase of progress depends not only on therapeutic innovation but on the systems that support their delivery. Roy Jensen, MD, vice chancellor and director of the University of Kansas Cancer Center and Kansas Masonic Cancer Research Institute, and William R. Jewell, MD, Distinguished Kansas Masonic Professor, gave a presentation at the ICE-T Congress that shifted the focus from drugs to infrastructure, emphasizing that long-term success in cellular therapy requires coordinated environments that integrate clinical care, research, and operational efficiency.

The vision presented is clear: Building a future beyond cancer requires aligning physical space, clinical workflows, and scientific collaboration into a unified ecosystem.

Centralization of Care

Improving Quality and Efficiency

A major initiative involves the proposed relocation of hematologic malignancies and cellular therapeutics services into a centralized cancer facility. Currently, care delivery is fragmented across multiple locations, requiring patients and providers to move between sites for laboratory testing, imaging, and treatment.

Centralization aims to address these inefficiencies by consolidating services within a single location. This model is expected to improve patient experience by reducing delays, minimizing logistical burden, and enhancing continuity of care. Importantly, it also strengthens infection control by limiting patient movement, which is especially critical for immunocompromised populations undergoing cellular therapy.

Enhancing Collaboration

Bridging Clinicians and Scientists

One of the key limitations identified in prior institutional reviews was the physical separation between clinicians and physician-scientists. This separation restricts opportunities for real-time collaboration and slows the translation of scientific discoveries into clinical practice.

The proposed model directly addresses this gap by clinical and research teams. By embedding research laboratories within the same facility as patient care areas, the institution aims to foster continuous interaction between disciplines, accelerate translational research, and increase clinical trial enrollment.

Capacity and Growth

Preparing for Expanding Demand

Cellular therapy programs are experiencing rapid growth, and existing infrastructure is increasingly insufficient to meet demand. Current facilities face capacity limitations, particularly in bone marrow transplantation and advanced cellular therapies.

Expansion into a larger, purpose-built facility is therefore positioned as a strategic necessity. This growth is not only about accommodating more patients but also about enabling more complex therapies, increasing throughput, and supporting future program development.

Integrated Research Environment

Building a Cellular Therapy Core

A central component of the vision is the development of a dedicated Cellular Therapy Core. This includes expansion of clinical and good manufacturing practices–compliant research laboratories, alongside infrastructure to support advanced translational and early-phase research.

The goal is to create an environment where laboratory discovery and clinical application are seamlessly connected. This includes recruiting funded investigators in tumor immunology and providing an embedded research space for clinical teams, enabling real-time integration of research into patient care.

Multidisciplinary Care

Strengthening the Patient Journey

The proposed model emphasizes coordinated, multidisciplinary care involving physicians, nursing teams, research staff, and support services. Improved communication between teams is expected to enhance continuity of care, streamline decision-making, and support patient education.

Strategic Vision

Toward National Leadership

Ultimately, the initiative is positioned as part of a broader strategy to establish the institution as a national leader in immunotherapy and cellular therapy. This includes expanding clinical programs, increasing research output, and strengthening the connection between scientific innovation and patient care.

The emphasis is not solely on growth, but on building a sustainable and scalable model that can support the next generation of cellular therapies.

Beyond REMS: Building a Scalable and Equitable CAR T-Cell Delivery Model

The Access Gap in a Transformative Era

A Regulatory Opening and What It Requires

CAR T-cell therapy has produced survival outcomes in relapsed/refractory multiple myeloma and aggressive lymphomas that were unimaginable a decade ago. Those outcomes remain largely inaccessible to patients who do not live near a FACT-accredited academic center, whose disease does not allow the weeks required for manufacturing, or whose geography or support system cannot meet the demands of the referral pathway. The FDA’s removal of REMS requirements for approved CAR T-cell products has removed 1 structural barrier to broader delivery. Whether that opening translates into wider access depends on whether the clinical and operational infrastructure to meet it is actually built.

At the ICE-T Congress, M. Bilal Abid, MD, MS, FACP, MRCP, FRCP, associate professor of medicine in hematology and medical oncology at Texas Tech University Health Sciences Center in Lubbock, as well as medical director of the Bone Marrow Transplantation and Cellular Therapy Program and associate director of the Hematology/Oncology Fellowship Program, made that case directly. His presentation argued that the removal of REMS is necessary but not sufficient, and that academic programs with the expertise to deliver these therapies safely carry a specific obligation to help community practices develop the capacity to do the same.

Two Therapies, Two Very Different Access Profiles

Abid drew a clear distinction between CAR T-cell therapy and bispecific antibodies as access propositions. CAR T-cell therapy requires FACT accreditation, prior authorization, an apheresis slot, and a manufacturing window typically measured in weeks, all before a single dose is administered. Out-of-specification product failures from cell quality issues, cold-chain errors, or process failures affect an estimated 10% to 20% of manufacturing runs. For a patient with rapidly progressing disease or limited performance reserve, that timeline is frequently unavailable.

Bispecific antibodies are off the shelf. With authorization secured, treatment can begin within 1 to 2 weeks. The step-up dosing period, when CRS risk is highest, can be managed with inpatient monitoring at an academic center, after which subsequent cycles transfer to a community infusion setting. The hybrid model academic center for ramp-up and community practice for maintenance is operational in Abid’s program, and the workflows he presented were built on direct institutional experience.

What REMS Removal Changes, and What It Does Not

REMS programs restricted CAR T-cell delivery to a certified set of institutions. Their removal expands who is legally permitted to administer these therapies. It does not build the staff training, toxicity protocols, electronic medical record infrastructure, or multidisciplinary team coordination required for safe delivery. Abid presented data from his published work mapping the spectrum of delivery models from fully academic to fully community-based, with an integrated hybrid model in between. REMS removal makes the community and hybrid categories accessible. The field must decide whether to build toward them.

His position was unambiguous: Academic programs have an obligation to share what they have built. That means making protocols available to community partners, providing in-service education for nursing and pharmacy teams, and, in some cases, deploying trained personnel to support new programs through early operations. Concentration of expertise in academic centers, while patients in rural and underserved communities go untreated, is not a neutral outcome. It is a choice.

Onboarding Workflows and the Architecture of Safe Administration

Abid presented detailed institutional workflows for bispecific antibody administration tools designed not for academic specialists but for care teams operating in settings where that specialist backup may not be immediately available.

New patient onboarding assigns clear, nonoverlapping responsibilities across the nurse, advanced practice provider, and treating physician. The nurse loads the correct bispecific regimen, schedules ramp-up doses to begin on Monday or Tuesday mornings so that the highest-risk 24 to 48 hours fall within standard inpatient coverage, preauthorizes the admission, confirms the pharmacy supply, and sends the patient home with educational materials and a thermometer. The advanced practice provider confirms bed availability before the patient leaves the clinic, briefs the inpatient team on administration timing, places admission orders with explicit vital sign and neurologic-check frequencies, and formally documents CRS and ICANS assessments on each follow-up day. The treating physician establishes the treatment commitment with the patient and family, communicates escalation preferences to the nurse, and maintains direct contact with the emergency department for after-hours concerns.

The initiation schedule follows a 3-week structure: first ramp-up dose with overnight admission in week 1, second ramp-up dose with the same monitoring in week 2, office-based full-dose treatment in week 3. Drug-specific decision trees at each step govern prophylactic antimicrobials, hydration, and fever management without requiring real-time physician input for each decision. The fever algorithm separates home-manageable grade 1 CRS acetaminophen, hourly checks, and a defined temperature threshold for emergency evaluation from higher-grade events requiring in-person assessment, with an explicit weekend pathway that does not depend on specialist availability.

Managing CRS and ICANS Outside the Academic Safety Net

Abid reviewed the 2024 consensus recommendations on CRS management for CD3 × CD20 bispecific antibodies from Crombie et al, published in Blood, which stratify management from home monitoring for grade 1 through intensive care unit–level care for grade 4. The practical implication for community delivery is significant: Grade 1 CRS does not require hospitalization if the patient has reliable caregiving support, monitoring equipment, and a clear escalation threshold in place. Community programs do not need to admit every patient with a fever, but they do need a protocol for every fever, and they need it before the first patient is treated.

ICANS monitoring used ICE score flowsheets embedded in Epic, allowing nursing staff to grade neurotoxicity and activate standardized responses within the electronic medical record without a physician order. These tools move decision support into the workflow rather than concentrating it on a clinician who may not be immediately reachable, which is ultimately the only model that scales.

Clinical Trials, Accrual, and the Evidence Gap

Abid closed with an argument about clinical trial enrollment that extends beyond the access question. When immune effector cell therapy trials enroll predominantly from academic centers, the efficacy and safety data they produce reflect patients in academic centers: On average, they are younger, fitter, more racially homogeneous, and more resourced than the populations who will receive these drugs once they enter routine practice. The disparity in trial accrual compounds the disparity in access and produces evidence whose generalizability to the broader patient population is uncertain.

Decentralized trial designs with hybrid and remote components, AI-assisted patient identification and matching, and deliberate community partnership programs address this not merely as a matter of equity but as a scientific necessity. The evidence base for cellular therapy is only as strong as the populations in which it was generated.

Gene Therapy in Sickle Cell Disease

Sickle Cell Disease

From Molecular Defect to Curative Therapy

Rajat Bansal, MD, assistant professor of medicine at the University of Kansas Cancer Center, delivered a presentation at the ICE-T Congress on gene therapy in sickle cell disease. Bansal’s clinical focus centers on the prevention and management of infectious complications after transplant, and that expertise shaped a presentation that moved from the molecular basis of the disease through the 2 FDA-approved gene therapy platforms to the practical and ethical realities that determine which patients receive them.

Pathophysiology and Disease Burden

Sickle cell disease results from a single-nucleotide substitution in HBB, producing the βS mutation (Glu6Val) that causes deoxygenated hemoglobin S (HbS) to polymerize. Polymerization deforms the erythrocyte, damages the cell membrane, and generates the rigid, irreversibly sickled cells responsible for vaso-occlusion. The severity of disease varies across genotypes HbSS, HbSC, HbSβ°-thalassemia, and HbSβ+-thalassemia; it is modulated by coinherited factors, most importantly fetal hemoglobin (HbF) levels. HbF and HbA both inhibit polymerization; HbS concentration, hypoxia, low pH, and cellular dehydration amplify it.

The downstream consequences span every organ system. Life expectancy in sickle cell disease averages 54 years, compared with 76 years in matched non–sickle cell disease cohorts. Quality-adjusted life expectancy is 33 years vs 67 years in matched controls. The socioeconomic toll compounds the clinical one: Lifetime income averages $1.23 million for patients with sickle cell disease vs $1.92 million for matched counterparts, reflecting both shorter lives and reduced annual earnings.

Lentiviral Gene Addition

Lovotibeglogene Autotemcel

Lovotibeglogene autotemcel (lovo-cel; Lyfgenia) is a lentiviral gene-addition therapy for sickle cell disease that uses autologous hematopoietic stem and progenitor cells transduced with the BB305 lentiviral vector encoding β^A-T87Q-globin. The resulting HbA^T87Q has antisickling properties and inhibits HbS polymerization under hypoxic conditions. In the phase 1/2 HGB-206 study (NCT02140554), the median total hemoglobin increased from 8.5 g/dL at baseline to at least 11 g/dL from 6 through 36 months after infusion. HbA^T87Q accounted for at least 40% of total hemoglobin and was distributed across a mean of 85% of circulating red blood cells, accompanied by reductions in markers of hemolysis. These findings demonstrated sustained production of antisickling hemoglobin and durable biochemical correction following lentiviral gene addition.

CRISPR-Based HbF Induction

Exagamglogene Autotemcel

The second approved approach, exagamglogene autotemcel (exa-cel), targets the developmental switch that suppresses fetal hemoglobin after birth. During fetal development, hemoglobin is predominantly HbF, composed of 2 α-globin and 2 γ-globin chains; HbF potently inhibits HbS polymerization. After birth, the transcription factor BCL11A progressively silences γ-globin production, driving the switch to adult β-globin and eliminating HbF’s protective effect. Exa-cel uses CRISPR/Cas9 to disrupt BCL11A, reversing this switch and restoring γ-globin expression. In clinical data, mean HbF comprised approximately 40% of total hemoglobin after treatment, with more than 90% of red blood cells expressing fetal hemoglobin, a proportion sufficient to substantially limit sickling across the erythrocyte population.

The Practical Realities of Delivery

Bansal drew extensively on institutional experience at the University of Kansas to describe the practical demands of gene therapy delivery, beginning with a point that distinguishes sickle cell disease from other conditions treated with autologous cellular therapies: The hematopoietic stem cell (HSC) compartment itself is damaged. Chronic hypoxia, high cell turnover, recurrent inflammation, and bone marrow microinfarction collectively compromise the HSC niche over years of disease. That damage makes collection difficult before the process even begins.

Patient evaluation before gene therapy is comprehensive, encompassing cardiac MRI for iron overload, MRI of the brain and vasculature for prior stroke and vasculopathy, pulmonary function testing, liver MRI, bone marrow biopsy with cytogenetics and next-generation sequencing for clonal hematopoiesis of indeterminate potential (CHIP) mutations, and dental clearance. Hydroxyurea is discontinued, and monthly exchange transfusions are initiated 2 to 3 months before mobilization, with the goal of reducing HbS to below 30% to minimize the risk of complications during apheresis.

Granulocyte colony-stimulating factor, the standard mobilizing agent in most autologous transplant settings, is contraindicated in sickle cell disease. It induces rapid hyperleukocytosis and neutrophil activation that can precipitate severe vaso-occlusive crises. Plerixafor, a CXCR4 antagonist, is used instead and is administered over 3 to 4 consecutive collection days per cycle, with cycles separated by 4 to 6 weeks. Even with this approach, collection frequently fails on the first attempt. Five of the first 7 patients treated at the University of Kansas required multiple collection cycles. Across clinical trials, 5% to 10% of patients failed manufacturing entirely; that is a failure rate with no analog in most other cellular therapy programs and one that reflects how profoundly the disease compromises the very cells being collected.

Following successful collection, CD34+ cells undergo isolation, either lentiviral transduction or CRISPR editing, and cryopreservation. Fertility preservation is completed before myeloablative busulfan conditioning begins. Monthly exchange transfusions continue through the manufacturing period, which typically spans several months. After conditioning and infusion of modified HSCs, patients remain hospitalized until engraftment, followed by long-term monitoring for late complications.

Secondary Malignancy

Parsing Risk and Attribution

Two patients in the lovotibeglogene autotemcel clinical trial developed AML, a finding that generated considerable concern and warranted careful analysis. The first case arose approximately 5.5 years after gene therapy. Neoplastic cells contained a lentiviral vector, but integration site analysis, RNA sequencing, and whole-genome sequencing found no evidence of insertional mutagenesis: The vector had integrated into VAMP4, a noncoding intronic region without transcriptional activation, and at a nononcogenic common integration site. Classical AML drivers—monosomy 7, a RUNX1 frameshift, PTPN11 mutation, and partial WT1 loss—were identified in the blasts. The second case developed approximately 3.5 years after infusion; neoplastic cells lacked lentiviral vector integration entirely. This patient’s busulfan area under the curve had fallen below the protocol target, raising the possibility that incompletely ablated HSCs with preexisting CHIP mutations had persisted and undergone malignant evolution under the selective pressure of reconstituting hematopoiesis.

Neither case provided evidence of vector-driven oncogenesis. What they did highlight is the baseline malignancy risk inherent to sickle cell disease itself. California registry data from 6423 patients with sickle cell disease followed over 27 years demonstrated a 3- to 4-fold increased risk of AML compared with the general population, most pronounced in women, individuals aged 15 to 39 years, and patients with severe disease. That background risk complicates attribution of any posttreatment leukemia to the gene therapy rather than to the underlying hematologic landscape.

Infection Risk

A Compounded Vulnerability

Patients with sickle cell disease enter gene therapy with a baseline infectious risk that most autologous transplant candidates do not carry. Functional asplenia, resulting from repeated splenic infarction, impairs opsonization and leaves patients particularly vulnerable to encapsulated organisms. Gene therapy compounds this vulnerability through 2 mechanisms: CD34+ selection, the first step of the manufacturing process, produces profound lymphodepletion by removing the bulk of lymphocytes from the collected product; and the myeloablative busulfan conditioning regimen eliminates residual immune effectors before the modified HSCs are infused. Data from autologous transplant cohorts demonstrate significantly slower immune reconstitution after CD34+-selected vs unselected autografts, a dynamic with direct relevance to gene therapy recipients who are, in effect, receiving a CD34+-selected product.

Current best practices from the American Society for Transplantation and Cellular Therapy recommend antifungal prophylaxis for 3 months post infusion and coverage against Pneumocystis jirovecii, viral pathogens, and encapsulated bacteria for 1 year. Guidelines are still evolving, and real-world data on immune reconstitution kinetics after gene therapy—specifically, the time to recovery of CD4+ counts, B-cell function, and immunoglobulin levels—remain limited. Bansal identified this as an area where prospective data collection is urgently needed.

Access, Cost, and the Allocation Problem

Approximately 15,000 individuals in the US are estimated to be eligible for gene therapy for sickle cell disease. The 2 approved manufacturers (Vertex Pharmaceuticals, exagamglogene autotemcel; bluebird bio, lovotibeglogene autotemcel) together have an annual production capacity of 150 to 200 patients. At that rate, clearing the eligible population would take roughly 100 years. Treatment costs of $2 million to $3 million per patient, the logistical complexity of delivery, and the small number of centers with the clinical expertise and financial infrastructure to administer gene therapy further narrow practical access. The result is a therapy that is transformative in clinical trials and functionally unavailable to the overwhelming majority of patients who need it.

Bansal argued that this disparity demands an explicit ethical framework for patient prioritization, one that does not emerge spontaneously from market forces or center-level discretion. He proposed looking to solid organ transplant allocation as a model: Prioritize patients with the greatest disease severity who retain sufficient organ function to undergo conditioning, and those without alternative curative options, such as a matched allogeneic donor. Lottery-based allocation was noted as a possible supplement while manufacturing capacity remains critically limited.

Allogeneic Transplant Remains Relevant

Gene therapy has not rendered allogeneic HSC transplantation obsolete. Contemporary transplant platforms continue to improve, with current data showing 93% donor chimerism, severe acute graft-vs-host disease (GVHD) in approximately 5% of patients, severe chronic GVHD in approximately 15%, and 97% of engrafted patients discontinuing immunosuppression within 1 year. Among engrafted patients, vaso-occlusive crises requiring opioid treatment were rare. For patients with a matched donor or for whom gene therapy access is not feasible given manufacturing constraints, cost, or center availability, allogeneic transplant offers durable disease control through a pathway that is already operational and whose long-term safety profile is well characterized.

Bispecific Antibodies and Chemoimmunotherapy in Relapsed Diffuse Large B-Cell Lymphoma

Choosing Between Agents, and Knowing When to Add Chemotherapy

Marc S. Hoffmann, MD, director of the lymphoma program and medical director for quality improvement at the University of Kansas Cancer Center, delivered a presentation at the ICE-T Congress around a practical framework for clinicians managing relapsed or refractory diffuse large B-cell lymphoma (DLBCL). He discussed how glofitamab and epcoritamab compare in day-to-day use, when combining either agent with chemotherapy is justified by the evidence, and how sequencing decisions shift depending on the individual patient’s history, fitness, and access to follow-on therapies.

Throughout the session, Hoffmann returned to a consistent principle: Response rates alone do not determine the right treatment. Dosing schedule, route of administration, steroid requirements, treatment duration, proximity to a treating center, and the trajectory of prior therapy all shape which option is actually deliverable, and deliverability is inseparable from benefit.

Glofitamab and Epcoritamab

Equal Efficacy, Different Logistics

Across pivotal studies, glofitamab and epcoritamab demonstrated comparable antitumor activity in relapsed/refractory DLBCL. ORRs ranged from approximately 52% to 63% and CR rates from 35% to 39% in heavily pretreated populations, including patients with prior CAR T-cell therapy. CRS occurred in approximately half of patients receiving either agent and was predominantly grade 1 or 2; grade 3 or higher events were uncommon. Rates of ICANS were similarly low across both drugs.

Where the agents diverge is in the operational details that govern how they are used. Glofitamab is administered intravenously following a single obinutuzumab pretreatment dose, uses weekly step-up dosing over 3 weeks before escalating to every-3-week full dosing, and is given for a fixed 12 cycles. Epcoritamab is administered subcutaneously, requires dexamethasone prophylaxis for the first 4 days after each injection in cycle 1, and continues indefinitely until progression. As Hoffmann summarized, the patients, the efficacy, and the toxicity are essentially the same. The decision between agents turns on route, steroid requirements, dosing cadence, and whether a patient’s circumstances favor a time-limited or continuous treatment course.

Rapid Step-Up Dosing

Compressing the Ramp-Up Window

For patients with bulky or rapidly progressive disease who cannot afford the 2 to 3 weeks a standard step-up protocol requires to reach full therapeutic dose, Hoffmann advocated for rapid escalation. For epcoritamab, the rapid protocol delivers step-up dose 1 on day 1, step-up dose 2 on day 3, and the first full 48-mg treatment dose on day 5, with standard weekly dosing resuming in cycle 2. For glofitamab, the rapid protocol achieves the first full 30-mg dose by day 12 of an 18-day compressed cycle 1, compared with the start of cycle 2 on the standard timeline.

Accelerated dosing does not relax monitoring requirements. It compresses the timeline in which observation and escalation infrastructure must be ready to act. Hoffmann emphasized that rapid step-up is appropriate when the institution has clear protocols for CRS recognition, tocilizumab administration, and escalation already embedded in the admission workflow, rather than as an improvisation.

STARGLO

A Meaningful Trial, a Difficult Approval

The phase 3 STARGLO trial (NCT04408638) enrolled 274 transplant-ineligible patients with relapsed/refractory DLBCL and compared glofitamab plus gemcitabine and oxaliplatin (GemOx) against rituximab plus GemOx (R-GemOx). In an updated analysis with a median follow-up of 20.7 months, glofitamab-GemOx extended the median OS from 12.9 to 25.5 months (HR, 0.62) and improved the median PFS from 3.6 to 13.8 months (HR, 0.40). The 24-month OS rate was 52.8% with the combination vs 33.5% with R-GemOx. CRS occurred in 44% of glofitamab-exposed patients and was grade 3 in 2.3%, with events clustering in cycle 1.

Despite these results, the FDA declined to approve the combination. Of 274 enrolled patients, only 25 (9%) were treated at North American sites, compared with 161 (59%) at Asian sites and 88 (32%) at European sites. When outcomes were broken down by region, the PFS HR was 0.27 in Asia, 0.84 in Europe, and 2.25 in North America. The trial’s benefit did not hold in the population for whom US approval was sought.

Hoffmann addressed this outcome directly. The University of Kansas Cancer Center was among the 11 US sites that activated, but more than 50 sites had been approached and declined, and he acknowledged that his own team would not open the same trial today. The issue is structural and extends beyond any single sponsor: Control arms considered acceptable at non-US institutions, where access to intensive salvage regimens is more limited, are often deemed inadequate by US academic centers, where CAR T-cell therapy and platinum-based salvage regimens are the standard. Non-US enrollment is easier to achieve, the enrolled patient populations differ meaningfully from US practice populations, and the observed benefit may not replicate domestically when data are aggregated. Pharmaceutical companies have incentives to enroll efficiently across global sites; those incentives and the need for US-relevant evidence do not always align.

Epcoritamab-Based Combinations

EPCORE NHL-2 Arms 4 and 5

The phase 1b/2 EPCORE NHL-2 trial (NCT04663347) evaluated epcoritamab in combination with chemotherapy in 2 distinct patient populations. Arm 5 enrolled 103 transplant-ineligible patients with relapsed/refractory DLBCL; they had a median age of 72 years, 28% had prior CAR T-cell therapy, and 70% were refractory to their last systemic therapy. The ORR by independent review committee assessment was 85%, with a CR rate of 61%, and the median time to response was 1.5 months. Among complete responders, 62% remained in CR at 18 months. Grade 3 or higher neutropenia occurred in approximately 45% of patients, and grade 3 or higher thrombocytopenia in 20%. Treatment-related mortality was 12.6%, driven substantially by COVID-19–related deaths in a high-risk older cohort.

Arm 4 enrolled 29 transplant-eligible patients with high-risk disease, more than 80% of whom had primary refractory disease or relapse within 12 months of initial therapy, and combined epcoritamab with rituximab, dexamethasone, cytarabine, and oxaliplatin or carboplatin (R-DHAX/C). Among the 13 high-risk patients who proceeded to autologous stem cell transplantation (ASCT), the 2-year PFS rate was 100%, compared with 61% in the broader high-risk cohort and 60% in the overall study population. CRS occurred in 45% of patients and was entirely grade 1 or 2. There were no fatal treatment-related adverse events. The contrast in treatment-related mortality between the 2 arms—0% in the transplant-eligible population vs 12.6% in the transplant-ineligible cohort—reflects the difference in patient selection rather than regimen toxicity.

When Chemotherapy Adds Value and When It Does Not

Adding chemotherapy to a bispecific antibody deepens responses and, in STARGLO, produced a meaningful OS benefit. It also adds myelosuppression, infection risk, and treatment-related mortality that ranges from 8% to 13% in older non–transplant-eligible patients. Hoffmann was unambiguous: For a 72-year-old patient with baseline cytopenias and no path to transplant, a 12.6% treatment-related mortality demands careful justification that an impressive response rate does not automatically supply. For a fit patient with high-risk disease and a transplant plan, the 2-year PFS data from the transplant-eligible arm make a compelling case.

Hoffmann also raised an unresolved concern specific to the post–CAR T-cell setting: Patients with prior CAR T-cell therapy are underrepresented in every combination study, and it remains unclear whether concurrent platinum or gemcitabine-based chemotherapy could impair a subsequent CAR T-cell construct if the patient remains eligible for cellular therapy. That question should not be assumed away when the combination is being considered as a bridge.

Clinical Decision-Making

Five Cases

Hoffmann grounded the evidence in 5 clinical cases that together spanned the range of scenarios encountered in practice.

A 36-year-old fit man with primary refractory DLBCL had not had success with rituximab plus ifosfamide, carboplatin, and etoposide (R-ICE). He relapsed 90 days after CAR T-cell therapy with post-CAR cytopenias that precluded clinical trial enrollment. With a matched sibling donor available and willingness to proceed to allogeneic transplant, he received epcoritamab plus rituximab plus dexamethasone, oxaliplatin, and cytarabine (R-DHAOx) using a rapid escalation protocol, the combination with the strongest supporting data in the high-risk, transplant-intent setting.

A 68-year-old man with DLBCL with multiple relapses after both prior ASCT and prior CAR T-cell therapy had stage IV chronic kidney disease and baseline cytopenias that rendered him ineligible for clinical trials and allogeneic transplant. Single-agent epcoritamab on an indefinite schedule was the appropriate choice. Adding chemotherapy in the context of renal insufficiency, existing marrow compromise, and extensive prior platinum exposure would have introduced unacceptable risk without a proportionate expectation of benefit.

A 76-year-old actively smoking man with 7 prior lines of therapy since a 2004 DLBCL diagnosis, including a prior autologous transplant, declined CAR T-cell therapy because he lived 2 hours from the treating center. Glofitamab was the right fit with its every-3-week schedule, fixed 12-cycle duration, and established activity in multiply relapsed disease.

A 60-year-old fit man who relapsed 15 months after rituximab plus cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) presented with progressive disease that was causing obstructive jaundice from a single periportal lymph node. His insurer required CAR T-cell therapy or transplant out of state. Hoffmann chose rapid-escalation glofitamab plus GemOx, a regimen that would preserve the future CAR T-cell option mandated by his insurer while providing immediate disease control, with consideration of consolidative radiation to the isolated periportal site.

A 74-year-old man with an ECOG performance status of 2 lived 2 hours from the center and had relapsed transformed DLBCL from marginal zone lymphoma with secondary central nervous system (CNS) involvement. His cerebrospinal fluid had cleared with R-ICE and intrathecal chemotherapy, but frailty precluded CAR T-cell therapy. Glofitamab, which has demonstrated CNS activity, was combined with zanubrutinib to enhance CNS penetration and activity in the underlying marginal zone histology. The every-3-week schedule reduced the logistical burden for a patient who could not make frequent trips.

Immune and Cellular Therapy in Myeloid Disorders: 2025 Updates

Myeloid Malignancies

Redefining the Immunotherapeutic Frontier in AML, MDS, and Myelofibrosis

Abdulraheem Yacoub, MD, professor of medicine and clinical director of ambulatory hematology clinics in the Division of Hematologic Malignancies and Cellular Therapeutics at the University of Kansas Cancer Center, presented a state-of-the-field review of immune and cellular therapy in myeloid malignancies. Yacoub’s research program is centered on the design and conduct of clinical and translational trials in myeloproliferative neoplasms (MPNs) and myeloid cancers. His presentation at the ICE-T Congress reflected both the breadth of that experience and his candid assessment of where the field has succeeded, where it has repeatedly failed, and where the most scientifically credible opportunities now lie.

The presentation was organized around 3 domains: the immune checkpoint blockade experience in AML and myelodysplastic syndromes (MDS), the emerging clinical data for TIM-3–directed therapy with sabatolimab, and the mutation-specific cellular immunotherapy programs targeting calreticulin (CALR)-mutant myelofibrosis.

Why Myeloid Disease Has Resisted Immune Therapy

Yacoub opened by identifying the structural obstacles that distinguish myeloid malignancies from the lymphoid and solid-tumor contexts in which immunotherapy has proved most effective. The core problem is antigen selection. Candidate targets such as CD123 and CD33 are expressed not only on leukemic blasts but on normal myeloid progenitors, creating a therapeutic index problem that has no analog in CD19-directed B-cell strategies, where the tolerable consequence of on-target normal tissue injury is manageable B-cell aplasia. A second obstacle is the state of the bone marrow at diagnosis: High-risk MDS and AML are frequently accompanied by depleted hematopoietic reserves and a stem cell pool that cannot reliably sustain recovery after intensive immunotoxic intervention. Third, and perhaps most consequential for immune checkpoint strategies, is T-cell exhaustion. Chronic antigenic stimulation, inflammatory signaling, and years of prior cytotoxic therapy can leave the endogenous T-cell compartment profoundly impaired, with upregulated coinhibitory molecules, dysfunctional effector activity, and limited capacity for sustained antileukemic response.

These individual barriers operate within a broader context of immune dysregulation that is already established at the time of diagnosis. The myeloid tumor microenvironment is characterized by tolerogenic dendritic cells that propagate T-cell anergy, expanded regulatory T-cell populations that suppress effector responses, dysfunctional natural killer cell cytotoxicity, and increased expression of multiple coinhibitory receptors on both CD4+ and CD8+ T cells. It is an environment that has blunted the activity of essentially every immune checkpoint agent studied in this space.

Checkpoint Blockade in AML and MDS

A Consistently Negative Data Set

The clinical history of CTLA-4 and PD-1/PD-L1 blockade in myeloid malignancies began with an early signal that has not been reproduced at scale. In a multicenter phase 1/1b study (NCT01822509), 28 patients with relapsed hematologic malignancies after allogeneic stem cell transplantation, including 12 with AML, received ipilimumab at 3 mg/kg or 10 mg/kg every 3 weeks at a median of 19.3 months after transplant. Among the 22 patients treated at 10 mg/kg, 5 achieved a complete response, including 4 patients with extramedullary AML and 1 patient with MDS that had evolved to AML, suggesting particular activity in extramedullary disease. In a separate single-arm study (NCT01757639), 29 patients with refractory MDS or AML treated across 7 US centers had a CR rate of 3.4%. Seven patients (24%) achieved prolonged stable disease lasting at least 46 weeks, including 3 with disease stability beyond 1 year, and 5 subsequently proceeded to allogeneic transplantation without apparent excess toxicity.

When these observations were subjected to randomized evaluation, the results were uniformly negative. The phase 2 FUSION-AML-001/HR-MDS-001 study (NCT02775903) enrolled 213 patients with untreated high-risk MDS (n = 84) or AML ineligible for intensive chemotherapy (n = 129) and randomly assigned them to azacitidine with or without the anti–PD-L1 antibody durvalumab. In both cohorts, adding durvalumab did not improve response rates, PFS, or OS. Translational correlative analyses explained the result biologically: Durvalumab did not promote T-cell expansion in peripheral blood or bone marrow, and no treatment-induced changes in PD-1 or TIM-3 expression on CD4+ or CD8+ T cells were detected in either arm. TP53-mutant disease has been hypothesized to be a potentially enriched subgroup, based on preclinical data showing elevated PD-L1 expression and downregulated miR-34 in TP53-mutant cells. In 61 patients with TP53-mutant disease, however, ORR was independent of mutation status, and the adverse prognostic impact of TP53 mutation on OS was not mitigated by durvalumab.

The phase 2 BLAST MRD AML-2 trial (NCT04284787) extended the negative data set to the venetoclax-based combination context. Fifty-eight patients 60 years or older with newly diagnosed AML ineligible for intensive induction were randomly assigned 1:1 to azacitidine plus venetoclax with or without pembrolizumab. The primary end point—MRD-negative CR or CR with incomplete count recovery within 6 treatment cycles (CR/CRi), assessed centrally by multicolor flow cytometry—was achieved by 44.8% of patients in both arms. No differences were observed in MRD-negative CR rates, best ORRs, or composite CR/CRi rates. Presented at the 2024 American Society of Hematology Annual Meeting and Exposition, these data close the case against PD-1/PD-L1 augmentation of hypomethylating agent–based backbones in unfit AML.

Sabatolimab and TIM-3

A More Encouraging Biology

TIM-3 differs from PD-1 and PD-L1 in a clinically relevant way. It is expressed on leukemic stem cells and myeloid blasts but absent from normal hematopoietic stem cells, offering a dual mechanism that the PD-1 axis cannot provide: direct targeting of the malignant clonal hierarchy alongside reactivation of exhausted T effector and natural killer cell populations. This biology supported the clinical development of sabatolimab, a TIM-3–directed antibody, in combination with azacitidine.

The phase 2 STIMULUS-MDS1 study (NCT03946670) randomly assigned patients with high-risk MDS to sabatolimab plus azacitidine vs placebo plus azacitidine. Neither co–primary end point reached statistical significance: The CR rate, respectively, was 21.5% vs 17.7% (P = .769), and the median PFS was 11.1 months vs 8.5 months (HR, 0.749; P = .102). Two findings within the data were worth carrying forward. First, among patients who achieved CR, median duration of response was 18.0 months in the sabatolimab arm vs 9.2 months with placebo, a near-doubling of CR durability that conventional response metrics do not capture. Second, an exploratory sex-stratified analysis found that women treated with sabatolimab survived a median of approximately 8 months longer than women on placebo plus azacitidine (26.8 vs 19.1 months), with no corresponding survival difference in men.

The phase 3 STIMULUS-MDS2 trial (NCT04266301) evaluated sabatolimab plus azacitidine vs placebo plus azacitidine in 530 patients with intermediate-, high-, or very high-risk MDS or chronic myelomonocytic leukemia-2 (CMML-2). At the primary analysis, the study did not meet its primary end point of OS. The median OS was 22.31 months with sabatolimab plus azacitidine vs 18.83 months with placebo plus azacitidine (HR, 0.85; 95% CI, 0.67-1.07; 1-sided P = .0825). Because the primary end point was not met, key secondary efficacy end points were not formally tested for statistical significance. Descriptively, the median PFS was 13.6 months vs 10.1 months (HR, 0.70; 95% CI, 0.56-0.86), the median leukemia-free survival was 19.4 vs 13.7 months (HR, 0.78; 95% CI, 0.61-1.00), CR rates were 19.6% vs 14.3%, and the ORRs were 42.6% vs 34.7%, respectively. Despite numerical trends favoring sabatolimab, the trial did not establish a statistically significant survival benefit, and Novartis subsequently discontinued the sabatolimab clinical development program.

Mutation-stratified analyses identified DNMT3A-mutant disease as a potentially biomarker-enriched subgroup, with an OS HR of 0.41 (95% CI, 0.22-0.78) favoring sabatolimab, while no consistent survival benefit was observed across ASXL1, TP53, RUNX1, TET2, or splicing factor mutations. The sex-specific survival differential from STIMULUS-MDS1 was independently replicated: Women treated with sabatolimab plus azacitidine survived a median of approximately 8 months longer than women on placebo plus azacitidine (25.1 months vs 17.2 months, respectively), with no meaningful OS separation in men. The biological basis of this sex-specific effect remains unknown and is the subject of ongoing investigation.

ENHANCE

A Definitive Negative for Anti-CD47 Therapy

Magrolimab, an anti-CD47 monoclonal antibody that disrupts the phagocytic “don’t eat me” signal on leukemic cells, was evaluated in the phase 3 ENHANCE trial (NCT04313881) in combination with azacitidine in patients with higher-risk MDS. Among 539 randomly assigned patients, magrolimab plus azacitidine failed to improve efficacy outcomes compared with placebo plus azacitidine. The ORR was 53.7% vs 58.7%, the CR rate was 21.3% vs 23.6%, and median OS was 15.9 vs 18.6 months (HR, 1.203; 95% CI, 0.947-1.528), respectively. Grade 3 or higher treatment-emergent adverse events occurred in 92.8% vs 79.2%, serious adverse events in 71.9% vs 51.5%, and treatment-emergent deaths in 15.2% vs 9.8%, respectively. Following the negative phase 3 results, development of magrolimab in MDS was discontinued.

Mutation-Specific Cellular Immunotherapy in Myelofibrosis

The mutCALR Frontier

The most scientifically compelling portion of Yacoub’s presentation addressed a domain central to his research program: mutation-directed immune therapy targeting the CALR neoepitope in MPNs. CALR mutations (mutCALR), predominantly a 52-bp deletion (type 1) or 5-bp insertion (type 2) in exon 9, generating a frameshift that produces a mutation-specific C-terminal neoantigen, are present in approximately 25% of essential thrombocythemia cases and 30% to 35% of primary myelofibrosis cases. The mutant C-terminus aberrantly binds nascent thrombopoietin receptor in the endoplasmic reticulum, trafficking to the cell surface and driving cytokine-independent JAK/STAT activation. Because this neoepitope is absent from normal hematopoietic stem cells, it offers a selectivity window that pan-myeloid antigens cannot provide.

Several platforms targeting mutCALR have entered clinical or advanced preclinical development. INCA033989 is a fully human, Fc-silent IgG1 monoclonal antibody with selective binding affinity for the mutCALR neoepitope; it antagonizes mutCALR-driven oncogenic signaling and inhibits malignant cell proliferation while sparing normal hematopoiesis in preclinical models. INCA035784 is a mutCALR N-domain–specific T-cell redirecting bispecific antibody being investigated in essential thrombocythemia and myelofibrosis in a phase 1 trial (NCT07008118). JNJ-88549968, a CD3 × mutCALR T-cell redirecting bispecific antibody, is in active phase 1 clinical evaluation (NCT06150157).

On the cellular therapy side, mutCALR-directed CAR T cells developed at University College London in England demonstrated high in vitro cytotoxicity against mutCALR-expressing cell lines, selective killing of mutCALR-positive cells isolated directly from myelofibrosis patient samples, and in vivo efficacy in murine models without unexpected toxicity. A critical feature distinguishes this construct from CAR T-cell programs targeting pan-myeloid antigens: Because approximately 50% of HSCs in affected patients carry the CALR mutation, mutCALR-directed CAR T cells are not expected to produce ablative pan-hematopoietic toxicity. SLAMF7-directed CAR T cells developed at Johns Hopkins University and Mayo Clinic exploit SLAMF7 expression on myelofibrosis monocytes and bone marrow cancer-associated fibroblasts, with preclinical data showing effective targeting of SLAMF7-positive stromal elements without significant toxicity to normal hematopoiesis.

Yacoub also reviewed donor-derived T cells engineered by TScan Therapeutics to target patient-specific minor histocompatibility antigens expressed selectively on hematopoietic cells. HA-1 and HA-2, lineage-restricted antigens with high and homogeneous hematopoietic expression, serve as targets for donor T cells that are inherently intolerant to recipient-specific alleles, a strategy designed to capture graft-vs-leukemia biology without the broad alloreactivity that complicates conventional donor lymphocyte infusions.

Advancing CAR T-Cell Delivery in Myeloid Disease

Yacoub outlined 4 strategic directions to improve CAR T-cell utility in myeloid neoplasms. Multitarget constructs engaging 2 or more myeloid antigens simultaneously address antigen escape, the dominant resistance mechanism that undermines single-target approaches. Post–allo-SCT consolidation with CAR T-cell therapy offers a platform to eliminate MRD in the setting of established donor chimerism, potentially reducing relapse in high-risk patients without the competing risks of primary transplant. Combinations with sensitizing agents, hypomethylating agents, BET inhibitors, and immune-priming strategies may modulate the tumor microenvironment to improve T-cell trafficking and persistence. Gamma-delta T cells, which mediate MHC-unrestricted cytotoxicity through innate recognition mechanisms, may be better suited than conventional αβ T cells to a profoundly immunosuppressive myeloid niche where sustained antigen-presenting cell function cannot be assumed.

Cellular Therapeutics in GI Oncology: Redefining the Boundaries of Treatment

Solid Tumors

Beyond the Hematology Blueprint

CAR T-cell therapy transformed outcomes in hematologic malignancies by targeting antigens that are uniformly expressed on malignant cells, absent from critical normal tissues, and accessible in an immunologically permissive environment. Solid tumors offer almost none of those conditions. Gastrointestinal (GI) cancers are characterized by antigen heterogeneity within and across tumor subclones, deep immunosuppression within the tumor microenvironment (TME), inadequate T-cell trafficking through dense stromal barriers, hypoxia and metabolic constraints that exhaust infiltrating lymphocytes, and on-target toxicity in normal gastrointestinal epithelium that shares antigen expression with the tumor. Progress in this space requires engineering solutions to each of those problems.

During the ICE-T Congress, Raed Al-Rajabi, MD, professor of medical oncology and GI medical oncology group lead at the University of Kansas Medical Center, and advanced clinical informatics director at the University of Kansas Cancer Center, presented the current clinical evidence for adoptive cell therapies in GI malignancies, spanning CAR T-cell therapy, tumor-infiltrating lymphocyte (TIL) therapy, and T-cell receptor (TCR) T cell therapy. Early signals have emerged across each platform. None has yet produced the level of evidence required to change standard practice.

CLDN18.2-Targeted CAR T-Cell Therapy

The Most Mature GI Program

Claudin 18.2 (CLDN18.2), a tight-junction protein expressed in gastric mucosa and retained across a meaningful proportion of gastric, gastroesophageal junction, and pancreatic tumors, is the most clinically advanced target for CAR T-cell therapy in GI oncology. Satricabtagene autoleucel (satri-cel; formerly CT041) was evaluated in a phase 1 study (NCT03874897) of 59 patients with gastric cancer. The ORR was 55% and the median PFS was 5.8 months. Patients with high CLDN18.2 expression and no liver or bone metastases fared considerably better than the overall cohort.

CRS occurred in 97% of patients but was grade 1 or 2 in all cases; no grade 3 events occurred. The more clinically distinctive toxicity was on-target injury to normal gastric epithelium: nausea in 67% of patients, vomiting in 53%, and gastric mucosal edema or erosion on endoscopy in 9%, including 1 grade 3 event. Sixty percent of patients received reinfusions, reflecting the limited persistence of the initial product and raising practical questions about the feasibility and cost of repeated autologous manufacturing as a routine component of care.

A subsequent randomized phase 2 (NCT04581473) analysis demonstrated a PFS improvement with satri-cel over standard chemotherapy, the first randomized efficacy signal for CAR T-cell therapy in any GI solid tumor. Al-Rajabi acknowledged the significance of that result while placing it in context. CLDN18.2 is now one of the most crowded target spaces in GI oncology, with more than 30 agents in clinical development including antibody-drug conjugates, bispecific antibodies, and first-line combinations incorporating zolbetuximab. Defining where CAR T-cell therapy belongs within that landscape will require larger survival data and clearer patient selection criteria.

Armored CAR T-Cell Therapy

Engineering Resistance to the TME

TGF-β–mediated immunosuppression is among the most potent mechanisms by which solid tumors neutralize infiltrating T cells. C-CAR031 addresses this by incorporating a dominant-negative TGF-β receptor into a GPC3-directed construct, enabling the engineered cells to resist TGF-β–mediated inactivation while maintaining cytotoxic function in hepatocellular carcinoma (HCC). In 22 evaluable patients, the ORR was 50% with a median duration of response of 7.4 months; earlier nonarmored GPC3-directed CAR T-cell constructs achieved ORRs of approximately 15%. CRS occurred in 92% of patients and was grade 3 in 4%. The construct is now being evaluated globally as AZD5851 in the ATHENA trial (NCT06084884), a global phase 1/2 study in patients with GPC3-positive advanced/recurrent HCC. Other GPC3-directed armored programs are also in early development. A phase 1 trial (NCT04405778) is evaluating TAK-102; researchers reported cytokine elevation and favorable TME changes in 11 early-phase patients, providing additional support for both the target and the armoring strategy.

On-Target, Off-Tumor Toxicity

The Price of Shared Antigen Expression

CLDN18.2, guanylate cyclase C (GUCY2C), carcinoembryonic antigen (CEA), and EGFR are all expressed in normal gastrointestinal epithelium. Engineered T cells directed at these antigens attack both the tumor and the normal tissue sharing the same surface protein. In a phase 1 study (NCT05287165) of GUCY2C-directed CAR T-cell therapy (IM96) in 19 evaluable patients with colorectal cancer, the ORR was 26%, but grade 3 diarrhea occurred in 55% of patients, rash of any grade in 70%, and mucositis in 35%. CEA-targeted therapy has produced colitis and pulmonary infiltration. EGFR-targeted approaches are associated with skin toxicity and mucositis. These are not idiosyncratic adverse events; they are mechanistic consequences of the target biology.

Several mitigation strategies are under investigation. Conventional cytokine-directed approaches, such as corticosteroids, IL-6 and IL-1 receptor blockade, and JAK-1 and Src inhibitors, manage inflammation after it develops. Programmed shutoff mechanisms, including suicide switches and antibody-directed elimination of CAR T cells when toxicity emerges, attempt to terminate effector activity before irreversible injury occurs. Hypoxia-responsive CAR constructs restrict cytotoxic activation to the low-oxygen environment of solid tumors, sparing normally oxygenated epithelial tissue. A phase 1 trial (NCT05396300) of hypoxia-responsive CEA-directed CAR T-cell therapy in 40 heavily pretreated patients reported an ORR of 25% with intraperitoneal delivery vs 8% intravenously. Diarrhea and colitis occurred in 31% of those in the intraperitoneal arm vs 4% in the intravenous arm, demonstrating that delivery route meaningfully modulates on-target toxicity.

Innate Immune Effectors and In Vivo Manufacturing

Two platforms beyond conventional CAR T-cell therapy have shown early activity in the solid tumor context. CAR-macrophage (CAR-M) therapy leverages the natural tumor-infiltrating capacity of macrophages, which penetrate dense stromal environments through matrix metalloproteinase secretion that conventional T cells cannot replicate. CT-0508, a HER2-directed CAR-M product, was detectable in the TME of 8 of 9 evaluated patients through day 28, with evidence of favorable microenvironmental remodeling. A combination with anti–PD-1 is under active investigation within the same phase 1 protocol (NCT04660929). In a phase 1 trial (NCT05108623), allogeneic invariant natural killer T-cell therapy with agenT-797 demonstrated 1 partial response and 9 cases of stable disease among 34 patients with advanced solid tumors, including gastric cancer, with increased CD8+ T-cell infiltration and T_H1 polarization confirmed on RNA sequencing.

Al-Rajabi also discussed in vivo CAR T-cell manufacturing—the direct delivery of CAR gene constructs via engineered nanocarriers or viral vectors to generate CAR T cells within the patient. If this approach matures, it will eliminate the need for centralized ex vivo manufacturing, remove the vein-to-vein delay, and substantially reduce cost. The logistical and financial barriers of conventional autologous production will not be resolved through process improvements alone. In vivo generation represents a fundamentally different model. Preclinical and early clinical work is ongoing.

TIL and TCR T-Cell Therapy

Reaching What CAR T Cells Cannot

CAR constructs bind surface proteins, which excludes most of the mutational landscape of GI tumors from their reach. TIL therapy and TCR-T cell therapy address this limitation by recognizing intracellular antigens presented by MHC molecules, including oncogenic driver mutations that define gastrointestinal tumor biology.

TIL therapy isolates tumor-reactive lymphocytes from resected tumor tissue, expands them ex vivo, and reinfuses them after lymphodepleting chemotherapy. The polyclonal T-cell receptor repertoire of TILs enables simultaneous recognition of multiple patient-specific epitopes, including neoantigens invisible to CAR constructs, providing a mechanistic advantage against antigen heterogeneity and clonal evolution. TIL therapy received FDA accelerated approval for metastatic melanoma in 2024. A proof-of-concept case in colorectal cancer demonstrated regression at most metastatic sites following infusion of 1.11 × 10¹¹ HLA-C*08:02–restricted TILs comprising 4 T-cell clonotypes targeting the KRAS G12D mutation. One lesion escaped through MHC class I downregulation, illustrating the resistance mechanisms that will need to be addressed as the platform scales to broader use.

TCR-T cell therapy applies an engineered receptor to the same principle, enabling recognition of intracellular peptide-MHC complexes and making oncogenic driver mutations such as KRAS G12D targetable in patients who cannot be reached by surface-binding CAR constructs. HLA restriction limits applicability to patients expressing the relevant haplotype. The platform received FDA accelerated approval in August 2024. Active trials at the University of Kansas Cancer Center include a phase 1 study of NT-112 targeting KRAS G12D in HLA-C*08:02–positive patients with advanced solid tumors (NCT06218914) and a phase 1/2 study (NCT05592626) of invikafusp alfa (STAR0602), a bifunctional TCR-targeting antibody-fusion molecule, in advanced solid tumors.