Chimeric antigen receptor (CAR) engineering of immune effector cells (IEC), such as T cells or Natural Killer (NK) cells, is a promising frontier of personalized cancer therapy, manifesting sustained remissions in specific populations of patients with relapsed or refractory (R/R) lymphoid malignancies (1). This approach entails reprogramming immune effector cells with synthetic receptors equipped with multiprong features enabling antigen recognition, downstream signaling, and costimulatory activation, ultimately resulting in targeted elimination of antigen-expressing tumor cells. The promise of CAR-T cell therapy has primarily been realized through CD19-directed therapies against B cell malignancies (2–6), and more recently through BCMA-directed therapies against multiple myeloma (7, 8), for which there are several commercial products now available. This success has spurred extensive preclinical investigation of cellular immunotherapy against other malignancies with an expectant dynamic and exciting pace of new indications. However, the clinical efficacy of CAR-based therapies against other malignancies, particularly solid tumors, has not yet been matched in many settings due to issues relating to trafficking, infiltration, clonal heterogeneity, off-target activity or on-target off-tumor activity against normal tissues, and the hostile immunosuppressive TME (9). There are additional logistical and clinical limitations to CAR-T cell therapy. The cumbersome and lengthy production associated with manufacturing an autologous product renders high expenditure and restrictive use (10). Even if previously generated allogeneic T cell products are HLA-matched with the recipient, genetic modification is still required to reduce the risk of graft-versus-host disease (GvHD) mediated by the native αβ T-cell receptor (TCR). Furthermore, CAR-T cell therapy can cause toxicities, such as immune effector cell–associated neurotoxicity syndrome (ICANS) and cytokine release syndrome (CRS), which can increase hospitalization duration and treatment cost (10). NK cells are innate immune cells that serve as an attractive alternative platform for CAR engineering and offer several advantages to the use of T cells, including avoiding the risk of GvHD, scalability in an off-the-shelf allogeneic cellular product, and a superior safety profile (10).

Enduring challenges with CAR engineered effector cells that may yield suboptimal response and disease relapse include i) antigen negative recurrence from immunoediting and antigen escape in the tumor, or ii) limited in vivo persistence or functional exhaustion of the CAR-engineered cells, resulting in antigen positive relapse (11). Strategies to combat antigen negative relapse thematically involve targeting multiple antigens either simultaneously or in a sequential manner. Approaches to overcome effector exhaustion or to increase proliferation and potency are varied and include i) altering the choice of co-stimulation (12), ii) cytokine cargo coupling (e.g. IL-15, IL-12, IL-18) with awareness of the detriment from unconstrained tonic signaling (13), and iii) synergizing CAR adoptive cell therapy with checkpoint blockade (14), among other tactics. Such advancements are being supplanted with pioneering, multiplexed gene editing tools including Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR) based editing to introduce deliberate genetic perturbations that enhance the potency of CAR effectors by, for example, reprograming immune cell metabolism, equipping cells with methods to resist exhaustion, disrupting pathways for immunosuppression, or improving survival (11). Such interventions and investigations show enormous promise in the marriage of sophisticated, hypothesis-driven modification of immune cells in a site-directed fashion to modulate cellular pathways that negatively impact the efficacy and fitness of CAR-modified effector cells.

In this review, through the prism of challenges the field has been facing with CAR-based immunotherapeutic strategies, we will discuss the evolution of innovative multiplex engineering strategies and precision gene editing tools that are designed to enhance the therapeutic scope and potency of this groundbreaking personalized therapy.

The expanding possibilities in personalized medicine coincide with the dramatic improvements in immune cell engineering, most notably the successful transition of CAR-T and CAR-NK cells as well as TCR gene therapy (15) into the clinic.

One of the most valuable features of cellular immunotherapy lies in the capacity of infused immune cells to expand and evolve in vivo. Indeed, some studies point towards peak CAR transcript levels as assessed in the patients’ peripheral blood to be a potential predictor of clinical response (16, 17). Contrary to antibody-based strategies, which often need to be administered continuously to mitigate rapid in vivo degradation, cell therapy products are ‘living drugs’ which exhibit striking plasticity with regard to their phenotypic, transcriptomic and epigenetic composition post-infusion. This dramatic evolution is shaped by the continuous selective pressures arising from the interaction with both the tumor cells as well as bystander host immune cells and may have profound effects on the persistence, memory formation and anti-tumor potency of infused immune effector cells (18–20). Seminal studies have in recent years investigated the single-cell trajectories of infused CAR-T cells and found key pre-infusion transcriptomic signatures which correlate with response and potential toxicity and may in the future serve as predictive biomarkers to help guide data-driven treatment decisions (21–24).

Novel high-dimensional assays with single-cell resolution have recently elucidated the complex epigenetic underpinnings of T cell exhaustion (25–30), and it is likely that other immune effector cells, including genetically engineered lymphocytes are subject to and governed by many of the same mechanisms although much is still to be learned in this regard.

Multiplexed gene engineering, ultimately, attempts to modulate these cellular signatures to enhance the safety and efficacy of cell therapies. By combining the effective delivery of synthetic receptors with additional cellular modifications brought about by the opportunities arising from precise gene editing, the next generation of multiplexed engineered cellular immunotherapies strives to enhance anti-tumor potency, increase in vivo persistence and overcome functional exhaustion by increasing cellular fitness and/or shielding engineered immune cells from the inhibitory cues received from endogenous checkpoints or immunosuppressive metabolites in the tumor microenvironment (TME) by introducing multiple genetic modifications ( Figure 1 ).

In the following section, we will outline the most pivotal examples of multiplexed engineered immune cells, a growing fraction of which have moved into clinical testing.

While the previous section focused on engineering strategies aiming at introducing synthetic transgenic payloads to enhance immune cell function, the following section aims to portray precision genome editing tools being employed to disrupt and modulate endogenous immune cell pathways which ultimately restrain anti-tumor potency.

Great strides have been made in recent years to endow immune cells not only with antigen specificity but also enhanced functionality and increased potency by multiplexed genetic engineering and gene editing strategies. Common to all these efforts is that they attempted to overcome previously established and recognized mechanisms of immune cell dysfunction and anergy. While the field is eagerly waiting to see the outcomes when these multi-engineered cell products eventually advance to the clinics, novel unbiased and large-scale discovery approaches have in recent years opened up new avenues to decode novel and undescribed mechanisms of immune cell regulation that might prove instrumental in designing the next generation of engineered immune effector cells.

Pooled loss-of-function CRISPR screens in cancer cell lines have become an invaluable tool to discern molecular mechanisms of therapeutic sensitivity and resistance. A large body of work across a number of different human cancer types has elucidated genetic dependencies that determine responses to anti-neoplastic treatments (154, 155). In recent years, combinatorial screens have further expanded these possibilities and allowed to identify systematically distinct genetic interactions, for instance, to discover synthetic lethal partners which might be exploited for pharmacological interventions. The unique molecular features of the DNA endonuclease Cas12a, which can simplify multiple simultaneous genetic edits, has greatly improved the editing efficiency compared to traditional Cas9 approaches and now permits successful identification of hits in combinatorial drop-out screens (156–159).

Whereas combinatorial screens can bring to light genetic interactions which may drive the onset, maintenance or progression of tumor cells, other platforms set out to investigate cancer-type dependent genetic dependencies. As is increasingly understood, CAR-T resistance mechanisms may vary dramatically across tumor types and explain some of the observed differential sensitivity for solid tumors versus hematologic malignancies (160). Novel platforms aim to investigate these distinct patterns of tumor-intrinsic resistance pathways in a massively parallel manner. For instance, CRISPR loss-of-function screening applied to a pool of hundreds of barcoded cancer cell lines (PRISM) in an orthogonal screening format recently helped elucidate shared cancer-intrinsic transcriptional signatures which correlate with NK cell-sensitivity (161). Noteworthy, this form of therapeutic sensitivity mapping may guide future biomarker-driven treatment algorithms.

In its latest iteration, precision genome engineering has been adapted for cell-type specific in vivo genetic editing to treat various human diseases including cancer. In a report from 2021, CRISPR/Cas9-mediated in vivo deletion of transthyretin (TTR), which is mutated in progressively fatal transthyretin amyloidosis, led to a striking decrease of circulating transthyretin protein levels in the peripheral blood of treated patients (191). Using an apolipoprotein-E (ApoE)-recruiting lipid nanoparticle (LNP) loaded with Cas9-encoding mRNA and a TTR-directed sgRNA, this platform allows to specifically correct this pathogenic variant in hepatocytes (192). Similarly, pre-clinical efforts have resulted in the in vivo generation of fibroblast activation protein (FAP)-directed CAR-T cells to treat cardiac injury, by selectively reprogramming T cells using a CD5-targeted LNP system (193). Besides LNP-mediated strategies, implantable bioinstructive scaffolds have been developed, which can generate CAR-T cells in vivo and, hence, circumvent the cost-intensive and lengthy vein-to-vein times required for ex vivo manufacturing of autologous CAR-T products (194).

Most recently, nuclease-inactive genome editing strategies have also transitioned to in vivo applications to correct pathogenic mutations in diseases like Hutchinson–Gilford progeria syndrome (195) or familial hypercholesterinemia using size-optimized adeno-associated viruses (AAV) as delivery vehicles (196).

These developments may have far-reaching implications, as they promise to dramatically reduce the time and cost of manufacturing cell products, which could ultimately increase patient access. Furthermore, the immense adaptability of precision genome editing strategies for designing and delivering DNA sequences to correct pathogenic allele variants make it likely that these powerful tools will play an increasingly important role in the treatment of many of the most devastating human diseases.

Despite the well-founded optimism, safety concerns arising from various engineering and genome editing strategies remain a preeminent concern, particularly when novel multiplex modified cellular therapeutics transition into the clinic. Recently, temporary halts of ongoing clinical trials investigating gene-edited cell products illustrated the delicate regulatory trade-off to find the right balance between granting patients access to innovative treatments while ensuring utmost patient safety.

Safety concerns may arise both from the vector system employed to deliver genetic cargo or from the induced genetic alterations, themselves. While overall vector safety has greatly improved over the last decades, nuclease-based genome editing strategies continue to carry the risk of inadvertent off-target cleavages. Off-target toxicities of CRISPR/Cas9 editing include mutational events or chromosomal rearrangements that may appear despite the first 20 nucleotide sequence of the sgRNA directing the efficiency of Cas9. Inadvertent off-target cleavage may lead to the deleterious loss of tumor suppressor genes (197). One such example of off-target mutations is the integration of DNA mismatches in the PAM-distal portion of the sgRNA sequence (134, 135, 198–200). Multiple unbiased sequencing platforms exist today and allow to detect these undesired genomic edits (201–203) and assess potential associated risks to decide whether a particular cell product can be safely infused. Global and sensitive detection of DSBs introduced by CRISPR RNA-guided nucleases is enabled through Genome-Wide, Unbiased Identification of DSBs enabled by Sequencing (GUIDE-seq) (201), Circularization for in vitro Reporting of Cleavage Effects by Sequencing (CIRCLE-seq) (202), or RNase H-dependent PCR Amplification Sequencing (rhamp-Seq) (203), which all function as assays to assess off-target cleavage. Approaches to diminish or prevent off-target effects of CRISPR/Cas9 editing include (i) identifying unique target sites without homology to other genomic regions, (ii) modifying the sgRNA by distal truncation of the 3’ end (204), (iii) substitution of regular Cas9 with high fidelity Cas9 (HiFi) or alternative nucleases such as Cas12 (205), (iv) lowering the concentration and optimizing the ratio of Cas9-sgRNA given to the cells (206), or (v) fusing dCas9 with Fokl nuclease (fCas9) to allow for higher sequence targeting specificity (207, 208). Alternatively, nuclease-inactive genome engineering including base editing and PRIME editing present potentially safer alternatives as they avoid DNA double-strand breaks and may well evolve to play a more important role for genome-edited cell products for clinical use going forward.

Today, clinical development programs aiming to translate promising cellular candidates into the clinic are in ever closer contact with regulatory bodies to jointly sketch a safe path forward for highly innovative products to enter the clinic. As the field matures, the most relevant learnings have recently been formalized in Food and Drug Administration (FDA)-issued guidance documents. One important document is focused on Considerations for the Development of Chimeric Antigen Receptor (CAR) Cell Products (209) and another important document is geared towards Human Gene Therapy Products Incorporating Human Genome Editing (210).

Within the framework of these resources, the FDA has issued recommendations regarding the chemistry, manufacturing, and control (CMC), as well as toxicology, pharmacology, and clinical study design involving CAR engineered immune effector cell therapies (209). A prominent guidance involves the choice of vector for gene delivery. Vectors integrating into cellular DNA (e.g., retroviral-based vectors or transposons) may increase the risk of delayed adverse events, including potentially by insertional mutagenesis, and this may require long-term follow-up. Such occurrences would be unexpected in non-integrating vectors, and thus such long-term follow-up for these products may not be warranted. The FDA advises that the necessity of each additional functional domain be justified in application reports, with a description of how these elements may affect CAR specificity, functionality, safety, and immunogenicity. Batch-based manufacturing and process optimization are recommended. If viral vectors are utilized, transduction reliability with batch-to-batch consistent viral multiplicity of infection (MOI) is critical. Aseptic technique should be carried out under current good manufacturing process (CGMP) conditions. Guidance also covers the manufacture and delivery of CAR products, with allowance decreed to both fresh and cryopreserved products depending on the desired shelf-life and quality of the products. Cryopreserved cell products may be more practical for delivery of the cells to different clinical sites, provided that controlled thawing procedures are carefully outlined, potential transport plans are established, and assurance regarding the safety of cryoprotectants has been assessed. Regarding study design, sponsors are recommended to arrange a priori analyses for all subjects, as well as for patients stratified by the receipt of bridging therapy or not. A plan for stopping rules from dose-limiting toxicities, adverse effect monitoring and attribution, and follow-up must be described.

The FDA has outlined separate guidelines for human gene therapy products incorporating human genome editing, encompassing CRISPR-engineered cell products (210). They elaborate on different genome editing tools, delivery methods, and conditions (e.g., ex vivo, in vivo), with stipulated goals to stabilize components, uphold aseptic processing, and minimize off-target genome alterations. The IND should contain documentation regarding genome editing efficiency and specificity. With appropriate controls, orthogonal and redundant models are recommended that incorporate in silico, biochemical, and cellular-based assays, for unbiased genome-wide analysis to detect potential off-target sites. Safety analyses should include assessments of the immunogenicity of genome editing components, viability assessment, evaluation for any residual gene editing components, and testing for selective survival advantage of edited cells, including dysregulated growth or clonal proliferation. It is essential to evaluate genomic integrity, including deletions or insertions, chromosomal rearrangements, exogenous DNA integration, and potential insertional mutagenesis or oncogenicity.

The intersection of multiplexed engineering, precision genome editing, and cellular immunotherapy has brought about an unprecedented wave of innovations which has over the last couple of years dramatically expanded the available treatment modalities to combat cancer and other devastating diseases. Iterative cycles of innovation have extended the scope of genome editing strategies and novel delivery methods have helped enable the design of multiplexed-engineered cell products, redrawing the treatment landscape in oncology and beyond. Nevertheless, important challenges remain, and the field is actively exploring opportunities to overcome them.

First, there is an imperative to better understand the mechanisms of therapeutic resistance and relapse following cellular immunotherapy. Devising strategies to overcome tumor immune escape and prevent functional exhaustion of infused immune cells while mitigating inadvertent toxicities are just two examples of how improved cellular therapeutics may evolve to become more powerful with time. As outlined, forward genetic screening approaches paired with information-rich multi-omic analyses may critically inform these novel strategies by uncovering the biological mechanisms which orchestrate the complex interplay between immune cells and tumors. Multiplexed gene engineering may, in turn, build on these learnings to enable the design of more powerful cell products which can produce more durable and potent anti-tumor immune responses.

Sketching a path forward for cellular immunotherapy to effectively combat solid tumors indisputably represents one of the most formidable challenges, the field seeks to address, and numerous highly differentiated strategies are being pursued to accomplish this goal. Besides tackling the challenges of shared antigen expression (35–37, 211), downregulation of MHC molecules (212–214) and impaired trafficking (39), the field is continuously branching out, for instance, towards aiming to restore the immunogenicity of immunologically ‘cold’ tumors, via various approached including epigenetic modulation (215, 216), tumor vaccination strategies (179) or by incorporating bacterial virulence factors to attract bystander immune cells (217). Researchers have also investigated the possibility of delivering CAR targets to solid tumors via oncolytic viruses, in the hope to render the transformed cells targetable using established CAR-T cell products (218). Furthermore, recent work has also pointed to other immune cell subsets for CAR modification (1, 219–223), including CAR-modified macrophages (CAR-M), which may provide unique advantages to better penetrate solid tumor capsules (224). Following the recent report of in vivo edited hepatocytes (191), future studies may also aim to reprogram the tumor cells directly in vivo. This approach might be leveraged to restore treatment sensitivity by augmenting the release of immune cell-attracting chemokines (99, 225) or by forcing expression of TAAs for enhanced immunogenicity (179). Other strategies might aim to selectively induce cell death, for instance, by targeted disruption of oncogenes or hybrid pharmaco-genetic interventions to suppress synthetic lethal gene interactions.

Lastly, a growing number of stakeholders in the field are aiming to expand the scope of eligible patients who can benefit from the tremendous advancements of cellular immunotherapy and precision genome engineering. The transition towards allogeneic off-the-shelf cell products (16, 31, 73, 226, 227) and in vivo-manufactured cell products (193, 194) is already in full swing. Bi- and multi-specific engager molecules are being explored to endow immune cells with CAR-like anti-tumor specificity (228–237), bypassing the complexities associated with cellular engineering entirely. Combinations of both approaches in the form of BiTE-secreting CAR-T cells epitomize the highly modular nature of next generation cellular immunotherapy (238).

Going forward, we expect that these and other emerging technologies will further advance and deepen our understanding of cancer biology and help engineer better, more potent and safer immunotherapies to combat various human cancers.

AB and MD conceived the outline for this manuscript. AB, GM, and MD wrote the manuscript and created figures. All authors contributed to the article and approved the submitted version.

AB received support from the German Research Foundation as Walter Benjamin Postdoctoral Fellow (464778766). GM receives support from the Radiological Society of North America (RSNA) and the American Society for Radiation Oncology (ASTRO). The project described was supported by RSNA Research & Education Foundation, through the Canon Medical Systems, USA/RSNA Research Resident Grant. The content is solely the responsibility of the authors and does not necessarily represent the official views of the RSNA R&E Foundation. MD acknowledges support from the Society for Immunotherapy of Cancer (SITC), The Leukemia SPORE developmental research program award, the Andrew Sabin Family Foundation and the MD Anderson AML and B-cell lymphoma moonshot programs.

MD and The University of Texas MD Anderson Cancer Center have an institutional financial conflict of interest with Takeda Pharmaceutical for the licensing of the CAR-NK cell technology.

The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.