Lentiviral Vectors for CAR-T and Other Cell Therapy Products: A Microbiological and Biosafety Perspective

1. Introduction

Viral vectors have become fundamental biotechnological tools in modern medicine, serving as highly efficient vehicles for delivering genetic material into human cells. Their ability to enable stable, targeted genomic modification has been a pillar of gene and cell therapy, opening new therapeutic strategies for monogenic diseases, malignancies, and immunological disorders previously considered incurable. Among these transformative applications, chimeric antigen receptor T-cell (CAR-T) therapy, first described in the late 20^th century, represents one of the most significant advances in oncological immunotherapy, reshaping the therapeutic paradigm for advanced hematological malignancies. [1,2,3]CAR-T cells are autologous or allogeneic T lymphocytes genetically engineered to express a CAR that redirects T-cell specificity toward tumor-associated antigens. Structurally, the CAR, defined as the main functional element of CAR-T cells, comprises four distinct domains: a ligand-binding domain, a spacer, a transmembrane domain, and a cytoplasmic domain.[4] Unlike natural T lymphocytes, CAR-T cells can recognize tumor antigens independently of the human major histocompatibility complex (MHC), enabling them to distinguish a broader spectrum of targets.[5]Upon binding tumor antigens, CAR-T cells undergo activation, expansion, cytokine secretion, metabolic reprogramming, and cytotoxic degranulation, releasing perforin and granzymes that mediate tumor cell lysis.[6]

The fundamental principle of gene therapy is the ability to administer a transgene that confers a therapeutic benefit. Although several gene delivery platforms have been developed, viral vectors remain the most widely adopted vehicles due to their high transduction efficiency and reproducibility.[7] Viral vectors are engineered to retain essential components required for cellular entry and transgene delivery while eliminating genes necessary for replication, thereby producing replication-defective viruses optimized for therapeutic use.[7]

Lentiviral vectors (LVV), γ-retroviruses, and adeno-associated viruses (AAV) are among the most used viral vectors in various applications.[1,2,3] Due to their superior transduction efficiency. γ-retroviral vectors were the first viral vectors used to produce CAR-T cells targeting CD19 and accounted for approximately one-fifth of gene therapy clinical trials. [2]However, their limited ability to transduce non-dividing cells constrains their broader application. AAVs offer a lower risk of toxicity and immunogenicity but are limited by their small packaging capacity (~50 kb), which makes them unsuitable for larger or more complex gene transgene constructs.[3]LVVs are the leading platform for CAR-T production.[1] In addition to their high gene transfer efficiency, they can transduce both non-proliferating and proliferating cells. Importantly, since the viral genome is transmitted to daughter cells, it can induce long-term, stable transgene expression and a sustained therapeutic effect.[1] Advanced vector designs have substantially enhanced biosafety by minimizing the risk of replication-competent lentivirus formation and reducing insertional activation of oncogenes. As a result, LVVs are currently the predominant vector system used in clinically approved and investigational CAR-T products worldwide.[4]

The ex vivo manufacturing process of CAR-T cells typically involves leukapheresis, T-cell enrichment and activation, viral transduction, expansion, formulation, and reinfusion into the patient. Automation and closed-system processing have become central to ensuring reproducibility, sterility, and regulatory compliance under Good Manufacturing Practice (GMP) standards. In Abu Dhabi, United Arab Emirates, our Good Manufacturing Practice (GMP) laboratory at the Abu Dhabi Stem Cells Center (ADSCC) has produced and applied clinical-grade autologous CAR-T cells for the treatment of hematological malignancies since 2024, using the CliniMacs Prodigy® platform (Miltenyi Biotec). The cells are stimulated and transduced with an LVV (Lentigen Technology, Miltenyi Biotec) encoding a CAR targeting CD19. [8] CliniMACS Prodigy® is among the most advanced automated platforms for CAR-T cell manufacturing and other applications, including stem cell enrichment and the preparation of virus-reactive T cells.[9] It integrates cell washing, magnetic cell separation, activation, transduction, and a cell culture device within a fully closed, sterile system. Its programmable interface enables standardized, GMP-compliant, multi-step processing with reduced contamination risk and operator variability.[10]

This review provides a comprehensive analysis of the evolution of viral vector technology, with particular emphasis on the biology, structural organization, and engineering of LVVs. It traces their development from foundational molecular insights into human immunodeficiency virus (HIV-1) genome organization to the design of next-generation vectors optimized for safety and efficiency. Furthermore, it examines transduction methodologies that enhance gene transfer to meet increasing clinical demand. Particular emphasis is placed on microbiological risk assessment, environmental containment, and biosafety governance in vector production and cell processing facilities. Additionally, the review explores the transition from conventional ex vivo manufacturing to emerging in vivo gene delivery, highlighting technological innovations and regulatory implications. This analysis will provide a comprehensive view of a technology that, despite its challenges, continues to evolve and redefine the limits of personalized medicine. The review may be of interest to laboratory scientists, engineers, and others responsible for biopharmaceutical processes and biosafety and quality; clinicians; facility designers; and regulatory and legislative bodies.

2. From the Biology of a Lentivirus to that of the Lentiviral Therapeutic Vector

2.1. What Are Lentiviruses?

Lentiviruses are of the genus Lentiviridae, of the family Retroviridae, characterized by their prolonged incubation period ("lenti-" from Latin, meaning "slow") and their ability to infect both dividing and resting (non-dividing) cells.[11] Unlike other retroviruses, lentiviruses can cross the intact nuclear membrane of the host cell thanks to specific pre-integration complexes, allowing them to integrate their genetic material into the host cell genome.[11]Lentiviruses possess a positive-sense single-stranded RNA genome (+ssRNA); Following entry into the host cell, reverse transcriptase converts the viral RNA into double-stranded DNA, which is subsequently integrated into the host genome by the viral integrase enzyme, forming a provirus. This stable integration allows viral genetic material to replicate alongside host DNA.[12] The lentivirus can transfer up to 10 kb of genes to the host cell, leading to the expression of encoded proteins.[13] The best-known example is HIV-1, which causes AIDS, but there are other examples, including Simian Immunodeficiency Virus (SIV) and Caprine Arthritis-Encephalitis Virus (CAEV).[14]

A breakthrough in retrovirology occurred in 1970 with the discovery of reverse transcriptase by Howard Temin and Satoshi Mizutani,[15] and independently by David Baltimore.[16] This discovery established the molecular basis of retroviral replication and stable genomic integration, ultimately enabling the conceptual development of retroviruses as gene delivery tools.[17] Subsequent natural dissection of this natural "genetic engineering" mechanism revealed the potential for retroviruses to be stripped of their pathogenic genes and adapted to transport therapeutic genes. These insights laid the foundations for gene therapy concepts that began to emerge in the early 80s. Although the field faced significant technical and regulatory challenges over the following forty years, HIV-based lentiviral vector technologies only fully matured in the late 90s. [18]

As illustrated in Figure 1, the genomic organization of wild-type HIV-1 (Figure 1A) and the virion structure (Figure 1B) provide the biological framework from which lentiviral therapeutic vectors were subsequently engineered.

2.2. Production of HIV-Derived LVVs

Modern lentiviral vectors (LVVs) are primarily derived from HIV- 1 but are extensively engineered to abolish pathogenicity and replication competence while preserving efficient gene transfer capacity. The third-generation LVV platform, currently the clinical standard for gene and cell therapy applications, employs a four-plasmid system designed to maximize biosafety. This system includes a transfer plasmid containing the "genetic cargo" of interest, flanked by self-inactivating (SIN) long terminal repeats (LTR). In SIN vectors, the U3 promoter/enhancer region of the native viral LTR is removed, thereby preventing activation of adjacent host genes after integration and reducing the risk of insertional mutagenesis. The transfer construct further includes the "therapeutic" or "reporter" transgene, an internal promoter that regulates transgene expression, and the packaging signal Ψ (psi), which ensures the selective encapsidation of vector RNA into assembling viral particles. The remaining viral components are supplied by separate plasmids encoding gag/pol, rev, and the envelope glycoprotein (Env). The gag/pol plasmid provides the structural and enzymatic proteins required for particle assembly and maturation. In contrast, the rev plasmid enables nuclear export of full-length vector RNA via interaction with the recv response element (RRe). An envelope plasmid encodes a heterologous glycoprotein that determines cellular tropism; most commonly, the vesicular stomatitis virus glycoprotein (VSV-G) is used because it confers broad tropism and enhances particle stability. These plasmids are co-transfected into producer cells, typically HEK 293T cells, where viral components are expressed and self-assemble. Importantly, only the transfer vector RNA containing the Ψ sequence is packaged, resulting in replication-incompetent lentiviral particles capable of a single round of infection without generating new virions.[18,19]

Consistent with earlier observations in murine leukemia virus (MLV) and Rous sarcoma virus (RSV), sequences within the 5′ untranslated region of HIV-1 were shown to be essential for selective RNA packaging into virions. [19,20,21] Initial studies demonstrated that heterologous transcripts containing these motifs could be efficiently encapsidated, establishing the first lentivirus-derived vectors. [22,23,24,25,26] Helper-free systems were subsequently developed by expressing gag/pol and env from separate plasmids, allowing gene transfer to CD4⁺ target cells while minimizing replication risk. [27] Contemporary HIV-1–based LVVs retain only the minimal cis-regulatory elements required for vector function, including the 5′ R/U5 region through the packaging signal (Ψ), the RRE, the central polypurine tract (cPPT), and a modified 3′ SIN LTR. Additional regulatory elements—such as heterologous internal promoters, the woodchuck hepatitis virus post-transcriptional regulatory element (WPRE), internal ribosome entry sites (IRES), or chromatin insulators—may be incorporated to optimize transgene expression and further enhance biosafety. [7]

2.3. Evolution of LVV Packaging Systems

To date, four generations of the lentivirus-derived LVV system have been developed; the first three are summarized in Figure 2. Generational classification is based on the progressive elimination of accessory genes and the distribution of the remaining essential genes across multiple plasmids, minimizing sequence overlap and reducing the probability of recombination that could form a replication-competent lentivirus (RCL). [27]

2.3.1. First-Generation LVV

Utilized a single packaging plasmid containing all HIV-1 genes (gag, pol, vif, vpr, vpu, tat, rev, nef), posing a considerable recombination risk. In 1996, Naldini and colleagues developed a three-plasmid system (Figure 2A) comprising a packaging construct with expression of gag/pol, and accessory proteins, driven by a heterologous human cytomegalovirus (CMV) promoter, an envelope plasmid encoding the VSV-G, and a transfer vector with the internal transgene cassette, flanked by the HIV-1 LTR and all necessary cis elements. [23,24,27] This vector system achieved in vivo gene transfer to rat neurons and stable transduction of non-proliferating cells, demonstrating the unique capabilities of HIV-1-based systems compared to MLV-based vectors.[23] Eliminating non-essential accessory genes from the packaging plasmid improved biosafety and created space for larger transgene inserts, while maintaining the viral life cycle.[21,24,28]

2.3.2. Second-Generation LVV

To enhance biosafety, the non-essential accessory genes (vif, vpr, vpu, and nef) were deleted from second-generation LVVs, retaining only the regulatory genes tat and rev. In addition, gag/pol and env were placed on separate plasmids to reduce further the risk of homologous recombination (Figure 2B). These modifications significantly improved the safety profile for potential in vivo applications without compromising transduction efficiency.[26,29,30]

2.3.3. Third-Generation LVV

This system is the current safety standard for LVV production and is based on a four-plasmid system: a SIN transfer vector, gag/pol, rev, and envelope. The tat gene was eliminated, and transcription of the transfer vector became Rev-dependent. At the same time, the 3′ ΔU3 LTR deletion conferred self-inactivation, reducing the likelihood of RCL formation[26,31,32,33]. Additional regulatory elements, including heterologous internal promoters, the woodchuck hepatitis virus post-transcriptional regulatory element (WPRE), and the central polypurine tract (cPPT), enhance RNA stability, transcription efficiency, and viral titer.[31] These systems provide high biosafety and long-term transgene expression in dividing and non-dividing cells, including terminally differentiated populations such as neurons and hepatocytes. [34,35,36] The tat gene was eliminated, and the transcriptional activity of the transfer plasmid was made dependent on the Rev protein (supplied in a fourth plasmid) via a chimeric promoter. This extreme segmentation makes the reconstitution of a wild-type virus statistically unlikely.[31] This advance in 1998 included self-inactivating LTRs (SIN): since the 3′ ΔU3 LTR, copied to the 5′ LTR upon reverse transcription, provided a transcriptionally inactive LTR in the provirus; these vector systems added a constitutively active heterologous promoter upstream of the vector transcript in the 5' LTR and reduced the packaging constructs to separate plasmids for the expression of gag/pol and rev [26] Initially characterized using the pRRL and pCCL vectors included in this analysis, these advanced vectors carried the RSV enhancer and the CMV promoter to regulate the expression of the transfer vector.[26] Since the tat function, which activates transcription of genomic RNA from the LTR, was replaced, eliminating tat from the system increased safety. Furthermore, the inclusion of rev as a separate rev plasmid reduced the probability of obtaining RCL. By combining these modifications with deletions in the 3' LTR to confer self-inactivation, [31,32,33] these third-generation LVVs brought unprecedented biosafety considerations to the platform. They became a highly safe delivery system for gene therapy.[31] Thus, with this complete system comprising four plasmids (SIN transfer + gag/pol + rev + env), the risk of chronic lymphocytic leukemia (CLL) is substantially reduced (Figure 2C).[24]

2.3.3. Fourth-Generation LVV

On the other hand, the variants considered fourth-generation have already been tested. With these LVV variants, greater viral safety has been achieved, as the fourth-generation packaging plasmid separates the gag/pol and rev sequences into two cassettes. The fourth-generation packaging systems are the safest to date, but they remain in the experimental phase.[24] Within it, various experimental transfer plasmids have been created, which account for the differences between numerous commercial vectors and those under clinical investigation. As simple retroviruses, HIV-1-derived LVVs can hijack the host cell machinery to maintain efficient nuclear import through the intact nuclear membrane. [34]

This feature has allowed them to transduce non-proliferating and terminally differentiated cells (e.g., post-mitotic neurons, hepatocytes, or macrophages) with high efficiency. [35] The prolonged effect of viral transduction enables long-term expression of the therapeutic gene of interest (providing constant, sustained dosing after a single viral administration), which is essential for gene therapy applications.[35] In Figure 2, a schematic representation of the plasmid structure of the first three generations of HIV-derived lentiviral systems is shown.[36]

On the other hand, as already mentioned, the creation of heterologous envelopes for pseudotyping these viral particles is considered a significant advance in the field, as it enabled a substantial diversification and expansion of transduction tropism. As a result, the host range of retroviral vectors, including LVVs, could be expanded or modified. Pseudotyped LVVs consist of vector particles carrying glycoproteins (GPs) derived from other enveloped viruses. Such particles possess the tropism of the virus from which the GP was derived. For example, to exploit the natural neuronal tropism of the rabies virus, vectors designed to target the central nervous system have been pseudotyped with GPs derived from rabies virus and other viruses, such as Mokola virus (MV) and Ross River virus (RRV).[36]Among the first GPs and still the most widely used for LVV pseudotyping is the VSV-G GP, due to its broad tropism and the stability of the resulting pseudotypes. Pseudotypes using VSV-G have become the standard for evaluating the efficacy of other viral vectors. Furthermore, the incorporation of heterologous envelopes into viral particles has improved vector safety for clinical use.[36,37] Still, the most widely used envelope for viral particle pseudotyping exhibits extremely broad tropism and can therefore be used for transduction in most cells and tissues.[35] Unlike wild-type HIV-1, which encodes a complete genome capable of autonomous replication and pathogenic propagation, LVVs used for CAR-T therapy manufacturing are highly modified, replication-deficient viral particles. They retain only the minimal cis sequences necessary for gene transfer, lack accessory and pathogenic genes, are pseudotyped with heterologous envelopes, and exclusively deliver the therapeutic CAR transgene without producing viral proteins in the transduced T cells.[37,38] Table 1. Presents a comparative summary between wild-type HIV-1 types and LVVs used for the manufacture of CAR-T therapies regarding genomic content, packaging and protein expression, replication capacity, biosafety classification, and clinical function. [38,39]

With the rapid development and refinement of new gene therapy techniques, such as CRISPR-Cas9 and chimeric antigen receptors (CAR), the safe and efficient administration of these technologies is fundamental. Advances in both LVVs and other retrovirus have been corroborated, and one of the most comprehensive safety studies to date analyzed the results of 17 lots of clinical vectors, 375 manufactured T-cell products, and 308 patients treated with infusions to develop RCL or retroviral (RCL/R) and integration-directed expansion.[40] This study supported the safety profile of these vectors in this application and their continued use in oncology, infectious diseases, autoimmunity, and inherited genetic disorders. It encouraged their adoption in other therapeutic fields. This analysis does not represent all vectors that have been developed or advanced to clinical trials. Critical follow-up studies will include direct comparisons of different regulatory elements in lentiviral vectors, such as the promoter or wPRE, in patients who have developed RCL or RCL/R.[29,40]

If optimizations yield a series of vectors with significantly better characteristics, the field could move towards standardized vectors providing a "plug-and-play" system. This level of standardization could streamline processing, evaluation, and approval by regulatory bodies and would also affect the financial viability of Good Manufacturing Practice (GMP)- compliant manufacturing. This advance would enable the development of multiple standardized vectors tailored to the target population's cell type, each with distinct safety requirements and therapeutic rationales.

While many approaches, including WAS and SCID-X1 trials, use gene transfer into stem cell populations, others, such as most CAR approaches, transduce differentiated lymphocytes. [27] On the other hand, it has been shown that the degree of cell lineage differentiation affects the penetrance of leukemia following insertional oncogenesis.[27] Stem cells, by necessity, possess active signaling pathways for growth and self-renewal; therefore, they are more susceptible to oncogenic transformation. In contrast, mature cells have this self-renewal pathways reduced, which must be reactivated to facilitate uncontrolled growth. These differences have implications for the choice of vector, internal promoter, or other regulatory elements. It is necessary to investigate the minimal transgene expression required to achieve therapeutic benefit and whether a specific cell type is suitable for safely harbouring such an enhancer/promoter, to guide the choice of vector and target cell population. [27]Table 2 presents a comparison between wild-type HIV-1 and the HIV-derived LVV used for CAR-T therapy manufacturing, including key elements genomic content, long terminal repeats (LTR), packaging system, envelope glycoproteins, protein expression in target cells, replication capacity, accessory/regulatory genes, biosafety level recommendations, and clinical effects.

It should be noted that the rev gene is present in all packaging systems, as its product, REV, plays a fundamental role in the export of full-length and partially spliced viral RNA (vRNA) from the nucleus to the cytoplasm.[29]

Advanced generations of packaging plasmids also contain a strong heterologous polyadenylation (poly-A) signal, derived from the SV40 virus or bovine/human growth hormone (bGH/hGH). These strong poly-A allow for high vRNA stability and, therefore, their inclusion is advantageous for packaging and viral titers. [24,29]Furthermore, it has been shown that the inclusion, in the third generation, of the woodchuck hepatitis virus post-transcriptional regulatory element (WPRE) and the central polypurine tract (cPPT) in the viral transfer cassette further improves viral RNA stability, transcription efficiency, and overall viral titer. [30] Table 3 summarizes the four generations of LVV, including their key genomic features, plasmid systems, safety profiles, and primary uses.

2.4. Other Applications of LVVs

LVV packaging can also be crucial for the development of gene and cell therapies. [41] Depending on the characteristics of the LVV system, it can be used in the following applications:

2.4.1. Basic Research

In molecular and cell biology, LVs are used in overexpression, knockdown, and gene insertion experiments. Gene knockdown involves replacing a gene with an artificial DNA fragment. In contrast, gene insertion experiments involve inserting a gene into a specific genomic location. These techniques facilitate the study of gene functions and disease mechanisms. Large-scale collaborations are underway to use LVVs to block the expression of specific genes using RNA interference technology in high-throughput formats. On the other hand, LVs are also employed to overexpress specific genes stably, allowing researchers to examine the effects of increased gene expression in a model system. For example, LVV-mediated gene editing technologies, such as CRISPR/Cas9, can repair or replace mutated genes.[42]

2.4.2. Construction of Stable Cell Lines

LVVs can be used to generate stable cell lines in the same way as standard retroviruses. [1] The process involves infecting host cells with recombinant or pseudotyped LVVs carrying selectable markers, such as the puromycin resistance gene. This gene confers antibiotic resistance to infected host cells. By adding these antibiotics to the host cell culture medium, cells that have not incorporated the LVV genome are eliminated. [1] These surviving cells can be expanded to create stable cell lines that include the lentiviral genome and harbor the genetic information it encodes.[1] For example, in vaccine development, LVVs can act as vectors to express pathogen antigens, inducing an immune response. [43] They are used in the development of vaccines against HIV and other infectious diseases. [17,38,44]

2.5. Comparison with Other Vector Platforms

LVs possess the unique ability to stably integrate into the genomes of dividing, non-dividing, and post-mitotic mammalian cells, a capability that γ retroviruses do not have to the same extent. While AAV can also transduce non-dividing cells, it cannot stably integrate into the host genome and requires much more time to design and prepare. Furthermore, LVs are much less immunogenic than γ retroviruses and AAVs, making them more suitable for a broader range of cell types and animal models.

In Table 4, the comparison highlights differences among LVVs, γ-retroviruses, and adeno-associated viruses, the most widely used in CAR-T cells, including genomic integration capacity, target cell types, immunogenicity, design and preparation time, application range, vector capacity, stability, regulatory aspects, and available products. [1]

3. Biology and Design of CAR-T Cells

3.1. What is the Structure of the CAR-T Cell?

CARs are structurally designed with a combination of three domains: an endodomain, an anchoring transmembrane domain, and an ectodomain. [49]The latter is a ligand-specific extracellular domain, composed of a single-chain variable fragment (scFv) region and a hinge region.[50] The scFv is a fusion protein formed by the variable regions of the heavy and light chains of immunoglobulins, joined by a short and flexible linker peptide.[51] The hinge region, also known as the spacer, separates the binding units from the transmembrane domain.[52] Most CAR-T cells are engineered with hinge regions similar to those of immunoglobulins, allowing them to access the target antigen. [53]The endodomain may consist of the intracellular activation domain of CD3ζ T cells, as a single entity, or of one or more intracellular co-stimulatory (or activation) domains. [54]While the scFv provides antigenic specificity, co-stimulatory domains are key for activating effector T cells. [55] Currently, CAR-T cells are classified into five generations according to the endodomain, as shown in Figure 3. [56]

First-generation CAR-T cells contained a single fragment of the CD3ζ chain.[57] These cells depended on exogenous cytokine production, showed insufficient persistence and T cell activation, and consequently did not achieve the desired results in most studies. [58,59]

Consequently, first-generation CARs have been replaced by second-generation CARs, which incorporate an intracellular signaling domain composed of various co-stimulatory receptors located in the cytoplasmic tail of the CARs, such as CD28 or CD137 (4-1BB). [4] These co-stimulatory proteins can enhance proliferation and cytotoxicity, as well as prolong persistence.[60,61]

Third-generation CARs integrate multiple signaling domains, including CD28, 4-1BB, ICOS, and/or OX40.[62,63,64]CAR-T cells, also known as TRUCK cells (T cells redirected for universal cytokine-mediated killing),[65] are designed to release cytokines within the tumor microenvironment (TME). Additionally, they can express additional proteins, such as chemokine receptors, isotype-switching receptors, bispecific T-cell engagers (BiTEs), and blockers/inducers of specific signaling pathways. [66]In this context, next-generation CAR-T cells are being developed. The fifth generation differs from the previous ones by the integration of an additional membrane receptor.

Various approaches are being explored; among the most promising is the incorporation of IL-2 receptor signaling to activate the antigen-dependent JAK/STAT pathway. [67,68]This signaling not only maintains CAR-T cell activity and promotes memory T cell formation but also reactivates and stimulates the immune system as a whole. Modification of T cells to express CAR generally involves transducing the cells with viral vectors carrying the transgene, resulting in semi-random DNA integration into the T cell genome. However, some fifth-generation strategies use specific receptors.

Site-specific integrations providing additional features can be achieved through CRISPR-mediated editing. An example is the insertion of the CAR into the TRAC locus (the T-cell receptor alpha constant region). The TRAC locus is a constant region of the T-cell receptor (TCR) alpha chain gene. This gene editing suppresses endogenous TCR expression to ensure specific antigen recognition and avoid potential interference from endogenous TCRs. By integrating TRAC, fifth-generation CAR-T cells maintain greater stability and identity over time, improving their ability to recognize and eliminate cancer cells. This genetic modification reduces the risk of T cell exhaustion and graft-versus-host effect. It improves the overall efficacy of CAR-T therapies, offering a more durable and potent treatment option.[69] Consistent with this, another innovative strategy has integrated the CAR cassette into the PDCD1 gene locus, demonstrating superior ability to eradicate cancer cells both in vitro and in xenograft models. [70]

Currently, six second-generation CAR-T cell constructs have been approved. Axicabtagene ciloleucel (Yescarta®) and brexucabtagene autoleucel (Tecartus®) are based on CD28, [71]While the other approved constructs are based on 4-1BB, most approved products employ a murine scFv, except ciltacabtagene autoleucel (Carvykti®), which uses a camelid-binding domain. [72] As explained in this section, to further improve the efficacy of CAR-T cell therapy, distinct CAR components have been designed, resulting in constructs with enhanced properties. [55,67,73]Therefore, in CAR-T terminology, the binding domain is not a "tumor antigen receptor". Still, it is the antigen recognition domain (ARD), typically a scFv or single-domain antibody (VHH), that is part of a larger synthetic receptor (the CAR). "Receptor" = signaling complex; the scFv/VHH alone cannot signal. By immunological convention, a receptor is the complete transmembrane signaling unit. In CARs, this unit includes the ectodomain (scFv/VHH + hinge/spacer), the transmembrane segment, and intracellular signaling (CD3ζ ± co-stimulatory molecules such as CD28 or 4-1BB). The isolated binding domain lacks the transmembrane/signaling architecture; therefore, it is not a receptor in itself. We refer to it as the CAR antigen recognition domain (scFv/VHH).[74] It is important to clarify terminology when referring to modified TCR therapies, as "tumor antigen receptor" is often interpreted (incorrectly) as a TCR specific to a tumor peptide-MHC complex. In the case of CARs, conversely, they use antibody-mediated binding and do not require MHC; confusing the ARD with a "tumor antigen receptor" risks conflating CAR approaches with TCR-T in texts and figures.[75] On the other hand, while the biology of many CAR targets is associated with tumors, it is not specific to them (CD19, BCMA, GD2, etc.). It is often found on some healthy cells, denominating the ARD as "tumor antigen receptor" erroneously implies exclusive tumor recognition. Clinically, this imprecision is necessary because on-target, off-tumor effects are expected (e.g., B-cell aplasia with CD19 CAR-T).[76]

While scFvs (human/mouse) and camelid VHHs (nanobodies) are common, CARs can also employ natural receptor or ligand domains (e.g., NKG2D ectodomain, based on cytokines/ligands) or adapter/switch platforms (anti-FITC/biotin "universal" CARs).[77] Affinity/format and signaling thresholds are decoupled, as the properties of the binding domain (affinity/epitope; scFv vs. VHH) influence synapse geometry and antigen density thresholds, but signal intensity also depends on the hinge/TM/costim region design, emphasizing again that the "receptor" is the complete construct, not just the ligand.[78] Figure 3 presents a schematic representation of the four generations of CAR-T structures.

3.2. Manufacturing of CAR-T Cells: Ex Vivo vs. In Vivo Methods, in a Brief Description

The ex vivo CAR-T cell manufacturing procedure remains constant despite various T-cell genetic modifications and requires patient participation. Patients eligible for CAR-T cell therapy typically undergo the following treatment process, as summarized below.(83)[81]

Lentiviral transduction of T cells to generate CAR T cells is an ex vivo gene-delivery process in which a CAR transgene is introduced into patient or donor T cells using integrating lentiviral vectors, enabling the cells to express a synthetic receptor that redirects their antigen specificity stably.

In this process, T cells are collected and placed in an ex vivo culture environment where they are rendered receptive to genetic modification, and a lentiviral vector carrying a chimeric antigen receptor (CAR) transgene is used to deliver the genetic information; because lentiviral vectors are integration-competent, the CAR sequence becomes stably incorporated into the T cell genome, enabling durable transcription and translation of the engineered receptor. The CAR construct typically encodes an extracellular antigen-binding domain fused to transmembrane and intracellular signaling modules, so that once expressed on the T cell surface, the receptor couples antigen recognition to activation and effector functions. After gene delivery, modified cells are maintained under conditions that support survival and selective outgrowth of transduced cells, and the resulting population is characterized by CAR expression, phenotype, and functional attributes to ensure potency and safety before clinical use.[82,83] Key conceptual advantages of lentiviral delivery include efficient gene transfer into nondividing lymphocytes and long-term transgene expression due to genomic integration, while considerations for clinical translation emphasize vector design, transgene expression control, and rigorous quality-control testing to minimize risks such as insertional mutagenesis or off-target activity. The final product is an autologous or allogeneic CAR-expressing T cell population that can be administered to the patient to mediate antigen-specific cytotoxicity against target cells expressing the cognate tumor antigen.[84]

With the rapid development and refinement of new gene therapy techniques, such as CRISPR-Cas9 and CARs, the safe and effective administration of these technologies is fundamental. With corroboration of improvements in both LVVs and other retroviral vectors, the most comprehensive safety study to date analyzed results from 17 lots of clinical vectors, 375 manufactured T-cell products, and 308 patients treated by infusion to evaluate RCL and RCL/R, as well as integration-directed expansion. [85]This analysis supports the safety profile of these vectors in this application and their continued use in oncology, infectious diseases, autoimmunity, and inherited genetic disorders, and encourages their adoption in other therapeutic fields. Vectors used in ex vivo CAR-T cell manufacturing also differ from those employed in vivo gene transfer. Workflows separate vector processing from cell processing, with validated cleaning. Additional controls apply to myeloid products derived from HSC, TCR T, Tregs, NK, and HSPC, as described..[86]

Table 5 below shows a comparison of vectors performance in aspects like transgene delivery, cells that can be transformed, transduction & most common editing enhancer, integration pattern, and others aspects for the most used vectors in the ex vivo process of CAR-T cells manufacturing: LVV, ɣ-retrovirus, and adeno-associated viral vectors, and Table 6 shows the advantages, disadvantages and examples for each class of vectors used for the CAR T manufacturing.

It is because ex vivo LVVs are optimized for controlled stable T cell engineering, minimal immunogenicity, and GMP compliance. In contrast, in vivo gene transfer vectors are being optimized for tissue-targeted delivery, immune evasion, and regulated expression, but they pose greater challenges for biodistribution and safety. New in vivo CAR-T technologies aim to combine the advantages of both approaches but remain in early clinical phases. If comparing LVVs for ex vivo and in vivo CAR-T manufacturing strategies, two strategies must be considered: ex vivo CAR-T manufacturing (e.g., patient T cells transduced outside the body) versus in vivo gene transfer (e.g., vector administered directly to patient tissues or immune cells).[87]

Figure 4 compares ex vivo gene transfer to generate CAR-T cells with in vivo gene therapy approaches. [88]

Although autologous CAR-T cells have demonstrated remarkable clinical results, significantly transforming the treatment of leukemias, problems persist in ensuring patients receive CAR-T cell therapy. In addition to the efficacy and safety issues of CAR-T therapy mentioned in previous sections, the high cost, complex process, and long waiting time of approximately three weeks required for personalized T-cell manufacturing hinder patient access to the development of universal allogeneic CAR-T cells (also known as "off-the-shelf" CAR-T cells) and other CARs employing alternative effector cells.[89]

Recently, various engineering approaches and alternative sourcing strategies have been developed to generate CAR-modified cells. Given the success of CAR-T therapy in oncology, the development of additional strategies based on new technologies and greater ease of use is justified to reduce costs and increase accessibility. Continuous advances in expanding cell sources and engineering techniques have revealed new ways to improve CAR immune cell supply and simplify CAR product manufacturing.[89]

Diverse sourcing strategies, encompassing autologous, donor-derived, third-party, and off-the-shelf cell products, have highlighted a broad spectrum of cell reserves with distinct attributes and potential. However, any basal transcription of the provirus could express this heterologous protein, which is potentially immunogenic. The function of the ISRE sequence in some LVVs is unclear. Understanding whether ISRE-mediated expression can occur during cell activation could be relevant in contexts requiring strict regulation. Understanding how these differences in the 5' UTR region affect packaging would be essential to determine the optimal packaging signal.[89]

These strategies have expanded the repertoire of therapeutic candidates and addressed limitations associated with cell availability and functionality. Furthermore, ingenious engineering techniques have driven the optimization of CAR-based immunotherapies. Techniques such as genomic editing, synthetic biology, and multigenic integration have enabled the customization of immune cells with greater persistence, specificity, and improved safety profiles. In parallel, advances in modular CAR designs, the incorporation of co-stimulatory domains, and switchable CAR systems have optimized the therapeutic response and mitigated adverse effects, underscoring remarkable progress in refining CAR-modified immune cells.[90] Therefore, the ability of LVVs to generate CAR-T cells in vivo has attracted significant interest, as it could eliminate the need to isolate and activate T cells ex vivo, thereby reducing both costs and in vitro production time.[71,91,92]

4. Landscape of Applications Where LVVs and Other Plasmids have Been Used

4.1. Late-Stage Approved Ex Vivo CAR-T Products

As of July 2025, 11 CAR-T structure products had been registered as approved by the three major regulatory bodies. Following the initial approval by the United States Food and Drug Administration (FDA) of the first two CAR-T therapies, Kymriah™ (Tisagenlecleucel) for B-cell Acute Lymphoblastic Leukemia (B-ALL), and diffuse large B-cell lymphoma (DLBCL) on August 30, 2017[93,94,95,96,97,98] and Yescarta™ (Axicabtagene ciloleucel) for the treatment of acute lymphoblastic leukemia (ALL) on October 18, 2017,[69,99,100,101,102,103,104,105] the approval of Terakus™ (Brexucabtagene autoleucel) for the treatment of adult patients with relapsed or refractory mantle cell lymphoma (R/R MCL) occurred on July 24, 2020,[106,107,108,109,110,111,112]These three products were also approved by the European Medicines Agency (EMA) on August 27, 2017, August 23 of the same year, and December 14, 2020, respectively. Next, the FDA approved Breyanzi® (Lisocabtagene maraleucel) for the treatment of adult patients with large B-cell lymphoma (LBCL) on February 5, 2021.[113,114,115,116,117]and Abecma® (Idecabtagene vicleucel) on March 26, 2021, for the treatment of adult patients with relapsed or refractory multiple myeloma (R/R MM) after four or more prior lines of treatment [5,118,119,120,121]and more recently, on February 28, 2022, gave the green light to Carvykti^TM (Ciltacabtagene autoleucel) for relapsed and R/R multiple myeloma in February 2022,[44,72,122,123,124,125,126,127]The latest initial FDA approval of a CAR-T product was obtained by Aucatzyl (Obecabtagene autoleucel) for patients with refractory and recurrent B-precursor acute lymphoblastic leukemia (R/R B-ALL).[128,129]

On the other hand, the National Medical Products Administration of China (NMPA) began by approving Yikaida® (Axicabtagene ciloleucel) on June 22, 2021, for CD19+ lymphoma, followed by the registration of Beinuoda® (Relmacabtagene autoleucel), also known internationally as Carteyva®, on September 6, 2021, with a CAR structure identical to the FDA's liso-cel, which has demonstrated high response rates and lower toxicity associated with CAR-T therapy in relapsed/refractory large B-cell lymphoma in the United States. Furthermore, the pivotal RELIANCE study confirmed the long-term efficacy of this product in relapsed/refractory large B-cell lymphoma. Thanks to its high target-elimination efficacy, long-term persistence, and low toxicity, Carteyva® became the second CAR-T product approved in China for three indications: relapsed/refractory large B-cell lymphoma (LBCL), relapsed/refractory follicular lymphoma (r/r), and relapsed/refractory mantle cell lymphoma (r/r). [129,130,131,132,133]

The NMPA also approved Fucaso® (Equacabtagene autoleucel) on June 30, 2023, as a third-line treatment for (R/R MM). [134,135,136,137]and Yuanruida or Yorwida® (Inaticabtagene autoleucel) on November 8, 2023, as therapy for adult R/R B-cell acute lymphoblastic leukemia (ALL),[138]and then on November 28, 2025, for adult patients with relapsed or refractory large B-cell lymphoma (r/r LBCL) after two or more prior systemic therapies, and the registration of Saikaize® or Zevor-cel (Zevorcabtagene autoleucel) was also published on February 23, 2024[139,140]also as 3rd line therapy for R/R MM. It should not be forgotten that the Chinese regulatory authority also approved Yescarta^TM since June 23, 2021, and Carvykti^TM since August 28, 2024.

As shown in Table 7, most CAR-T therapies approved in the US (FDA), the European Medicines Agency (EMA), and China (National Medical Products Administration [NMPA]) use LVV as a gene delivery strategy, indicating its widespread use. [140,141]

This table demonstrates that clinical use of CAR technology has progressed more rapidly in recent years. CAR-T immunotherapy has demonstrated impressive clinical efficacy in relapsed and refractory (r/r) hematological malignancies, including CD19+ leukemia and lymphoma, and BCMA+ multiple myeloma. Motivated by these achievements, researchers have expanded CAR technology, including applications such as CAR-NK, CAR-CIK, and CAR-MΦ, and have employed cell therapy with modified CAR cells to treat a broader spectrum of aggressive diseases. [3,141,145]

4.2. Other Cell Therapy Products Obtained by Using HIV-Derived LVVs

The field of personalized medicine, particularly gene therapy, is undergoing an exciting and auspicious period. New therapeutic strategies under development offer a glimpse of the horizon of treatments for genetic diseases, far more tangible than what was visible just a few years ago. However, the path to developing gene therapy strategies has not been simple or free of complications and failures. Nevertheless, recent advances, particularly the emergence and development of CRISPR-Cas9, place gene therapy in an auspicious position. Gene therapy has ceased to be a discipline associated with science fiction, and, with some products on the market and many others in various phases of clinical trials, it is realistic to expect new therapies to become available.

Among the cell therapies obtained with the use of LVV for the genetic manipulation of other cells without the CAR structure are the following examples:

4.2.1. Approved Therapies with LVV-Modified Hematopoietic Stem Cells (HSC)

Several FDA-approved products meet this condition, such as Zynteglo (betibeglogene autotemcel), whose regulatory action was approved in 2022.[146] It was also published on December 8, 2023, about the product Lyfgenia (lovotibeglogene marcelpivlecel) from Bluebird Bio. [147]and the product Skysona (elivaldogene autotemcel), which received successive approvals between 2022 and 2025[148] and, finally, about Lenmeldy (atidarsagene autotemcel) on March 18, 2024.[149]

4.2.2. T lymphocytes Whose TCR has Been Modified by LVV Transduction

A phase 1 clinical trial reported on patients treated with T cells whose TCR was modified to recognize the NY-ESO_1 antigen (TAEST16001), consisting of autologous T cells transduced with LVV. [150,151]On the other hand, Sangano Biosciences reported that it obtained a product, TX200-TR101, which used a third-generation LVV-SIN in a first-in-human clinical trial. [152]

4.2.3. CAR-NK Platforms Obtained via NK Cells Transduced with LVV

A report about the efficient and robust transduction of NK cells with LVV pseudotyped with BaEV was done in 2019, [153]In 2024, another group reported production protocols for high-titer BaEV-R-free LVVs as an advanced manufacturing approach.[154]

4.2.4. HSC-Derived Cellular Drug Factories

Refers to a clinical trial reported in ClinicalTrials.gov with the bioproduct Temferon^TM (NCT03866109): based on autologous HSPCs transduced with an LVV encoding IFN-α2 and echoed by the publication of Mazzoleni S. et al. [155] Which describes the Temferon platform of the company Genenta.com. This same Temferon-modified HSPC product (LVV) gives rise to myeloid cells with tumor tropism that release IFN-2alpha, according to a study summarized at ASH. [156,157] Table 8 below summarizes the mentioned products.

5. Limitations and Risks Associated with the Use of LVVs

5.1. Risk is Associated with Randomly Inserting into the Cell Genome and Insertional Oncogenesis

While LVVs are primarily used as research tools to introduce gene products into in vitro systems or animal models, their applications in gene therapy development are significant, particularly for adoptive cell transfer therapies. These include CAR-T, CAR-NK, TCR-T, and TIL therapies, which constitute the primary use of LVVs in this field.[155] Appreciating how these LVVs have been developed and how they generate new opportunities in regenerative medicine and tissue engineering, one must not overlook that limitations persist in their use for CAR-T therapies.[158] Among other challenges is that, during CAR gene integration into T cells, LVVs often integrate randomly into the cell genome, which can cause adverse effects on the host genome. It can lead to unwanted gene silencing, overexpression, or genetic mutations, increasing safety risks.[159] Furthermore, the limited transcriptional capacity of LVVs restricts the size and complexity of CAR gene payloads, as well as the incorporation of associated regulatory elements. Finally, large-scale production of LVVs for clinical applications requires GMP-certified laboratory and manufacturing reagents and a biosafety level 2+ or 3 (BSL2+ or BSL3), entailing practically prohibitive manufacturing costs and regulatory hurdles. [92,160,161]

On the other hand, these vectors can also alter regular cell gene expression programs and increase cancer risk by stably integrating into the genome. Serious adverse events have been reported with the use of both gamma and safer self-inactivating LVV (SIN-LVs).[162,163,164] Intensive research, combined with field experience, revealed that the main determinants of insertional oncogenesis are the vector class (γ-retroviral, LVV, or α-retroviral) and the chosen internal promoter in SIN-LV configurations. [165,166,167]It has also been shown that synergistic interactions between these aspects and the specific transgene contribute to mutagenicity.[167] These insights have led to significant improvements in vector design, which have been crucial to the field's progress. Yet, there is still an unmet need for sensitive and reliable preclinical genotoxicity assays.[88,168]

But it should not be forgotten that, by the end of 2023, the FDA had received 22 reports of T-cell malignancies linked to CAR-T treatments.(87) [88]Among them, three cases in which sequencing was performed detected the CAR gene in the malignant clone,(87) [88]suggesting a possible connection between malignancies and the integration properties of retroviral vectors. In 2024, several reports described rare secondary T-cell lymphomas following CAR-T therapy, which can be attributed to multiple factors, including lentiviral insertions within T-cell homeostasis genes, occasional progression of CAR-T cells themselves to T-cell lymphoma, and genetic mutations associated with clonal hematopoiesis.[169,170,171,172,173]

Therefore, other methods for in vitro T cell genetic modification, considered non-viral as they do not depend on this type of vector, have already been tested. But these methods, including electroporation via the transposon system, prove ineffective in achieving stable CAR gene expression, although the produced CAR-T cells exhibit faster cytotoxicity in vitro. This inefficiency highlights the need for safer and cheaper virus-free gene transfer vectors, typically represented by transposon systems, CRISPR/Cas9 systems, and mRNA electroporation platforms.[32,36,43] Further research is required to thoroughly elucidate the long-term safety and efficacy of these advanced therapies.

The integration of CRISPR/Cas-based editing techniques into these approaches provides unmatched precision in genetic engineering, enabling genes to be inserted or disrupted. Furthermore, various processing techniques facilitate the automated manufacturing of non-viral CAR-T cells, and the importance of implementing measures to ensure the safety and genetic stability of these therapies was emphasized. These measures include the use of sequencing technologies to identify unwanted cut sites and computational methods to deconvolve translocations. Given the growing importance of non-viral transgene administration and CRISPR/Cas-based editing in gene therapy, these technologies must continue to evolve and undergo rigorous testing to ensure their safe and practical application in clinical practice.[89]

5.2. Biological Risk Assessment When Working with LVV

Viral vectors have generally proven to be efficient tools for gene delivery to target cells/tissue, a critical aspect of achieving therapeutic efficacy.[174]As already mentioned, viral vector expression cassettes can be engineered with elements to enhance target specificity and increase transgene expression, [175] These features have been explored and exploited to develop efficient methods for delivering genes of interest into mammalian cells. However, a vast repertoire of viral vectors exists with profoundly different properties. To date, five main classes of viral vectors have been tested for clinical applications, [176] including, apart from LVV, other retroviruses (RV), adenoviruses (AV), AAV, and herpes simplex viruses (HSV). Although many viral vectors are safe and effective delivery vehicles for clinical gene therapy, some are considered risky due to potential oncogenicity. [3,29] These features have been explored and exploited to develop efficient methods for delivering genes of interest into mammalian cells. However, a vast repertoire of viral vectors exists with profoundly different properties. To date, five main classes of viral vectors have been tested for clinical applications,[177] They may pose risks, such as replication-competent retroviruses (RCR) and insertional mutagenesis, which can activate oncogenes or cause secondary malignancies, especially with 1st- and 2nd-generation LVVs.[178]

Despite these safety improvements, working with LVVs requires strict compliance with biosafety protocols to prevent laboratory-acquired infections and ensure patient safety in clinical applications. Risks associated with HIV-1-derived LVVs include the potential to generate RCVs that could make them pathogens and the potential for oncogenesis. These risks can be mitigated according to the nature of the vector system (and its safety characteristics) or those exacerbated by the nature of the transgene insert encoded by the vector.[160] General criteria for LVV risk assessment must take into account a series of parameters/considerations, including: the nature of the vector system and the potential for regeneration of RCV from vector components; the nature of the transgene insert (e.g., known oncogenes or genes with high oncogenic potential may require special precautions); the vector titer and the total amount of the vector, the inherent biological containment of the animal host, if relevant; a negative RCL testing. Modern clinical CAR-T LVVs are designed so that the probability of generating an RCL is extremely low, although not mathematically zero. Therefore, design, process controls, and batch release testing collectively reduce residual risk to the level "as low as reasonably achievable (ALARA)". [179]Since the risk is not literally zero, regulatory bodies require RCL control strategies and testing.[160,179,180]

In particular, in those LVVs where that of VSV-G has replaced the wild-type HIV-1 envelope, this modification increases the number of potential cells and tissues the particle can penetrate, and its mode of transmission could include direct exposure to infected body fluids, sexual contact, splashes, or percutaneous injection.[24] Signs and symptoms associated with this type of infection, often accidental, are fever, nasal discharge, sore throat, diarrhea, cough, vomiting, nausea, lethargy, myalgia, anorexia, and dyspnea, but this depends on infectious dose, which in turn would be determined by the source of infection, and taking into account the incubation period which can be 1 to 6 months, surveillance and reporting of accidental exposure must be recorded and followed, as biosafety practice.[181] This step is necessary, as prophylaxis exists and must be performed after documented exposure to any HIV-1-derived LVV, including the use of antiretroviral drugs, as no vaccines are available for these LVVs. [182]Surveillance of affected individuals should be conducted by monitoring for influenza-like symptoms and confirming diagnoses with RT-PCR testing and initial rapid tests.[182]These concerns have motivated more than one regulatory organization, including the Department of Health of Abu Dhabi (DOH), to mandate rigorous containment measures, thorough risk assessment, and comprehensive personnel training to ensure safe handling practices in both laboratory and clinical settings.[182,183,184]

Biosafety practices for HIV-derived LVVs have evolved as viral vector generation technology and understanding of potential risks have improved. Current standards emphasize evidence-based approaches that focus on actual risks rather than predetermined prescriptions, thereby enabling more flexible and effective safety protocols. International harmonization of biosafety guidelines facilitates global collaboration in the research and development of LVV therapies. As these technologies are increasingly incorporated into clinical applications, integrating GMP standards with traditional biosafety practices creates a comprehensive framework to ensure safety from the laboratory to the patient's bedside.[185]

Among the new challenges in LVV biosafety is the growing use of CRISPR/Cas9 systems delivered via LVV, which demand greater containment due to their genotoxic potential. The increase in production for clinical applications also raises new biosafety considerations distinct from those for research-scale work. It is essential to ensure that future developments incorporate increasingly effective vector systems with even higher safety profiles, such as targeted integration systems that minimize the risk of insertional mutagenesis.[181]

5.3. Laboratory Containment Level

Given that risk group 2 classification applies to LVVs, containment level 2 facilities, equipment, and operational practices must be used when working with these vectors; it is recommended that work not be performed on open benchtops with this type of viral particle. [181,185]All work must be performed inside a biosafety cabinet. The use of sharps with needle safety features is also recommended. Centrifuge rotors must have lids; samples must be loaded and unloaded inside the biosafety cabinet, and the centrifuge must be disinfected with a suitable disinfectant after use. If the vector is not replication-competent, animals infected with the vector will remain in BSL-2 for 72 hours, after which they will be moved to BSL-1. [181,185]If the vector is replication-competent, animals will be kept in BSL-2 for the entire duration of the experiment.[185] Other authors recommend using Biosafety Level 2+ (BSL-2+) containment. [186,187] The BSL-2+ level is the term commonly used for laboratories working with microorganisms that employ biosafety practices and procedures proper to BSL-3 laboratories. Most research institutions still struggle to determine when to use this approach and what BSL-3 practices to utilize, as BSL-2+ is not a recognized containment level. There is no standardized list of microorganisms, viral vectors, or research projects that must be conducted in BSL-2+ environments; however, it is assumed that, when generating new vectors whose potential is still under experimental evaluation, this would be appropriate practice.[187] Each decision to use specific BSL-3 practices in a BSL-2 laboratory must be based on a risk assessment. [181,185]Some examples of when BSL-2+ use may be appropriate include viral vectors with gene inserts containing oncogenes or genes of unknown function, or 1st- and 2nd-generation LVVs that present a higher risk of recombination, which can lead to RCL. [160]

Risk assessment includes several steps before work can be approved and started: the laboratory supervisor or Principal Investigator (PI) must complete and submit the project registration document to the Biosafety Officer (BSO). It is a document that must clearly describe the project's purpose and detail the steps for handling biohazardous materials. It is reviewed and discussed with the laboratory supervisor or PI, the BSO, and, in some cases, selected members of the Institutional Biosafety Committee (IBC) with experience in BSL-2+ research. [185]For example, a virologist may be asked to review a project involving viral vectors. Suitable BSL-2 laboratory space must be proposed, and BSL-3 practices described. IBC review and consensus must occur before project initiation. At the IBC meeting, the BSO describes the proposed project and the BSL-3 practices to be used in the BSL-2 laboratory space. IBC members must agree on appropriate Biosafety Level 3 (BSL-3) practices for the proposed work and determine a suitable Biosafety Level 2 (BSL-2) laboratory space. Risk communication and training must occur after IBC approval and before any laboratory work is performed. The Biomedical Safety Officer (BSO) must review required BSL-3 procedures with the laboratory supervisor or principal investigator (PI) and their staff. Ideally, these procedures should be documented in a Standard Operating Procedure (SOP). Additionally, it is essential to review the laboratory space to ensure required BSL-2 elements are present, including, among others, biological waste containers, a sink with soap and paper towels, and certified biosafety cabinets (BSCs).[182]

The laboratory supervisor or principal investigator must designate a laboratory meeting facility that meets the requirements outlined in U.S. CDC/NIH publications on biosafety in microbiological and biomedical laboratories.[182]It must be an inner laboratory with two doors separating it from the biosafety cabinet (BSC) and the corridor. Airflow must be from the corridor into this laboratory (i.e., in a negative direction). All air must be exhausted to the outside of the building; it must not be recirculated. An Environmental Health and Safety (EHS) officer can assess the laboratory's negative-pressure status.[187]

5.4. Considerations on Personal Protective Measures

The following safety equipment must be used when working with LVVs: at a minimum, personnel must wear gloves, closed-toe shoes, a lab coat, and appropriate eye and face protection before working with lentiviruses or their derived vectors.

Additional protective equipment may be required according to laboratory-specific standard operating procedures (SOPs). And utmost care must be taken to avoid spills and/or splashes of infected materials. It must be presumed that LVV is present on all equipment and devices coming into direct contact with infected materials. While decontamination within the laboratory is preferable, removing materials from the decontamination facility is also an option. All materials coming into contact with LVV must be disinfected with a 1:10 bleach solution before disposal. Additionally, all work surfaces must be disinfected with a 1:10 bleach solution at the end of work and at the end of the workday. (Note: A contact time of at least 15 minutes is required for decontamination. [188]

When working with LVV, the following personal protective equipment must be worn: gloves (consider double gloving depending on procedures being performed), lab coat, safety glasses, and face shield. Waste management is also highly regulated. Examples: Non-sharp waste: All cultures, stocks, and cell culture materials must be disinfected and autoclaved before disposal using double-walled biohazard waste containers. Sharp waste, including needles, syringes, razor blades, scalpels, Pasteur pipettes, and tips, must be disposed of in an approved puncture-resistant sharps container. Sharps containers must not be filled to more than 2/3 of their capacity.[188]

5.5. Response and Decontamination Procedures for Spills, and in Cases of Accidental Exposure

In the event of a minor spill containing LVV (<1 liter) within a BSC, the panel must be closed and the cabinet allowed to operate for 15 minutes before proceeding with spill cleanup. Lentiviruses, such as HIV-1, are not commonly transmitted through aerosols, but extra precautions must be taken during a spill outside a BSC to reduce exposure risk. Leave the room immediately and allow aerosols to dissipate for 15 minutes. Notify other people working in the laboratory. Put on appropriate personal protective equipment.

Cover the spill area with paper towels or another absorbent material, and apply an EPA-registered disinfectant effective against the viruses involved (bleach, glutaraldehyde, hypochlorite, iodine, and phenols), working from the perimeter toward the center. Decontamination protocols for LVV require the use of a suitable disinfectant.[188]

An immediate response to potential LVV exposures is crucial to minimize risk. In case of exposure, personnel must immediately: (1) Rinse mucous membranes (eyes, nose, and mouth) with plenty of water for 15 minutes at an eyewash station; (2) Wash exposed skin with soap and water for at least 15 minutes (5 minutes if skin is intact); And (3) In case of puncture wounds or cuts, encourage bleeding by gently pressing the area while washing with soap and water. After providing first aid, the affected person must seek medical attention immediately, even if the exposure seems mild. Healthcare professionals must be informed of the specific nature of the LVV, including the generation method, the expressed transgene, and any other relevant details. [187]

In case of significant exposures to HIV-based LVVs, post-exposure prophylaxis (PEP) with antiretroviral drugs may be considered. Some guidelines recommend that physicians consider initiating a 7-day course of a nucleoside reverse transcriptase inhibitor (NRTI), as soon as possible after exposure (within 72 hours). The need for PEP must be determined by a medical assessment based on the type and magnitude of exposure. All exposure incidents must be reported immediately to the supervisor and the institution's biosafety office, and appropriate incident report forms must be completed, with follow-up required by institutional policies. [187,189]

5.6. Other Facility Infrastructure Requirements

Optimizing workflows, improving process robustness, and adopting closed, automated systems are recommended to facilitate scalability and reduce product costs while maintaining the efficiency of CAR-T cells or other products. To achieve commercial-scale production of CAR-T cell therapies, comprehensive automation is desirable, including fully automated and partially automated closed systems. [190]Fully automated closed systems can be used in less strict cleanrooms. However, the final product stage must be performed in an ISO 7 cleanroom (Class 10,000, Clean C). Cleanroom classification systems are based on the maximum number of suspended particles per cubic meter of air, with defined particle size ranges (≥0.5 μm and ≥5 μm). It applies according to ISO classification, defined by ISO 14644-1 and ISO 14644-2 standards, and/or according to Good Manufacturing Practice (GMP) classification, characterized by Annex 1 of the European Union (EU) GMP and corresponding FDA regulations, with grades A, B, C, and D. [158,190,191]

To maintain classification, cleanrooms must have: HEPA filtration (typically H14 or higher for Grade A), unidirectional (laminar) airflow in critical Grade A zones, positive pressure differentials (≥10–15 Pa between cleanroom classes), continuous particle count monitoring in Grade A, periodic particle count monitoring in lower grades, microbiological monitoring: settle plates, contact plates, air samplers, glove prints. [158]

For cell isolation involving open manipulation, steps must be performed in an ISO 5 (Class 100) BSC. The environment surrounding BSCs must remain sterile during cell therapy handling and manufacturing processes, in compliance with Good Manufacturing Practices (GMP). The environment of ISO 5 BSCs (formerly Class 100) must be controlled and classified as an ISO 7 cleanroom (Class 10,000, Clean C). ISO 5 BSCs (formerly Class 100) must comply with ISO 5, which includes, at a minimum, HEPA-filtered laminar airflow over the work area and a UV sterilization lamp.[189]

Table 9 serves as a guide for classifications according to ISO standards, EU GMP grades, and FDA cleanroom classes. [176,177,178,179]

5.7. Design of Outpatient Rooms for CAR-T Therapy Application in Human Subjects

There must be a designated area for outpatient care that protects patients from exposure to infectious agents, allows for isolation, and supports confidential examination and assessment, as well as the administration of intravenous fluids, medications, and blood products. For procedures performed in an outpatient setting, a designated area with appropriate space and design to minimize the risk of microbial contamination must be available. Immediate assessment and treatment mechanisms must be in place, staffed by an on-call physician available 24 hours a day, 7 days a week. [192]

6. Conclusions

As described in this review, several CAR-T cell therapies have received FDA, EMA, and NMPA approvals as second- and third-line treatments, and ongoing clinical trials are evaluating their efficacy as first-line therapies. Immune cell therapies have transformed the landscape of cancer and other disease treatment, offering unprecedented clinical results. The global T-cell therapy market is projected to grow rapidly, increasing from USD 10.30 billion in 2025 to USD 161.21 billion by 2034, with a compound annual growth rate of 35.74%. [193]

Current engineering approaches focus on developing the next generation of CAR-T cells to improve potency and safety. [194,195,196]including strategies aimed at enhancing CAR-T cell fitness (e.g., signaling, killing capacity), overcoming antigen loss (e.g., via dual-targeting CARs), and modifying the tumor microenvironment (e.g., interleukin-18 secreting CARs) [194,195,196]including strategies aimed at enhancing CAR-T cell fitness (e.g., signaling, killing capacity), overcoming antigen loss (e.g., via dual-targeting CARs), and modifying the tumor microenvironment (e.g., interleukin-18 secreting CARs) [194,196,197,198]particularly as progress is made in evaluating the efficacy of CAR-T cell therapy in solid tumors.[198]

The development of standardized protocols for obtaining, engineering, and characterizing CAR-modified immune cells will be fundamental to ensuring reproducibility and facilitating regulatory approval. [91]The field of non-viral transgene delivery to T cells has experienced significant advances in recent years, with several platforms already used in clinical trials or showing promise for future clinical application. These platforms offer a substantial advantage over viral vectors, as they are more cost-effective, thereby accelerating the development of innovative gene therapies.

Continuous personnel training remains an essential component of an effective biosafety program, as human factors remain the most crucial variable in preventing laboratory accidents and exposures. By rigorously applying current guidelines and adapting to new challenges, the scientific community can continue to harness the potential of LVV technology while maintaining the highest safety standards for both laboratory personnel and patients. Intensifying collaborative efforts between medical centers, the biotechnology industry, pharmaceutical companies, and nanotechnology institutes is poised to accelerate the translation of these cutting-edge approaches into transformative clinical interventions, such as in vivo generation of CAR-T cells, which could reshape the landscape of cancer and other immune-related diseases. As we explore these frontiers, a promising era of precision immunotherapy emerges, offering personalized, potent, and durable treatments for patients in need. This potential motivates us to continue to expand the boundaries of scientific research.