Gene Therapy

Gene therapy is the therapeutic modification of genetic material or gene expression in a person’s cells. It can introduce a functional gene, suppress a harmful gene’s activity, edit a DNA sequence, or give cells a new therapeutic function. Treatment may be delivered directly into the body or performed on cells outside the body before they are administered to the patient. Genome editing and genetically engineered cell therapies therefore overlap with gene therapy, but neither term is synonymous with it. (fda.gov)

Gene therapy provides ways to address the molecular causes of some inherited disorders and to engineer cells to act against cancer. Approved applications include treatments for selected rare diseases and genetically modified immune-cell therapies. Its development brings together genetics, vector engineering, transplantation medicine, and clinical research; effectiveness depends on reaching appropriate cells and achieving a sufficient therapeutic effect. Safety assessment, manufacturing quality, and long-term observation are central because some interventions produce persistent biological changes. (fda.gov)

Definition, Scope, and Terminology

Gene therapy is the therapeutic use of transferred genetic material or targeted genetic modification to change gene expression or the biological properties of living cells. Its scope includes supplying a functional gene, suppressing harmful gene activity, modifying an existing sequence, and introducing a new cellular function. The US Food and Drug Administration (FDA), in its January 2020 guidance on long-term follow-up, describes gene therapy products as acting through transcription or translation of transferred genetic material, or through specific alteration of human genetic sequences. This definition encompasses nucleic-acid preparations, genetically modified microorganisms, genome-editing tools, and genetically modified human cells; it is not restricted to treatment of inherited disorders. (fda.gov)

The defining distinction from conventional protein replacement is what supplies the therapeutic activity. In hemophilia, factor replacement administers the missing clotting protein itself, whereas a gene-transfer approach supplies genetic instructions that enable recipient cells to produce a clotting factor. Manufacturing a replacement protein using recombinant DNA does not make its administration gene therapy: the patient receives the protein, not the genetic production system. Likewise, transplantation of unmodified donor blood-forming stem cells is not gene therapy merely because the donor cells contain functional genes. Genetic modification of those cells before administration creates an intervention that overlaps cell therapy and gene therapy. (nhlbi.nih.gov)

“Gene replacement” often denotes functional compensation rather than literal replacement at the original chromosomal location. An additional gene can supply a missing function while the disease-associated allele remains unchanged. Transferred DNA may remain separate from the cell’s chromosomes, so integration is not a defining requirement. Nor must gene therapy permanently alter DNA: the FDA’s examples include laboratory-transcribed RNA, whose translation can provide therapeutic activity without a lasting genomic change. Conversely, transiently delivered editing machinery can produce an enduring DNA alteration. The persistence of the delivery material, the persistence of the genetic change, and the duration of the therapeutic effect are therefore distinct properties. (nhlbi.nih.gov)

Genome editing overlaps with gene therapy when targeted sequence changes serve a therapeutic purpose, but the terms are not interchangeable: gene transfer need not edit the genome, and editing is also used in nontherapeutic research. Therapeutic editing need not correct the disease-causing mutation directly. For example, Casgevy modifies a regulatory region of BCL11A in blood-forming stem and progenitor cells, increasing fetal hemoglobin rather than repairing the sickle-cell mutation itself. CAR T-cell therapy illustrates another overlap: genetically modified T cells acquire an engineered antigen-recognition function. Such cells are both a cellular treatment and a gene-therapy product, even though their purpose is not to replace a defective inherited gene. (fda.gov)

“Genetic medicines” is a broader descriptive category, rather than a universally standardized regulatory class. It can include gene transfer, therapeutic genome editing, and RNA-directed treatments that change how genetic information is used. Some RNA therapies act on existing messenger RNA to reduce protein production or alter RNA processing, rather than supplying an expressed gene or modifying DNA. Consequently, treatment of a genetic disease, use of a nucleic acid, and classification as gene therapy are separate questions; the product’s mechanism and the applicable regulatory definition determine the boundary. (medlineplus.gov)

Jurisdiction-specific terminology further qualifies that boundary. In the European Union, Commission Directive 2009/120/EC defines a gene therapy medicinal product through two linked criteria: an active substance containing or consisting of recombinant nucleic acid intended to regulate, repair, add, or delete a genetic sequence, and an effect directly related to that sequence or its expression product. The legal definition includes therapeutic, prophylactic, and diagnostic effects, but expressly excludes vaccines against infectious diseases. This wording differs from the FDA’s mechanism-based description; scientific discussion of a platform therefore does not, by itself, establish the product’s legal classification in every jurisdiction. (eur-lex.europa.eu)

Historical Development

In November 1973, Stanley Cohen, Annie Chang, Herbert Boyer, and Robert Helling reported the construction of biologically functional recombinant DNA plasmids by joining DNA fragments and introducing the resulting molecules into bacteria. These experiments established methods for assembling and propagating genetic sequences, rather than treating disease. The transition toward therapeutic transfer required delivery into human cells and expression of a functional gene: in 1985, W. French Anderson, R. Michael Blaese, and colleagues at the US National Institutes of Health (NIH) used a retroviral vector to introduce a human adenosine deaminase (ADA) gene into patient-derived cells in culture. (pubmed.ncbi.nlm.nih.gov)

The NIH human gene-marking study began on May 22, 1989, under Steven Rosenberg and colleagues, following institutional, NIH Recombinant DNA Advisory Committee, and Food and Drug Administration (FDA) review. Researchers removed tumor-infiltrating lymphocytes from patients with advanced melanoma, introduced a bacterial neomycin-resistance gene using a retroviral vector, and returned the modified cells to the patients. The added gene served as a detectable marker for tracking the cells’ distribution and survival; it was not intended to correct an inherited defect or enhance their anticancer activity. The initial report described five patients and demonstrated persistence of marked lymphocytes in the circulation and at tumor sites for months. Thus, the experiment established clinically supervised human gene transfer, but the genetic modification itself was not a therapeutic intervention. (irp.nih.gov)

On September 14, 1990, the NIH team initiated an explicitly therapeutic trial for ADA-deficient severe combined immunodeficiency (ADA-SCID). The first participant was four-year-old Ashanthi DeSilva; a second child began treatment on January 31, 1991. Blaese, Anderson, Kenneth Culver, and collaborators collected the children’s own T lymphocytes, cultured them, transferred a functional ADA gene with a retroviral vector, and reinfused the modified cells. This was gene addition, not replacement of the defective sequence at its original chromosomal location. Both children continued polyethylene-glycol-conjugated ADA enzyme replacement, known as PEG-ADA, as a concurrent treatment. Repeated cell infusions continued for approximately two years. The four-year report, published in 1995, documented persistence of the integrated vector and ADA expression, together with improvements in immune measurements; continued PEG-ADA treatment limited attribution of clinical benefit to gene transfer alone. (findingaids.nlm.nih.gov)

A later authorization milestone was Glybera (alipogene tiparvovec), granted European Union marketing authorization on October 25, 2012. Its indication was restricted to adults with genetically confirmed familial lipoprotein lipase deficiency, detectable lipoprotein lipase protein, and severe or recurrent pancreatitis despite dietary fat restriction. The authorization was issued under exceptional circumstances because comprehensive evidence could not be supplied under normal conditions of use. It expired on October 28, 2017, after the manufacturer declined to seek renewal, citing lack of demand. (ema.europa.eu)

In the United States, engineered immune cells and directly administered gene-transfer products reached approval in 2017. On August 30, the FDA approved Kymriah (tisagenlecleucel), its first approved gene therapy, for patients up to age 25 with refractory or second-or-later-relapsed B-cell precursor acute lymphoblastic leukemia. The product consists of patients’ T cells genetically modified to express a CD19-directed chimeric antigen receptor, an application of CAR T-cell therapy. On December 19, the FDA approved Luxturna (voretigene neparvovec) for retinal dystrophy associated with confirmed biallelic RPE65 mutations, marking a distinct milestone for directly administered gene transfer targeting an inherited disorder. (fda.gov)

On December 8, 2023, Casgevy (exagamglogene autotemcel) received its initial US approval for sickle-cell disease in patients aged 12 years or older with recurrent vaso-occlusive crises. It was the first FDA-approved therapy employing CRISPR/Cas9 genome editing. Patients’ own blood-forming stem cells are edited outside the body and returned after conditioning; the modification increases fetal hemoglobin production rather than directly correcting the sickle-cell mutation. The FDA simultaneously approved Lyfgenia, a lentiviral gene-addition treatment for sickle-cell disease, placing genome editing and gene transfer alongside one another as licensed therapeutic approaches. (fda.gov)

Therapeutic Mechanisms and Molecular Targets

Gene addition, or gene augmentation, supplies a functional coding sequence so that treated cells produce a missing, deficient, or therapeutically useful protein. This provides functional complementation: the introduced sequence compensates for inadequate biological activity without necessarily removing the disease-causing allele. It is therefore distinct from replacement at the original genomic locus, which requires a targeted change to the endogenous DNA. Voretigene neparvovec (Luxturna), initially approved in the United States in 2017, illustrates complementation. It delivers a functional RPE65 sequence to retinal cells, enabling production of an enzyme involved in generating the visual pigment component 11-cis-retinal. Its intended action is restoration of enzyme activity, not correction of the patient’s original RPE65 variants. The treatment also requires viable retinal cells: supplying a gene does not recreate cells already lost to degeneration. (fda.gov)

Gene disruption and expression suppression instead reduce an unwanted biological activity. Disruption changes a DNA sequence so that a coding region or regulatory element no longer functions as intended; suppression can act on messenger RNA without changing the corresponding DNA. In RNA interference, short regulatory RNAs direct sequence-dependent degradation or repression of target transcripts. These approaches are particularly relevant when a dominant variant produces a toxic protein or interferes with the normal protein’s function: adding a healthy copy may leave the harmful product present. Selective suppression can target the mutant allele, whereas mutation-independent suppression can reduce both normal and mutant transcripts and therefore require simultaneous functional supplementation. In a 2018 study, Artur Cideciyan and colleagues combined a rhodopsin-targeting short hairpin RNA with an RNA-interference-resistant human RHO sequence in dogs with inherited dominant retinal degeneration. Endogenous rhodopsin RNA was nearly completely suppressed, replacement protein reached up to 30% of normal levels, and treated retinal regions retained structure and function over more than eight months. This was preclinical evidence, not an established human treatment. (nature.com)

The molecular target need not be the gene containing the causal variant. Exagamglogene autotemcel (Casgevy), initially approved in the United States in 2023, uses genome editing to disrupt an erythroid-specific enhancer of BCL11A in hematopoietic stem cells. The modification reduces expression of BCL11A, a repressor of fetal hemoglobin, in their erythroid descendants. Increased γ-globin production raises fetal hemoglobin rather than repairing the disease-causing HBB sequence. In sickle-cell disease, this counteracts hemoglobin-S-associated sickling; in transfusion-dependent β-thalassemia, it helps compensate for deficient β-globin production and the resulting globin-chain imbalance. Targeting a lineage-specific regulatory element links the intervention to the blood-cell population in which the desired activity is required. (fda.gov)

Introduction of a new therapeutic function changes what a cell can produce or recognize, rather than restoring an inherited deficiency. Nadofaragene firadenovec (Adstiladrin), initially approved in the United States in 2022, delivers an interferon-alfa-2b gene to bladder urothelial cells, producing transient local expression of an antitumor cytokine. Engineered antigen recognition provides a different example. In tisagenlecleucel (Kymriah), initially approved in 2017, T cells express a chimeric antigen receptor that combines an antibody-derived CD19-binding domain with intracellular CD3-zeta and 4-1BB signaling domains. Recognition of the CD19 antigen activates the modified cells; CD3-zeta initiates activation and antitumor activity, while 4-1BB supports expansion and persistence. This CAR T-cell therapy redirects immune-cell behavior rather than correcting a cancer-associated mutation. (dailymed.nlm.nih.gov)

Mechanism selection consequently depends on the variant’s effect, the relevant tissue, and the biological activity needed. A deficiency that can be compensated by additional protein supports an augmentation strategy; a harmful dominant product may require suppression as well as supplementation. A compensatory pathway can sometimes bypass the original defect, as fetal hemoglobin induction illustrates. Tissue requirements are equally specific: retinal enzyme complementation depends on surviving local cells, globin regulation must affect erythroid descendants, and engineered tumor recognition requires immune cells capable of executing an effector response. Molecular success is therefore assessed through the intended functional output—not simply the presence of an introduced sequence or an edited DNA site. (nature.com)

In Vivo and Ex Vivo Treatment

In vivo treatment delivers a genetic payload directly to the patient, so that gene transfer or editing occurs within the body. Ex vivo treatment instead collects cells, modifies them outside the body, and administers the resulting cellular product. The distinction concerns where the genetic intervention occurs, not its molecular mechanism: gene addition and genome editing can use either approach. In vivo treatment must deliver its payload to cells within existing tissues; ex vivo treatment allows a selected cell population to be manipulated before administration. The administered product is therefore typically a delivery formulation in the first case and genetically modified living cells in the second. (fda.gov)

For in vivo therapy, local delivery places the product near its intended target, whereas systemic delivery, commonly through intravenous infusion, exposes it to the circulation. Luxturna (voretigene neparvovec), initially approved in the United States in 2017, illustrates local administration: subretinal injection places a vector carrying functional RPE65 near retinal cells in patients with biallelic RPE65-associated retinal dystrophy. Zolgensma (onasemnogene abeparvovec), initially approved in 2019, illustrates systemic administration: an intravenous vector delivers an SMN1 gene to support production of survival motor neuron protein. Systemic exposure is not restricted to the intended therapeutic target; Zolgensma’s prescribing information reports vector DNA in multiple organs in two postmortem examinations, with the highest levels in the liver. Thus, administration route determines initial exposure but does not by itself guarantee tissue specificity. (fda.gov)

Ex vivo treatment usually begins with collection of blood or bone marrow and isolation of the required cells. Hematopoietic stem and progenitor cells provide a source of modified blood-cell lineages, while T cells can be engineered for antigen recognition in CAR T-cell therapy. Processing includes genetic modification, expansion where required, and assessment of the cellular product before administration. Casgevy (exagamglogene autotemcel), initially approved in the United States in 2023, provides a concrete example: a patient’s CD34-positive cells are collected by apheresis, edited outside the body using CRISPR/Cas9, and returned by intravenous infusion. Its treatment pathway also includes collection of unmodified backup cells for possible rescue if engraftment fails. (fda.gov)

Autologous products use the recipient’s own cells, making collection and manufacturing patient-specific. Treatment scheduling must accommodate this interval, and the identity of the collected material must remain linked to the recipient throughout processing and administration. Allogeneic products use donor-derived cells and can potentially be manufactured in advance for multiple recipients. Donor-derived CAR T-cell approaches are being evaluated in clinical trials as an alternative to individually manufactured products, which can take weeks to produce. Their logistics substitute donor collection and advance production for recipient-specific collection, but also introduce immune-compatibility constraints; additional engineering may be used to reduce recognition of the donor cells as foreign. (dailymed.nlm.nih.gov)

Conditioning, when required, is a separate intervention between cell collection and cell administration. In stem-cell treatments such as Casgevy, myeloablative chemotherapy reduces existing marrow cells and permits the modified cells to engraft and restore blood-cell production. CAR T-cell treatment commonly uses lymphodepleting chemotherapy before infusion to support the transferred cells. These regimens have different purposes and are not interchangeable, nor is conditioning a defining requirement of every ex vivo therapy. After administration, stem-cell recipients require assessment of engraftment and blood-count recovery, whereas engineered T-cell recipients are monitored for expansion, activity, and treatment-related effects. The delivered cells, rather than the original laboratory editing or transfer reagent, constitute the therapeutic product entering the patient. (dailymed.nlm.nih.gov)

Viral Delivery Systems

Viral vectors retain a virus’s capacity to enter cells and deliver genetic material while modifying or removing functions unnecessary for treatment. Their tropism—the range of cells and tissues they can transduce—depends partly on interactions between viral surface proteins and cellular receptors. Changing a capsid or envelope can therefore change target-cell access; for example, lentiviral vectors can be “pseudotyped” with vesicular stomatitis virus glycoprotein rather than their original envelope. Tropism is distinct from expression specificity: a vector may enter several cell types while an internal promoter restricts where its therapeutic sequence is expressed. Administration route and tissue exposure also influence the distribution of transduced cells. (ema.europa.eu)

Adeno-associated virus (AAV) vectors generally package single-stranded DNA, with a practical genome-size limit near 4.7 kilobases (kb); terminal repeats and expression-control sequences consume part of that space. Recombinant AAV replaces the viral rep and cap genes with the therapeutic expression cassette, while production systems supply the missing functions separately. Delivered DNA persists predominantly outside the chromosomes, as episomal forms, although integration can occur at low frequency. This supports prolonged expression in relatively long-lived cells but allows vector genomes to be diluted during cell proliferation. Capsid selection affects tissue targeting, while pre-existing neutralizing antibodies and responses to capsid proteins can limit transduction. A clinical example is voretigene neparvovec, approved in the United States on December 19, 2017, which uses non-replicating AAV2 to deliver RPE65 to retinal cells through subretinal administration. (pmc.ncbi.nlm.nih.gov)

Adenoviral vectors carry double-stranded DNA and normally maintain it outside host chromosomes rather than integrating it. Conventional E1/E3-deleted vectors accommodate approximately 7.5–8 kb of foreign DNA; deleting E1 removes functions essential for replication, which producer cells supply separately. Helper-dependent adenoviral vectors remove essentially all viral protein-coding sequences and can accommodate about 36 kb, making them suitable for larger genes or complex regulatory constructs. They require helper functions during production. Adenoviral delivery can produce strong expression, but immune recognition of the capsid and, in less extensively deleted vectors, residual viral gene expression can cause inflammation and shorten expression. Removing viral coding sequences reduces these liabilities without eliminating capsid-mediated responses. (pubmed.ncbi.nlm.nih.gov)

Gammaretroviral and lentiviral vectors are both derived from retroviruses. They package RNA, reverse-transcribe it into DNA, and ordinarily integrate the resulting sequence into the cellular genome. Their usable insert capacity is roughly 8 kb, including regulatory sequences. Integration allows the introduced sequence to pass to daughter cells, making these platforms useful for genetically modifying proliferating cell populations. An important difference concerns nuclear access: conventional Moloney murine leukemia virus–derived gammaretroviral vectors require passage through mitosis for efficient integration, whereas lentiviral vectors can transduce non-dividing cells. Thomas Roe and colleagues demonstrated the mitotic requirement experimentally in 1993. Stable integration nevertheless introduces the possibility of disrupting a gene or changing the activity of nearby genes. (ema.europa.eu)

Replication-defective vector design separates gene delivery from viral propagation. A particle can enter a target cell without enabling that cell to produce infectious progeny. Retroviral production systems separate the therapeutic transfer genome from genes encoding structural, enzymatic, and envelope proteins, reducing opportunities to reconstruct a replication-competent virus. Self-inactivating designs additionally delete long-terminal-repeat regulatory activity; Romain Zufferey and colleagues reported a lentiviral design with a 400-nucleotide deletion in 1998. This modification reduces unwanted viral promoter activity but does not abolish integration. Replication-defective delivery vectors also differ from deliberately replication-competent or replication-conditional therapeutic viruses, which require a different safety assessment. (ema.europa.eu)

Platform-specific evaluation consequently distinguishes immune and tissue-exposure risks from integration and propagation risks. AAV development examines capsid-directed immunity, tissue distribution, persistence, and unintended integration; adenoviral development additionally addresses residual viral functions and helper-virus contamination where applicable. Integrating retroviral vectors require assessment of insertion-related effects and testing for replication-competent contaminants. The ability to sustain expression is therefore not interchangeable with the absence of replication, and neither property alone establishes a vector’s safety. (ema.europa.eu)

Nonviral Delivery Systems

Nonviral delivery separates the therapeutic payload from a virus-derived vehicle. Payloads include circular plasmid DNA, messenger RNA (mRNA), regulatory RNAs, and RNA–protein complexes used for genome editing. Plasmids carry an expression cassette containing a therapeutic coding sequence and regulatory elements; DNA must reach the cell nucleus before transcription, whereas mRNA can be translated after release into the cytoplasm. Neither lipid encapsulation nor electroporation is itself the therapeutic mechanism: these methods provide intracellular access for a payload that produces a protein, suppresses expression, or enables an edit. The FDA explicitly includes engineered plasmid DNA among gene-therapy product types. (fda.gov)

Unformulated, or “naked,” nucleic acids can produce expression in accessible tissues, but this does not imply efficient delivery throughout the body. In 1990, Jon A. Wolff and colleagues demonstrated protein expression after injecting DNA and RNA constructs directly into mouse skeletal muscle without a specialized carrier; luciferase activity following DNA delivery remained detectable for at least two months. This experiment established that viral infection was not necessary for gene transfer, while illustrating the importance of tissue and administration route. Nonviral DNA need not integrate into chromosomes, and its expression duration depends on the construct and recipient cells rather than on the absence of a viral carrier alone. (doi.org)

Lipid-based systems package nucleic acids in liposomes, lipid–nucleic-acid complexes, or lipid nanoparticles. Their principal tasks are to protect the cargo and support cellular uptake and release; entry into a cell does not necessarily deliver an active payload to the cytoplasm. Many particles enter membrane-bound endosomes, from which the nucleic acid must escape. In a 2013 imaging study, Jérôme Gilleron and colleagues estimated that only 1–2% of nanoparticle-delivered small interfering RNA escaped from endosomes in the experimental systems examined—not a universal efficiency for all lipid formulations. A gene-therapy example was the 2021 initial report of NTLA-2001, which delivered Cas9 mRNA and a guide RNA intravenously in lipid nanoparticles to target liver cells. Among six participants, mean serum transthyretin reductions at day 28 were 52% and 87% in the two dose groups. These were pharmacodynamic results, not measures of nanoparticle uptake or proof of clinical cure. (nature.com)

Polymer carriers provide another means of assembling and protecting nucleic acids. Positively charged polymers associate with negatively charged DNA to form complexes commonly called polyplexes. Polyethylenimine, reported as a gene-transfer vector by Otmane Boussif and colleagues in 1995, is an established experimental example. Biodegradable poly(β-amino ester) systems offer adjustable chemistry and breakdown through ester-bond hydrolysis; a 2013 study demonstrated delivery of minicircle DNA using this platform in cultured cells and animals. Effective polymer delivery requires both sufficient association to protect the payload and subsequent release that permits its biological activity. Carrier composition, degradation, and cell compatibility therefore matter alongside the amount of DNA taken up. (pmc.ncbi.nlm.nih.gov)

Physical methods bypass some carrier-dependent barriers by temporarily altering the cell membrane. Electroporation uses electrical pulses to permit entry of DNA, RNA, or ribonucleoprotein complexes and is particularly useful for cells modified outside the body. Casgevy manufacturing, for example, introduces a CRISPR–Cas9 ribonucleoprotein complex into collected CD34-positive hematopoietic cells by electroporation. Mechanical membrane disruption provides a related experimental approach. These platforms accommodate different payload formats, but delivery efficiency must be assessed together with cell survival and function: a 2020 continuous-flow electroporation study obtained approximately 70% mRNA transfection in primary human T cells under selected conditions, with viability declining as electrical exposure increased. Transient mRNA or editing-protein delivery can nevertheless produce a lasting genomic change; conversely, delivery of an expression plasmid does not by itself guarantee durable expression. (fda.gov)

Therapeutic Genome Editing

Programmable nucleases combine sequence recognition with DNA cleavage. Zinc-finger nucleases (ZFNs) use engineered zinc-finger protein arrays, whereas transcription activator-like effector nucleases (TALENs) use arrays of DNA-binding repeats; both commonly employ paired FokI cleavage domains to produce a double-strand break between their binding sites. In CRISPR-associated systems, a guide RNA directs an enzyme such as Cas9 to complementary DNA beside a compatible protospacer-adjacent motif, or PAM. Retargeting therefore generally involves changing the guide sequence rather than redesigning DNA-binding proteins. In 2012, Martin Jinek and colleagues, including Jennifer Doudna and Emmanuelle Charpentier, demonstrated programmable Cas9 cleavage and engineered a single-guide RNA from its two natural RNA components. (nature.com)

The cleavage event does not itself specify the final edit: cellular DNA repair determines the resulting sequence. Non-homologous end joining reconnects broken ends without a homologous template and can generate small insertions or deletions that disrupt a coding sequence or regulatory element. Homology-directed repair uses a homologous sequence, potentially supplied as donor DNA, to support a specified correction or insertion; it is most active during the S and G2 phases of the cell cycle. A therapeutic edit need not repair the disease-causing variant. Casgevy, initially approved in the United States on December 8, 2023, uses CRISPR–Cas9 editing of a BCL11A erythroid enhancer in blood-forming stem and progenitor cells to increase fetal hemoglobin, rather than correcting the sickle-cell variant itself. (fda.gov)

Base editing couples a targeting protein to an enzyme that chemically changes a DNA base, without requiring an intended double-strand break or a separate donor template. Alexis Komor and colleagues in David Liu’s laboratory reported cytosine base editing in 2016: cytosine is deaminated to uracil, and subsequent processing establishes a $C\!:\!G \rightarrow T\!:\!A$ substitution. Nicole Gaudelli and colleagues introduced adenine base editing in 2017, using an evolved deaminase to produce $A\!:\!T \rightarrow G\!:\!C$ substitutions. These foundational editors act within a limited editing window rather than exclusively on one nucleotide. Other susceptible bases within that window can therefore undergo “bystander” conversion, making the local sequence important to the resulting edit. (nature.com)

Prime editing, reported by Andrew Anzalone and colleagues in 2019, combines a Cas9 nickase with an engineered reverse transcriptase. Its prime-editing guide RNA both identifies the target and carries a primer-binding sequence and a template encoding the desired alteration. After one DNA strand is nicked, reverse transcription extends the exposed DNA end using that RNA template; cellular processing incorporates the newly written sequence. The original study demonstrated all 12 possible single-base substitution types, small insertions and small deletions without a separate donor DNA molecule or an intended double-strand break. Some configurations additionally nick the opposite strand to favor incorporation of the edit, so avoiding a programmed double-strand break does not mean eliminating DNA cleavage or repair-dependent byproducts. (nature.com)

Editing-tool delivery and genomic modification are separate operations. Components can enter cells as DNA, RNA, protein or preassembled protein–RNA complexes, either outside the body or through direct administration. Transient machinery can leave a lasting genomic change; prolonged editor activity is not necessary for that change to persist. Delivery duration nevertheless affects the opportunity for unintended editing. (fda.gov)

Specificity assessment distinguishes edits at unintended sites from unwanted outcomes at the intended site. The FDA’s January 2024 genome-editing guidance recommends complementary computational, biochemical and cellular methods, including genome-wide analysis, followed by sensitive verification of candidate off-target sites in relevant human cells. Characterization also examines insertions, deletions and chromosomal rearrangements at or associated with the target. Editing frequency alone therefore does not establish product quality: the identity, distribution and biological consequences of the resulting sequence changes also require assessment. (fda.gov)

Expression, Persistence, and Tissue Response

Therapeutic gene expression depends on the regulatory architecture of the delivered sequence, not simply on its presence in a cell. A promoter controls initiation of transcription, while enhancers influence its strength and cell-type specificity. Broadly active promoters support expression across several cell types; tissue-selective elements concentrate expression in particular lineages. For example, etranacogene dezaparvovec uses the liver-specific LP1 promoter to drive production of the Padua variant of factor IX in hepatocytes, which release the clotting protein into the circulation. Regulatory targeting therefore distinguishes the tissue producing a therapeutic protein from the tissues benefiting from it. The European Medicines Agency’s 2018 gene-therapy guideline treats the expression construct and the duration of its biological activity as important components of product evaluation. (fda.gov)

Persistence of genetic material has three distinct forms. An episome is DNA maintained outside the host chromosomes: recombinant adeno-associated viral vectors predominantly use this route, although low-frequency integration can occur. Episomal persistence can support prolonged expression without replacement of the original genomic sequence. By contrast, integrating retroviral and lentiviral vectors insert the therapeutic sequence into chromosomal DNA, allowing transmission to daughter cells during division. Genome editing provides a third arrangement: the delivered editing machinery can disappear while the genomic alteration remains. In exagamglogene autotemcel, a Cas9–guide-RNA ribonucleoprotein complex is introduced into blood stem cells outside the body. It disrupts a GATA1-binding site in the erythroid-specific enhancer of BCL11A, reducing BCL11A expression in erythroid descendants and increasing fetal hemoglobin. Continued production of the editing machinery is not required to maintain that alteration. (fda.gov)

Cell turnover determines whether persistent sequences yield persistent tissue function. Nonintegrating episomal DNA may be diluted as cells divide, making growing or regenerating tissues different from relatively stable adult tissues. Experimental liver studies illustrate this limitation: hepatocyte proliferation can progressively reduce episomal vector abundance, whereas chromosomal alterations are inherited by daughter cells. Integration or editing nevertheless cannot preserve a therapeutic population if the modified cells themselves disappear. For blood-cell therapies, durability depends on successful engraftment of modified, long-term repopulating hematopoietic stem cells in bone marrow. Their self-renewal and differentiation can replenish short-lived blood cells; sustained gene marking in granulocytes therefore provides evidence of continuing production from modified progenitors rather than survival of the originally infused mature cells. (insight.jci.org)

Selective survival can further change the proportion of modified cells after treatment. In adenosine deaminase deficiency, correction reduces the metabolic toxicity affecting lymphocytes, giving corrected cells a survival advantage. A 2009 study by Alessandro Aiuti and colleagues reported that, one year after treatment, mean vector-positive proportions were 88.0% in circulating T cells but 5.1% in marrow CD34-positive cells. These measurements describe different cellular compartments, not a uniform correction rate throughout the body. A 2021 lentiviral study likewise found increasing vector copy numbers in peripheral-blood mononuclear cells, interpreted as evidence of selective accumulation of corrected lymphocytes. Tissue response thus reflects both the initial genetic modification and subsequent population dynamics. (nejm.org)

Dose–response assessment connects the administered product to molecular expression and biological function. Relevant measurements include vector genomes or modified cells administered, vector copies per cell, editing frequency, therapeutic protein concentration, enzyme activity, and lineage-specific cellular function. These quantities are not interchangeable: detecting vector DNA establishes gene marking, but does not itself demonstrate expression or therapeutic activity. For factor IX gene transfer, the Hemgenix development programme measured liver transduction, plasma factor IX protein, and clotting activity; precursor-vector animal studies showed dose-dependent increases at four weeks. Clinical expression need not remain proportional to the initial dose. In a 2021 SPK-8011 study of factor VIII gene transfer, initial expression peaks occurred at 6–12 weeks, but a clear dose response in steady-state activity was not observed beyond 52 weeks. Serial biomarker measurements therefore distinguish early expression, establishment of a plateau, and subsequent loss of activity. (ema.europa.eu)

Applications in Inherited Blood and Immune Disorders

Hematopoietic stem cells are central targets because they can engraft in bone marrow, self-renew, and generate successive generations of blood cells. Ex vivo treatments modify a patient’s own CD34-positive stem and progenitor cells, which are administered after conditioning chemotherapy. The therapeutic effect depends on the descendants of the modified cells: erythroid cells produce therapeutic hemoglobin in hemoglobin disorders, whereas immune-cell lineages can regain functions impaired by inherited immunodeficiencies. Mature circulating blood cells are therefore not the principal long-term genetic target; the intervention establishes a renewable source of modified cells. (fda.gov)

In sickle-cell disease, gene addition introduces a β-globin sequence whose product counteracts sickling without repairing the original disease-causing allele. Lovotibeglogene autotemcel, marketed as Lyfgenia, uses a lentiviral vector to produce HbA$^{T87Q}$, an anti-sickling hemoglobin. Its initial US approval on December 8, 2023, covered patients aged 12 years or older with a history of vaso-occlusive events. In the approval analysis, 28 of 32 evaluable participants experienced complete resolution of vaso-occlusive events between months 6 and 18 after infusion. For transfusion-dependent β-thalassemia, betibeglogene autotemcel, marketed as Zynteglo and approved in the United States in 2022, supplies the same modified β-globin to compensate for inadequate β-globin production. Across its two pivotal studies, 32 of 36 evaluable participants achieved transfusion independence, defined as at least 12 continuous months without red-cell transfusions while maintaining a weighted average hemoglobin concentration of at least 9 g/dL. (fda.gov)

Exagamglogene autotemcel, marketed as Casgevy, instead uses CRISPR–Cas9 genome editing to disrupt an erythroid-specific enhancer of BCL11A. Reduced BCL11A expression in erythroid descendants releases repression of γ-globin production, increasing fetal hemoglobin rather than adding a β-globin gene or correcting the original mutation. Fetal hemoglobin reduces sickling in sickle-cell disease and compensates for deficient β-globin production in β-thalassemia. Casgevy received its initial US sickle-cell approval on December 8, 2023, followed by approval for transfusion-dependent β-thalassemia on January 16, 2024; both initially covered patients aged 12 years or older. The FDA’s subsequent indication includes patients aged two years or older. In the original sickle-cell approval dataset, 29 of 31 evaluable participants remained free of severe vaso-occlusive crises for at least 12 consecutive months. This endpoint differed from Lyfgenia’s, so the percentages do not constitute a head-to-head comparison. (fda.gov)

Inherited immune disorders provide a different rationale: restoring a missing enzyme or signaling component can permit immune-cell development and function. In adenosine deaminase–deficient severe combined immunodeficiency, or ADA-SCID, failure to metabolize deoxyadenosine allows toxic metabolites to accumulate and damage lymphocytes. Strimvelis introduces a functional ADA gene into autologous CD34-positive cells using a retroviral vector; it received European Union authorization on May 26, 2016, for patients lacking a suitable matched related stem-cell donor. A separate investigational lentiviral approach, reported in 2021 by Donald Kohn and colleagues, treated 50 patients: 48 had sustained restoration of ADA activity and immune function at two to three years. Other genotype-specific approaches have entered clinical studies, including lentiviral IL2RG gene addition for X-linked severe combined immunodeficiency, reported in eight infants in 2019. These studies are distinct from authorization of a particular product and indication. (ema.europa.eu)

Hemophilia contrasts with these stem-cell applications because genetically modified liver cells produce a circulating clotting factor. Intravenously administered adeno-associated viral vectors deliver a factor IX sequence for hemophilia B or a factor VIII sequence for hemophilia A. Hemgenix received US approval on November 22, 2022, for adults with hemophilia B who use factor IX prophylaxis, have current or historical life-threatening hemorrhage, or experience repeated serious spontaneous bleeding. Roctavian received US approval on June 29, 2023, for adults with severe hemophilia A without pre-existing antibodies to AAV5 detected by an FDA-approved test. In its 112-patient efficacy cohort, mean annualized bleeding decreased from 5.4 to 2.6 episodes over follow-up of at least three years. These adult indications do not establish efficacy in children or confer approval on other investigational factor-delivery or editing strategies. (content.govdelivery.com)

Applications in Other Inherited Disorders

Retinal gene delivery is exemplified by voretigene neparvovec-rzyl (Luxturna), approved by the US Food and Drug Administration (FDA) on December 19, 2017, for retinal dystrophy caused by pathogenic variants in both copies of RPE65. An adeno-associated virus serotype 2 vector delivers functional RPE65 complementary DNA into retinal pigment epithelial cells through subretinal injection. The encoded enzyme supports regeneration of the light-sensitive chromophore required for the visual cycle; treatment therefore restores a biochemical function rather than replacing lost retinal tissue. Eligibility requires surviving, viable retinal cells. The pivotal study assessed functional vision using a multi-luminance mobility test: participants navigated an obstacle course under illumination ranging from 400 to 1 lux. Light sensitivity, visual acuity, and visual fields were additional outcomes. This genotype-specific indication does not extend to inherited retinal degeneration caused by other genes. (fda.gov)

In spinal muscular atrophy (SMA), deficient survival motor neuron (SMN) protein leads to progressive loss of motor neurons. Onasemnogene abeparvovec-xioi (Zolgensma), initially approved in the United States on May 24, 2019, uses an AAV9 vector to introduce a gene encoding functional SMN protein through intravenous administration. Its US indication covers children younger than two years with biallelic SMN1 mutations. This is gene addition, not correction of the endogenous mutation. A separate formulation, onasemnogene abeparvovec-brve (Itvisma), received FDA approval on November 24, 2025, for patients aged two years and older with a confirmed SMN1 mutation and is administered intrathecally. These routes illustrate distinct approaches to neurological tissue access: systemic vector exposure versus administration into cerebrospinal fluid. (fda.gov)

Treatment timing is particularly consequential in SMA because motor-neuron loss is irreversible. In the presymptomatic SPR1NT study, published in 2022, 14 infants with biallelic SMN1 deletions and two SMN2 copies received intravenous onasemnogene abeparvovec at a median age of 21 days. All achieved independent sitting for at least 30 seconds by 18 months, and all survived without permanent ventilation at 14 months; none of 23 untreated children in the matched natural-history cohort achieved independent sitting. The study was single-arm rather than a randomized comparison of treatment ages. Its outcomes concern preservation and acquisition of motor function before substantial degeneration, not regeneration of neurons already lost. (nature.com)

Inherited metabolic disease can instead be approached through genetically modified cells that supply a missing enzyme. In metachromatic leukodystrophy (MLD), deficiency of arylsulfatase A (ARSA) causes sulfatide accumulation and damage to the central and peripheral nervous systems. Atidarsagene autotemcel (Lenmeldy), approved by the FDA on March 18, 2024, consists of the patient’s own CD34-positive hematopoietic stem and progenitor cells, modified outside the body with a lentiviral vector carrying functional ARSA. Following myeloablative conditioning and infusion, modified cells engraft in the bone marrow and generate myeloid descendants that produce ARSA. Unlike subretinal delivery, which places a vector directly beside its target cells, this approach uses an engrafted cellular population as a continuing enzyme source. Its indication is restricted to presymptomatic late-infantile, presymptomatic early-juvenile, and early-symptomatic early-juvenile MLD. (fda.gov)

The evidence supporting Lenmeldy included 37 treated children from two single-arm trials and an expanded-access program, compared with untreated natural-history patients. The principal endpoint was survival without severe motor impairment, defined by loss of locomotion and unsupported sitting, or death. Among children treated before symptoms of late-infantile MLD appeared, all were alive at six years, compared with 58% in the natural-history group; 71% could walk without assistance at five years. Motor and cognitive outcomes therefore complement measurement of enzyme activity: restoration of ARSA production alone does not describe whether neurological function has been preserved. The restriction to specified early disease stages is integral to the clinical application, rather than an indication for all forms or stages of MLD. (fda.gov)

Cancer and Acquired-Disease Applications

Chimeric antigen receptor (CAR) T-cell therapy combines genetic modification with cancer immunotherapy: the introduced gene gives T cells a receptor that recognizes a selected tumor-associated antigen, rather than correcting an inherited defect. A typical CAR contains an extracellular antibody-derived binding domain, a membrane-spanning region, and intracellular signaling and costimulatory domains. Antigen binding activates the engineered cells, supporting their expansion and destruction of antigen-bearing targets. Licensed applications include CD19-directed therapies for specified B-cell leukemias and lymphomas, and B-cell maturation antigen (BCMA)-directed therapies for multiple myeloma. The first US CAR T-cell approval, for tisagenlecleucel, occurred in 2017. These targets are also present on normal cells of the corresponding blood-cell lineage; recognition is therefore not inherently exclusive to malignant cells. (cancer.gov)

Other engineered immune-cell approaches differ in what they recognize. Modified T-cell receptors can recognize fragments of intracellular proteins presented on the cell surface by human leukocyte antigen (HLA) molecules, whereas conventional CARs bind accessible surface antigens directly. Afamitresgene autoleucel (Tecelra), initially approved by the FDA on August 1, 2024, uses autologous T cells engineered to recognize MAGE-A4-derived peptides presented by specified HLA-A*02 variants in synovial sarcoma. Its activity depends on both the tumor antigen and the compatible HLA molecule, making molecular eligibility testing integral to this approach. Antigen selection also constrains solid-tumor therapies: expression may vary between patients and among cells within the same tumor, leaving some malignant cells without a sufficiently abundant target. (fda.gov)

Natural killer (NK) cells can also be genetically equipped with CARs. In a phase I/II study published in 2020, Katy Rezvani and colleagues investigated donor cord-blood-derived NK cells engineered to express a CD19-specific CAR and interleukin-15, which supported cellular proliferation and survival. Eight of 11 patients with relapsed or refractory CD19-positive lymphoid cancers responded, including seven complete remissions. Five responding patients subsequently received additional post-remission treatment, limiting attribution of longer-term disease control to CAR-NK cells alone. This was an experimental donor-derived cellular approach, not evidence of a licensed treatment indication. (mdanderson.org)

Tumor-directed gene delivery instead introduces therapeutic sequences into the tumor itself. Investigational adenoviral delivery of the tumor-suppressor gene TP53, for example, has been studied through intratumoral administration combined with immune-checkpoint inhibitors. Oncolytic viruses add a distinct mechanism: viral replication within tumors causes cell lysis and can release tumor antigens that contribute to immune recognition. Talimogene laherparepvec (T-VEC; Imlygic), a genetically modified herpes simplex virus type 1, was approved in the United States on October 27, 2015. It is injected into unresectable cutaneous, subcutaneous, or nodal melanoma lesions recurring after initial surgery. The virus also expresses granulocyte–macrophage colony-stimulating factor (GM-CSF); this cytokine, together with antigen release, may promote an antitumor immune response, although the complete mechanism remains unresolved. Its label does not establish improved overall survival or an effect on visceral metastases. (cancer.gov)

Infectious-disease research uses genetic interventions to protect host cells or alter their response to infection, rather than to attack tumors. In HIV research, the investigational product SB-728-T consists of a patient’s T cells modified using zinc-finger nucleases to disrupt CCR5, which encodes a cellular coreceptor used by CCR5-dependent HIV strains. Following nuclease cleavage, error-prone DNA repair disrupts receptor expression, with the aim of producing a population of cells resistant to that route of viral entry. Early clinical studies have examined safety, modified-cell persistence, and effects on viral load. SB-728-T remains an investigational strategy, not a licensed HIV treatment; persistence of modified cells is not itself evidence of eradication of infection. (clinicalinfo.hiv.gov)

Preclinical Development and Product Characterization

Target validation establishes whether the proposed genetic intervention can change a disease-relevant biological process, rather than merely deliver a sequence into cells. Proof-of-concept studies test successive steps: access to the intended cells, transfer or modification of genetic material, production of the intended biological activity, and improvement in a relevant disease measure. Cellular assays can establish whether an introduced enzyme is functional or whether engineered cells recognize their intended target; animal disease models assess activity in the context of tissue architecture and whole-body physiology. Model selection depends on biological responsiveness, not simply anatomical resemblance to humans. Healthy animals can complement disease models when the latter have shortened survival or other features that obscure toxicity. FDA’s November 2013 preclinical guidance describes these complementary roles. (downloads.regulations.gov)

Biodistribution studies determine where a product and its transferred genetic material travel, how long they remain, and how they are cleared. ICH S12, issued by FDA in May 2023, recommends clinically representative material and, where feasible, the intended administration route. Sampling includes target and nontarget tissues, with particular attention to unintended exposure of organs such as the liver, brain, and gonads. Quantitative or digital polymerase chain reaction measures vector or transgene sequences over time; protein assays and tissue-localization methods can additionally establish where expression occurs. Detection of genetic material alone does not establish functional expression. Persistent detection in gonads may warrant investigation of whether the signal originates from germ cells or other resident cells, rather than being interpreted as evidence of heritable modification. (fda.gov)

Vector shedding is distinct from distribution within the body: it is release through feces, urine, saliva, respiratory secretions, or skin lesions. FDA’s August 2015 shedding guidance addresses whether such release could expose untreated contacts or health-care personnel. The distinction between detectable genetic sequences and infectious material is important because a nucleic-acid assay can detect fragments or noninfectious particles. Shedding assessment therefore considers the vector’s replication capacity and, where relevant, the infectivity of recovered material. These studies address transmission risk, whereas biodistribution studies principally characterize exposure inside the treated organism. (fda.gov)

Biological-activity measurements connect product characterization to dose justification. Gene-transfer efficiency, expression level, and functional activity are related but separate quantities; a high sequence-copy count is not itself evidence of therapeutic function. Proof-of-concept studies help identify an active dose range, the dose–response relationship, and the duration of activity. Toxicology studies then examine local and systemic effects, their onset, persistence, potential reversibility, and dependence on dose. FDA’s rare-disease guidance specifies that these studies should reflect the proposed clinical dose range, route, schedule, and evaluation endpoints as closely as feasible. Safety assessment can also encompass the delivery device or procedure and immune responses to the vector and expressed protein. The tested material must adequately represent the clinical product, since formulation or manufacturing changes can alter activity and exposure. (fda.gov)

Animal findings remain conditional on the model. Species-specific receptors, vector tropism, and regulation of gene expression can produce tissue targeting unlike that in humans. Preexisting antibodies can change vector distribution, while an animal immune response against a human therapeutic protein can prematurely eliminate expression. Conversely, immunodeficient models used to accommodate human cells cannot reproduce an intact human immune system. ICH S12 recognizes that a relevant animal species may not exist when a vector’s human cellular target is absent in animals; alternative approaches then require biological justification. Such limitations affect interpretation of negative results as well as apparent success: failure to express a transgene may reflect species incompatibility, while absent toxicity may reflect inadequate exposure of the tissue at risk. (fda.gov)

Clinical Trials and Interpretation of Evidence

Early-phase trials primarily evaluate safety, administration feasibility, and the relationship between dose and biological activity; preliminary efficacy measurements help select regimens for subsequent studies rather than establish definitive benefit. The FDA’s June 2015 guidance describes dose escalation and staggered enrollment, with observation intervals intended to detect acute and subacute adverse events before further participants are treated. Dose characterization is product-specific: vector titer may describe a gene-transfer product, whereas the number and composition of administered cells matter for engineered-cell products. Because genetic interventions can have prolonged activity, dose selection cannot simply reproduce the assumptions used for rapidly cleared conventional drugs. (fda.gov)

Later studies evaluate prespecified outcomes that reflect the disease’s consequences, not merely successful gene delivery. Functional vision provides one example. In the phase 3 randomized controlled trial reviewed for Luxturna in 2017, 31 participants were assigned to treatment or control groups, comprising 21 and 10 participants respectively; controls could receive treatment after one year. The multi-luminance mobility test assessed navigation through a standardized course at different illumination levels, capturing functional performance rather than visual acuity alone. Masked grading reduced assessment bias despite the open-label treatment assignment. Such endpoints require validation, including evidence that their measurement properties and changes are meaningful for the affected population. (fda.gov)

For inherited blood disorders, endpoints quantify dependence on supportive treatment or occurrence of clinical events. The 2022 Zynteglo prescribing information defined transfusion independence as at least 12 continuous months without packed-red-cell transfusions while maintaining weighted-average hemoglobin of at least 9 g/dL. Across two phase 3 studies, 32 of 36 evaluable participants, or 89%, achieved that endpoint. For Casgevy’s initial sickle-cell disease approval on December 8, 2023, the primary efficacy outcome was freedom from severe vaso-occlusive crises for at least 12 consecutive months during a 24-month follow-up period. Among 31 participants with sufficient follow-up for evaluation, 29, or 93.5%, achieved it. These percentages describe defined outcomes in evaluable populations, not lifetime cure rates. (fda.gov)

Rare-disease trials often confront small eligible populations, variable progression, and limited knowledge of untreated disease. These constraints affect statistical power and the precision of effect estimates. Randomization strengthens interpretation, but a single-arm study may compare participants with their own pretreatment course or with an external natural-history cohort. Such comparisons require sufficiently similar populations, outcome definitions, and observation schedules; differences in disease severity, supportive care, or ascertainment can introduce confounding. FDA’s January 2020 rare-disease guidance favors concurrent controls where feasible and recognizes historical controls particularly when disease progression is predictable and the treatment effect is substantial. (fda.gov)

Interpretation also depends on who enters the analysis and when outcomes are measured. The Casgevy approval analysis distinguished 44 treated participants from the 31 with sufficient follow-up for its primary efficacy assessment. The Zynteglo studies similarly distinguished 41 treated participants from 36 evaluable for transfusion independence. Reporting these denominators alongside the observation period prevents an endpoint response proportion from being mistaken for a result established in every treated participant. A twelve-month event-free interval and a twelve-month transfusion-free interval are clinically important, but each establishes benefit over a specified period and under specified criteria. (fda.gov)

Molecular activity, surrogate response, and durable clinical benefit constitute different levels of evidence. Detection of a therapeutic sequence or altered gene expression can demonstrate biological activity without establishing improved function or survival. A biomarker becomes a surrogate endpoint only with evidence supporting its ability to predict clinical benefit; mechanistic plausibility alone is insufficient. A claim of cure additionally requires a disease-specific meaning and evidence of sustained freedom from relevant manifestations or treatment dependence. It cannot follow solely from molecular correction or a short observation window. Repeated longitudinal clinical measurements help distinguish persistent benefit from an initially favorable response. (fda.gov)

Adverse Effects and Long-Term Surveillance

Immune toxicity can arise from recognition of the delivery vector, the therapeutic protein, or genetically modified cells. With adeno-associated virus (AAV) vectors, pre-existing or treatment-induced antibodies can neutralize vector particles and reduce gene transfer; cytotoxic T cells can attack transduced cells, reducing expression and damaging tissue. Responses against the newly expressed therapeutic protein constitute a separate immunogenicity concern. These effects are product-dependent rather than universal consequences of gene transfer. The prescribing information for onasemnogene abeparvovec (Zolgensma), for example, identifies serious liver injury and acute liver failure, including fatal cases. It also records postmarketing thrombotic microangiopathy, generally within two weeks of infusion, characterized by thrombocytopenia, hemolytic anemia, and acute kidney injury. (fda.gov)

Engineered immune cells have additional toxicities related to their biological activity. In CAR T-cell therapy, activation and expansion of infused cells can produce cytokine release syndrome: excessive inflammatory signaling causes fever, hypotension, capillary leak, hypoxia, and sometimes organ dysfunction. Immune effector cell–associated neurotoxicity syndrome can involve confusion, impaired speech, or seizures. Recognition of an antigen shared by malignant and normal cells can also damage healthy populations; depletion of normal B cells by CD19-directed therapies contributes to impaired antibody production and infection risk. These inflammatory, neurological, and target-mediated effects are distinct from delayed genetic injury. (cancer.gov)

Insertional mutagenesis occurs when integrated vector DNA disrupts a gene or changes nearby gene regulation, potentially conferring a growth advantage. A 2008 report by Salima Hacein-Bey-Abina and colleagues documented T-cell leukemia in four of nine successfully treated patients in a ten-patient X-linked severe combined immunodeficiency trial, arising 31–68 months after gammaretroviral gene transfer. Leukemic cells contained activating integrations near genes including LMO2 and CCND2, together with additional acquired genetic abnormalities. Clonal expansion—the increasing contribution of descendants of one modified cell—is an observation rather than a diagnosis of cancer. Integration-site analysis can identify persistent or expanding clones, which require interpretation alongside blood counts, clinical findings, and molecular abnormalities. (pubmed.ncbi.nlm.nih.gov)

Unintended genome editing is a different risk category: it includes changes at off-target sequences and unwanted outcomes at the intended site, such as large deletions or chromosomal rearrangements. These changes can impair cellular function without involving vector integration. Assessment therefore includes edit-site characterization, off-target analysis, and examination of chromosomal integrity. Separately, secondary malignancies are a postmarketing surveillance concern for engineered-cell products. On April 18, 2024, the US Food and Drug Administration described boxed-warning changes for BCMA- and CD19-directed autologous CAR T-cell therapies following reports of T-cell malignancies, including CAR-positive tumors, and specified lifelong monitoring for secondary malignancies. Such reports do not establish insertional mutagenesis as the mechanism in every case. (fda.gov)

Adverse events following treatment must also be separated from risks of conditioning and hematopoietic stem-cell transplantation. Myeloablative busulfan can contribute to profound cytopenias, infections, mucosal injury, hepatic complications, and infertility. In the Casgevy prescribing information, a 44-patient sickle-cell disease cohort had grade 3–4 mucositis in 86% and febrile neutropenia in 48% after conditioning and infusion. The label explicitly attributes the tabulated events to the combined treatment course and describes the profile as generally consistent with busulfan conditioning and transplantation; these percentages are not estimates of toxicity caused specifically by the genomic edit. (fda.gov)

Long-term surveillance is risk-based because vector persistence, integration, permanent editing, target-cell longevity, and disease context differ. FDA’s January 2020 guidance recommends 15 years of follow-up for integrating gammaretroviral and lentiviral vectors, up to 15 years for genome-editing products, and up to five years for AAV vectors, subject to product-specific assessment rather than a universal schedule. Follow-up can record new malignancies, hematological abnormalities, neurological disease, and autoimmune events, with molecular investigations where indicated. After authorization, dedicated observational studies, patient registries, and adverse-event reporting provide complementary surveillance: registries systematically track treated populations, while spontaneous reports identify potential safety signals requiring further investigation. (fda.gov)

Manufacturing, Quality Control, and Distribution

Manufacturing controls both the genetic material and the biological systems used to produce or deliver it. Viral-vector production involves characterized producer-cell banks, vector production, harvesting, purification, and formulation; genetically modified cell products additionally require collection, selection, culture, genetic modification, and final harvesting of cells. Reagents, including selection antibodies and culture supplements, require qualification because their composition and contaminants can affect the finished product. Quality therefore depends on control of intermediate stages as well as final testing: a satisfactory end-product assay does not substitute for a consistently controlled manufacturing process. (fda.gov)

Biological variability complicates expansion from development batches to commercial production. In CAR T-cell therapy, patient-derived starting material varies in cell number, viability, and cellular composition; disease and previous treatment may also affect manufacturing performance. Changes in culture duration or growth factors can alter the resulting cell population. Scaling production or transferring it to another facility consequently requires assessment of product quality, not merely increased output. The FDA’s January 2024 CAR T-cell guidance gives a specific example: changing vector production from adherent to suspension culture calls for comparisons of both vectors and the cells modified with them. Splitting the same donor’s starting material between the old and new processes helps separate manufacturing effects from donor variability. (fda.gov)

Batch specifications distinguish identity, purity, and potency. Identity tests establish that the intended genetic construct and, where applicable, cell population are present. Purity testing measures process-related residues, such as host-cell DNA, proteins, and culture reagents, alongside product-related impurities such as empty viral capsids or unwanted cell populations. Potency concerns relevant biological activity rather than particle or cell count alone. The FDA’s January 2020 manufacturing guidance also recommends a minimum acceptable viability of 70% for ex vivo genetically modified cells, while allowing justification of a lower specification with supporting data. Release specifications define the tests and acceptance criteria against which each batch is assessed. (fda.gov)

Microbiological controls address sterility, mycoplasma, endotoxin, and adventitious agents. Living cell products cannot undergo terminal sterilization or sterilizing filtration without losing the cells themselves, making controlled aseptic processing essential. (fda.gov) Replication-defective viral products also require platform-appropriate testing for unwanted replication-competent agents: recombination during manufacture can generate replication-competent virus despite the intended vector design. The FDA’s January 2020 retroviral-vector guidance addresses testing during manufacture, including selection of materials, sampling amounts, and assay methods. These tests investigate contamination or replication capacity, not the therapeutic activity measured by potency assays. (fda.gov)

Stability studies establish supported storage periods and conditions by following attributes such as identity, purity, container integrity, and biological activity. Compatibility with diluents and delivery devices also matters because adsorption or loss of activity can change the delivered dose. (fda.gov) Distribution conditions are product-specific: the prescribing information for tisagenlecleucel, marketed as Kymriah, specifies temperature-monitored storage at or below −120°C and a 30-minute administration window after thawing and reaching room temperature. (dailymed.nlm.nih.gov) For patient-specific products, physical preservation must accompany an uninterrupted chain of identity linking collection, manufacture, shipment, and administration to the intended recipient. The FDA’s 2024 CAR T-cell guidance recommends at least two unique identifiers and label checks before each processing step, with independent patient-and-label checks at the bedside. (fda.gov)

Regulation and Authorization

In the United States, permission to investigate a gene therapy is distinct from licensure to market it. A sponsor submits an Investigational New Drug application (IND) containing manufacturing information, preclinical evidence, and proposed clinical protocols. The IND ordinarily becomes effective 30 calendar days after the Food and Drug Administration (FDA) receives it, unless FDA imposes a clinical hold; FDA may also notify the sponsor earlier that the investigation may proceed. This permits investigation under the specified protocols, not commercial marketing or a claim that efficacy has been established. Marketing requires approval of a Biologics License Application (BLA), supported by evidence of safety and effectiveness and the ability to manufacture the product consistently to appropriate quality standards. Gene therapy products fall within the remit of FDA’s Center for Biologics Evaluation and Research (CBER). (fda.gov)

In the European Union, gene therapy medicinal products are a category of advanced therapy medicinal products (ATMPs), alongside somatic-cell therapies and tissue-engineered products. Regulation (EC) No 1394/2007 establishes the ATMP framework. The gene-therapy definition introduced through Commission Directive 2009/120/EC requires a biological medicine containing an active substance that contains or consists of recombinant nucleic acid, administered to regulate, repair, replace, add, or delete a genetic sequence; its therapeutic, prophylactic, or diagnostic effect must relate directly to that sequence or its expression product. Vaccines against infectious diseases are expressly excluded from this category. For centralized marketing authorization, the European Medicines Agency’s Committee for Advanced Therapies (CAT) prepares a draft opinion, the Committee for Medicinal Products for Human Use (CHMP) adopts the final scientific opinion, and the European Commission makes the legally binding authorization decision. An EMA recommendation is therefore not itself the authorization. (ema.europa.eu)

Scientific guidance, binding legislation, and product-specific authorization perform different functions. FDA guidance generally describes recommended approaches rather than creating legally enforceable obligations; an alternative approach may be used if it satisfies applicable statutes and regulations. EU scientific and procedural guidance likewise operates within the legislative framework rather than replacing it. Authorization can also impose continuing evidence obligations. EU conditional marketing authorization permits earlier authorization on less comprehensive data when the benefit–risk balance is positive, an unmet medical need is addressed, comprehensive evidence is expected subsequently, and the benefit of immediate availability outweighs the risk of awaiting further evidence. It is reviewed annually while the evidence remains incomplete. This mechanism is marketing authorization subject to conditions, not merely permission to conduct a trial. (fda.gov)

Casgevy illustrates why dates and indications must be recorded by jurisdiction. On December 8, 2023, FDA initially approved this genome-editing therapy for sickle-cell disease in patients aged 12 years or older with recurrent vaso-occlusive crises. On January 16, 2024, FDA approved a second indication, transfusion-dependent beta-thalassemia, for patients aged 12 years or older. The EU granted conditional marketing authorization on February 9, 2024, covering transfusion-dependent beta-thalassemia and severe sickle-cell disease with recurrent vaso-occlusive crises in patients aged 12 years or older for whom hematopoietic stem-cell transplantation was appropriate but a suitable human-leukocyte-antigen-matched related donor was unavailable. These are historical authorization milestones, not interchangeable descriptions of worldwide eligibility. (fda.gov)

Regulatory approval does not establish commercial availability in every territory. EMA distinguishes EU-wide authorization from national pricing and reimbursement decisions; an authorized medicine may consequently remain unavailable in some Member States. Regulatory inventories also require classification rather than simple counting. FDA’s “Approved Cellular and Gene Therapy Products” list includes both genetically modified products and other cellular products, including hematopoietic progenitor-cell preparations from cord blood. Its total therefore cannot be treated as the number of licensed gene therapies. Product-specific approval documents and labeling identify the authorized indication, while classification must establish whether the intervention actually involves therapeutic genetic modification. (ema.europa.eu)

Consent, Heritability, and Research Governance

Somatic treatment modifies non-reproductive cells, with genetic effects confined to the treated person rather than transmitted through reproduction. A permanent edit is therefore not necessarily heritable: modifying blood-forming cells differs from modifying eggs, sperm, or their precursors. The World Health Organization (WHO) distinguishes somatic genome editing, germline editing undertaken without reproductive use, and heritable editing involving reproduction. Editing embryos in laboratory research without initiating a pregnancy is consequently a different governance category from using edited embryos to establish pregnancies. On July 26, 2019, WHO published a statement opposing progression to clinical applications of human germline genome editing at that time; this was a policy recommendation, not itself an internationally binding law. (who.int)

Informed consent in therapeutic research covers the experimental purpose, reasonably foreseeable risks, possible benefits, available alternatives, confidentiality, and voluntary participation. For gene-transfer studies, consent also addresses the extended relationship between participants and investigators after administration. The US Food and Drug Administration’s January 2020 long-term follow-up guidance specifies that consent documents explain the purpose and duration of surveillance, contact intervals, visit locations, and what follow-up contacts involve. Intended storage of blood or tissue for future testing must also be disclosed, and documents must be updated when relevant new adverse-reaction information emerges. Follow-up is risk-based rather than identical for every product: the guidance includes periods extending to 15 years for certain persistent genetic interventions. Agreement to participate does not turn an experimental intervention into an established treatment or guarantee benefit. (hhs.gov)

Pediatric research distinguishes parental or guardian permission from the child’s assent. Under US Department of Health and Human Services regulations, assent means affirmative agreement, not merely failure to object; the institutional review board assesses capacity using age, maturity, and psychological state. The regulations provide specified exceptions, including circumstances in which children cannot reasonably be consulted or a potentially beneficial intervention is available only through the research. For research presenting more than minimal risk but offering prospective direct benefit, the board must determine that anticipated benefit justifies the risk and that the benefit–risk relationship is at least as favorable as available alternatives. Participants who reach legal adulthood during continuing research generally provide their own legally effective consent unless an applicable waiver is approved; earlier parental permission and childhood assent do not automatically substitute for adult consent. (hhs.gov)

Protection of genetic information requires attention to both identifiable clinical records and research datasets. NIH guidance explains that removing names does not guarantee anonymity because genomic information may permit re-identification, potentially affecting participants, families, or groups. Its consent guidance addresses future research use, data sharing, and limits to confidentiality. Controlled-access repositories restrict distribution to approved investigators operating under institutional agreements and participant-consent limitations. Institutional oversight also separates human-subject protection from biosafety review: institutional review boards examine consent and participant protections, while institutional biosafety committees operating under the NIH Guidelines review research involving recombinant or synthetic nucleic acids. NIH’s gene-transfer review checklist includes vector origin and modifications, genetic payload, replication-competent virus testing, target cells, and delivery method. (grants.nih.gov)

WHO released Human genome editing: a framework for governance and companion recommendations on July 12, 2021, following an expert committee established in December 2018. The framework addresses institutional, national, regional, and international oversight; the recommendations cover nine areas, including research registries, international research and medical travel, reporting unsafe or unregistered activities, intellectual property, and public engagement. Its illustrative scenarios distinguish a hypothetical somatic-editing trial for sickle-cell disease in West Africa from proposed somatic or epigenetic editing to enhance athletic performance and from a clinic offering heritable editing to international clients. These distinctions concern both biological heritability and the intervention’s purpose: a non-heritable intervention can still be an enhancement proposal rather than disease treatment. The scenarios are governance exercises, not endorsements or clinical authorizations of the proposed activities. (who.int)

Access, Health-System Requirements, and Economic Evaluation

Delivery depends on health-system capacity as well as product availability. Autologous blood-stem-cell therapies require coordinated hematology and transplant services, cell collection, laboratory and pharmacy support, inpatient nursing, and follow-up care. Centers must accommodate manufacturing intervals while maintaining the patient’s disease management. For example, FDA patient information for Casgevy describes potentially repeated stem-cell collections, storage of backup cells, shipment to a manufacturing site, and hospitalization for chemotherapy and infusion; recovery after infusion can require four to six weeks in hospital. A “one-time” infusion therefore represents one component of a prolonged care pathway, rather than a single outpatient encounter. These requirements are particularly relevant to therapies delivered through a hematopoietic stem-cell transplantation infrastructure. (fda.gov)

Geographic access also depends on referral networks, travel, temporary accommodation, caregiver availability, and coordination between the specialist center and local clinicians. Fertility-preservation and behavioral-health services may form part of the treatment pathway, while repeated return visits create continuing logistical demands. Insurance authorization and arrangements for treatment outside the patient’s home state can constitute separate administrative steps. The US Centers for Medicare & Medicaid Services (CMS) explicitly addresses these issues in its Cell and Gene Therapy Access Model: beginning January 1, 2026, participating states must maintain a contract with at least one qualified treatment center for each selected model drug, ensure enrollment of qualified providers, and monitor waiting times and network adequacy. Contracting arrangements may include out-of-state centers and must account for associated care, not merely the infusion. (cms.gov)

Acquisition price, total treatment cost, and payer expenditure are distinct quantities. CMS documents US list prices of $2.2 million per patient for Casgevy and $3.1 million for Lyfgenia following their December 2023 approvals. These figures are product prices, not comprehensive estimates of the treatment episode. Additional costs include assessment, collection, conditioning, hospitalization, supportive care, and subsequent monitoring. Payer expenditure also depends on rebates and negotiated terms, so it need not equal the list price. Economic analyses separately account for future disease-management costs that treatment might avert; reductions in painful crises do not necessarily eliminate expenditure on pre-existing organ damage or chronic pain. Patient and caregiver travel and time costs are another category, whose inclusion depends on the analytical perspective. (cms.gov)

Cost-effectiveness analysis compares additional costs with additional health benefits, often expressed as quality-adjusted life-years. Because much of the expenditure occurs initially while benefits may accrue over decades, estimates depend strongly on assumptions about durability, survival, complications, and future care. Short-term clinical response cannot by itself establish lifelong economic benefit. The Institute for Clinical and Economic Review’s 2023 sickle-cell assessment, for example, modeled an annual 0.27% return to standard-care costs and outcomes after year seven; this was an extrapolation assumption, not an observed failure rate. Analyses therefore test alternative durability scenarios. Budget impact addresses a different question: how treating an eligible population affects expenditure over a specified period. A therapy can offer favorable lifetime value yet create substantial near-term financing demands. (icer.org)

CMS’s voluntary Cell and Gene Therapy Access Model provides a documented example of outcomes-based payment. Its initial focus is Medicaid access to gene therapies for sickle-cell disease. The model opened a rolling participation window in January 2025, and the first state began participation on March 1, 2025. CMS negotiates common terms with manufacturers on participating states’ behalf, including discounts and rebates linked to specified health outcomes. If treatment is less successful than agreed, manufacturers reimburse part of the payer’s expenditure. CMS establishes financial and clinical measures, reconciles data, and evaluates results; states retain responsibility for their share of treatment costs. These arrangements redistribute some financial risk, but their operation requires outcome measurement and sustained data collection rather than an assumption that every treated patient obtains permanent benefit. (cms.gov)

References

1. What is Gene Therapy?fda.gov
2. Long Term Follow-Up After Administration of Human Gene Therapy Products; Guidance for Industryfda.gov
3. Bleeding Disorders - Treatmentnhlbi.nih.gov
4. Viral Gene Therapy Shows Promise as Treatment for Bleeding Disordernhlbi.nih.gov
5. Sickle Cell Disease - Treatmentnhlbi.nih.gov
6. Genetic Therapies - Typesnhlbi.nih.gov
7. December 8, 2023 Summary Basis for Regulatory Action - CASGEVYfda.gov
8. What are CAR T cell therapy, RNA therapy, and other genetic therapies?: MedlinePlus Geneticsmedlineplus.gov
9. Commission Directive 2009/120/ECeur-lex.europa.eu
10. Construction of biologically functional bacterial plasmids in vitro.pubmed.ncbi.nlm.nih.gov
11. The NIH Catalyst, May–June 2019: Gene Therapy Turns 30 Years Oldirp.nih.gov
12. NLM History of Medicine Finding Aidsfindingaids.nlm.nih.gov
13. Results From First Human Gene Therapy Clinical Trialgenome.gov
14. T lymphocyte-directed gene therapy for ADA-SCID: initial trial results after 4 yearspubmed.ncbi.nlm.nih.gov
15. Glyberaema.europa.eu
16. FDA Media Briefing on the First Gene Therapy Approved in the U.S.fda.gov
17. Search Orphan Drug Designations and Approvals: Kymriahaccessdata.fda.gov
18. LUXTURNAfda.gov
19. Clinical Development of Gene Therapy Productsfda.gov
20. FDA Approves First Gene Therapies to Treat Patients with Sickle Cell Diseasefda.gov
21. Package Insert - LUXTURNAfda.gov
22. Mirtron-mediated RNA knockdown/replacement therapy for the treatment of dominant retinitis pigmentosanature.com
23. Mutation-independent rhodopsin gene therapy by knockdown and replacement with a single AAV vectorpubmed.ncbi.nlm.nih.gov
24. Package Insert - CASGEVYfda.gov
25. ADSTILADRIN prescribing informationdailymed.nlm.nih.gov
26. Package Insert and Medication Guide - KYMRIAHfda.gov
27. How Gene Therapy Can Cure or Treat Diseasesfda.gov
28. Clinical Development of Gene Therapy Productsfda.gov
29. Package Insert - ZOLGENSMAfda.gov
30. CASGEVY Prescribing Informationdailymed.nlm.nih.gov
31. CAR T Cells: Engineering Immune Cells to Treat Cancercancer.gov
32. T-cell Transfer Therapy - Immunotherapycancer.gov
33. Guideline on the quality, non-clinical and clinical aspects of gene therapy medicinal productsema.europa.eu
34. Guideline on Development and Manufacture of Lentiviral vectorema.europa.eu
35. Reflection paper on quality, non-clinical and clinical issues related to the development of recombinant adeno-associated viral vectorsema.europa.eu
36. Packaging Capacity of Adeno-Associated Virus Serotypes: Impact of Larger Genomes on Infectivity and Postentry Stepspmc.ncbi.nlm.nih.gov
37. LUXTURNA prescribing informationdailymed.nlm.nih.gov
38. Development of a helper-independent human adenovirus vector and its use in the transfer of the herpes simplex virus thymidine kinase genepubmed.ncbi.nlm.nih.gov
39. System for efficient helper-dependent minimal adenovirus construction and rescuepubmed.ncbi.nlm.nih.gov
40. Integration of murine leukemia virus DNA depends on mitosispubmed.ncbi.nlm.nih.gov
41. Self-inactivating lentivirus vector for safe and efficient in vivo gene deliverypubmed.ncbi.nlm.nih.gov
42. Image-based analysis of lipid nanoparticle–mediated siRNA delivery, intracellular trafficking and endosomal escapenature.com
43. A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethyleniminepmc.ncbi.nlm.nih.gov
44. Development of Poly (β-amino esters)-Based Biodegradable Nanoparticles for Non-Viral Delivery of Minicircle DNApmc.ncbi.nlm.nih.gov
45. Scalable continuous-flow electroporation platform enabling T cell transfection for cellular therapy manufacturingnature.com
46. Package Insert - CASGEVY (STN 125787)fda.gov
47. High throughput intracellular delivery by viscoelastic mechanoporationnature.com
48. High-throughput continuous-flow microfluidic electroporation of mRNA into primary human T cells for applications in cellular therapy manufacturingnature.com
49. Non-transgenic genome modifications in a hemimetabolous insect using zinc-finger and TAL effector nucleasesnature.com
50. Human Gene Therapy Products Incorporating Human Genome Editing; Guidance for Industryfda.gov
51. December 8, 2023 Summary Basis for Regulatory Action - CASGEVYfda.gov
52. Programmable editing of a target base in genomic DNA without double-stranded DNA cleavagenature.com
53. Programmable base editing of A•T to G•C in genomic DNA without DNA cleavagenature.com
54. Search-and-replace genome editing without double-strand breaks or donor DNAnature.com
55. Package Insert - HEMGENIXfda.gov
56. Package Insert - HEMGENIXfda.gov
57. Long Term Follow-Up After Administration of Human Gene Therapy Products; Guidance for Industryfda.gov
58. Genome editing without nucleases confers proliferative advantage to edited hepatocytes and corrects Wilson diseaseinsight.jci.org
59. Package Insert - ZYNTEGLOfda.gov
60. CASGEVYfda.gov
61. FDA Roundup: January 16, 2024fda.gov
62. Strimvelisema.europa.eu
63. New gene therapy for the treatment of children with ultra-rare immune disorder recommended for approvalema.europa.eu
64. FDA Approves First Gene Therapy to Treat Adults with Hemophilia Bcontent.govdelivery.com
65. FDA Approves First Gene Therapy for Adults with Severe Hemophilia Afda.gov
66. Clinical Review, December 16, 2017 - LUXTURNAfda.gov
67. ZOLGENSMAfda.gov
68. Package Insert - ZOLGENSMAfda.gov
69. FDA Approves Gene Therapy for Treatment of Spinal Muscular Atrophyfda.gov
70. Package Insert and Information for Patients - ITVISMAfda.gov
71. Onasemnogene abeparvovec for presymptomatic infants with two copies of SMN2 at risk for spinal muscular atrophy type 1: the Phase III SPR1NT trialnature.com
72. FDA Approves First Gene Therapy for Children with Metachromatic Leukodystrophyfda.gov
73. Package Insert - LENMELDYfda.gov
74. LENMELDYfda.gov
75. One CAR T-Cell Therapy for Blood Cancers? - NCIcancer.gov
76. June 17, 2026 Approval Letter - TECELRAfda.gov
77. August 1, 2024 Summary Basis for Regulatory Action - TECELRAfda.gov
78. First Cancer TCR Cell Therapy Approved by FDA - NCIcancer.gov
79. CD19 CAR NK-cell therapy achieves 73% response rate in patients with leukemia and lymphomamdanderson.org
80. Safety and Efficacy of p53 Gene Therapy Combined With Immune Checkpoint Inhibitors in Solid Tumors. - NCIcancer.gov
81. Purple Book Search: Imlygicpurplebooksearch.fda.gov
82. Package Insert - IMLYGICfda.gov
83. Preclinical Assessment of Investigational Cellular and Gene Therapy Products: Guidance for Industrydownloads.regulations.gov
84. S12: Nonclinical Biodistribution Considerations for Gene Therapy Productsfda.gov
85. Design and Analysis of Shedding Studies for Virus or Bacteria-Based Gene Therapy and Oncolytic Productsfda.gov
86. Human Gene Therapy for Rare Diseases; Guidance for Industryfda.gov
87. Considerations for the Design of Early-Phase Clinical Trials of Cellular and Gene Therapy Products; Guidance for Industryfda.gov
88. Package Insert - ZYNTEGLOfda.gov
89. FDA Approves First Gene Therapies to Treat Patients with Sickle Cell Diseasefda.gov
90. Human Gene Therapy for Rare Diseases; Guidance for Industryfda.gov
91. Immunogenicity of Gene Therapy Productsfda.gov
92. Assessing Immunogenicity of Products for Gene Therapy and T cell Therapyfda.gov
93. FDA Technology: Identification and Elimination of Helper and Cytotoxic T Cell Epitopes in AAV9fda.gov
94. Pediatric CAR T-Cell Therapy (PDQ®)cancer.gov
95. Insertional oncogenesis in 4 patients after retrovirus-mediated gene therapy of SCID-X1pubmed.ncbi.nlm.nih.gov
96. Human Gene Therapy Products Incorporating Human Genome Editingfda.gov
97. Safety Assessment of Genome Editing in Human Gene Therapy Products Using Next-Generation Sequencing; Draft Guidance for Industryfda.gov
98. FDA Requires Boxed Warning for T cell Malignancies Following Treatment with BCMA-Directed or CD19-Directed Autologous Chimeric Antigen Receptor (CAR) T cell Immunotherapiesfda.gov
99. CAR T-Cell Therapy and Secondary T-Cell Cancer Riskcancer.gov
100. Manufacturing of Gene Therapies: Ensuring Product Safety and Qualityfda.gov
101. Chemistry, Manufacturing, and Control Information for Human Gene Therapy Investigational New Drug Applications; Guidance for Industryfda.gov
102. Considerations for the Development of Chimeric Antigen Receptor T Cell Products; Guidance for Industryfda.gov
103. Testing of Retroviral Vector-Based Human Gene Therapy Products for Replication Competent Retrovirus During Product Manufacture and Patient Follow-upfda.gov
104. KYMRIAH (tisagenlecleucel) Prescribing Informationdailymed.nlm.nih.gov
105. Frequently Asked Questions — Developing Potential Cellular and Gene Therapy Products; Guidance for Industryfda.gov
106. Biologics License Applications (BLAs) for CBER-Regulated Productsfda.gov
107. Cellular & Gene Therapy Productsfda.gov
108. Advanced therapy medicinal products: Overviewema.europa.eu
109. Directive - 2009/120 - EN - EUR-Lexeur-lex.europa.eu
110. Committee for Advanced Therapies (CAT)ema.europa.eu
111. Committee for Advanced Therapies (CAT) agenda - 19-21 June 2024ema.europa.eu
112. Innovative Designs for Clinical Trials of Cellular and Gene Therapy Products in Small Populations: Draft Guidance for Industryfda.gov
113. Marketing-authorisation procedures for advanced-therapy medicinal productsema.europa.eu
114. Conditional marketing authorisationema.europa.eu
115. CBER 2023 Orphan Approvals (new BLAs)fda.gov
116. Ensuring responsible use of life sciences researchwho.int
117. Statement on governance and oversight of human genome editingwho.int
118. Informed Consent FAQshhs.gov
119. Research with Children FAQshhs.gov
120. Children: Information on Special Protections for Children as Research Subjectshhs.gov
121. NIH Guidance on Elements of Consent under the Genomic Data Sharing Policygrants.nih.gov
122. Using Genomic Data Responsibly Under the NIH Genomic Data Sharing Policygrants.nih.gov
123. Human genome editing: a framework for governancewho.int
124. WHO issues new recommendations on human genome editing for the advancement of public healthwho.int
125. Patient Information - CASGEVYfda.gov
126. Cell & Gene Therapy Access Model: Request for Applications for Statescms.gov
127. CGT Access Model Frequently Asked Questionscms.gov
128. Gene Therapies for Sickle Cell Diseaseicer.org
129. CGT (Cell and Gene Therapy Access) Modelcms.gov