Viral and non-viral vectors in gene therapy: current state and clinical perspectives

Summary

Advancements in gene therapy have achieved significant milestones in treating human diseases, offering renewed hope to patients with limited options. Key to this progress are vectors, which include both viral and non-viral methodologies that impact the success of gene therapy. Over the past two decades, three widely used viral vectors—lentiviruses (LV), adenoviruses (Ad), and adeno-associated viruses (AAV)—have enabled notable preclinical and clinical successes, including the approval of Luxturna for a genetic retinal disease and CAR-T therapies for blood cancers. Recently, the first-in-human dual AAV therapy for hereditary hearing loss, which overcomes large gene delivery, has showcased the restoration of auditory function for patients. Additionally, non-viral vectors such as lipid nanoparticles (LNP) and N-acetylgalactosamine (GalNAc) have led to successful gene therapy products. This review focuses on both viral and non-viral delivery systems in gene therapy, highlighting their current state and future perspectives in treating human diseases.

Search strategy and selection criteria.

Data for this review were identified by searches of MEDLINE, Current Contents, PubMed, and references from relevant articles using the search terms “viral vectors”, “non-viral vectors”, and “gene therapy”. Information of clinical trials was searched in clinicaltrials.gov. Only articles published in English between 1956 and 2025 were included.

Introduction

Gene therapy is a medical intervention that involves the modification or manipulation of genetic material with a person's cells to treat or prevent disease. This technique aims to correct or replace defective genes, introduce new or modified genes, or alter the expression of genes to achieve therapeutic effects. Currently, this approach primarily targets severe or life-threatening conditions with limited or no effective treatments, aiming to address genetic defects at their root cause. Gene therapy could be performed either in vivo or ex vivo. In vivo gene therapy delivers therapeutic genes directly into the target tissues or cells of patients, while ex vivo gene therapy modifies cells extracted from patients, cultures them outside of the body, and then reintroduces the transgene-expressing cells back into the patients.

Vectors, both viral and non-viral, are critical tools for therapeutic gene delivery and are broadly used in gene therapy. Viral vectors have been used as gene therapy vectors since 1975¹ due to their high infection efficiency in vivo and the diversity of targeted tissues, and they remain the primary gene delivery vectors to this day. Among them, lentiviruses (LV), adenoviruses (Ad), and adeno-associated viruses (AAV) are the most widely used, accounting for over 80% of approved viral-based gene therapy products. Other viral and non-viral vectors are also used in gene therapy due to their unique considerations. To date, a total of 14 ex vivo and 21 in vivo vector-based gene therapies have granted market approval, primarily from the United States Food and Drug Administration (FDA). Among these approved therapies, 29 are viral based, while 6 are non-viral based (Table 1).

Table 1.

Gene therapies based on viral and non-viral vectors that have been approved for market (until Apr. 30, 2025).

Drug name	Company	Condition	Year first approved	Regulatory agency
LV-based ex vivo gene therapy (including CAR-T therapy)
Kymriah	Novartis	Acute lymphocytic leukaemia; diffuse large B-cell lymphoma; follicular lymphoma	2017	FDA
Zynteglo	bluebird bio	Transfusion-dependent beta thalassaemia	2019/2022	EMA/FDA
Breyanzi	Celgene (Bristol Myers Squibb)	Diffuse large B-cell lymphoma; follicular lymphoma; chronic lymphocytic leukaemia; mantle cell lymphoma	2021	FDA
Abecma	bluebird bio	Multiple myeloma	2021	FDA
Carteyva (Relma-cel)	JW Therapeutics (Shanghai)	Diffuse large B-cell lymphoma; follicular lymphoma; mantle cell lymphoma	2021	NMPA
Skysona	bluebird bio	Early cerebral adrenoleukodystrophy (CALD)	2021	FDA
Carvykti	Legend Biotech	Multiple myeloma	2022	FDA
Lyfgenia	bluebird bio	Sickle cell disease	2023	FDA
Libmeldy/Lenmeldy	Orchard Therapeutics	Metachromatic leukodystrophy	2020/2024	EMA/FDA
Breyanzi	Bristol Myers Squibb	Relapsed/Refractory Chronic Lymphocytic Leukaemia	2024	FDA
Tecelra (Afami-cel)	Adaptimmune	Synovial sarcoma	2024	FDA
Ad-based gene therapy
Gendicine	Shenzhen SiBiono GeneTech	Head and neck cancer	2003	CFDA
Oncorine	Shanghai Sunway Biotech	Head and neck cancer; nasopharyngeal cancer	2005	CFDA
Adstiladrin	Merck & Co.	Bladder cancer	2022	FDA
AAV-based gene therapy
Glybera	uniQure	Lipoprotein lipase deficiency	2012	EMA
Luxturna	Spark Therapeutics (Roche)	Leber congenital amaurosis; retinitis pigmentosa	2017	FDA/EMA
Zolgensma	Norvartis	Spinal muscular atrophy	2019	EMA
Upstaza	PTC Therapeutics	Aromatic l-amino acid decarboxylase (AADC) deficiency	2022	EMA
Hemgenix	uniQure	Hemophilia B	2022	FDA
Elevidys	Sarepta Therapeutics	Duchenne muscular dystrophy	2023	FDA
Roctavian	BioMarin	Hemophilia A	2022	FDA
Beqvez	Pfizer	Hemophilia B	2024	FDA
BBM-H901	Belief BioMed	Hemophilia B	2025	NMPA
HSV-based gene therapy
Imlygic	Amgen	Melanoma	2015	FDA/EMA
Vyjuvek	Krystal Biotech	Dystrophic epidermolysis bullosa	2023	FDA
Delytact	Daiichi Sankyo	Malignant glioma	2021	MHLW
RV-based gene therapy (including CAR-T therapy)
Rexin-G	Epeius Biotechnologies	Solid tumours	2006	BFAD
Strimvelis	Orchard Therapeutics	Adenosine deaminase deficiency	2016	EMA
Yescarta (ex vivo)	Kite Pharma (Gilead)	Diffuse large B-cell lymphoma; Hodgkin's lymphoma; follicular lymphoma	2017	FDA
Tecartus (ex vivo)	Kite pharma (Gilead)	Metachromatic leukodystrophy	2020	FDA
LNP-based gene therapy
Onpattro	Alnylam&Sanofi	Human transthyretin amyloidosis (hATTR)	2018	FDA
GalNAc-based gene therapy
Givlaari	Alnylam	Acute hepatic porphyria (AHP)	2019	FDA
Oxlumo	Alnylam	Primary hyperoxaluria type1 (PH1)	2020	FDA
Leqvio	Novartis	Hypercholesterolaemia	2020	FDA
Amvuttra	Alnylam	hATTR	2022	FDA
Rivfloza	Novo Nordisk A/S	Primary hyperoxaluria	2023	FDA

FDA: Food and Drug Administration (US); EMA: European Medicines Agency; CFDA: China Food and Drug Administration (2003–2018); NMPA: National Medical Products Administration (China); BFDA; Bureau of Food and Drugs (Philippines); MHLW: Ministry of Health, Labour and Welfare (Japan)

The administration route is a crucial aspect of drug delivery, as different methods have distinct technical applicability and drawbacks. Systemic injection is suitable for treating haematological diseases, liver-related conditions, and for applications involving CAR-T therapies; however, it often requires high doses and can have off-target effects, potentially triggering severe immune reactions and other toxicities. In contrast, local injection serves as an alternative delivery method for relatively isolated organs, such as the eyes, inner ears, and brain, with specific routes like subretinal and intravitreal for the eyes, round window membrane and posterior semicircular canal (PSCC) routes for the inner ears, and intracerebroventricular and intraparenchymal for the brain. These localized approaches reduce the required doses of gene therapy products and decrease the risk of severe immune responses, although they pose challenges such as invasive administration and the potential for uneven distribution (Fig. 1).

Fig. 1.

Summary of gene therapy modalities. Commonly used viral and non-viral vectors in gene therapy (left), as well as their administration routes (right). IP, Intraparenchymal delivery; ICV,Intracerebroventricular delivery; ICM, Intracisterna magna delivery; SR, Subretinal injection; IVT, Intravitreal injection; RWM, Round Window Membrane route; IT, intrathecal injection.

Gene therapy holds promise for treating a wide range of genetic disorders, including cystic fibrosis, muscular dystrophy, central nervous system (CNS)-related diseases, as well as ocular and hearing disorders and other acquired conditions (Fig. 2). To date, more than 3000 clinical trials of gene therapeutics have been registered, and thousands more are in the preclinical trial stage.² In this review, we discuss general compositions of viral and non-viral vectors and their biological properties, feature their uses as commercialized drugs or in clinical trials, and finally discuss challenges and future outlooks.

Fig. 2.

Current clinical applications ofviral and non-viral gene therapy in major human diseases.

Viral vectors in gene therapy

The powerful in vivo infection capability of viruses allows them to be engineered into efficient therapeutic vehicles. An ideal viral vector needs to (1) be able to carry therapeutic genes of a certain size, (2) be easy to package and produce, (3) be replication-deficient to ensure safety, and (4) exhibit good transduction efficiency in vivo. Several types of viral vectors, including LV, Ad, AAV, and herpes simplex virus (HSV), are applied in the treatment of human diseases. Novel viral vectors have also constantly been developed, providing new options for therapies.

The choice of viral vector depends on various factors such as the specific therapeutic goal, target cell type, duration of gene expression, safety profile, and immune response. Each viral vector has its advantages and limitations, and extensive research and clinical trials are conducted to ensure safety and efficacy before their use in human patients.

There are several successful gene therapies based on viral vectors. In the following sections, we will discuss the characteristics of these viral vectors, and their applications in gene therapy, with a particular focus on approved gene therapy products (Table 1) and drugs in phase 3 clinical trials (Table 2).

Table 2.

A selection of ongoing Phase 3 clinical trials employing viral and non-viral vectors.

Vector	Conditions	Sponsor	Interventions	NCT Number
LV	Lysosomal Storage Diseases| Metachromatic Leukodystrophy	Orchard Therapeutics	OTL-200	NCT04283227
	Sickle Cell Disease	bluebird bio	BB1111	NCT04293185
	Haematological Malignancies	Galapagos NV	GLPG CAR T-cell therapy	NCT06652633
	Hurler Syndrome	Orchard Therapeutics	OTL-203	NCT06149403
	Neoplasms	Celgene	Gene-modified (GM) T cell therapy	NCT03435796
	Long Term Safety of Patients Receiving CAR-T in an Eligible Clinical Trial or Managed Access Program	Novartis Pharmaceuticals	Previously treated CAR-T patients	NCT02445222
AdV	COVID-19 Prevention	Gamaleya Research Institute of Epidemiology and Microbiology, Health Ministry of the Russian Federation	Sputnik Light	NCT04741061
	Heart Failure	Renova Therapeutics	Ad5.hAC6|	NCT03360448
	Hepatocellular Carcinoma	Huazhong University of Science and Technology	ADV-Tk	NCT03313596
	Angina, Stable	Angionetics Inc.	Ad5FGF-4	NCT02928094
	Head and Neck Neoplasms	Guangzhou Double Bioproducts Co., Ltd	Endostatins injection	NCT02630264
	Angina, Stable	Cardium Therapeutics	Alferminogene tadenovec	NCT01550614
	Transitional Cell Carcinoma| Bladder Cancer|Carcinoma in Situ| Carcinoma in Situ Concurrent With Papillary Tumours	CG Oncology, Inc.	Adenovirus Vector	NCT01438112
	End Stage Renal Disease	Ark Therapeutics Ltd	Graft placement surgery plus Trinam therapy	NCT00895479
	Angina Pectoris	Cardium Therapeutics	Ad5FGF-4	NCT00438867
	Ovarian Cancer|Primary Peritoneal Cavity Cancer	Merck Sharp & Dohme LLC	AdV-p53 SCH-58500	NCT00003880
	Pancreatic Cancer	GenVec	TNFerade	NCT00051467
	Carcinoma, Squamous Cell	Introgen Therapeutics	INGN 201	NCT00041626
	Carcinoma, Squamous Cell	Introgen Therapeutics	INGN 201	NCT00041613
AAV	Hemophilia A|Hemophilia B	Pfizer	Testing of hepatic AAV Vector integration	NCT05568719
	Hemophilia A|Hemophilia B	Pfizer	Standard of Care FIX Replacement therapy	NCT03587116
	Hemophilia A	Pfizer	Recombinant AAV2/6 Human Factor VIII Gene Therapy	NCT04370054
	Hemophilia B	Shanghai Belief-Delivery BioMed Co., Ltd	Single dose intravenous injection of BBM-H901	NCT05203679
	Hemophilia B	Pfizer	PF-06838435	NCT03861273
	Hemophilia B	CSL Behring	AAV5-hFIXco-Padua	NCT03569891
	Neovascular Age-Related Macular Degeneration (nAMD)|Wet AMD	Adverum Biotechnologies, Inc.	Ixo-vec (anti-VEGF)	NCT06856577
	X-Linked Retinitis Pigmentosa	Beacon Therapeutics	rAAV2tYF-GRK1-hRPGRco	NCT04850118
	Choroideremia |X-Linked Retinitis Pigmentosa	NightstaRx Ltd, a Biogen Company	BIIB111| BIIB112	NCT03584165
	Choroideremia	Biogen	BIIB111	NCT03496012
	Inherited Retinal Dystrophy Due to RPE65 Mutations |Leber Congenital Amaurosis	Spark Therapeutics, Inc.	AAV2-hRPE65v2	NCT00999609
	ornithine transcarboxylase (OTC) Deficiency	Ultragenyx Pharmaceutical Inc	DTX301	NCT05345171
	Glycogen Storage Disease Type IA	Ultragenyx Pharmaceutical Inc	DTX401	NCT05139316
	Mucopolysaccharidosis Type IIIA	LYSOGENE	LYS-SAF302	NCT03612869
GalNAc	Primary Hyperoxaluria Type 1 (PH1)|Primary Hyperoxaluria Type 2 (PH2)|Kidney Diseases|Urologic Diseases|Genetic Disease|Primary Hyperoxaluria Type 3 (PH3)	Dicerna Pharmaceuticals, Inc., a Novo Nordisk company	DCR-PHXC	NCT04042402
	Homozygous Familial Hypercholesterolaemia	Novartis Pharmaceuticals	Inclisiran Sodium for injection	NCT03851705
	Hereditary Angioedema	Intellia Therapeutics	NTLA-2002	NCT06634420
LNP	SARS-CoV-2 Infection	Everest Medicines (Singapore) Pte. Ltd.	PTX-COVID19-B	NCT05534048 NCT05534035
	SARS-CoV-2 Infection|HIV Infections|COVID-19	COVID-19 Prevention Network	Moderna mRNA-1273	NCT05168813
	SARS-CoV-2 Infection|COVID-19	BioNTech SE	BNT162b2	NCT04816669
	SARS-CoV-2 Infection	National Institute of Allergy and Infectious Diseases (NIAID)	Moderna COVID-19 Vaccine	NCT04811664
	Transthyretin Amyloidosis (ATTR) With Cardiomyopathy	Alnylam Pharmaceuticals	Patisiran	NCT03997383

Lentivirus

Genome and structure

Lentivirus is a type of RNA virus that belongs to the Retroviridae family.¹ They are enveloped, single-stranded RNA viruses with a diameter of approximately 100 nm.² LV vectors generally consist of several essential components, including the 5′ long terminal repeat (LTR), the Ψ packaging signal, the central polypurine tract/chain termination sequence (cPPT/CTS), the Rev responsive element (RRE), and the 3′ LTR, which contains the poly(A) signal.³ Core protein genes of LV include gag, pol and env. The gag encodes for capsid and matrix proteins, the env responsible for transmembrane and envelope glycoproteins expression that dictate the virus’ cellular tropism, and the pol encodes for reverse transcriptase and integrase. There are also several auxiliary genes exist in lentiviral genome, such as Vif, Vpr, Vpu, and Nef, which are important for viral entrance into host cells or viral release. Lentiviruses recognize specific cell surface receptors to enter the target cell. The viral RNA is reverse transcribed into DNA, which then transports to the nucleus. Upon entering the nucleus, the viral DNA integrates into the host genome, ensures that the gene is passed on during cell division, facilitating long-term gene expression in the host cells (Fig. 3). Their genetic cargo capacity is up to 9 kb.⁴ Lentivirus is highly versatile for gene delivery as it can infect both dividing and non-dividing cells. The first generation of lentiviral vectors is derived from human immunodeficiency virus 1 (HIV-1) and has proven its capability to efficiently transduce CNS organs in vivo.⁵ To improve safety, the new generations of lentiviral vectors are devoid of auxiliary genes such as Vif, Vpr, Vpu, and Nef, which are not essencial for viral production and infection. Additionally, the 3′ LTR is also truncated to create a self-inactivating viral packaging system.⁶

Fig. 3.

Transduction pathway of viral and non-viral vectors. Viral vectors attach to specific receptors on the surface of target cells. Adenoviral and adeno-associated viral vectors utilize endocytosis to deliver their genetic material, whereas lentiviral and herpes simplex viral vectors employ membrane fusion mechanisms to efficiently release their genetic material into the cytoplasm. In contrast, non-viral vectors enter cells independently of receptors and rely on endocytosis to deliver DNA or RNA.

Lentivirus as a vector in ex vivo gene therapy

Lentiviral vectors have become highly popular for clinical applications due to their ability to transduce nonproliferating or slowly proliferating cells. These vectors have been used to alter T cells by introducing genes to generate immunity against cancer through the delivery of chimeric antigen receptors (CARs) or cloned T-cell receptors. Recently, multiple clinical trials used third-generation, self-inactivating LVs to introduce genes into haematopoietic stem cells to correct primary immunodeficiencies and haemoglobinopathies.

As of November 15, 2024, there are a total of 11 approved LV-based ex vivo gene therapies, including 3 haematopoietic stem-cell transplantation (HSCT) therapies and 8 CAR-T therapies for the treatment of refractory haematologic malignancies and other diseases (Table 1). Kymriah, a CD19-directed genetically modified autologous T-cell immunotherapy, was the first drug approved by the FDA to treat paediatric B-cell acute lymphoblastic leukaemia (ALL) in 2017 and was also the first approved CAR-T therapy globally.⁷ Zynteglo utilizes autologous CD34⁺ haematopoietic stem cells transduced with a lentiviral vector containing the β-globin gene to treat β-thalassaemia. It was first approved in the European Union in 2019 and received FDA approval in 2022.⁸ Libmeldy is an autologous haematopoietic stem cell gene therapy (HSC-GT) product based on autologous CD34⁺ stem cells transduced ex vivo with a lentiviral vector encoding the human ARSA gene, which received marketing authorisation for the treatment of metachromatic leukodystrophy (MLD) in December 2020 in the EU⁹ and approved by the FDA in March 2024.¹⁰ Additionally, Skysona, a gene therapy developed by bluebird bio for the treatment of cerebral adrenoleukodystrophy (CALD), was approved by the FDA in 2022.⁸ Clinical trials based on lentiviral gene therapy are mainly focus on primary immunodeficiencies, haemoglobinopathies, leukaemia, and neurodegenerative diseases.¹¹ Phase III clinical trials of lentiviral gene therapies targeting MLD, Sickle Cell Disease, Haematological Malignancies, and cerebral metabolic disorder such as Hurler Syndromes are currently underway (Table 2).

Integration provides long-term stable expression of therapeutic genes, which is of great benefit in certain circumstances e.g. early onset liver disorders because of rapidly expanding liver cell population. However, integration remains the most significant risk due to the potential for insertional mutagenesis, which can lead to oncogenesis or other adverse effects.¹² Skysona has been reported to cause blood cancers; among 67 children who received Skysona treatment, 6 were diagnosed with myelodysplastic syndromes and 1 with acute myeloid leukaemia (AML).¹³ To conquer this problem, integrase-defective lentiviral vectors have been designed, which are already being applied in vaccine development,¹⁴ providing wider options for the treatment of human diseases.

Adenovirus

Genome and structure

Adenovirus (Ad) is one of the first viral vectors explored for gene therapy applications.¹⁵ Adenoviruses are a type of DNA virus distinguished by their genome size, which ranges from 34 to 43 kb. These viruses are encapsulated within a non-enveloped icosahedral viral particle, measuring approximately 70–90 nm in diameter. The Ad genome consists of five “early-phase” genes: E1A, E1B, E2, E3, and E4, and five“late-phase” protein L1-L5.¹⁶ There are over 50 serotypes of adenovirus, with Ad5 and Ad26 being the most used in gene therapy. Infection of Ad starts with recognition of cell surface receptors, then Ad enter the cell and release genetic material via endocytosis. The viral DNA then enters the nucleus (Fig. 3). Three generations of Ad vectors have been developed to meet various therapeutic needs. The first generation involves the deletion of the E1 and E3 regions, enabling the vector to carry up to 7.5 kb of foreign DNA. In the second generation, both E1 and E4 regions are removed, significantly reducing immunogenicity. The third generation deletes all viral genes, allowing the vector to accommodate more than 30 kb of foreign DNA. Key features of Ad vectors include: (1) the ability to deliver multiple genome copies into a single host cell, resulting in high levels of gene expression; (2) transient gene expression due to the episomal nature of the DNA cargo, which is not integrated into the host genome; (3) the capability to transduce both dividing and non-dividing cells; and (4) a high level of immunogenicity.¹⁷ Due to these characteristics, Ad vector-based gene therapies are primarily employed for cancer treatment.

Adenovirus in gene therapy

In the early 1990s, Ad vectors were first utilized for in vivo therapy to deliver and express the α-1 antitrypsin (A1AT) gene in rat hepatocytes and lung tissues.18, 19, 20 Following this initial demonstration, numerous Ad-based gene delivery trials were conducted to treat human monogenic diseases. Unfortunately, a series of studies revealed that Ad vectors are highly immunogenic, which not only limited the delivery and expression of exogenous genes but also elicited unwanted immune responses in treated individuals.²¹

It is now widely recognized that Ad vectors can trigger robust immune responses in humans. High immunogenicity in adenoviral vectors presents a dual-edged characteristic that can be both a limitation and an advantage, depending on the therapeutic context. In the realm of gene therapy, elevated immunogenicity is often seen as a drawback, as it can provoke robust immune responses that impede the effective delivery and sustain expression of transgenes within target cells. Conversely, in applications such as oncolytic virotherapy, high immunogenicity plays a beneficial role.²² In this context, the strong immune response elicited by oncolytic adenoviruses could enhance therapeutic efficacy by stimulating anti-tumour immunity, promoting the recognition and elimination of cancer cells. For instance, in 2003, Gendicine, an Ad vector carrying a Rous sarcoma virus promoter-driven p53 gene, became the world's first commercialized gene therapy drug for head and neck cancer and was the first conditionally replicative adenovirus (CRAd) to enter clinical trials.²³ This was followed by H101 (Oncorine), another replication-selective Ad vector that received commercial approval in China.²⁴ Although numerous clinical results have confirmed the safety of H101, this drug ultimately proved to be not as good as expected. The primary reason for its reduced efficacy is that the deletion of E1B–55K leads to diminished viral replication, even within tumour cells in vivo.²⁵ Currently, clinical trials utilizing AdV-based gene therapy primarily focus on developing new vaccines, including those for COVID-19, as well as targeting various types of cancers (Table 2).

Adeno-associated virus

Genome and structure

Adeno-associated virus (AAV) was first identified as a contaminant of adenovirus in 1965.²⁶ It is a non-pathogenic parvovirus featuring a single-stranded DNA genome of 4.7 kb, encased in a non-enveloped icosahedral capsid with an outer diameter of approximately 20 nm. The genome structure of AAV is simply composed of two primary regions: the rep gene and the cap gene, flanked by two 145-nucleotide-long inverted terminal repeats (ITRs).²⁷ The open reading frame (ORF) of rep gene encoded Rep40, Rep52, Rep68, and Rep72, which play important roles in viral replication and packaging. The ORF of cap gene encodes for three viral capsid proteins: VP1, VP2, and VP3, the assembly-activating protein (AAP) and the membrane-associated accessory protein MAAP. AAP provides a scaffold for capsid assembly²⁸ and MAAP has been identified to play a role in the release of AAV particles from the cell.²⁹ The simple structure of AAV makes it a suitable candidate for gene delivery vector development. Subsequent engineering of recombinant AAV (rAAV) showed that the genome could be modified by removing all but the ITR elements, allowing for the insertion of any desired promoter and gene of interest. The rAAV system enables efficient delivery of genes smaller than 4.7 kb into target tissues for therapeutic purposes.

AAV replication depends on a heterologous helper virus. It has specific integration sites in the genome, with a particularly low likelihood of integration. Typically, AAV vectors remain in target tissues as episomal entities, allowing for long-term expression. Additionally, AAV eliciting a mild immune response compared with other viruses. These characteristics make AAV a highly safe delivery vector for gene therapy.

AAVs as vectors in gene therapy

Due to their efficiency and safety in vivo, AAVs are the most suitable viral vectors for gene therapies, particularly for conditions that require long-term gene modifications. Over 13 natural serotypes of AAV have been identified and developed for gene therapy, leading to approval of 8 AAV-based gene therapy products, namely Glybera, Luxturna, Zolgensma, Upstaza, Hemgenix, Roctavian, Elevidys and Beqvez. In 2012, the first AAV product, Glybera, was approved by the European Medicines Agency (EMA).³⁰ It was an AAV-1 based therapy expresses lipoprotein lipase in muscle to treat lipoprotein lipase deficiency. Glybera was withdrawn from the market in 2017 due to its ineffectiveness and difficulty in administration, and a lack of patients. Despite these challenges, Glybera demonstrated the potential for gene therapies to provide long-term solutions for genetic disorders and marked a significant milestone in the field of gene therapy, paving the way for further research and advancements in this area. In 2017, Luxturna, based on AAV2, was first approved by the FDA as an AAV bioproduct for Leber congenital amaurosis.³¹ It is also the first AAV-based drug available for the treatment of retinal dystrophy.³² Clinical trials have demonstrated that Luxturna significantly enhances functional vision in patients with RPE65 mutation-associated inherited retinal dystrophy. Both phase 1 and phase 3 trials have shown sustained improvements in ambulatory navigation, light sensitivity, and visual fields, with effects lasting up to 4 years post-treatment. In 2019, Zolgensma, an AAV9-based gene therapy, was approved by the FDA for the treatment of type I spinal muscular atrophy (SMA) in children, which is the most common single-gene disease causing infant mortality.³¹ Hemgenix is an AAV5-based vector that expresses highly active factor IX (FIX) protein variant Padua for the treatment of hemophilia B. This gene therapy product from CSL Behring was approved by the FDA in 2022 based on safety and efficacy studies in adult male patients with severe or moderately severe diseases.³³ Roctavian is an AAV5 vector-based gene therapy for treating patients with severe hemophilia A who do not have a history of coagulation factor VIII inhibitors or detectable antibodies to AAV5. This therapy was conditionally approved by the EMA in 2022 and approved by the FDA in 2023.³⁴ Elevidys, a therapy based on AAVrh74, became the world's first single-use gene therapy for Duchenne muscular dystrophy (DMD) and received FDA approval to be marketed in the United States for the treatment of independently ambulatory children aged 4–5 years.³⁵ Another AAVrh74 based gene therapy, Beqvez, was approved by the FDA in April 2024, for the treatment of adults with moderate-to-severe hemophilia B (congenital factor IX deficiency).³⁶ In April 2025, China's first AAV gene therapy drug, BBM-H901, which is based on AAV843—a capsid independently developed by Belief BioMed—was approved by the NMPA for the treatment of Hemophilia B.³⁷

So far, there are over 200 clinical trials based on AAV underway, primarily focused on the treatment of visual and auditory disorders, neurological diseases, cardiovascular diseases, muscle conditions, hemophilia, and tumours. With the success of Luxturna, the potential of gene therapy is particularly evident for the treatment of relatively isolated organs such as the eyes and ears. Their small volumes require only low doses of rAAV. Many ocular and hearing conditions are monogenic, making them ideal candidates for gene therapy. Furthermore, the accessibility of the eyes and ears allows for multiple administration routes for rAAV delivery. Subsequently, a series of gene therapies has been developed for the treatment of ocular and hearing disorders. Several gene therapies for hereditary ocular diseases have entered clinical phase III trials (Table 2).

The increasing number of clinical trials utilizing rAAVs is an encouraging indicator of the potential of these vector-based gene therapies. However, numerous challenges remain within the field. Since the wild-type genome of AAV is approximately 4.7 kb, its gene load capacity is limited. Oversized genes and other elements need to be truncated or minimized for effective delivery, which may reduce the activity of transgenes. Another strategy is to use dual- or multiple-vector systems, by homology,³⁸ RNA trans-splicing,³⁹ or split intein designs.⁴⁰ (Fig. 4). It is worth mentioning that the first clinical trial using dual AAV delivery of OTOF gene has proven to be safe and effective in children diagnosed with DFNB9 (a genetic hearing loss caused by OTOF mutation) in 2022.⁴¹ This was the first AAV-OTOF therapy used for hereditary hearing loss treatment and the first dual-AAV gene replacement therapy applied in human. The Lancet commented that this work provides a “paradigm shift” in the treatment of hereditary deafness. Both unilateral and bilateral ear injections have shown significant efficacy.^41,42 Subsequently, other groups also confirmed the success of AAV-OTOF therapy in hereditary hearing loss treatment.^43,44

Fig. 4.

Dual- or multiple-AAV vector systems. Dual-AAV vectors have been developed to transfer large transgenes into various tissues. There are three split-AAV vector strategies: the overlapping, the trans-splicing, and the hybrid (AP and AK) split-AAV approch.

Adaptive immunity against the capsid and the foreign transgene are significant factors contributing to reduced efficacy.⁴⁵ Pre-existing neutralizing antibodies (NAbs) against the vector capsids also influence transduction efficacy. Plasmaphaeresis offers a viable method for removing anti-AAV antibodies from the bloodstream,⁴⁶ while pre-treatment with IgG-cleaving endopeptidases can nonspecifically reduce IgG antibodies in patients’ sera.⁴⁷ Although AAV has a lower immunogenicity compared to other viruses, it can still trigger severe immune reactions at high doses, including hepatotoxicity, complement activation, and acute kidney injury.⁴⁸ The pursuit of new capsids with lower immunogenicity is another option. To mitigate transgene triggered immunity, using tissue-specific promoters to detarget professional antigen-presenting cells (APCs) could be helpful.⁴⁹

Other viral vectors

Other viral vectors are also being studied in the clinical settings for in vivo gene delivery, albeit to a lesser extent. These include retrovirus (RV), herpes simplex virus (HSV), vesicular stomatitis virus (VSV), Modified vaccinia virus Ankara (MVA), arenavirus, Sendai virus, and measles virus, among others. Of note, the clinical applications of these vectors are based on their unique properties. For example, VSV is being studied for treating liver cancer and advanced solid tumours because of its intrinsic oncolytic capability. MVA is widely exploited for vaccination applications attributed to its tropism toward antigen presenting cells.

There are four RV-based and two HSV-based treatments approved for commercialization (Table 1). Rexin-G was the world's first gene therapy drug based on RV; it was used as a monotherapy to treat advanced chemotherapy-resistant cancer and was approved by the Philippine Food and Drug Administration (BFAD) for all solid tumours in 2006.⁵⁰ Strimvelis is an RV-based gene therapy for the treatment of severe combined immunodeficiency (ADA-SCID), introducing a normal ADA gene into the patient's haematopoietic stem cells, and was approved by the European Union in 2016. RV is also used in two CAR-T therapies, Yescarta (approved by the FDA in 2017) and Tecartus (approved by the FDA in 2020) for lymphoma treatment. Imlygic, produced by Amgen, is a melanoma treatment product based on HSV that was approved in 2015, marking the first oncolytic viral therapy approved by the FDA.⁵¹ Vyjuvek, another HSV-based gene therapy for dystrophic epidermolysis bullosa treatment, was approved by FDA in 2023.

Challenges and prospects of viral vectors

Early successes in treating monogenic diseases have demonstrated the safety and effectiveness of viral vector-based gene therapy. However, challenges persist that must be addressed for a broader range of diseases. One prominent challenge is the effective delivery of viral vectors to target tissues, where the choice of administration route can greatly impact the success of therapy. Addressing concerns related to specific cell and tissue targeting while overcoming physical barriers will be essential for expanding the range of targeted diseases. The immunogenicity poses another challenge, impacting the safety, durability, and efficacy of gene therapy. Strategies to modulate and mitigate immune responses, including pre-existing immunity in patients, innate immune reactions upon vector administration, and adaptive immune response triggered by the viral capsid and transgene, are essential for patient safety. Understanding the long-term consequences of immune reactions and their potential impact on repeated dosing will be critical. Finally, gaining a better understanding of integration can inform potential implications for tumour formation, hepatotoxicity, neurotoxicity, and newly detected adverse effects associated with high doses of viral vectors. Addressing these challenges is necessary to improve safety and accessibility for patients to this promising area of therapies.

Non-viral vectors in gene therapy

Although viral vectors are widely used for delivering therapeutic gene fragments, they have drawbacks such as risks of cancer, immunogenicity, broad tropism, and limitations in gene packaging length.⁵² In contrast, non-viral vectors, produced through chemical synthesis or physical processing, can overcome many of the limitations of viral vectors. Non-viral vectors do not have the ability to replicate or infect, thereby avoiding the risks of viral infection or carcinogenicity associated with genome integration, while also exhibiting lower immunogenicity that improves their safety.⁵³ Additionally, non-viral vectors can carry larger gene loads and are easier to produce compared to viral vectors.⁵⁴ However, they still face limitations, such as low transfection efficiency and lack of targeting, which restrict their clinical applications.

Currently, various non-viral vectors are used to deliver gene editing tools, mostly DNA fragments, mRNA and siRNA. Many drugs using non-viral carriers for delivery have been approved by the FDA, among them, the most used carriers are lipid nanoparticles (LNPs) and N-acetylgalactosamine (GalNAc). Additionally, various new non-viral carriers are being applied in the development and clinical research of new drugs. At present, a variety of gene therapy tools (ASO, siRNA, and Crispr/Cas9) using non-viral vectors as carriers have entered or completed phase III clinical trials, providing solutions for various types of diseases such as neuromuscular diseases, metabolic diseases, cardiovascular diseases, haematological diseases and infectious diseases (Table 2).

Lipid nanoparticle (LNP)

LNP delivery technology represents a revolutionary advancement for the clinical translation of gene therapy, capable of delivering small interfering RNA (siRNA), messenger RNA (mRNA), or CRISPR/Cas9.⁵⁵ LNPs are typically composed of four main components: a cationic or ionizable lipid, a helper lipid, polyethylene glycol (PEG)-lipid, and cholesterol, all of which influence the delivery efficiency, fluidity, colloidal stability, and structural stability of LNPs.⁵⁴ In 2013, Teresa Coelho et al. found that Dlin-MC3-DMA was an highly effective LNP vector and was well tolerated in preclinical models, leading to clinical development of Patisiran, an LNP siRNA drug for the treatment of hereditary transthyretin-mediated amyloidosis (hATTR).⁵⁶ Patisiran is the first LNP encapsulated siRNA drug, approved by the FDA in 2018, and it is the first siRNA therapy for a rare peripheral nerve disease.⁵⁷ NTLA-2002, by Intellia Therapeutics, uses lipid nanoparticles to deliver CRISPR-Cas9 mRNA targeting the KLKB1 gene to treat hereditary angioedema (HAE). Phase I/II trials in 2024 showed that it was effective with no adverse events.⁵⁸

N-acetylgalactosamine (GalNAc)

GalNAc coupling modification is the most used small nucleic acid drug delivery system, mainly for the delivery of antisense oligonucleotides (ASO) and siRNA. GalNAc conjugates with siRNA to form a polysaccharide-siRNA monoconjugate, which can be specifically delivered to liver cells. The drug enters the cells to exert its function through cellular endocytosis.⁵⁹ Givosiran is a subcutaneously administered RNA interference therapy that targets hepatic ALAS1 mRNA, thereby preventing the accumulation of δ-aminolevulinic acid (ALA) and porphobilinogen (PBG). In 2019, the FDA approved Givosiran for the treatment of adults with acute hepatic porphyria (AHP).⁶⁰ Another siRNA drug using GalNAc as a carrier, Lumasiran, reduces oxalate production by targeting inhibition of the enzyme encoding hydroxyacid oxidase 1 (HAO1), and was approved by the FDA and EMA in 2020 for the treatment of primary hyperoxaluria type 1 (PH1).⁶¹ Inclisiran is primarily used to treat patients with hypercholesterolaemia and atherosclerotic cardiovascular disease and was approved by the FDA in 2021.⁶² Vutrisiran, a subcutaneous transthyretin-directed siRNA therapy delivered to liver cells via a GalNAc carrier, was approved in the United States in June 2022 for the treatment of polyneuropathy in adults with hATTR amyloidosis.⁶³ Nedosiran is a synthetic double-stranded siRNA that was first approved in the United States on September 29, 2023, for reducing urinary oxalate levels in children aged 9 years and older and adults with primary hyperoxaluria type 1 (PH1) who have relatively preserved kidney function.⁶⁴

Challenge and prospects of non-viral vector

The design of non-viral vectors needs to ensure the protection of genetic material from endonuclease degradation, effective transport to the nucleus, and successful nuclear uptake. Different therapeutic mechanisms require variations in the design of these delivery systems. Currently, non-viral vectors used for gene therapy offer a certain degree of safety and loading capacity.

Although non-viral vectors offer improved safety and loading capacity over viral vectors, currently approved non-viral vectors still face several limitations. One major drawback is their restricted biodistribution. These vectors are primarily taken up by the liver, limiting their effectiveness in delivering therapeutics to other tissues and organs, which narrows their application scope. Additionally, the stability of LNPs can be compromised under varying environmental conditions, affecting their storage and transportation. The cationic lipid components in LNPs may also exhibit cytotoxicity, and high-dose or repeated administrations could lead to cell damage and adverse effects. Meanwhile, GalNAc delivery is dependent on the ASGPR receptor in the liver, meaning that repeated dosing could potentially reduce drug delivery efficiency over time. Despite having a higher loading capacity, non-viral vectors currently lack the ability to effectively deliver CRISPR/Cas9, and the transportation of this type of therapeutics remains a challenge.

Emerging types of non-viral vectors, such as proteolipid vehicles (PLVs) and exosomes, have shown promise in improving the transfection efficiency of gene editing tools and enhancing the delivery efficiency to extrahepatic cells.^65,66 Inorganic nanoparticles are also being considered as prospective vectors, as their physicochemical properties are more easily adjustable, and they can have greater stability and lower manufacturing costs compared to naturally derived or organic materials.⁶⁷ Although some new attempts that deliver CRISPR/Cas9 by non-viral vectors, such as LNP, are emerging, the encapsulate efficiency is restricted. The efficient transportation of this therapeutic system remains a challenge.^68,69

Future research should focus on optimizing vector design, to improve their transfection efficiency and targeting capabilities, as well as developing new screening and characterization methods to promote the clinical application of non-viral vectors. Further exploration of the immunogenic responses and toxicity of non-viral vectors will be essential for the successful development of gene therapy products.

Outstanding questions

With the development of gene therapy, both viral and non-viral vectors have undergone iterative technological advancements. However, each type has its own limitations. Viral vectors have high in vivo infection efficiency but can provoke severe immune responses. Additionally, using viral vectors as delivery vehicles for gene editing tools can increase off-target effects due to prolonged expression. Non-viral vectors, on the other hand, have low immunogenicity and short expression duration, making them more suitable for transient gene expression or suppression; however, their infection efficiency still needs improvement. Many pathogenic genes exhibit tissue or cell specificity in their expression, and the widespread expression of both viral and non-viral vectors can result in the therapeutic genes being expressed in non-target organs, leading to toxicity. Overcoming these challenges is essential for advancing the understanding of viral–host interaction and improvements of vectors. Otherwise, to achieve specific expression of target genes, specific regulatory elements such as promoters, enhancers, repressors, and miRNA binding sites are used. The combination of novel vectors with tissue- and cell-specific regulatory elements to construct a toolbox for targeted delivery is a promising strategy.

Immunology response is another problem. From the first unsuccessful Ad-based gene therapy, most clinical incidents related to gene therapy are caused by immune responses. Due to the nature of viruses, viral-based gene therapy vectors can elicit immune responses in the body. Even adeno-associated viruses (AAV), which are known for their low immunogenicity, can trigger severe immune reactions in cases of high-dose administration. Immune suppress reagents are introduced into treatments. Compared with systemic drug delivery, precise local injection is more beneficial for reducing immunogenicity. Highly efficient and specific non-viral vectors may be another key to solving this problem.

Contributors

G.G. and Y.X. conducted the literature review and wrote the original manuscript. Z. H, H.W. and X.C. edited the manuscript. Y.S. and W.Y. contributed to the conception of this review and edited the manuscript. G.G. and Y.X. contributed equally to this article. All authors approved the final version.

Declaration of interests

The authors declare no competing interests.