How to Choose Gene Therapy Delivery: AAV, Lipid Nanoparticles, and Beyond
Gene therapy delivery is the biggest technical challenge in genetic medicine. This guide compares AAV vectors, lipid nanoparticles, lentiviral vectors, and emerging delivery technologies.
Introduction
Gene therapy delivery is the biggest technical challenge in genetic medicine. This guide compares AAV vectors, lipid nanoparticles, lentiviral vectors, and emerging delivery technologies.
Prerequisites
- ✓ Molecular biology (DNA, RNA, viruses)
- ✓ Understanding of gene therapy concepts
- ✓ Familiarity with clinical trial processes
Key Concepts
Step-by-Step Guide
- 1
Understand the Delivery Challenge
Gene therapy requires delivering genetic material into target cells safely and efficiently. The delivery system must: protect the genetic material from degradation, reach the target tissue, enter the cells, release the payload, and avoid immune detection. No single delivery system is perfect — each has trade-offs in capacity, tropism, immunogenicity, and duration of expression.
Warning: Delivery, not gene editing, is the biggest bottleneck in gene therapy. The editing tools (CRISPR, base editors) work well in vitro — getting them to the right cells in vivo is the challenge. - 2
Master AAV Vectors
AAV is the most widely used gene therapy vector. Advantages: low immunogenicity, non-integrating (safe), tissue-specific serotypes, FDA-approved (Luxturna, Zolgensma, Hemgenix). Disadvantages: small packaging capacity (~4.7kb), pre-existing immunity in ~70% of adults, expensive manufacturing ($1-2M per dose). Serotypes determine tropism: AAV2 (eye), AAV5 (airway), AAV8 (liver), AAV9 (CNS, heart), AAVrh10 (CNS).
textAAV Serotype Tropism: AAV2: Eye, CNS AAV5: Airway, retina AAV8: Liver, pancreas AAV9: CNS, heart, muscle AAVrh10: CNS, lung AAV-PHP.eB: CNS (crosses BBB) Packaging limit: ~4.7 kb Cost: $1-2M per dose (commercial) - 3
Understand Lentiviral Vectors
Lentiviruses (derived from HIV) integrate into the host genome, providing long-term expression. Used primarily for ex vivo therapy: extract patient cells, transduce with lentivirus, reinfuse. FDA-approved therapies: CAR-T cells (Kymriah, Yescarta), sickle cell (Lyfgenia, Casgevy). Advantages: large packaging capacity (~8kb), stable integration. Disadvantages: insertional mutagenesis risk, complex manufacturing.
Tip: Lentiviral vectors are the workhorse of ex vivo gene therapy (CAR-T, sickle cell). AAV dominates in vivo gene therapy (eye, liver, CNS). - 4
Compare Lipid Nanoparticle Delivery
LNPs deliver mRNA and DNA payloads without viral components. Advantages: no pre-existing immunity, large payload capacity, tunable tropism, cheaper manufacturing. Disadvantages: transient expression (mRNA degrades in days), liver accumulation (hard to target other organs), less efficient than viruses for some tissues. LNPs are the basis for mRNA vaccines and some CRISPR therapies.
LNP and viral vector manufacturing require different facilities — LNP is chemical, viral is biological. - 5
Evaluate Non-Viral Delivery Methods
Beyond LNPs: Electroporation (electrical pulses open cell membranes — used in ex vivo CAR-T), naked DNA/plasmids (low efficiency but simple), polymer nanoparticles (biodegradable, tunable), exosomes (natural vesicles, low immunogenicity but hard to load), and cell-penetrating peptides (for small payloads). Each has niche applications but none match AAV or LNP for general use.
Tip: Electroporation is the standard for ex vivo T-cell modification — it is simpler and cheaper than lentiviral transduction for some applications. - 6
Target Specific Tissues
Tissue targeting is the frontier of gene therapy delivery. Approaches: AAV serotype selection (natural tropism), AAV capsid engineering (directed evolution, rational design), LNP surface modification (targeting ligands, antibodies), and tissue-specific promoters (gene only expressed in target cells). The blood-brain barrier is the biggest challenge for CNS delivery — AAV-PHP.eB crosses it in mice but not yet reliably in humans.
Warning: Most systemically delivered gene therapies accumulate in the liver. Liver toxicity is the most common dose-limiting adverse event in gene therapy trials. - 7
Manage Immunogenicity
Immune responses to delivery vectors limit gene therapy. Pre-existing antibodies to AAV (70% of adults) exclude many patients. Anti-drug antibodies (ADAs) after first dose prevent re-dosing. Mitigation: immunosuppression (rituximab, sirolimus), engineered capsids that evade antibodies, tolerization (gradual exposure to reduce immune response), and alternative serotypes for re-dosing.
textImmunogenicity Management: 1. Screen patients for pre-existing antibodies 2. Exclude seropositive patients (current approach) 3. Immunosuppression regimen (rituximab + sirolimus) 4. Engineered capsids (immune evasion) 5. Alternative serotypes for re-dosing 6. Tolerization protocols (experimental) Challenge: 30-70% of adults have AAV antibodies - 8
Scale Manufacturing
AAV manufacturing is the cost bottleneck: 1 dose of Zolgensma = $2.1M. Current process: transient transfection in HEK293 cells (low yield, ~1e14 vg per batch). Improvements: baculovirus/Sf9 system (higher yield, lower cost), producer cell lines (stable, scalable), and suspension culture (scalable bioreactors). Target: $100K-500K per dose by 2030.
Warning: AAV manufacturing yield is 10-100x lower than other biologics. A single dose may require an entire bioreactor batch. This is why gene therapies cost $1-2M per dose. - 9
Navigate Clinical and Regulatory Pathways
Gene therapy regulatory requirements: long-term follow-up (15 years for integrating vectors, 5 years for AAV), specific toxicity monitoring (liver, thrombocytopenia, neurotoxicity), and REMS (Risk Evaluation and Mitigation Strategy) programs. FDA has approved 12+ gene therapies as of 2026. The pathway is established but expensive — Phase 3 trials cost $100-300M.
- 10
Assess Emerging Delivery Technologies
Next-generation delivery: AAV capsid engineering (AAV-PHP.eB for CNS, AAV-LK03 for liver), targeted LNPs (SORT technology for lung, spleen, liver), virus-like particles (VLPs — non-replicating, low immunogenicity), and in vivo CRISPR delivery (LNP-CRISPR for liver, AAV-CRISPR for other tissues). Intellia's NTLA-2001 delivers CRISPR via LNP to liver for transthyretin amyloidosis — the first in vivo CRISPR therapy.
- 11
Choose the Right Delivery System
Decision framework: Ex vivo (extract, modify, reinfuse) → lentivirus or electroporation. In vivo, long-term expression → AAV (serotype by tissue). In vivo, transient expression → LNP (mRNA or DNA). CNS target → AAV9 or AAV-PHP.eB. Liver target → AAV8 or LNP. Eye → AAV2 (subretinal). Large payload (>5kb) → lentivirus or LNP. Repeat dosing needed → LNP or engineered AAV.
Summary
Gene therapy delivery is the biggest technical challenge in genetic medicine. AAV dominates in vivo gene therapy (tissue-specific serotypes, FDA-approved, but small capacity and expensive). Lentivirus dominates ex vivo therapy (CAR-T, stable integration). LNPs deliver mRNA/CRISPR (no pre-existing immunity, large capacity, but transient). The key challenges: tissue targeting (most vectors accumulate in liver), immunogenicity (70% of adults have AAV antibodies), and manufacturing cost ($1-2M per dose for AAV). Emerging technologies — capsid engineering, targeted LNPs, in vivo CRISPR — are expanding what is possible.
Frequently Asked Questions
AAV manufacturing yield is very low — a single dose may require an entire bioreactor batch. Plus, gene therapies treat rare diseases (small patient populations), so the cost per patient is high. Zolgensma: $2.1M, Hemgenix: $3.5M. Manufacturing improvements and larger patient populations will reduce costs.
With AAV, typically no — the first dose creates antibodies that neutralize subsequent doses. Immunosuppression or engineered capsids may enable redosing in the future. LNPs can be redosed (mRNA vaccines demonstrated this). Lentiviral ex vivo therapy doesn't need redosing (one-time modification).
In vivo: the vector is injected directly into the patient (AAV for eye, liver, CNS). Ex vivo: cells are removed from the patient, genetically modified in the lab, and reinfused (CAR-T, sickle cell). Ex vivo is safer (controlled modification) but more complex logistically.
Three approaches: 1) AAV-CRISPR (viral delivery of Cas9 + gRNA), 2) LNP-CRISPR (lipid delivery of mRNA for Cas9 + gRNA), 3) Ex vivo electroporation of CRISPR ribonucleoprotein (RNP). LNP-CRISPR is emerging as the preferred approach for liver targets; AAV for other tissues.
Test Your Knowledge
1. What is the main advantage of AAV over lentiviral vectors?
AAV has lower immunogenicity and does not integrate into the host genome, reducing the risk of insertional mutagenesis. However, AAV has a smaller packaging capacity (~4.7kb vs ~8kb for lentivirus) and is more expensive to manufacture.
2. Why do most systemically delivered gene therapies accumulate in the liver?
After intravenous injection, LNPs and AAV are cleared from blood by the liver — Kupffer cells and hepatocytes take up particles via receptor-mediated endocytosis. This is both a challenge (for non-liver targets) and an advantage (for liver-targeted therapies).
3. What limits repeat dosing of AAV gene therapy?
The first AAV dose induces antibodies that neutralize subsequent doses. ~70% of adults already have pre-existing AAV antibodies from natural infections, excluding many patients entirely. Engineered capsids and immunosuppression are being developed to overcome this.