By Victor Lien

In vivo gene editing is positioned to fundamentally reshape the treatment of cancer, moving beyond conventional chemotherapy and radiation to enable direct, precise modifications of a patient’s DNA of the target cells of the organs. This approach eliminates the complex, time-consuming steps of ex vivo manipulation, which requires cell extraction, laboratory modification, and reinfusion, such as autologous cell therapy. At the core of this revolutionary potential is the CRISPR-Cas9 system, programmable molecular scissors with the promise of curing previously intractable cancers.

This is part 1 of a series exploring the latest in gene therapy delivery. Part 2 explores the critical role CDMOs play in clearing obstacles in manufacturing supply.

However, the journey from laboratory breakthrough to universally available therapy is constrained by a triad of formidable challenges: biological complexity, the need for sophisticated and safe delivery systems, and the creation of a reliable, high-volume manufacturing infrastructure capable of scaling production. These challenges are intrinsically linked; the choice of delivery dictates the complexity of manufacturing, which in turn determines the eventual cost and accessibility of the therapy.

This expanded analysis explores the foundational technologies, details the leading delivery platforms, dissects the innovative therapeutic strategies against cancer, and analyzes the critical manufacturing and supply chain bottlenecks that currently define and will shape the future trajectory of in vivo gene editing in oncology.

CRISPR-Cas9: The Engine For Precision Editing

The system operates based on two essential components, which together form an active ribonucleoprotein (RNP) complex:

1. Guide RNA (gRNA): This short synthetic ribonucleic acid sequence, typically about 20 nucleotides long, acts as the molecular homing device. It is designed to be fully complementary to a specific unique 20-base pair target sequence in the vastness of the host cell’s genome. This complementary base pairing guides the enzyme to the exact chromosomal location designated for editing.
2. Cas9 Nuclease: This programmable enzyme is derived from Streptococcus pyogenes (known as SpCas9) or other bacteria. It functions as molecular scissors. Once the gRNA has successfully located and annealed to the target DNA sequence, the Cas9 enzyme introduces a clean double-strand break (DSB) at that precise location. The cut only occurs if a short adjacent sequence known as the protospacer adjacent motif (PAM) is present, which ensures the enzyme only cuts foreign DNA and not the host genome.

The introduction of the DSB is the signal for the cell’s natural DNA repair mechanisms to take over. This is the stage where researchers direct the editing outcome, which is crucial for achieving different therapeutic goals in oncology:

• Non-homologous end joining (NHEJ): This spontaneous cellular mechanism rapidly repairs DNA errors by ligating the two severed DNA ends together but often introduces small insertions or deletions, or indels, at the break site. This process effectively results in a gene knockout, disabling the function of the targeted gene, which is the mechanism used for blocking genes like programmed death-1 (PD-1) in T cells or disrupting oncogenes in tumor cells.
• Homology-directed repair (HDR): This high-fidelity precise repair mechanism occurs less frequently in non-dividing cells. By providing a synthetic donor DNA template alongside the RNP complex, researchers can instruct the cell to use the template as a guide to accurately repair the DSB. This allows for the precise correction of a mutation or the insertion of an entirely new genetic sequence, such as a chimeric antigen receptor (CAR) gene, into a T cell genome.

The modularity, specificity, and efficiency of the CRISPR-Cas9 system have firmly established it as the leading platform for gene editing. Furthermore, its ongoing evolution, including the development of base editors, which correct single-base point mutations without a DSB, and prime editors, which enable versatile short sequence insertions and deletions, promises even higher precision with reduced off-target risks, expanding its therapeutic application from monogenic disease correction to sophisticated immuno-oncology strategies.

The Delivery Mission: Getting The Tool To The Worksite With Safety And Efficiency

The therapeutic success of in vivo gene editing hinges on one critical factor: the efficient and safe delivery of the CRISPR-Cas9 components into the nucleus of target cells. The human body is inherently hostile to foreign genetic material, deploying immune responses and enzymatic degradation.

While numerous delivery platforms are in development — including polymer nanoparticles, virus-like particles, and electroporation — two have emerged as clinically advanced leaders: adeno-associated virus (AAV) and messenger RNA (mRNA), delivered via lipid nanoparticles (LNPs). Their dominance is not accidental; it stems from a powerful combination of regulatory familiarity, manufacturing maturity, and compelling clinical precedent that alternative platforms currently lack.

AAV vectors have a long track record in gene therapy, with multiple FDA-approved products for monogenic diseases. This has created a well-understood, though complex, regulatory pathway for their use. Their primary advantage is the ability to provide long-lasting expression from a single dose, making them ideal for targeting slowly-dividing tissues.

The mRNA/LNP platform, while newer to gene editing, gained global validation and unprecedented manufacturing scale through the COVID-19 vaccines. This demonstrated the platform’s safety, its suitability for rapid and scalable chemical synthesis, and its key safety feature for gene editing: transient expression that reduces off-target risks and allows for repeat dosing.

This confluence of proven clinical utility, scalable production, and a clear regulatory road map has made AAV and mRNA/LNP the default choices for most early-stage in vivo gene editing programs.

AAV: Durable Expression With Packaging Limitations

AAV vectors utilize engineered non-pathogenic viral shells to deliver a DNA blueprint encoding the Cas9 protein and its gRNA. The DNA cargo, once delivered to the nucleus, typically persists as a stable non-integrating circular piece of DNA known as an episome. This mechanism results in long-term, potentially permanent, expression of the editing machinery from a single dose, making AAV the ideal choice for targeting slowly dividing or non-dividing tissues, such as the liver or the retina, for sustained therapeutic effect. The long track record of AAV in traditional gene therapy, with multiple FDA-approved products for monogenic diseases, provides a known, though complex, regulatory path.

The most critical limitation of AAV, however, is its cramped packaging capacity of approximately 4.7 kb. The commonly used SpCas9 gene, along with its requisite promoter and gRNA, often exceeds this limit, posing a major logistical challenge. Developers employ sophisticated genetic engineering strategies to overcome this packaging constraint:

• All-in-one vector with smaller Cas9 orthologs: Researchers utilize compact Cas9 enzymes from other bacterial species, such as Staphylococcus aureus Cas9 (SaCas9), which is significantly smaller at about 3.2 kb. This reduction in size allows all necessary components — the gene, the promoter, and the gRNA — to fit within a single AAV vector, simplifying the delivery process.
• Dual AAV system: When the full SpCas9 sequence is required, the coding sequence, along with the gRNA and promoter, is often separated into two distinct AAV vectors. The success of the edit hinges on the cell being co-transduced by both vectors simultaneously, which significantly lowers the overall editing efficiency in vivo due to the probabilistic nature of dual infection.
• Split-Cas9 system: In an even more technically complex approach, the Cas9 gene itself is divided into two nonfunctional segments, which are packaged into separate AAVs. Inside the target cell, the full functional Cas9 protein is reassembled through a self-splicing biological process called intein-mediated protein splicing.

While AAV provides the advantage of durable expression, its immunogenicity remains a major obstacle. Many patients possess preexisting neutralizing antibodies to common AAV serotypes from prior environmental exposure, which can neutralize the therapy before it reaches the target tissue. Furthermore, the initial dose itself can trigger a robust cellular immune response against the viral capsid, which effectively prevents any future re-dosing — a severe limitation in oncology where repeated or maintenance interventions may be essential. Ongoing research into capsid engineering is focused on creating novel serotypes that evade the immune system and exhibit enhanced tissue tropism.

mRNA/LNP: Transient, Repeatable, And Non-Viral

In contrast to AAV, the mRNA/LNP platform is a non-viral transient delivery system. Instead of delivering a DNA blueprint that requires transcription in the nucleus, LNPs carry preformed mRNA that directly encodes the Cas9 protein. This Cas9 mRNA bypasses the nucleus and is translated directly into the functional protein by ribosomes in the cytoplasm.

This transient expression, typically lasting from a few days to a few weeks before the mRNA and protein are naturally degraded, is a paramount safety feature. It drastically shortens the window during which the Cas9 enzyme is active, thereby reducing the probability of off-target editing (unintended cuts at non-target sites). Crucially, the non-viral nature allows for repeat dosing, as the immune response to the LNP itself is generally less severe and easier to manage than the robust antiviral response elicited by AAV. The LNP platform gained global validation and unprecedented manufacturing scale through its use in COVID-19 vaccines, proving its safety, suitability for rapid chemical synthesis, and regulatory acceptability.

The LNP delivery mechanism is a multi-stage process:

1. Systemic administration and protection: LNP-mRNA/sgRNA formulations are administered intravenously. The LNP’s structure is critical for survival and function. It consists of ionizable lipids for binding RNA and enabling endosomal escape, structural lipids like cholesterol for integrity, and PEG-lipids that form a hydrophilic exterior to prevent aggregation.
2. Cellular uptake: LNPs are naturally preferentially taken up by hepatocytes in the liver via endocytosis, though targeting ligands can be added for other tissues.
3. Endosomal escape: The LNP is trapped in an endosome, which acidifies. This acidic environment re-ionizes the lipids, disrupting the endosomal membrane and releasing the mRNA cargo into the cytoplasm — the most critical and often rate-limiting step.
4. Editing machinery assembly and nuclear translocation: The released Cas9 mRNA is translated into protein by the host cell’s ribosomes. This Cas9 protein then binds to the separately delivered single-guide RNA (sgRNA) to form the active RNP complex. Finally, the complex enters the nucleus through the nuclear pore complex to perform the targeted genetic edit.

Optimization strategies for mRNA/LNP systems are rapidly advancing and include co-encapsulation of Cas9 mRNA and gRNA, chemical modification of RNA to enhance stability, and the engineering of novel ionizable lipids for more efficient endosomal escape. The use of self-amplifying RNA (saRNA) is also being explored to extend the duration of Cas9 expression from a single smaller dose, bridging the gap toward the durability offered by AAV.

Conquering Cancer: Therapeutic Strategies And Delivery Hurdles

In vivo gene editing offers a transformative approach to cancer therapy by enabling precise genetic modifications directly within the patient’s body. These interventions primarily fall into two categories: enhancing the immune system’s ability to fight cancer and directly attacking the genetic drivers of the tumor itself.

Immuno-oncology: editing immune cells to fight cancer

The most clinically advanced applications of gene editing in oncology currently involve ex vivo manipulation — where a patient’s T cells are extracted, edited in a lab, and reinfused. However, the field is aggressively pursuing in vivo methods to achieve the same results more simply and scalable.

Key in vivo editing strategies targeting the patient's immune cells include -

• PD-1 knockout: PD-1 is an immune checkpoint protein that, when activated by cancer cells, acts as a brake on T cell activity. Using CRISPR-Cas9 to knock out the PDCD1 gene in T cells can prevent this dampening, leading to sustained cytotoxic activity against tumors.
• In vivo CAR-T engineering: Instead of ex vivo manufacturing, the genes necessary for a CAR are delivered in vivo. The Cas9 RNP complex is programmed to insert the CAR gene directly into a specific location within the T cell genome, often at a highly expressed locus like the T cell receptor alpha constant (TRAC) locus. This not only rewires the T cell to recognize tumor-associated antigens but can simultaneously knock out the endogenous T cell receptors (TCRs), reducing the risk of unwanted immune reactions, such as graft-versus-host disease (GvHD) in allogeneic, or off-the-shelf, therapies.

The central delivery challenge for in vivo immuno-oncology is targeting specificity. To be successful, LNPs or AAVs must be engineered to home exclusively to T cells or other desired immune cell populations, such as natural killer (NK) cells, within the vast circulating volume of the human body. Achieving this level of cellular exclusivity while avoiding off-target editing of other highly metabolically active cells, like hepatocytes, is paramount. Solutions being explored involve attaching cell-specific targeting ligands (e.g., antibodies or peptides that bind to unique surface receptors) to the exterior of LNPs or engineering novel T cell-tropic AAV capsids that naturally prefer lymphocytes.

Direct tumor editing: correcting genetic drivers in vivo

Beyond immune modulation, in vivo gene editing is being explored as a direct intervention within tumor cells, aiming to disrupt the very genes that cause malignancy.

Emerging tactics include:

• Oncogene disruption: Driver mutations in genes like KRAS, EGFR, and MYC propel uncontrolled cell growth. CRISPR-Cas9 can be programmed to introduce disruptive double-strand breaks into these oncogenes, effectively halting tumor growth.
• Tumor suppressor restoration: Genes like TP53 and RB1 are frequently inactivated in cancer. Restoring their function through targeted gene correction can reestablish cell cycle control and trigger apoptosis, or programmed cell death, in malignant cells.
• Synthetic lethality: This approach involves editing a gene that is nonessential in healthy cells but becomes essential for survival in the context of a specific cancer mutation. This induces selective cell death only in tumor cells, sparing healthy tissue.

The delivery challenges for direct tumor editing are formidable, centered on the hostile tumor microenvironment (TME).

• Physical barriers: Dense stroma, chaotic and leaky vasculature, and elevated interstitial pressure physically hinder the uniform penetration and distribution of editing components.
• Chemical barriers: Hypoxic, or low oxygen, and acidic conditions within the TME can impair LNP stability and function.
• Cellular kinetics and heterogeneity: Rapid tumor cell division can dilute the therapeutic payload, and the immense genetic diversity within a single tumor, known as heterogeneity, means that a single edit may not be effective against all cells. The lack of universal surface markers further complicates ligand-based targeting strategies.

In part 1 of this article, we outlined the gene editing and delivery mechanisms for in vivo gene therapy and the MOA of how they can be applied to various cancerous indications. In the second part of this article, we will explain how these intrinsic therapeutic complexities would translate to manufacturing bottlenecks and CDMO constraints.

About The Author:

Victor Lien, Ph.D., is a CMC strategy and platform developer with extensive experience across several advanced modalities, including antibody-drug conjugates, cell and gene therapies, and mRNA. He most recently worked as a director of biotherapeutic pharmaceutical sciences at Pfizer. Before that, he was a director of CMC and tech transfer at Gracell Biotechnologies, now a part of AstraZeneca. Other experience includes leadership positions at Bristol Myers Squibb, DowDuPont, Fluidigm, and Vertex Pharmaceuticals. He was also a systems biology instructor at Harvard Medical School. He received his Ph.D. from the University of California San Diego.