The original CRISPR-Cas9 gene editing system that captured public imagination in the 2010s works by making double-stranded breaks in DNA at targeted sequences. The cell's own repair machinery then closes the break, typically introducing small insertions or deletions that disrupt the targeted gene. This works for many therapeutic applications, but it has meaningful limitations — the DNA break creates opportunities for off-target editing, and the repair outcome is imprecise, producing variable end results.

Two newer approaches — base editing and prime editing — have addressed these limitations by making more precise molecular changes without full DNA breaks. As these technologies have matured and entered clinical trials, they've become the preferred approaches for applications requiring precise nucleotide-level changes. This piece examines what technically distinguishes these approaches, why the distinctions matter for therapeutic applications, and where the different technologies excel.

The Original CRISPR-Cas9 Approach

Traditional CRISPR-Cas9 gene editing follows a simple mechanism: a guide RNA directs the Cas9 protein to a specific DNA sequence, where Cas9 creates a double-stranded break. The cell repairs the break through one of two pathways — non-homologous end joining (NHEJ), which typically introduces small errors that disrupt the targeted gene, or homology-directed repair (HDR), which can incorporate a template DNA sequence if one is provided.

This approach works well for gene knockout applications where disrupting a target gene is the goal. It's the mechanism used in Casgevy for sickle cell disease — the treatment disrupts BCL11A gene function to allow fetal hemoglobin expression. The imprecise repair outcome is acceptable because the therapeutic effect requires only gene disruption, not specific nucleotide changes.

The limitations become significant when specific nucleotide changes are needed. Correcting a specific point mutation, for example, requires precise base substitution that Cas9's blunt-force approach doesn't provide efficiently.

Base Editing: Precise Single-Nucleotide Changes

Base editing, developed in David Liu's lab at Harvard beginning around 2016, offers a fundamentally different approach. Rather than making DNA breaks, base editors use a modified Cas9 (which finds the target sequence but doesn't cut) fused to an enzyme that chemically converts one DNA base into another.

Two main types have emerged:

Cytosine base editors (CBE) convert cytosine (C) to thymine (T). This addresses point mutations where a C-to-T change corrects a disease-causing mutation.

Adenine base editors (ABE) convert adenine (A) to guanine (G). This addresses a complementary set of point mutations.

The therapeutic implications are significant. Approximately half of known disease-causing point mutations could theoretically be corrected by base editing without the off-target risks of full DNA breaks. Verve Therapeutics' familial hypercholesterolemia program uses base editing to correct the PCSK9 gene, and several other clinical programs use similar approaches for specific hereditary diseases.

Base Editing Limitations

Base editing has meaningful limitations. The available base conversions are limited to specific chemistries. Delivery challenges remain similar to traditional CRISPR. Some off-target editing still occurs, though at substantially lower rates than double-strand-break approaches. And some target sequences don't fit the necessary PAM (protospacer adjacent motif) requirements for current editor variants.

Prime Editing: The Molecular Word Processor

Prime editing, developed in the Liu lab beginning 2019, offers even greater precision. Prime editors use a Cas9 nickase (which cuts one DNA strand rather than both) fused to a reverse transcriptase. The system uses an extended guide RNA that includes both the target sequence and the desired edit sequence, allowing the reverse transcriptase to write in the specified edit.

The technical advantages are substantial. Prime editing can make all 12 possible base-to-base conversions (not just the two available with base editing). It can perform small insertions and deletions with high precision. It reduces off-target effects further than base editing. And it works on target sequences that base editing cannot access.

The mechanism is more complex than either traditional CRISPR or base editing, which has translated to slower clinical development. But the therapeutic potential is broader. Prime Medicine and others have advanced multiple clinical programs, with the first prime editing trials beginning in 2024-2025 for diseases including chronic granulomatous disease and Wilson disease.

The gene editing field has evolved from "we can cut DNA at specific locations" to "we can make specific molecular changes at specific locations" — a substantial capability increase that translates to broader therapeutic potential.

The Clinical Development Comparison

As of mid-2026, the three approaches have different clinical development profiles:

Traditional CRISPR-Cas9 has the most mature clinical presence with Casgevy approved and several late-stage trials. Application is limited to gene disruption approaches (like sickle cell) or knockin approaches requiring template DNA.

Base editing has advanced to Phase 3 trials for familial hypercholesterolemia (Verve) and multiple mid-stage trials for other conditions. Regulatory approval expected in 2027-2028 for lead indications.

Prime editing is in earlier clinical development with Phase 1/2 trials beginning through 2025. First regulatory approvals not expected before 2029.

The clinical timeline gap partly reflects when each technology was developed but also reflects increasing manufacturing and delivery complexity. Newer approaches have more moving parts and more sophisticated delivery requirements.

The Off-Target Question

The most-scrutinized safety concern across gene editing approaches involves off-target effects — editing at unintended DNA sequences that share similarity with the target. Even low-frequency off-target editing raises concerns about long-term consequences including potential cancer risk.

The off-target rates by approach:

The therapeutic implications matter. For treating patients with life-threatening diseases where the alternative is death or severe disability, higher off-target rates may be acceptable. For treating milder conditions or elective applications, lower off-target rates become more important. This creates a natural sorting: aggressive treatments for severe conditions use whatever tool works, while more speculative applications (including any anti-aging use) will require the most precise approaches.

The Delivery Question Remains Central

All three editing approaches share the same fundamental delivery challenge: getting the editing components into the specific cells that need editing, at high enough efficiency to produce therapeutic effect, without triggering excessive immune response or off-target effects.

Current solutions vary by target tissue:

Blood cells — ex vivo editing followed by autologous transplantation. Works well but treatment complex.

Liver — lipid nanoparticle delivery. Increasingly successful, enabling in vivo treatment approaches.

Eye — direct intraocular injection. Accessible tissue with immune-privileged environment.

Muscle — AAV vector delivery. Works for some applications but immune response limitations.

Brain, most other tissues — no reliable delivery method for widespread therapeutic use.

The delivery limitations affect all three editing approaches equally. Newer editing chemistries don't help if the components can't reach target cells.

What This Means for Therapeutic Landscape

The evolution from traditional CRISPR to base editing to prime editing represents genuine capability expansion but shouldn't be conflated with treatment availability expansion. Most patients with genetic diseases in 2026 still don't have access to gene editing treatments regardless of which editing approach might theoretically work for their specific mutation.

The most impactful near-term applications remain in categories where existing approaches work well: hemoglobinopathies (traditional CRISPR), specific liver-targeted diseases (base editing), and eye-targeted conditions (multiple approaches). Broader disease targets await delivery innovations that haven't yet emerged.

The Realistic Timeline

For patients hoping that emerging gene editing approaches will address their specific condition:

These timelines assume continued technical progress and no major setbacks. Real-world timelines have consistently run longer than initial projections in gene therapy development.

The gene editing field has matured substantially since 2020. The therapeutic possibilities are real. The timeline to broad access is longer than enthusiasm suggests. Both truths coexist, and honest engagement requires acknowledging both.

Dr. Sarah Whitcomb has no financial relationships with any gene editing company mentioned in this article. TimesWriter editorial standards require disclosure of author conflicts of interest.