The Off-Target Problem the Gene Editing Industry Can No Longer Afford to Ignore

Gene editing has changed dramatically over the last decade. Early innovation focused on proving technologies such as CRISPR could edit DNA accurately enough to become viable therapeutic tools. However, today with hundreds of programmes progressing through discovery and preclinical development, new editing modalities continue to emerge, and the focus is no longer simply can we edit DNA? It's can we do it safely? 

One persistent challenge remains and that’s off-target activity. Despite remarkable advances in genome editing, unintended DNA modifications remain one of the biggest obstacles facing developers. While editing efficiency often dominates early programme decisions, understanding where else an editor may interact with the genome has become just as important. As more therapies move towards the clinic, comprehensive off-target characterisation is no longer simply good scientific practice, it’s becoming fundamental to programme success. 

Why off-target activity still matters 

Every genome editor works by recognising a DNA sequence and making a precise genetic change. However, no editor is perfectly specific. 

For nuclease-based systems such as CRISPR-Cas9 and Cas12, unintended DNA double-strand breaks can occur at sites that closely resemble the intended target sequence. Although many of these events may have little biological consequence, others can disrupt gene function, alter gene regulation or contribute to larger structural changes within the genome. 

Importantly, these risks are not confined to traditional CRISPR nucleases. Base editors and prime editors were developed to improve precision and avoid creating double-strand breaks, yet they introduce their own mechanisms of unintended editing. Different technologies present different safety challenges, but none remove the need for careful characterisation. For therapeutic developers, the question is no longer whether off-target effects exist, it's whether they have been characterized well enough to make confident decisions. 

The science has moved on. Some screening methods haven't. 

Gene editing technologies have evolved rapidly over the past decade, and the methods used to evaluate off-target activity haven't always kept pace. Many established workflows were developed for an earlier generation of editors and rely on indirect measurements of editing activity. Some identify candidate sites in purified DNA outside the cellular environment. Others depend on DNA repair events or PCR amplification to infer where breaks may have occurred. 

The existing methods aren't poor science. Many were developed to answer specific questions and have made important contributions to the field. The challenge is that gene editing has moved on and researchers are now asking different questions from the ones they were trying to answer ten years ago. 

The biological factors that influence where editing occurs cannot always be reproduced using cell-free systems. Amplification-based workflows can introduce bias that makes low-frequency events more difficult to interpret. Finding off-target sites is only part of the challenge. Researchers also need confidence that the data reflects what actually happened inside the cells they intend to treat.  

Why timing is becoming just as important as sensitivity 

Historically, off-target assessment has often been treated as a late-stage exercise, performed once lead candidates have already been selected and this approach is beginning to look outdated. 

Modern discovery programmes may compare hundreds of guide RNAs, multiple nuclease variants and different delivery strategies before selecting a development candidate. If comprehensive off-target data isn't available until much later, teams risk investing significant time and resources in candidates that ultimately prove less suitable than alternatives. 

Early access to reliable genome-wide data changes that process. Instead of selecting candidates primarily on editing efficiency, researchers can balance efficacy with specificity throughout development. Potential safety concerns can be identified earlier, guide RNAs refined more efficiently, and stronger candidates progressed with greater confidence and with that off-target analysis becomes part of discovery rather than simply a regulatory requirement. 

What should modern off-target screening look like? 

While no single assay answers every question, there is growing agreement about what developers need from an off-target screening method. 

Researchers increasingly need approaches that are: 

• Genome-wide and unbiased, capable of identifying unexpected editing events rather than relying solely on prediction.  

• Cell-based, so editing is measured within biologically relevant systems.  

• Direct, detecting DNA breaks themselves rather than downstream proxies.  

• Free from PCR amplification bias, allowing more quantitative interpretation of editing frequency.  

• Fast enough to support iterative development, with results available while decisions can still influence programme design.  

These same principles are reflected in recent regulatory guidance, which places growing emphasis on genome-wide characterisation performed in therapeutically relevant cells using next-generation sequencing approaches.  

Bringing better data into earlier decisions 

This is exactly the challenge that guided the development of INDUCE-seq®. 

Rather than inferring editing activity through indirect readouts, INDUCE-seq® directly captures and sequences DNA break ends in situ within intact cells. Because the workflow is PCR-free, each sequencing read corresponds to a single labelled DNA break, providing a direct and quantitative measure of editing activity across the genome. 

One advantage of analysing on- and off-target activity within the same workflow is that it provides a more complete picture of editor behaviour. Combined with integrated bioinformatics and decision-ready outputs, INDUCE-seq® enables researchers to compare guide RNAs, evaluate nuclease variants and understand specificity without stitching together results from multiple independent assays. 

Looking ahead 

Gene editing continues to evolve rapidly. New editing systems are emerging, therapeutic programmes are becoming more ambitious and regulatory expectations continue to evolve alongside them. 

By producing actionable data in days rather than weeks, INDUCE-seq® makes it practical to incorporate comprehensive off-target analysis into routine discovery workflows, when those insights can still influence programme decisions. 

Off-target analysis has become much more than a box to tick before an IND submission. It now informs decisions throughout discovery, from guide RNA selection to candidate prioritisation and editor optimisation. As gene editing technologies continue to evolve, the value of generating reliable, biologically relevant off-target data early in development will only increase.