What the FDA’s New Guidance Means for CRISPR off-Target Analysis

When the FDA released its draft guidance in April 2026, it marked an important step for the gene editing field. For the first time, developers had a detailed view of how the agency expects off-target risk to be assessed as programmes move towards the clinic. 

The guidance doesn't prescribe a single technology, nor does it dismiss the methods already used across the industry. Instead, it shifts the conversation. Rather than asking whether off-target sites can be found, it asks how confidently developers can demonstrate what is actually happening in the cells they intend to treat. This distinction has implications for study design, data interpretation and, ultimately, the evidence submitted to regulators. 

Why now?

Gene editing has moved on quickly. What was once largely confined to academic research is now underpinning an increasing number of clinical programmes across cell and gene therapy. Alongside CRISPR-Cas9, developers are working with Cas12a, base editors, prime editors and engineered Cas nuclease variants, each bringing different opportunities and different safety considerations. 

As these technologies mature, regulators need confidence that developers understand whether editing has occurred, where unintended DNA breaks may have happened, and what that means for patient safety. 

The FDA's guidance reflects that shift. It acknowledges the progress that has been made in off-target detection while recognising that different methods answer different scientific questions. Rather than recommending a single assay, it encourages developers to select approaches that are appropriate for their programme and to understand the strengths and limitations of the data they generate.  

What does the guidance say?

One of the clearest themes running through the document is the importance of generating evidence that reflects biologically relevant conditions. 

Historically, many off-target workflows have relied on computational prediction, biochemical assays or targeted validation of candidate sites. These approaches continue to have value and remain useful in many development programmes. However, prediction alone cannot confirm whether editing has occurred in living cells, while targeted approaches can only investigate the sites selected for analysis. 

The FDA places particular emphasis on genome-wide assessment using next-generation sequencing (NGS)-based methods capable of identifying off-target editing events in therapeutically relevant cells. This reflects a broader move towards experimental evidence rather than relying solely on prediction. 

The guidance also recognises that no single assay answers every question. Developers should understand the limitations of each method they use and, where appropriate, combine complementary approaches to build a robust evidence package. 

What does this mean for developers?

For many organisations, the biggest change is not necessarily adopting a new technology but changing when and how off-target assessment is performed. 

Generating high-quality off-target data earlier in development can influence decisions throughout a programme. It can help compare guide RNAs, evaluate different editor variants, understand the impact of delivery methods and support selection of the most suitable candidate before significant resources are committed. 

By the time a programme reaches IND-enabling studies, developers are no longer simply asking whether an editor works. They need confidence that its safety profile has been characterised using evidence that reflects the biology of the intended therapeutic system. 

This places greater importance on study design. Choosing relevant cell models, selecting appropriate controls and understanding what each assay can and cannot measure all become essential parts of building a regulatory data package.

Looking beyond compliance

The FDA guidance should not be viewed as a checklist of regulatory requirements. Instead, it provides a useful indication of the direction in which the field is moving. 

The emphasis is increasingly on unbiased, genome-wide evidence generated in biologically relevant systems and supported by a clear understanding of assay performance. These principles are valuable not only for regulatory submissions but throughout discovery and preclinical development, where better data can improve decision making and reduce downstream risk. As gene editing technologies continue to evolve, confidence in safety data will become just as important as confidence in editing efficiency.

Turning guidance into action

For therapeutic developers, meeting evolving regulatory expectations starts with generating data they can trust. That means selecting methods that produce reliable, genome-wide evidence in therapeutically relevant cells and understanding both the strengths and limitations of the data those methods generate. 

This is where technologies such as INDUCE-seq® can play an important role. Unlike approaches that rely on prediction or indirect measurements, INDUCE-seq directly maps DNA double-strand breaks in edited cells, providing an unbiased, genome-wide view of both on- and off-target activity. Because it captures DNA breaks in situ without PCR amplification, it enables researchers to evaluate editing outcomes with high sensitivity while avoiding biases introduced by amplification-based workflows. 

From early guide RNA screening and editor optimisation through to IND-enabling studies, this type of evidence can help researchers compare editing strategies, understand off-target profiles in therapeutically relevant cells and build confidence in the decisions they make as programmes progress. 

Importantly, the FDA does not recommend one specific assay over another. Instead, the guidance encourages developers to select methods that are appropriate for their application and to understand the quality and limitations of the resulting data. Technologies capable of generating unbiased, genome-wide datasets in biologically relevant systems are therefore well placed to support these expectations.

Looking ahead

The publication of the FDA's draft guidance is unlikely to be the last evolution in regulatory thinking around genome editing safety. As more therapies enter clinical development, expectations around the quality, reproducibility and biological relevance of off-target data will continue to develop. 

For researchers, that presents an opportunity rather than simply another regulatory hurdle. Generating robust off-target data earlier in development can improve candidate selection, strengthen preclinical decision making and reduce uncertainty before programmes reach the clinic. 

Ultimately, the guidance reinforces a principle that has always underpinned good science: the more accurately we understand what is happening inside edited cells, the better equipped we are to develop safer and more effective gene editing therapies.