Abstract
Since 2006, the carcinogenic potential of antisense oligonucleotides (ASOs) has primarily been evaluated using standard 2-year rodent bioassays. To date, no human-relevant tumor findings have been identified across nonclinical ASO studies. As the field matures and more therapeutics advance into development, there is increasing support for carcinogenicity strategies that built on prior data, leverage platform-based risk assessments, and align with evolving regulatory guidance. In several programs, traditional 2-year mouse studies have been replaced by short-term assessments, such as the 6-month study in transgenic rasH2 mouse study, often conducted alongside a 2-year rat study in accordance with ICH S1B(R1). Historically, deviations from the standard two-species approach have been uncommon and typically limited to severe neurological indications or cases lacking an endogenous rodent target. Nevertheless, for eplontersen and olezarsen, carcinogenicity assessment was limited to a single Tg.rasH2 mouse study or fully waived, supported by prior two-species carcinogenicity data from inotersen and volanesorsen. With the recent ICH S1B(R1) addendum, sponsors now have a defined pathway to pursue carcinogenicity strategies, including potential waivers of the 2-year rat study when supported by the totality of evidence. This article reviews the evolving landscape of ASO carcinogenicity testing and highlights opportunities to reduce or waive long-term studies.
Introduction
Antisense oligonucleotides (ASOs) are synthetic, single-stranded nucleic acids designed to bind to specific mRNA sequences and modulate gene expression. Over the past three decades, they have emerged as a targeted therapeutic class with broad potential across genetic, neurological, metabolic, and cardiovascular diseases. Advances in chemical modifications, such as the incorporation of 2′-O-methoxyethyl (2′-MOE), constrained ethyl (cEt) nucleotides, and GalNAc conjugates, have improved pharmacokinetic properties, reduced off-target effects, and enabled tissue-specific delivery. These developments have not only enhanced therapeutic efficacy but also reduced the occurrence of rodent-specific proinflammatory responses, which are not predictive of human risk.
Traditional carcinogenicity testing has required 2-year studies in rats and mice, as based on ICH M3(R2) guidance. However, the value of doing a two-species carcinogenicity assessment for ASOs has always been unclear since the subcutaneous route of administration provides complete bioavailability, and the pharmacokinetics, toxicities, and metabolism are very similar between mice and rats. Furthermore, rodent-specific findings, particularly those lacking mechanistic relevance, can complicate risk interpretation without providing meaningful insight into human safety.
Carcinogenicity Assessment Experience With ASOs
Across multiple ASO programs, nonclinical studies have consistently demonstrated a low potential for carcinogenicity. Notably, while dose-dependent or treatment-related increases in tumor incidence have occasionally been observed in long-term rodent studies, these findings have been determined to be rodent-specific and not relevant to humans based on increased sensitivity of mice to specific effects, such as pro-inflammatory effects. 3 Overall, no consistent or biologically meaningful tumor responses have been identified across studies.
Examples of approved ASO products include mipomersen, volanesorsen, olezarsen, inotersen, eplontersen, and donidalorsen. For these ASOs, repeat-dose chronic toxicity studies, including 6- and 9-month studies in non-human primates, revealed no evidence of preneoplastic or neoplastic lesions. However, in some earlier ASO programs, rodent carcinogenicity studies identified hepatocellular tumors, underscoring the importance of a case-by-case evaluation when assessing carcinogenicity risk. These findings are generally considered to have low human relevance, based on mechanistic and species-specific factors. 2 Furthermore, ASOs have shown no genotoxic potential in either standard in vitro or in vivo assays. While genotoxicity is a common mechanism of carcinogenesis, these findings support the view that ASOs are unlikely to act through DNA-reactive carcinogenic pathways. This is further supported by a recent study evaluating the carcinogenic potential of inotersen in both Tg.rasH2 mice and Sprague-Dawley rats, which found no evidence of treatment-related neoplastic changes. 6 Additional nonclinical studies, including those with donidalorsen, have shown similar results, further supporting the lack of carcinogenic potential for ASOs with similar 2'-MOE chemistry.
In several regulatory interactions, agencies have granted waivers for the standard 2-year rat bioassay or accepted deferral to the post marketing requirements (PMRs). These decisions were informed by the overall strength of the nonclinical safety data and the absence of tumor-promoting signals in chronic toxicity studies as well as the context of rare neurological disease indications with limited patient populations and/or lack of proliferative target biology, supporting a lower concern for carcinogenic potential in those specific cases. However, for agencies such as the FDA, such deferrals are typically indication-driven and should not be interpreted as a general conclusion that ASOs carry inherently low tumorigenic risk. In fact, the FDA’s draft guidance on nonclinical safety assessment of oligonucleotide-based therapeutics outlines a potential path to single-species carcinogenicity assessment, but only for ASOs with well-characterized chemistries and adequate supporting data, reinforcing the need for a case-by-case, risk-based approach.
Weight of Evidence and Regulatory Strategy
The Weight of Evidence (WoE) approach has become a cornerstone of modern carcinogenicity assessment for ASOs, in alignment with evolving regulatory guidance. Both global and draft regulatory guidelines (ICH S1B(R1) and S1B(R1) Addendum) increasingly support the use of integrated, science-based assessments to determine the need for long-term carcinogenicity studies, particularly for modalities like ASOs that lack structural alerts or mechanistic concerns.
This framework integrates multiple lines of data to assess carcinogenic potential and determine whether carcinogenicity studies are scientifically warranted.
This strategic integration of mechanistic understanding, chronic safety data, and prior regulatory experience enables Sponsors to justify alternative carcinogenicity testing strategies. These often involve the use of a single-species model, such as the Tg.rasH2 mouse model, or exclusion of a study altogether in the absence of risk indicators.
Ethical and Scientific Considerations
Beyond scientific justification, evolving carcinogenicity strategies also reflect ethical considerations. In line with the 3Rs principle (Reduce, Replace, Refine), avoiding unnecessary, long-term rodent studies helps reduce animal use while preserving the integrity of safety evaluation. Long-term bioassays are not only resource-intensive but may also produce ambiguous or misleading results, especially when unrelated to human biology.
In many cases, the totality of chronic toxicity data, combined with genotoxicity and pharmacological profiling, provides a more predictive and relevant safety signal than 2-year carcinogenicity studies in rats. A growing number of regulatory authorities have acknowledged this, encouraging a more rational, risk-based approach to carcinogenicity testing for oligonucleotides and other non-traditional modalities. When appropriate, use of the 6-month Tg.rasH2 mouse model, alongside robust chronic toxicity and mechanistic data, can offer a more timely and scientifically relevant assessment of carcinogenic risk than defaulting to a 2-year rodent bioassay.
Conclusions
The evolution of ASO design, chemistry, and accumulated safety data supports a progressive refinement in the approach to carcinogenicity assessment. While traditional 2-year rodent bioassays have historically contributed negative data that aided regulatory decision-making, a growing body of evidence suggests that their predictive value for human risk assessment may be limited, particularly in the context of ASOs lacking genotoxicity, hormonal disruption, or immune-mediated carcinogenic potential.
Only in recent years has the field accumulated sufficient nonclinical and regulatory experience to justify platform-based risk assessments. This has enabled more efficient and scientifically grounded strategies that minimize unnecessary animal use while preserving confidence in safety evaluation.
A WoE approach that integrates target biology, genotoxicity testing, chronic toxicity data, and mechanistic insight offers a more predictive and ethical framework for evaluating carcinogenic potential. Regulatory precedent, including accepted waivers and use of the Tg.rasH2 mouse model, reinforces that long-term studies in two rodent species are not always required, particularly when supported by a robust and consistent nonclinical package.
As regulatory expectations continue to evolve, platform-based safety assessments and case-by-case evaluation will remain essential for ensuring scientific integrity, patient safety, and alignment with the 3Rs principle in nonclinical development of ASO therapeutics.
Footnotes
Acknowledgements
The author thanks Drs Henry, Wange, Chen, You, and Zanardi for their critical reviews.
Author Contributions
The author contributed to Writing – original draft; Investigation; and Conceptualization (TWK).
Declaration of Conflicting Interests
The author declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The author is an employee of Ionis Pharmaceuticals, Inc. The research described was supported by Ionis Pharmaceuticals.
Funding
The author received no financial support for the research, authorship, and/or publication of this article.
