Oligonucleotide Therapeutics (ONTs) are a rapidly expanding class of drugs and may hold the key to unlocking personalized medicine. Recent advancements, particularly in stability and delivery systems, have accelerated the development of oligo-based therapies, offering hope to patients with genetic disorders and other conditions once considered “undruggable.”
RNA-targeting ONTs are short, synthetic strands of nucleic acids, typically 15-30 nucleotides in length. They intervene at the level of genetic information rather than protein activity, treating conditions by reshaping cellular messages instead of reacting to downstream damage.
To achieve this effect, RNA-targeting ONTs rely on programmable specificity, meaning they are designed to bind precisely to a specific complementary RNA target through Watson-Crick base pairing. Chemical modifications to their structure can improve resistance to degradation and increase potency. These include backbone changes (such as phosphorothioate linkages) and sugar modifications (such as 2′-O-methoxyethyl, or 2′-O-MOE), which enhance molecular stability. Effective delivery is also critical, with platforms such as N-acetylgalactosamine (GalNAc) conjugates and lipid nanoparticles (LNPs) serving as key examples.
Yet the same features that make ONTs so promising—their programmable specificity, chemical complexity, and, in many cases, the need for effective delivery systems—also introduce distinct safety considerations that must be addressed throughout development. Understanding and anticipating these risks is essential to translating oligonucleotide innovation into safe, effective therapies.
Understanding How ONTs Work
To assess the safety of oligonucleotide therapeutics effectively, it is first necessary to understand how they work. The two major classes of RNA-targeting ONTs are Antisense Oligonucleotides (ASOs) and Small Interfering RNAs (siRNAs). While they differ in structure and mechanism, both act through sequence-directed RNA modulation. Consequently, the same features that enable efficacy can also drive toxicity.
- ASOs are single-stranded, chemically modified synthetic nucleic acids. Their primary mechanisms include steric blockage of pre-mRNA splicing or translation, and RNase H1 recruitment for mRNA Both actions rely on sequence complementarity to nucleic acids, reducing or altering protein production.
- siRNAs are double-stranded RNA duplexes. Their primary mechanism is RNA Interference (RNAi) via the RISC for catalytic mRNA cleavage. They also act via sequence complementarity to nucleic acids and offer a potent, sustained reduction of a specific protein.
Many RNA-targeting ONTs exhibit tissue-centric, non-linear pharmacokinetics. They’re characterized by rapid plasma clearance, minimal systemic exposure, and fast uptake into highly perfused tissues, where they accumulate intracellularly and are metabolized slowly. As a result, ONTs can have very long tissue half-lives (weeks to months), which enables infrequent dosing but also delays the resolution of toxicity linked to prolonged intacellular persistence of both parent drug and metabolites. This has important implications for nonclinical study design.
Potential clinical manifestations of ONTs include injection-site reactions, flu-like symptoms, thrombocytopenia, transaminase elevations, complement activation, and glomerulonephropathy. To understand and predict these signals, a toxicity assessment framework is needed that assess both sequence-dependent and chemistry-driven risks. Effective safety assessment requires strategies targeting both types of liabilities.
Addressing Translational Challenges
Across ONTs, nonclinical studies can capture the major organ toxicities, particularly in the liver, kidney, and hematologic systems. But translation to clinical outcomes remains a critical ONT development challenge. ASOs often show broader systemic and immune-mediated effects, while siRNA therapies tend to exhibit more targeted, delivery-driven toxicities, highlighting how modality and platform influence safety profiles.
These differences are rooted in a shared pharmacological paradigm in which efficacy and toxicity are driven by prolonged tissue exposure (rather than plasma levels) and involve multiple toxicity pathways including on-target effects and both sequence- and chemistry-driven off-target mechanisms.
To address these challenges, development strategies must account for species differences (e.g. using surrogate oligonucleotides when human target isn’t conserved in standard nonclinical sepcies), adapt study designs to accommodate long tissue half-lives, and incorporate approaches that better predict immune and local effects, which are often underpredicted in nonclinical models.
Navigating Regulatory Pathways
The regulation of ONTs has evolved over three defining eras. Before 2024, foundational principles were established through key documents including ICH S6(R1) (2011) and ICH M3(R2) (2009), OSWG White Papers (2012-2025), Japan PMDA Guideline (2020), and Japanese Research WG (2021) report. . The year 2024 marked a pivotal shift, FDA Draft Guidances (Nonclinical Safety + Clinical Pharmacology), EMA Guideline, and the ICH S13 Concept Paper introduced draft guidance for the United States and the European Union, manufacturing standards, and the beginnings of harmonization. In the future, ICH S13 (2027) is expected to establish a global standard, reduce variability, and enable platform data leverage.
The FDA 2024 draft nonclinical guidance offers three core mandates.
- Comprehensive screening (in silico, in vitro, in vivo when needed)
- Pharmacological relevance (tissue over plasma)
- Platform efficiency (leverage class experience)
Overall, there has been a shift from template toxicology to risk-based, ONT-specific development. Species selection is also being redefined, emphasizing relevance over quantity.
Designing a Comprehensive Nonclinical Safety Package (per FDA 2024 guidance)
To build an effective nonclinical safety package for an ONT, developers should start with a risk-based, fit-for-purpose strategy rather than a conventional checklist. That strategy should be anchored in the molecule’s mechanism of action, chemistry-driven liabilities, tissue distribution, and expected duration of tissue residence so that each study meaningfully informs human risk. Each element below contributes to the translation of those characteristics into a more predictive and efficient development program.
Integrated IND-enabling strategy: ONT program is designed as a single integrated, strategy-driven program instead of a series of stand alone studies. As such, safety pharmacology endpoints (CV, CNS and Respiratory) are integrated into general toxicology study, dosing is aligned with tissue exposure and regulatory guidance. Beyond standard toxicology endpoints, inclusion of enhanced hematology and coagulation panels, immunotoxicity assessments including cytokine panel, immunophenotyping and complement activation, and immunogenicity. More critically, ONTs specific endpoints including tissue concentrations, PD biomarkers and focused histopathological examination are incorporated into the study, improving the ability to predict clinical safety.
Regulatory flexibility for rare diseases: For ONTs with well characterized platform, risk-proportionate approaches allow streamlined packages, potentially reduced study requirements, and early regulatory alignment for low-prevalence indications.
Assay waivers based on platform knowledge: Standard assays (e.g., hERG, genotoxicity, phototoxicity) may be waived when justified by mechanism and chemistry, focusing resources on relevant risks for ONTs with well characterized platform.
Chronic toxicity driven by tissue exposure: Long-term studies ( 6 month rodent and 9 month nonrodent, at least one species should be pharmacologically relevant) are designed around sustained tissue exposure, not plasma PK, to capture accumulation and delayed toxicities.
Tailored DART strategy: For a chronic ONT, DART assessment is required. Reproductive and developmental studies follow a clinical tissue exposure-driven approach, using relevant or surrogate species and accounting for long tissue half-life. Adapting study design to tissue residence: Species selection, dosing frequency, and recovery periods are optimized based on tissue PK, with traditional designs often insufficient. DART program should be regulatory compliant, mechanistically informative and clinically relevant.
Risk-based carcinogenicity assessment: Carcinogenicity testing is conducted only when triggered, using a weight-of-evidence approach rather than routine studies.
A Final Word
ONTs hold enormous potential, and recent innovations and breakthroughs position them to make a profound difference in personalized medicine and the treatment of difficult-to-treat conditions. The emerging standard for oligonucleotide safety assessment is a mechanistic, tissue-driven framework that integrates species-relevant models, translational biomarkers, and a proactive de-risking strategy from in silico through to in vivo systems. Supported by evolving regulatory guidance and platform-based efficiencies, this approach shifts nonclinical safety from a compliance exercise toward a predictive, risk-based engine.
