Considerations in the Preclinical Assessment of the Safety of Antisense Oligonucleotides

Abstract

The nucleic acid therapeutics field has made tremendous progress in the past decades. Continuous advances in chemistry and design have led to many successful clinical applications, eliciting even more interest from researchers including both academic groups and drug development companies. Many preclinical studies in the field focus on improving the delivery of antisense oligonucleotide drugs (ONDs) and/or assessing their efficacy in target tissues, often neglecting the evaluation of toxicity, at least in early phases of development. A series of consensus recommendations regarding regulatory considerations and expectations have been generated by the Oligonucleotide Safety Working Group and the Japanese Research Working Group for the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use S6 and Related Issues (WGS6) in several white papers. However, safety aspects should also be kept in sight in earlier phases while screening and designing OND to avoid subsequent failure in the development phase. Experts and members of the network "DARTER," a COST Action funded by the Cooperation in Science and Technology of the EU, have utilized their collective experience working with OND, as well as their insights into OND-mediated toxicities, to generate a series of consensus recommendations to assess OND toxicity in early stages of preclinical research. In the past few years, several publications have described predictive assays, which can be used to assess OND-mediated toxicity in vitro or ex vivo to filter out potential toxic candidates before moving to in vivo phases of preclinical development, that is, animal toxicity studies. These assays also have the potential to provide translational insight since they allow a safety evaluation in human in vitro systems. Yet, small preliminary in vivo studies should also be considered to complement this early assessment. In this study, we summarize the state of the art and provide guidelines and recommendations on the different tests available for these early stage preclinical assessments.

Keywords: antisense oligonucleotides; early preclinical safety assessment; predictive assays; toxicity.

FIG. 1.

OND-associated toxicities: schematic representation of the most common OND-mediated toxicities. Some of these effects are strictly chemistry and design specific (sequence independent, white boxes), some are mostly class specific but can be influenced by the sequence (light gray boxes), and others are sequence dependent (dark gray boxes). The arrows represent the impact of a specific effect on another, for example, immunostimulatory effects play a role on thrombocytopenia and toxicities observed in high exposure organs. OND, oligonucleotide drug.

FIG. 2.

Simplified recommendation path to assess OND-associated toxicities depending on the design and chemistry: for all types of ONDs, it is highly recommended to perform in silico analysis during sequence selection to select the candidates with the least predicted OTE, followed by in vitro testing of the predictive OTE at a suitable time in the screening cascade. *Depending on the resources available, the in vitro screening may be run on a smaller set of candidates, for example, after proinflammatory assessment in the cascade. The overall cytotoxicity of OND can then be assessed in vitro using caspase assays, which offer good throughput to select safe compounds. Depending on the application and chemistry selected, the priority for predictive assay is then different since high-affinity ONDs present higher probabilities of inducing subacute hepatotoxicity or nephrotoxicity than low-affinity PS-ONDs or even ASO such as PMO, which have never been reported to induce such toxicities (when unconjugated). ASO, antisense oligonucleotide; cET, constrained ethyl; LNA, locked nucleic acid; MOE, methoxyethyl; OTE, off-target effects; PMO, phosphorodiamidate morpholino oligomer; PS, phosphorothioate.