Oligonucleotide-based therapies, such as antisense oligonucleotides (ASOs), small interfering RNAs (siRNAs), represent a class of therapeutic agents that specifically target gene transcription or translation mechanisms through sequence specificity. These pharmaceuticals exhibit significant promise in the treatment of genetic disorders, including spinal muscular atrophy, as well as malignancies, viral infections, and metabolic diseases. Nonetheless, unintended toxicity continues to pose a considerable challenge and remain a critical safety concern in the development of oligonucleotide therapeutics (ONTs). Off-target toxicity may be caused by hybridization to sequences that are similar but not identical to the target, hybridization-independent sequence related, or sequence- and hybridization-independent effects. The effects may result in diminished transcript levels, decreased translation rates, or anomalous splicing, employing same molecular pathways and protein machinery as the desired on-target effects. Currently, there exists no established methodology for the systematic identification and evaluation of off-target toxicity, which may hinder the optimization of safety approaches. This review delineates significant nonclinical toxicities and clinical adverse effects by summarizing and analyzing approved oligonucleotides with their off-target assays, encompassing the limitations of nonclinical off-target effects and the potential off-target mechanisms. Plus, it discusses and emphasizes the factors that lead to the off target of ONTs, systematically offers approaches and workflows of preclinical assessments to enhance the transfer value of oligonucleotide therapies from nonclinical to clinical trials by managing unavoidable off-target effects.
The retinal pigment epithelium (RPE) is essential for outer retinal homeostasis and a key target for drug-induced retinal toxicity. However, conventional preclinical models, including immortalized cell lines and animal systems, may incompletely recapitulate human RPE biology because of limited representation of specialized RPE functions and species-specific differences in retinal physiology. Induced pluripotent stem cell-derived RPE (iPSC-RPE) has emerged as a candidate human-relevant platform for New Approach Methodology (NAM) development, with the potential to improve mechanistic relevance by reproducing key structural and functional features of native human RPE. Its translational utility depends on a clearly defined context of use (COU), reproducible differentiation and functional characterization, and benchmark-based frameworks for future assay qualification and validation. This review evaluates the potential of iPSC-RPE as a COU-defined NAM framework for RPE-centered drug-induced retinal toxicity, with particular emphasis on toxicities in which RPE dysfunction represents a primary or mechanistically important component. We summarize differentiation strategies and critical quality attributes that may inform future assay qualification, evaluate disease-modeling evidence supporting physiological relevance, and discuss applications in ocular safety assessment, including chloroquine/hydroxychloroquine-, MEK inhibitor-, and anti-VEGF-associated retinal effects. We further discuss limitations, including cellular immaturity, donor-to-donor variability, incomplete microenvironmental modeling, and the limited validation of long-term exposure paradigms. Current evidence supports mechanistic concordance for selected RPE-centered toxicities, whereas prospective predictive performance across diverse compounds and clinically relevant exposure conditions remains to be established. Further COU-specific qualification and prospective validation will be required to establish the reliability and decision relevance of iPSC-RPE for defined ocular toxicology applications.