Nanoparticle-based lateral flow immunoassays for detecting marine biotoxins in seafood: a systematic review of trends in development over two decades

• A systematic overview of the summary of findings of lateral flow assays development and performance to date. • Shows how nanoparticle design, biorecognition chemistry, workflow simplification, and device architecture impact sensitivity, specificity, and limit of detection. • Identifies critical gaps (validation, reproducibility, matrix effects) and outlines standardized strategies for performance optimization. • Provides evidence-based guidance for regulators and policymakers on evaluation and public-health deployment. Marine and freshwater biotoxins threaten public health and seafood safety due to their tendency to accumulate in fish and shellfish. This systematic literature review summarises findings from 51 peer-reviewed studies published between 2004 and 2025 that focus on nanoparticle-enhanced lateral flow immunoassays (LFIAs) for toxin detection. Gold nanoparticles (AuNPs) were by far the most used label, employed in ∼75 % of studies, which consistently delivered the lowest limits of detection (LODs) as low as 0.03 ng/ml for okadaic acid detection. Despite increasing concerns over the co-occurrence of multiple toxins, ∼88 % of the studies were based on single-analyte detection, only ∼12 % employing multiplexed approaches. Statistical analyses of the studies, namely Kruskal–Wallis, Dunn’s post hoc, correlation, and regression analysis, revealed meaningful trends. LODs have significantly improved over time, a marked reduction being observed after 2016. Kruskal–Wallis and Dunn’s post hoc tests showed significant differences across variables such as nanoparticle size, assay duration, and toxin recovery. Spearman correlation results identified a moderately significant negative correlation of LOD with publication year (ρ = –0.458, p < 0.001), and a weaker significant correlation of LOD with assay time (ρ = –0.301, p = 0.018), while linear regression analysis identified publication year (R² = 0.2441, p < 0.0001) and nanoparticle size (R² = 0.1072, p = 0.0264) as significant predictors of assay sensitivity. Overall, this review underscores the central role of nanoparticle engineering, simplified assay workflows, and optimized design strategies in improving LFIA performance. The findings provide a valuable roadmap for advancing the development of portable, sensitive, and scalable biosensing platforms for marine toxin monitoring and seafood safety surveillance.