Toward Resolving Heterogeneous Mixtures of Nanocarriers in Drug Delivery Systems through Light Scattering and Machine Learning

Nanocarriers (NCs) have emerged as a revolutionary approach in targeted drug delivery, promising to enhance drug efficacy and reduce toxicity through precise targeting and controlled release mechanisms. Despite their potential, the clinical adoption of NCs is hindered by challenges in their physicochemical characterization, essential for ensuring drug safety, efficacy, and quality control. Traditional characterization methods, such as dynamic light scattering and nanoparticle tracking analysis, offer limited insights, primarily focusing on particle size and concentration, while techniques like high-performance liquid chromatography and mass spectrometry are hampered by extensive sample preparation, high costs, and potential sample degradation. Addressing these limitations, this work presents a cost-effective methodology leveraging light scattering and optical forces, combined with machine learning algorithms, to characterize polydisperse nanoparticle mixtures, including lipid-based NCs. We prove that our approach provides quantification of the relative concentration of complex nanoparticle suspensions by detecting changes in refractive index and polydispersity without extensive sample preparation or destruction, offering a high-throughput solution for NC characterization in drug delivery systems. Experimental validation demonstrates the method's efficacy in characterizing commercially available synthetic nanoparticles and Doxoves, a liposomal formulation of Doxorubicin used in cancer treatment, marking a significant advancement toward reliable, noninvasive characterization techniques that can accelerate the clinical translation of nanocarrier-based therapeutics.