Nanoparticle-aided glycovariant assays to bridge biomarker performance and ctDNA results

Abstract Numerous immunoassay based cancer biomarkers established in the 1970 and 1980′ies are widely used in clinical routine. Initial expectations of biomarkers such as CEA, CA125, CA19-9, AFP to provide decisive help in the diagnosis of early stage, pre-symptomatic cancers have not been realized. Thus, they are primarily used for monitoring disease progression and occasionally being useful as prognostic indicators. This limitation is due to the marker also being measurable in healthy individuals and frequently at elevated concentrations in common benign conditions. Most conventional tumor markers are glycosylated and interestingly specific alterations of the glycostructure part can often be seen early in the cancerous process. Conventional double monoclonal immunoassays are however blind to such changes as they are based on peptide epitope recognition. Wide selections of carbohydrate recognizing macromolecules, lectins, but also glycan structure recognizing antibodies are potentially useful for detecting such changes. Despite numerous attempts generating proof-of-principle evidence for this, such assays have generally not been successfully introduced into clinical routine. The affinity constants of lectin and glycan specific antibodies for their corresponding carbohydrate structures may be up to several orders too low to provide the detection limits and robustness expected from routine tumor markers. In this review, we describe an approach based on the use of highly fluorescent Eu3+--chelate dyed nanoparticles onto which lectins or glycan specific antibodies are coated to provide the necessary binding strength and signal amplification to provide low detection limits, while maintaining the original glycan-structure specificity. This concept applied to three markers, PSA, CA125 and CA15-3 provide glycoform assays of greatly enhanced cancer specificity using sample volumes similar or lower than corresponding traditional ELISAs. For ovarian cancer, we show that this new approach when applied to ovarian cyst fluid samples provide results similar to the performance obtained with ctDNA determinations of a set of 17 driver mutations and greatly superior compared to corresponding conventional immunoassays. Based on our results, we predict that the nanoparticle-lectin concept will enable a new generation of simple, low-cost biomarker assays of highly improved cancer specificity. Such tools should ideally be evaluated together with determination of ctDNA to establish early detection schemes for cancers e.g. ovarian, pancreas, lung where the detection rate of early stage disease is presently unacceptably low.