Towards therapeutic drug monitoring of methotrexate using SERS

Therapeutic drug monitoring (TDM) is a clinical practice that primarily measures medication levels in the bloodstream. With such measurements, TDM aims to provide essential information that can be used to perform safe and reliable treatment. Thus TDM is utilized for dose adjustment of certain drugs to prevent toxic side effects and uneffective treatment. Nowadays, highly automated immunoassay platforms are routinely used for drug measurement in blood samples as part of TDM practices. When immunoassay kits are not commercially available or when their selectivity falls short, high-pressure liquid chromatography-based analytical techniques are still employed. Due to the time consuming, expensive and cumbersome nature of these analytical approaches, TDM can only be performed to its full potential in well-equipped centralized clinics. New analytical methods should be developed to allow TDM to be performed more quickly and in resource-constrained environments while maintaining sensitivity and selectivity.
Surface-enhanced Raman spectroscopy (SERS) is a promising candidate for the next generation of TDM platforms. Molecule-specific SERS signals can be collected rapidly from various samples to use in health science applications. Furthermore, compact and relatively inexpensive Raman spectrometers can pave the way to on-site SERS detections. However, SERS commonly requires sample pretreatment to enable detection of molecules in complex sample matrices such as serum, plasma and urine. Moreover, reproducibility is a well known problem in SERS due to the variations observed in SERS substrates.
In this PhD project, we developed and optimized label-free SERS-based detection methods for TDM of the methotrexate (MTX) drug. MTX is a commonly used antineoplastic agent and frequently monitored in pediatric leukemia treatments. TDM of MTX is performed to prevent toxic effects caused by the slow elimination of MTX, with TDM results used to adjust MTX antidote dosage. We developed a method called nanopillar assisted separation (NPAS) that uses nanopillar SERS substrates both as signal enhancers and as a simple sample pretreatment procedure for serum samples. We were able to detect MTX in serum samples at concentrations as low as 0.7 μM using this NPAS method. In addition, the implementation of the NPAS approach in a centrifugal microfluidic disc was investigated to achieve automation and better control over the method application.
In order to reach lower detection limits, we used an electrochemically assisted SERS approach to promote affinity between MTX molecules and nanopillar SERS substrates. The development and optimization of the assay were performed in aqueous buffered samples, while detection in a serum sample matrix was also investigated. For refining the serum sample matrix, various sample pretreatment steps were investigated such as protein precipitation and centrifugal filtration or gel filtration. This allowed us to detect analytes in serum samples with the electrochemically assisted SERS method for first time in the literature, down to a concentration of 2 μM for MTX.
Physical and chemical quality control measure were implemented to eliminate outlier SERS substrates after production to improve the reproducibility of both NPAS and electrochemically assisted SERS approaches. Additionally, the performances of a compact and low-cost Raman spectrometer prototype developed in our group were evaluated with both methods. By combining the developed assays with this compact detection setup, these SERS-based sensing methods could contribute to make TDM practices more widespread and accessible.