(M1530-10-63) Quantum Electrochemical Spectroscopy (QES): A Novel Approach for Enhanced Detection and Evaluation of Drug-Induced Liver Injury (DILI)

Purpose: Drug-induced liver injury (DILI) is universally evaluated in drug discovery and development and is one of the most common causes of attrition in early drug discovery. Approximately 18% of commercial medicines carry a significant DILI risk. DILI evaluations occur throughout preclinical and clinical development as well as post-marketing. Evaluations are standardized and, in combination with other risk assessment assays, factor significantly in decision-making in molecule progression, regulatory approval, and post-approval surveillance. Current standard methods to assess DILI are based on general liver injury serum biomarkers such as alanine aminotransferase (ALT), aspartate aminotransferase (AST), alkaline phosphatase (ALP), and total bilirubin (TBL)¹. These biomarkers are often assessed in a centralized and industrialized manner, using costly reagents and significant biological sample amount. In this study we evaluate the ability of a new analytical technique, Quantum Electrochemical Spectroscopy (QES)² to perform liver and immune biomarker analysis as well as digital phenotyping of DILI and compare it with gold-standard blood analyzer-derived data (Advia 1800, Siemens Healthineers). QES utilizes electronic means to measure molecular vibrations from a minuscule sample ( < 10ul of serum) at room temperature without reagents or sample preparations, and in a few minutes. QES data can be used to identify, quantify and classify analytes and digital phenotypes without the need to select the analyte before data acquisition.

Methods: Experimental design: a rat preclinical toxicity study of 12 days was conducted using 5 male and 5 Female individuals per group (Control and 3 dose groups). A total of 40 serum samples were collected from Dec 2019 to Jan 2020 and stored at -20C. One 100 ul aliquot was analyzed for blood biomarkers using the Advia 1800 analyzer (gold standard) and three 4 ul aliquots were used with the QES Method. QES Data Acquisition: 4 ul of rat serum or 4ul of analyte standard solution in PBS were pipetted into a consumable electrochemical sensor coupled to a Probius™ QES instrument. The QES instrument performs a 30 min voltage scan of the sample to create a high dimensional vibrational signature of the molecular species therein. All samples were assessed in triplicates. The high-dimensional dataset was uploaded to the software (Probius BCS) as a digital representation of the sample, a digital twin. Then, using reference standards (see below), the high-dimensional vibrational spectrum of the samples was deconvoluted to individual molecule contributions in the bulk signal and parsed for inferencing using trained models. To create and train these models, 2 different approaches were used, depending on the tasks, Quantification, and classification (digital phenotyping): Quantification: The quantification of analytes requires the creation of standard curves following the method of standard additions. In brief, for each analyte (AST, ALT, IL-6, Albumin) standard solutions were created using commercially available synthetic or purified molecules (Biotechne, USA and Sigma-Aldrich, USA) and spiked at different concentrations into a PBS solution or the biological matrix (pooled Wistar Rat Serum, Sigma Aldrich). Electronic vibrational signal intensity was plotted vs standard concentration to create a calibration curve, where freshly prepared blinded standard aliquots were used to verify the method. (Figure 1) Digital phenotyping: QES signatures acquired from 50% of the randomized samples were used to create a classifier based on QES signatures for the given category (control and treatment groups) or analyte level. The remaining samples were used for validation of the trained k-nearest neighbor's classifier.

Results: in this study, we demonstrate the ability of QES to quantify and assess common blood biomarkers used in toxicity studies. Quantification Curves were created by the method of standard addition for common Liver toxicity biomarkers (AST, ALT, ALB, Figure 2). We also demonstrate the capacity of QES to add analytes on demand, and a-posteriori of data acquisition to assess other physiological functions (i.e., Inflammation, IL06 and TNF-a) without the need to re-acquire the data or consume additional sample volume. QES can also create classifiers based on digital phenotyping cues like dose or biomarker level to predict toxicity response and assess digital phenotypes that are invisible to the traditional blood analyzer techniques (Figure 3). QES demonstrate strong correlation between AST and ALT (Figure 4) as well as creatinine, which are not evident with the traditional analytical techniques.

Conclusion: Despite being a key step in the assessment of molecule progression, there are major limitations in DILI detection and prediction due to the current lack of specific biomarkers. Current traditional liver biomarkers do not provide a broad view of DILI and are cumbersome to assess. In this proof of Principle study, we demonstrate the ability of QES, a novel analytical technique, to identify and measure traditional liver biomarkers, assess inflammation-related signals and create digital phenotyping signatures with a very convenient workflow that only requires 4ul of a serum sample, one pipetting step, and no reagents or sample preparation.

References: 1) Robles-Díaz M, et al. Biomarkers in DILI: One More Step Forward. Front Pharmacol. 2016 Aug 22; 7:267
2) Gupta, Chaitanya, et al. "Quantum tunneling currents in a nanoengineered electrochemical system." The Journal of Physical Chemistry C 121.28 (2017): 15085-15105.

Acknowledgements: All Authors of this abstract are either employees and/or shareholders of ProbiusDx Inc. or Merck KGaA.

Figure 1: Method of Analyte Addition for Quantitation of Analytes

Figure 2: Examples of AST, ALT and ALB Calibration Curves and Blind Verification Results

Figure 3: Comparison of Correlation between ALT and AST values from Rat Serum Samples by 4 different methods (standard addition with concentration, with vs enzymatic activity, QES classifier, and Gold standard data)