Abstract LB157: Electrochemical biosensor in the detection of pancreatic cancer biomarkers: AI driven model & patient cohort study

Abstract Introduction: Pancreatic cancer stands as the 12th most common occurring cancer worldwide. The increasing mortality is due to the diagnosis of pancreatic cancer done at a very late stage as it is harder to pinpoint until the symptoms have advanced, causing rapid spread of the tumour affecting the pancreas and other surrounding organs. Late stage causes treatment to be less effective. There are various diagnostic tools for the detection of pancreatic cancer, the most common being scans like CT, MRI, blood tests, and biopsy, we have come to understand that utilization of electrochemical biosensors can be an efficient and sound point-of-care system in the early diagnosis of pancreatic cancer. Biosensors utilize electrode interfaces to transduce molecular recognition events into electrical signals. In our research, we use electrochemical biosensors to detect multiple protein biomarkers such as CA19-9, CA242, CEA, and CA125 in serum samples. To increase the sensitivity of the biosensors, we nanoengineer graphene oxide onto the surface to increase sensitivity and surface area. We validate our results with existing biological assays in detecting biomarkers like ELISA and Western Blot. We aim to analyze our data in an AI model to predict the risk levels. Methods: Nanoengineering of Sensor: The gold working electrode surface of the biosensor is coated with graphene oxide nano colloidal solution and kept in UV overnight for crosslinking. Electrochemical Biosensor:The nanoengineered sensors are coated with a DSP crosslinker, followed by the addition of biomarker-specific antibodies. A blocking agent, TBS, is added to prevent non-specific binding. The patient samples are then added, and the sensors are run. Electrochemical impedance spectroscopy (EIS) and cyclic voltammetry are analyzed based on the interactions between the sample and the antibody. Data from the analysis is added to the AI model. ELISA is performed: A capture antibody is coated onto a plate and binds to the target antigen from the sample. Western Blot: Proteins in the serum sample are separated in gel electrophoresis and transferred to a membrane where specific antibodies to detect the biomarker are added and visualized using chemiluminescence. Results: Patient samples were individually run on the graphene-coated sensors for all four pancreatic biomarkers (IRB no:360124-28). Electrochemical spectroscopy calculated each patient sample's impedance, capacitance, resistance, and oxidation (CV). Each parameter was compared to a risk table for each biomarker, labelling the results as high risk, low risk, and normal concentrations. Patients were divided into two categories: control non-cancer (n=5) and cancer (n=5). Comparatively, most control samples showed low concentrations of the cancer biomarkers than the cancer samples and this data was justified by ELISA and Western Blot. Further, the data were analyzed using a developed AI model. Conclusion: Pancreatic cancer diagnosis is challenging as detection is harder during the earlier stages in patients. To solve this problem, targeting multiple cancer biomarkers by electrochemical sensors is a valuable tool in early diagnosis due to its portability, rapid detection and sensitivity. We further increase the sensitivity and surface area of the sensor by nanoengineering graphene oxide nano colloidal particles. Citation Format: Rashwitha Julia Saldanha, Mareeswari Paramasivan, Gnanasekar Munirathinam, Christopher Sumeet Gondi Gondi, Lusine Demirkhanyan, Mathew. T Mathew. Electrochemical biosensor in the detection of pancreatic cancer biomarkers: AI driven model & patient cohort study [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 2 (Late-Breaking, Clinical Trial, and Invited Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_2):Abstract nr LB157.