Multiplexed Inflammatory Cytokine Monitoring via a Vertical Flow Assay during Normothermic Liver Perfusion

Introduction: : Liver failure poses a growing threat to global health, mainly driven by alcohol misuse, metabolic disorders, and viral hepatitis. Despite recent medical advancements, liver transplantation remains the only curative treatment for end-stage liver disease. During the transplant process, donor liver preservation is a critical determinant of transplant success, influencing graft function and post-transplant complications. Recently, normothermic machine perfusion (NMP) has emerged as a promising preservation technique, offering extended preservation times, salvage of marginal grafts, and reduced ischemia-reperfusion injury. Despite these advantages, NMP lacks a definitive, real-time, and non-invasive biomarker to instruct clinicians during transplantation. Emerging evidence indicates that raised inflammatory cytokine concentrations during perfusion correlate with adverse post-transplant outcomes, including early allograft dysfunction. This observation highlights the need for rapid diagnostic methods to evaluate liver inflammation during machine perfusion. In response, we propose a novel approach for point-of-care cytokine detection to enhance decision-making during liver transplantation procedures.

Materials and

Methods: : Here, we developed a paper-based platform for rapid, multiplexed detection of inflammatory cytokines in liver perfusate samples. The platform leverages a microfluidic paper device embedded with antibody arrays, detecting IL-6, TNFα, and IL-8, which are implicated in liver injury. Samples are applied directly onto the device and flow through membrane capillary channels, enabling antigen-antibody interactions where colorimetric signals are produced in response to cytokine binding. The analytical performance of the device was validated using spiked samples and clinical perfusate collected during normothermic machine perfusion. Data from the platform were integrated with machine learning algorithms to classify inflammation status, providing a real-time assessment tool to guide intraoperative decisions.

Results, Conclusions, and Discussions:: During normothermic machine perfusion, our paper-based platform demonstrated robust detection of inflammatory cytokines in perfusate samples at clinically relevant nanogram concentrations. By employing a spatially multiplexed assay format, we achieved simultaneous quantification of several cytokines with negligible nonspecific binding. Specifically, IL-6, TNFα, and IL-8 were able to be detected at limits of detection of < 2.5 ng/mL. Integration of a machine-learning classifier then enabled precise, real-time stratification of perfusate inflammation status, differentiating non-inflamed, intermediate, and inflamed samples, supporting intraoperative decision-making for graft selection and ultimately improving patient outcomes. These findings highlight the platform’s practical advantages over conventional immunoassays and its strong potential as an intraoperative diagnostic tool.

In summary, we present a rapid, point-of-care, paper-based diagnostic device for multiplexed cytokine detection during normothermic machine perfusion. Its clinically relevant sensitivity, combined with machine-learning-driven classification, addresses the throughput limitations of standard immunoassays and enables real-time assessment of liver inflammation—a critical predictor of post-transplant success. Future work will focus on extensive validation and refinement to support clinical translation and widespread adoption in liver transplantation settings.

Acknowledgements and/or References (Optional)::