Metabolomic and Lipidomic Profiling for Pre-Transplant Assessment of Delayed Graft Function Risk Using Chemical Biopsy with Microextraction Probes

Abstract

Organ shortage remains a significant challenge in transplantology, prompting efforts to maximize the use of available organs and expand the donor pool, including through extended criteria donors (ECDs). However, ECD kidney recipients often face poorer outcomes, including a higher incidence of delayed graft function (DGF), which is linked to worse graft performance, reduced long-term survival, and an increased need for interventions like dialysis. This underscores the urgent need for strategies to improve early DGF risk assessment and optimize post-transplant management for high-risk patients. This study conducted multi-time point metabolomic and lipidomic analyses of donor kidney tissue and recipient plasma to identify compounds predicting DGF risk and assess the translational potential of solid-phase microextraction (SPME) for graft evaluation and early complication detection. The SPME-based chemical biopsy enabled a direct kidney analysis, while thin-film microextraction facilitated high-throughput plasma preparation. Following high-performance liquid chromatography coupled with a mass spectrometry analysis, the random forest algorithm was applied to identify compounds with predictive potential for assessing DGF risk before transplantation. Additionally, a comparison of metabolomic and lipidomic profiles of recipient plasma during the early post-operative days identified metabolites that distinguish between DGF and non-DGF patients. The selected compounds primarily included amino acids and their derivatives, nucleotides, organic acids, peptides, and lipids, particularly phospholipids and triacylglycerols. In conclusion, this study highlights the significant translational potential of chemical biopsies and plasma metabolite analyses for risk assessments and the non-invasive monitoring of DGF. The identified metabolites provide a foundation for developing a comprehensive DGF assessment and monitoring method, with potential integration into routine clinical practice.

Keywords: DGF; LC-MS; graft quality assessment; kidney transplantation; lipidomics; metabolomics; solid-phase microextraction.

Figure 1

Receiver operating characteristic (ROC) curves of random forest models for selected compound sets distinguishing between DGF and non-DGF groups. AUC: area under curve.

Figure 2

Pathway analysis of metabolites selected for random forest models from metabolomic and lipidomic analyses.

Figure 3

Selected metabolites differentiating the DGF and non-DGF groups on post-operative days 1 and 5. The rectangle’s height represents the normalized peak areas in the interquartile range (Q1 and Q3). The upper whisker denotes the largest data point (excluding any outliers), while the lower whisker denotes the lowest data point (excluding any outliers). The median normalized peak area of each group is indicated with a yellow square.

Figure 4

Chemical biopsy sampling during simultaneous surgical preparation of the graft.

Figure 5

Study design. The study involved samples collected from kidneys recovered from deceased donors and plasma samples from organ recipients.