PRESERVATION FLUID FTIR FINGERPRINTING AS A PRE-IMPLANT READOUT OF EARLY KIDNEY GRAFT FUNCTION

Introduction: Delayed graft function remains a clinically relevant early complication after kidney transplantation and reflects, at least in part, the cumulative biological stress experienced by the graft before implantation. Current risk assessment relies mainly on donor, recipient and perioperative variables, but these parameters may not fully capture the biochemical status of the organ at the end of static cold storage.

Objectives: To investigate whether Fourier-transform infrared spectroscopy (FTIR) of kidney preservation fluid can provide a rapid pre-implant biochemical fingerprint associated with delayed graft function.

Methods: In this retrospective proof-of-concept study, preservation fluid samples from 56 kidney transplants were analyzed using a dried-film FTIR workflow. Spectral modelling focused on the 900–1800 cm⁻¹ fingerprint region, including phosphate- and carbohydrate-associated signals within approximately 900–1200 cm⁻¹ and the amide II and amide I domains at approximately 1500–1600 cm⁻¹ and 1600–1700 cm⁻¹, respectively. Spectral variables were selected independently within each donor-blinded training fold rather than from a fixed preselected band panel. Three predictive strategies were compared: a clinical model based on conventional variables, an FTIR-only model, and a combined clinical-plus-FTIR model.

Results: Delayed graft function occurred in 14 of 56 transplants. The FTIR-only model achieved the highest discrimination, with an area under the receiver operating characteristic curve of 0.814. The clinical model reached an AUC of 0.775, while the combined model reached an AUC of 0.796. These findings suggest that preservation fluid contains measurable biochemical information related to early graft function, not fully captured by standard clinical variables alone.

Conclusions: FTIR fingerprinting of kidney preservation fluid may offer a fast, low-volume and scalable approach for pre-implant graft assessment. Although exploratory, these results support further multicenter validation and recalibration before clinical implementation.