Introduction: The incidence of primary non-function (PNF) after liver transplantation (LT) remains a major concern with the increasing use of marginal grafts. Indocyanine green (ICG) fluorescence is an imaging technique used in hepatobiliary surgery and LT. Because few early predictors are available, we aimed to quantify in real time the fluorescence of grafts during LT to predict their 3 months survival.
Patients and methods: After graft revascularization, ICG was intravenously injected, then the fluorescence of the graft was captured with a near infrared camera and postoperatively quantified. A multi-parametric modelling of the parenchymal fluorescence intensity curve was proposed, and a predictive model of graft survival was tested.
Results: Between July 2017 and May 2019, 76 LT were included, among which 6 recipients were re-transplanted (re-LT+ ). No side effects of ICG injection were observed. The parameter a150(temporal course of fluorescence intensity) was significantly higher in the group re-LT+ group (0.022s-1 [0.016-0.035] versus 0.012s-1 [0.009-0.015], p=0.01). This parameter was the only independent predictive factor of graft survival at 3 months (OR=2.4 CI95% [1.05-5.5] p=0.03). The best cut-off for the parameter a150 (0.0155s-1 ) predicted the graft survival at 3 months with a sensitivity of 83% and a specificity of 79% (AUC=0.82 [0.67-0.98], p=0.01).
Conclusion: Quantitative assessment of intraoperative ICG fluorescence on the graft was feasible to predict graft survival at 3 months with a good sensitivity and specificity. Further prospective studies should validate these results over larger cohorts and evaluate the clinical impact of this tool.
Keywords: Graft function; Indocyanine green dye; Modelling; Near-infrared fluorescence; Primary non-function.