NADH fluorescence and photolysis indices for differentiating myocardial ischemia from artifacts

Background: Visualization of reduced nicotinamide adenine dinucleotide (NADH) fluorescence is a promising method for assessing the metabolic status of the myocardium during ischemia. However, its widespread clinical application is hindered by the fundamental ambiguity in signal interpretation, as it depends both on the metabolic concentration of this coenzyme and the illumination conditions that prevents reliable distinction of ischemia state from technical artifacts. Objective: The study was aimed to develop and validate a method for unambiguous interpretation of the fluorescence signal to differentiate ischemia from artifacts, based on the combined analysis of fluorescence intensity and the rate of NADH photolysis. Methods: A series of experiments was conducted on isolated rat hearts (n = 8) perfused using the Langendorff technique. In 4 experiments, ischemia was modeled due to cessation of perfusion; in the other 4 cases, a technical artifact was induced by change in power and position of the ultraviolet source. For each pair of recordings, namely before and after intervention, the relative changes in fluorescence intensity (dF) and the rate of photolysis (dN), calculated by approximating the kinetics with a mathematical model, were assessed. The primary endpoint was the accuracy of classifying the intervention type based on the combined analysis of dF and dN. Results: In induced ischemia, a significant increase in dF (294 ± 89 a.u.) was revealed with no change in dN (1.31 ± 1.17 a.u., p > 0.05). In the case of the artifact assessing, an increase in dF was also recorded (289 ± 83 a.u.) but it was accompanied by a significant increment in dN (17.46 ± 5.25 a.u., p < 0.01). The combined analysis of dF and dN enabled to accurately classify technical artifacts across the entire test sample (n = 8). Conclusion: The combined analysis of NADH fluorescence and photolysis eliminates signal interpretation ambiguity and reliably differentiates myocardial ischemia state from technical artifacts. The proposed method, which does not require a reference recording or prolonged scanning, could facilitate the adoption of metabolic imaging to intraoperative monitoring and the assessment of graft viability.