Assessing ATR-FTIR spectroscopy for steatosis quantification in liver biopsies in a long-duration cross-sectional study

Abstract

Evaluating the presence of fat in liver biopsies is of relevance in the diagnosis of liver steatosis and liver transplantation due to its potential negative impact on the success of transplantation procedures. Currently, the standard method for assessing steatosis relies on a semi-quantitative grading system, which involves visual examination of liver biopsies stained with hematoxylin and eosin (H&E). Nonetheless, this method is time-consuming, subjective, and yields inconsistent results. Recent research has shown that Attenuated Total Reflectance-Fourier Transform Infrared (ATR-FTIR) spectroscopy has the potential to be used as a real-time quantitative tool for the direct evaluation of steatosis in liver tissues. This study reports the results of an extensive, cross-sectional clinical investigation conducted over a 40-month period, with the primary goal of evaluating the stability and precision of a pre-established method that utilizes first derivative ATR-FTIR on dry film for quantifying steatosis in the 1810–1052 cm−1 region, with H&E digital image analysis as reference method. Samples used to train the PLS model (n = 22) were gathered and analyzed during the initial 15 months of the clinical study, while the samples in the test set (n = 77) were collected and analyzed during the subsequent 25 months. Results yielded a mean RMSEP = 1.01 % in the validation set, and showed that accuracy remained stable over time, demonstrating the method's robustness without the necessity for recalibration, or method re-optimization. Furthermore, with a total sample analysis time of less than 2.5 min, its quick turnaround time could be particularly valuable in clinical settings where time is often a critical factor in decision making. These outcomes validate the potential of dry-film ATR-FTIR as a low-cost tool for the initial graft evaluation in the operating room, and support for future investments to further validate and expand this approach.