Quantitative Assessment of Liver Steatosis on Tissue Section Using Infrared Spectroscopy

Abstract

Liver transplantation (LT) constitutes a major treatment of patients with cirrhosis or hepatocellular carcinoma. Due to the shortage of grafts, transplantation teams are using marginal grafts from expanded criteria donors.1Durand F. et al.Liver Transpl. 2008; 14: 1694-1707Crossref PubMed Scopus (234) Google Scholar, 2Noujaim H.M. et al.Transplantation. 2009; 87: 919-925Crossref PubMed Scopus (56) Google Scholar, 3Briceno J. et al.Transplantation. 2010; 90: 530-539Crossref PubMed Scopus (74) Google Scholar These marginal grafts constitute an additional source of organs but also an increased risk factor of primary nonfunction or dysfunction of the graft mainly due to their poor quality. Steatosis is one of the most important factors affecting liver allograft function, mostly because of more severe ischemia-reperfusion injury. More than 30% of the population in Western countries is estimated to exhibit this lesion. Steatosis is characterized by the intra-cellular accumulation of triacylglycerides (TG) resulting in the formation of lipid vesicles in the hepatocytes. The gold standard to assess hepatic steatosis in liver grafts during the transplantation procedure is the histologic examination of frozen sections by a pathologist. The quantitative histologic evaluation of steatosis is based on the percentage of hepatocytes containing cytoplasm fat inclusions. Most transplant surgeons currently discard grafts when the total amount of hepatic steatosis is more than 60% because of an expected high risk of graft failure. However, while there is general agreement that mild steatosis (<30%) causes minor graft injury, the use of grafts with moderate steatosis (30%-60%) remains a challenging issue with primary non-function rates ranging between 0% and 75% after LT.1Durand F. et al.Liver Transpl. 2008; 14: 1694-1707Crossref PubMed Scopus (234) Google Scholar, 2Noujaim H.M. et al.Transplantation. 2009; 87: 919-925Crossref PubMed Scopus (56) Google Scholar, 3Briceno J. et al.Transplantation. 2010; 90: 530-539Crossref PubMed Scopus (74) Google Scholar, 4Spitzer A.L. et al.Liver Transpl. 2010; 16: 874-884Crossref PubMed Scopus (241) Google Scholar, 5McCormack L. et al.J Hepatol. 2011; 54: 1055-1062Abstract Full Text Full Text PDF PubMed Scopus (198) Google Scholar The major issue is that assessment of hepatic steatosis on histologic sections is an imperfect and not reproducible method. A recent study demonstrated poor inter-observer agreement among expert pathologists for quantitative assessment of steatosis on paraffin sections.6El-Badry A.M. et al.Ann Surg. 2009; 250: 691-697Crossref PubMed Scopus (21) Google Scholar Despite improvements using stereology, the emergency timing of liver transplantation imposes a rapid process such as the use of frozen sections leading to some damages in the tissue morphology and thus to more imprecise assessment of steatosis (Figure 1A). Finally, we have demonstrated that the histological estimation of steatosis is poorly correlated with the true lipid content of the liver tissue after lipid extraction from the same surgical specimen followed by gas phase chromatography coupled to mass spectrometry (GC-MS) (Figure 1B). Thus, an alternative method to histopathology providing an objective assessment of steatosis is an unmet need. Infrared (IR) microspectroscopy is based on the determination of absorption of infrared light due to resonance with vibrational motions of functional molecular groups. As such IR microspectroscopy constitutes a valuable tool for biochemical investigations.7Barth A, et al. Biological and Biomedical Infrared Spectroscopy. Advances in Biomedical Spectroscopy. Haris, ed. Volume 2. Amsterdam, Netherlands: IOS Press BV; 2009.Google Scholar, 8Le Naour F. et al.PLoS ONE. 2009; 4: e7408Crossref PubMed Scopus (67) Google Scholar We have performed IR microspectroscopy analysis on frozen liver samples harvested from 27 human liver surgical specimens exhibiting various levels of steatosis (0–90%) based on histological examination. Acquisition of IR spectra on unstained frozen sections was achieved first using a commercially available laboratory IR microscope that allows scanning a whole tissue section ranking between 1 to 5 mm2 (up to 10 000 spectra with 50 x 50 μm2 aperture size) in less than 5 minutes. Important variations in intensity of bands attributed to lipids (3000–2800 cm-1) were observed even in the non-steatotic part of fatty livers despite normal histological aspect thus strengthening that IR microspectroscopy allows detecting subtle chemical changes occurring in the liver tissue. In contrast, the intensity of bands attributed to proteins (Amide I and Amide II: 1485–1710 cm-1) was not significantly affected by steatosis. We further demonstrated that an average IR spectrum from an area 500 x 500 μm2 was representative of the whole tissue section. The quantification of the lipid content was addressed from the IR spectra by calculating the ratio of integrated intensity of bands attributed to lipids (2800–3000 cm-1) related to proteins (Amide II: 1485-1595 cm-1) (Figure 1C). This ratio allows normalizing the intensity variation of the bands due to variations in thickness of the tissue section. For each patient, the average ratio lipids/proteins measured by IR microspectroscopy was plotted as a function of the concentration of TG leading to exhibit a marked linearity (Figure 1D). A standard curve was further established to quantify the lipid content and thus the level of steatosis. The method was further applied to 12 surgical biopsies obtained at harvesting of liver grafts right before clamping that exhibited steatosis in the range of 30-60%. The theoretical amount of TG estimated by IR microspectroscopy was comparable to the true TG content (Figure 1E). The video (video 1) illustrates the power of the method by investigating a questionable liver graft exhibiting 50% steatosis based on histopathology. A frozen tissue section with 5 μm thickness is performed and deposited on a regular glass slide. The analysis takes advantage of attenuated total reflection-Fourier transform infrared (ATR-FTIR) spectroscopy that allows by a direct contact of the sample with the crystal detector (Figure 2A) the acquisition of an average IR spectrum corresponding to a large area of 2 x 2 mm2 of the tissue section. The analysis takes less than 1 minute for acquiring the spectrum, calculating the ratio lipids/proteins, extrapolating the amount of TG based on the established standard curve and thus the related virtual percentage of steatosis (approximately 20%). A symbol, as green light, appears on the computer screen (Figure 2B) when the graft is suitable for liver transplantation. Infrared microspectroscopy allows the quantification of lipid content on tissue section. This rapid method can be easily used at the hospital for reliable assessment of graft quality control in liver transplantation. We thank Benoît Trousselle for skillful tissue processing. We are grateful to Justine Bertrand-Michel for her expertise in lipidomic analysis and to Oriana Ciacio, Raffaela Pittau, Colette Danet, Catherine Danguy des Déserts, Elisabeth Pasdeloup, Gael Berthelot and Franck Chiappini for the assistance in liver graft samples collect. eyJraWQiOiI4ZjUxYWNhY2IzYjhiNjNlNzFlYmIzYWFmYTU5NmZmYyIsImFsZyI6IlJTMjU2In0.eyJzdWIiOiJkZjg2ZDQwNDA5OGYxZGE1MTFiZTBhZWRhMmY2ZDg4YyIsImtpZCI6IjhmNTFhY2FjYjNiOGI2M2U3MWViYjNhYWZhNTk2ZmZjIiwiZXhwIjoxNjgxNDY0MzUyfQ.iu71T5r2k6nhyaoNmPM9PSOkPFjK3R8qgkJmanD3j7S7Sf3_T8weL3bMP1uXoggSRFizzHsnPx8uTJNX_046h7oLUVMAVHwdwVe0PV0QxZwZiaVSnNCp4cRV1gNnmpeWXGQoaprkK2LoxkUOXTLp7O_DKPtP8u3EfLN8KahbPpd3-zhRVumX06M6WrHruY2HBv1PEG17NMdEHRqbnPjOHyn9Q6PsUdk9A-j0T83TA-smMNnNz8YZFauQFNRIjduQx6pnhJVY77ZX5pIphnQRqlvjEzR1r3v1ufwhTAWt4bzjFaEK9h-bPElEth7G8XpJuWDILKstMjjCvL4vgGynmA Download .mp4 (25.46 MB) Help with .mp4 files Video 1