Hepatocellular carcinoma (HCC) is the third leading cause of cancer-related death with a five-year survival rate of less than 7%. The HCC incidence is rising throughout the world as a consequence of hepatitis virus infections, chronic aflatoxin exposure, non-alcoholic and alcoholic fatty liver diseases. However, the relationship between HCC tumorgenesis and changes in the amounts or composition of circulating plasma lipids and urinary volatile organic metabolites (VOMs) has been not fully addressed yet. The present study aimed to investigate the plasma phospholipid fatty acids and urinary volatile organic metabolites aiding in finding potential biomarker for diagnosis of HCC. Two analytical methods, both invasive and noninvasive, were applied in our study. In invasive way, A total of 37 plasma samples from healthy controls and HCC patients were collected and their phospholipid fatty acid distribution were profiled by gas chromatography-mass spectrometry (GC-MS) followed by multivariate statistical analysis. In noninvasive way, A total of 77 urine samples from healthy controls, HCC patients and medicine treatment group were collected and their urinary VOMs were characterized by headspace-gas chromatography-mass spectrometry (HS-GC-MS) followed by central composite design (CCD) and multivariate statistical analysis. In invasive way, linoleic acid (18:2 n-6) and oleic acid (18:0 n-9) were found to be common biomarkers in plasma phospholipid fatty acids and liver tissue fatty acids for the abnormalities. In noninvasive way, under the optimized conditions by means of CCD, seven urinary VOMs were characterized and compared between healthy controls and HCC patients and between drug-treatment and non-treatment group. Taken together, these data indicate that the GC-MS-based plasma phospholipid fatty acids in invasive way and urinary volatile organic metabolites in noninvasive way both exert an extraordinary effect on screening for diagnostic biomarkers of HCC. Our findings suggested that GC-MS has the potential to be developed as an effective diagnostic tool for hepatocelluar carcinoma in metabolomics.

肝細胞癌是以其五年生存率少於 7%佔據癌症引起死亡的第三大殺手。肝細 胞癌的發生率於全球正在上升，主要由肝炎病毒感染、長期接觸黃曲霉素、非酒 精性和酒精性脂肪肝疾病引起。然而，在循環血漿脂質的含量與組成以及尿液揮 發性有機物方面，肝細胞癌的腫瘤發生與改變的關係仍然沒有充分報導。本文主 要目的在於研究血漿磷脂脂肪酸和尿液揮發性代謝產物是否有助於肝細胞癌診 斷中潛在生物標記物的發現。 兩種分析方法包括侵入性和非侵入性方式，均得以在本研究中應用。侵入性 方式實驗中，一共來自于健康人群組和肝細胞癌病人組的 37 個血漿樣本已經收 集，並且其血漿磷脂脂肪酸分佈通過氣相色譜質譜聯用及多元分析統計用得以鑒 定和解析。非侵入性方式實驗中，一共來自于健康人群組和肝細胞癌病人，肝細 胞癌病人組以及索拉菲尼給藥組的 77 個尿液樣本收集完成。尿液揮發性代謝產 物通過頂空氣相色譜質譜聯用，以及中心複合設計及多元分析統計得到解析。在 侵入性方式實驗，亞油酸和油酸被發現作為血漿磷脂脂肪酸及肝組織脂肪酸異常 變化的共有生物標記物。在非侵入性方式實驗，在中心複合設計優化實驗條件後， 九種尿液揮發性代謝產物得到鑒定與解析，並且健康人群組和肝細胞癌病人組， 肝細胞癌病人組和索拉菲尼給藥組之間優選的生物標記物得到對比比較和解 析。 綜上所述，這些資料指示氣相色譜質譜聯用技術為基礎的侵入性血漿磷脂脂 肪酸實驗以及非侵入性尿液揮發代謝產物實驗，能夠充分發揮其在篩選作為診斷 肝細胞癌的生物標記物的作用。我們的發現表明氣相色譜質譜聯用技術結合代謝 組學思路可被開發作為肝細胞癌有效診斷工具。 關鍵字：肝細胞癌；氣相質譜；血漿磷脂脂肪酸；尿液揮發性代謝產物；頂空進 樣；中心複合設計；多元統計分析