以液相層析質譜分析法評估voriconazole引起肝臟損傷之潛力生物標記

Abstract

Voriconazole屬於第二代triazole類藥物，經常使用在具有血液疾病的病人以治療或預防黴菌感染，由voriconazole引起的肝臟損傷從轉胺酶濃度短暫升高到肝臟機能不良都可能發生。儘管傳統的肝臟功能檢測與因果推論的評分表是評估是否為藥物引起肝臟損傷的標準流程，但仍需具更高敏感度及特異性的生物標記。代謝體分析研究指出voriconazole引起肝臟損傷的機轉與氧化壓力有關。其中，tryptophan、cytidine、inosine、pyruvate、α-ketoglutarate、glycocholate、和β-N-acetylglucosamine這7個代謝物的濃度，在voriconazole引起肝臟損傷的病人組別中具有顯著的變化。將α-ketoglutarate、glycocholate、和β-N-acetylglucosamine作為一組檢測的套組時，比傳統的肝功能檢測有更好的偵測voriconazole引起肝損傷的效果，有潛力作為生物標記。本篇研究目的在於以更多的人數 (樣本數=129) 評估該生物標記之應用潛力。我們將2016年1月1日至2017年9月30日於臺大醫院接受voriconazole療法的病人，經納入及排除條件篩選後，根據Common Terminology Criteria for Adverse Events (CTCAE) 傳統的肝功能檢測指標之濃度標準，分為控制組與產生Grade 2以上之肝毒性組別。利用液相層析四極柱質譜儀分析病人血漿檢體的上述目標代謝物，將得到的代謝物層析峰面積 (peak area) 進行統計學上的差異倍數評估、單變項統計以及多變項羅吉斯迴歸。有肝毒性組別與控制組的病人，其血中α-ketoglutarate和glycocholate濃度都呈現顯著的差異。由α-ketoglutarate和glycocholate作為一組檢測套組後對於肝損傷的偵測能力也高於傳統的肝功能檢測。以1.250為分界值偵測肝毒性時，具有78.6%的敏感度與72.4%的特異性。然而，由於本研究難以使用傳統因果推論評量評估表篩選因voriconazole引起肝臟損傷的病人，因此，未來需利用更多經評估表篩選後，確認為voriconazole引起肝臟損傷的病人進行研究，以進一步證實α-ketoglutarate及glycocholate作為voriconazole引起肝臟損傷生物標記之適用性。

Voriconazole is a second-generation triazole. It is commonly used for prophylaxis and treatment of fungal infections in hematology patients. Voriconazole-induced liver injury has been reported from transient elevation in serum aminotransferase levels to severe hepatic dysfunction. Although conventional liver tests and scoring systems of causality were currently the standard assessment for drug-induced liver injury (DILI), ideal biomarkers with higher sensitivity and specificity were still in need. Voriconazole-induced liver injury was shown to be associated with oxidative stress by using metabolomics analysis. Levels of seven metabolites, including tryptophan, cytidine, inosine, pyruvate, α-ketoglutarate, glycocholate, and β-N-acetylglucosamine, were significantly changed in patients with voriconazole-induced liver injury. A metabolite panel consisting of α-ketoglutarate, glycocholate, and β-N-acetylglucosamine was previously proposed as potential biomarkers to identify voriconazole-induced liver injury with better performance than that of conventional liver tests. The present study aimed to evaluate the practicality of these biomarkers in a larger number of patients (n=129). After application of the inclusion and exclusion criteria, patients receiving voriconazole therapy at the National Taiwan University Hospital from 1 January 2016 to 30 September 2017 were categorized into the control group and the hepatotoxicity group (with at least Grade 2 adverse events) based on the levels of liver tests according to the Common Terminology Criteria for Adverse Events (CTCAE). The analysis of plasma samples for the aforementioned metabolites by ultra-high-performance liquid chromatography coupled to triple quadrupole mass spectrometry was performed. The obtained peak areas of metabolites were analyzed for fold change evaluation, univariate statistics, and multivariate logistic regression. Levels of α-ketoglutarate and glycocholate were significantly different between the control group and the hepatotoxicity group. The panel consisting of the two metabolites also had better performance for detecting liver injury in comparison with the individual biomarkers in the conventional liver tests. With the cutoff value of 1.250, the sensitivity and specificity to identify hepatotoxicity were 78.6% and 72.4%, respectively. However, there was difficulty in the application of conventional scales for the identification of patients with voriconazole-induced liver injury in the present study. Therefore, further investigation with more patients meeting the “probable” criteria of voriconazole-induced liver injury is needed for confirming the suitability of α-ketoglutarate and glycocholate as the biomarkers for voriconazole-induced liver injury.