摘要目的:总结高原地区急性胰腺炎（AP）患者临床特征，并通过代谢组学技术筛查高原AP发病的潜在生物标志物。方法:前瞻性纳入2023年12月至2024年5月间拉萨市人民医院消化内科收治的42例AP患者（AP组），同期纳入健康对照者33例（健康对照组），收集患者的人口学资料及临床资料。基于超高效液相色谱质谱联用系统行血清非靶向代谢组学检测，主成分分析评估AP组与健康对照组代谢谱，正交-偏最小二乘判别分析构建回归模型评估代谢产物解释率（ R2）和预测能力（ Q2）。基于OPLS-DA降维法计算变量投影重要度（VIP），基于差异倍数值（FC）计算代谢物在AP组和健康对照组间的差异倍数。对样本和差异代谢物进行双向聚类绘制聚类热图，评估组内样本相似度。绘制差异代谢物火山图，筛选区分AP与健康对照组上调和下调前5位差异代谢物。绘制血清差异代谢物的受试者工作特征曲线（ROC），计算曲线下面积（AUC）、灵敏度和特异度，评估差异代谢物的鉴别诊断价值。 结果:AP组（37例，88.1%）与健康对照组（27例，81.8%）均以藏族为主。AP患者年龄（50.45±17.85）岁，男女比例1.1∶1.0，病因主要为胆源性（33例，78.6%），多为中度重症（26例，61.9%）。局部并发症发生率为83.3%，主要为胰性胸腹腔积液及胰周积液（30例，71.4%）。全身并发症发生率为59.5%，主要为全身炎症反应综合征（22例，52.4%）和呼吸衰竭（15例，35.7%）。血清代谢组学主成分分析显示两组血清代谢产物差异明显。正交-偏最小二乘判别分析显示代谢产物可有效区分AP患者与健康对照者，正离子模式下 R2=0.992， Q2=0.913；负离子模式下 R2=0.983， Q2=0.914。AP组共有450种上调的差异代谢物和366种下调的差异代谢物，其中上调最显著的前5位代谢物分别为γ-谷氨酰-L-亮氨酸、可的松、4(15)-copaen-11-ol、黏菌黄素、吲哚-3-乙醛醇，下调最显著的前5位代谢物分别为N-乙酰-L-色氨酸、犬尿酸、脱氧尿嘧啶核苷酸、假尿苷、乙酸法尼酯。ROC分析显示这些标志物用于区分AP患者与健康对照者的AUC值均>0.8且 P值均<0.001，灵敏度及特异度均>80%，其中N-乙酰-L-色氨酸与乙酸法尼酯的鉴别诊断性能最佳。 结论:高原地区AP患者以胆源性病因多见，多为中度重症，局部并发症及全身并发症发生率较高。部分氨基酸类和异戊烯醇酯类代谢物可显著区分AP患者与健康对照者，可能参与高原AP的发病。

abstractsObjective:To summarize the clinical characteristics of patients with acute pancreatitis (AP) in plateau areas, and screen potential biomarkers for the pathogenesis of AP at high altitudes by metabolomics.Methods:A total of 42 patients with AP admitted to the Department of Gastroenterology in Lhasa People's Hospital from December 2023 to May 2024 were prospectively enrolled (AP group), and 33 healthy controls (Control group) were included during the same period. Demographic and clinical data were collected, and serum non-targeted metabolomics was performed based on ultra-performance liquid chromatography-mass spectrometry techniques. The serum metabolites map was evaluated by using principal component analysis, and a regression model with orthogonal partial least squares-discriminant analysis (OPLS-DA) was constructed to evaluate the explanatory power ( R2) and predictive power ( Q2) of these metabolites. The OPLS-DA-based dimensionality reduction was applied to compute variable importance in projection (VIP), while fold change (FC) values were used to assess the difference of metabolites between groups. A bidirectional clustering heat map was generated for samples and differential metabolites to evaluate sample similarity within groups. Additionally, a volcano plot was created to visualize differential metabolites, and the top five up-regulated and down-regulated metabolites distinguishing AP from healthy controls were selected. The receiver operating characteristic curve (ROC) was drawn, and the area under the curve (AUC), sensitivity and specificity based on ROC analysis were calculated to evaluate the differential power of differential metabolites. Results:The majority of participants were Tibetans in both the AP group (37 cases, 88.1%) and the control group (27 cases, 81.8%). The average age of AP patients was (50.45±17.85) years old, and the male to female ratio was 1.1∶1.0. The leading etiology was biliary disease (33 cases, 78.6%), and most patients encountered moderately severe disease (26 cases, 61.9%). The incidence of local complications was 83.3%, mainly thoracoabdominal effusion and peripancreatic effusion (30 cases, 71.4%). The incidence of systemic complications was 59.5%, mainly systemic inflammatory response syndrome (22 cases, 52.4%) and respiratory failure (15 cases, 35.7%). Principal component analysis showed significant differences in serum metabolites between groups. OPLS-DA showed that these metabolites effectively distinguished AP patients from healthy controls: R2=0.992 and Q2=0.913 in the positive ion mode, R2=0.983 and Q2=0.914 in the negative ion mode. There were 450 up-regulated and 366 down-regulated differential metabolites in AP group, respectively. Among them, gamma-glutamylleucine, cortisone, 4(15)-Copaen-11-ol, mytiloxanthin, and indole-3-glycol aldehyde were the top five up-regulated metabolites, while N-Acetyltryptophan, kynurenic acid, deoxyuridine monophosphate, pseudouridine, and farnesyl acetate were the top five down-regulated metabolites. ROC analysis of these markers showed that all AUC values were >0.8, with all P values <0.001, with both sensitivity and specificity exceeding 80%. Among them, N-Acetyltryptophan and farnesyl acetate possessed the best differential performance. Conclusions:Biliary causes are most frequent among AP patients in plateau area. The disease severity is mainly moderately severe, accompanied by high incidences of local and systemic complications. Some amino acids and prenol lipids could significantly distinguish AP patients from healthy controls, and might be involved in the pathogenesis of AP at high altitudes.