高參數免疫分析血液檢測為人類胰臟癌提供早期的診斷策略

Abstract

胰腺導管腺癌（PDAC）通常在晚期被診斷，而且五年存活率低。我們與國立臺灣大學醫院合作，招募了49位參與者：23位健康捐贈者、24位PDAC患者和2位其他疾病患者。我們建立了血液冷凍保存和細胞恢復的臨床工作流程，並使用光譜流式細胞儀開發了一個用於單一樣本的40種螢光標記的高參數免疫分析模型。我們的數據顯示，與PDAC患者相比，健康個體的周邊血中有較多的naive CD4+和CD8+ T 細胞，而PDAC組的效應CD4、CD8和PD-1表達則較高。我們進一步運用了先進的機器學習模型，包括隨機森林、梯度提升和貝葉斯加法回歸樹（BART）來分類不同組別之間的差異，我們確定了naive CD8免疫細胞群在PDAC和健康組之間最具差異，而CD95表達則成一個關鍵的生物標記。我們還觀察到某些腸道微生物群可以調節免疫表達，這表示介入微生物群的給予並結合免疫療法在胰腺導管腺癌（PDAC）患者的治療中有潛在的效果。在血液中CD95表達和naïve CD8+ T細胞的表達與腸道中的Erysipelotrichales和Akkermansia和Lancefieldella的存在之間發現了正相關。總結來說，腸道微生物群的分析和免疫反應的綜合應用強調了腸道微生物群與系統性免疫指標之間的關聯，我們的研究結果顯示了在癌症的檢測和診斷中廣泛的應用。

Pancreatic ductal adenocarcinoma (PDAC) is frequently identified at advanced stages, contributing to its dismal five-year survival statistics. This investigation, in collaboration with National Taiwan University Hospital, incorporated 49 subjects: 23 healthy controls, 24 PDAC patients, and 2 individuals with distinct pathologies. We optimized protocols for blood cryopreservation and cell recovery, and employed spectral flow cytometry for high-parameter immune profiling, simultaneously analyzing 40 fluorescence markers per sample. Our data indicate elevated naive CD4+ and CD8+ T-cells presents in the peripheral blood of healthy participants relative to PDAC patients. Conversely, PDAC patients exhibited increased levels of effector CD4+ and CD8+ T cells and PD-1 expression. We further employed advanced machine learning models, including Random Forest, Gradient Boosting, and Bayesian Additive Regression Trees (BART), to classify the differences across cohorts and identified the distinct features in naive CD8+ T-cell populations between PDAC and healthy cohorts, with CD95 expression emerging as a pivotal biomarker. We also noted that certain gut microbiota compositions modulate immune expression, suggesting the potential efficacy of combined immunotherapeutic and microbiome-targeted interventions in PDAC treatment. A positive association was identified between CD95 expression and naïve CD8+ T cell occurrence in the blood and the presence of Erysipelotrichales, Akkermansia, and Lancefieldella in the gut. Overall, the integrated application of microbial community analyses and immunological assessments emphasizes the connection between gut microbiota and systemic immunological indicators, suggesting the utility of a comprehensive omics strategy in the detection and diagnosis of cancer.