中文摘要（第一部分）

微生物群在局部晚期或晚期肺癌诊断中的价值及相关性研究

目的：人类微生物群是由栖息在人体口腔、肺、肠道等部位的以细菌为主、还包括病毒和真菌等组成的一个复杂生态系统，微生物失调等改变与包括恶性肿瘤在内的多种疾病有关。但微生物群在肺癌发生发展中的作用尚不明确，目前国内及世界上仍缺乏针对局部晚期或晚期肺癌患者特异性微生物群构成的研究数据。本研究通过比较局部晚期或晚期肺癌患者与非肿瘤人群的肠道及口腔唾液菌群构成差异，旨在挖掘与肺癌发生发展密切相关的关键菌种，探寻局部晚期或晚期肺癌辅助诊断的微生物标志物。

方法：收集2018年9月至2019年9月在中国医学科学院肿瘤医院确诊的初治局部晚期或晚期肺癌患者及其无肿瘤病史家属作为非肿瘤人群的粪便及口腔唾液标本作为发现队列，应用16S核糖体核糖核酸（16S ribosomal ribonucleic acid, 16S rRNA）基因测序分析肺癌患者及非肿瘤人群的粪便及口腔唾液标本中微生物菌群构成。运用R统计编程语言分析局部晚期或晚期肺癌患者对比非肿瘤人群的肠道及口腔唾液菌群多样性、构成成分等差异，采用随机森林机器学习10折交叉验证方法构建辅助局部晚期或晚期肺癌诊断的肠道及唾液菌群模型，并在验证队列中进行验证。

结果：研究共纳入60例初诊局部晚期或晚期肺癌患者和35例非肿瘤人群对照组作为发现队列。与非肿瘤人群相比，肺癌患者唾液菌群α多样性明显降低，并且唾液及肠道微生物构成成分均发生了显著变化。采用10折交叉验证方法，在唾液菌群方面，发现具有10种基于操作分类单元（operational taxonomic unit, OTU）的微生物菌种标记物的预测模型在肺癌预测中具有较高的准确性，曲线下面积（area under the curve, AUC）达 85％，敏感性和特异性分别达82%和99%。在粪便菌群方面，另10种基于OTU的微生物菌种标记物的预测模型亦可协助肺癌的诊断（AUC = 78％，敏感性 = 65%，特异性 = 99%）。为了验证发现队列的研究结果，我们招募了另一个由41名肺癌患者和26名非肿瘤人群对照组构成的独立队列，作为验证队列。在验证队列中，唾液菌群模型预测肺癌准确率达77%，敏感性及特异性分别为53%和95%；粪便菌群模型预测准确率达67%，敏感性及特异性分别为47%和86%。

结论：局部晚期或晚期肺癌患者较非肿瘤人群具有独特的微生物群谱，肠道及唾液特异性菌群模型可潜在辅助肺癌的预测诊断。

关键词：微生物群；16S rRNA; 肺癌；诊断

中文摘要（第二部分）

肠道菌群在局部晚期或晚期肺癌一线化疗疗效及生存获益中的相关性研究

目的：越来越多的研究证据表明，肠道微生物在调节肺癌患者免疫治疗反应中具有重要作用。然而，肠道菌群与化疗结果之间的关系尚不明确。本研究为单中心前瞻性观察性研究，旨在探讨局部晚期或晚期肺癌患者肠道菌群与一线化疗疗效及生存获益的相关性，初步探索预测一线化疗临床结局的无创性肠道微生物标志物，并评估其临床效能。

方法：收集2018年9月至2019年9月在中国医学科学院肿瘤医院确诊并拟接受一线化疗的局部晚期或晚期肺癌患者接受化疗前的基线粪便样本。利用宏基因组测序方法分析患者的肠道菌群构成。随访统计患者一线化疗疗效及生存获益等临床特征，运用R统计编程语言分析不同化疗疗效及生存获益患者的肠道菌群多样性、构成成分、代谢通路等差异。

结果：共计64例接受一线化疗的局部晚期或晚期肺癌患者纳入本研究并提供一线化疗前的基线粪便样本。其中，33例患者化疗疗效为疾病缓解（response, R），均为部分缓解；另外31例未出现疾病缓解（non-response, NR），表现为疾病稳定或疾病进展。患者中位无进展生存期（progression free survival, PFS）为7个月（1.5-14.5月）。一线化疗后疾病缓解组肺癌患者的肠道菌群富集变形链球菌（Streptococcus mutans）（P = 0.026）和酪黄肠球菌（Enterococcus casseliflavus） （P = 0.049），未缓解组患者富集乳明串珠菌（Leuconostoc lactis）（P = 0.002）和惰性真杆菌（Eubacterium siraeum）（P = 0.006）等11种菌种。代谢通路功能分析显示疾病缓解组肺癌患者富集L-谷氨酸降解VIII代谢通路（P = 0.014），未缓解组富集C4光合碳同化循环，还原性三羧酸循环I，己糖醇发酵成乳酸、甲酸盐、乙醇和乙酸盐代谢通路（P < 0.05）。另一方面，12种菌种在不同程度无进展生存获益组间差异具有统计学意义，玫瑰蔷薇菌（Roseburia inulinivorans）（P = 0.004）、变形链球菌（Streptococcus mutans）（P = 0.048）等6种菌种在无进展生存获益较长组（PFS≥7月）富集（P < 0.05）；斯特科雷氏普氏菌（Prevotella stercorea）、黏性阿克曼菌（Akkermansia muciniphila）、约翰逊乳杆菌（Lactobacillus johnsonii）、细线虫菌（Alistipes indistinctus）、中间链球菌（Streptococcus intermedius）、交叉丁酸弧菌（Butyrivibrio crossotus）在无进展生存获益较短患者组（PFS＜7月）富集（P < 0.05）。代谢功能上，无进展生存获益较长组富集嘌呤核碱基降解、乙酰辅酶A发酵制丁酸酯等代谢通路（P < 0.05）。此外，通过Spearman相关分析显示某些特异性细菌种类与年龄、体重指数（body mass index, BMI）、病理类型、转移部位等临床表型之间存在显著相关性。

结论：特异性肠道菌群与肺癌一线化疗疗效及生存获益具有一定的相关性，是预测化疗结局的潜在标志物。

关键词：肠道菌群；宏基因组；肺癌；化疗；临床结局

Abstract (Part I)

The value and correlation of microbiota in the diagnosis for locally advanced or advanced lung cancer

Background: The human microbiota is a complex ecosystem composed of mainly bacteria, viruses and fungi that inhabit the human mouth, lung, and intestines. Changes such as microbial disorders are related to many diseases including malignant tumors. However, the role of the microbiota in the occurrence and development of lung cancer is still unclear. At present, there is still a lack of research data on the specific microbiota composition of locally advanced or advanced lung cancer patients in China and the world. This study aims to compare the differences in intestinal and oral flora in locally advanced or advanced lung cancer patients compared to non-tumor populations, for screening key bacteria in the occurrence and development of lung cancer, and discovering microbial biomarkers that assist the diagnosis of locally advanced or advanced lung cancer.

Methods: We analyzed stool and oral saliva specimens from the discovery cohort consisted of newly diagnosed locally advanced or advanced lung cancer patients in Cancer Hospital, Chinese Academy of Medical Sciences from September, 2018 to September, 2019 and their family members without tumor history, through 16S ribosomal ribonucleic acid (16S rRNA) of microbiota. The R statistical programming language was used to analyze the differences in the diversity and composition of the intestinal and oral flora of locally advanced or advanced lung cancer patients compared with non-tumor populations. The random forest machine learning 10-fold cross-validation method was used to construct the model of intestinal and salivary flora that assisted the diagnosis of lung cancer, and validated in the validation cohort.

Results: A total of 60 newly diagnosed locally advanced or advanced lung cancer patients and 35 non-tumor population controls were included in the discovery cohort. Compared with the non-tumor populations, the α diversity of salivary microflora was significantly decreased in lung cancer patients, and the compositions of salivary and intestinal microflora were significantly changed. Using the 10-fold cross-validation method, in terms of saliva flora, it was found that the prediction model with 10 microbial species biomarkers based on the operational taxonomic unit (OTU) has high accuracy in the prediction of lung cancer, with the area under the curve (AUC) as 85%, sensitivity and specificity reaching 82% and 99%, respectively. In terms of intestinal flora, another prediction model based on OTU-based 10 microbial species biomarkers can also assist in the diagnosis of lung cancer (AUC = 78%, sensitivity = 65%, specificity = 99%). To verify the findings of the discovery cohort, we recruited another independent cohort consisting of 41 lung cancer patients and 26 non-tumor control groups as the verification cohort. In the verification cohort, the prediction accuracy of the salivary flora prediction lung cancer model was 77%, and the sensitivity and specificity were 53% and 95%, respectively; the intestinal flora model had a prediction accuracy of 67%, with the sensitivity and specificity as 47% and 86%, respectively.

Conclusions: Patients with locally advanced or advanced lung cancer have the special microbiota profile compared to non-tumor populations, and the specific intestinal and salivary microbiota biomarker models can be used to assist in the potential predictive diagnosis of lung cancer.

Keywords: Microbiota; 16S rRNA; Lung cancer; Diagnosis

Abstract (Part II)

The correlation study of gut microbiome in the clinical efficacy and survival benefits of first-line chemotherapy for patients with locally advanced or advanced lung cancer

Background: Accumulating evidences have disclosed the important role of gut microbiome in modulating response of immunotherapy in the patients with lung cancer. However, research exploring the relationship of intestinal flora and chemotherapy is still limited. The study is to investigate the correlation between intestinal flora and chemotherapy efficacy and survival benefits in locally advanced or advanced lung cancer, preliminarily exploring non-innovative intestinal microbial biomarkers to predict the clinical outcomes of first-line chemotherapy.

Methods: We analyzed baseline stool samples from patients with locally advanced or advanced lung cancer before chemotherapy treatment in Cancer Hospital, Chinese Academy of Medical Sciences from September, 2018 to September, 2019, through metagenomics of gut microbiota. The diversity, composition, function and metabolic pathway of microbial communities were compared among patients with different chemotherapy outcomes.

Results: 64 consecutive patients with locally advanced or advanced lung cancer treated with chemotherapy were included into this study. All patients provided the stool samples. 33 of 64 patients responded to treatment (responders) and another 31 patients didn’t (non-responders). The median progression-free survival was 7 months (range, 1.5-14.5). Streptococcus mutans (P = 0.026) and Enterococcus casseliflavus (P = 0.049) were enriched in responders, while 11 bacteria including Leuconostoc lactis (P = 0.002) and Eubacterium siraeum (P = 0.006) were enriched in non-responders. Functional analysis of metabolic pathways revealed that L-glutamate degradation VIII pathway was enriched in responders (P = 0.014), and C4 photosynthetic carbon assimilation cycle, reductive TCA cycle I, and hexitol fermentation to lactate, formate, ethanol and acetate were enriched in non-responders (P < 0.05). On the other hand, 12 species were statistically significantly different among the progression-free survival benefit groups. Patients enriched with 6 bacterial species, such as Roseburia inulinivorans (P = 0.004), and Streptococcus mutans (P = 0.048) had longer progression-free survival than those enriched in the other 6 bacteria including Prevotella stercorea, Akkermansia muciniphila, Lactobacillus johnsonii, Alistipes indistinctus, Streptococcus intermedius and Butyrivibrio crossotus (P < 0.05). Purine nucleobases degradation I and other 4 metabolic pathways were enriched in lung cancer patients with longer progression-free survival (P < 0.05). In addition, significant associations of certain bacterial species with clinical parameters such as age, body mass index (BMI), pathological pattern and metastatic sites were observed by spearman correlation analysis.

Conclusions: Specific intestinal bacteria may be associated with the clinical outcomes such as efficacy and survival benefits of locally advanced or advanced lung cancer patients with chemotherapy, which may be the potential biomarkers to predict the chemotherapy outcomes.

Keywords: Gut microbiome; Metagenomics; Lung cancer; Chemotherapy; Clinical outcomes