临床肿瘤学杂志

目的 探讨局部晚期口腔鳞状细胞癌（OSCC）患者的唾液微生物群落结构特征，分析唾液微生物群落组成与新辅助免疫化疗应答间的关系。方法 纳入17例接受替雷利珠单抗联合紫杉醇及顺铂（TP方案）新辅助治疗的局部晚期OSCC患者为OSCC组，10例健康体检者为正常对照（NC）组。OSCC组患者均接受替雷利珠单抗联合TP方案新辅助治疗3个周期，依据影像学及病理学疗效分为应答组与非应答组。采集受试者非刺激性唾液样本，采用Illumina MiSeq平台对细菌16S rRNA基因V3～V4区进行高通量测序，通过QIIME 2分析扩增子序列变体（ASV）。α多样性采用Ace指数、Chao1指数、Shannon指数及Simpson指数进行评估；β多样性基于Bray-Curtis距离行主坐标分析（PCoA）和非度量多维尺度分析（NMDS），并以相似性分析（ANOSIM）检验组间差异；菌群差异物种采用线性判别分析（LDA）效应量分析筛选。以P<0.05且LDA>4为差异标志物筛选标准，并构建Logistic回归模型及绘制受试者工作特征（ROC）曲线评估预测价值。结果 17例OSCC患者中应答组10例、非应答组7例，两组基线资料差异均无统计学意义（P>0.05）。与NC组相比，OSCC组唾液微生物Ace、Chao1、Shannon指数显著升高，Simpson指数显著降低（均P<0.05）；基于Bray-Curtis距离的PCoA、NMDS分析显示两组菌群结构显著分离（ANOSIM:R=0.165,P<0.05）。OSCC组与NC组共有的核心ASVs为422个，OSCC组特有ASVs 4 467个，NC组1 508个。门水平两组均以厚壁菌门、拟杆菌门等为优势菌门；属水平上，OSCC组微小单胞菌属、卟啉单胞菌属、梭杆菌属显著富集。不同治疗应答患者中，以中位数（四分位距）表示，非应答组梭杆菌属相对丰度为0.088(0.069～0.129)，显著高于应答组0.046(0.039～0.062)；卟啉单胞菌属相对丰度0.099(0.068～0.126)亦显著高于应答组0.052(0.040～0.064)（均P=0.043）。Logistic回归显示，卟啉单胞菌属（OR=1.899,95%CI:1.097～3.292,P=0.022）、梭杆菌属（OR=3.081,95%CI:1.025～9.253,P=0.045）为治疗应答独立影响因素，联合预测模型的ROC曲线下面积为0.886(95%CI:0.673～1.000)。结论 OSCC患者唾液微生物群落与健康人群存在显著差异，梭杆菌属和卟啉单胞菌属丰度升高与新辅助免疫化疗应答不佳相关，二者联合模型预测效能良好，有望成为无创预测标志物。

Objective To investigate the structural characteristics of salivary microbial community in patients with locally advanced oral squamous cell carcinoma(OSCC), and to analyze the relationship between salivary microbial composition and response to neoadjuvant immunochemotherapy. Methods A total of 17 patients with locally advanced OSCC who received neoadjuvant therapy with tislelizumab combined with paclitaxel and cisplatin(TP regimen) were enrolled as the OSCC group, and 10 healthy individuals undergoing physical examination served as the normal control(NC) group. Patients in the OSCC group received 3 cycles of neoadjuvant therapy with tislelizumab plus TP regimen, and were divided into responders and nonresponders according to imaging and pathological efficacy. Unstimulated saliva samples were collected from all subjects. Highthroughput sequencing of the V3–V4 region of the bacterial 16S rRNA gene was performed on the Illumina MiSeq platform, and amplicon sequence variants(ASVs) were analyzed using QIIME 2. Alpha diversity was evaluated using the Ace index, Chao1 index, Shannon index and Simpson index. Beta diversity was assessed by principal coordinate analysis(PCoA) and nonmetric multidimensional scaling(NMDS) based on BrayCurtis distance, and intergroup differences were tested by analysis of similarities(ANOSIM). Differentially abundant taxa were screened by linear discriminant analysis(LDA) effect size analysis. A threshold of P<0.05 and LDA score>4 was used for identifying differential biomarkers. A Logistic regression model was constructed and receiver operating characteristic(ROC) curves were plotted to assess the predictive value. Results Among the 17 OSCC patients, 10 were in the responder group and 7 in the non-responder group. There were no statistically significant differences in baseline data between the two groups(P>0.05). Compared with the NC group, the Ace, Chao1 and Shannon indices of salivary microbiota in the OSCC group were significantly increased, while the Simpson index was significantly decreased(all P<0.05). PCoA and NMDS analyses based on Bray-Curtis distance showed distinct separation of microbial community structure between the two groups(ANOSIM: R=0.165, P<0.05). There were 422 core ASVs shared by the OSCC group and NC group, with 4 467 ASVs unique to the OSCC group and 1 508 unique to the NC group. At the phylum level, Firmicutes, Bacteroidetes and others were the dominant phyla in both groups. At the genus level, Parvimonas, Porphyromonas and Fusobacterium were significantly enriched in the OSCC group. In patients with different treatment responses, expressed as median(interquartile range), the relative abundance of Fusobacterium in the non-responder group was 0.088(0.069–0.129), which was significantly higher than 0.046(0.039–0.062) in the responder group. The relative abundance of Porphyromonas was 0.099(0.068–0.126), also significantly higher than 0.052(0.040–0.064) in the responder group(both P=0.043). Logistic regression showed that Porphyromonas(OR=1.899, 95% CI: 1.097–3.292, P=0.022) and Fusobacterium(OR=3.081, 95% CI: 1.025–9.253, P=0.045) were independent influencing factors for treatment response. The area under the ROC curve of the combined predictive model was 0.886(95% CI: 0.673–1.000). Conclusion Salivary microbial community structure differs significantly between OSCC patients and healthy individuals. Elevated abundances of Fusobacterium and Porphyromonas are associated with poor response to neoadjuvant immunochemotherapy. The combined model of these two genera shows favorable predictive performance and is expected to serve as noninvasive predictive biomarkers.