Multicohort Analysis Reveals Colorectal Cancer Tumor Locationassociated Fecal Microbiota And Their Clinical Impact Yufeng Lin Harry Cheukhay Lau Chuanfa Liu Xiao Ding Yang Sun Jiamei Rong Xiang Zhang Luyao Wang Kai Yuan Yinglei Miao William Kakei Wu Sunny Hei Wong Joseph Jaoyiu Sung Jun Yu by Yufeng Lin & Harry Cheuk-hay Lau & Chuanfa Liu & Xiao Ding & Yang Sun & Jiamei Rong & Xiang Zhang & Luyao Wang & Kai Yuan & Yinglei Miao & William Ka-kei Wu & Sunny Hei Wong & Joseph Jao-yiu Sung & Jun Yu instant download after payment.

SUMMARYMicrobial alterations in different tumor locations of colorectal cancer (CRC) remain unclear. Here, 1,375fecal metagenomes from six in-house and published datasets were analyzed, including 128 right-sidedCRC (rCRC), 168 left-sided CRC (lCRC), 250 rectal cancer (RC), and 829 controls. Firmicutes progressively increase from rCRC, lCRC, to RC, in contrast to the gradual decrease of Bacteroidetes. Tumorlocation-associated fecal microbes are identified, including Veillonella parvula for rCRC, Streptococcusangionosus for lCRC, and Peptostreptococcus anaerobius for RC, while Fusobacterium nucleatum isenriched in all tumor locations. Tumor location-associated bacteria correlate with patient survival. Clinically, we establish a microbial biomarker panel for each tumor location that accurately diagnosesrCRC (area under the receiver operating characteristic curve [AUC] = 91.59%), lCRC (AUC = 91.69%),or RC (AUC = 90.53%) from controls. Tumor location-specific biomarkers also have higher diagnostic accuracy (AUC = 91.38%) than location-non-specific biomarkers (AUC = 82.92%). Overall, we characterizefecal microbes associated with different CRC tumor locations, highlighting that tumor location should beconsidered in non-invasive diagnosis.