Table 2.

Included studies assessing microbiome-derived biomarkers for early CRC detection

Reference	Comparison of interest1	Biomarker	Performance for early CRC detection		Analytical method2
AUC (95% CI)	Sens/spec
Zhang Y, 2021 [32]	AP vs HC	13 OTU3	0.61 (0.55–0.66)	34 / 80	16S
	AP vs HC	13 OTU + FIT	0.64 (0.58–0.69)	41 / 80
	AP vs HC	13 OTU + FIT + APCS4	0.71 (0.65-0.75)	46 / 80
Zhang S, 2020 [31]	AP vs HC	5 OTU	0.95 (0.91-0.99)		16S
Yachida, 2019 [19]	Stage 0 vs HC	29 species	0.73		Metagenomics
		16 KO genes	0.75
		29 species + 24 metabolites + 16 KO genes	0.78
Yao, 2021 [30]	Stage I/II vs HC	5 species	N/A	I 62 / -, II 72 / -	qPCR
	Stage I/II vs HC	5 species + FIT	N/A	I 70 / -, II 89 / -
Xie, 2017 [29]	Stage I/II vs HC	Fusobacteria nucleatum	0.59 (0.54-0.64)		qPCR
		Clostridium symbiosum	0.73 (0.68-0.77)
		Clostridium symbiosum + FIT	0.81 (0.75-0.86)
		Clostridium symbiosum and Fusobacteria nucleatum + FIT + CEA	0.82 (0.77-0.87)
Wei, 2022 [28]	AP vs HC	5 OTUs	0.71 (0.63-0.80)		16S
Mo, 2021 [27]	AP vs HC	Fusobacterium nucleatum and Parvimonas micra	0.57	39 / 83	qPCR
		Fusobacterium nucleatum and Parvimonas micra + DNA methylation/mutation5 + FIT	0.73	28 / 94
	Stage I/II	Fusobacterium nucleatum and Parvimonas micra + DNA methylation/mutation + FIT	N/A	I 60 / -, II 85 / -
Liu, 2021 [41]	AP vs HC	Fusobacterium nucleatum	0.74 (0.56-0.91)	71 / 65	qPCR
		pks+ Escherichia coli	0.82 (0.64-0.98)	91 / 75
		Fusobacterium nucleatum, pks+ Escherichia coli + CEA + CA19-9 + FOBT	0.85 (0.57-1.0)	67 / 90
		4 species + FIT	N/A	49 / 81
Liang, 2017 [40]	Stage I/II vs HC	4 species + FIT	N/A	I 77 / -, II 100 / -	qPCR
Liang, 2021 [39]	AP vs HC	4 species	0.67 (0.62-0.72)	39 / 83	qPCR
		4 species + FIT	N/A	49 / 81
Konishi, 2022 [38]	AP vs HC	50 OTUs	0.52 (0.45 – 0.60)		16S
	Stage 0 vs HC		0.54 (0.45 – 0.62)
	Stage I vs HC		0.74 (0.68 – 0.80)
	Stage II vs HC		0.86 (0.81 – 0.90)
Guo, 2018 [37]	Stage I vs HC	3 taxa	0.80	90 / 60	qPCR
	Stage I/II vs HC		0.89	90 / 76
Goedert, 2015 [47]	AP vs HC	5 taxa + 7 genera	0.77		16S
Gao, 2022 [36]	AP vs HC	12 species	0.66 (0.57-0.75)		Metagenomics
	Stage I vs HC		0.88 (0.72-1.00)
	Stage II vs HC		0.96 (0.91-1.00)
Gao, 2020 [35]	AP vs HC	18 genera	0.62 (0.52-0.71)	84 / 39	Metagenomics
		18 genera + FIT	0.72 (0.63-0.81)
Coker, 2022 [34]	AP vs HC	14 species	0.84 (0.80-0.89)		Metagenomics
		14 species + 2 metabolites	0.88 (0.84-0.92)
Ai, 2017 [33]	AP vs HC	6 species	0.87		16S
Rezasoltani, 2018 [43]	AP + Stage I vs HC	5 taxa	0.97	100 / 77	qPCR
Zeller, 2014 [50]	CRC vs HC^	22 species	0.84		Metagenomics
	CRC vs HC^	22 species + FOBT	0.87
Young, 2021 [49]	AP vs HC*	15 taxa	0.72 (0.68-0.75)		16S
	AP vs HC⁎⁎	15 taxa	0.82 (0.79-0.84)
	AP vs HC*	15 taxa + age + sex	0.71 (0.67-0.74)
	AP vs HC⁎⁎	15 taxa + age + sex	0.84 (0.82-0.86)
Tarallo, 2019 [48]	AP vs HC	hsa-miRNA + bsRNA + mbDNA6	0.47		sRNA7/Metagenomics
Flemer, 2017 [46]	AP vs HC	16 OTUs (faecal)	0.87		16S
		12 OTUs (oral)	0.89 (0.80-0.89)	67 / -
		16 OTUs (faecal) + 12 OTUs (oral)	0.98 (0.95-0.98)	88 / 94
Eklöf, 2017 [45]	AP vs HC	clbA+ bacteria, Fusobacterium nucleatum	N/A	47 / 63	qPCR
		clbA+ bacteria, Fusobacterium nucleatum + FIT	N/A	61 / 63
Clos-Garcia, 2020 [44]	AP vs HC	16 genera	0.28		16S
Zackular, 2014 [53]	AP vs HC	5 OTUs	0.84 (0.74–0.94)		16S
		5 OTUs + age + race + BMI	0.90 (0.82–0.98)	90 / 80
Baxter, 2016 [52]	AP vs HC	22 OTUs	0.67	18 / 97	16S
		23 OTUs + FIT	0.76

¹CRC = colorectal cancer, AP = adenoma and/or polyp, HC = healthy controls

²Analytical method: 16S = 16S rRNA gene sequence, qPCR = real-time quantitative PCR, metagenomics

³OTU = operational taxonomic unit

⁴APCS = Asia-Pacific Colorectal Screening, a validated risk-stratification tool

⁵Methylated genes: Septin9, NDRG4, BMP3. Mutated genes: KRAS, BRAF, PI3KCA

⁶hsa = Homo sapiens microRNAs, bsRNA = bacterial small RNAs, mbDNA = microbial DNA

⁷sRNA = small RNA sequencing

^{^}Based on CRC vs HC, but similar changes in microbial abundance observed in early-stages of CRC

^⁎Healthy controls with normal colonoscopy

^⁎⁎Healthy controls, blood negative in stool