Modelling the impact of colorectal cancer screening strategies: A systematic review and meta-analysis

Table 1 Summary of screening effectiveness based on novel screening strategies¹

Screening strategies2 (modality, age range, interval)	No. of studies	No. of scenarios	Risk ratios of CRC incidence	Risk ratios of CRC mortality
Direct visualization tests
AI-assisted colonoscopy, 50-80, 10[55]	1	2	0.18	0.13
			0.513	0.483
AI-assisted colonoscopy, 45-75, 10[19]	1	1	0.24	0.24
MRC, 50-80, 5/10[45]	1	4	NR	0.55-0.753
Stool-based tests
ngMT-sDNA (Exact Sciences), 45-75, 1[26,28,35]	3	4	0.27 (0.21-0.36)4	0.23 (0.14-0.37)4
			0.293	0.303
ngMT-sDNA (Exact Sciences), 45-75, 2[28]	1	1	0.28	0.24
ngMT-sDNA (Exact Sciences), 45-75, 3[26,28,35]	3	5	0.35 (0.28-0.44)4	0.26 (0.18-0.39)4
			0.533	0.343
mt-sRNA, 45-75, 1[35]	1	2	0.29	0.26
			0.323	0.283
mt-sRNA, 45-75, 3[35]	1	2	0.35	0.27
			0.493	0.323
FIT-RNA (Geneoscopy ColoSense), 45-75, 3[28]	1	2	0.32-0.34	0.28-0.29
M3CRC, 50-75, 1[58]5	1	1	0.503	NR
Blood-based tests
cf-bDNA (Guardant Shield), 45-75, 1[28]	1	1	0.35	0.26
cf-bDNA (Guardant Shield), 45-75, 2[28]	1	1	0.50	0.36
cf-bDNA (Guardant Shield), 45-75, 3[27,28,35,36,48,56,59]	76	12	0.56 (0.48-0.64)4	0.43 (0.34-0.54)4
			0.67 (0.57-0.78)3,4	0.56 (0.44-0.73)3,4
cf-bDNA (Freenome), 45-75, 3[28]	1	2	0.58-0.59	0.46-0.47
Liquid biopsy, 45-75, 3[24]7	1	1	0.98	0.94
Urine metabolomic-based tests
UMT, 50-75, 3[47]	1	2	0.51	0.41
			0.583	0.503
UMT, 50-75, 1[57]	1	1	NR	0.503

¹Lifetime effects compared with no screening under perfect adherence.

²Screening strategies are presented as screening modality, age to begin-age to end screening, screening interval.

³Lifetime effects compared with no screening under realistic adherence.

⁴Pooled estimates from meta-analysis are reported when data are available; otherwise, the range of values is reported.

⁵A novel fecal bacterial biomarker with the combination of Lachnoclostridium species marker m³, Fusobacterium nucleatum, Clostridium hathewayi, Bacteroides clarus, and a faecal immunochemical test.

⁶One study that only reported colorectal cancer incidence and mortality joint range derived from the three CISNET models is not represented in this table[33].

⁷Circulating tumor DNA-based markers developed by GRAIL for their MCED test Galleri.

AI: Artificial intelligence; cf-bDNA: Cell-free DNA blood test; MRC: MR colonography; mt-sRNA: Multitarget stool RNA; ngMT-sDNA: Next-generation MT-sDNA; NR: Not reported; UMT: Urine metabolomic-based test; CRC: Colorectal cancer; FIT: Faecal immunochemical testing.

Table 2 Summary of screening effectiveness based on hybrid strategies¹

Hybrid strategies	No. of studies	No. of scenarios	Screening modalities	Risk ratios of CRC incidence	Risk ratios of CRC mortality
Concurrent dual-modality screening2	8	27	FS, FIT	0.17-0.48 (100% adherence); 0.58-0.93 (realistic adherence)	0.11-0.29 (100% adherence); 0.34-0.86 (realistic adherence)
			FS, FOBT
			Colonoscopy, liquid biopsy
			FIT, mSEPT9
Age-stratified sequential screening3	8	26	FIT, colonoscopy	0.19-0.47 (100% adherence); 0.39-0.88 (realistic adherence)	0.11-0.37 (100% adherence); 0.38-0.82 (realistic adherence)
			FOBT, colonoscopy
			FS, FIT
			FS, colonoscopy
Gender-specific strategies	2	5	FIT, colonoscopy	0.36 (100% adherence); 0.50-0.85 (realistic adherence)	0.28 (100% adherence); 0.74 (realistic adherence)
			FS, colonoscopy
Risk-stratified strategies	1	1	FIT, colonoscopy	0.36 (100% adherence)	NR

¹Screening scenarios varied by model, natural history assumptions, adherence, screening age, and interval for each strategy.

²Concurrent dual-modality screening refers to combining two distinct screening tests simultaneously.

³Age-stratified sequential screening refers to initiating with a primary screening method during earlier ages and transitioning to an alternative modality at older ages.

CRC: Colorectal cancer; FIT: Fecal immunochemical test; FOBT: Fecal occult blood testing; FS: Flexible sigmoidoscopy; mSEPT9: Methylated Septin 9 DNA blood test; NR: Not reported.